the puzzle of regional brain activity in depression and anxiety: … · 2020-02-19 · wendy heller...
TRANSCRIPT
The Puzzle of Regional Brain Activity in
Depression and Anxiety The Importance of
Subtypes and Comorbidity
Wendy Heller and Jack B NitschkeUniversity of Illinois at Urbanaplusmn Champaign USA
The literature on brain activity in depression and anxiety is reviewed with anemphasis on highlighting discrepancies and inconsistencies In particularanterior and posterior asymmetries have been reported for both depressionand anxiety but the magnitude and direction of these asymmetries has beenvariable We propose that by identifying subtypes of depression and anxietysome of these inconsistencies can be explained In addition we reviewevidence suggesting that issues of comorbidity are important to consider inattempting to account for regional brain activity in depression and anxiety
INTRODUCTION
A fundamental assumption in cognitive neuroscience is that various regionsof the brain perform different roles in the organisation of human behaviourIt is also assumed that most tasks require multiple cognitive operationseach of which depends on a set of neural computations localised to someregion(s) of the brain Thus it follows that most activities involve a set ofbrain regions that are recruited by the particular information processing andbehavioural demands of the task In addition the neural computationsinvolved are viewed as components or subroutines that contribute to butdo not constitute the reg nal output (ie there is no ` grandmother cellrsquo rsquo or` Wisconsin card sort arearsquo rsquo Petersen Corbetta Miezin amp Shulman1994) Rather these elementary processing mechanisms can be calledupon by a variety of tasks and the degree to which they are activatedwill depend on the unique characteristics of the task Metaphorically
COGNITION AND EMOTION 1998 12 (3) 421plusmn 447
Requests for reprints should be sent to Dr Wendy Heller Department of PsychologyUniversity of Illinois at Urbanaplusmn Champaign 603 E Daniel St Champaign IL 61820 USA
Wendy Heller was supported by NIMH grant MH52079 and by a grant from the Universityof Illinois Research Board Jack B Nitschke was supported by NIMH training grantMH14257 to the University of Illinois The authors gratefully acknowledge the commentsof Marie T Banich on an earlier draft of this paper
Oacute 1998 Psychology Press Ltd
speaking behaviour like a symphony is the output of an orchestra ofneurones Each region of the brain like each section of the orchestra has aparticular role to play and a particular manner in which it affects the overallcomposition To understand the making of a symphony we must identifythe component parts
Emotion like many cognitive processes is not a unidimensional con-struct Rather the term subsumes a variety of distinct phenomena includingphysiological events (eg changes in heart rate or skin conductance) feelingstate and modes of information processing Thus emotion is unlikely to` liversquorsquo in some specireg c location in the brain instead various brain regionsare likely to be engaged in different aspects of it It is therefore crucial thatbrain activity in emotion be examined in the light of a thorough under-standing of the multiple processes and brain regions involved The degree towhich we can disentangle the various subcomponents of the emotionalphenomena we are investigating determines the degree to which we candiscern the precise neural mechanisms of importance
In this article we will address the issues of depression and anxiety fromthe foregoing perspective We will emphasise the argument that differentsubcomponents of the symptomatology of affective and anxiety disordersare associated with distinct regions of the brain The degree to which thesesubcomponents are conmacr ated or co-occur will obscure our ability toidentify the neural mechanisms associated with particular disordersFurthermore the more we can be specireg c about the quality or nature ofthe symptomatology the more we are likely to be able to specify thecritical brain mechanisms
In turn a better understanding of the neural mechanisms involved caninform our understanding of the symptoms of depression and anxiety Forexample if right posterior regions of the brain prove to be dysfunctional indepression we might expect that depressed people would have difreg cultyprocessing narrative information a task that has been shown to rely inpart on this area of the brain (Heller amp Nitschke 1997) Because narrativeinformation processing is likely to play an important role in many ther-apeutic interventions the signireg cance of a dereg cit in this area is clear
Our approach is exemplireg ed by a neuropsychological model developedby Heller (1986 1990 1993ab) in which psychological theories of emo-tion were used to decompose emotional states into two componentsvalence and arousal This model integrates the dimensional circumplextheory of emotion (based primarily on self-report for a review see Larsonamp Diener 1992) with neuropsychological data on cognitive emotional andautonomic functioning during different affective states Based on evidencefrom electroencephalographic (EEG) blood macr ow and lesion studies themodel posits that the valence dimension (pleasant unpleasant) is dependenton functions of the anterior regions Bias toward a particular valence pole
422 HELLER AND NITSCHKE
is associated with the relative activity level of the left and right hemi-spheres When the left frontal region is active relative to the right affectivevalence is pleasant whereas when the right frontal region is active relativeto the left affective valence is unpleasant The association of valence withasymmetric activity of the anterior regions has been extensively studied(for reviews see Davidson 1992ab Heller 1990) and will be furtherreviewed in the course of this article
In contrast the arousal dimension (or activation dimension Larsen ampDiener 1992) is posited to depend on right parietotemporal regions ofthe brain More activity in this region is associated with higher self-reported arousal whereas less activity is associated with lower self-reported arousal This aspect of the model was based on theoreticalwork by numerous authors (eg Heilman Schwartz amp Watson 1978Levy Heller Banich amp Burton 1983 Tucker 1981) suggesting a specialrole for the right hemisphere in emotion-related arousal functions Thearousal component of Hellerrsquo s model derives empirical support fromnumerous additional studies examining both brain-damaged patients andnormal subjects (for reviews see Gainotti Caltagirone amp Zoccolotti1993 Wittling 1995)
To briemacr y review this evidence self-reports of emotion-related arousaland the right hemisphere have been investigated directly by HellerNitschke and Lindsay (1997b) Greater emphasis on right hemisphereprocessing as remacr ected in a left hemispatial bias on a free-vision task offace processing was associated with higher levels of self-reported arousalon several indices derived from the Proreg le of Mood States (an emotionaladjective checklist) Self-reported arousal has also been shown to beassociated with a number of autonomic functions particularly skin con-ductance (Greenwald Cook amp Lang 1989 Lang Greenwald Bradly ampHamm 1993) In turn numerous studies have suggested a link between theright hemisphere and skin conductance right brain-damaged patients haveconsistently been found to display reduced skin conductance responses toemotional stimuli (Caltagirone Zoccolotti Originale Daniele amp Mammu-cari 1989 Heilman Schwartz amp Watson 1978 Morrow Vrtunski Kimamp Boller 1981 Myslobodsky amp Horesh 1978 Valenstein amp Heilman1984 Zoccolotti Caltagirone Benedetti amp Gainotti 1986 ZoccolottiScabini amp Violani 1982) Compatible with these results there is consider-able evidence for greater autonomic conditioning when conditioned emo-tional stimuli are presented to the right versus the left hemisphere (Dawsonamp Schell 1982 Johnsen amp Hugdahl 1991 1993) There is also evidencethat cardiovascular functions associated with emotional responses are moredependent on the right than on the left hemisphere (Caltagirone et al1989 Hugdahl Franzon Andersson amp Walldebo 1983 Wittling 1990Zoccolotti et al 1986)
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 423
At the time of its conception Hellerrsquo s model attempted to address anumber of conmacr icts in the literature Which hemisphere was involved inemotion and in what manner was a matter of much debate Based onevidence that the right hemisphere was specialised to process emotionalinformation some researchers were arguing that all emotional functionswere dependent on the right hemisphere Representative of this perspec-tive Levy et al (1983) had obtained behavioural evidence on tasksassociated with parieto-temporal functions that led us to conclude thatindividuals displaying relatively reduced right hemisphere activationwere predisposed to unpleasant emotions whereas individuals displayingenhanced right hemisphere activation were predisposed to pleasant emo-tions Based on these reg ndings we posited that reduced activation of theright hemisphere was associated with depression This hypothesis wassupported by evidence that depressed people display dereg cits in cognitivefunctions associated with the right hemisphere on neuropsychological testsand in lateralised tachistoscopic paradigms (for reviews see Heller 19901993ab Tucker 1981)
In direct contrast EEG studies indicated that compared to the lefthemisphere right hemisphere activation was increased in depressedmood states particularly for anterior regions (for a review see David-son 1984) When mood was euthymic however activation of the righthemisphere was decreased relative to the left hemisphere These resultsbuilt on a set of observations systematically described by Gainotti(1972) noting very different emotional behaviours in people with rightversus left brain damage Left hemisphere damage was associated withthe so-called ` catastrophic reactionrsquorsquo in which patients were described tobe emotionally volatile and especially prone to depression and crying Incontrast right hemisphere damage was observed to be frequently accom-panied by a ` euphoricrsquo rsquo or ` indifferencersquo rsquo reaction Similar emotionalresponses have been described in patients undergoing the WADA orsodium amytal test Injections of sodium amytal that deactivate the lefthemisphere are associated with the ` catastrophicrsquo rsquo reaction whereasthose that deactivate the right hemisphere are associated with the` indifferenceeuphoriarsquo rsquo reaction (eg Lee Loring Meador Flanigin ampBrooks 1988)
The EEG and lesion studies inspired what have been referred to as` valencersquo rsquo theories (for a review see Borod 1993) In one variant ofthis type of theory the left hemisphere was thought to be specialised forpleasant affect and the right hemisphere was thought to be specialisedfor unpleasant affect (eg Sackeim et al 1982) To account for the lesiondata it was assumed that brain damage immobilises or diminishes thefunctioning of the damaged hemisphere and that the behaviour observedremacr ects the functioning of the intact hemisphere more or less on its own
424 HELLER AND NITSCHKE
In another variant of a valence theory Tucker (1981) concluded thatunpleasant and pleasant emotions were associated with the left and righthemispheres respectively a reversal of the valence theory described ear-lier Here the lesion data were explained by arguing that the observedbehaviour was not a function of the intact hemisphere but rather remacr ectedemotional processes of the subcortical areas on the damaged side that werereleased from cortical inhibition because of the lesion The EEG datawould be explained by assuming that increased anterior activity on oneside or the other would inhibit the affect associated with the correspondingipsilateral subcortical region (unpleasant for the left pleasant for the right)By this reasoning it follows that the affective effects of a lateralisedsubcortical lesion should be opposite to those of a cortical lesion (eg alesion to left subcortical structures should be associated with a suppressionof activity in this area leading to a decrease in unpleasant affect) How-ever several studies have found left anterior subcortical lesions to beassociated with depression (for a review see Liotti amp Tucker 1995)
In attempting to account for the co-occurrence in depression of increasedactivation in right anterior regions and dereg cits on cognitive tasks specia-lised to the right hemisphere Tucker (1981) argued that frontal regions havebeen traditionally viewed as having inhibitory regulatory effects on poster-ior regions By distinguishing valence from arousal as components ofemotion and localising these dimensions to anterior and posterior regionsof the brain respectively Hellerrsquo s (1986 1990) model also emphasised acaudal distinction However the model has additional explanatory powerAlthough inhibitory mechanisms of the anterior regions may well be play-ing a role in affective regulation (also see Tomarken amp Keener this Issue)the concept of a (relatively) independent arousal dimension associated withthe posterior right hemisphere allows us to model the patterns of brainactivity associated with all the emotions as represented on the circum-plex Indeed the patterns of activity that have been reported for depressionand anxiety cannot be fully accounted for by invoking an inhibitorymechanism For example if the inhibitory function of the right anteriorregion is the driving force behind right posterior activity in differentaffective states we would expect anxiety (which like depression is oftenaccompanied by increased right anterior activity) to be similarly accompa-nied by decreased right posterior activity However the evidence suggeststhat anxiety is often associated with increased activity in this region (asreviewed later) Thus by positing an independent arousal dimensionHellerrsquo s model provides an explanation for the patterns of brain activitythat have been observed for anxiety as well as depression
Tucker and colleagues have furthermore emphasised the important roleof cortical-subcortical interactions in producing the asymmetries in corticalactivity Hellerrsquo s model does not address the mechanisms by which the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 425
asymmetries associated with affective valence are generated indeed itcould well be the case that the relationship between activity in the anteriorregions and valence is mediated largely via interaction with subcorticalstructures Furthermore inhibitory mechanisms may be important in thisregard
Note that the approach we are emphasising in this article is compatiblewith Davidson and colleaguesrsquo attempts to decompose emotional statesaccording to the degree that they involve approach versus withdrawalbehaviours (Davidson 1992 Davidson amp Tomarken 1989 Tomarken ampKeener this Issue see also Kinsbourne 1988) These authors have arguedthat approach and withdrawal are the fundamental behavioural substratesassociated with affective valence However this theoretical formulationclearly emphasises the anterior cortical asymmetries associated with plea-santunpleasant valence Although Davidson has addressed posterior asym-metries in his work (eg Davidson Schaffer amp Saron 1985 Davidson1992a) the approach-withdrawal model does not focus on the patterns ofposterior brain activity that have been documented for depression andanxiety Thus considering the arousal dimension in addition to theapproach-withdrawal distinction may address the data for the posteriorregions more comprehensively
The present article has several goals We will highlight persistent andemerging discrepancies in the reg eld of brain function in depression andanxiety and suggest that turning to the psychological literature on thespecireg c symptomatology and comorbidity of these disorders can help usresolve at least some of them In the process we will revisit Hellerrsquo s modelof emotion with a critical look at how well it has been able to account forthe wealth of recent reg ndings in the reg eld Some modireg cations to this modelwill be proposed and continuing questions will be explicated
REGIONAL BRAIN ACTIVITY
Anterior and posterior asymmetries in cortical activity are often but notalways reported in depression and anxiety As a result it has been difreg cultto identify precisely the patterns of brain activity that are important In thesections following we highlight the discrepancies in the reg ndings foranterior and posterior asymmetries We then discuss possible causes forthese discrepancies and suggest ways to disentangle the results
Depression
Anterior asymmetries for EEG alpha have often been found in studies ofdepression (eg Allen Iacono Depue amp Arbisi 1993 Henriques ampDavidson 1990 1991 Schaffer Davidson amp Saron 1983) Across these
426 HELLER AND NITSCHKE
studies depression has been associated with more right than left anterioractivity Numerous positron emission tomography (PET) studies have alsoreported less left anterior activity (eg Baxter et al 1985 1989 Bench etal 1992 Bench Friston Brown Frackowiak amp Dolan 1993 George etal 1994 Martinot et al 1990) which is consistent with the EEG reg ndingsearlier However other studies have reported reg ndings in the oppositedirection for depression (Drevets et al 1992) and sad mood (George etal 1995) Null effects for depression have also been found (eg NitschkeHeller Etienne amp Miller 1995 Tomarken amp Davidson 1994)
The reg ndings for the parieto-temporal regons have been even morevariable In EEG and blood macr ow studies that report positive reg ndingsdepression is typically associated with less right posterior activity (egFlor-Henry 1979 Henriques amp Davidson 1990 Post et al 1987 Uytden-hoef et al 1983) Numerous studies however report no effects for thisregion of the brain (for a review see Davidson amp Tomarken 1989)
Reports of reduced right parieto-temporal activity although not alwayspresent are consistent with neuropsychological studies often indicatingimpaired or suppressed right hemisphere function in depressed peopleOn neuropsychological tasks thought to depend on right parietotemporalfunctioning depressed people have often been found to display decrementsin performance (eg Berndt amp Berndt 1980 Flor-Henry 1976 GoldsteinFilskov Weaver amp Ives 1977 Gruzelier Seymour Wilson Jolley ampHirsch 1988 Kronfol Hamsher Digre amp Waziri 1978) Similar findingsfor sad mood in nondepressed people have also been reported for righthemisphere tasks (Tucker Stenslie Roth amp Shearer 1981)
Studies that have directed information to one or the other hemisphereusing lateralised paradigms such as dichotic listening or visual half-reg eldpresentation have also found specireg c dereg cits for the right hemisphere indepressed people (for a review see Bruder 1995) Left-ear decrements orabsent left-ear (right hemisphere) advantages have been reported fordepression on nonverbal dichotic listening tasks (eg Bruder et al1989) Similarly left visual reg eld (right hemisphere) dereg cits in reactiontime and accuracy for nonverbal stimuli have been reported for depressedpeople (Bruder et al 1989 Liotti Sava Rizzolati amp Caffarra 1991)Comparable results have also been obtained in studies of nondepressedpeople who undergo mood inductions (Banich Stolar Heller amp Goldman1992 Ladavas Nicoletti Umilta amp Rizzolatti 1984) Reaction time andaccuracy for the right hemisphere are selectively impaired in sad moods
On the Chimeric Faces Task (CFT) a free-vision task of face processingthat typically elicits a left hemispatial bias suggesting greater right hemi-sphere activation Jaeger Borod and Peselow (1987) found smaller lefthemispatial biases for depressed patients than nonpsychiatric controlsUsing the same task Heller Etienne and Miller (1995) reported that
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 427
high-depressed participants had smaller left hemispatial biases than low-depressed participants Due to the nature of the CFT hemispatial biases arenonspecireg c as to whether they remacr ect a decrease in activity for the con-tralateral hemisphere or an increase in activity for the ipsilateral hemi-sphere (ie a reduction in the magnitude of the left hemispatial bias couldremacr ect either reduced right hemisphere or increased left hemisphere activ-ity) However based on the evidence (reviewed earlier) for selectivedereg cits in performance on right hemisphere tasks relatively poor perfor-mance for the left-visual reg eld on tachistoscopic tasks and reduced EEGactivity for right posterior regions the most likely interpretation of the CFTreg ndings is that they remacr ect diminished right posterior activity
In summary although anterior asymmetries have often been reported indepression the more common reg nding of higher left than right activation isnot invariably found Furthermore there is a signireg cant discrepancybetween the number of studies that have reported parieto-temporal asym-metries in activation which have been relatively few and the number ofstudies that have reported dereg cits in right hemisphere information proces-sing which have been many Also worthy of note is whether studies ofclinical depression psychometrically dereg ned depression and sad mood areaddressing the same phenomenon The number of consistencies acrossthese populations in the physiological literature reviewed here suggestssome commonality on the other hand the inconsistencies could be in partdue to the heterogeneity between the samples studied
Anxiety
Research on brain activity in anxious states has also been plagued byinconsistencies Again although anterior and posterior asymmetries havebeen reported no consistent pattern has emerged Studies of multisite EEGand regional cerebral blood macr ow (rCBF) using xenon-133 or PET havereported right hemisphere left hemisphere and bilateral increases inactivity during anxiety (for a review see Heller Nitschke Etienne ampMiller 1997a) Other studies have found no differences in brain activity(Gur et al 1988 Mountz et al 1989 Nordhal et al 1990 Tomarken ampDavidson 1994)
Having comprehensively reviewed this literature elsewhere (Heller etal 1997a) we will provide an overview of those studies reporting differ-ential hemispheric involvement in anxiety A subset of these suggest anasymmetry in favour of right hemisphere activation in various regions(frontal lateral prefrontal precentral frontal anterior temporal parietaloccipital and parahippocampal) Heightened right hemisphere activity hasbeen reported for panic disorder patients (Reiman Raichle Butler Hers-covitch amp Robins 1984) panic disorder patients experiencing a sodium
428 HELLER AND NITSCHKE
lactate-induced panic attack (Stewart Devous Rush Lane amp Bonte1988) generalised anxiety disorder (GAD) patients during an anxiety-stimulating task (Wu et al 1991) psychometrically dereg ned high-anxiousparticipants immediately following arterial and venous catheterisation andagain after completion of the PET scans (Reivich Gur amp Alavi 1983) andsocial phobics prior to making a public speech (Davidson MarshallTomarken amp Henriques submitted) Consistent with this pattern anti-anxiety medication was found to reduce the rate of glucose metabolismin the right hemisphere especially for the occipital and frontal regions forpatients with GAD (Buchsbaum et al 1987) Similarly in a population ofnonpsychiatric volunteers those administered diazepam intravenouslyshowed maximal decreases in rCBF in the right frontal area whereasthose injected with a placebo showed no rCBF changes (Mathew Wil-son amp Daniel 1985)
Regional brain activity can also be inferred from performance on tasksthat depend on particular brain regions On the CFT Heller et al (1995)reported that high-anxious participants had larger left hemispatial biasesthan low-anxious participants suggested to remacr ect greater right posterioractivity Similarly on a neuropsychological test battery panic disorderpatients performed less well than normal controls on left but not righthemisphere tasks (Yeudall et al 1983) Data obtained from a series ofstudies using behavioural measures to examine right versus left hemispherefunction during an experimental condition producing transient (ie state)anxiety (Tucker Antes Stenslie amp Barnhardt 1978 Tucker Roth Arne-son amp Buckingham 1977 Tyler amp Tucker 1982) are also consistent withan interpretation of greater right than left hemisphere activity in anxiety(Heller et al 1995)
However when subjects were classireg ed according to trait anxietyTucker and colleagues (Tucker et al 1978 Tyler amp Tucker 1982)interpreted the results as strongly suggestive of greater left than righthemisphere activity This interpretation is consistent with reports ofnumerous other labs reg nding increased left hemisphere activity associatedwith anxiety in various cortical and subcortical regions (orbital frontalinferior frontal anterior cingulate caudate putamen and thalamus)Heightened left hemisphere activity has been found in obsessive-compul-sive disorder (OCD) patients (Baxter et al 1987 Swedo et al 1989)GAD patients (Buchsbaum et al 1985 Wu et al 1991) and studentsclassireg ed as ` worriersrsquo rsquo on the basis of self-reports when asked to ` worryabout a specireg c topic of personal concernrsquo rsquo (Carter Johnson amp Borkovec1986)
In summary some reg ndings for anxiety indicate increased right hemi-sphere activation whereas others indicate increased left hemisphere acti-vation Indeed it may be the case that they could co-occur which could
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 429
account for the reg ndings of bilateral increases or no asymmetries In themeantime we are lacking a model to explain these discrepancies
ACCOUNTING FOR INCONSISTENCIES INBRAIN ACTIVITY
We are thus confronted with inconsistent results with regard to regionalbrain activity in both depression and anxiety In previous work we haveargued that in order to resolve these discrepancies it is helpful to turn tothe psychological literature on symptomatology in depression and anxietyIn the sections that follow we make some general observations that applyto both depression and anxiety before focusing on each more exclusively
In addressing the inconsistencies for the blood macr ow reg ndings severalconcerns about the literature are worthy of note First several of the mostfrequently cited studies in this literature sampled a restricted set of brainregions (eg Reiman et al 1984 Reivich et al 1983) Furthermore theregions of interest were not consistent from one study to the nextSecond statistical procedures employing multiple t-tests rather thanrepeated-measures ANOVAs are questionable especially with regard todrawing conclusions about hemispheric differences Third although somestudies examined selected diagnostic groups (eg major depressive dis-order OCD panic disorder GAD) most treated depression or anxiety asa unitary construct and did not consider that different types of each mightbe associated with different patterns of brain activity Moreover with theexception of recent OCD studies most did not take into account thedegree of comorbidity with depression or anxiety which may be criticalin determining the patterns of brain activity that occur (Heller et al1995 Tucker 1988)
As emphasised by Davidson Tomarken and colleagues (Davidson1993 Tomarken Davidson Wheeler amp Doss 1992 Tomarken David-son Wheeler amp Kinney 1992) there are a number of methodogicalapproaches that need to be considered when asymmetric patterns ofEEG activity do not emerge Some of these considerations are relevantto the design of blood macr ow studies as well One issue is that of stability(ie the extent to which brain activity remacr ects a stable individual pattern)This can be addressed by obtaining a sufreg cient number of baseline samplesas well as test-retest data For example Tomarken et al (1992) found thatonly participants with a stable pattern of anterior asymmetry across twomeasurement occasions showed an association between left frontalhypoactivation and more dispositional negative affect as measured bythe PANAS
It is also important to note that for the most part EEG studies haveexamined resting or baseline activity Davidson (1993) has pointed out
430 HELLER AND NITSCHKE
that baseline levels of frontal activation asymmetry may remacr ect individualtendencies toward a valenced affective response but this relationship maynot be observed in the absence of a specireg c environmental elicitor Thushe emphasises the need to examine the relationship between individualdifferences in patterns of asymmetry and affective response to emotionallychallenging tasks As applied to the pattern of posterior reg ndings theseconsiderations suggest the possibility that more robust differences inposterior brain activity might be yielded by studies measuring brainactivity while participants perform tasks for which depressed peopleshow decrements Additional methodological approaches highlighted byDavidson and colleagues such as the use of extreme groups and classifica-tion of participants according to the affect measures as opposed to the brainasymmetry measures would also be useful in future studies of brainactivity
A number of other possible explanations for the failure to reg nd con-sistent decreases in parieto-temporal activity during depression have beenproposed (Heller 1990 1993b) One of these explanations focused on theseverity of the depression This supposition was based on the observationthat posterior asymmetries were more likely to be reported when depres-sion was relatively severe Heller suggested that under less extremeaffective conditions task-induced activity predominates masking theaffect-induced activity More recently Heller et al (1995) suggestedthat severity per se is not the issue rather because the comorbiditywith anxiety decreases as severity of depression increases (Hiller Zau-dig amp Rose 1989) the opposing effects of anxiety on posterior brainactivity diminishes
Heller (1993b) also suggested that the inconsistencies in the depressionand anxiety literatures may be due to signireg cant differences in regionalbrain activity that covary more strongly with dispositional characteristicsassociated with emotional functioning (eg extraversion) than with situa-tionally induced emotions Several studies have reported perceptual(Berenbaum amp Williams 1994 Heller amp Nisenson 1993) and EEG(Nitschke et al 1994) asymmetries for extraversion Nitschke et al alsofound that extraversion moderated the relationship between anxiety andEEG activity across both frontal central and parietal regions For anxiousparticipants high extraversion was associated with less activity (morealpha) whereas for nonanxious (and nondepressed) controls high extra-version was associated with more activity (less alpha) These two sets ofreg ndings albeit preliminary provide evidence for the need to acknowledgeregional differences in brain function for personality characteristics suchas extraversion (for a review see Nitschke Heller amp Miller in prep) andpoint to the possible signireg cance of personality in understanding regionalbrain activity in affective and anxiety disorders
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 431
Subtypes of Depression and Anxiety
Another promising avenue for elucidating the discrepancies reported in theliterature for both depression and anxiety is the analysis of subtypes
Depression Many distinctions for depression have been proposed overthe years including most notably bipolar versus unipolar endogenousversus exogenous and neurotic versus reactive The current Diagnostic andstatistical manual of mental disordersplusmn fourth edition (DSMIV APA 1994)subdivides unipolar depression into major depressive disorder and dysthy-mic disorder and provides further specireg city with criteria for severity andfor psychotic melancholic catatonic or atypical features
Perhaps the most useful distinction in physiological research on depres-sion has been the comparison of melancholic to nonmelancholic depres-sion The dereg ning feature of melancholic depression is anhedonia which isthe inability to experience pleasure Current consensus as indicated by theDSMIV (APA 1994) suggests that melancholia is most strongly charac-terised by a loss of pleasure in all activities and a lack of responsivity topleasurable stimuli Many other symptoms may be shared by both melan-cholic and nonmelancholic depression including depressed mood distur-bances in eating and sleeping patterns low self-esteem inability toconcentrate psychomotor retardation fatigue indecisiveness feelings ofguilt and thoughts about death There is current controversy over whethermelancholic depression is qualitatively different from nonmelancholicdepression with some authors arguing that the primary distinction is oneof severity (Zimmerman Coryell amp Pfohl 1986) Nonetheless neuropsy-chological and neurophysiological studies have been consistent in demon-strating differences between melancholic and nonmelancholic depression
In a comprehensive series of studies Bruder has shown that perceptualasymmetries differ for different subtypes of depression (for a review seeBruder 1995) On a perceptual asymmetry task melancholic patientsshowed an abnormally large right ear (left hemisphere) advantage for adichotic syllables task but no left ear (right hemisphere) advantage for acomplex tone task (Bruder 1995) The lack of a left ear advantage was dueto poor left ear accuracy consistent with other studies demonstrating righthemisphere dereg cits in depressed people However nonmelancholicdepressed patients did not differ from normal controls on either test Alongthese lines in a PET study of melancholic depression Bench et al (1992)reported less activity in the left dorsolateral prefrontal cortex compared tonondepressed controls Note that these data could be consistent with eithera qualitative or a quantitative distinction between melancholic and non-melancholic depression However other data may be more consistent witha qualitative difference between the two types of depression For example
432 HELLER AND NITSCHKE
Davidson Larson and Abercrombie (1995) reported that depression sever-ity in melancholic subjects was predicted by the extent to which activationwas decreased in the left lateral prefrontal region In contrast amongnonmelancholics depression severity was predicted by the extent to whichactivation was increased in the right medial prefrontal area
Davidson et alrsquo s (1995) data have implications for theories regardingthe relationship between pleasantunpleasant affect and anterior regions ofthe brain Research on the psychological characteristics of depression andanxiety has suggested that lack of positive affect (or more preciselyactivated pleasant affect Larsen amp Diener 1992) is unique to depression(Clark amp Watson 1991) Indeed in designing a questionnaire to measureboth general and specireg c components of depression and anxiety Watsonand colleagues (Watson 1995ab) developed a specireg c depression scalethat primarily measures anhedonia Because anhedonia is a fundamentalcharacteristic of melancholia it seems reasonable to assume that decreasedpleasant affect would accompany this type of depression If so Hellerrsquo smodel which predicts less left than right anterior activity for decreasedpleasant affect is completely consistent with the available data
However Davidson et alrsquo s (1995) data suggests the possibility thataffective valence may depend less on the relative balance of activationbetween the left and right anterior regions and more on specireg c systemsassociated with each hemisphere Whereas melancholic depression ischaracterised by decreased pleasant affect depression without melan-cholic features may be better characterised by increased unpleasantaffect Following this line of reasoning Hellerrsquo s model which in itscurrent form argues for a link between anterior asymmetry and the valencedimension would not distinguish between melancholic and nonmelan-cholic depression If affective valence depends more on specireg c systemsassociated with each hemisphere as Davidson et alrsquo s data suggest Hellerrsquo smodel may need to be expanded to remacr ect the unique contributions of theleft and right hemispheres to pleasantunpleasant affect For example itmight be more accurate to link the left anterior region with the activatedpleasant affect (ie positive affect) octant of the circumplex model ofemotion and the right anterior region with the activated unpleasant affect(ie negative affect) octant (see Larson amp Diener 1992)
Because few studies have specireg cally examined the full range of poten-tial affective descriptors in relation to patterns of brain activity theseissues remain to be resolved In most research only isolated portions ofthe circumplex have been sampled For example because positive andnegative affect as measured by the PANAS provide information regardingonly two of the four quadrants in the circumplex model studies that do notcollect data regarding the other two quadrants cannot shed light on howwell they might have accounted for patterns of brain activity
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 433
Anxiety As reviewed earlier studies that examine regional brainactivity in anxiety are even more plagued by inconsistencies than studiesof depression In keeping with the theme of this article we have argued thatby decomposing anxiety into subtypes on the basis of the psychologicalliterature we can potentially resolve some of these discrepancies
In previous work we have suggested that it may be important toconsider the type of anxiety that is being examined (Heller et al 19951997a) In particular we argued that regional brain activity might bedifferent for anxious arousal (eg panic) and anxious apprehension (egworry) types of anxiety that appear to differ signireg cantly in psychologicaland physiological characteristics (eg Barlow 1988 1991 Eysenck 1957Klein 1987 Watson et al 1995b) Anxious arousal has been described byWatson et al as being distinguished by symptoms of physiological hyper-arousal and somatic tension Panic attacks and high-stress states would fallinto this category In contrast anxious apprehension involves worry and ischaracterised by verbal rumination typically about future events This typeof anxiety would be characteristic of obsessive-compulsiveness general-ised anxiety states and trait anxiety (as identireg ed by self-reports of anxiousapprehension and worry on various questionnaires) As noted by Heller etal (1997a) the two kinds of anxiety are not mutually exclusive Extremedegrees of anxious apprehension (eg being worried or fearful about thefuture) may prompt experiences of anxious arousal (eg somatic symptomsand exaggerated physiological responses to stressful events)
We have argued that when the literature on brain activity in anxiety isreviewed in the light of this distinction between anxious arousal andanxious apprehension most of the studies reporting greater right hemi-sphere activity have either studied panic attacks (eg Reiman et al 1984)or high-stress situations likely to entail anxious arousal (eg Tucker et al1977) On the other hand studies indicating greater left hemisphere activity(primarily anterior) have examined OCD (eg Baxter et al 1987) GAD(Wu et al 1991) or questionnaire-dereg ned trait anxiety indexing anxiousapprehension (eg Tucker et al 1978) Indeed as mentioned earlierstudies have highlighted various regions in both hemispheres As aresult it is difreg cult to pinpoint a particular region that is clearly a focalpoint in anxiety Nonetheless the data do suggest that right hemisphereregions emerge as particularly salient in some types of anxiety (ie anxiousarousal) whereas left hemisphere regions emerge as particularly salient inother types of anxiety (ie anxious apprehension)
In a direct test of this hypothesis (Heller et al 1997a) we selectedparticipants on the basis of self-reported trait anxiety using a measureindexing anxious apprehension in a between-subjects design We thenmanipulated anxious arousal on a within-subjects basis Specireg cally wecontrasted brain activity (EEG alpha) during rest periods to brain activity
434 HELLER AND NITSCHKE
during a task in which participants listened to emotional narrativesdesigned to elicit anxious arousal (ie fear and sad narratives that hadpreviously received unpleasant valence and high arousal ratings)
For frontal activity group differences were unaffected by the narrativesAnxious participants showed a larger asymmetry in favour of the lefthemisphere than controls during the rest and listen periods of all narrativetypes as well as for eight minutes of resting baseline preceding thenarrative task Examination of the data indicated that the increased asym-metry was primarily due to a decrease in right hemisphere activity for theanxious participants Nonetheless the increase in magnitude of the asym-metry favouring the left hemisphere found for anxious participants isconsistent with the studies reviewed earlier that reported or inferredrelatively greater left than right anterior activity in anxious subjectsclassireg ed on the basis of worry or trait anxiety (eg Carter et al 1986Tucker et al 1978 Tyler amp Tucker 1982) Our frontal data suggest that itwill be important to investigate whether anxious apprehension is bestcharacterised by the magnitude of the asymmetry favouring the left hemi-sphere or by the level of left hemisphere activity in isolation
In contrast the parietal data supported our prediction regarding theassociation of anxious arousal with greater right hemisphere activityWhen listening to the fear and sad narratives anxious participants showeda selective increase in right parietal activity that was not demonstrated bycontrols The right parietal effect however was not obtained for neutraland happy narratives that had previously received pleasant valence and lowarousal ratings These reg ndings are consistent with those studies that havereported or inferred greater right hemisphere activity during anxious statescharacterised by panic and stress-inducing conditions (eg Reiman et al1984 Stewart et al 1988 Tucker et al 1977)
An important question here has to do with the discrepancies in frontalasymmetries that have been reported Like others (eg Baxter et al 1987Gur et al 1988 Wu et al 1991) we did not reg nd heightened right anterioractivity in anxiety However a number of other studies have (eg Davidsonet al 1996 Reivich et al 1983 Tomarken amp Davidson 1994) Asdiscussed earlier we have previously argued that anterior regions areuniquely involved in the valence dimension of emotion with pleasantaffect associated with greater left activity and unpleasant affect associatedwith greater right activity (eg Heller 1990) Therefore we would expectheightened right anterior activity in an anxious group only if they differedfrom controls on measures indexing affective valence In our recent studythe nondepressed anxious group did not differ from controls on either theGBI Dysthymia scale or the PANAS PA and NA scales
It may also be important to measure the level of anxious apprehension inthe sample studied For example Tomarken and Davidson (1994) found
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 435
that with Marloweplusmn Crowne scores controlled trait anxiety failed to predictfrontal asymmetry However their groups were divided according to amedian split
In summary our results directly support the proposed distinctionbetween anxious apprehension and anxious arousal as types of anxietythat may account for signireg cant variance in patterns of brain activityThey suggest furthermore that Hellerrsquo s model should be rereg ned to reflectthe likelihood that parietal activation will covary with anxious arousal butnot with anxious apprehension If these distinctions are accurate we wouldexpect to observe increases in right parietotemporal activity only underconditions when anxiety is accompanied by increased arousal
Thus in homodynamic and EEG studies a failure to reg nd increased rightposterior activity in anxiety is not surprising as variations in anxiousarousal are likely to be large Some individuals may experience anxiousarousal frequently especially those with disorders more characteristic ofanxious arousal such as panic disorder phobias or post-traumatic stressdisorder Others although apprehensive may rarely experience anxiousarousal In addition the situation may vary in the degree to which it elicitsanxious arousal Being forced to give a speech is likely to induce a greatdeal of anxious arousal in social phobics thus we would have predicted asignireg cant increase in right parietal activity precisely as found by David-son et al (submitted) Note that this effect was only found during theAnticipation condition not during baseline Planning or after thespeech During the Anticipation condition subjects listened while a voiceannounced every 30 seconds for three minutes the time remaining beforethey would be given a topic and allowed to plan for their speech Pre-sumably this period is the most stressful and the most likely to result inanxious arousal Indeed the correlation between heart rate acceleration tothe Anticipation condition and the increase in state anxiety was signireg cant
More recently Nitschke Heller Imig and Miller (submitted) haveattempted to identify anxious arousal and anxious apprehension usingquestionnaires Data from 783 students for a battery of anxiety and depres-sion questionnaires provided strong support for the anxiety distinctionproposed here at least on a psychological level of analysis The Moodand Anxiety Symptom Questionnaire-Anxious Arousal scale (MASQWatson et al 1995ab) was chosen as a measure of anxious arousalwhereas the Penn State Worry Questionnaire (PSWQ Meyer MillerMetzger amp Borkovec 1990 Molina amp Borkovec 1994) was included asa measure of anxious apprehension The correlation between the two scales(r = 20) was much lower than those involving the other more generalscales of anxiety included in the study (MASQ-General Anxiety scaleBeck Anxiety Inventory and the State-Trait Anxiety Inventory-Traitscale ranging from r = 47 to r = 65) These results further support the
436 HELLER AND NITSCHKE
argument that anxious arousal and anxious apprehension are distinctanxiety states
However new questions are highlighted by this distinction betweenanxious arousal and anxious apprehension particularly in the light ofHellerrsquo s model If anxious apprehension is associated with left anterioractivity the overall pattern of regional brain activity appears to be identicalto the brain activity associated with pleasant emotional states (ie more leftthan right anterior activity) However anxious apprehension might beexpected to fall in the portion of the circumplex model of emotionassociated with activated unpleasant affect (as suggested by adjectivessuch as fearful nervous jittery anxious) Below we suggest a numberof theoretical possibilities for resolving this apparent discrepancy all ofwhich remain to be examined empirically
In general it will be important to clarify the degree to which anxiousapprehension is in fact characterised by unpleasant versus pleasant valenceas well as high versus low arousal In addition it may be helpful toconsider these anterior asymmetries from the approach-withdrawal frame-work employed by Davidson (1992ab Davidson amp Tomarken 1989)Under conditions where anxious apprehension leads to approach tenden-cies one might expect an asymmetry in favour of the left anterior region Incontrast when anxious apprehension is associated with withdrawal tenden-cies (perhaps as anxious arousal increases) an asymmetry may emerge infavour of the right anterior region Conceivably the degree to whichanxious apprehension is characterised by approach versus withdrawaltendencies would depend on the degree of stress and discomfortinvolved the personrsquo s coping style and other factors With regard tovalence we would predict that anxious apprehension characterised byapproach tendencies would be accompanied by relatively pleasant affectAnxious apprehension accompanied by withdrawal tendencies in contrastwould be accompanied by relatively unpleasant affect
It is also possible that instead of a direct link between anxious appre-hension and approach behaviour they share more fundamental character-istics associated with left anterior functions For example one might arguethat they both involve sequential thinking and the development of hier-archically structured plans focusing on future concerns (for a review ofrelevant left anterior functions see Tomarken amp Keener this Issue)
Another important issue has to do with the nature of the asymmetricpattern of activity As reviewed earlier we found that the asymmetry infavour of the left hemisphere for participants classireg ed on the basis of traitanxiety was due to a relative decrease in right hemisphere activity Thusthe left frontal region of the anxious participants was not signireg cantly moreactive than that of controls It will therefore be important to examine thepossibility that absolute activity of the left anterior region plays a more
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 437
important role in the modulation of valence (eg Davidson et al 1995)than of anxious apprehension
Another possibility is that independent left anterior systems are asso-ciated with anxious apprehension versus pleasant affect Liotti and Tucker(1995) review theories that there may be multiple functionally segregatedparallel frontal cortico-striato-thalamic circuits some associated withmotor functions some with set shifting and behavioural inhibition andsome with emotional information processing and hedonic responsivity tostimuli Possibly one of these circuits remacr ects activation associated withpleasant affect and another remacr ects activation associated with anxiousapprehension andor approach Different neurotransmitter systems couldbe involved More precise methods of localisation would therefore benecessary to distinguish the unique brain activity associated with each ofthese psychologicalbehavioural phenomena
Comorbidity of Depression and Anxiety
In considering the issue of decomposing emotional states with regard to thesymptomatology in affective and anxiety disorders we have previouslyargued that issues of comorbidity of depression and anxiety are crucial(Heller et al 1995) A review of the literature reveals that rates of comor-bidity are extremely high (Alloy Kelly Mineka amp Clements 1990 Helleret al 1995 Hiller et al 1989 Katon amp Roy-Byrne 1991) Even in our veryrecent work (Nitschke et al submitted) using the MASQ a measuredesigned by Watson and colleagues to avoid symptom overlap in themeasurement of anxiety and depression (Watson et al 1995ab) comor-bidity was fairly pronounced in a large college student sample The MASQAnxious Arousal (AA) and Anhedonic Depression (AD) scales were used asmeasures of those constructs and the PSWQ provided the measure ofanxious apprehension Of the 125 (out of 783) participants scoring abovethe 80th percentile on at least one of the three scales of interest and belowthe 50th percentile on any where the 80th percentile cut-off was not metfully 50 demonstrated comorbidity by scoring above the 80th percentileon at least two of the scales This rate of comorbidity emerged despite thelow correlations found between the three scales (MASQ AA and MASQAD r = 20 MASQ AA and PSWQ r = 20 MASQ AD and PSWQ r =36) relative to the correlations found for the other more general measuresof depression including the MASQ General Mixed General Depression andGeneral Anxiety scales Beck Depression Inventory Beck Anxiety Inven-tory General Behaviour Inventory-Dysthymia scale and the State-TraitAnxiety Inventory-Trait scale (ranging from r = 47 to r = 76)
Because at least some studies have found opposite patterns of activityfor the right posterior region during depression versus during anxiety (ie
438 HELLER AND NITSCHKE
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
speaking behaviour like a symphony is the output of an orchestra ofneurones Each region of the brain like each section of the orchestra has aparticular role to play and a particular manner in which it affects the overallcomposition To understand the making of a symphony we must identifythe component parts
Emotion like many cognitive processes is not a unidimensional con-struct Rather the term subsumes a variety of distinct phenomena includingphysiological events (eg changes in heart rate or skin conductance) feelingstate and modes of information processing Thus emotion is unlikely to` liversquorsquo in some specireg c location in the brain instead various brain regionsare likely to be engaged in different aspects of it It is therefore crucial thatbrain activity in emotion be examined in the light of a thorough under-standing of the multiple processes and brain regions involved The degree towhich we can disentangle the various subcomponents of the emotionalphenomena we are investigating determines the degree to which we candiscern the precise neural mechanisms of importance
In this article we will address the issues of depression and anxiety fromthe foregoing perspective We will emphasise the argument that differentsubcomponents of the symptomatology of affective and anxiety disordersare associated with distinct regions of the brain The degree to which thesesubcomponents are conmacr ated or co-occur will obscure our ability toidentify the neural mechanisms associated with particular disordersFurthermore the more we can be specireg c about the quality or nature ofthe symptomatology the more we are likely to be able to specify thecritical brain mechanisms
In turn a better understanding of the neural mechanisms involved caninform our understanding of the symptoms of depression and anxiety Forexample if right posterior regions of the brain prove to be dysfunctional indepression we might expect that depressed people would have difreg cultyprocessing narrative information a task that has been shown to rely inpart on this area of the brain (Heller amp Nitschke 1997) Because narrativeinformation processing is likely to play an important role in many ther-apeutic interventions the signireg cance of a dereg cit in this area is clear
Our approach is exemplireg ed by a neuropsychological model developedby Heller (1986 1990 1993ab) in which psychological theories of emo-tion were used to decompose emotional states into two componentsvalence and arousal This model integrates the dimensional circumplextheory of emotion (based primarily on self-report for a review see Larsonamp Diener 1992) with neuropsychological data on cognitive emotional andautonomic functioning during different affective states Based on evidencefrom electroencephalographic (EEG) blood macr ow and lesion studies themodel posits that the valence dimension (pleasant unpleasant) is dependenton functions of the anterior regions Bias toward a particular valence pole
422 HELLER AND NITSCHKE
is associated with the relative activity level of the left and right hemi-spheres When the left frontal region is active relative to the right affectivevalence is pleasant whereas when the right frontal region is active relativeto the left affective valence is unpleasant The association of valence withasymmetric activity of the anterior regions has been extensively studied(for reviews see Davidson 1992ab Heller 1990) and will be furtherreviewed in the course of this article
In contrast the arousal dimension (or activation dimension Larsen ampDiener 1992) is posited to depend on right parietotemporal regions ofthe brain More activity in this region is associated with higher self-reported arousal whereas less activity is associated with lower self-reported arousal This aspect of the model was based on theoreticalwork by numerous authors (eg Heilman Schwartz amp Watson 1978Levy Heller Banich amp Burton 1983 Tucker 1981) suggesting a specialrole for the right hemisphere in emotion-related arousal functions Thearousal component of Hellerrsquo s model derives empirical support fromnumerous additional studies examining both brain-damaged patients andnormal subjects (for reviews see Gainotti Caltagirone amp Zoccolotti1993 Wittling 1995)
To briemacr y review this evidence self-reports of emotion-related arousaland the right hemisphere have been investigated directly by HellerNitschke and Lindsay (1997b) Greater emphasis on right hemisphereprocessing as remacr ected in a left hemispatial bias on a free-vision task offace processing was associated with higher levels of self-reported arousalon several indices derived from the Proreg le of Mood States (an emotionaladjective checklist) Self-reported arousal has also been shown to beassociated with a number of autonomic functions particularly skin con-ductance (Greenwald Cook amp Lang 1989 Lang Greenwald Bradly ampHamm 1993) In turn numerous studies have suggested a link between theright hemisphere and skin conductance right brain-damaged patients haveconsistently been found to display reduced skin conductance responses toemotional stimuli (Caltagirone Zoccolotti Originale Daniele amp Mammu-cari 1989 Heilman Schwartz amp Watson 1978 Morrow Vrtunski Kimamp Boller 1981 Myslobodsky amp Horesh 1978 Valenstein amp Heilman1984 Zoccolotti Caltagirone Benedetti amp Gainotti 1986 ZoccolottiScabini amp Violani 1982) Compatible with these results there is consider-able evidence for greater autonomic conditioning when conditioned emo-tional stimuli are presented to the right versus the left hemisphere (Dawsonamp Schell 1982 Johnsen amp Hugdahl 1991 1993) There is also evidencethat cardiovascular functions associated with emotional responses are moredependent on the right than on the left hemisphere (Caltagirone et al1989 Hugdahl Franzon Andersson amp Walldebo 1983 Wittling 1990Zoccolotti et al 1986)
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 423
At the time of its conception Hellerrsquo s model attempted to address anumber of conmacr icts in the literature Which hemisphere was involved inemotion and in what manner was a matter of much debate Based onevidence that the right hemisphere was specialised to process emotionalinformation some researchers were arguing that all emotional functionswere dependent on the right hemisphere Representative of this perspec-tive Levy et al (1983) had obtained behavioural evidence on tasksassociated with parieto-temporal functions that led us to conclude thatindividuals displaying relatively reduced right hemisphere activationwere predisposed to unpleasant emotions whereas individuals displayingenhanced right hemisphere activation were predisposed to pleasant emo-tions Based on these reg ndings we posited that reduced activation of theright hemisphere was associated with depression This hypothesis wassupported by evidence that depressed people display dereg cits in cognitivefunctions associated with the right hemisphere on neuropsychological testsand in lateralised tachistoscopic paradigms (for reviews see Heller 19901993ab Tucker 1981)
In direct contrast EEG studies indicated that compared to the lefthemisphere right hemisphere activation was increased in depressedmood states particularly for anterior regions (for a review see David-son 1984) When mood was euthymic however activation of the righthemisphere was decreased relative to the left hemisphere These resultsbuilt on a set of observations systematically described by Gainotti(1972) noting very different emotional behaviours in people with rightversus left brain damage Left hemisphere damage was associated withthe so-called ` catastrophic reactionrsquorsquo in which patients were described tobe emotionally volatile and especially prone to depression and crying Incontrast right hemisphere damage was observed to be frequently accom-panied by a ` euphoricrsquo rsquo or ` indifferencersquo rsquo reaction Similar emotionalresponses have been described in patients undergoing the WADA orsodium amytal test Injections of sodium amytal that deactivate the lefthemisphere are associated with the ` catastrophicrsquo rsquo reaction whereasthose that deactivate the right hemisphere are associated with the` indifferenceeuphoriarsquo rsquo reaction (eg Lee Loring Meador Flanigin ampBrooks 1988)
The EEG and lesion studies inspired what have been referred to as` valencersquo rsquo theories (for a review see Borod 1993) In one variant ofthis type of theory the left hemisphere was thought to be specialised forpleasant affect and the right hemisphere was thought to be specialisedfor unpleasant affect (eg Sackeim et al 1982) To account for the lesiondata it was assumed that brain damage immobilises or diminishes thefunctioning of the damaged hemisphere and that the behaviour observedremacr ects the functioning of the intact hemisphere more or less on its own
424 HELLER AND NITSCHKE
In another variant of a valence theory Tucker (1981) concluded thatunpleasant and pleasant emotions were associated with the left and righthemispheres respectively a reversal of the valence theory described ear-lier Here the lesion data were explained by arguing that the observedbehaviour was not a function of the intact hemisphere but rather remacr ectedemotional processes of the subcortical areas on the damaged side that werereleased from cortical inhibition because of the lesion The EEG datawould be explained by assuming that increased anterior activity on oneside or the other would inhibit the affect associated with the correspondingipsilateral subcortical region (unpleasant for the left pleasant for the right)By this reasoning it follows that the affective effects of a lateralisedsubcortical lesion should be opposite to those of a cortical lesion (eg alesion to left subcortical structures should be associated with a suppressionof activity in this area leading to a decrease in unpleasant affect) How-ever several studies have found left anterior subcortical lesions to beassociated with depression (for a review see Liotti amp Tucker 1995)
In attempting to account for the co-occurrence in depression of increasedactivation in right anterior regions and dereg cits on cognitive tasks specia-lised to the right hemisphere Tucker (1981) argued that frontal regions havebeen traditionally viewed as having inhibitory regulatory effects on poster-ior regions By distinguishing valence from arousal as components ofemotion and localising these dimensions to anterior and posterior regionsof the brain respectively Hellerrsquo s (1986 1990) model also emphasised acaudal distinction However the model has additional explanatory powerAlthough inhibitory mechanisms of the anterior regions may well be play-ing a role in affective regulation (also see Tomarken amp Keener this Issue)the concept of a (relatively) independent arousal dimension associated withthe posterior right hemisphere allows us to model the patterns of brainactivity associated with all the emotions as represented on the circum-plex Indeed the patterns of activity that have been reported for depressionand anxiety cannot be fully accounted for by invoking an inhibitorymechanism For example if the inhibitory function of the right anteriorregion is the driving force behind right posterior activity in differentaffective states we would expect anxiety (which like depression is oftenaccompanied by increased right anterior activity) to be similarly accompa-nied by decreased right posterior activity However the evidence suggeststhat anxiety is often associated with increased activity in this region (asreviewed later) Thus by positing an independent arousal dimensionHellerrsquo s model provides an explanation for the patterns of brain activitythat have been observed for anxiety as well as depression
Tucker and colleagues have furthermore emphasised the important roleof cortical-subcortical interactions in producing the asymmetries in corticalactivity Hellerrsquo s model does not address the mechanisms by which the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 425
asymmetries associated with affective valence are generated indeed itcould well be the case that the relationship between activity in the anteriorregions and valence is mediated largely via interaction with subcorticalstructures Furthermore inhibitory mechanisms may be important in thisregard
Note that the approach we are emphasising in this article is compatiblewith Davidson and colleaguesrsquo attempts to decompose emotional statesaccording to the degree that they involve approach versus withdrawalbehaviours (Davidson 1992 Davidson amp Tomarken 1989 Tomarken ampKeener this Issue see also Kinsbourne 1988) These authors have arguedthat approach and withdrawal are the fundamental behavioural substratesassociated with affective valence However this theoretical formulationclearly emphasises the anterior cortical asymmetries associated with plea-santunpleasant valence Although Davidson has addressed posterior asym-metries in his work (eg Davidson Schaffer amp Saron 1985 Davidson1992a) the approach-withdrawal model does not focus on the patterns ofposterior brain activity that have been documented for depression andanxiety Thus considering the arousal dimension in addition to theapproach-withdrawal distinction may address the data for the posteriorregions more comprehensively
The present article has several goals We will highlight persistent andemerging discrepancies in the reg eld of brain function in depression andanxiety and suggest that turning to the psychological literature on thespecireg c symptomatology and comorbidity of these disorders can help usresolve at least some of them In the process we will revisit Hellerrsquo s modelof emotion with a critical look at how well it has been able to account forthe wealth of recent reg ndings in the reg eld Some modireg cations to this modelwill be proposed and continuing questions will be explicated
REGIONAL BRAIN ACTIVITY
Anterior and posterior asymmetries in cortical activity are often but notalways reported in depression and anxiety As a result it has been difreg cultto identify precisely the patterns of brain activity that are important In thesections following we highlight the discrepancies in the reg ndings foranterior and posterior asymmetries We then discuss possible causes forthese discrepancies and suggest ways to disentangle the results
Depression
Anterior asymmetries for EEG alpha have often been found in studies ofdepression (eg Allen Iacono Depue amp Arbisi 1993 Henriques ampDavidson 1990 1991 Schaffer Davidson amp Saron 1983) Across these
426 HELLER AND NITSCHKE
studies depression has been associated with more right than left anterioractivity Numerous positron emission tomography (PET) studies have alsoreported less left anterior activity (eg Baxter et al 1985 1989 Bench etal 1992 Bench Friston Brown Frackowiak amp Dolan 1993 George etal 1994 Martinot et al 1990) which is consistent with the EEG reg ndingsearlier However other studies have reported reg ndings in the oppositedirection for depression (Drevets et al 1992) and sad mood (George etal 1995) Null effects for depression have also been found (eg NitschkeHeller Etienne amp Miller 1995 Tomarken amp Davidson 1994)
The reg ndings for the parieto-temporal regons have been even morevariable In EEG and blood macr ow studies that report positive reg ndingsdepression is typically associated with less right posterior activity (egFlor-Henry 1979 Henriques amp Davidson 1990 Post et al 1987 Uytden-hoef et al 1983) Numerous studies however report no effects for thisregion of the brain (for a review see Davidson amp Tomarken 1989)
Reports of reduced right parieto-temporal activity although not alwayspresent are consistent with neuropsychological studies often indicatingimpaired or suppressed right hemisphere function in depressed peopleOn neuropsychological tasks thought to depend on right parietotemporalfunctioning depressed people have often been found to display decrementsin performance (eg Berndt amp Berndt 1980 Flor-Henry 1976 GoldsteinFilskov Weaver amp Ives 1977 Gruzelier Seymour Wilson Jolley ampHirsch 1988 Kronfol Hamsher Digre amp Waziri 1978) Similar findingsfor sad mood in nondepressed people have also been reported for righthemisphere tasks (Tucker Stenslie Roth amp Shearer 1981)
Studies that have directed information to one or the other hemisphereusing lateralised paradigms such as dichotic listening or visual half-reg eldpresentation have also found specireg c dereg cits for the right hemisphere indepressed people (for a review see Bruder 1995) Left-ear decrements orabsent left-ear (right hemisphere) advantages have been reported fordepression on nonverbal dichotic listening tasks (eg Bruder et al1989) Similarly left visual reg eld (right hemisphere) dereg cits in reactiontime and accuracy for nonverbal stimuli have been reported for depressedpeople (Bruder et al 1989 Liotti Sava Rizzolati amp Caffarra 1991)Comparable results have also been obtained in studies of nondepressedpeople who undergo mood inductions (Banich Stolar Heller amp Goldman1992 Ladavas Nicoletti Umilta amp Rizzolatti 1984) Reaction time andaccuracy for the right hemisphere are selectively impaired in sad moods
On the Chimeric Faces Task (CFT) a free-vision task of face processingthat typically elicits a left hemispatial bias suggesting greater right hemi-sphere activation Jaeger Borod and Peselow (1987) found smaller lefthemispatial biases for depressed patients than nonpsychiatric controlsUsing the same task Heller Etienne and Miller (1995) reported that
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 427
high-depressed participants had smaller left hemispatial biases than low-depressed participants Due to the nature of the CFT hemispatial biases arenonspecireg c as to whether they remacr ect a decrease in activity for the con-tralateral hemisphere or an increase in activity for the ipsilateral hemi-sphere (ie a reduction in the magnitude of the left hemispatial bias couldremacr ect either reduced right hemisphere or increased left hemisphere activ-ity) However based on the evidence (reviewed earlier) for selectivedereg cits in performance on right hemisphere tasks relatively poor perfor-mance for the left-visual reg eld on tachistoscopic tasks and reduced EEGactivity for right posterior regions the most likely interpretation of the CFTreg ndings is that they remacr ect diminished right posterior activity
In summary although anterior asymmetries have often been reported indepression the more common reg nding of higher left than right activation isnot invariably found Furthermore there is a signireg cant discrepancybetween the number of studies that have reported parieto-temporal asym-metries in activation which have been relatively few and the number ofstudies that have reported dereg cits in right hemisphere information proces-sing which have been many Also worthy of note is whether studies ofclinical depression psychometrically dereg ned depression and sad mood areaddressing the same phenomenon The number of consistencies acrossthese populations in the physiological literature reviewed here suggestssome commonality on the other hand the inconsistencies could be in partdue to the heterogeneity between the samples studied
Anxiety
Research on brain activity in anxious states has also been plagued byinconsistencies Again although anterior and posterior asymmetries havebeen reported no consistent pattern has emerged Studies of multisite EEGand regional cerebral blood macr ow (rCBF) using xenon-133 or PET havereported right hemisphere left hemisphere and bilateral increases inactivity during anxiety (for a review see Heller Nitschke Etienne ampMiller 1997a) Other studies have found no differences in brain activity(Gur et al 1988 Mountz et al 1989 Nordhal et al 1990 Tomarken ampDavidson 1994)
Having comprehensively reviewed this literature elsewhere (Heller etal 1997a) we will provide an overview of those studies reporting differ-ential hemispheric involvement in anxiety A subset of these suggest anasymmetry in favour of right hemisphere activation in various regions(frontal lateral prefrontal precentral frontal anterior temporal parietaloccipital and parahippocampal) Heightened right hemisphere activity hasbeen reported for panic disorder patients (Reiman Raichle Butler Hers-covitch amp Robins 1984) panic disorder patients experiencing a sodium
428 HELLER AND NITSCHKE
lactate-induced panic attack (Stewart Devous Rush Lane amp Bonte1988) generalised anxiety disorder (GAD) patients during an anxiety-stimulating task (Wu et al 1991) psychometrically dereg ned high-anxiousparticipants immediately following arterial and venous catheterisation andagain after completion of the PET scans (Reivich Gur amp Alavi 1983) andsocial phobics prior to making a public speech (Davidson MarshallTomarken amp Henriques submitted) Consistent with this pattern anti-anxiety medication was found to reduce the rate of glucose metabolismin the right hemisphere especially for the occipital and frontal regions forpatients with GAD (Buchsbaum et al 1987) Similarly in a population ofnonpsychiatric volunteers those administered diazepam intravenouslyshowed maximal decreases in rCBF in the right frontal area whereasthose injected with a placebo showed no rCBF changes (Mathew Wil-son amp Daniel 1985)
Regional brain activity can also be inferred from performance on tasksthat depend on particular brain regions On the CFT Heller et al (1995)reported that high-anxious participants had larger left hemispatial biasesthan low-anxious participants suggested to remacr ect greater right posterioractivity Similarly on a neuropsychological test battery panic disorderpatients performed less well than normal controls on left but not righthemisphere tasks (Yeudall et al 1983) Data obtained from a series ofstudies using behavioural measures to examine right versus left hemispherefunction during an experimental condition producing transient (ie state)anxiety (Tucker Antes Stenslie amp Barnhardt 1978 Tucker Roth Arne-son amp Buckingham 1977 Tyler amp Tucker 1982) are also consistent withan interpretation of greater right than left hemisphere activity in anxiety(Heller et al 1995)
However when subjects were classireg ed according to trait anxietyTucker and colleagues (Tucker et al 1978 Tyler amp Tucker 1982)interpreted the results as strongly suggestive of greater left than righthemisphere activity This interpretation is consistent with reports ofnumerous other labs reg nding increased left hemisphere activity associatedwith anxiety in various cortical and subcortical regions (orbital frontalinferior frontal anterior cingulate caudate putamen and thalamus)Heightened left hemisphere activity has been found in obsessive-compul-sive disorder (OCD) patients (Baxter et al 1987 Swedo et al 1989)GAD patients (Buchsbaum et al 1985 Wu et al 1991) and studentsclassireg ed as ` worriersrsquo rsquo on the basis of self-reports when asked to ` worryabout a specireg c topic of personal concernrsquo rsquo (Carter Johnson amp Borkovec1986)
In summary some reg ndings for anxiety indicate increased right hemi-sphere activation whereas others indicate increased left hemisphere acti-vation Indeed it may be the case that they could co-occur which could
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 429
account for the reg ndings of bilateral increases or no asymmetries In themeantime we are lacking a model to explain these discrepancies
ACCOUNTING FOR INCONSISTENCIES INBRAIN ACTIVITY
We are thus confronted with inconsistent results with regard to regionalbrain activity in both depression and anxiety In previous work we haveargued that in order to resolve these discrepancies it is helpful to turn tothe psychological literature on symptomatology in depression and anxietyIn the sections that follow we make some general observations that applyto both depression and anxiety before focusing on each more exclusively
In addressing the inconsistencies for the blood macr ow reg ndings severalconcerns about the literature are worthy of note First several of the mostfrequently cited studies in this literature sampled a restricted set of brainregions (eg Reiman et al 1984 Reivich et al 1983) Furthermore theregions of interest were not consistent from one study to the nextSecond statistical procedures employing multiple t-tests rather thanrepeated-measures ANOVAs are questionable especially with regard todrawing conclusions about hemispheric differences Third although somestudies examined selected diagnostic groups (eg major depressive dis-order OCD panic disorder GAD) most treated depression or anxiety asa unitary construct and did not consider that different types of each mightbe associated with different patterns of brain activity Moreover with theexception of recent OCD studies most did not take into account thedegree of comorbidity with depression or anxiety which may be criticalin determining the patterns of brain activity that occur (Heller et al1995 Tucker 1988)
As emphasised by Davidson Tomarken and colleagues (Davidson1993 Tomarken Davidson Wheeler amp Doss 1992 Tomarken David-son Wheeler amp Kinney 1992) there are a number of methodogicalapproaches that need to be considered when asymmetric patterns ofEEG activity do not emerge Some of these considerations are relevantto the design of blood macr ow studies as well One issue is that of stability(ie the extent to which brain activity remacr ects a stable individual pattern)This can be addressed by obtaining a sufreg cient number of baseline samplesas well as test-retest data For example Tomarken et al (1992) found thatonly participants with a stable pattern of anterior asymmetry across twomeasurement occasions showed an association between left frontalhypoactivation and more dispositional negative affect as measured bythe PANAS
It is also important to note that for the most part EEG studies haveexamined resting or baseline activity Davidson (1993) has pointed out
430 HELLER AND NITSCHKE
that baseline levels of frontal activation asymmetry may remacr ect individualtendencies toward a valenced affective response but this relationship maynot be observed in the absence of a specireg c environmental elicitor Thushe emphasises the need to examine the relationship between individualdifferences in patterns of asymmetry and affective response to emotionallychallenging tasks As applied to the pattern of posterior reg ndings theseconsiderations suggest the possibility that more robust differences inposterior brain activity might be yielded by studies measuring brainactivity while participants perform tasks for which depressed peopleshow decrements Additional methodological approaches highlighted byDavidson and colleagues such as the use of extreme groups and classifica-tion of participants according to the affect measures as opposed to the brainasymmetry measures would also be useful in future studies of brainactivity
A number of other possible explanations for the failure to reg nd con-sistent decreases in parieto-temporal activity during depression have beenproposed (Heller 1990 1993b) One of these explanations focused on theseverity of the depression This supposition was based on the observationthat posterior asymmetries were more likely to be reported when depres-sion was relatively severe Heller suggested that under less extremeaffective conditions task-induced activity predominates masking theaffect-induced activity More recently Heller et al (1995) suggestedthat severity per se is not the issue rather because the comorbiditywith anxiety decreases as severity of depression increases (Hiller Zau-dig amp Rose 1989) the opposing effects of anxiety on posterior brainactivity diminishes
Heller (1993b) also suggested that the inconsistencies in the depressionand anxiety literatures may be due to signireg cant differences in regionalbrain activity that covary more strongly with dispositional characteristicsassociated with emotional functioning (eg extraversion) than with situa-tionally induced emotions Several studies have reported perceptual(Berenbaum amp Williams 1994 Heller amp Nisenson 1993) and EEG(Nitschke et al 1994) asymmetries for extraversion Nitschke et al alsofound that extraversion moderated the relationship between anxiety andEEG activity across both frontal central and parietal regions For anxiousparticipants high extraversion was associated with less activity (morealpha) whereas for nonanxious (and nondepressed) controls high extra-version was associated with more activity (less alpha) These two sets ofreg ndings albeit preliminary provide evidence for the need to acknowledgeregional differences in brain function for personality characteristics suchas extraversion (for a review see Nitschke Heller amp Miller in prep) andpoint to the possible signireg cance of personality in understanding regionalbrain activity in affective and anxiety disorders
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 431
Subtypes of Depression and Anxiety
Another promising avenue for elucidating the discrepancies reported in theliterature for both depression and anxiety is the analysis of subtypes
Depression Many distinctions for depression have been proposed overthe years including most notably bipolar versus unipolar endogenousversus exogenous and neurotic versus reactive The current Diagnostic andstatistical manual of mental disordersplusmn fourth edition (DSMIV APA 1994)subdivides unipolar depression into major depressive disorder and dysthy-mic disorder and provides further specireg city with criteria for severity andfor psychotic melancholic catatonic or atypical features
Perhaps the most useful distinction in physiological research on depres-sion has been the comparison of melancholic to nonmelancholic depres-sion The dereg ning feature of melancholic depression is anhedonia which isthe inability to experience pleasure Current consensus as indicated by theDSMIV (APA 1994) suggests that melancholia is most strongly charac-terised by a loss of pleasure in all activities and a lack of responsivity topleasurable stimuli Many other symptoms may be shared by both melan-cholic and nonmelancholic depression including depressed mood distur-bances in eating and sleeping patterns low self-esteem inability toconcentrate psychomotor retardation fatigue indecisiveness feelings ofguilt and thoughts about death There is current controversy over whethermelancholic depression is qualitatively different from nonmelancholicdepression with some authors arguing that the primary distinction is oneof severity (Zimmerman Coryell amp Pfohl 1986) Nonetheless neuropsy-chological and neurophysiological studies have been consistent in demon-strating differences between melancholic and nonmelancholic depression
In a comprehensive series of studies Bruder has shown that perceptualasymmetries differ for different subtypes of depression (for a review seeBruder 1995) On a perceptual asymmetry task melancholic patientsshowed an abnormally large right ear (left hemisphere) advantage for adichotic syllables task but no left ear (right hemisphere) advantage for acomplex tone task (Bruder 1995) The lack of a left ear advantage was dueto poor left ear accuracy consistent with other studies demonstrating righthemisphere dereg cits in depressed people However nonmelancholicdepressed patients did not differ from normal controls on either test Alongthese lines in a PET study of melancholic depression Bench et al (1992)reported less activity in the left dorsolateral prefrontal cortex compared tonondepressed controls Note that these data could be consistent with eithera qualitative or a quantitative distinction between melancholic and non-melancholic depression However other data may be more consistent witha qualitative difference between the two types of depression For example
432 HELLER AND NITSCHKE
Davidson Larson and Abercrombie (1995) reported that depression sever-ity in melancholic subjects was predicted by the extent to which activationwas decreased in the left lateral prefrontal region In contrast amongnonmelancholics depression severity was predicted by the extent to whichactivation was increased in the right medial prefrontal area
Davidson et alrsquo s (1995) data have implications for theories regardingthe relationship between pleasantunpleasant affect and anterior regions ofthe brain Research on the psychological characteristics of depression andanxiety has suggested that lack of positive affect (or more preciselyactivated pleasant affect Larsen amp Diener 1992) is unique to depression(Clark amp Watson 1991) Indeed in designing a questionnaire to measureboth general and specireg c components of depression and anxiety Watsonand colleagues (Watson 1995ab) developed a specireg c depression scalethat primarily measures anhedonia Because anhedonia is a fundamentalcharacteristic of melancholia it seems reasonable to assume that decreasedpleasant affect would accompany this type of depression If so Hellerrsquo smodel which predicts less left than right anterior activity for decreasedpleasant affect is completely consistent with the available data
However Davidson et alrsquo s (1995) data suggests the possibility thataffective valence may depend less on the relative balance of activationbetween the left and right anterior regions and more on specireg c systemsassociated with each hemisphere Whereas melancholic depression ischaracterised by decreased pleasant affect depression without melan-cholic features may be better characterised by increased unpleasantaffect Following this line of reasoning Hellerrsquo s model which in itscurrent form argues for a link between anterior asymmetry and the valencedimension would not distinguish between melancholic and nonmelan-cholic depression If affective valence depends more on specireg c systemsassociated with each hemisphere as Davidson et alrsquo s data suggest Hellerrsquo smodel may need to be expanded to remacr ect the unique contributions of theleft and right hemispheres to pleasantunpleasant affect For example itmight be more accurate to link the left anterior region with the activatedpleasant affect (ie positive affect) octant of the circumplex model ofemotion and the right anterior region with the activated unpleasant affect(ie negative affect) octant (see Larson amp Diener 1992)
Because few studies have specireg cally examined the full range of poten-tial affective descriptors in relation to patterns of brain activity theseissues remain to be resolved In most research only isolated portions ofthe circumplex have been sampled For example because positive andnegative affect as measured by the PANAS provide information regardingonly two of the four quadrants in the circumplex model studies that do notcollect data regarding the other two quadrants cannot shed light on howwell they might have accounted for patterns of brain activity
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 433
Anxiety As reviewed earlier studies that examine regional brainactivity in anxiety are even more plagued by inconsistencies than studiesof depression In keeping with the theme of this article we have argued thatby decomposing anxiety into subtypes on the basis of the psychologicalliterature we can potentially resolve some of these discrepancies
In previous work we have suggested that it may be important toconsider the type of anxiety that is being examined (Heller et al 19951997a) In particular we argued that regional brain activity might bedifferent for anxious arousal (eg panic) and anxious apprehension (egworry) types of anxiety that appear to differ signireg cantly in psychologicaland physiological characteristics (eg Barlow 1988 1991 Eysenck 1957Klein 1987 Watson et al 1995b) Anxious arousal has been described byWatson et al as being distinguished by symptoms of physiological hyper-arousal and somatic tension Panic attacks and high-stress states would fallinto this category In contrast anxious apprehension involves worry and ischaracterised by verbal rumination typically about future events This typeof anxiety would be characteristic of obsessive-compulsiveness general-ised anxiety states and trait anxiety (as identireg ed by self-reports of anxiousapprehension and worry on various questionnaires) As noted by Heller etal (1997a) the two kinds of anxiety are not mutually exclusive Extremedegrees of anxious apprehension (eg being worried or fearful about thefuture) may prompt experiences of anxious arousal (eg somatic symptomsand exaggerated physiological responses to stressful events)
We have argued that when the literature on brain activity in anxiety isreviewed in the light of this distinction between anxious arousal andanxious apprehension most of the studies reporting greater right hemi-sphere activity have either studied panic attacks (eg Reiman et al 1984)or high-stress situations likely to entail anxious arousal (eg Tucker et al1977) On the other hand studies indicating greater left hemisphere activity(primarily anterior) have examined OCD (eg Baxter et al 1987) GAD(Wu et al 1991) or questionnaire-dereg ned trait anxiety indexing anxiousapprehension (eg Tucker et al 1978) Indeed as mentioned earlierstudies have highlighted various regions in both hemispheres As aresult it is difreg cult to pinpoint a particular region that is clearly a focalpoint in anxiety Nonetheless the data do suggest that right hemisphereregions emerge as particularly salient in some types of anxiety (ie anxiousarousal) whereas left hemisphere regions emerge as particularly salient inother types of anxiety (ie anxious apprehension)
In a direct test of this hypothesis (Heller et al 1997a) we selectedparticipants on the basis of self-reported trait anxiety using a measureindexing anxious apprehension in a between-subjects design We thenmanipulated anxious arousal on a within-subjects basis Specireg cally wecontrasted brain activity (EEG alpha) during rest periods to brain activity
434 HELLER AND NITSCHKE
during a task in which participants listened to emotional narrativesdesigned to elicit anxious arousal (ie fear and sad narratives that hadpreviously received unpleasant valence and high arousal ratings)
For frontal activity group differences were unaffected by the narrativesAnxious participants showed a larger asymmetry in favour of the lefthemisphere than controls during the rest and listen periods of all narrativetypes as well as for eight minutes of resting baseline preceding thenarrative task Examination of the data indicated that the increased asym-metry was primarily due to a decrease in right hemisphere activity for theanxious participants Nonetheless the increase in magnitude of the asym-metry favouring the left hemisphere found for anxious participants isconsistent with the studies reviewed earlier that reported or inferredrelatively greater left than right anterior activity in anxious subjectsclassireg ed on the basis of worry or trait anxiety (eg Carter et al 1986Tucker et al 1978 Tyler amp Tucker 1982) Our frontal data suggest that itwill be important to investigate whether anxious apprehension is bestcharacterised by the magnitude of the asymmetry favouring the left hemi-sphere or by the level of left hemisphere activity in isolation
In contrast the parietal data supported our prediction regarding theassociation of anxious arousal with greater right hemisphere activityWhen listening to the fear and sad narratives anxious participants showeda selective increase in right parietal activity that was not demonstrated bycontrols The right parietal effect however was not obtained for neutraland happy narratives that had previously received pleasant valence and lowarousal ratings These reg ndings are consistent with those studies that havereported or inferred greater right hemisphere activity during anxious statescharacterised by panic and stress-inducing conditions (eg Reiman et al1984 Stewart et al 1988 Tucker et al 1977)
An important question here has to do with the discrepancies in frontalasymmetries that have been reported Like others (eg Baxter et al 1987Gur et al 1988 Wu et al 1991) we did not reg nd heightened right anterioractivity in anxiety However a number of other studies have (eg Davidsonet al 1996 Reivich et al 1983 Tomarken amp Davidson 1994) Asdiscussed earlier we have previously argued that anterior regions areuniquely involved in the valence dimension of emotion with pleasantaffect associated with greater left activity and unpleasant affect associatedwith greater right activity (eg Heller 1990) Therefore we would expectheightened right anterior activity in an anxious group only if they differedfrom controls on measures indexing affective valence In our recent studythe nondepressed anxious group did not differ from controls on either theGBI Dysthymia scale or the PANAS PA and NA scales
It may also be important to measure the level of anxious apprehension inthe sample studied For example Tomarken and Davidson (1994) found
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 435
that with Marloweplusmn Crowne scores controlled trait anxiety failed to predictfrontal asymmetry However their groups were divided according to amedian split
In summary our results directly support the proposed distinctionbetween anxious apprehension and anxious arousal as types of anxietythat may account for signireg cant variance in patterns of brain activityThey suggest furthermore that Hellerrsquo s model should be rereg ned to reflectthe likelihood that parietal activation will covary with anxious arousal butnot with anxious apprehension If these distinctions are accurate we wouldexpect to observe increases in right parietotemporal activity only underconditions when anxiety is accompanied by increased arousal
Thus in homodynamic and EEG studies a failure to reg nd increased rightposterior activity in anxiety is not surprising as variations in anxiousarousal are likely to be large Some individuals may experience anxiousarousal frequently especially those with disorders more characteristic ofanxious arousal such as panic disorder phobias or post-traumatic stressdisorder Others although apprehensive may rarely experience anxiousarousal In addition the situation may vary in the degree to which it elicitsanxious arousal Being forced to give a speech is likely to induce a greatdeal of anxious arousal in social phobics thus we would have predicted asignireg cant increase in right parietal activity precisely as found by David-son et al (submitted) Note that this effect was only found during theAnticipation condition not during baseline Planning or after thespeech During the Anticipation condition subjects listened while a voiceannounced every 30 seconds for three minutes the time remaining beforethey would be given a topic and allowed to plan for their speech Pre-sumably this period is the most stressful and the most likely to result inanxious arousal Indeed the correlation between heart rate acceleration tothe Anticipation condition and the increase in state anxiety was signireg cant
More recently Nitschke Heller Imig and Miller (submitted) haveattempted to identify anxious arousal and anxious apprehension usingquestionnaires Data from 783 students for a battery of anxiety and depres-sion questionnaires provided strong support for the anxiety distinctionproposed here at least on a psychological level of analysis The Moodand Anxiety Symptom Questionnaire-Anxious Arousal scale (MASQWatson et al 1995ab) was chosen as a measure of anxious arousalwhereas the Penn State Worry Questionnaire (PSWQ Meyer MillerMetzger amp Borkovec 1990 Molina amp Borkovec 1994) was included asa measure of anxious apprehension The correlation between the two scales(r = 20) was much lower than those involving the other more generalscales of anxiety included in the study (MASQ-General Anxiety scaleBeck Anxiety Inventory and the State-Trait Anxiety Inventory-Traitscale ranging from r = 47 to r = 65) These results further support the
436 HELLER AND NITSCHKE
argument that anxious arousal and anxious apprehension are distinctanxiety states
However new questions are highlighted by this distinction betweenanxious arousal and anxious apprehension particularly in the light ofHellerrsquo s model If anxious apprehension is associated with left anterioractivity the overall pattern of regional brain activity appears to be identicalto the brain activity associated with pleasant emotional states (ie more leftthan right anterior activity) However anxious apprehension might beexpected to fall in the portion of the circumplex model of emotionassociated with activated unpleasant affect (as suggested by adjectivessuch as fearful nervous jittery anxious) Below we suggest a numberof theoretical possibilities for resolving this apparent discrepancy all ofwhich remain to be examined empirically
In general it will be important to clarify the degree to which anxiousapprehension is in fact characterised by unpleasant versus pleasant valenceas well as high versus low arousal In addition it may be helpful toconsider these anterior asymmetries from the approach-withdrawal frame-work employed by Davidson (1992ab Davidson amp Tomarken 1989)Under conditions where anxious apprehension leads to approach tenden-cies one might expect an asymmetry in favour of the left anterior region Incontrast when anxious apprehension is associated with withdrawal tenden-cies (perhaps as anxious arousal increases) an asymmetry may emerge infavour of the right anterior region Conceivably the degree to whichanxious apprehension is characterised by approach versus withdrawaltendencies would depend on the degree of stress and discomfortinvolved the personrsquo s coping style and other factors With regard tovalence we would predict that anxious apprehension characterised byapproach tendencies would be accompanied by relatively pleasant affectAnxious apprehension accompanied by withdrawal tendencies in contrastwould be accompanied by relatively unpleasant affect
It is also possible that instead of a direct link between anxious appre-hension and approach behaviour they share more fundamental character-istics associated with left anterior functions For example one might arguethat they both involve sequential thinking and the development of hier-archically structured plans focusing on future concerns (for a review ofrelevant left anterior functions see Tomarken amp Keener this Issue)
Another important issue has to do with the nature of the asymmetricpattern of activity As reviewed earlier we found that the asymmetry infavour of the left hemisphere for participants classireg ed on the basis of traitanxiety was due to a relative decrease in right hemisphere activity Thusthe left frontal region of the anxious participants was not signireg cantly moreactive than that of controls It will therefore be important to examine thepossibility that absolute activity of the left anterior region plays a more
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 437
important role in the modulation of valence (eg Davidson et al 1995)than of anxious apprehension
Another possibility is that independent left anterior systems are asso-ciated with anxious apprehension versus pleasant affect Liotti and Tucker(1995) review theories that there may be multiple functionally segregatedparallel frontal cortico-striato-thalamic circuits some associated withmotor functions some with set shifting and behavioural inhibition andsome with emotional information processing and hedonic responsivity tostimuli Possibly one of these circuits remacr ects activation associated withpleasant affect and another remacr ects activation associated with anxiousapprehension andor approach Different neurotransmitter systems couldbe involved More precise methods of localisation would therefore benecessary to distinguish the unique brain activity associated with each ofthese psychologicalbehavioural phenomena
Comorbidity of Depression and Anxiety
In considering the issue of decomposing emotional states with regard to thesymptomatology in affective and anxiety disorders we have previouslyargued that issues of comorbidity of depression and anxiety are crucial(Heller et al 1995) A review of the literature reveals that rates of comor-bidity are extremely high (Alloy Kelly Mineka amp Clements 1990 Helleret al 1995 Hiller et al 1989 Katon amp Roy-Byrne 1991) Even in our veryrecent work (Nitschke et al submitted) using the MASQ a measuredesigned by Watson and colleagues to avoid symptom overlap in themeasurement of anxiety and depression (Watson et al 1995ab) comor-bidity was fairly pronounced in a large college student sample The MASQAnxious Arousal (AA) and Anhedonic Depression (AD) scales were used asmeasures of those constructs and the PSWQ provided the measure ofanxious apprehension Of the 125 (out of 783) participants scoring abovethe 80th percentile on at least one of the three scales of interest and belowthe 50th percentile on any where the 80th percentile cut-off was not metfully 50 demonstrated comorbidity by scoring above the 80th percentileon at least two of the scales This rate of comorbidity emerged despite thelow correlations found between the three scales (MASQ AA and MASQAD r = 20 MASQ AA and PSWQ r = 20 MASQ AD and PSWQ r =36) relative to the correlations found for the other more general measuresof depression including the MASQ General Mixed General Depression andGeneral Anxiety scales Beck Depression Inventory Beck Anxiety Inven-tory General Behaviour Inventory-Dysthymia scale and the State-TraitAnxiety Inventory-Trait scale (ranging from r = 47 to r = 76)
Because at least some studies have found opposite patterns of activityfor the right posterior region during depression versus during anxiety (ie
438 HELLER AND NITSCHKE
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
is associated with the relative activity level of the left and right hemi-spheres When the left frontal region is active relative to the right affectivevalence is pleasant whereas when the right frontal region is active relativeto the left affective valence is unpleasant The association of valence withasymmetric activity of the anterior regions has been extensively studied(for reviews see Davidson 1992ab Heller 1990) and will be furtherreviewed in the course of this article
In contrast the arousal dimension (or activation dimension Larsen ampDiener 1992) is posited to depend on right parietotemporal regions ofthe brain More activity in this region is associated with higher self-reported arousal whereas less activity is associated with lower self-reported arousal This aspect of the model was based on theoreticalwork by numerous authors (eg Heilman Schwartz amp Watson 1978Levy Heller Banich amp Burton 1983 Tucker 1981) suggesting a specialrole for the right hemisphere in emotion-related arousal functions Thearousal component of Hellerrsquo s model derives empirical support fromnumerous additional studies examining both brain-damaged patients andnormal subjects (for reviews see Gainotti Caltagirone amp Zoccolotti1993 Wittling 1995)
To briemacr y review this evidence self-reports of emotion-related arousaland the right hemisphere have been investigated directly by HellerNitschke and Lindsay (1997b) Greater emphasis on right hemisphereprocessing as remacr ected in a left hemispatial bias on a free-vision task offace processing was associated with higher levels of self-reported arousalon several indices derived from the Proreg le of Mood States (an emotionaladjective checklist) Self-reported arousal has also been shown to beassociated with a number of autonomic functions particularly skin con-ductance (Greenwald Cook amp Lang 1989 Lang Greenwald Bradly ampHamm 1993) In turn numerous studies have suggested a link between theright hemisphere and skin conductance right brain-damaged patients haveconsistently been found to display reduced skin conductance responses toemotional stimuli (Caltagirone Zoccolotti Originale Daniele amp Mammu-cari 1989 Heilman Schwartz amp Watson 1978 Morrow Vrtunski Kimamp Boller 1981 Myslobodsky amp Horesh 1978 Valenstein amp Heilman1984 Zoccolotti Caltagirone Benedetti amp Gainotti 1986 ZoccolottiScabini amp Violani 1982) Compatible with these results there is consider-able evidence for greater autonomic conditioning when conditioned emo-tional stimuli are presented to the right versus the left hemisphere (Dawsonamp Schell 1982 Johnsen amp Hugdahl 1991 1993) There is also evidencethat cardiovascular functions associated with emotional responses are moredependent on the right than on the left hemisphere (Caltagirone et al1989 Hugdahl Franzon Andersson amp Walldebo 1983 Wittling 1990Zoccolotti et al 1986)
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 423
At the time of its conception Hellerrsquo s model attempted to address anumber of conmacr icts in the literature Which hemisphere was involved inemotion and in what manner was a matter of much debate Based onevidence that the right hemisphere was specialised to process emotionalinformation some researchers were arguing that all emotional functionswere dependent on the right hemisphere Representative of this perspec-tive Levy et al (1983) had obtained behavioural evidence on tasksassociated with parieto-temporal functions that led us to conclude thatindividuals displaying relatively reduced right hemisphere activationwere predisposed to unpleasant emotions whereas individuals displayingenhanced right hemisphere activation were predisposed to pleasant emo-tions Based on these reg ndings we posited that reduced activation of theright hemisphere was associated with depression This hypothesis wassupported by evidence that depressed people display dereg cits in cognitivefunctions associated with the right hemisphere on neuropsychological testsand in lateralised tachistoscopic paradigms (for reviews see Heller 19901993ab Tucker 1981)
In direct contrast EEG studies indicated that compared to the lefthemisphere right hemisphere activation was increased in depressedmood states particularly for anterior regions (for a review see David-son 1984) When mood was euthymic however activation of the righthemisphere was decreased relative to the left hemisphere These resultsbuilt on a set of observations systematically described by Gainotti(1972) noting very different emotional behaviours in people with rightversus left brain damage Left hemisphere damage was associated withthe so-called ` catastrophic reactionrsquorsquo in which patients were described tobe emotionally volatile and especially prone to depression and crying Incontrast right hemisphere damage was observed to be frequently accom-panied by a ` euphoricrsquo rsquo or ` indifferencersquo rsquo reaction Similar emotionalresponses have been described in patients undergoing the WADA orsodium amytal test Injections of sodium amytal that deactivate the lefthemisphere are associated with the ` catastrophicrsquo rsquo reaction whereasthose that deactivate the right hemisphere are associated with the` indifferenceeuphoriarsquo rsquo reaction (eg Lee Loring Meador Flanigin ampBrooks 1988)
The EEG and lesion studies inspired what have been referred to as` valencersquo rsquo theories (for a review see Borod 1993) In one variant ofthis type of theory the left hemisphere was thought to be specialised forpleasant affect and the right hemisphere was thought to be specialisedfor unpleasant affect (eg Sackeim et al 1982) To account for the lesiondata it was assumed that brain damage immobilises or diminishes thefunctioning of the damaged hemisphere and that the behaviour observedremacr ects the functioning of the intact hemisphere more or less on its own
424 HELLER AND NITSCHKE
In another variant of a valence theory Tucker (1981) concluded thatunpleasant and pleasant emotions were associated with the left and righthemispheres respectively a reversal of the valence theory described ear-lier Here the lesion data were explained by arguing that the observedbehaviour was not a function of the intact hemisphere but rather remacr ectedemotional processes of the subcortical areas on the damaged side that werereleased from cortical inhibition because of the lesion The EEG datawould be explained by assuming that increased anterior activity on oneside or the other would inhibit the affect associated with the correspondingipsilateral subcortical region (unpleasant for the left pleasant for the right)By this reasoning it follows that the affective effects of a lateralisedsubcortical lesion should be opposite to those of a cortical lesion (eg alesion to left subcortical structures should be associated with a suppressionof activity in this area leading to a decrease in unpleasant affect) How-ever several studies have found left anterior subcortical lesions to beassociated with depression (for a review see Liotti amp Tucker 1995)
In attempting to account for the co-occurrence in depression of increasedactivation in right anterior regions and dereg cits on cognitive tasks specia-lised to the right hemisphere Tucker (1981) argued that frontal regions havebeen traditionally viewed as having inhibitory regulatory effects on poster-ior regions By distinguishing valence from arousal as components ofemotion and localising these dimensions to anterior and posterior regionsof the brain respectively Hellerrsquo s (1986 1990) model also emphasised acaudal distinction However the model has additional explanatory powerAlthough inhibitory mechanisms of the anterior regions may well be play-ing a role in affective regulation (also see Tomarken amp Keener this Issue)the concept of a (relatively) independent arousal dimension associated withthe posterior right hemisphere allows us to model the patterns of brainactivity associated with all the emotions as represented on the circum-plex Indeed the patterns of activity that have been reported for depressionand anxiety cannot be fully accounted for by invoking an inhibitorymechanism For example if the inhibitory function of the right anteriorregion is the driving force behind right posterior activity in differentaffective states we would expect anxiety (which like depression is oftenaccompanied by increased right anterior activity) to be similarly accompa-nied by decreased right posterior activity However the evidence suggeststhat anxiety is often associated with increased activity in this region (asreviewed later) Thus by positing an independent arousal dimensionHellerrsquo s model provides an explanation for the patterns of brain activitythat have been observed for anxiety as well as depression
Tucker and colleagues have furthermore emphasised the important roleof cortical-subcortical interactions in producing the asymmetries in corticalactivity Hellerrsquo s model does not address the mechanisms by which the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 425
asymmetries associated with affective valence are generated indeed itcould well be the case that the relationship between activity in the anteriorregions and valence is mediated largely via interaction with subcorticalstructures Furthermore inhibitory mechanisms may be important in thisregard
Note that the approach we are emphasising in this article is compatiblewith Davidson and colleaguesrsquo attempts to decompose emotional statesaccording to the degree that they involve approach versus withdrawalbehaviours (Davidson 1992 Davidson amp Tomarken 1989 Tomarken ampKeener this Issue see also Kinsbourne 1988) These authors have arguedthat approach and withdrawal are the fundamental behavioural substratesassociated with affective valence However this theoretical formulationclearly emphasises the anterior cortical asymmetries associated with plea-santunpleasant valence Although Davidson has addressed posterior asym-metries in his work (eg Davidson Schaffer amp Saron 1985 Davidson1992a) the approach-withdrawal model does not focus on the patterns ofposterior brain activity that have been documented for depression andanxiety Thus considering the arousal dimension in addition to theapproach-withdrawal distinction may address the data for the posteriorregions more comprehensively
The present article has several goals We will highlight persistent andemerging discrepancies in the reg eld of brain function in depression andanxiety and suggest that turning to the psychological literature on thespecireg c symptomatology and comorbidity of these disorders can help usresolve at least some of them In the process we will revisit Hellerrsquo s modelof emotion with a critical look at how well it has been able to account forthe wealth of recent reg ndings in the reg eld Some modireg cations to this modelwill be proposed and continuing questions will be explicated
REGIONAL BRAIN ACTIVITY
Anterior and posterior asymmetries in cortical activity are often but notalways reported in depression and anxiety As a result it has been difreg cultto identify precisely the patterns of brain activity that are important In thesections following we highlight the discrepancies in the reg ndings foranterior and posterior asymmetries We then discuss possible causes forthese discrepancies and suggest ways to disentangle the results
Depression
Anterior asymmetries for EEG alpha have often been found in studies ofdepression (eg Allen Iacono Depue amp Arbisi 1993 Henriques ampDavidson 1990 1991 Schaffer Davidson amp Saron 1983) Across these
426 HELLER AND NITSCHKE
studies depression has been associated with more right than left anterioractivity Numerous positron emission tomography (PET) studies have alsoreported less left anterior activity (eg Baxter et al 1985 1989 Bench etal 1992 Bench Friston Brown Frackowiak amp Dolan 1993 George etal 1994 Martinot et al 1990) which is consistent with the EEG reg ndingsearlier However other studies have reported reg ndings in the oppositedirection for depression (Drevets et al 1992) and sad mood (George etal 1995) Null effects for depression have also been found (eg NitschkeHeller Etienne amp Miller 1995 Tomarken amp Davidson 1994)
The reg ndings for the parieto-temporal regons have been even morevariable In EEG and blood macr ow studies that report positive reg ndingsdepression is typically associated with less right posterior activity (egFlor-Henry 1979 Henriques amp Davidson 1990 Post et al 1987 Uytden-hoef et al 1983) Numerous studies however report no effects for thisregion of the brain (for a review see Davidson amp Tomarken 1989)
Reports of reduced right parieto-temporal activity although not alwayspresent are consistent with neuropsychological studies often indicatingimpaired or suppressed right hemisphere function in depressed peopleOn neuropsychological tasks thought to depend on right parietotemporalfunctioning depressed people have often been found to display decrementsin performance (eg Berndt amp Berndt 1980 Flor-Henry 1976 GoldsteinFilskov Weaver amp Ives 1977 Gruzelier Seymour Wilson Jolley ampHirsch 1988 Kronfol Hamsher Digre amp Waziri 1978) Similar findingsfor sad mood in nondepressed people have also been reported for righthemisphere tasks (Tucker Stenslie Roth amp Shearer 1981)
Studies that have directed information to one or the other hemisphereusing lateralised paradigms such as dichotic listening or visual half-reg eldpresentation have also found specireg c dereg cits for the right hemisphere indepressed people (for a review see Bruder 1995) Left-ear decrements orabsent left-ear (right hemisphere) advantages have been reported fordepression on nonverbal dichotic listening tasks (eg Bruder et al1989) Similarly left visual reg eld (right hemisphere) dereg cits in reactiontime and accuracy for nonverbal stimuli have been reported for depressedpeople (Bruder et al 1989 Liotti Sava Rizzolati amp Caffarra 1991)Comparable results have also been obtained in studies of nondepressedpeople who undergo mood inductions (Banich Stolar Heller amp Goldman1992 Ladavas Nicoletti Umilta amp Rizzolatti 1984) Reaction time andaccuracy for the right hemisphere are selectively impaired in sad moods
On the Chimeric Faces Task (CFT) a free-vision task of face processingthat typically elicits a left hemispatial bias suggesting greater right hemi-sphere activation Jaeger Borod and Peselow (1987) found smaller lefthemispatial biases for depressed patients than nonpsychiatric controlsUsing the same task Heller Etienne and Miller (1995) reported that
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 427
high-depressed participants had smaller left hemispatial biases than low-depressed participants Due to the nature of the CFT hemispatial biases arenonspecireg c as to whether they remacr ect a decrease in activity for the con-tralateral hemisphere or an increase in activity for the ipsilateral hemi-sphere (ie a reduction in the magnitude of the left hemispatial bias couldremacr ect either reduced right hemisphere or increased left hemisphere activ-ity) However based on the evidence (reviewed earlier) for selectivedereg cits in performance on right hemisphere tasks relatively poor perfor-mance for the left-visual reg eld on tachistoscopic tasks and reduced EEGactivity for right posterior regions the most likely interpretation of the CFTreg ndings is that they remacr ect diminished right posterior activity
In summary although anterior asymmetries have often been reported indepression the more common reg nding of higher left than right activation isnot invariably found Furthermore there is a signireg cant discrepancybetween the number of studies that have reported parieto-temporal asym-metries in activation which have been relatively few and the number ofstudies that have reported dereg cits in right hemisphere information proces-sing which have been many Also worthy of note is whether studies ofclinical depression psychometrically dereg ned depression and sad mood areaddressing the same phenomenon The number of consistencies acrossthese populations in the physiological literature reviewed here suggestssome commonality on the other hand the inconsistencies could be in partdue to the heterogeneity between the samples studied
Anxiety
Research on brain activity in anxious states has also been plagued byinconsistencies Again although anterior and posterior asymmetries havebeen reported no consistent pattern has emerged Studies of multisite EEGand regional cerebral blood macr ow (rCBF) using xenon-133 or PET havereported right hemisphere left hemisphere and bilateral increases inactivity during anxiety (for a review see Heller Nitschke Etienne ampMiller 1997a) Other studies have found no differences in brain activity(Gur et al 1988 Mountz et al 1989 Nordhal et al 1990 Tomarken ampDavidson 1994)
Having comprehensively reviewed this literature elsewhere (Heller etal 1997a) we will provide an overview of those studies reporting differ-ential hemispheric involvement in anxiety A subset of these suggest anasymmetry in favour of right hemisphere activation in various regions(frontal lateral prefrontal precentral frontal anterior temporal parietaloccipital and parahippocampal) Heightened right hemisphere activity hasbeen reported for panic disorder patients (Reiman Raichle Butler Hers-covitch amp Robins 1984) panic disorder patients experiencing a sodium
428 HELLER AND NITSCHKE
lactate-induced panic attack (Stewart Devous Rush Lane amp Bonte1988) generalised anxiety disorder (GAD) patients during an anxiety-stimulating task (Wu et al 1991) psychometrically dereg ned high-anxiousparticipants immediately following arterial and venous catheterisation andagain after completion of the PET scans (Reivich Gur amp Alavi 1983) andsocial phobics prior to making a public speech (Davidson MarshallTomarken amp Henriques submitted) Consistent with this pattern anti-anxiety medication was found to reduce the rate of glucose metabolismin the right hemisphere especially for the occipital and frontal regions forpatients with GAD (Buchsbaum et al 1987) Similarly in a population ofnonpsychiatric volunteers those administered diazepam intravenouslyshowed maximal decreases in rCBF in the right frontal area whereasthose injected with a placebo showed no rCBF changes (Mathew Wil-son amp Daniel 1985)
Regional brain activity can also be inferred from performance on tasksthat depend on particular brain regions On the CFT Heller et al (1995)reported that high-anxious participants had larger left hemispatial biasesthan low-anxious participants suggested to remacr ect greater right posterioractivity Similarly on a neuropsychological test battery panic disorderpatients performed less well than normal controls on left but not righthemisphere tasks (Yeudall et al 1983) Data obtained from a series ofstudies using behavioural measures to examine right versus left hemispherefunction during an experimental condition producing transient (ie state)anxiety (Tucker Antes Stenslie amp Barnhardt 1978 Tucker Roth Arne-son amp Buckingham 1977 Tyler amp Tucker 1982) are also consistent withan interpretation of greater right than left hemisphere activity in anxiety(Heller et al 1995)
However when subjects were classireg ed according to trait anxietyTucker and colleagues (Tucker et al 1978 Tyler amp Tucker 1982)interpreted the results as strongly suggestive of greater left than righthemisphere activity This interpretation is consistent with reports ofnumerous other labs reg nding increased left hemisphere activity associatedwith anxiety in various cortical and subcortical regions (orbital frontalinferior frontal anterior cingulate caudate putamen and thalamus)Heightened left hemisphere activity has been found in obsessive-compul-sive disorder (OCD) patients (Baxter et al 1987 Swedo et al 1989)GAD patients (Buchsbaum et al 1985 Wu et al 1991) and studentsclassireg ed as ` worriersrsquo rsquo on the basis of self-reports when asked to ` worryabout a specireg c topic of personal concernrsquo rsquo (Carter Johnson amp Borkovec1986)
In summary some reg ndings for anxiety indicate increased right hemi-sphere activation whereas others indicate increased left hemisphere acti-vation Indeed it may be the case that they could co-occur which could
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 429
account for the reg ndings of bilateral increases or no asymmetries In themeantime we are lacking a model to explain these discrepancies
ACCOUNTING FOR INCONSISTENCIES INBRAIN ACTIVITY
We are thus confronted with inconsistent results with regard to regionalbrain activity in both depression and anxiety In previous work we haveargued that in order to resolve these discrepancies it is helpful to turn tothe psychological literature on symptomatology in depression and anxietyIn the sections that follow we make some general observations that applyto both depression and anxiety before focusing on each more exclusively
In addressing the inconsistencies for the blood macr ow reg ndings severalconcerns about the literature are worthy of note First several of the mostfrequently cited studies in this literature sampled a restricted set of brainregions (eg Reiman et al 1984 Reivich et al 1983) Furthermore theregions of interest were not consistent from one study to the nextSecond statistical procedures employing multiple t-tests rather thanrepeated-measures ANOVAs are questionable especially with regard todrawing conclusions about hemispheric differences Third although somestudies examined selected diagnostic groups (eg major depressive dis-order OCD panic disorder GAD) most treated depression or anxiety asa unitary construct and did not consider that different types of each mightbe associated with different patterns of brain activity Moreover with theexception of recent OCD studies most did not take into account thedegree of comorbidity with depression or anxiety which may be criticalin determining the patterns of brain activity that occur (Heller et al1995 Tucker 1988)
As emphasised by Davidson Tomarken and colleagues (Davidson1993 Tomarken Davidson Wheeler amp Doss 1992 Tomarken David-son Wheeler amp Kinney 1992) there are a number of methodogicalapproaches that need to be considered when asymmetric patterns ofEEG activity do not emerge Some of these considerations are relevantto the design of blood macr ow studies as well One issue is that of stability(ie the extent to which brain activity remacr ects a stable individual pattern)This can be addressed by obtaining a sufreg cient number of baseline samplesas well as test-retest data For example Tomarken et al (1992) found thatonly participants with a stable pattern of anterior asymmetry across twomeasurement occasions showed an association between left frontalhypoactivation and more dispositional negative affect as measured bythe PANAS
It is also important to note that for the most part EEG studies haveexamined resting or baseline activity Davidson (1993) has pointed out
430 HELLER AND NITSCHKE
that baseline levels of frontal activation asymmetry may remacr ect individualtendencies toward a valenced affective response but this relationship maynot be observed in the absence of a specireg c environmental elicitor Thushe emphasises the need to examine the relationship between individualdifferences in patterns of asymmetry and affective response to emotionallychallenging tasks As applied to the pattern of posterior reg ndings theseconsiderations suggest the possibility that more robust differences inposterior brain activity might be yielded by studies measuring brainactivity while participants perform tasks for which depressed peopleshow decrements Additional methodological approaches highlighted byDavidson and colleagues such as the use of extreme groups and classifica-tion of participants according to the affect measures as opposed to the brainasymmetry measures would also be useful in future studies of brainactivity
A number of other possible explanations for the failure to reg nd con-sistent decreases in parieto-temporal activity during depression have beenproposed (Heller 1990 1993b) One of these explanations focused on theseverity of the depression This supposition was based on the observationthat posterior asymmetries were more likely to be reported when depres-sion was relatively severe Heller suggested that under less extremeaffective conditions task-induced activity predominates masking theaffect-induced activity More recently Heller et al (1995) suggestedthat severity per se is not the issue rather because the comorbiditywith anxiety decreases as severity of depression increases (Hiller Zau-dig amp Rose 1989) the opposing effects of anxiety on posterior brainactivity diminishes
Heller (1993b) also suggested that the inconsistencies in the depressionand anxiety literatures may be due to signireg cant differences in regionalbrain activity that covary more strongly with dispositional characteristicsassociated with emotional functioning (eg extraversion) than with situa-tionally induced emotions Several studies have reported perceptual(Berenbaum amp Williams 1994 Heller amp Nisenson 1993) and EEG(Nitschke et al 1994) asymmetries for extraversion Nitschke et al alsofound that extraversion moderated the relationship between anxiety andEEG activity across both frontal central and parietal regions For anxiousparticipants high extraversion was associated with less activity (morealpha) whereas for nonanxious (and nondepressed) controls high extra-version was associated with more activity (less alpha) These two sets ofreg ndings albeit preliminary provide evidence for the need to acknowledgeregional differences in brain function for personality characteristics suchas extraversion (for a review see Nitschke Heller amp Miller in prep) andpoint to the possible signireg cance of personality in understanding regionalbrain activity in affective and anxiety disorders
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 431
Subtypes of Depression and Anxiety
Another promising avenue for elucidating the discrepancies reported in theliterature for both depression and anxiety is the analysis of subtypes
Depression Many distinctions for depression have been proposed overthe years including most notably bipolar versus unipolar endogenousversus exogenous and neurotic versus reactive The current Diagnostic andstatistical manual of mental disordersplusmn fourth edition (DSMIV APA 1994)subdivides unipolar depression into major depressive disorder and dysthy-mic disorder and provides further specireg city with criteria for severity andfor psychotic melancholic catatonic or atypical features
Perhaps the most useful distinction in physiological research on depres-sion has been the comparison of melancholic to nonmelancholic depres-sion The dereg ning feature of melancholic depression is anhedonia which isthe inability to experience pleasure Current consensus as indicated by theDSMIV (APA 1994) suggests that melancholia is most strongly charac-terised by a loss of pleasure in all activities and a lack of responsivity topleasurable stimuli Many other symptoms may be shared by both melan-cholic and nonmelancholic depression including depressed mood distur-bances in eating and sleeping patterns low self-esteem inability toconcentrate psychomotor retardation fatigue indecisiveness feelings ofguilt and thoughts about death There is current controversy over whethermelancholic depression is qualitatively different from nonmelancholicdepression with some authors arguing that the primary distinction is oneof severity (Zimmerman Coryell amp Pfohl 1986) Nonetheless neuropsy-chological and neurophysiological studies have been consistent in demon-strating differences between melancholic and nonmelancholic depression
In a comprehensive series of studies Bruder has shown that perceptualasymmetries differ for different subtypes of depression (for a review seeBruder 1995) On a perceptual asymmetry task melancholic patientsshowed an abnormally large right ear (left hemisphere) advantage for adichotic syllables task but no left ear (right hemisphere) advantage for acomplex tone task (Bruder 1995) The lack of a left ear advantage was dueto poor left ear accuracy consistent with other studies demonstrating righthemisphere dereg cits in depressed people However nonmelancholicdepressed patients did not differ from normal controls on either test Alongthese lines in a PET study of melancholic depression Bench et al (1992)reported less activity in the left dorsolateral prefrontal cortex compared tonondepressed controls Note that these data could be consistent with eithera qualitative or a quantitative distinction between melancholic and non-melancholic depression However other data may be more consistent witha qualitative difference between the two types of depression For example
432 HELLER AND NITSCHKE
Davidson Larson and Abercrombie (1995) reported that depression sever-ity in melancholic subjects was predicted by the extent to which activationwas decreased in the left lateral prefrontal region In contrast amongnonmelancholics depression severity was predicted by the extent to whichactivation was increased in the right medial prefrontal area
Davidson et alrsquo s (1995) data have implications for theories regardingthe relationship between pleasantunpleasant affect and anterior regions ofthe brain Research on the psychological characteristics of depression andanxiety has suggested that lack of positive affect (or more preciselyactivated pleasant affect Larsen amp Diener 1992) is unique to depression(Clark amp Watson 1991) Indeed in designing a questionnaire to measureboth general and specireg c components of depression and anxiety Watsonand colleagues (Watson 1995ab) developed a specireg c depression scalethat primarily measures anhedonia Because anhedonia is a fundamentalcharacteristic of melancholia it seems reasonable to assume that decreasedpleasant affect would accompany this type of depression If so Hellerrsquo smodel which predicts less left than right anterior activity for decreasedpleasant affect is completely consistent with the available data
However Davidson et alrsquo s (1995) data suggests the possibility thataffective valence may depend less on the relative balance of activationbetween the left and right anterior regions and more on specireg c systemsassociated with each hemisphere Whereas melancholic depression ischaracterised by decreased pleasant affect depression without melan-cholic features may be better characterised by increased unpleasantaffect Following this line of reasoning Hellerrsquo s model which in itscurrent form argues for a link between anterior asymmetry and the valencedimension would not distinguish between melancholic and nonmelan-cholic depression If affective valence depends more on specireg c systemsassociated with each hemisphere as Davidson et alrsquo s data suggest Hellerrsquo smodel may need to be expanded to remacr ect the unique contributions of theleft and right hemispheres to pleasantunpleasant affect For example itmight be more accurate to link the left anterior region with the activatedpleasant affect (ie positive affect) octant of the circumplex model ofemotion and the right anterior region with the activated unpleasant affect(ie negative affect) octant (see Larson amp Diener 1992)
Because few studies have specireg cally examined the full range of poten-tial affective descriptors in relation to patterns of brain activity theseissues remain to be resolved In most research only isolated portions ofthe circumplex have been sampled For example because positive andnegative affect as measured by the PANAS provide information regardingonly two of the four quadrants in the circumplex model studies that do notcollect data regarding the other two quadrants cannot shed light on howwell they might have accounted for patterns of brain activity
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 433
Anxiety As reviewed earlier studies that examine regional brainactivity in anxiety are even more plagued by inconsistencies than studiesof depression In keeping with the theme of this article we have argued thatby decomposing anxiety into subtypes on the basis of the psychologicalliterature we can potentially resolve some of these discrepancies
In previous work we have suggested that it may be important toconsider the type of anxiety that is being examined (Heller et al 19951997a) In particular we argued that regional brain activity might bedifferent for anxious arousal (eg panic) and anxious apprehension (egworry) types of anxiety that appear to differ signireg cantly in psychologicaland physiological characteristics (eg Barlow 1988 1991 Eysenck 1957Klein 1987 Watson et al 1995b) Anxious arousal has been described byWatson et al as being distinguished by symptoms of physiological hyper-arousal and somatic tension Panic attacks and high-stress states would fallinto this category In contrast anxious apprehension involves worry and ischaracterised by verbal rumination typically about future events This typeof anxiety would be characteristic of obsessive-compulsiveness general-ised anxiety states and trait anxiety (as identireg ed by self-reports of anxiousapprehension and worry on various questionnaires) As noted by Heller etal (1997a) the two kinds of anxiety are not mutually exclusive Extremedegrees of anxious apprehension (eg being worried or fearful about thefuture) may prompt experiences of anxious arousal (eg somatic symptomsand exaggerated physiological responses to stressful events)
We have argued that when the literature on brain activity in anxiety isreviewed in the light of this distinction between anxious arousal andanxious apprehension most of the studies reporting greater right hemi-sphere activity have either studied panic attacks (eg Reiman et al 1984)or high-stress situations likely to entail anxious arousal (eg Tucker et al1977) On the other hand studies indicating greater left hemisphere activity(primarily anterior) have examined OCD (eg Baxter et al 1987) GAD(Wu et al 1991) or questionnaire-dereg ned trait anxiety indexing anxiousapprehension (eg Tucker et al 1978) Indeed as mentioned earlierstudies have highlighted various regions in both hemispheres As aresult it is difreg cult to pinpoint a particular region that is clearly a focalpoint in anxiety Nonetheless the data do suggest that right hemisphereregions emerge as particularly salient in some types of anxiety (ie anxiousarousal) whereas left hemisphere regions emerge as particularly salient inother types of anxiety (ie anxious apprehension)
In a direct test of this hypothesis (Heller et al 1997a) we selectedparticipants on the basis of self-reported trait anxiety using a measureindexing anxious apprehension in a between-subjects design We thenmanipulated anxious arousal on a within-subjects basis Specireg cally wecontrasted brain activity (EEG alpha) during rest periods to brain activity
434 HELLER AND NITSCHKE
during a task in which participants listened to emotional narrativesdesigned to elicit anxious arousal (ie fear and sad narratives that hadpreviously received unpleasant valence and high arousal ratings)
For frontal activity group differences were unaffected by the narrativesAnxious participants showed a larger asymmetry in favour of the lefthemisphere than controls during the rest and listen periods of all narrativetypes as well as for eight minutes of resting baseline preceding thenarrative task Examination of the data indicated that the increased asym-metry was primarily due to a decrease in right hemisphere activity for theanxious participants Nonetheless the increase in magnitude of the asym-metry favouring the left hemisphere found for anxious participants isconsistent with the studies reviewed earlier that reported or inferredrelatively greater left than right anterior activity in anxious subjectsclassireg ed on the basis of worry or trait anxiety (eg Carter et al 1986Tucker et al 1978 Tyler amp Tucker 1982) Our frontal data suggest that itwill be important to investigate whether anxious apprehension is bestcharacterised by the magnitude of the asymmetry favouring the left hemi-sphere or by the level of left hemisphere activity in isolation
In contrast the parietal data supported our prediction regarding theassociation of anxious arousal with greater right hemisphere activityWhen listening to the fear and sad narratives anxious participants showeda selective increase in right parietal activity that was not demonstrated bycontrols The right parietal effect however was not obtained for neutraland happy narratives that had previously received pleasant valence and lowarousal ratings These reg ndings are consistent with those studies that havereported or inferred greater right hemisphere activity during anxious statescharacterised by panic and stress-inducing conditions (eg Reiman et al1984 Stewart et al 1988 Tucker et al 1977)
An important question here has to do with the discrepancies in frontalasymmetries that have been reported Like others (eg Baxter et al 1987Gur et al 1988 Wu et al 1991) we did not reg nd heightened right anterioractivity in anxiety However a number of other studies have (eg Davidsonet al 1996 Reivich et al 1983 Tomarken amp Davidson 1994) Asdiscussed earlier we have previously argued that anterior regions areuniquely involved in the valence dimension of emotion with pleasantaffect associated with greater left activity and unpleasant affect associatedwith greater right activity (eg Heller 1990) Therefore we would expectheightened right anterior activity in an anxious group only if they differedfrom controls on measures indexing affective valence In our recent studythe nondepressed anxious group did not differ from controls on either theGBI Dysthymia scale or the PANAS PA and NA scales
It may also be important to measure the level of anxious apprehension inthe sample studied For example Tomarken and Davidson (1994) found
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 435
that with Marloweplusmn Crowne scores controlled trait anxiety failed to predictfrontal asymmetry However their groups were divided according to amedian split
In summary our results directly support the proposed distinctionbetween anxious apprehension and anxious arousal as types of anxietythat may account for signireg cant variance in patterns of brain activityThey suggest furthermore that Hellerrsquo s model should be rereg ned to reflectthe likelihood that parietal activation will covary with anxious arousal butnot with anxious apprehension If these distinctions are accurate we wouldexpect to observe increases in right parietotemporal activity only underconditions when anxiety is accompanied by increased arousal
Thus in homodynamic and EEG studies a failure to reg nd increased rightposterior activity in anxiety is not surprising as variations in anxiousarousal are likely to be large Some individuals may experience anxiousarousal frequently especially those with disorders more characteristic ofanxious arousal such as panic disorder phobias or post-traumatic stressdisorder Others although apprehensive may rarely experience anxiousarousal In addition the situation may vary in the degree to which it elicitsanxious arousal Being forced to give a speech is likely to induce a greatdeal of anxious arousal in social phobics thus we would have predicted asignireg cant increase in right parietal activity precisely as found by David-son et al (submitted) Note that this effect was only found during theAnticipation condition not during baseline Planning or after thespeech During the Anticipation condition subjects listened while a voiceannounced every 30 seconds for three minutes the time remaining beforethey would be given a topic and allowed to plan for their speech Pre-sumably this period is the most stressful and the most likely to result inanxious arousal Indeed the correlation between heart rate acceleration tothe Anticipation condition and the increase in state anxiety was signireg cant
More recently Nitschke Heller Imig and Miller (submitted) haveattempted to identify anxious arousal and anxious apprehension usingquestionnaires Data from 783 students for a battery of anxiety and depres-sion questionnaires provided strong support for the anxiety distinctionproposed here at least on a psychological level of analysis The Moodand Anxiety Symptom Questionnaire-Anxious Arousal scale (MASQWatson et al 1995ab) was chosen as a measure of anxious arousalwhereas the Penn State Worry Questionnaire (PSWQ Meyer MillerMetzger amp Borkovec 1990 Molina amp Borkovec 1994) was included asa measure of anxious apprehension The correlation between the two scales(r = 20) was much lower than those involving the other more generalscales of anxiety included in the study (MASQ-General Anxiety scaleBeck Anxiety Inventory and the State-Trait Anxiety Inventory-Traitscale ranging from r = 47 to r = 65) These results further support the
436 HELLER AND NITSCHKE
argument that anxious arousal and anxious apprehension are distinctanxiety states
However new questions are highlighted by this distinction betweenanxious arousal and anxious apprehension particularly in the light ofHellerrsquo s model If anxious apprehension is associated with left anterioractivity the overall pattern of regional brain activity appears to be identicalto the brain activity associated with pleasant emotional states (ie more leftthan right anterior activity) However anxious apprehension might beexpected to fall in the portion of the circumplex model of emotionassociated with activated unpleasant affect (as suggested by adjectivessuch as fearful nervous jittery anxious) Below we suggest a numberof theoretical possibilities for resolving this apparent discrepancy all ofwhich remain to be examined empirically
In general it will be important to clarify the degree to which anxiousapprehension is in fact characterised by unpleasant versus pleasant valenceas well as high versus low arousal In addition it may be helpful toconsider these anterior asymmetries from the approach-withdrawal frame-work employed by Davidson (1992ab Davidson amp Tomarken 1989)Under conditions where anxious apprehension leads to approach tenden-cies one might expect an asymmetry in favour of the left anterior region Incontrast when anxious apprehension is associated with withdrawal tenden-cies (perhaps as anxious arousal increases) an asymmetry may emerge infavour of the right anterior region Conceivably the degree to whichanxious apprehension is characterised by approach versus withdrawaltendencies would depend on the degree of stress and discomfortinvolved the personrsquo s coping style and other factors With regard tovalence we would predict that anxious apprehension characterised byapproach tendencies would be accompanied by relatively pleasant affectAnxious apprehension accompanied by withdrawal tendencies in contrastwould be accompanied by relatively unpleasant affect
It is also possible that instead of a direct link between anxious appre-hension and approach behaviour they share more fundamental character-istics associated with left anterior functions For example one might arguethat they both involve sequential thinking and the development of hier-archically structured plans focusing on future concerns (for a review ofrelevant left anterior functions see Tomarken amp Keener this Issue)
Another important issue has to do with the nature of the asymmetricpattern of activity As reviewed earlier we found that the asymmetry infavour of the left hemisphere for participants classireg ed on the basis of traitanxiety was due to a relative decrease in right hemisphere activity Thusthe left frontal region of the anxious participants was not signireg cantly moreactive than that of controls It will therefore be important to examine thepossibility that absolute activity of the left anterior region plays a more
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 437
important role in the modulation of valence (eg Davidson et al 1995)than of anxious apprehension
Another possibility is that independent left anterior systems are asso-ciated with anxious apprehension versus pleasant affect Liotti and Tucker(1995) review theories that there may be multiple functionally segregatedparallel frontal cortico-striato-thalamic circuits some associated withmotor functions some with set shifting and behavioural inhibition andsome with emotional information processing and hedonic responsivity tostimuli Possibly one of these circuits remacr ects activation associated withpleasant affect and another remacr ects activation associated with anxiousapprehension andor approach Different neurotransmitter systems couldbe involved More precise methods of localisation would therefore benecessary to distinguish the unique brain activity associated with each ofthese psychologicalbehavioural phenomena
Comorbidity of Depression and Anxiety
In considering the issue of decomposing emotional states with regard to thesymptomatology in affective and anxiety disorders we have previouslyargued that issues of comorbidity of depression and anxiety are crucial(Heller et al 1995) A review of the literature reveals that rates of comor-bidity are extremely high (Alloy Kelly Mineka amp Clements 1990 Helleret al 1995 Hiller et al 1989 Katon amp Roy-Byrne 1991) Even in our veryrecent work (Nitschke et al submitted) using the MASQ a measuredesigned by Watson and colleagues to avoid symptom overlap in themeasurement of anxiety and depression (Watson et al 1995ab) comor-bidity was fairly pronounced in a large college student sample The MASQAnxious Arousal (AA) and Anhedonic Depression (AD) scales were used asmeasures of those constructs and the PSWQ provided the measure ofanxious apprehension Of the 125 (out of 783) participants scoring abovethe 80th percentile on at least one of the three scales of interest and belowthe 50th percentile on any where the 80th percentile cut-off was not metfully 50 demonstrated comorbidity by scoring above the 80th percentileon at least two of the scales This rate of comorbidity emerged despite thelow correlations found between the three scales (MASQ AA and MASQAD r = 20 MASQ AA and PSWQ r = 20 MASQ AD and PSWQ r =36) relative to the correlations found for the other more general measuresof depression including the MASQ General Mixed General Depression andGeneral Anxiety scales Beck Depression Inventory Beck Anxiety Inven-tory General Behaviour Inventory-Dysthymia scale and the State-TraitAnxiety Inventory-Trait scale (ranging from r = 47 to r = 76)
Because at least some studies have found opposite patterns of activityfor the right posterior region during depression versus during anxiety (ie
438 HELLER AND NITSCHKE
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
At the time of its conception Hellerrsquo s model attempted to address anumber of conmacr icts in the literature Which hemisphere was involved inemotion and in what manner was a matter of much debate Based onevidence that the right hemisphere was specialised to process emotionalinformation some researchers were arguing that all emotional functionswere dependent on the right hemisphere Representative of this perspec-tive Levy et al (1983) had obtained behavioural evidence on tasksassociated with parieto-temporal functions that led us to conclude thatindividuals displaying relatively reduced right hemisphere activationwere predisposed to unpleasant emotions whereas individuals displayingenhanced right hemisphere activation were predisposed to pleasant emo-tions Based on these reg ndings we posited that reduced activation of theright hemisphere was associated with depression This hypothesis wassupported by evidence that depressed people display dereg cits in cognitivefunctions associated with the right hemisphere on neuropsychological testsand in lateralised tachistoscopic paradigms (for reviews see Heller 19901993ab Tucker 1981)
In direct contrast EEG studies indicated that compared to the lefthemisphere right hemisphere activation was increased in depressedmood states particularly for anterior regions (for a review see David-son 1984) When mood was euthymic however activation of the righthemisphere was decreased relative to the left hemisphere These resultsbuilt on a set of observations systematically described by Gainotti(1972) noting very different emotional behaviours in people with rightversus left brain damage Left hemisphere damage was associated withthe so-called ` catastrophic reactionrsquorsquo in which patients were described tobe emotionally volatile and especially prone to depression and crying Incontrast right hemisphere damage was observed to be frequently accom-panied by a ` euphoricrsquo rsquo or ` indifferencersquo rsquo reaction Similar emotionalresponses have been described in patients undergoing the WADA orsodium amytal test Injections of sodium amytal that deactivate the lefthemisphere are associated with the ` catastrophicrsquo rsquo reaction whereasthose that deactivate the right hemisphere are associated with the` indifferenceeuphoriarsquo rsquo reaction (eg Lee Loring Meador Flanigin ampBrooks 1988)
The EEG and lesion studies inspired what have been referred to as` valencersquo rsquo theories (for a review see Borod 1993) In one variant ofthis type of theory the left hemisphere was thought to be specialised forpleasant affect and the right hemisphere was thought to be specialisedfor unpleasant affect (eg Sackeim et al 1982) To account for the lesiondata it was assumed that brain damage immobilises or diminishes thefunctioning of the damaged hemisphere and that the behaviour observedremacr ects the functioning of the intact hemisphere more or less on its own
424 HELLER AND NITSCHKE
In another variant of a valence theory Tucker (1981) concluded thatunpleasant and pleasant emotions were associated with the left and righthemispheres respectively a reversal of the valence theory described ear-lier Here the lesion data were explained by arguing that the observedbehaviour was not a function of the intact hemisphere but rather remacr ectedemotional processes of the subcortical areas on the damaged side that werereleased from cortical inhibition because of the lesion The EEG datawould be explained by assuming that increased anterior activity on oneside or the other would inhibit the affect associated with the correspondingipsilateral subcortical region (unpleasant for the left pleasant for the right)By this reasoning it follows that the affective effects of a lateralisedsubcortical lesion should be opposite to those of a cortical lesion (eg alesion to left subcortical structures should be associated with a suppressionof activity in this area leading to a decrease in unpleasant affect) How-ever several studies have found left anterior subcortical lesions to beassociated with depression (for a review see Liotti amp Tucker 1995)
In attempting to account for the co-occurrence in depression of increasedactivation in right anterior regions and dereg cits on cognitive tasks specia-lised to the right hemisphere Tucker (1981) argued that frontal regions havebeen traditionally viewed as having inhibitory regulatory effects on poster-ior regions By distinguishing valence from arousal as components ofemotion and localising these dimensions to anterior and posterior regionsof the brain respectively Hellerrsquo s (1986 1990) model also emphasised acaudal distinction However the model has additional explanatory powerAlthough inhibitory mechanisms of the anterior regions may well be play-ing a role in affective regulation (also see Tomarken amp Keener this Issue)the concept of a (relatively) independent arousal dimension associated withthe posterior right hemisphere allows us to model the patterns of brainactivity associated with all the emotions as represented on the circum-plex Indeed the patterns of activity that have been reported for depressionand anxiety cannot be fully accounted for by invoking an inhibitorymechanism For example if the inhibitory function of the right anteriorregion is the driving force behind right posterior activity in differentaffective states we would expect anxiety (which like depression is oftenaccompanied by increased right anterior activity) to be similarly accompa-nied by decreased right posterior activity However the evidence suggeststhat anxiety is often associated with increased activity in this region (asreviewed later) Thus by positing an independent arousal dimensionHellerrsquo s model provides an explanation for the patterns of brain activitythat have been observed for anxiety as well as depression
Tucker and colleagues have furthermore emphasised the important roleof cortical-subcortical interactions in producing the asymmetries in corticalactivity Hellerrsquo s model does not address the mechanisms by which the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 425
asymmetries associated with affective valence are generated indeed itcould well be the case that the relationship between activity in the anteriorregions and valence is mediated largely via interaction with subcorticalstructures Furthermore inhibitory mechanisms may be important in thisregard
Note that the approach we are emphasising in this article is compatiblewith Davidson and colleaguesrsquo attempts to decompose emotional statesaccording to the degree that they involve approach versus withdrawalbehaviours (Davidson 1992 Davidson amp Tomarken 1989 Tomarken ampKeener this Issue see also Kinsbourne 1988) These authors have arguedthat approach and withdrawal are the fundamental behavioural substratesassociated with affective valence However this theoretical formulationclearly emphasises the anterior cortical asymmetries associated with plea-santunpleasant valence Although Davidson has addressed posterior asym-metries in his work (eg Davidson Schaffer amp Saron 1985 Davidson1992a) the approach-withdrawal model does not focus on the patterns ofposterior brain activity that have been documented for depression andanxiety Thus considering the arousal dimension in addition to theapproach-withdrawal distinction may address the data for the posteriorregions more comprehensively
The present article has several goals We will highlight persistent andemerging discrepancies in the reg eld of brain function in depression andanxiety and suggest that turning to the psychological literature on thespecireg c symptomatology and comorbidity of these disorders can help usresolve at least some of them In the process we will revisit Hellerrsquo s modelof emotion with a critical look at how well it has been able to account forthe wealth of recent reg ndings in the reg eld Some modireg cations to this modelwill be proposed and continuing questions will be explicated
REGIONAL BRAIN ACTIVITY
Anterior and posterior asymmetries in cortical activity are often but notalways reported in depression and anxiety As a result it has been difreg cultto identify precisely the patterns of brain activity that are important In thesections following we highlight the discrepancies in the reg ndings foranterior and posterior asymmetries We then discuss possible causes forthese discrepancies and suggest ways to disentangle the results
Depression
Anterior asymmetries for EEG alpha have often been found in studies ofdepression (eg Allen Iacono Depue amp Arbisi 1993 Henriques ampDavidson 1990 1991 Schaffer Davidson amp Saron 1983) Across these
426 HELLER AND NITSCHKE
studies depression has been associated with more right than left anterioractivity Numerous positron emission tomography (PET) studies have alsoreported less left anterior activity (eg Baxter et al 1985 1989 Bench etal 1992 Bench Friston Brown Frackowiak amp Dolan 1993 George etal 1994 Martinot et al 1990) which is consistent with the EEG reg ndingsearlier However other studies have reported reg ndings in the oppositedirection for depression (Drevets et al 1992) and sad mood (George etal 1995) Null effects for depression have also been found (eg NitschkeHeller Etienne amp Miller 1995 Tomarken amp Davidson 1994)
The reg ndings for the parieto-temporal regons have been even morevariable In EEG and blood macr ow studies that report positive reg ndingsdepression is typically associated with less right posterior activity (egFlor-Henry 1979 Henriques amp Davidson 1990 Post et al 1987 Uytden-hoef et al 1983) Numerous studies however report no effects for thisregion of the brain (for a review see Davidson amp Tomarken 1989)
Reports of reduced right parieto-temporal activity although not alwayspresent are consistent with neuropsychological studies often indicatingimpaired or suppressed right hemisphere function in depressed peopleOn neuropsychological tasks thought to depend on right parietotemporalfunctioning depressed people have often been found to display decrementsin performance (eg Berndt amp Berndt 1980 Flor-Henry 1976 GoldsteinFilskov Weaver amp Ives 1977 Gruzelier Seymour Wilson Jolley ampHirsch 1988 Kronfol Hamsher Digre amp Waziri 1978) Similar findingsfor sad mood in nondepressed people have also been reported for righthemisphere tasks (Tucker Stenslie Roth amp Shearer 1981)
Studies that have directed information to one or the other hemisphereusing lateralised paradigms such as dichotic listening or visual half-reg eldpresentation have also found specireg c dereg cits for the right hemisphere indepressed people (for a review see Bruder 1995) Left-ear decrements orabsent left-ear (right hemisphere) advantages have been reported fordepression on nonverbal dichotic listening tasks (eg Bruder et al1989) Similarly left visual reg eld (right hemisphere) dereg cits in reactiontime and accuracy for nonverbal stimuli have been reported for depressedpeople (Bruder et al 1989 Liotti Sava Rizzolati amp Caffarra 1991)Comparable results have also been obtained in studies of nondepressedpeople who undergo mood inductions (Banich Stolar Heller amp Goldman1992 Ladavas Nicoletti Umilta amp Rizzolatti 1984) Reaction time andaccuracy for the right hemisphere are selectively impaired in sad moods
On the Chimeric Faces Task (CFT) a free-vision task of face processingthat typically elicits a left hemispatial bias suggesting greater right hemi-sphere activation Jaeger Borod and Peselow (1987) found smaller lefthemispatial biases for depressed patients than nonpsychiatric controlsUsing the same task Heller Etienne and Miller (1995) reported that
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 427
high-depressed participants had smaller left hemispatial biases than low-depressed participants Due to the nature of the CFT hemispatial biases arenonspecireg c as to whether they remacr ect a decrease in activity for the con-tralateral hemisphere or an increase in activity for the ipsilateral hemi-sphere (ie a reduction in the magnitude of the left hemispatial bias couldremacr ect either reduced right hemisphere or increased left hemisphere activ-ity) However based on the evidence (reviewed earlier) for selectivedereg cits in performance on right hemisphere tasks relatively poor perfor-mance for the left-visual reg eld on tachistoscopic tasks and reduced EEGactivity for right posterior regions the most likely interpretation of the CFTreg ndings is that they remacr ect diminished right posterior activity
In summary although anterior asymmetries have often been reported indepression the more common reg nding of higher left than right activation isnot invariably found Furthermore there is a signireg cant discrepancybetween the number of studies that have reported parieto-temporal asym-metries in activation which have been relatively few and the number ofstudies that have reported dereg cits in right hemisphere information proces-sing which have been many Also worthy of note is whether studies ofclinical depression psychometrically dereg ned depression and sad mood areaddressing the same phenomenon The number of consistencies acrossthese populations in the physiological literature reviewed here suggestssome commonality on the other hand the inconsistencies could be in partdue to the heterogeneity between the samples studied
Anxiety
Research on brain activity in anxious states has also been plagued byinconsistencies Again although anterior and posterior asymmetries havebeen reported no consistent pattern has emerged Studies of multisite EEGand regional cerebral blood macr ow (rCBF) using xenon-133 or PET havereported right hemisphere left hemisphere and bilateral increases inactivity during anxiety (for a review see Heller Nitschke Etienne ampMiller 1997a) Other studies have found no differences in brain activity(Gur et al 1988 Mountz et al 1989 Nordhal et al 1990 Tomarken ampDavidson 1994)
Having comprehensively reviewed this literature elsewhere (Heller etal 1997a) we will provide an overview of those studies reporting differ-ential hemispheric involvement in anxiety A subset of these suggest anasymmetry in favour of right hemisphere activation in various regions(frontal lateral prefrontal precentral frontal anterior temporal parietaloccipital and parahippocampal) Heightened right hemisphere activity hasbeen reported for panic disorder patients (Reiman Raichle Butler Hers-covitch amp Robins 1984) panic disorder patients experiencing a sodium
428 HELLER AND NITSCHKE
lactate-induced panic attack (Stewart Devous Rush Lane amp Bonte1988) generalised anxiety disorder (GAD) patients during an anxiety-stimulating task (Wu et al 1991) psychometrically dereg ned high-anxiousparticipants immediately following arterial and venous catheterisation andagain after completion of the PET scans (Reivich Gur amp Alavi 1983) andsocial phobics prior to making a public speech (Davidson MarshallTomarken amp Henriques submitted) Consistent with this pattern anti-anxiety medication was found to reduce the rate of glucose metabolismin the right hemisphere especially for the occipital and frontal regions forpatients with GAD (Buchsbaum et al 1987) Similarly in a population ofnonpsychiatric volunteers those administered diazepam intravenouslyshowed maximal decreases in rCBF in the right frontal area whereasthose injected with a placebo showed no rCBF changes (Mathew Wil-son amp Daniel 1985)
Regional brain activity can also be inferred from performance on tasksthat depend on particular brain regions On the CFT Heller et al (1995)reported that high-anxious participants had larger left hemispatial biasesthan low-anxious participants suggested to remacr ect greater right posterioractivity Similarly on a neuropsychological test battery panic disorderpatients performed less well than normal controls on left but not righthemisphere tasks (Yeudall et al 1983) Data obtained from a series ofstudies using behavioural measures to examine right versus left hemispherefunction during an experimental condition producing transient (ie state)anxiety (Tucker Antes Stenslie amp Barnhardt 1978 Tucker Roth Arne-son amp Buckingham 1977 Tyler amp Tucker 1982) are also consistent withan interpretation of greater right than left hemisphere activity in anxiety(Heller et al 1995)
However when subjects were classireg ed according to trait anxietyTucker and colleagues (Tucker et al 1978 Tyler amp Tucker 1982)interpreted the results as strongly suggestive of greater left than righthemisphere activity This interpretation is consistent with reports ofnumerous other labs reg nding increased left hemisphere activity associatedwith anxiety in various cortical and subcortical regions (orbital frontalinferior frontal anterior cingulate caudate putamen and thalamus)Heightened left hemisphere activity has been found in obsessive-compul-sive disorder (OCD) patients (Baxter et al 1987 Swedo et al 1989)GAD patients (Buchsbaum et al 1985 Wu et al 1991) and studentsclassireg ed as ` worriersrsquo rsquo on the basis of self-reports when asked to ` worryabout a specireg c topic of personal concernrsquo rsquo (Carter Johnson amp Borkovec1986)
In summary some reg ndings for anxiety indicate increased right hemi-sphere activation whereas others indicate increased left hemisphere acti-vation Indeed it may be the case that they could co-occur which could
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 429
account for the reg ndings of bilateral increases or no asymmetries In themeantime we are lacking a model to explain these discrepancies
ACCOUNTING FOR INCONSISTENCIES INBRAIN ACTIVITY
We are thus confronted with inconsistent results with regard to regionalbrain activity in both depression and anxiety In previous work we haveargued that in order to resolve these discrepancies it is helpful to turn tothe psychological literature on symptomatology in depression and anxietyIn the sections that follow we make some general observations that applyto both depression and anxiety before focusing on each more exclusively
In addressing the inconsistencies for the blood macr ow reg ndings severalconcerns about the literature are worthy of note First several of the mostfrequently cited studies in this literature sampled a restricted set of brainregions (eg Reiman et al 1984 Reivich et al 1983) Furthermore theregions of interest were not consistent from one study to the nextSecond statistical procedures employing multiple t-tests rather thanrepeated-measures ANOVAs are questionable especially with regard todrawing conclusions about hemispheric differences Third although somestudies examined selected diagnostic groups (eg major depressive dis-order OCD panic disorder GAD) most treated depression or anxiety asa unitary construct and did not consider that different types of each mightbe associated with different patterns of brain activity Moreover with theexception of recent OCD studies most did not take into account thedegree of comorbidity with depression or anxiety which may be criticalin determining the patterns of brain activity that occur (Heller et al1995 Tucker 1988)
As emphasised by Davidson Tomarken and colleagues (Davidson1993 Tomarken Davidson Wheeler amp Doss 1992 Tomarken David-son Wheeler amp Kinney 1992) there are a number of methodogicalapproaches that need to be considered when asymmetric patterns ofEEG activity do not emerge Some of these considerations are relevantto the design of blood macr ow studies as well One issue is that of stability(ie the extent to which brain activity remacr ects a stable individual pattern)This can be addressed by obtaining a sufreg cient number of baseline samplesas well as test-retest data For example Tomarken et al (1992) found thatonly participants with a stable pattern of anterior asymmetry across twomeasurement occasions showed an association between left frontalhypoactivation and more dispositional negative affect as measured bythe PANAS
It is also important to note that for the most part EEG studies haveexamined resting or baseline activity Davidson (1993) has pointed out
430 HELLER AND NITSCHKE
that baseline levels of frontal activation asymmetry may remacr ect individualtendencies toward a valenced affective response but this relationship maynot be observed in the absence of a specireg c environmental elicitor Thushe emphasises the need to examine the relationship between individualdifferences in patterns of asymmetry and affective response to emotionallychallenging tasks As applied to the pattern of posterior reg ndings theseconsiderations suggest the possibility that more robust differences inposterior brain activity might be yielded by studies measuring brainactivity while participants perform tasks for which depressed peopleshow decrements Additional methodological approaches highlighted byDavidson and colleagues such as the use of extreme groups and classifica-tion of participants according to the affect measures as opposed to the brainasymmetry measures would also be useful in future studies of brainactivity
A number of other possible explanations for the failure to reg nd con-sistent decreases in parieto-temporal activity during depression have beenproposed (Heller 1990 1993b) One of these explanations focused on theseverity of the depression This supposition was based on the observationthat posterior asymmetries were more likely to be reported when depres-sion was relatively severe Heller suggested that under less extremeaffective conditions task-induced activity predominates masking theaffect-induced activity More recently Heller et al (1995) suggestedthat severity per se is not the issue rather because the comorbiditywith anxiety decreases as severity of depression increases (Hiller Zau-dig amp Rose 1989) the opposing effects of anxiety on posterior brainactivity diminishes
Heller (1993b) also suggested that the inconsistencies in the depressionand anxiety literatures may be due to signireg cant differences in regionalbrain activity that covary more strongly with dispositional characteristicsassociated with emotional functioning (eg extraversion) than with situa-tionally induced emotions Several studies have reported perceptual(Berenbaum amp Williams 1994 Heller amp Nisenson 1993) and EEG(Nitschke et al 1994) asymmetries for extraversion Nitschke et al alsofound that extraversion moderated the relationship between anxiety andEEG activity across both frontal central and parietal regions For anxiousparticipants high extraversion was associated with less activity (morealpha) whereas for nonanxious (and nondepressed) controls high extra-version was associated with more activity (less alpha) These two sets ofreg ndings albeit preliminary provide evidence for the need to acknowledgeregional differences in brain function for personality characteristics suchas extraversion (for a review see Nitschke Heller amp Miller in prep) andpoint to the possible signireg cance of personality in understanding regionalbrain activity in affective and anxiety disorders
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 431
Subtypes of Depression and Anxiety
Another promising avenue for elucidating the discrepancies reported in theliterature for both depression and anxiety is the analysis of subtypes
Depression Many distinctions for depression have been proposed overthe years including most notably bipolar versus unipolar endogenousversus exogenous and neurotic versus reactive The current Diagnostic andstatistical manual of mental disordersplusmn fourth edition (DSMIV APA 1994)subdivides unipolar depression into major depressive disorder and dysthy-mic disorder and provides further specireg city with criteria for severity andfor psychotic melancholic catatonic or atypical features
Perhaps the most useful distinction in physiological research on depres-sion has been the comparison of melancholic to nonmelancholic depres-sion The dereg ning feature of melancholic depression is anhedonia which isthe inability to experience pleasure Current consensus as indicated by theDSMIV (APA 1994) suggests that melancholia is most strongly charac-terised by a loss of pleasure in all activities and a lack of responsivity topleasurable stimuli Many other symptoms may be shared by both melan-cholic and nonmelancholic depression including depressed mood distur-bances in eating and sleeping patterns low self-esteem inability toconcentrate psychomotor retardation fatigue indecisiveness feelings ofguilt and thoughts about death There is current controversy over whethermelancholic depression is qualitatively different from nonmelancholicdepression with some authors arguing that the primary distinction is oneof severity (Zimmerman Coryell amp Pfohl 1986) Nonetheless neuropsy-chological and neurophysiological studies have been consistent in demon-strating differences between melancholic and nonmelancholic depression
In a comprehensive series of studies Bruder has shown that perceptualasymmetries differ for different subtypes of depression (for a review seeBruder 1995) On a perceptual asymmetry task melancholic patientsshowed an abnormally large right ear (left hemisphere) advantage for adichotic syllables task but no left ear (right hemisphere) advantage for acomplex tone task (Bruder 1995) The lack of a left ear advantage was dueto poor left ear accuracy consistent with other studies demonstrating righthemisphere dereg cits in depressed people However nonmelancholicdepressed patients did not differ from normal controls on either test Alongthese lines in a PET study of melancholic depression Bench et al (1992)reported less activity in the left dorsolateral prefrontal cortex compared tonondepressed controls Note that these data could be consistent with eithera qualitative or a quantitative distinction between melancholic and non-melancholic depression However other data may be more consistent witha qualitative difference between the two types of depression For example
432 HELLER AND NITSCHKE
Davidson Larson and Abercrombie (1995) reported that depression sever-ity in melancholic subjects was predicted by the extent to which activationwas decreased in the left lateral prefrontal region In contrast amongnonmelancholics depression severity was predicted by the extent to whichactivation was increased in the right medial prefrontal area
Davidson et alrsquo s (1995) data have implications for theories regardingthe relationship between pleasantunpleasant affect and anterior regions ofthe brain Research on the psychological characteristics of depression andanxiety has suggested that lack of positive affect (or more preciselyactivated pleasant affect Larsen amp Diener 1992) is unique to depression(Clark amp Watson 1991) Indeed in designing a questionnaire to measureboth general and specireg c components of depression and anxiety Watsonand colleagues (Watson 1995ab) developed a specireg c depression scalethat primarily measures anhedonia Because anhedonia is a fundamentalcharacteristic of melancholia it seems reasonable to assume that decreasedpleasant affect would accompany this type of depression If so Hellerrsquo smodel which predicts less left than right anterior activity for decreasedpleasant affect is completely consistent with the available data
However Davidson et alrsquo s (1995) data suggests the possibility thataffective valence may depend less on the relative balance of activationbetween the left and right anterior regions and more on specireg c systemsassociated with each hemisphere Whereas melancholic depression ischaracterised by decreased pleasant affect depression without melan-cholic features may be better characterised by increased unpleasantaffect Following this line of reasoning Hellerrsquo s model which in itscurrent form argues for a link between anterior asymmetry and the valencedimension would not distinguish between melancholic and nonmelan-cholic depression If affective valence depends more on specireg c systemsassociated with each hemisphere as Davidson et alrsquo s data suggest Hellerrsquo smodel may need to be expanded to remacr ect the unique contributions of theleft and right hemispheres to pleasantunpleasant affect For example itmight be more accurate to link the left anterior region with the activatedpleasant affect (ie positive affect) octant of the circumplex model ofemotion and the right anterior region with the activated unpleasant affect(ie negative affect) octant (see Larson amp Diener 1992)
Because few studies have specireg cally examined the full range of poten-tial affective descriptors in relation to patterns of brain activity theseissues remain to be resolved In most research only isolated portions ofthe circumplex have been sampled For example because positive andnegative affect as measured by the PANAS provide information regardingonly two of the four quadrants in the circumplex model studies that do notcollect data regarding the other two quadrants cannot shed light on howwell they might have accounted for patterns of brain activity
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 433
Anxiety As reviewed earlier studies that examine regional brainactivity in anxiety are even more plagued by inconsistencies than studiesof depression In keeping with the theme of this article we have argued thatby decomposing anxiety into subtypes on the basis of the psychologicalliterature we can potentially resolve some of these discrepancies
In previous work we have suggested that it may be important toconsider the type of anxiety that is being examined (Heller et al 19951997a) In particular we argued that regional brain activity might bedifferent for anxious arousal (eg panic) and anxious apprehension (egworry) types of anxiety that appear to differ signireg cantly in psychologicaland physiological characteristics (eg Barlow 1988 1991 Eysenck 1957Klein 1987 Watson et al 1995b) Anxious arousal has been described byWatson et al as being distinguished by symptoms of physiological hyper-arousal and somatic tension Panic attacks and high-stress states would fallinto this category In contrast anxious apprehension involves worry and ischaracterised by verbal rumination typically about future events This typeof anxiety would be characteristic of obsessive-compulsiveness general-ised anxiety states and trait anxiety (as identireg ed by self-reports of anxiousapprehension and worry on various questionnaires) As noted by Heller etal (1997a) the two kinds of anxiety are not mutually exclusive Extremedegrees of anxious apprehension (eg being worried or fearful about thefuture) may prompt experiences of anxious arousal (eg somatic symptomsand exaggerated physiological responses to stressful events)
We have argued that when the literature on brain activity in anxiety isreviewed in the light of this distinction between anxious arousal andanxious apprehension most of the studies reporting greater right hemi-sphere activity have either studied panic attacks (eg Reiman et al 1984)or high-stress situations likely to entail anxious arousal (eg Tucker et al1977) On the other hand studies indicating greater left hemisphere activity(primarily anterior) have examined OCD (eg Baxter et al 1987) GAD(Wu et al 1991) or questionnaire-dereg ned trait anxiety indexing anxiousapprehension (eg Tucker et al 1978) Indeed as mentioned earlierstudies have highlighted various regions in both hemispheres As aresult it is difreg cult to pinpoint a particular region that is clearly a focalpoint in anxiety Nonetheless the data do suggest that right hemisphereregions emerge as particularly salient in some types of anxiety (ie anxiousarousal) whereas left hemisphere regions emerge as particularly salient inother types of anxiety (ie anxious apprehension)
In a direct test of this hypothesis (Heller et al 1997a) we selectedparticipants on the basis of self-reported trait anxiety using a measureindexing anxious apprehension in a between-subjects design We thenmanipulated anxious arousal on a within-subjects basis Specireg cally wecontrasted brain activity (EEG alpha) during rest periods to brain activity
434 HELLER AND NITSCHKE
during a task in which participants listened to emotional narrativesdesigned to elicit anxious arousal (ie fear and sad narratives that hadpreviously received unpleasant valence and high arousal ratings)
For frontal activity group differences were unaffected by the narrativesAnxious participants showed a larger asymmetry in favour of the lefthemisphere than controls during the rest and listen periods of all narrativetypes as well as for eight minutes of resting baseline preceding thenarrative task Examination of the data indicated that the increased asym-metry was primarily due to a decrease in right hemisphere activity for theanxious participants Nonetheless the increase in magnitude of the asym-metry favouring the left hemisphere found for anxious participants isconsistent with the studies reviewed earlier that reported or inferredrelatively greater left than right anterior activity in anxious subjectsclassireg ed on the basis of worry or trait anxiety (eg Carter et al 1986Tucker et al 1978 Tyler amp Tucker 1982) Our frontal data suggest that itwill be important to investigate whether anxious apprehension is bestcharacterised by the magnitude of the asymmetry favouring the left hemi-sphere or by the level of left hemisphere activity in isolation
In contrast the parietal data supported our prediction regarding theassociation of anxious arousal with greater right hemisphere activityWhen listening to the fear and sad narratives anxious participants showeda selective increase in right parietal activity that was not demonstrated bycontrols The right parietal effect however was not obtained for neutraland happy narratives that had previously received pleasant valence and lowarousal ratings These reg ndings are consistent with those studies that havereported or inferred greater right hemisphere activity during anxious statescharacterised by panic and stress-inducing conditions (eg Reiman et al1984 Stewart et al 1988 Tucker et al 1977)
An important question here has to do with the discrepancies in frontalasymmetries that have been reported Like others (eg Baxter et al 1987Gur et al 1988 Wu et al 1991) we did not reg nd heightened right anterioractivity in anxiety However a number of other studies have (eg Davidsonet al 1996 Reivich et al 1983 Tomarken amp Davidson 1994) Asdiscussed earlier we have previously argued that anterior regions areuniquely involved in the valence dimension of emotion with pleasantaffect associated with greater left activity and unpleasant affect associatedwith greater right activity (eg Heller 1990) Therefore we would expectheightened right anterior activity in an anxious group only if they differedfrom controls on measures indexing affective valence In our recent studythe nondepressed anxious group did not differ from controls on either theGBI Dysthymia scale or the PANAS PA and NA scales
It may also be important to measure the level of anxious apprehension inthe sample studied For example Tomarken and Davidson (1994) found
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 435
that with Marloweplusmn Crowne scores controlled trait anxiety failed to predictfrontal asymmetry However their groups were divided according to amedian split
In summary our results directly support the proposed distinctionbetween anxious apprehension and anxious arousal as types of anxietythat may account for signireg cant variance in patterns of brain activityThey suggest furthermore that Hellerrsquo s model should be rereg ned to reflectthe likelihood that parietal activation will covary with anxious arousal butnot with anxious apprehension If these distinctions are accurate we wouldexpect to observe increases in right parietotemporal activity only underconditions when anxiety is accompanied by increased arousal
Thus in homodynamic and EEG studies a failure to reg nd increased rightposterior activity in anxiety is not surprising as variations in anxiousarousal are likely to be large Some individuals may experience anxiousarousal frequently especially those with disorders more characteristic ofanxious arousal such as panic disorder phobias or post-traumatic stressdisorder Others although apprehensive may rarely experience anxiousarousal In addition the situation may vary in the degree to which it elicitsanxious arousal Being forced to give a speech is likely to induce a greatdeal of anxious arousal in social phobics thus we would have predicted asignireg cant increase in right parietal activity precisely as found by David-son et al (submitted) Note that this effect was only found during theAnticipation condition not during baseline Planning or after thespeech During the Anticipation condition subjects listened while a voiceannounced every 30 seconds for three minutes the time remaining beforethey would be given a topic and allowed to plan for their speech Pre-sumably this period is the most stressful and the most likely to result inanxious arousal Indeed the correlation between heart rate acceleration tothe Anticipation condition and the increase in state anxiety was signireg cant
More recently Nitschke Heller Imig and Miller (submitted) haveattempted to identify anxious arousal and anxious apprehension usingquestionnaires Data from 783 students for a battery of anxiety and depres-sion questionnaires provided strong support for the anxiety distinctionproposed here at least on a psychological level of analysis The Moodand Anxiety Symptom Questionnaire-Anxious Arousal scale (MASQWatson et al 1995ab) was chosen as a measure of anxious arousalwhereas the Penn State Worry Questionnaire (PSWQ Meyer MillerMetzger amp Borkovec 1990 Molina amp Borkovec 1994) was included asa measure of anxious apprehension The correlation between the two scales(r = 20) was much lower than those involving the other more generalscales of anxiety included in the study (MASQ-General Anxiety scaleBeck Anxiety Inventory and the State-Trait Anxiety Inventory-Traitscale ranging from r = 47 to r = 65) These results further support the
436 HELLER AND NITSCHKE
argument that anxious arousal and anxious apprehension are distinctanxiety states
However new questions are highlighted by this distinction betweenanxious arousal and anxious apprehension particularly in the light ofHellerrsquo s model If anxious apprehension is associated with left anterioractivity the overall pattern of regional brain activity appears to be identicalto the brain activity associated with pleasant emotional states (ie more leftthan right anterior activity) However anxious apprehension might beexpected to fall in the portion of the circumplex model of emotionassociated with activated unpleasant affect (as suggested by adjectivessuch as fearful nervous jittery anxious) Below we suggest a numberof theoretical possibilities for resolving this apparent discrepancy all ofwhich remain to be examined empirically
In general it will be important to clarify the degree to which anxiousapprehension is in fact characterised by unpleasant versus pleasant valenceas well as high versus low arousal In addition it may be helpful toconsider these anterior asymmetries from the approach-withdrawal frame-work employed by Davidson (1992ab Davidson amp Tomarken 1989)Under conditions where anxious apprehension leads to approach tenden-cies one might expect an asymmetry in favour of the left anterior region Incontrast when anxious apprehension is associated with withdrawal tenden-cies (perhaps as anxious arousal increases) an asymmetry may emerge infavour of the right anterior region Conceivably the degree to whichanxious apprehension is characterised by approach versus withdrawaltendencies would depend on the degree of stress and discomfortinvolved the personrsquo s coping style and other factors With regard tovalence we would predict that anxious apprehension characterised byapproach tendencies would be accompanied by relatively pleasant affectAnxious apprehension accompanied by withdrawal tendencies in contrastwould be accompanied by relatively unpleasant affect
It is also possible that instead of a direct link between anxious appre-hension and approach behaviour they share more fundamental character-istics associated with left anterior functions For example one might arguethat they both involve sequential thinking and the development of hier-archically structured plans focusing on future concerns (for a review ofrelevant left anterior functions see Tomarken amp Keener this Issue)
Another important issue has to do with the nature of the asymmetricpattern of activity As reviewed earlier we found that the asymmetry infavour of the left hemisphere for participants classireg ed on the basis of traitanxiety was due to a relative decrease in right hemisphere activity Thusthe left frontal region of the anxious participants was not signireg cantly moreactive than that of controls It will therefore be important to examine thepossibility that absolute activity of the left anterior region plays a more
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 437
important role in the modulation of valence (eg Davidson et al 1995)than of anxious apprehension
Another possibility is that independent left anterior systems are asso-ciated with anxious apprehension versus pleasant affect Liotti and Tucker(1995) review theories that there may be multiple functionally segregatedparallel frontal cortico-striato-thalamic circuits some associated withmotor functions some with set shifting and behavioural inhibition andsome with emotional information processing and hedonic responsivity tostimuli Possibly one of these circuits remacr ects activation associated withpleasant affect and another remacr ects activation associated with anxiousapprehension andor approach Different neurotransmitter systems couldbe involved More precise methods of localisation would therefore benecessary to distinguish the unique brain activity associated with each ofthese psychologicalbehavioural phenomena
Comorbidity of Depression and Anxiety
In considering the issue of decomposing emotional states with regard to thesymptomatology in affective and anxiety disorders we have previouslyargued that issues of comorbidity of depression and anxiety are crucial(Heller et al 1995) A review of the literature reveals that rates of comor-bidity are extremely high (Alloy Kelly Mineka amp Clements 1990 Helleret al 1995 Hiller et al 1989 Katon amp Roy-Byrne 1991) Even in our veryrecent work (Nitschke et al submitted) using the MASQ a measuredesigned by Watson and colleagues to avoid symptom overlap in themeasurement of anxiety and depression (Watson et al 1995ab) comor-bidity was fairly pronounced in a large college student sample The MASQAnxious Arousal (AA) and Anhedonic Depression (AD) scales were used asmeasures of those constructs and the PSWQ provided the measure ofanxious apprehension Of the 125 (out of 783) participants scoring abovethe 80th percentile on at least one of the three scales of interest and belowthe 50th percentile on any where the 80th percentile cut-off was not metfully 50 demonstrated comorbidity by scoring above the 80th percentileon at least two of the scales This rate of comorbidity emerged despite thelow correlations found between the three scales (MASQ AA and MASQAD r = 20 MASQ AA and PSWQ r = 20 MASQ AD and PSWQ r =36) relative to the correlations found for the other more general measuresof depression including the MASQ General Mixed General Depression andGeneral Anxiety scales Beck Depression Inventory Beck Anxiety Inven-tory General Behaviour Inventory-Dysthymia scale and the State-TraitAnxiety Inventory-Trait scale (ranging from r = 47 to r = 76)
Because at least some studies have found opposite patterns of activityfor the right posterior region during depression versus during anxiety (ie
438 HELLER AND NITSCHKE
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
In another variant of a valence theory Tucker (1981) concluded thatunpleasant and pleasant emotions were associated with the left and righthemispheres respectively a reversal of the valence theory described ear-lier Here the lesion data were explained by arguing that the observedbehaviour was not a function of the intact hemisphere but rather remacr ectedemotional processes of the subcortical areas on the damaged side that werereleased from cortical inhibition because of the lesion The EEG datawould be explained by assuming that increased anterior activity on oneside or the other would inhibit the affect associated with the correspondingipsilateral subcortical region (unpleasant for the left pleasant for the right)By this reasoning it follows that the affective effects of a lateralisedsubcortical lesion should be opposite to those of a cortical lesion (eg alesion to left subcortical structures should be associated with a suppressionof activity in this area leading to a decrease in unpleasant affect) How-ever several studies have found left anterior subcortical lesions to beassociated with depression (for a review see Liotti amp Tucker 1995)
In attempting to account for the co-occurrence in depression of increasedactivation in right anterior regions and dereg cits on cognitive tasks specia-lised to the right hemisphere Tucker (1981) argued that frontal regions havebeen traditionally viewed as having inhibitory regulatory effects on poster-ior regions By distinguishing valence from arousal as components ofemotion and localising these dimensions to anterior and posterior regionsof the brain respectively Hellerrsquo s (1986 1990) model also emphasised acaudal distinction However the model has additional explanatory powerAlthough inhibitory mechanisms of the anterior regions may well be play-ing a role in affective regulation (also see Tomarken amp Keener this Issue)the concept of a (relatively) independent arousal dimension associated withthe posterior right hemisphere allows us to model the patterns of brainactivity associated with all the emotions as represented on the circum-plex Indeed the patterns of activity that have been reported for depressionand anxiety cannot be fully accounted for by invoking an inhibitorymechanism For example if the inhibitory function of the right anteriorregion is the driving force behind right posterior activity in differentaffective states we would expect anxiety (which like depression is oftenaccompanied by increased right anterior activity) to be similarly accompa-nied by decreased right posterior activity However the evidence suggeststhat anxiety is often associated with increased activity in this region (asreviewed later) Thus by positing an independent arousal dimensionHellerrsquo s model provides an explanation for the patterns of brain activitythat have been observed for anxiety as well as depression
Tucker and colleagues have furthermore emphasised the important roleof cortical-subcortical interactions in producing the asymmetries in corticalactivity Hellerrsquo s model does not address the mechanisms by which the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 425
asymmetries associated with affective valence are generated indeed itcould well be the case that the relationship between activity in the anteriorregions and valence is mediated largely via interaction with subcorticalstructures Furthermore inhibitory mechanisms may be important in thisregard
Note that the approach we are emphasising in this article is compatiblewith Davidson and colleaguesrsquo attempts to decompose emotional statesaccording to the degree that they involve approach versus withdrawalbehaviours (Davidson 1992 Davidson amp Tomarken 1989 Tomarken ampKeener this Issue see also Kinsbourne 1988) These authors have arguedthat approach and withdrawal are the fundamental behavioural substratesassociated with affective valence However this theoretical formulationclearly emphasises the anterior cortical asymmetries associated with plea-santunpleasant valence Although Davidson has addressed posterior asym-metries in his work (eg Davidson Schaffer amp Saron 1985 Davidson1992a) the approach-withdrawal model does not focus on the patterns ofposterior brain activity that have been documented for depression andanxiety Thus considering the arousal dimension in addition to theapproach-withdrawal distinction may address the data for the posteriorregions more comprehensively
The present article has several goals We will highlight persistent andemerging discrepancies in the reg eld of brain function in depression andanxiety and suggest that turning to the psychological literature on thespecireg c symptomatology and comorbidity of these disorders can help usresolve at least some of them In the process we will revisit Hellerrsquo s modelof emotion with a critical look at how well it has been able to account forthe wealth of recent reg ndings in the reg eld Some modireg cations to this modelwill be proposed and continuing questions will be explicated
REGIONAL BRAIN ACTIVITY
Anterior and posterior asymmetries in cortical activity are often but notalways reported in depression and anxiety As a result it has been difreg cultto identify precisely the patterns of brain activity that are important In thesections following we highlight the discrepancies in the reg ndings foranterior and posterior asymmetries We then discuss possible causes forthese discrepancies and suggest ways to disentangle the results
Depression
Anterior asymmetries for EEG alpha have often been found in studies ofdepression (eg Allen Iacono Depue amp Arbisi 1993 Henriques ampDavidson 1990 1991 Schaffer Davidson amp Saron 1983) Across these
426 HELLER AND NITSCHKE
studies depression has been associated with more right than left anterioractivity Numerous positron emission tomography (PET) studies have alsoreported less left anterior activity (eg Baxter et al 1985 1989 Bench etal 1992 Bench Friston Brown Frackowiak amp Dolan 1993 George etal 1994 Martinot et al 1990) which is consistent with the EEG reg ndingsearlier However other studies have reported reg ndings in the oppositedirection for depression (Drevets et al 1992) and sad mood (George etal 1995) Null effects for depression have also been found (eg NitschkeHeller Etienne amp Miller 1995 Tomarken amp Davidson 1994)
The reg ndings for the parieto-temporal regons have been even morevariable In EEG and blood macr ow studies that report positive reg ndingsdepression is typically associated with less right posterior activity (egFlor-Henry 1979 Henriques amp Davidson 1990 Post et al 1987 Uytden-hoef et al 1983) Numerous studies however report no effects for thisregion of the brain (for a review see Davidson amp Tomarken 1989)
Reports of reduced right parieto-temporal activity although not alwayspresent are consistent with neuropsychological studies often indicatingimpaired or suppressed right hemisphere function in depressed peopleOn neuropsychological tasks thought to depend on right parietotemporalfunctioning depressed people have often been found to display decrementsin performance (eg Berndt amp Berndt 1980 Flor-Henry 1976 GoldsteinFilskov Weaver amp Ives 1977 Gruzelier Seymour Wilson Jolley ampHirsch 1988 Kronfol Hamsher Digre amp Waziri 1978) Similar findingsfor sad mood in nondepressed people have also been reported for righthemisphere tasks (Tucker Stenslie Roth amp Shearer 1981)
Studies that have directed information to one or the other hemisphereusing lateralised paradigms such as dichotic listening or visual half-reg eldpresentation have also found specireg c dereg cits for the right hemisphere indepressed people (for a review see Bruder 1995) Left-ear decrements orabsent left-ear (right hemisphere) advantages have been reported fordepression on nonverbal dichotic listening tasks (eg Bruder et al1989) Similarly left visual reg eld (right hemisphere) dereg cits in reactiontime and accuracy for nonverbal stimuli have been reported for depressedpeople (Bruder et al 1989 Liotti Sava Rizzolati amp Caffarra 1991)Comparable results have also been obtained in studies of nondepressedpeople who undergo mood inductions (Banich Stolar Heller amp Goldman1992 Ladavas Nicoletti Umilta amp Rizzolatti 1984) Reaction time andaccuracy for the right hemisphere are selectively impaired in sad moods
On the Chimeric Faces Task (CFT) a free-vision task of face processingthat typically elicits a left hemispatial bias suggesting greater right hemi-sphere activation Jaeger Borod and Peselow (1987) found smaller lefthemispatial biases for depressed patients than nonpsychiatric controlsUsing the same task Heller Etienne and Miller (1995) reported that
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 427
high-depressed participants had smaller left hemispatial biases than low-depressed participants Due to the nature of the CFT hemispatial biases arenonspecireg c as to whether they remacr ect a decrease in activity for the con-tralateral hemisphere or an increase in activity for the ipsilateral hemi-sphere (ie a reduction in the magnitude of the left hemispatial bias couldremacr ect either reduced right hemisphere or increased left hemisphere activ-ity) However based on the evidence (reviewed earlier) for selectivedereg cits in performance on right hemisphere tasks relatively poor perfor-mance for the left-visual reg eld on tachistoscopic tasks and reduced EEGactivity for right posterior regions the most likely interpretation of the CFTreg ndings is that they remacr ect diminished right posterior activity
In summary although anterior asymmetries have often been reported indepression the more common reg nding of higher left than right activation isnot invariably found Furthermore there is a signireg cant discrepancybetween the number of studies that have reported parieto-temporal asym-metries in activation which have been relatively few and the number ofstudies that have reported dereg cits in right hemisphere information proces-sing which have been many Also worthy of note is whether studies ofclinical depression psychometrically dereg ned depression and sad mood areaddressing the same phenomenon The number of consistencies acrossthese populations in the physiological literature reviewed here suggestssome commonality on the other hand the inconsistencies could be in partdue to the heterogeneity between the samples studied
Anxiety
Research on brain activity in anxious states has also been plagued byinconsistencies Again although anterior and posterior asymmetries havebeen reported no consistent pattern has emerged Studies of multisite EEGand regional cerebral blood macr ow (rCBF) using xenon-133 or PET havereported right hemisphere left hemisphere and bilateral increases inactivity during anxiety (for a review see Heller Nitschke Etienne ampMiller 1997a) Other studies have found no differences in brain activity(Gur et al 1988 Mountz et al 1989 Nordhal et al 1990 Tomarken ampDavidson 1994)
Having comprehensively reviewed this literature elsewhere (Heller etal 1997a) we will provide an overview of those studies reporting differ-ential hemispheric involvement in anxiety A subset of these suggest anasymmetry in favour of right hemisphere activation in various regions(frontal lateral prefrontal precentral frontal anterior temporal parietaloccipital and parahippocampal) Heightened right hemisphere activity hasbeen reported for panic disorder patients (Reiman Raichle Butler Hers-covitch amp Robins 1984) panic disorder patients experiencing a sodium
428 HELLER AND NITSCHKE
lactate-induced panic attack (Stewart Devous Rush Lane amp Bonte1988) generalised anxiety disorder (GAD) patients during an anxiety-stimulating task (Wu et al 1991) psychometrically dereg ned high-anxiousparticipants immediately following arterial and venous catheterisation andagain after completion of the PET scans (Reivich Gur amp Alavi 1983) andsocial phobics prior to making a public speech (Davidson MarshallTomarken amp Henriques submitted) Consistent with this pattern anti-anxiety medication was found to reduce the rate of glucose metabolismin the right hemisphere especially for the occipital and frontal regions forpatients with GAD (Buchsbaum et al 1987) Similarly in a population ofnonpsychiatric volunteers those administered diazepam intravenouslyshowed maximal decreases in rCBF in the right frontal area whereasthose injected with a placebo showed no rCBF changes (Mathew Wil-son amp Daniel 1985)
Regional brain activity can also be inferred from performance on tasksthat depend on particular brain regions On the CFT Heller et al (1995)reported that high-anxious participants had larger left hemispatial biasesthan low-anxious participants suggested to remacr ect greater right posterioractivity Similarly on a neuropsychological test battery panic disorderpatients performed less well than normal controls on left but not righthemisphere tasks (Yeudall et al 1983) Data obtained from a series ofstudies using behavioural measures to examine right versus left hemispherefunction during an experimental condition producing transient (ie state)anxiety (Tucker Antes Stenslie amp Barnhardt 1978 Tucker Roth Arne-son amp Buckingham 1977 Tyler amp Tucker 1982) are also consistent withan interpretation of greater right than left hemisphere activity in anxiety(Heller et al 1995)
However when subjects were classireg ed according to trait anxietyTucker and colleagues (Tucker et al 1978 Tyler amp Tucker 1982)interpreted the results as strongly suggestive of greater left than righthemisphere activity This interpretation is consistent with reports ofnumerous other labs reg nding increased left hemisphere activity associatedwith anxiety in various cortical and subcortical regions (orbital frontalinferior frontal anterior cingulate caudate putamen and thalamus)Heightened left hemisphere activity has been found in obsessive-compul-sive disorder (OCD) patients (Baxter et al 1987 Swedo et al 1989)GAD patients (Buchsbaum et al 1985 Wu et al 1991) and studentsclassireg ed as ` worriersrsquo rsquo on the basis of self-reports when asked to ` worryabout a specireg c topic of personal concernrsquo rsquo (Carter Johnson amp Borkovec1986)
In summary some reg ndings for anxiety indicate increased right hemi-sphere activation whereas others indicate increased left hemisphere acti-vation Indeed it may be the case that they could co-occur which could
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 429
account for the reg ndings of bilateral increases or no asymmetries In themeantime we are lacking a model to explain these discrepancies
ACCOUNTING FOR INCONSISTENCIES INBRAIN ACTIVITY
We are thus confronted with inconsistent results with regard to regionalbrain activity in both depression and anxiety In previous work we haveargued that in order to resolve these discrepancies it is helpful to turn tothe psychological literature on symptomatology in depression and anxietyIn the sections that follow we make some general observations that applyto both depression and anxiety before focusing on each more exclusively
In addressing the inconsistencies for the blood macr ow reg ndings severalconcerns about the literature are worthy of note First several of the mostfrequently cited studies in this literature sampled a restricted set of brainregions (eg Reiman et al 1984 Reivich et al 1983) Furthermore theregions of interest were not consistent from one study to the nextSecond statistical procedures employing multiple t-tests rather thanrepeated-measures ANOVAs are questionable especially with regard todrawing conclusions about hemispheric differences Third although somestudies examined selected diagnostic groups (eg major depressive dis-order OCD panic disorder GAD) most treated depression or anxiety asa unitary construct and did not consider that different types of each mightbe associated with different patterns of brain activity Moreover with theexception of recent OCD studies most did not take into account thedegree of comorbidity with depression or anxiety which may be criticalin determining the patterns of brain activity that occur (Heller et al1995 Tucker 1988)
As emphasised by Davidson Tomarken and colleagues (Davidson1993 Tomarken Davidson Wheeler amp Doss 1992 Tomarken David-son Wheeler amp Kinney 1992) there are a number of methodogicalapproaches that need to be considered when asymmetric patterns ofEEG activity do not emerge Some of these considerations are relevantto the design of blood macr ow studies as well One issue is that of stability(ie the extent to which brain activity remacr ects a stable individual pattern)This can be addressed by obtaining a sufreg cient number of baseline samplesas well as test-retest data For example Tomarken et al (1992) found thatonly participants with a stable pattern of anterior asymmetry across twomeasurement occasions showed an association between left frontalhypoactivation and more dispositional negative affect as measured bythe PANAS
It is also important to note that for the most part EEG studies haveexamined resting or baseline activity Davidson (1993) has pointed out
430 HELLER AND NITSCHKE
that baseline levels of frontal activation asymmetry may remacr ect individualtendencies toward a valenced affective response but this relationship maynot be observed in the absence of a specireg c environmental elicitor Thushe emphasises the need to examine the relationship between individualdifferences in patterns of asymmetry and affective response to emotionallychallenging tasks As applied to the pattern of posterior reg ndings theseconsiderations suggest the possibility that more robust differences inposterior brain activity might be yielded by studies measuring brainactivity while participants perform tasks for which depressed peopleshow decrements Additional methodological approaches highlighted byDavidson and colleagues such as the use of extreme groups and classifica-tion of participants according to the affect measures as opposed to the brainasymmetry measures would also be useful in future studies of brainactivity
A number of other possible explanations for the failure to reg nd con-sistent decreases in parieto-temporal activity during depression have beenproposed (Heller 1990 1993b) One of these explanations focused on theseverity of the depression This supposition was based on the observationthat posterior asymmetries were more likely to be reported when depres-sion was relatively severe Heller suggested that under less extremeaffective conditions task-induced activity predominates masking theaffect-induced activity More recently Heller et al (1995) suggestedthat severity per se is not the issue rather because the comorbiditywith anxiety decreases as severity of depression increases (Hiller Zau-dig amp Rose 1989) the opposing effects of anxiety on posterior brainactivity diminishes
Heller (1993b) also suggested that the inconsistencies in the depressionand anxiety literatures may be due to signireg cant differences in regionalbrain activity that covary more strongly with dispositional characteristicsassociated with emotional functioning (eg extraversion) than with situa-tionally induced emotions Several studies have reported perceptual(Berenbaum amp Williams 1994 Heller amp Nisenson 1993) and EEG(Nitschke et al 1994) asymmetries for extraversion Nitschke et al alsofound that extraversion moderated the relationship between anxiety andEEG activity across both frontal central and parietal regions For anxiousparticipants high extraversion was associated with less activity (morealpha) whereas for nonanxious (and nondepressed) controls high extra-version was associated with more activity (less alpha) These two sets ofreg ndings albeit preliminary provide evidence for the need to acknowledgeregional differences in brain function for personality characteristics suchas extraversion (for a review see Nitschke Heller amp Miller in prep) andpoint to the possible signireg cance of personality in understanding regionalbrain activity in affective and anxiety disorders
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 431
Subtypes of Depression and Anxiety
Another promising avenue for elucidating the discrepancies reported in theliterature for both depression and anxiety is the analysis of subtypes
Depression Many distinctions for depression have been proposed overthe years including most notably bipolar versus unipolar endogenousversus exogenous and neurotic versus reactive The current Diagnostic andstatistical manual of mental disordersplusmn fourth edition (DSMIV APA 1994)subdivides unipolar depression into major depressive disorder and dysthy-mic disorder and provides further specireg city with criteria for severity andfor psychotic melancholic catatonic or atypical features
Perhaps the most useful distinction in physiological research on depres-sion has been the comparison of melancholic to nonmelancholic depres-sion The dereg ning feature of melancholic depression is anhedonia which isthe inability to experience pleasure Current consensus as indicated by theDSMIV (APA 1994) suggests that melancholia is most strongly charac-terised by a loss of pleasure in all activities and a lack of responsivity topleasurable stimuli Many other symptoms may be shared by both melan-cholic and nonmelancholic depression including depressed mood distur-bances in eating and sleeping patterns low self-esteem inability toconcentrate psychomotor retardation fatigue indecisiveness feelings ofguilt and thoughts about death There is current controversy over whethermelancholic depression is qualitatively different from nonmelancholicdepression with some authors arguing that the primary distinction is oneof severity (Zimmerman Coryell amp Pfohl 1986) Nonetheless neuropsy-chological and neurophysiological studies have been consistent in demon-strating differences between melancholic and nonmelancholic depression
In a comprehensive series of studies Bruder has shown that perceptualasymmetries differ for different subtypes of depression (for a review seeBruder 1995) On a perceptual asymmetry task melancholic patientsshowed an abnormally large right ear (left hemisphere) advantage for adichotic syllables task but no left ear (right hemisphere) advantage for acomplex tone task (Bruder 1995) The lack of a left ear advantage was dueto poor left ear accuracy consistent with other studies demonstrating righthemisphere dereg cits in depressed people However nonmelancholicdepressed patients did not differ from normal controls on either test Alongthese lines in a PET study of melancholic depression Bench et al (1992)reported less activity in the left dorsolateral prefrontal cortex compared tonondepressed controls Note that these data could be consistent with eithera qualitative or a quantitative distinction between melancholic and non-melancholic depression However other data may be more consistent witha qualitative difference between the two types of depression For example
432 HELLER AND NITSCHKE
Davidson Larson and Abercrombie (1995) reported that depression sever-ity in melancholic subjects was predicted by the extent to which activationwas decreased in the left lateral prefrontal region In contrast amongnonmelancholics depression severity was predicted by the extent to whichactivation was increased in the right medial prefrontal area
Davidson et alrsquo s (1995) data have implications for theories regardingthe relationship between pleasantunpleasant affect and anterior regions ofthe brain Research on the psychological characteristics of depression andanxiety has suggested that lack of positive affect (or more preciselyactivated pleasant affect Larsen amp Diener 1992) is unique to depression(Clark amp Watson 1991) Indeed in designing a questionnaire to measureboth general and specireg c components of depression and anxiety Watsonand colleagues (Watson 1995ab) developed a specireg c depression scalethat primarily measures anhedonia Because anhedonia is a fundamentalcharacteristic of melancholia it seems reasonable to assume that decreasedpleasant affect would accompany this type of depression If so Hellerrsquo smodel which predicts less left than right anterior activity for decreasedpleasant affect is completely consistent with the available data
However Davidson et alrsquo s (1995) data suggests the possibility thataffective valence may depend less on the relative balance of activationbetween the left and right anterior regions and more on specireg c systemsassociated with each hemisphere Whereas melancholic depression ischaracterised by decreased pleasant affect depression without melan-cholic features may be better characterised by increased unpleasantaffect Following this line of reasoning Hellerrsquo s model which in itscurrent form argues for a link between anterior asymmetry and the valencedimension would not distinguish between melancholic and nonmelan-cholic depression If affective valence depends more on specireg c systemsassociated with each hemisphere as Davidson et alrsquo s data suggest Hellerrsquo smodel may need to be expanded to remacr ect the unique contributions of theleft and right hemispheres to pleasantunpleasant affect For example itmight be more accurate to link the left anterior region with the activatedpleasant affect (ie positive affect) octant of the circumplex model ofemotion and the right anterior region with the activated unpleasant affect(ie negative affect) octant (see Larson amp Diener 1992)
Because few studies have specireg cally examined the full range of poten-tial affective descriptors in relation to patterns of brain activity theseissues remain to be resolved In most research only isolated portions ofthe circumplex have been sampled For example because positive andnegative affect as measured by the PANAS provide information regardingonly two of the four quadrants in the circumplex model studies that do notcollect data regarding the other two quadrants cannot shed light on howwell they might have accounted for patterns of brain activity
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 433
Anxiety As reviewed earlier studies that examine regional brainactivity in anxiety are even more plagued by inconsistencies than studiesof depression In keeping with the theme of this article we have argued thatby decomposing anxiety into subtypes on the basis of the psychologicalliterature we can potentially resolve some of these discrepancies
In previous work we have suggested that it may be important toconsider the type of anxiety that is being examined (Heller et al 19951997a) In particular we argued that regional brain activity might bedifferent for anxious arousal (eg panic) and anxious apprehension (egworry) types of anxiety that appear to differ signireg cantly in psychologicaland physiological characteristics (eg Barlow 1988 1991 Eysenck 1957Klein 1987 Watson et al 1995b) Anxious arousal has been described byWatson et al as being distinguished by symptoms of physiological hyper-arousal and somatic tension Panic attacks and high-stress states would fallinto this category In contrast anxious apprehension involves worry and ischaracterised by verbal rumination typically about future events This typeof anxiety would be characteristic of obsessive-compulsiveness general-ised anxiety states and trait anxiety (as identireg ed by self-reports of anxiousapprehension and worry on various questionnaires) As noted by Heller etal (1997a) the two kinds of anxiety are not mutually exclusive Extremedegrees of anxious apprehension (eg being worried or fearful about thefuture) may prompt experiences of anxious arousal (eg somatic symptomsand exaggerated physiological responses to stressful events)
We have argued that when the literature on brain activity in anxiety isreviewed in the light of this distinction between anxious arousal andanxious apprehension most of the studies reporting greater right hemi-sphere activity have either studied panic attacks (eg Reiman et al 1984)or high-stress situations likely to entail anxious arousal (eg Tucker et al1977) On the other hand studies indicating greater left hemisphere activity(primarily anterior) have examined OCD (eg Baxter et al 1987) GAD(Wu et al 1991) or questionnaire-dereg ned trait anxiety indexing anxiousapprehension (eg Tucker et al 1978) Indeed as mentioned earlierstudies have highlighted various regions in both hemispheres As aresult it is difreg cult to pinpoint a particular region that is clearly a focalpoint in anxiety Nonetheless the data do suggest that right hemisphereregions emerge as particularly salient in some types of anxiety (ie anxiousarousal) whereas left hemisphere regions emerge as particularly salient inother types of anxiety (ie anxious apprehension)
In a direct test of this hypothesis (Heller et al 1997a) we selectedparticipants on the basis of self-reported trait anxiety using a measureindexing anxious apprehension in a between-subjects design We thenmanipulated anxious arousal on a within-subjects basis Specireg cally wecontrasted brain activity (EEG alpha) during rest periods to brain activity
434 HELLER AND NITSCHKE
during a task in which participants listened to emotional narrativesdesigned to elicit anxious arousal (ie fear and sad narratives that hadpreviously received unpleasant valence and high arousal ratings)
For frontal activity group differences were unaffected by the narrativesAnxious participants showed a larger asymmetry in favour of the lefthemisphere than controls during the rest and listen periods of all narrativetypes as well as for eight minutes of resting baseline preceding thenarrative task Examination of the data indicated that the increased asym-metry was primarily due to a decrease in right hemisphere activity for theanxious participants Nonetheless the increase in magnitude of the asym-metry favouring the left hemisphere found for anxious participants isconsistent with the studies reviewed earlier that reported or inferredrelatively greater left than right anterior activity in anxious subjectsclassireg ed on the basis of worry or trait anxiety (eg Carter et al 1986Tucker et al 1978 Tyler amp Tucker 1982) Our frontal data suggest that itwill be important to investigate whether anxious apprehension is bestcharacterised by the magnitude of the asymmetry favouring the left hemi-sphere or by the level of left hemisphere activity in isolation
In contrast the parietal data supported our prediction regarding theassociation of anxious arousal with greater right hemisphere activityWhen listening to the fear and sad narratives anxious participants showeda selective increase in right parietal activity that was not demonstrated bycontrols The right parietal effect however was not obtained for neutraland happy narratives that had previously received pleasant valence and lowarousal ratings These reg ndings are consistent with those studies that havereported or inferred greater right hemisphere activity during anxious statescharacterised by panic and stress-inducing conditions (eg Reiman et al1984 Stewart et al 1988 Tucker et al 1977)
An important question here has to do with the discrepancies in frontalasymmetries that have been reported Like others (eg Baxter et al 1987Gur et al 1988 Wu et al 1991) we did not reg nd heightened right anterioractivity in anxiety However a number of other studies have (eg Davidsonet al 1996 Reivich et al 1983 Tomarken amp Davidson 1994) Asdiscussed earlier we have previously argued that anterior regions areuniquely involved in the valence dimension of emotion with pleasantaffect associated with greater left activity and unpleasant affect associatedwith greater right activity (eg Heller 1990) Therefore we would expectheightened right anterior activity in an anxious group only if they differedfrom controls on measures indexing affective valence In our recent studythe nondepressed anxious group did not differ from controls on either theGBI Dysthymia scale or the PANAS PA and NA scales
It may also be important to measure the level of anxious apprehension inthe sample studied For example Tomarken and Davidson (1994) found
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 435
that with Marloweplusmn Crowne scores controlled trait anxiety failed to predictfrontal asymmetry However their groups were divided according to amedian split
In summary our results directly support the proposed distinctionbetween anxious apprehension and anxious arousal as types of anxietythat may account for signireg cant variance in patterns of brain activityThey suggest furthermore that Hellerrsquo s model should be rereg ned to reflectthe likelihood that parietal activation will covary with anxious arousal butnot with anxious apprehension If these distinctions are accurate we wouldexpect to observe increases in right parietotemporal activity only underconditions when anxiety is accompanied by increased arousal
Thus in homodynamic and EEG studies a failure to reg nd increased rightposterior activity in anxiety is not surprising as variations in anxiousarousal are likely to be large Some individuals may experience anxiousarousal frequently especially those with disorders more characteristic ofanxious arousal such as panic disorder phobias or post-traumatic stressdisorder Others although apprehensive may rarely experience anxiousarousal In addition the situation may vary in the degree to which it elicitsanxious arousal Being forced to give a speech is likely to induce a greatdeal of anxious arousal in social phobics thus we would have predicted asignireg cant increase in right parietal activity precisely as found by David-son et al (submitted) Note that this effect was only found during theAnticipation condition not during baseline Planning or after thespeech During the Anticipation condition subjects listened while a voiceannounced every 30 seconds for three minutes the time remaining beforethey would be given a topic and allowed to plan for their speech Pre-sumably this period is the most stressful and the most likely to result inanxious arousal Indeed the correlation between heart rate acceleration tothe Anticipation condition and the increase in state anxiety was signireg cant
More recently Nitschke Heller Imig and Miller (submitted) haveattempted to identify anxious arousal and anxious apprehension usingquestionnaires Data from 783 students for a battery of anxiety and depres-sion questionnaires provided strong support for the anxiety distinctionproposed here at least on a psychological level of analysis The Moodand Anxiety Symptom Questionnaire-Anxious Arousal scale (MASQWatson et al 1995ab) was chosen as a measure of anxious arousalwhereas the Penn State Worry Questionnaire (PSWQ Meyer MillerMetzger amp Borkovec 1990 Molina amp Borkovec 1994) was included asa measure of anxious apprehension The correlation between the two scales(r = 20) was much lower than those involving the other more generalscales of anxiety included in the study (MASQ-General Anxiety scaleBeck Anxiety Inventory and the State-Trait Anxiety Inventory-Traitscale ranging from r = 47 to r = 65) These results further support the
436 HELLER AND NITSCHKE
argument that anxious arousal and anxious apprehension are distinctanxiety states
However new questions are highlighted by this distinction betweenanxious arousal and anxious apprehension particularly in the light ofHellerrsquo s model If anxious apprehension is associated with left anterioractivity the overall pattern of regional brain activity appears to be identicalto the brain activity associated with pleasant emotional states (ie more leftthan right anterior activity) However anxious apprehension might beexpected to fall in the portion of the circumplex model of emotionassociated with activated unpleasant affect (as suggested by adjectivessuch as fearful nervous jittery anxious) Below we suggest a numberof theoretical possibilities for resolving this apparent discrepancy all ofwhich remain to be examined empirically
In general it will be important to clarify the degree to which anxiousapprehension is in fact characterised by unpleasant versus pleasant valenceas well as high versus low arousal In addition it may be helpful toconsider these anterior asymmetries from the approach-withdrawal frame-work employed by Davidson (1992ab Davidson amp Tomarken 1989)Under conditions where anxious apprehension leads to approach tenden-cies one might expect an asymmetry in favour of the left anterior region Incontrast when anxious apprehension is associated with withdrawal tenden-cies (perhaps as anxious arousal increases) an asymmetry may emerge infavour of the right anterior region Conceivably the degree to whichanxious apprehension is characterised by approach versus withdrawaltendencies would depend on the degree of stress and discomfortinvolved the personrsquo s coping style and other factors With regard tovalence we would predict that anxious apprehension characterised byapproach tendencies would be accompanied by relatively pleasant affectAnxious apprehension accompanied by withdrawal tendencies in contrastwould be accompanied by relatively unpleasant affect
It is also possible that instead of a direct link between anxious appre-hension and approach behaviour they share more fundamental character-istics associated with left anterior functions For example one might arguethat they both involve sequential thinking and the development of hier-archically structured plans focusing on future concerns (for a review ofrelevant left anterior functions see Tomarken amp Keener this Issue)
Another important issue has to do with the nature of the asymmetricpattern of activity As reviewed earlier we found that the asymmetry infavour of the left hemisphere for participants classireg ed on the basis of traitanxiety was due to a relative decrease in right hemisphere activity Thusthe left frontal region of the anxious participants was not signireg cantly moreactive than that of controls It will therefore be important to examine thepossibility that absolute activity of the left anterior region plays a more
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 437
important role in the modulation of valence (eg Davidson et al 1995)than of anxious apprehension
Another possibility is that independent left anterior systems are asso-ciated with anxious apprehension versus pleasant affect Liotti and Tucker(1995) review theories that there may be multiple functionally segregatedparallel frontal cortico-striato-thalamic circuits some associated withmotor functions some with set shifting and behavioural inhibition andsome with emotional information processing and hedonic responsivity tostimuli Possibly one of these circuits remacr ects activation associated withpleasant affect and another remacr ects activation associated with anxiousapprehension andor approach Different neurotransmitter systems couldbe involved More precise methods of localisation would therefore benecessary to distinguish the unique brain activity associated with each ofthese psychologicalbehavioural phenomena
Comorbidity of Depression and Anxiety
In considering the issue of decomposing emotional states with regard to thesymptomatology in affective and anxiety disorders we have previouslyargued that issues of comorbidity of depression and anxiety are crucial(Heller et al 1995) A review of the literature reveals that rates of comor-bidity are extremely high (Alloy Kelly Mineka amp Clements 1990 Helleret al 1995 Hiller et al 1989 Katon amp Roy-Byrne 1991) Even in our veryrecent work (Nitschke et al submitted) using the MASQ a measuredesigned by Watson and colleagues to avoid symptom overlap in themeasurement of anxiety and depression (Watson et al 1995ab) comor-bidity was fairly pronounced in a large college student sample The MASQAnxious Arousal (AA) and Anhedonic Depression (AD) scales were used asmeasures of those constructs and the PSWQ provided the measure ofanxious apprehension Of the 125 (out of 783) participants scoring abovethe 80th percentile on at least one of the three scales of interest and belowthe 50th percentile on any where the 80th percentile cut-off was not metfully 50 demonstrated comorbidity by scoring above the 80th percentileon at least two of the scales This rate of comorbidity emerged despite thelow correlations found between the three scales (MASQ AA and MASQAD r = 20 MASQ AA and PSWQ r = 20 MASQ AD and PSWQ r =36) relative to the correlations found for the other more general measuresof depression including the MASQ General Mixed General Depression andGeneral Anxiety scales Beck Depression Inventory Beck Anxiety Inven-tory General Behaviour Inventory-Dysthymia scale and the State-TraitAnxiety Inventory-Trait scale (ranging from r = 47 to r = 76)
Because at least some studies have found opposite patterns of activityfor the right posterior region during depression versus during anxiety (ie
438 HELLER AND NITSCHKE
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
asymmetries associated with affective valence are generated indeed itcould well be the case that the relationship between activity in the anteriorregions and valence is mediated largely via interaction with subcorticalstructures Furthermore inhibitory mechanisms may be important in thisregard
Note that the approach we are emphasising in this article is compatiblewith Davidson and colleaguesrsquo attempts to decompose emotional statesaccording to the degree that they involve approach versus withdrawalbehaviours (Davidson 1992 Davidson amp Tomarken 1989 Tomarken ampKeener this Issue see also Kinsbourne 1988) These authors have arguedthat approach and withdrawal are the fundamental behavioural substratesassociated with affective valence However this theoretical formulationclearly emphasises the anterior cortical asymmetries associated with plea-santunpleasant valence Although Davidson has addressed posterior asym-metries in his work (eg Davidson Schaffer amp Saron 1985 Davidson1992a) the approach-withdrawal model does not focus on the patterns ofposterior brain activity that have been documented for depression andanxiety Thus considering the arousal dimension in addition to theapproach-withdrawal distinction may address the data for the posteriorregions more comprehensively
The present article has several goals We will highlight persistent andemerging discrepancies in the reg eld of brain function in depression andanxiety and suggest that turning to the psychological literature on thespecireg c symptomatology and comorbidity of these disorders can help usresolve at least some of them In the process we will revisit Hellerrsquo s modelof emotion with a critical look at how well it has been able to account forthe wealth of recent reg ndings in the reg eld Some modireg cations to this modelwill be proposed and continuing questions will be explicated
REGIONAL BRAIN ACTIVITY
Anterior and posterior asymmetries in cortical activity are often but notalways reported in depression and anxiety As a result it has been difreg cultto identify precisely the patterns of brain activity that are important In thesections following we highlight the discrepancies in the reg ndings foranterior and posterior asymmetries We then discuss possible causes forthese discrepancies and suggest ways to disentangle the results
Depression
Anterior asymmetries for EEG alpha have often been found in studies ofdepression (eg Allen Iacono Depue amp Arbisi 1993 Henriques ampDavidson 1990 1991 Schaffer Davidson amp Saron 1983) Across these
426 HELLER AND NITSCHKE
studies depression has been associated with more right than left anterioractivity Numerous positron emission tomography (PET) studies have alsoreported less left anterior activity (eg Baxter et al 1985 1989 Bench etal 1992 Bench Friston Brown Frackowiak amp Dolan 1993 George etal 1994 Martinot et al 1990) which is consistent with the EEG reg ndingsearlier However other studies have reported reg ndings in the oppositedirection for depression (Drevets et al 1992) and sad mood (George etal 1995) Null effects for depression have also been found (eg NitschkeHeller Etienne amp Miller 1995 Tomarken amp Davidson 1994)
The reg ndings for the parieto-temporal regons have been even morevariable In EEG and blood macr ow studies that report positive reg ndingsdepression is typically associated with less right posterior activity (egFlor-Henry 1979 Henriques amp Davidson 1990 Post et al 1987 Uytden-hoef et al 1983) Numerous studies however report no effects for thisregion of the brain (for a review see Davidson amp Tomarken 1989)
Reports of reduced right parieto-temporal activity although not alwayspresent are consistent with neuropsychological studies often indicatingimpaired or suppressed right hemisphere function in depressed peopleOn neuropsychological tasks thought to depend on right parietotemporalfunctioning depressed people have often been found to display decrementsin performance (eg Berndt amp Berndt 1980 Flor-Henry 1976 GoldsteinFilskov Weaver amp Ives 1977 Gruzelier Seymour Wilson Jolley ampHirsch 1988 Kronfol Hamsher Digre amp Waziri 1978) Similar findingsfor sad mood in nondepressed people have also been reported for righthemisphere tasks (Tucker Stenslie Roth amp Shearer 1981)
Studies that have directed information to one or the other hemisphereusing lateralised paradigms such as dichotic listening or visual half-reg eldpresentation have also found specireg c dereg cits for the right hemisphere indepressed people (for a review see Bruder 1995) Left-ear decrements orabsent left-ear (right hemisphere) advantages have been reported fordepression on nonverbal dichotic listening tasks (eg Bruder et al1989) Similarly left visual reg eld (right hemisphere) dereg cits in reactiontime and accuracy for nonverbal stimuli have been reported for depressedpeople (Bruder et al 1989 Liotti Sava Rizzolati amp Caffarra 1991)Comparable results have also been obtained in studies of nondepressedpeople who undergo mood inductions (Banich Stolar Heller amp Goldman1992 Ladavas Nicoletti Umilta amp Rizzolatti 1984) Reaction time andaccuracy for the right hemisphere are selectively impaired in sad moods
On the Chimeric Faces Task (CFT) a free-vision task of face processingthat typically elicits a left hemispatial bias suggesting greater right hemi-sphere activation Jaeger Borod and Peselow (1987) found smaller lefthemispatial biases for depressed patients than nonpsychiatric controlsUsing the same task Heller Etienne and Miller (1995) reported that
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 427
high-depressed participants had smaller left hemispatial biases than low-depressed participants Due to the nature of the CFT hemispatial biases arenonspecireg c as to whether they remacr ect a decrease in activity for the con-tralateral hemisphere or an increase in activity for the ipsilateral hemi-sphere (ie a reduction in the magnitude of the left hemispatial bias couldremacr ect either reduced right hemisphere or increased left hemisphere activ-ity) However based on the evidence (reviewed earlier) for selectivedereg cits in performance on right hemisphere tasks relatively poor perfor-mance for the left-visual reg eld on tachistoscopic tasks and reduced EEGactivity for right posterior regions the most likely interpretation of the CFTreg ndings is that they remacr ect diminished right posterior activity
In summary although anterior asymmetries have often been reported indepression the more common reg nding of higher left than right activation isnot invariably found Furthermore there is a signireg cant discrepancybetween the number of studies that have reported parieto-temporal asym-metries in activation which have been relatively few and the number ofstudies that have reported dereg cits in right hemisphere information proces-sing which have been many Also worthy of note is whether studies ofclinical depression psychometrically dereg ned depression and sad mood areaddressing the same phenomenon The number of consistencies acrossthese populations in the physiological literature reviewed here suggestssome commonality on the other hand the inconsistencies could be in partdue to the heterogeneity between the samples studied
Anxiety
Research on brain activity in anxious states has also been plagued byinconsistencies Again although anterior and posterior asymmetries havebeen reported no consistent pattern has emerged Studies of multisite EEGand regional cerebral blood macr ow (rCBF) using xenon-133 or PET havereported right hemisphere left hemisphere and bilateral increases inactivity during anxiety (for a review see Heller Nitschke Etienne ampMiller 1997a) Other studies have found no differences in brain activity(Gur et al 1988 Mountz et al 1989 Nordhal et al 1990 Tomarken ampDavidson 1994)
Having comprehensively reviewed this literature elsewhere (Heller etal 1997a) we will provide an overview of those studies reporting differ-ential hemispheric involvement in anxiety A subset of these suggest anasymmetry in favour of right hemisphere activation in various regions(frontal lateral prefrontal precentral frontal anterior temporal parietaloccipital and parahippocampal) Heightened right hemisphere activity hasbeen reported for panic disorder patients (Reiman Raichle Butler Hers-covitch amp Robins 1984) panic disorder patients experiencing a sodium
428 HELLER AND NITSCHKE
lactate-induced panic attack (Stewart Devous Rush Lane amp Bonte1988) generalised anxiety disorder (GAD) patients during an anxiety-stimulating task (Wu et al 1991) psychometrically dereg ned high-anxiousparticipants immediately following arterial and venous catheterisation andagain after completion of the PET scans (Reivich Gur amp Alavi 1983) andsocial phobics prior to making a public speech (Davidson MarshallTomarken amp Henriques submitted) Consistent with this pattern anti-anxiety medication was found to reduce the rate of glucose metabolismin the right hemisphere especially for the occipital and frontal regions forpatients with GAD (Buchsbaum et al 1987) Similarly in a population ofnonpsychiatric volunteers those administered diazepam intravenouslyshowed maximal decreases in rCBF in the right frontal area whereasthose injected with a placebo showed no rCBF changes (Mathew Wil-son amp Daniel 1985)
Regional brain activity can also be inferred from performance on tasksthat depend on particular brain regions On the CFT Heller et al (1995)reported that high-anxious participants had larger left hemispatial biasesthan low-anxious participants suggested to remacr ect greater right posterioractivity Similarly on a neuropsychological test battery panic disorderpatients performed less well than normal controls on left but not righthemisphere tasks (Yeudall et al 1983) Data obtained from a series ofstudies using behavioural measures to examine right versus left hemispherefunction during an experimental condition producing transient (ie state)anxiety (Tucker Antes Stenslie amp Barnhardt 1978 Tucker Roth Arne-son amp Buckingham 1977 Tyler amp Tucker 1982) are also consistent withan interpretation of greater right than left hemisphere activity in anxiety(Heller et al 1995)
However when subjects were classireg ed according to trait anxietyTucker and colleagues (Tucker et al 1978 Tyler amp Tucker 1982)interpreted the results as strongly suggestive of greater left than righthemisphere activity This interpretation is consistent with reports ofnumerous other labs reg nding increased left hemisphere activity associatedwith anxiety in various cortical and subcortical regions (orbital frontalinferior frontal anterior cingulate caudate putamen and thalamus)Heightened left hemisphere activity has been found in obsessive-compul-sive disorder (OCD) patients (Baxter et al 1987 Swedo et al 1989)GAD patients (Buchsbaum et al 1985 Wu et al 1991) and studentsclassireg ed as ` worriersrsquo rsquo on the basis of self-reports when asked to ` worryabout a specireg c topic of personal concernrsquo rsquo (Carter Johnson amp Borkovec1986)
In summary some reg ndings for anxiety indicate increased right hemi-sphere activation whereas others indicate increased left hemisphere acti-vation Indeed it may be the case that they could co-occur which could
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 429
account for the reg ndings of bilateral increases or no asymmetries In themeantime we are lacking a model to explain these discrepancies
ACCOUNTING FOR INCONSISTENCIES INBRAIN ACTIVITY
We are thus confronted with inconsistent results with regard to regionalbrain activity in both depression and anxiety In previous work we haveargued that in order to resolve these discrepancies it is helpful to turn tothe psychological literature on symptomatology in depression and anxietyIn the sections that follow we make some general observations that applyto both depression and anxiety before focusing on each more exclusively
In addressing the inconsistencies for the blood macr ow reg ndings severalconcerns about the literature are worthy of note First several of the mostfrequently cited studies in this literature sampled a restricted set of brainregions (eg Reiman et al 1984 Reivich et al 1983) Furthermore theregions of interest were not consistent from one study to the nextSecond statistical procedures employing multiple t-tests rather thanrepeated-measures ANOVAs are questionable especially with regard todrawing conclusions about hemispheric differences Third although somestudies examined selected diagnostic groups (eg major depressive dis-order OCD panic disorder GAD) most treated depression or anxiety asa unitary construct and did not consider that different types of each mightbe associated with different patterns of brain activity Moreover with theexception of recent OCD studies most did not take into account thedegree of comorbidity with depression or anxiety which may be criticalin determining the patterns of brain activity that occur (Heller et al1995 Tucker 1988)
As emphasised by Davidson Tomarken and colleagues (Davidson1993 Tomarken Davidson Wheeler amp Doss 1992 Tomarken David-son Wheeler amp Kinney 1992) there are a number of methodogicalapproaches that need to be considered when asymmetric patterns ofEEG activity do not emerge Some of these considerations are relevantto the design of blood macr ow studies as well One issue is that of stability(ie the extent to which brain activity remacr ects a stable individual pattern)This can be addressed by obtaining a sufreg cient number of baseline samplesas well as test-retest data For example Tomarken et al (1992) found thatonly participants with a stable pattern of anterior asymmetry across twomeasurement occasions showed an association between left frontalhypoactivation and more dispositional negative affect as measured bythe PANAS
It is also important to note that for the most part EEG studies haveexamined resting or baseline activity Davidson (1993) has pointed out
430 HELLER AND NITSCHKE
that baseline levels of frontal activation asymmetry may remacr ect individualtendencies toward a valenced affective response but this relationship maynot be observed in the absence of a specireg c environmental elicitor Thushe emphasises the need to examine the relationship between individualdifferences in patterns of asymmetry and affective response to emotionallychallenging tasks As applied to the pattern of posterior reg ndings theseconsiderations suggest the possibility that more robust differences inposterior brain activity might be yielded by studies measuring brainactivity while participants perform tasks for which depressed peopleshow decrements Additional methodological approaches highlighted byDavidson and colleagues such as the use of extreme groups and classifica-tion of participants according to the affect measures as opposed to the brainasymmetry measures would also be useful in future studies of brainactivity
A number of other possible explanations for the failure to reg nd con-sistent decreases in parieto-temporal activity during depression have beenproposed (Heller 1990 1993b) One of these explanations focused on theseverity of the depression This supposition was based on the observationthat posterior asymmetries were more likely to be reported when depres-sion was relatively severe Heller suggested that under less extremeaffective conditions task-induced activity predominates masking theaffect-induced activity More recently Heller et al (1995) suggestedthat severity per se is not the issue rather because the comorbiditywith anxiety decreases as severity of depression increases (Hiller Zau-dig amp Rose 1989) the opposing effects of anxiety on posterior brainactivity diminishes
Heller (1993b) also suggested that the inconsistencies in the depressionand anxiety literatures may be due to signireg cant differences in regionalbrain activity that covary more strongly with dispositional characteristicsassociated with emotional functioning (eg extraversion) than with situa-tionally induced emotions Several studies have reported perceptual(Berenbaum amp Williams 1994 Heller amp Nisenson 1993) and EEG(Nitschke et al 1994) asymmetries for extraversion Nitschke et al alsofound that extraversion moderated the relationship between anxiety andEEG activity across both frontal central and parietal regions For anxiousparticipants high extraversion was associated with less activity (morealpha) whereas for nonanxious (and nondepressed) controls high extra-version was associated with more activity (less alpha) These two sets ofreg ndings albeit preliminary provide evidence for the need to acknowledgeregional differences in brain function for personality characteristics suchas extraversion (for a review see Nitschke Heller amp Miller in prep) andpoint to the possible signireg cance of personality in understanding regionalbrain activity in affective and anxiety disorders
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 431
Subtypes of Depression and Anxiety
Another promising avenue for elucidating the discrepancies reported in theliterature for both depression and anxiety is the analysis of subtypes
Depression Many distinctions for depression have been proposed overthe years including most notably bipolar versus unipolar endogenousversus exogenous and neurotic versus reactive The current Diagnostic andstatistical manual of mental disordersplusmn fourth edition (DSMIV APA 1994)subdivides unipolar depression into major depressive disorder and dysthy-mic disorder and provides further specireg city with criteria for severity andfor psychotic melancholic catatonic or atypical features
Perhaps the most useful distinction in physiological research on depres-sion has been the comparison of melancholic to nonmelancholic depres-sion The dereg ning feature of melancholic depression is anhedonia which isthe inability to experience pleasure Current consensus as indicated by theDSMIV (APA 1994) suggests that melancholia is most strongly charac-terised by a loss of pleasure in all activities and a lack of responsivity topleasurable stimuli Many other symptoms may be shared by both melan-cholic and nonmelancholic depression including depressed mood distur-bances in eating and sleeping patterns low self-esteem inability toconcentrate psychomotor retardation fatigue indecisiveness feelings ofguilt and thoughts about death There is current controversy over whethermelancholic depression is qualitatively different from nonmelancholicdepression with some authors arguing that the primary distinction is oneof severity (Zimmerman Coryell amp Pfohl 1986) Nonetheless neuropsy-chological and neurophysiological studies have been consistent in demon-strating differences between melancholic and nonmelancholic depression
In a comprehensive series of studies Bruder has shown that perceptualasymmetries differ for different subtypes of depression (for a review seeBruder 1995) On a perceptual asymmetry task melancholic patientsshowed an abnormally large right ear (left hemisphere) advantage for adichotic syllables task but no left ear (right hemisphere) advantage for acomplex tone task (Bruder 1995) The lack of a left ear advantage was dueto poor left ear accuracy consistent with other studies demonstrating righthemisphere dereg cits in depressed people However nonmelancholicdepressed patients did not differ from normal controls on either test Alongthese lines in a PET study of melancholic depression Bench et al (1992)reported less activity in the left dorsolateral prefrontal cortex compared tonondepressed controls Note that these data could be consistent with eithera qualitative or a quantitative distinction between melancholic and non-melancholic depression However other data may be more consistent witha qualitative difference between the two types of depression For example
432 HELLER AND NITSCHKE
Davidson Larson and Abercrombie (1995) reported that depression sever-ity in melancholic subjects was predicted by the extent to which activationwas decreased in the left lateral prefrontal region In contrast amongnonmelancholics depression severity was predicted by the extent to whichactivation was increased in the right medial prefrontal area
Davidson et alrsquo s (1995) data have implications for theories regardingthe relationship between pleasantunpleasant affect and anterior regions ofthe brain Research on the psychological characteristics of depression andanxiety has suggested that lack of positive affect (or more preciselyactivated pleasant affect Larsen amp Diener 1992) is unique to depression(Clark amp Watson 1991) Indeed in designing a questionnaire to measureboth general and specireg c components of depression and anxiety Watsonand colleagues (Watson 1995ab) developed a specireg c depression scalethat primarily measures anhedonia Because anhedonia is a fundamentalcharacteristic of melancholia it seems reasonable to assume that decreasedpleasant affect would accompany this type of depression If so Hellerrsquo smodel which predicts less left than right anterior activity for decreasedpleasant affect is completely consistent with the available data
However Davidson et alrsquo s (1995) data suggests the possibility thataffective valence may depend less on the relative balance of activationbetween the left and right anterior regions and more on specireg c systemsassociated with each hemisphere Whereas melancholic depression ischaracterised by decreased pleasant affect depression without melan-cholic features may be better characterised by increased unpleasantaffect Following this line of reasoning Hellerrsquo s model which in itscurrent form argues for a link between anterior asymmetry and the valencedimension would not distinguish between melancholic and nonmelan-cholic depression If affective valence depends more on specireg c systemsassociated with each hemisphere as Davidson et alrsquo s data suggest Hellerrsquo smodel may need to be expanded to remacr ect the unique contributions of theleft and right hemispheres to pleasantunpleasant affect For example itmight be more accurate to link the left anterior region with the activatedpleasant affect (ie positive affect) octant of the circumplex model ofemotion and the right anterior region with the activated unpleasant affect(ie negative affect) octant (see Larson amp Diener 1992)
Because few studies have specireg cally examined the full range of poten-tial affective descriptors in relation to patterns of brain activity theseissues remain to be resolved In most research only isolated portions ofthe circumplex have been sampled For example because positive andnegative affect as measured by the PANAS provide information regardingonly two of the four quadrants in the circumplex model studies that do notcollect data regarding the other two quadrants cannot shed light on howwell they might have accounted for patterns of brain activity
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 433
Anxiety As reviewed earlier studies that examine regional brainactivity in anxiety are even more plagued by inconsistencies than studiesof depression In keeping with the theme of this article we have argued thatby decomposing anxiety into subtypes on the basis of the psychologicalliterature we can potentially resolve some of these discrepancies
In previous work we have suggested that it may be important toconsider the type of anxiety that is being examined (Heller et al 19951997a) In particular we argued that regional brain activity might bedifferent for anxious arousal (eg panic) and anxious apprehension (egworry) types of anxiety that appear to differ signireg cantly in psychologicaland physiological characteristics (eg Barlow 1988 1991 Eysenck 1957Klein 1987 Watson et al 1995b) Anxious arousal has been described byWatson et al as being distinguished by symptoms of physiological hyper-arousal and somatic tension Panic attacks and high-stress states would fallinto this category In contrast anxious apprehension involves worry and ischaracterised by verbal rumination typically about future events This typeof anxiety would be characteristic of obsessive-compulsiveness general-ised anxiety states and trait anxiety (as identireg ed by self-reports of anxiousapprehension and worry on various questionnaires) As noted by Heller etal (1997a) the two kinds of anxiety are not mutually exclusive Extremedegrees of anxious apprehension (eg being worried or fearful about thefuture) may prompt experiences of anxious arousal (eg somatic symptomsand exaggerated physiological responses to stressful events)
We have argued that when the literature on brain activity in anxiety isreviewed in the light of this distinction between anxious arousal andanxious apprehension most of the studies reporting greater right hemi-sphere activity have either studied panic attacks (eg Reiman et al 1984)or high-stress situations likely to entail anxious arousal (eg Tucker et al1977) On the other hand studies indicating greater left hemisphere activity(primarily anterior) have examined OCD (eg Baxter et al 1987) GAD(Wu et al 1991) or questionnaire-dereg ned trait anxiety indexing anxiousapprehension (eg Tucker et al 1978) Indeed as mentioned earlierstudies have highlighted various regions in both hemispheres As aresult it is difreg cult to pinpoint a particular region that is clearly a focalpoint in anxiety Nonetheless the data do suggest that right hemisphereregions emerge as particularly salient in some types of anxiety (ie anxiousarousal) whereas left hemisphere regions emerge as particularly salient inother types of anxiety (ie anxious apprehension)
In a direct test of this hypothesis (Heller et al 1997a) we selectedparticipants on the basis of self-reported trait anxiety using a measureindexing anxious apprehension in a between-subjects design We thenmanipulated anxious arousal on a within-subjects basis Specireg cally wecontrasted brain activity (EEG alpha) during rest periods to brain activity
434 HELLER AND NITSCHKE
during a task in which participants listened to emotional narrativesdesigned to elicit anxious arousal (ie fear and sad narratives that hadpreviously received unpleasant valence and high arousal ratings)
For frontal activity group differences were unaffected by the narrativesAnxious participants showed a larger asymmetry in favour of the lefthemisphere than controls during the rest and listen periods of all narrativetypes as well as for eight minutes of resting baseline preceding thenarrative task Examination of the data indicated that the increased asym-metry was primarily due to a decrease in right hemisphere activity for theanxious participants Nonetheless the increase in magnitude of the asym-metry favouring the left hemisphere found for anxious participants isconsistent with the studies reviewed earlier that reported or inferredrelatively greater left than right anterior activity in anxious subjectsclassireg ed on the basis of worry or trait anxiety (eg Carter et al 1986Tucker et al 1978 Tyler amp Tucker 1982) Our frontal data suggest that itwill be important to investigate whether anxious apprehension is bestcharacterised by the magnitude of the asymmetry favouring the left hemi-sphere or by the level of left hemisphere activity in isolation
In contrast the parietal data supported our prediction regarding theassociation of anxious arousal with greater right hemisphere activityWhen listening to the fear and sad narratives anxious participants showeda selective increase in right parietal activity that was not demonstrated bycontrols The right parietal effect however was not obtained for neutraland happy narratives that had previously received pleasant valence and lowarousal ratings These reg ndings are consistent with those studies that havereported or inferred greater right hemisphere activity during anxious statescharacterised by panic and stress-inducing conditions (eg Reiman et al1984 Stewart et al 1988 Tucker et al 1977)
An important question here has to do with the discrepancies in frontalasymmetries that have been reported Like others (eg Baxter et al 1987Gur et al 1988 Wu et al 1991) we did not reg nd heightened right anterioractivity in anxiety However a number of other studies have (eg Davidsonet al 1996 Reivich et al 1983 Tomarken amp Davidson 1994) Asdiscussed earlier we have previously argued that anterior regions areuniquely involved in the valence dimension of emotion with pleasantaffect associated with greater left activity and unpleasant affect associatedwith greater right activity (eg Heller 1990) Therefore we would expectheightened right anterior activity in an anxious group only if they differedfrom controls on measures indexing affective valence In our recent studythe nondepressed anxious group did not differ from controls on either theGBI Dysthymia scale or the PANAS PA and NA scales
It may also be important to measure the level of anxious apprehension inthe sample studied For example Tomarken and Davidson (1994) found
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 435
that with Marloweplusmn Crowne scores controlled trait anxiety failed to predictfrontal asymmetry However their groups were divided according to amedian split
In summary our results directly support the proposed distinctionbetween anxious apprehension and anxious arousal as types of anxietythat may account for signireg cant variance in patterns of brain activityThey suggest furthermore that Hellerrsquo s model should be rereg ned to reflectthe likelihood that parietal activation will covary with anxious arousal butnot with anxious apprehension If these distinctions are accurate we wouldexpect to observe increases in right parietotemporal activity only underconditions when anxiety is accompanied by increased arousal
Thus in homodynamic and EEG studies a failure to reg nd increased rightposterior activity in anxiety is not surprising as variations in anxiousarousal are likely to be large Some individuals may experience anxiousarousal frequently especially those with disorders more characteristic ofanxious arousal such as panic disorder phobias or post-traumatic stressdisorder Others although apprehensive may rarely experience anxiousarousal In addition the situation may vary in the degree to which it elicitsanxious arousal Being forced to give a speech is likely to induce a greatdeal of anxious arousal in social phobics thus we would have predicted asignireg cant increase in right parietal activity precisely as found by David-son et al (submitted) Note that this effect was only found during theAnticipation condition not during baseline Planning or after thespeech During the Anticipation condition subjects listened while a voiceannounced every 30 seconds for three minutes the time remaining beforethey would be given a topic and allowed to plan for their speech Pre-sumably this period is the most stressful and the most likely to result inanxious arousal Indeed the correlation between heart rate acceleration tothe Anticipation condition and the increase in state anxiety was signireg cant
More recently Nitschke Heller Imig and Miller (submitted) haveattempted to identify anxious arousal and anxious apprehension usingquestionnaires Data from 783 students for a battery of anxiety and depres-sion questionnaires provided strong support for the anxiety distinctionproposed here at least on a psychological level of analysis The Moodand Anxiety Symptom Questionnaire-Anxious Arousal scale (MASQWatson et al 1995ab) was chosen as a measure of anxious arousalwhereas the Penn State Worry Questionnaire (PSWQ Meyer MillerMetzger amp Borkovec 1990 Molina amp Borkovec 1994) was included asa measure of anxious apprehension The correlation between the two scales(r = 20) was much lower than those involving the other more generalscales of anxiety included in the study (MASQ-General Anxiety scaleBeck Anxiety Inventory and the State-Trait Anxiety Inventory-Traitscale ranging from r = 47 to r = 65) These results further support the
436 HELLER AND NITSCHKE
argument that anxious arousal and anxious apprehension are distinctanxiety states
However new questions are highlighted by this distinction betweenanxious arousal and anxious apprehension particularly in the light ofHellerrsquo s model If anxious apprehension is associated with left anterioractivity the overall pattern of regional brain activity appears to be identicalto the brain activity associated with pleasant emotional states (ie more leftthan right anterior activity) However anxious apprehension might beexpected to fall in the portion of the circumplex model of emotionassociated with activated unpleasant affect (as suggested by adjectivessuch as fearful nervous jittery anxious) Below we suggest a numberof theoretical possibilities for resolving this apparent discrepancy all ofwhich remain to be examined empirically
In general it will be important to clarify the degree to which anxiousapprehension is in fact characterised by unpleasant versus pleasant valenceas well as high versus low arousal In addition it may be helpful toconsider these anterior asymmetries from the approach-withdrawal frame-work employed by Davidson (1992ab Davidson amp Tomarken 1989)Under conditions where anxious apprehension leads to approach tenden-cies one might expect an asymmetry in favour of the left anterior region Incontrast when anxious apprehension is associated with withdrawal tenden-cies (perhaps as anxious arousal increases) an asymmetry may emerge infavour of the right anterior region Conceivably the degree to whichanxious apprehension is characterised by approach versus withdrawaltendencies would depend on the degree of stress and discomfortinvolved the personrsquo s coping style and other factors With regard tovalence we would predict that anxious apprehension characterised byapproach tendencies would be accompanied by relatively pleasant affectAnxious apprehension accompanied by withdrawal tendencies in contrastwould be accompanied by relatively unpleasant affect
It is also possible that instead of a direct link between anxious appre-hension and approach behaviour they share more fundamental character-istics associated with left anterior functions For example one might arguethat they both involve sequential thinking and the development of hier-archically structured plans focusing on future concerns (for a review ofrelevant left anterior functions see Tomarken amp Keener this Issue)
Another important issue has to do with the nature of the asymmetricpattern of activity As reviewed earlier we found that the asymmetry infavour of the left hemisphere for participants classireg ed on the basis of traitanxiety was due to a relative decrease in right hemisphere activity Thusthe left frontal region of the anxious participants was not signireg cantly moreactive than that of controls It will therefore be important to examine thepossibility that absolute activity of the left anterior region plays a more
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 437
important role in the modulation of valence (eg Davidson et al 1995)than of anxious apprehension
Another possibility is that independent left anterior systems are asso-ciated with anxious apprehension versus pleasant affect Liotti and Tucker(1995) review theories that there may be multiple functionally segregatedparallel frontal cortico-striato-thalamic circuits some associated withmotor functions some with set shifting and behavioural inhibition andsome with emotional information processing and hedonic responsivity tostimuli Possibly one of these circuits remacr ects activation associated withpleasant affect and another remacr ects activation associated with anxiousapprehension andor approach Different neurotransmitter systems couldbe involved More precise methods of localisation would therefore benecessary to distinguish the unique brain activity associated with each ofthese psychologicalbehavioural phenomena
Comorbidity of Depression and Anxiety
In considering the issue of decomposing emotional states with regard to thesymptomatology in affective and anxiety disorders we have previouslyargued that issues of comorbidity of depression and anxiety are crucial(Heller et al 1995) A review of the literature reveals that rates of comor-bidity are extremely high (Alloy Kelly Mineka amp Clements 1990 Helleret al 1995 Hiller et al 1989 Katon amp Roy-Byrne 1991) Even in our veryrecent work (Nitschke et al submitted) using the MASQ a measuredesigned by Watson and colleagues to avoid symptom overlap in themeasurement of anxiety and depression (Watson et al 1995ab) comor-bidity was fairly pronounced in a large college student sample The MASQAnxious Arousal (AA) and Anhedonic Depression (AD) scales were used asmeasures of those constructs and the PSWQ provided the measure ofanxious apprehension Of the 125 (out of 783) participants scoring abovethe 80th percentile on at least one of the three scales of interest and belowthe 50th percentile on any where the 80th percentile cut-off was not metfully 50 demonstrated comorbidity by scoring above the 80th percentileon at least two of the scales This rate of comorbidity emerged despite thelow correlations found between the three scales (MASQ AA and MASQAD r = 20 MASQ AA and PSWQ r = 20 MASQ AD and PSWQ r =36) relative to the correlations found for the other more general measuresof depression including the MASQ General Mixed General Depression andGeneral Anxiety scales Beck Depression Inventory Beck Anxiety Inven-tory General Behaviour Inventory-Dysthymia scale and the State-TraitAnxiety Inventory-Trait scale (ranging from r = 47 to r = 76)
Because at least some studies have found opposite patterns of activityfor the right posterior region during depression versus during anxiety (ie
438 HELLER AND NITSCHKE
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
studies depression has been associated with more right than left anterioractivity Numerous positron emission tomography (PET) studies have alsoreported less left anterior activity (eg Baxter et al 1985 1989 Bench etal 1992 Bench Friston Brown Frackowiak amp Dolan 1993 George etal 1994 Martinot et al 1990) which is consistent with the EEG reg ndingsearlier However other studies have reported reg ndings in the oppositedirection for depression (Drevets et al 1992) and sad mood (George etal 1995) Null effects for depression have also been found (eg NitschkeHeller Etienne amp Miller 1995 Tomarken amp Davidson 1994)
The reg ndings for the parieto-temporal regons have been even morevariable In EEG and blood macr ow studies that report positive reg ndingsdepression is typically associated with less right posterior activity (egFlor-Henry 1979 Henriques amp Davidson 1990 Post et al 1987 Uytden-hoef et al 1983) Numerous studies however report no effects for thisregion of the brain (for a review see Davidson amp Tomarken 1989)
Reports of reduced right parieto-temporal activity although not alwayspresent are consistent with neuropsychological studies often indicatingimpaired or suppressed right hemisphere function in depressed peopleOn neuropsychological tasks thought to depend on right parietotemporalfunctioning depressed people have often been found to display decrementsin performance (eg Berndt amp Berndt 1980 Flor-Henry 1976 GoldsteinFilskov Weaver amp Ives 1977 Gruzelier Seymour Wilson Jolley ampHirsch 1988 Kronfol Hamsher Digre amp Waziri 1978) Similar findingsfor sad mood in nondepressed people have also been reported for righthemisphere tasks (Tucker Stenslie Roth amp Shearer 1981)
Studies that have directed information to one or the other hemisphereusing lateralised paradigms such as dichotic listening or visual half-reg eldpresentation have also found specireg c dereg cits for the right hemisphere indepressed people (for a review see Bruder 1995) Left-ear decrements orabsent left-ear (right hemisphere) advantages have been reported fordepression on nonverbal dichotic listening tasks (eg Bruder et al1989) Similarly left visual reg eld (right hemisphere) dereg cits in reactiontime and accuracy for nonverbal stimuli have been reported for depressedpeople (Bruder et al 1989 Liotti Sava Rizzolati amp Caffarra 1991)Comparable results have also been obtained in studies of nondepressedpeople who undergo mood inductions (Banich Stolar Heller amp Goldman1992 Ladavas Nicoletti Umilta amp Rizzolatti 1984) Reaction time andaccuracy for the right hemisphere are selectively impaired in sad moods
On the Chimeric Faces Task (CFT) a free-vision task of face processingthat typically elicits a left hemispatial bias suggesting greater right hemi-sphere activation Jaeger Borod and Peselow (1987) found smaller lefthemispatial biases for depressed patients than nonpsychiatric controlsUsing the same task Heller Etienne and Miller (1995) reported that
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 427
high-depressed participants had smaller left hemispatial biases than low-depressed participants Due to the nature of the CFT hemispatial biases arenonspecireg c as to whether they remacr ect a decrease in activity for the con-tralateral hemisphere or an increase in activity for the ipsilateral hemi-sphere (ie a reduction in the magnitude of the left hemispatial bias couldremacr ect either reduced right hemisphere or increased left hemisphere activ-ity) However based on the evidence (reviewed earlier) for selectivedereg cits in performance on right hemisphere tasks relatively poor perfor-mance for the left-visual reg eld on tachistoscopic tasks and reduced EEGactivity for right posterior regions the most likely interpretation of the CFTreg ndings is that they remacr ect diminished right posterior activity
In summary although anterior asymmetries have often been reported indepression the more common reg nding of higher left than right activation isnot invariably found Furthermore there is a signireg cant discrepancybetween the number of studies that have reported parieto-temporal asym-metries in activation which have been relatively few and the number ofstudies that have reported dereg cits in right hemisphere information proces-sing which have been many Also worthy of note is whether studies ofclinical depression psychometrically dereg ned depression and sad mood areaddressing the same phenomenon The number of consistencies acrossthese populations in the physiological literature reviewed here suggestssome commonality on the other hand the inconsistencies could be in partdue to the heterogeneity between the samples studied
Anxiety
Research on brain activity in anxious states has also been plagued byinconsistencies Again although anterior and posterior asymmetries havebeen reported no consistent pattern has emerged Studies of multisite EEGand regional cerebral blood macr ow (rCBF) using xenon-133 or PET havereported right hemisphere left hemisphere and bilateral increases inactivity during anxiety (for a review see Heller Nitschke Etienne ampMiller 1997a) Other studies have found no differences in brain activity(Gur et al 1988 Mountz et al 1989 Nordhal et al 1990 Tomarken ampDavidson 1994)
Having comprehensively reviewed this literature elsewhere (Heller etal 1997a) we will provide an overview of those studies reporting differ-ential hemispheric involvement in anxiety A subset of these suggest anasymmetry in favour of right hemisphere activation in various regions(frontal lateral prefrontal precentral frontal anterior temporal parietaloccipital and parahippocampal) Heightened right hemisphere activity hasbeen reported for panic disorder patients (Reiman Raichle Butler Hers-covitch amp Robins 1984) panic disorder patients experiencing a sodium
428 HELLER AND NITSCHKE
lactate-induced panic attack (Stewart Devous Rush Lane amp Bonte1988) generalised anxiety disorder (GAD) patients during an anxiety-stimulating task (Wu et al 1991) psychometrically dereg ned high-anxiousparticipants immediately following arterial and venous catheterisation andagain after completion of the PET scans (Reivich Gur amp Alavi 1983) andsocial phobics prior to making a public speech (Davidson MarshallTomarken amp Henriques submitted) Consistent with this pattern anti-anxiety medication was found to reduce the rate of glucose metabolismin the right hemisphere especially for the occipital and frontal regions forpatients with GAD (Buchsbaum et al 1987) Similarly in a population ofnonpsychiatric volunteers those administered diazepam intravenouslyshowed maximal decreases in rCBF in the right frontal area whereasthose injected with a placebo showed no rCBF changes (Mathew Wil-son amp Daniel 1985)
Regional brain activity can also be inferred from performance on tasksthat depend on particular brain regions On the CFT Heller et al (1995)reported that high-anxious participants had larger left hemispatial biasesthan low-anxious participants suggested to remacr ect greater right posterioractivity Similarly on a neuropsychological test battery panic disorderpatients performed less well than normal controls on left but not righthemisphere tasks (Yeudall et al 1983) Data obtained from a series ofstudies using behavioural measures to examine right versus left hemispherefunction during an experimental condition producing transient (ie state)anxiety (Tucker Antes Stenslie amp Barnhardt 1978 Tucker Roth Arne-son amp Buckingham 1977 Tyler amp Tucker 1982) are also consistent withan interpretation of greater right than left hemisphere activity in anxiety(Heller et al 1995)
However when subjects were classireg ed according to trait anxietyTucker and colleagues (Tucker et al 1978 Tyler amp Tucker 1982)interpreted the results as strongly suggestive of greater left than righthemisphere activity This interpretation is consistent with reports ofnumerous other labs reg nding increased left hemisphere activity associatedwith anxiety in various cortical and subcortical regions (orbital frontalinferior frontal anterior cingulate caudate putamen and thalamus)Heightened left hemisphere activity has been found in obsessive-compul-sive disorder (OCD) patients (Baxter et al 1987 Swedo et al 1989)GAD patients (Buchsbaum et al 1985 Wu et al 1991) and studentsclassireg ed as ` worriersrsquo rsquo on the basis of self-reports when asked to ` worryabout a specireg c topic of personal concernrsquo rsquo (Carter Johnson amp Borkovec1986)
In summary some reg ndings for anxiety indicate increased right hemi-sphere activation whereas others indicate increased left hemisphere acti-vation Indeed it may be the case that they could co-occur which could
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 429
account for the reg ndings of bilateral increases or no asymmetries In themeantime we are lacking a model to explain these discrepancies
ACCOUNTING FOR INCONSISTENCIES INBRAIN ACTIVITY
We are thus confronted with inconsistent results with regard to regionalbrain activity in both depression and anxiety In previous work we haveargued that in order to resolve these discrepancies it is helpful to turn tothe psychological literature on symptomatology in depression and anxietyIn the sections that follow we make some general observations that applyto both depression and anxiety before focusing on each more exclusively
In addressing the inconsistencies for the blood macr ow reg ndings severalconcerns about the literature are worthy of note First several of the mostfrequently cited studies in this literature sampled a restricted set of brainregions (eg Reiman et al 1984 Reivich et al 1983) Furthermore theregions of interest were not consistent from one study to the nextSecond statistical procedures employing multiple t-tests rather thanrepeated-measures ANOVAs are questionable especially with regard todrawing conclusions about hemispheric differences Third although somestudies examined selected diagnostic groups (eg major depressive dis-order OCD panic disorder GAD) most treated depression or anxiety asa unitary construct and did not consider that different types of each mightbe associated with different patterns of brain activity Moreover with theexception of recent OCD studies most did not take into account thedegree of comorbidity with depression or anxiety which may be criticalin determining the patterns of brain activity that occur (Heller et al1995 Tucker 1988)
As emphasised by Davidson Tomarken and colleagues (Davidson1993 Tomarken Davidson Wheeler amp Doss 1992 Tomarken David-son Wheeler amp Kinney 1992) there are a number of methodogicalapproaches that need to be considered when asymmetric patterns ofEEG activity do not emerge Some of these considerations are relevantto the design of blood macr ow studies as well One issue is that of stability(ie the extent to which brain activity remacr ects a stable individual pattern)This can be addressed by obtaining a sufreg cient number of baseline samplesas well as test-retest data For example Tomarken et al (1992) found thatonly participants with a stable pattern of anterior asymmetry across twomeasurement occasions showed an association between left frontalhypoactivation and more dispositional negative affect as measured bythe PANAS
It is also important to note that for the most part EEG studies haveexamined resting or baseline activity Davidson (1993) has pointed out
430 HELLER AND NITSCHKE
that baseline levels of frontal activation asymmetry may remacr ect individualtendencies toward a valenced affective response but this relationship maynot be observed in the absence of a specireg c environmental elicitor Thushe emphasises the need to examine the relationship between individualdifferences in patterns of asymmetry and affective response to emotionallychallenging tasks As applied to the pattern of posterior reg ndings theseconsiderations suggest the possibility that more robust differences inposterior brain activity might be yielded by studies measuring brainactivity while participants perform tasks for which depressed peopleshow decrements Additional methodological approaches highlighted byDavidson and colleagues such as the use of extreme groups and classifica-tion of participants according to the affect measures as opposed to the brainasymmetry measures would also be useful in future studies of brainactivity
A number of other possible explanations for the failure to reg nd con-sistent decreases in parieto-temporal activity during depression have beenproposed (Heller 1990 1993b) One of these explanations focused on theseverity of the depression This supposition was based on the observationthat posterior asymmetries were more likely to be reported when depres-sion was relatively severe Heller suggested that under less extremeaffective conditions task-induced activity predominates masking theaffect-induced activity More recently Heller et al (1995) suggestedthat severity per se is not the issue rather because the comorbiditywith anxiety decreases as severity of depression increases (Hiller Zau-dig amp Rose 1989) the opposing effects of anxiety on posterior brainactivity diminishes
Heller (1993b) also suggested that the inconsistencies in the depressionand anxiety literatures may be due to signireg cant differences in regionalbrain activity that covary more strongly with dispositional characteristicsassociated with emotional functioning (eg extraversion) than with situa-tionally induced emotions Several studies have reported perceptual(Berenbaum amp Williams 1994 Heller amp Nisenson 1993) and EEG(Nitschke et al 1994) asymmetries for extraversion Nitschke et al alsofound that extraversion moderated the relationship between anxiety andEEG activity across both frontal central and parietal regions For anxiousparticipants high extraversion was associated with less activity (morealpha) whereas for nonanxious (and nondepressed) controls high extra-version was associated with more activity (less alpha) These two sets ofreg ndings albeit preliminary provide evidence for the need to acknowledgeregional differences in brain function for personality characteristics suchas extraversion (for a review see Nitschke Heller amp Miller in prep) andpoint to the possible signireg cance of personality in understanding regionalbrain activity in affective and anxiety disorders
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 431
Subtypes of Depression and Anxiety
Another promising avenue for elucidating the discrepancies reported in theliterature for both depression and anxiety is the analysis of subtypes
Depression Many distinctions for depression have been proposed overthe years including most notably bipolar versus unipolar endogenousversus exogenous and neurotic versus reactive The current Diagnostic andstatistical manual of mental disordersplusmn fourth edition (DSMIV APA 1994)subdivides unipolar depression into major depressive disorder and dysthy-mic disorder and provides further specireg city with criteria for severity andfor psychotic melancholic catatonic or atypical features
Perhaps the most useful distinction in physiological research on depres-sion has been the comparison of melancholic to nonmelancholic depres-sion The dereg ning feature of melancholic depression is anhedonia which isthe inability to experience pleasure Current consensus as indicated by theDSMIV (APA 1994) suggests that melancholia is most strongly charac-terised by a loss of pleasure in all activities and a lack of responsivity topleasurable stimuli Many other symptoms may be shared by both melan-cholic and nonmelancholic depression including depressed mood distur-bances in eating and sleeping patterns low self-esteem inability toconcentrate psychomotor retardation fatigue indecisiveness feelings ofguilt and thoughts about death There is current controversy over whethermelancholic depression is qualitatively different from nonmelancholicdepression with some authors arguing that the primary distinction is oneof severity (Zimmerman Coryell amp Pfohl 1986) Nonetheless neuropsy-chological and neurophysiological studies have been consistent in demon-strating differences between melancholic and nonmelancholic depression
In a comprehensive series of studies Bruder has shown that perceptualasymmetries differ for different subtypes of depression (for a review seeBruder 1995) On a perceptual asymmetry task melancholic patientsshowed an abnormally large right ear (left hemisphere) advantage for adichotic syllables task but no left ear (right hemisphere) advantage for acomplex tone task (Bruder 1995) The lack of a left ear advantage was dueto poor left ear accuracy consistent with other studies demonstrating righthemisphere dereg cits in depressed people However nonmelancholicdepressed patients did not differ from normal controls on either test Alongthese lines in a PET study of melancholic depression Bench et al (1992)reported less activity in the left dorsolateral prefrontal cortex compared tonondepressed controls Note that these data could be consistent with eithera qualitative or a quantitative distinction between melancholic and non-melancholic depression However other data may be more consistent witha qualitative difference between the two types of depression For example
432 HELLER AND NITSCHKE
Davidson Larson and Abercrombie (1995) reported that depression sever-ity in melancholic subjects was predicted by the extent to which activationwas decreased in the left lateral prefrontal region In contrast amongnonmelancholics depression severity was predicted by the extent to whichactivation was increased in the right medial prefrontal area
Davidson et alrsquo s (1995) data have implications for theories regardingthe relationship between pleasantunpleasant affect and anterior regions ofthe brain Research on the psychological characteristics of depression andanxiety has suggested that lack of positive affect (or more preciselyactivated pleasant affect Larsen amp Diener 1992) is unique to depression(Clark amp Watson 1991) Indeed in designing a questionnaire to measureboth general and specireg c components of depression and anxiety Watsonand colleagues (Watson 1995ab) developed a specireg c depression scalethat primarily measures anhedonia Because anhedonia is a fundamentalcharacteristic of melancholia it seems reasonable to assume that decreasedpleasant affect would accompany this type of depression If so Hellerrsquo smodel which predicts less left than right anterior activity for decreasedpleasant affect is completely consistent with the available data
However Davidson et alrsquo s (1995) data suggests the possibility thataffective valence may depend less on the relative balance of activationbetween the left and right anterior regions and more on specireg c systemsassociated with each hemisphere Whereas melancholic depression ischaracterised by decreased pleasant affect depression without melan-cholic features may be better characterised by increased unpleasantaffect Following this line of reasoning Hellerrsquo s model which in itscurrent form argues for a link between anterior asymmetry and the valencedimension would not distinguish between melancholic and nonmelan-cholic depression If affective valence depends more on specireg c systemsassociated with each hemisphere as Davidson et alrsquo s data suggest Hellerrsquo smodel may need to be expanded to remacr ect the unique contributions of theleft and right hemispheres to pleasantunpleasant affect For example itmight be more accurate to link the left anterior region with the activatedpleasant affect (ie positive affect) octant of the circumplex model ofemotion and the right anterior region with the activated unpleasant affect(ie negative affect) octant (see Larson amp Diener 1992)
Because few studies have specireg cally examined the full range of poten-tial affective descriptors in relation to patterns of brain activity theseissues remain to be resolved In most research only isolated portions ofthe circumplex have been sampled For example because positive andnegative affect as measured by the PANAS provide information regardingonly two of the four quadrants in the circumplex model studies that do notcollect data regarding the other two quadrants cannot shed light on howwell they might have accounted for patterns of brain activity
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 433
Anxiety As reviewed earlier studies that examine regional brainactivity in anxiety are even more plagued by inconsistencies than studiesof depression In keeping with the theme of this article we have argued thatby decomposing anxiety into subtypes on the basis of the psychologicalliterature we can potentially resolve some of these discrepancies
In previous work we have suggested that it may be important toconsider the type of anxiety that is being examined (Heller et al 19951997a) In particular we argued that regional brain activity might bedifferent for anxious arousal (eg panic) and anxious apprehension (egworry) types of anxiety that appear to differ signireg cantly in psychologicaland physiological characteristics (eg Barlow 1988 1991 Eysenck 1957Klein 1987 Watson et al 1995b) Anxious arousal has been described byWatson et al as being distinguished by symptoms of physiological hyper-arousal and somatic tension Panic attacks and high-stress states would fallinto this category In contrast anxious apprehension involves worry and ischaracterised by verbal rumination typically about future events This typeof anxiety would be characteristic of obsessive-compulsiveness general-ised anxiety states and trait anxiety (as identireg ed by self-reports of anxiousapprehension and worry on various questionnaires) As noted by Heller etal (1997a) the two kinds of anxiety are not mutually exclusive Extremedegrees of anxious apprehension (eg being worried or fearful about thefuture) may prompt experiences of anxious arousal (eg somatic symptomsand exaggerated physiological responses to stressful events)
We have argued that when the literature on brain activity in anxiety isreviewed in the light of this distinction between anxious arousal andanxious apprehension most of the studies reporting greater right hemi-sphere activity have either studied panic attacks (eg Reiman et al 1984)or high-stress situations likely to entail anxious arousal (eg Tucker et al1977) On the other hand studies indicating greater left hemisphere activity(primarily anterior) have examined OCD (eg Baxter et al 1987) GAD(Wu et al 1991) or questionnaire-dereg ned trait anxiety indexing anxiousapprehension (eg Tucker et al 1978) Indeed as mentioned earlierstudies have highlighted various regions in both hemispheres As aresult it is difreg cult to pinpoint a particular region that is clearly a focalpoint in anxiety Nonetheless the data do suggest that right hemisphereregions emerge as particularly salient in some types of anxiety (ie anxiousarousal) whereas left hemisphere regions emerge as particularly salient inother types of anxiety (ie anxious apprehension)
In a direct test of this hypothesis (Heller et al 1997a) we selectedparticipants on the basis of self-reported trait anxiety using a measureindexing anxious apprehension in a between-subjects design We thenmanipulated anxious arousal on a within-subjects basis Specireg cally wecontrasted brain activity (EEG alpha) during rest periods to brain activity
434 HELLER AND NITSCHKE
during a task in which participants listened to emotional narrativesdesigned to elicit anxious arousal (ie fear and sad narratives that hadpreviously received unpleasant valence and high arousal ratings)
For frontal activity group differences were unaffected by the narrativesAnxious participants showed a larger asymmetry in favour of the lefthemisphere than controls during the rest and listen periods of all narrativetypes as well as for eight minutes of resting baseline preceding thenarrative task Examination of the data indicated that the increased asym-metry was primarily due to a decrease in right hemisphere activity for theanxious participants Nonetheless the increase in magnitude of the asym-metry favouring the left hemisphere found for anxious participants isconsistent with the studies reviewed earlier that reported or inferredrelatively greater left than right anterior activity in anxious subjectsclassireg ed on the basis of worry or trait anxiety (eg Carter et al 1986Tucker et al 1978 Tyler amp Tucker 1982) Our frontal data suggest that itwill be important to investigate whether anxious apprehension is bestcharacterised by the magnitude of the asymmetry favouring the left hemi-sphere or by the level of left hemisphere activity in isolation
In contrast the parietal data supported our prediction regarding theassociation of anxious arousal with greater right hemisphere activityWhen listening to the fear and sad narratives anxious participants showeda selective increase in right parietal activity that was not demonstrated bycontrols The right parietal effect however was not obtained for neutraland happy narratives that had previously received pleasant valence and lowarousal ratings These reg ndings are consistent with those studies that havereported or inferred greater right hemisphere activity during anxious statescharacterised by panic and stress-inducing conditions (eg Reiman et al1984 Stewart et al 1988 Tucker et al 1977)
An important question here has to do with the discrepancies in frontalasymmetries that have been reported Like others (eg Baxter et al 1987Gur et al 1988 Wu et al 1991) we did not reg nd heightened right anterioractivity in anxiety However a number of other studies have (eg Davidsonet al 1996 Reivich et al 1983 Tomarken amp Davidson 1994) Asdiscussed earlier we have previously argued that anterior regions areuniquely involved in the valence dimension of emotion with pleasantaffect associated with greater left activity and unpleasant affect associatedwith greater right activity (eg Heller 1990) Therefore we would expectheightened right anterior activity in an anxious group only if they differedfrom controls on measures indexing affective valence In our recent studythe nondepressed anxious group did not differ from controls on either theGBI Dysthymia scale or the PANAS PA and NA scales
It may also be important to measure the level of anxious apprehension inthe sample studied For example Tomarken and Davidson (1994) found
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 435
that with Marloweplusmn Crowne scores controlled trait anxiety failed to predictfrontal asymmetry However their groups were divided according to amedian split
In summary our results directly support the proposed distinctionbetween anxious apprehension and anxious arousal as types of anxietythat may account for signireg cant variance in patterns of brain activityThey suggest furthermore that Hellerrsquo s model should be rereg ned to reflectthe likelihood that parietal activation will covary with anxious arousal butnot with anxious apprehension If these distinctions are accurate we wouldexpect to observe increases in right parietotemporal activity only underconditions when anxiety is accompanied by increased arousal
Thus in homodynamic and EEG studies a failure to reg nd increased rightposterior activity in anxiety is not surprising as variations in anxiousarousal are likely to be large Some individuals may experience anxiousarousal frequently especially those with disorders more characteristic ofanxious arousal such as panic disorder phobias or post-traumatic stressdisorder Others although apprehensive may rarely experience anxiousarousal In addition the situation may vary in the degree to which it elicitsanxious arousal Being forced to give a speech is likely to induce a greatdeal of anxious arousal in social phobics thus we would have predicted asignireg cant increase in right parietal activity precisely as found by David-son et al (submitted) Note that this effect was only found during theAnticipation condition not during baseline Planning or after thespeech During the Anticipation condition subjects listened while a voiceannounced every 30 seconds for three minutes the time remaining beforethey would be given a topic and allowed to plan for their speech Pre-sumably this period is the most stressful and the most likely to result inanxious arousal Indeed the correlation between heart rate acceleration tothe Anticipation condition and the increase in state anxiety was signireg cant
More recently Nitschke Heller Imig and Miller (submitted) haveattempted to identify anxious arousal and anxious apprehension usingquestionnaires Data from 783 students for a battery of anxiety and depres-sion questionnaires provided strong support for the anxiety distinctionproposed here at least on a psychological level of analysis The Moodand Anxiety Symptom Questionnaire-Anxious Arousal scale (MASQWatson et al 1995ab) was chosen as a measure of anxious arousalwhereas the Penn State Worry Questionnaire (PSWQ Meyer MillerMetzger amp Borkovec 1990 Molina amp Borkovec 1994) was included asa measure of anxious apprehension The correlation between the two scales(r = 20) was much lower than those involving the other more generalscales of anxiety included in the study (MASQ-General Anxiety scaleBeck Anxiety Inventory and the State-Trait Anxiety Inventory-Traitscale ranging from r = 47 to r = 65) These results further support the
436 HELLER AND NITSCHKE
argument that anxious arousal and anxious apprehension are distinctanxiety states
However new questions are highlighted by this distinction betweenanxious arousal and anxious apprehension particularly in the light ofHellerrsquo s model If anxious apprehension is associated with left anterioractivity the overall pattern of regional brain activity appears to be identicalto the brain activity associated with pleasant emotional states (ie more leftthan right anterior activity) However anxious apprehension might beexpected to fall in the portion of the circumplex model of emotionassociated with activated unpleasant affect (as suggested by adjectivessuch as fearful nervous jittery anxious) Below we suggest a numberof theoretical possibilities for resolving this apparent discrepancy all ofwhich remain to be examined empirically
In general it will be important to clarify the degree to which anxiousapprehension is in fact characterised by unpleasant versus pleasant valenceas well as high versus low arousal In addition it may be helpful toconsider these anterior asymmetries from the approach-withdrawal frame-work employed by Davidson (1992ab Davidson amp Tomarken 1989)Under conditions where anxious apprehension leads to approach tenden-cies one might expect an asymmetry in favour of the left anterior region Incontrast when anxious apprehension is associated with withdrawal tenden-cies (perhaps as anxious arousal increases) an asymmetry may emerge infavour of the right anterior region Conceivably the degree to whichanxious apprehension is characterised by approach versus withdrawaltendencies would depend on the degree of stress and discomfortinvolved the personrsquo s coping style and other factors With regard tovalence we would predict that anxious apprehension characterised byapproach tendencies would be accompanied by relatively pleasant affectAnxious apprehension accompanied by withdrawal tendencies in contrastwould be accompanied by relatively unpleasant affect
It is also possible that instead of a direct link between anxious appre-hension and approach behaviour they share more fundamental character-istics associated with left anterior functions For example one might arguethat they both involve sequential thinking and the development of hier-archically structured plans focusing on future concerns (for a review ofrelevant left anterior functions see Tomarken amp Keener this Issue)
Another important issue has to do with the nature of the asymmetricpattern of activity As reviewed earlier we found that the asymmetry infavour of the left hemisphere for participants classireg ed on the basis of traitanxiety was due to a relative decrease in right hemisphere activity Thusthe left frontal region of the anxious participants was not signireg cantly moreactive than that of controls It will therefore be important to examine thepossibility that absolute activity of the left anterior region plays a more
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 437
important role in the modulation of valence (eg Davidson et al 1995)than of anxious apprehension
Another possibility is that independent left anterior systems are asso-ciated with anxious apprehension versus pleasant affect Liotti and Tucker(1995) review theories that there may be multiple functionally segregatedparallel frontal cortico-striato-thalamic circuits some associated withmotor functions some with set shifting and behavioural inhibition andsome with emotional information processing and hedonic responsivity tostimuli Possibly one of these circuits remacr ects activation associated withpleasant affect and another remacr ects activation associated with anxiousapprehension andor approach Different neurotransmitter systems couldbe involved More precise methods of localisation would therefore benecessary to distinguish the unique brain activity associated with each ofthese psychologicalbehavioural phenomena
Comorbidity of Depression and Anxiety
In considering the issue of decomposing emotional states with regard to thesymptomatology in affective and anxiety disorders we have previouslyargued that issues of comorbidity of depression and anxiety are crucial(Heller et al 1995) A review of the literature reveals that rates of comor-bidity are extremely high (Alloy Kelly Mineka amp Clements 1990 Helleret al 1995 Hiller et al 1989 Katon amp Roy-Byrne 1991) Even in our veryrecent work (Nitschke et al submitted) using the MASQ a measuredesigned by Watson and colleagues to avoid symptom overlap in themeasurement of anxiety and depression (Watson et al 1995ab) comor-bidity was fairly pronounced in a large college student sample The MASQAnxious Arousal (AA) and Anhedonic Depression (AD) scales were used asmeasures of those constructs and the PSWQ provided the measure ofanxious apprehension Of the 125 (out of 783) participants scoring abovethe 80th percentile on at least one of the three scales of interest and belowthe 50th percentile on any where the 80th percentile cut-off was not metfully 50 demonstrated comorbidity by scoring above the 80th percentileon at least two of the scales This rate of comorbidity emerged despite thelow correlations found between the three scales (MASQ AA and MASQAD r = 20 MASQ AA and PSWQ r = 20 MASQ AD and PSWQ r =36) relative to the correlations found for the other more general measuresof depression including the MASQ General Mixed General Depression andGeneral Anxiety scales Beck Depression Inventory Beck Anxiety Inven-tory General Behaviour Inventory-Dysthymia scale and the State-TraitAnxiety Inventory-Trait scale (ranging from r = 47 to r = 76)
Because at least some studies have found opposite patterns of activityfor the right posterior region during depression versus during anxiety (ie
438 HELLER AND NITSCHKE
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
high-depressed participants had smaller left hemispatial biases than low-depressed participants Due to the nature of the CFT hemispatial biases arenonspecireg c as to whether they remacr ect a decrease in activity for the con-tralateral hemisphere or an increase in activity for the ipsilateral hemi-sphere (ie a reduction in the magnitude of the left hemispatial bias couldremacr ect either reduced right hemisphere or increased left hemisphere activ-ity) However based on the evidence (reviewed earlier) for selectivedereg cits in performance on right hemisphere tasks relatively poor perfor-mance for the left-visual reg eld on tachistoscopic tasks and reduced EEGactivity for right posterior regions the most likely interpretation of the CFTreg ndings is that they remacr ect diminished right posterior activity
In summary although anterior asymmetries have often been reported indepression the more common reg nding of higher left than right activation isnot invariably found Furthermore there is a signireg cant discrepancybetween the number of studies that have reported parieto-temporal asym-metries in activation which have been relatively few and the number ofstudies that have reported dereg cits in right hemisphere information proces-sing which have been many Also worthy of note is whether studies ofclinical depression psychometrically dereg ned depression and sad mood areaddressing the same phenomenon The number of consistencies acrossthese populations in the physiological literature reviewed here suggestssome commonality on the other hand the inconsistencies could be in partdue to the heterogeneity between the samples studied
Anxiety
Research on brain activity in anxious states has also been plagued byinconsistencies Again although anterior and posterior asymmetries havebeen reported no consistent pattern has emerged Studies of multisite EEGand regional cerebral blood macr ow (rCBF) using xenon-133 or PET havereported right hemisphere left hemisphere and bilateral increases inactivity during anxiety (for a review see Heller Nitschke Etienne ampMiller 1997a) Other studies have found no differences in brain activity(Gur et al 1988 Mountz et al 1989 Nordhal et al 1990 Tomarken ampDavidson 1994)
Having comprehensively reviewed this literature elsewhere (Heller etal 1997a) we will provide an overview of those studies reporting differ-ential hemispheric involvement in anxiety A subset of these suggest anasymmetry in favour of right hemisphere activation in various regions(frontal lateral prefrontal precentral frontal anterior temporal parietaloccipital and parahippocampal) Heightened right hemisphere activity hasbeen reported for panic disorder patients (Reiman Raichle Butler Hers-covitch amp Robins 1984) panic disorder patients experiencing a sodium
428 HELLER AND NITSCHKE
lactate-induced panic attack (Stewart Devous Rush Lane amp Bonte1988) generalised anxiety disorder (GAD) patients during an anxiety-stimulating task (Wu et al 1991) psychometrically dereg ned high-anxiousparticipants immediately following arterial and venous catheterisation andagain after completion of the PET scans (Reivich Gur amp Alavi 1983) andsocial phobics prior to making a public speech (Davidson MarshallTomarken amp Henriques submitted) Consistent with this pattern anti-anxiety medication was found to reduce the rate of glucose metabolismin the right hemisphere especially for the occipital and frontal regions forpatients with GAD (Buchsbaum et al 1987) Similarly in a population ofnonpsychiatric volunteers those administered diazepam intravenouslyshowed maximal decreases in rCBF in the right frontal area whereasthose injected with a placebo showed no rCBF changes (Mathew Wil-son amp Daniel 1985)
Regional brain activity can also be inferred from performance on tasksthat depend on particular brain regions On the CFT Heller et al (1995)reported that high-anxious participants had larger left hemispatial biasesthan low-anxious participants suggested to remacr ect greater right posterioractivity Similarly on a neuropsychological test battery panic disorderpatients performed less well than normal controls on left but not righthemisphere tasks (Yeudall et al 1983) Data obtained from a series ofstudies using behavioural measures to examine right versus left hemispherefunction during an experimental condition producing transient (ie state)anxiety (Tucker Antes Stenslie amp Barnhardt 1978 Tucker Roth Arne-son amp Buckingham 1977 Tyler amp Tucker 1982) are also consistent withan interpretation of greater right than left hemisphere activity in anxiety(Heller et al 1995)
However when subjects were classireg ed according to trait anxietyTucker and colleagues (Tucker et al 1978 Tyler amp Tucker 1982)interpreted the results as strongly suggestive of greater left than righthemisphere activity This interpretation is consistent with reports ofnumerous other labs reg nding increased left hemisphere activity associatedwith anxiety in various cortical and subcortical regions (orbital frontalinferior frontal anterior cingulate caudate putamen and thalamus)Heightened left hemisphere activity has been found in obsessive-compul-sive disorder (OCD) patients (Baxter et al 1987 Swedo et al 1989)GAD patients (Buchsbaum et al 1985 Wu et al 1991) and studentsclassireg ed as ` worriersrsquo rsquo on the basis of self-reports when asked to ` worryabout a specireg c topic of personal concernrsquo rsquo (Carter Johnson amp Borkovec1986)
In summary some reg ndings for anxiety indicate increased right hemi-sphere activation whereas others indicate increased left hemisphere acti-vation Indeed it may be the case that they could co-occur which could
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 429
account for the reg ndings of bilateral increases or no asymmetries In themeantime we are lacking a model to explain these discrepancies
ACCOUNTING FOR INCONSISTENCIES INBRAIN ACTIVITY
We are thus confronted with inconsistent results with regard to regionalbrain activity in both depression and anxiety In previous work we haveargued that in order to resolve these discrepancies it is helpful to turn tothe psychological literature on symptomatology in depression and anxietyIn the sections that follow we make some general observations that applyto both depression and anxiety before focusing on each more exclusively
In addressing the inconsistencies for the blood macr ow reg ndings severalconcerns about the literature are worthy of note First several of the mostfrequently cited studies in this literature sampled a restricted set of brainregions (eg Reiman et al 1984 Reivich et al 1983) Furthermore theregions of interest were not consistent from one study to the nextSecond statistical procedures employing multiple t-tests rather thanrepeated-measures ANOVAs are questionable especially with regard todrawing conclusions about hemispheric differences Third although somestudies examined selected diagnostic groups (eg major depressive dis-order OCD panic disorder GAD) most treated depression or anxiety asa unitary construct and did not consider that different types of each mightbe associated with different patterns of brain activity Moreover with theexception of recent OCD studies most did not take into account thedegree of comorbidity with depression or anxiety which may be criticalin determining the patterns of brain activity that occur (Heller et al1995 Tucker 1988)
As emphasised by Davidson Tomarken and colleagues (Davidson1993 Tomarken Davidson Wheeler amp Doss 1992 Tomarken David-son Wheeler amp Kinney 1992) there are a number of methodogicalapproaches that need to be considered when asymmetric patterns ofEEG activity do not emerge Some of these considerations are relevantto the design of blood macr ow studies as well One issue is that of stability(ie the extent to which brain activity remacr ects a stable individual pattern)This can be addressed by obtaining a sufreg cient number of baseline samplesas well as test-retest data For example Tomarken et al (1992) found thatonly participants with a stable pattern of anterior asymmetry across twomeasurement occasions showed an association between left frontalhypoactivation and more dispositional negative affect as measured bythe PANAS
It is also important to note that for the most part EEG studies haveexamined resting or baseline activity Davidson (1993) has pointed out
430 HELLER AND NITSCHKE
that baseline levels of frontal activation asymmetry may remacr ect individualtendencies toward a valenced affective response but this relationship maynot be observed in the absence of a specireg c environmental elicitor Thushe emphasises the need to examine the relationship between individualdifferences in patterns of asymmetry and affective response to emotionallychallenging tasks As applied to the pattern of posterior reg ndings theseconsiderations suggest the possibility that more robust differences inposterior brain activity might be yielded by studies measuring brainactivity while participants perform tasks for which depressed peopleshow decrements Additional methodological approaches highlighted byDavidson and colleagues such as the use of extreme groups and classifica-tion of participants according to the affect measures as opposed to the brainasymmetry measures would also be useful in future studies of brainactivity
A number of other possible explanations for the failure to reg nd con-sistent decreases in parieto-temporal activity during depression have beenproposed (Heller 1990 1993b) One of these explanations focused on theseverity of the depression This supposition was based on the observationthat posterior asymmetries were more likely to be reported when depres-sion was relatively severe Heller suggested that under less extremeaffective conditions task-induced activity predominates masking theaffect-induced activity More recently Heller et al (1995) suggestedthat severity per se is not the issue rather because the comorbiditywith anxiety decreases as severity of depression increases (Hiller Zau-dig amp Rose 1989) the opposing effects of anxiety on posterior brainactivity diminishes
Heller (1993b) also suggested that the inconsistencies in the depressionand anxiety literatures may be due to signireg cant differences in regionalbrain activity that covary more strongly with dispositional characteristicsassociated with emotional functioning (eg extraversion) than with situa-tionally induced emotions Several studies have reported perceptual(Berenbaum amp Williams 1994 Heller amp Nisenson 1993) and EEG(Nitschke et al 1994) asymmetries for extraversion Nitschke et al alsofound that extraversion moderated the relationship between anxiety andEEG activity across both frontal central and parietal regions For anxiousparticipants high extraversion was associated with less activity (morealpha) whereas for nonanxious (and nondepressed) controls high extra-version was associated with more activity (less alpha) These two sets ofreg ndings albeit preliminary provide evidence for the need to acknowledgeregional differences in brain function for personality characteristics suchas extraversion (for a review see Nitschke Heller amp Miller in prep) andpoint to the possible signireg cance of personality in understanding regionalbrain activity in affective and anxiety disorders
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 431
Subtypes of Depression and Anxiety
Another promising avenue for elucidating the discrepancies reported in theliterature for both depression and anxiety is the analysis of subtypes
Depression Many distinctions for depression have been proposed overthe years including most notably bipolar versus unipolar endogenousversus exogenous and neurotic versus reactive The current Diagnostic andstatistical manual of mental disordersplusmn fourth edition (DSMIV APA 1994)subdivides unipolar depression into major depressive disorder and dysthy-mic disorder and provides further specireg city with criteria for severity andfor psychotic melancholic catatonic or atypical features
Perhaps the most useful distinction in physiological research on depres-sion has been the comparison of melancholic to nonmelancholic depres-sion The dereg ning feature of melancholic depression is anhedonia which isthe inability to experience pleasure Current consensus as indicated by theDSMIV (APA 1994) suggests that melancholia is most strongly charac-terised by a loss of pleasure in all activities and a lack of responsivity topleasurable stimuli Many other symptoms may be shared by both melan-cholic and nonmelancholic depression including depressed mood distur-bances in eating and sleeping patterns low self-esteem inability toconcentrate psychomotor retardation fatigue indecisiveness feelings ofguilt and thoughts about death There is current controversy over whethermelancholic depression is qualitatively different from nonmelancholicdepression with some authors arguing that the primary distinction is oneof severity (Zimmerman Coryell amp Pfohl 1986) Nonetheless neuropsy-chological and neurophysiological studies have been consistent in demon-strating differences between melancholic and nonmelancholic depression
In a comprehensive series of studies Bruder has shown that perceptualasymmetries differ for different subtypes of depression (for a review seeBruder 1995) On a perceptual asymmetry task melancholic patientsshowed an abnormally large right ear (left hemisphere) advantage for adichotic syllables task but no left ear (right hemisphere) advantage for acomplex tone task (Bruder 1995) The lack of a left ear advantage was dueto poor left ear accuracy consistent with other studies demonstrating righthemisphere dereg cits in depressed people However nonmelancholicdepressed patients did not differ from normal controls on either test Alongthese lines in a PET study of melancholic depression Bench et al (1992)reported less activity in the left dorsolateral prefrontal cortex compared tonondepressed controls Note that these data could be consistent with eithera qualitative or a quantitative distinction between melancholic and non-melancholic depression However other data may be more consistent witha qualitative difference between the two types of depression For example
432 HELLER AND NITSCHKE
Davidson Larson and Abercrombie (1995) reported that depression sever-ity in melancholic subjects was predicted by the extent to which activationwas decreased in the left lateral prefrontal region In contrast amongnonmelancholics depression severity was predicted by the extent to whichactivation was increased in the right medial prefrontal area
Davidson et alrsquo s (1995) data have implications for theories regardingthe relationship between pleasantunpleasant affect and anterior regions ofthe brain Research on the psychological characteristics of depression andanxiety has suggested that lack of positive affect (or more preciselyactivated pleasant affect Larsen amp Diener 1992) is unique to depression(Clark amp Watson 1991) Indeed in designing a questionnaire to measureboth general and specireg c components of depression and anxiety Watsonand colleagues (Watson 1995ab) developed a specireg c depression scalethat primarily measures anhedonia Because anhedonia is a fundamentalcharacteristic of melancholia it seems reasonable to assume that decreasedpleasant affect would accompany this type of depression If so Hellerrsquo smodel which predicts less left than right anterior activity for decreasedpleasant affect is completely consistent with the available data
However Davidson et alrsquo s (1995) data suggests the possibility thataffective valence may depend less on the relative balance of activationbetween the left and right anterior regions and more on specireg c systemsassociated with each hemisphere Whereas melancholic depression ischaracterised by decreased pleasant affect depression without melan-cholic features may be better characterised by increased unpleasantaffect Following this line of reasoning Hellerrsquo s model which in itscurrent form argues for a link between anterior asymmetry and the valencedimension would not distinguish between melancholic and nonmelan-cholic depression If affective valence depends more on specireg c systemsassociated with each hemisphere as Davidson et alrsquo s data suggest Hellerrsquo smodel may need to be expanded to remacr ect the unique contributions of theleft and right hemispheres to pleasantunpleasant affect For example itmight be more accurate to link the left anterior region with the activatedpleasant affect (ie positive affect) octant of the circumplex model ofemotion and the right anterior region with the activated unpleasant affect(ie negative affect) octant (see Larson amp Diener 1992)
Because few studies have specireg cally examined the full range of poten-tial affective descriptors in relation to patterns of brain activity theseissues remain to be resolved In most research only isolated portions ofthe circumplex have been sampled For example because positive andnegative affect as measured by the PANAS provide information regardingonly two of the four quadrants in the circumplex model studies that do notcollect data regarding the other two quadrants cannot shed light on howwell they might have accounted for patterns of brain activity
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 433
Anxiety As reviewed earlier studies that examine regional brainactivity in anxiety are even more plagued by inconsistencies than studiesof depression In keeping with the theme of this article we have argued thatby decomposing anxiety into subtypes on the basis of the psychologicalliterature we can potentially resolve some of these discrepancies
In previous work we have suggested that it may be important toconsider the type of anxiety that is being examined (Heller et al 19951997a) In particular we argued that regional brain activity might bedifferent for anxious arousal (eg panic) and anxious apprehension (egworry) types of anxiety that appear to differ signireg cantly in psychologicaland physiological characteristics (eg Barlow 1988 1991 Eysenck 1957Klein 1987 Watson et al 1995b) Anxious arousal has been described byWatson et al as being distinguished by symptoms of physiological hyper-arousal and somatic tension Panic attacks and high-stress states would fallinto this category In contrast anxious apprehension involves worry and ischaracterised by verbal rumination typically about future events This typeof anxiety would be characteristic of obsessive-compulsiveness general-ised anxiety states and trait anxiety (as identireg ed by self-reports of anxiousapprehension and worry on various questionnaires) As noted by Heller etal (1997a) the two kinds of anxiety are not mutually exclusive Extremedegrees of anxious apprehension (eg being worried or fearful about thefuture) may prompt experiences of anxious arousal (eg somatic symptomsand exaggerated physiological responses to stressful events)
We have argued that when the literature on brain activity in anxiety isreviewed in the light of this distinction between anxious arousal andanxious apprehension most of the studies reporting greater right hemi-sphere activity have either studied panic attacks (eg Reiman et al 1984)or high-stress situations likely to entail anxious arousal (eg Tucker et al1977) On the other hand studies indicating greater left hemisphere activity(primarily anterior) have examined OCD (eg Baxter et al 1987) GAD(Wu et al 1991) or questionnaire-dereg ned trait anxiety indexing anxiousapprehension (eg Tucker et al 1978) Indeed as mentioned earlierstudies have highlighted various regions in both hemispheres As aresult it is difreg cult to pinpoint a particular region that is clearly a focalpoint in anxiety Nonetheless the data do suggest that right hemisphereregions emerge as particularly salient in some types of anxiety (ie anxiousarousal) whereas left hemisphere regions emerge as particularly salient inother types of anxiety (ie anxious apprehension)
In a direct test of this hypothesis (Heller et al 1997a) we selectedparticipants on the basis of self-reported trait anxiety using a measureindexing anxious apprehension in a between-subjects design We thenmanipulated anxious arousal on a within-subjects basis Specireg cally wecontrasted brain activity (EEG alpha) during rest periods to brain activity
434 HELLER AND NITSCHKE
during a task in which participants listened to emotional narrativesdesigned to elicit anxious arousal (ie fear and sad narratives that hadpreviously received unpleasant valence and high arousal ratings)
For frontal activity group differences were unaffected by the narrativesAnxious participants showed a larger asymmetry in favour of the lefthemisphere than controls during the rest and listen periods of all narrativetypes as well as for eight minutes of resting baseline preceding thenarrative task Examination of the data indicated that the increased asym-metry was primarily due to a decrease in right hemisphere activity for theanxious participants Nonetheless the increase in magnitude of the asym-metry favouring the left hemisphere found for anxious participants isconsistent with the studies reviewed earlier that reported or inferredrelatively greater left than right anterior activity in anxious subjectsclassireg ed on the basis of worry or trait anxiety (eg Carter et al 1986Tucker et al 1978 Tyler amp Tucker 1982) Our frontal data suggest that itwill be important to investigate whether anxious apprehension is bestcharacterised by the magnitude of the asymmetry favouring the left hemi-sphere or by the level of left hemisphere activity in isolation
In contrast the parietal data supported our prediction regarding theassociation of anxious arousal with greater right hemisphere activityWhen listening to the fear and sad narratives anxious participants showeda selective increase in right parietal activity that was not demonstrated bycontrols The right parietal effect however was not obtained for neutraland happy narratives that had previously received pleasant valence and lowarousal ratings These reg ndings are consistent with those studies that havereported or inferred greater right hemisphere activity during anxious statescharacterised by panic and stress-inducing conditions (eg Reiman et al1984 Stewart et al 1988 Tucker et al 1977)
An important question here has to do with the discrepancies in frontalasymmetries that have been reported Like others (eg Baxter et al 1987Gur et al 1988 Wu et al 1991) we did not reg nd heightened right anterioractivity in anxiety However a number of other studies have (eg Davidsonet al 1996 Reivich et al 1983 Tomarken amp Davidson 1994) Asdiscussed earlier we have previously argued that anterior regions areuniquely involved in the valence dimension of emotion with pleasantaffect associated with greater left activity and unpleasant affect associatedwith greater right activity (eg Heller 1990) Therefore we would expectheightened right anterior activity in an anxious group only if they differedfrom controls on measures indexing affective valence In our recent studythe nondepressed anxious group did not differ from controls on either theGBI Dysthymia scale or the PANAS PA and NA scales
It may also be important to measure the level of anxious apprehension inthe sample studied For example Tomarken and Davidson (1994) found
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 435
that with Marloweplusmn Crowne scores controlled trait anxiety failed to predictfrontal asymmetry However their groups were divided according to amedian split
In summary our results directly support the proposed distinctionbetween anxious apprehension and anxious arousal as types of anxietythat may account for signireg cant variance in patterns of brain activityThey suggest furthermore that Hellerrsquo s model should be rereg ned to reflectthe likelihood that parietal activation will covary with anxious arousal butnot with anxious apprehension If these distinctions are accurate we wouldexpect to observe increases in right parietotemporal activity only underconditions when anxiety is accompanied by increased arousal
Thus in homodynamic and EEG studies a failure to reg nd increased rightposterior activity in anxiety is not surprising as variations in anxiousarousal are likely to be large Some individuals may experience anxiousarousal frequently especially those with disorders more characteristic ofanxious arousal such as panic disorder phobias or post-traumatic stressdisorder Others although apprehensive may rarely experience anxiousarousal In addition the situation may vary in the degree to which it elicitsanxious arousal Being forced to give a speech is likely to induce a greatdeal of anxious arousal in social phobics thus we would have predicted asignireg cant increase in right parietal activity precisely as found by David-son et al (submitted) Note that this effect was only found during theAnticipation condition not during baseline Planning or after thespeech During the Anticipation condition subjects listened while a voiceannounced every 30 seconds for three minutes the time remaining beforethey would be given a topic and allowed to plan for their speech Pre-sumably this period is the most stressful and the most likely to result inanxious arousal Indeed the correlation between heart rate acceleration tothe Anticipation condition and the increase in state anxiety was signireg cant
More recently Nitschke Heller Imig and Miller (submitted) haveattempted to identify anxious arousal and anxious apprehension usingquestionnaires Data from 783 students for a battery of anxiety and depres-sion questionnaires provided strong support for the anxiety distinctionproposed here at least on a psychological level of analysis The Moodand Anxiety Symptom Questionnaire-Anxious Arousal scale (MASQWatson et al 1995ab) was chosen as a measure of anxious arousalwhereas the Penn State Worry Questionnaire (PSWQ Meyer MillerMetzger amp Borkovec 1990 Molina amp Borkovec 1994) was included asa measure of anxious apprehension The correlation between the two scales(r = 20) was much lower than those involving the other more generalscales of anxiety included in the study (MASQ-General Anxiety scaleBeck Anxiety Inventory and the State-Trait Anxiety Inventory-Traitscale ranging from r = 47 to r = 65) These results further support the
436 HELLER AND NITSCHKE
argument that anxious arousal and anxious apprehension are distinctanxiety states
However new questions are highlighted by this distinction betweenanxious arousal and anxious apprehension particularly in the light ofHellerrsquo s model If anxious apprehension is associated with left anterioractivity the overall pattern of regional brain activity appears to be identicalto the brain activity associated with pleasant emotional states (ie more leftthan right anterior activity) However anxious apprehension might beexpected to fall in the portion of the circumplex model of emotionassociated with activated unpleasant affect (as suggested by adjectivessuch as fearful nervous jittery anxious) Below we suggest a numberof theoretical possibilities for resolving this apparent discrepancy all ofwhich remain to be examined empirically
In general it will be important to clarify the degree to which anxiousapprehension is in fact characterised by unpleasant versus pleasant valenceas well as high versus low arousal In addition it may be helpful toconsider these anterior asymmetries from the approach-withdrawal frame-work employed by Davidson (1992ab Davidson amp Tomarken 1989)Under conditions where anxious apprehension leads to approach tenden-cies one might expect an asymmetry in favour of the left anterior region Incontrast when anxious apprehension is associated with withdrawal tenden-cies (perhaps as anxious arousal increases) an asymmetry may emerge infavour of the right anterior region Conceivably the degree to whichanxious apprehension is characterised by approach versus withdrawaltendencies would depend on the degree of stress and discomfortinvolved the personrsquo s coping style and other factors With regard tovalence we would predict that anxious apprehension characterised byapproach tendencies would be accompanied by relatively pleasant affectAnxious apprehension accompanied by withdrawal tendencies in contrastwould be accompanied by relatively unpleasant affect
It is also possible that instead of a direct link between anxious appre-hension and approach behaviour they share more fundamental character-istics associated with left anterior functions For example one might arguethat they both involve sequential thinking and the development of hier-archically structured plans focusing on future concerns (for a review ofrelevant left anterior functions see Tomarken amp Keener this Issue)
Another important issue has to do with the nature of the asymmetricpattern of activity As reviewed earlier we found that the asymmetry infavour of the left hemisphere for participants classireg ed on the basis of traitanxiety was due to a relative decrease in right hemisphere activity Thusthe left frontal region of the anxious participants was not signireg cantly moreactive than that of controls It will therefore be important to examine thepossibility that absolute activity of the left anterior region plays a more
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 437
important role in the modulation of valence (eg Davidson et al 1995)than of anxious apprehension
Another possibility is that independent left anterior systems are asso-ciated with anxious apprehension versus pleasant affect Liotti and Tucker(1995) review theories that there may be multiple functionally segregatedparallel frontal cortico-striato-thalamic circuits some associated withmotor functions some with set shifting and behavioural inhibition andsome with emotional information processing and hedonic responsivity tostimuli Possibly one of these circuits remacr ects activation associated withpleasant affect and another remacr ects activation associated with anxiousapprehension andor approach Different neurotransmitter systems couldbe involved More precise methods of localisation would therefore benecessary to distinguish the unique brain activity associated with each ofthese psychologicalbehavioural phenomena
Comorbidity of Depression and Anxiety
In considering the issue of decomposing emotional states with regard to thesymptomatology in affective and anxiety disorders we have previouslyargued that issues of comorbidity of depression and anxiety are crucial(Heller et al 1995) A review of the literature reveals that rates of comor-bidity are extremely high (Alloy Kelly Mineka amp Clements 1990 Helleret al 1995 Hiller et al 1989 Katon amp Roy-Byrne 1991) Even in our veryrecent work (Nitschke et al submitted) using the MASQ a measuredesigned by Watson and colleagues to avoid symptom overlap in themeasurement of anxiety and depression (Watson et al 1995ab) comor-bidity was fairly pronounced in a large college student sample The MASQAnxious Arousal (AA) and Anhedonic Depression (AD) scales were used asmeasures of those constructs and the PSWQ provided the measure ofanxious apprehension Of the 125 (out of 783) participants scoring abovethe 80th percentile on at least one of the three scales of interest and belowthe 50th percentile on any where the 80th percentile cut-off was not metfully 50 demonstrated comorbidity by scoring above the 80th percentileon at least two of the scales This rate of comorbidity emerged despite thelow correlations found between the three scales (MASQ AA and MASQAD r = 20 MASQ AA and PSWQ r = 20 MASQ AD and PSWQ r =36) relative to the correlations found for the other more general measuresof depression including the MASQ General Mixed General Depression andGeneral Anxiety scales Beck Depression Inventory Beck Anxiety Inven-tory General Behaviour Inventory-Dysthymia scale and the State-TraitAnxiety Inventory-Trait scale (ranging from r = 47 to r = 76)
Because at least some studies have found opposite patterns of activityfor the right posterior region during depression versus during anxiety (ie
438 HELLER AND NITSCHKE
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
lactate-induced panic attack (Stewart Devous Rush Lane amp Bonte1988) generalised anxiety disorder (GAD) patients during an anxiety-stimulating task (Wu et al 1991) psychometrically dereg ned high-anxiousparticipants immediately following arterial and venous catheterisation andagain after completion of the PET scans (Reivich Gur amp Alavi 1983) andsocial phobics prior to making a public speech (Davidson MarshallTomarken amp Henriques submitted) Consistent with this pattern anti-anxiety medication was found to reduce the rate of glucose metabolismin the right hemisphere especially for the occipital and frontal regions forpatients with GAD (Buchsbaum et al 1987) Similarly in a population ofnonpsychiatric volunteers those administered diazepam intravenouslyshowed maximal decreases in rCBF in the right frontal area whereasthose injected with a placebo showed no rCBF changes (Mathew Wil-son amp Daniel 1985)
Regional brain activity can also be inferred from performance on tasksthat depend on particular brain regions On the CFT Heller et al (1995)reported that high-anxious participants had larger left hemispatial biasesthan low-anxious participants suggested to remacr ect greater right posterioractivity Similarly on a neuropsychological test battery panic disorderpatients performed less well than normal controls on left but not righthemisphere tasks (Yeudall et al 1983) Data obtained from a series ofstudies using behavioural measures to examine right versus left hemispherefunction during an experimental condition producing transient (ie state)anxiety (Tucker Antes Stenslie amp Barnhardt 1978 Tucker Roth Arne-son amp Buckingham 1977 Tyler amp Tucker 1982) are also consistent withan interpretation of greater right than left hemisphere activity in anxiety(Heller et al 1995)
However when subjects were classireg ed according to trait anxietyTucker and colleagues (Tucker et al 1978 Tyler amp Tucker 1982)interpreted the results as strongly suggestive of greater left than righthemisphere activity This interpretation is consistent with reports ofnumerous other labs reg nding increased left hemisphere activity associatedwith anxiety in various cortical and subcortical regions (orbital frontalinferior frontal anterior cingulate caudate putamen and thalamus)Heightened left hemisphere activity has been found in obsessive-compul-sive disorder (OCD) patients (Baxter et al 1987 Swedo et al 1989)GAD patients (Buchsbaum et al 1985 Wu et al 1991) and studentsclassireg ed as ` worriersrsquo rsquo on the basis of self-reports when asked to ` worryabout a specireg c topic of personal concernrsquo rsquo (Carter Johnson amp Borkovec1986)
In summary some reg ndings for anxiety indicate increased right hemi-sphere activation whereas others indicate increased left hemisphere acti-vation Indeed it may be the case that they could co-occur which could
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 429
account for the reg ndings of bilateral increases or no asymmetries In themeantime we are lacking a model to explain these discrepancies
ACCOUNTING FOR INCONSISTENCIES INBRAIN ACTIVITY
We are thus confronted with inconsistent results with regard to regionalbrain activity in both depression and anxiety In previous work we haveargued that in order to resolve these discrepancies it is helpful to turn tothe psychological literature on symptomatology in depression and anxietyIn the sections that follow we make some general observations that applyto both depression and anxiety before focusing on each more exclusively
In addressing the inconsistencies for the blood macr ow reg ndings severalconcerns about the literature are worthy of note First several of the mostfrequently cited studies in this literature sampled a restricted set of brainregions (eg Reiman et al 1984 Reivich et al 1983) Furthermore theregions of interest were not consistent from one study to the nextSecond statistical procedures employing multiple t-tests rather thanrepeated-measures ANOVAs are questionable especially with regard todrawing conclusions about hemispheric differences Third although somestudies examined selected diagnostic groups (eg major depressive dis-order OCD panic disorder GAD) most treated depression or anxiety asa unitary construct and did not consider that different types of each mightbe associated with different patterns of brain activity Moreover with theexception of recent OCD studies most did not take into account thedegree of comorbidity with depression or anxiety which may be criticalin determining the patterns of brain activity that occur (Heller et al1995 Tucker 1988)
As emphasised by Davidson Tomarken and colleagues (Davidson1993 Tomarken Davidson Wheeler amp Doss 1992 Tomarken David-son Wheeler amp Kinney 1992) there are a number of methodogicalapproaches that need to be considered when asymmetric patterns ofEEG activity do not emerge Some of these considerations are relevantto the design of blood macr ow studies as well One issue is that of stability(ie the extent to which brain activity remacr ects a stable individual pattern)This can be addressed by obtaining a sufreg cient number of baseline samplesas well as test-retest data For example Tomarken et al (1992) found thatonly participants with a stable pattern of anterior asymmetry across twomeasurement occasions showed an association between left frontalhypoactivation and more dispositional negative affect as measured bythe PANAS
It is also important to note that for the most part EEG studies haveexamined resting or baseline activity Davidson (1993) has pointed out
430 HELLER AND NITSCHKE
that baseline levels of frontal activation asymmetry may remacr ect individualtendencies toward a valenced affective response but this relationship maynot be observed in the absence of a specireg c environmental elicitor Thushe emphasises the need to examine the relationship between individualdifferences in patterns of asymmetry and affective response to emotionallychallenging tasks As applied to the pattern of posterior reg ndings theseconsiderations suggest the possibility that more robust differences inposterior brain activity might be yielded by studies measuring brainactivity while participants perform tasks for which depressed peopleshow decrements Additional methodological approaches highlighted byDavidson and colleagues such as the use of extreme groups and classifica-tion of participants according to the affect measures as opposed to the brainasymmetry measures would also be useful in future studies of brainactivity
A number of other possible explanations for the failure to reg nd con-sistent decreases in parieto-temporal activity during depression have beenproposed (Heller 1990 1993b) One of these explanations focused on theseverity of the depression This supposition was based on the observationthat posterior asymmetries were more likely to be reported when depres-sion was relatively severe Heller suggested that under less extremeaffective conditions task-induced activity predominates masking theaffect-induced activity More recently Heller et al (1995) suggestedthat severity per se is not the issue rather because the comorbiditywith anxiety decreases as severity of depression increases (Hiller Zau-dig amp Rose 1989) the opposing effects of anxiety on posterior brainactivity diminishes
Heller (1993b) also suggested that the inconsistencies in the depressionand anxiety literatures may be due to signireg cant differences in regionalbrain activity that covary more strongly with dispositional characteristicsassociated with emotional functioning (eg extraversion) than with situa-tionally induced emotions Several studies have reported perceptual(Berenbaum amp Williams 1994 Heller amp Nisenson 1993) and EEG(Nitschke et al 1994) asymmetries for extraversion Nitschke et al alsofound that extraversion moderated the relationship between anxiety andEEG activity across both frontal central and parietal regions For anxiousparticipants high extraversion was associated with less activity (morealpha) whereas for nonanxious (and nondepressed) controls high extra-version was associated with more activity (less alpha) These two sets ofreg ndings albeit preliminary provide evidence for the need to acknowledgeregional differences in brain function for personality characteristics suchas extraversion (for a review see Nitschke Heller amp Miller in prep) andpoint to the possible signireg cance of personality in understanding regionalbrain activity in affective and anxiety disorders
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 431
Subtypes of Depression and Anxiety
Another promising avenue for elucidating the discrepancies reported in theliterature for both depression and anxiety is the analysis of subtypes
Depression Many distinctions for depression have been proposed overthe years including most notably bipolar versus unipolar endogenousversus exogenous and neurotic versus reactive The current Diagnostic andstatistical manual of mental disordersplusmn fourth edition (DSMIV APA 1994)subdivides unipolar depression into major depressive disorder and dysthy-mic disorder and provides further specireg city with criteria for severity andfor psychotic melancholic catatonic or atypical features
Perhaps the most useful distinction in physiological research on depres-sion has been the comparison of melancholic to nonmelancholic depres-sion The dereg ning feature of melancholic depression is anhedonia which isthe inability to experience pleasure Current consensus as indicated by theDSMIV (APA 1994) suggests that melancholia is most strongly charac-terised by a loss of pleasure in all activities and a lack of responsivity topleasurable stimuli Many other symptoms may be shared by both melan-cholic and nonmelancholic depression including depressed mood distur-bances in eating and sleeping patterns low self-esteem inability toconcentrate psychomotor retardation fatigue indecisiveness feelings ofguilt and thoughts about death There is current controversy over whethermelancholic depression is qualitatively different from nonmelancholicdepression with some authors arguing that the primary distinction is oneof severity (Zimmerman Coryell amp Pfohl 1986) Nonetheless neuropsy-chological and neurophysiological studies have been consistent in demon-strating differences between melancholic and nonmelancholic depression
In a comprehensive series of studies Bruder has shown that perceptualasymmetries differ for different subtypes of depression (for a review seeBruder 1995) On a perceptual asymmetry task melancholic patientsshowed an abnormally large right ear (left hemisphere) advantage for adichotic syllables task but no left ear (right hemisphere) advantage for acomplex tone task (Bruder 1995) The lack of a left ear advantage was dueto poor left ear accuracy consistent with other studies demonstrating righthemisphere dereg cits in depressed people However nonmelancholicdepressed patients did not differ from normal controls on either test Alongthese lines in a PET study of melancholic depression Bench et al (1992)reported less activity in the left dorsolateral prefrontal cortex compared tonondepressed controls Note that these data could be consistent with eithera qualitative or a quantitative distinction between melancholic and non-melancholic depression However other data may be more consistent witha qualitative difference between the two types of depression For example
432 HELLER AND NITSCHKE
Davidson Larson and Abercrombie (1995) reported that depression sever-ity in melancholic subjects was predicted by the extent to which activationwas decreased in the left lateral prefrontal region In contrast amongnonmelancholics depression severity was predicted by the extent to whichactivation was increased in the right medial prefrontal area
Davidson et alrsquo s (1995) data have implications for theories regardingthe relationship between pleasantunpleasant affect and anterior regions ofthe brain Research on the psychological characteristics of depression andanxiety has suggested that lack of positive affect (or more preciselyactivated pleasant affect Larsen amp Diener 1992) is unique to depression(Clark amp Watson 1991) Indeed in designing a questionnaire to measureboth general and specireg c components of depression and anxiety Watsonand colleagues (Watson 1995ab) developed a specireg c depression scalethat primarily measures anhedonia Because anhedonia is a fundamentalcharacteristic of melancholia it seems reasonable to assume that decreasedpleasant affect would accompany this type of depression If so Hellerrsquo smodel which predicts less left than right anterior activity for decreasedpleasant affect is completely consistent with the available data
However Davidson et alrsquo s (1995) data suggests the possibility thataffective valence may depend less on the relative balance of activationbetween the left and right anterior regions and more on specireg c systemsassociated with each hemisphere Whereas melancholic depression ischaracterised by decreased pleasant affect depression without melan-cholic features may be better characterised by increased unpleasantaffect Following this line of reasoning Hellerrsquo s model which in itscurrent form argues for a link between anterior asymmetry and the valencedimension would not distinguish between melancholic and nonmelan-cholic depression If affective valence depends more on specireg c systemsassociated with each hemisphere as Davidson et alrsquo s data suggest Hellerrsquo smodel may need to be expanded to remacr ect the unique contributions of theleft and right hemispheres to pleasantunpleasant affect For example itmight be more accurate to link the left anterior region with the activatedpleasant affect (ie positive affect) octant of the circumplex model ofemotion and the right anterior region with the activated unpleasant affect(ie negative affect) octant (see Larson amp Diener 1992)
Because few studies have specireg cally examined the full range of poten-tial affective descriptors in relation to patterns of brain activity theseissues remain to be resolved In most research only isolated portions ofthe circumplex have been sampled For example because positive andnegative affect as measured by the PANAS provide information regardingonly two of the four quadrants in the circumplex model studies that do notcollect data regarding the other two quadrants cannot shed light on howwell they might have accounted for patterns of brain activity
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 433
Anxiety As reviewed earlier studies that examine regional brainactivity in anxiety are even more plagued by inconsistencies than studiesof depression In keeping with the theme of this article we have argued thatby decomposing anxiety into subtypes on the basis of the psychologicalliterature we can potentially resolve some of these discrepancies
In previous work we have suggested that it may be important toconsider the type of anxiety that is being examined (Heller et al 19951997a) In particular we argued that regional brain activity might bedifferent for anxious arousal (eg panic) and anxious apprehension (egworry) types of anxiety that appear to differ signireg cantly in psychologicaland physiological characteristics (eg Barlow 1988 1991 Eysenck 1957Klein 1987 Watson et al 1995b) Anxious arousal has been described byWatson et al as being distinguished by symptoms of physiological hyper-arousal and somatic tension Panic attacks and high-stress states would fallinto this category In contrast anxious apprehension involves worry and ischaracterised by verbal rumination typically about future events This typeof anxiety would be characteristic of obsessive-compulsiveness general-ised anxiety states and trait anxiety (as identireg ed by self-reports of anxiousapprehension and worry on various questionnaires) As noted by Heller etal (1997a) the two kinds of anxiety are not mutually exclusive Extremedegrees of anxious apprehension (eg being worried or fearful about thefuture) may prompt experiences of anxious arousal (eg somatic symptomsand exaggerated physiological responses to stressful events)
We have argued that when the literature on brain activity in anxiety isreviewed in the light of this distinction between anxious arousal andanxious apprehension most of the studies reporting greater right hemi-sphere activity have either studied panic attacks (eg Reiman et al 1984)or high-stress situations likely to entail anxious arousal (eg Tucker et al1977) On the other hand studies indicating greater left hemisphere activity(primarily anterior) have examined OCD (eg Baxter et al 1987) GAD(Wu et al 1991) or questionnaire-dereg ned trait anxiety indexing anxiousapprehension (eg Tucker et al 1978) Indeed as mentioned earlierstudies have highlighted various regions in both hemispheres As aresult it is difreg cult to pinpoint a particular region that is clearly a focalpoint in anxiety Nonetheless the data do suggest that right hemisphereregions emerge as particularly salient in some types of anxiety (ie anxiousarousal) whereas left hemisphere regions emerge as particularly salient inother types of anxiety (ie anxious apprehension)
In a direct test of this hypothesis (Heller et al 1997a) we selectedparticipants on the basis of self-reported trait anxiety using a measureindexing anxious apprehension in a between-subjects design We thenmanipulated anxious arousal on a within-subjects basis Specireg cally wecontrasted brain activity (EEG alpha) during rest periods to brain activity
434 HELLER AND NITSCHKE
during a task in which participants listened to emotional narrativesdesigned to elicit anxious arousal (ie fear and sad narratives that hadpreviously received unpleasant valence and high arousal ratings)
For frontal activity group differences were unaffected by the narrativesAnxious participants showed a larger asymmetry in favour of the lefthemisphere than controls during the rest and listen periods of all narrativetypes as well as for eight minutes of resting baseline preceding thenarrative task Examination of the data indicated that the increased asym-metry was primarily due to a decrease in right hemisphere activity for theanxious participants Nonetheless the increase in magnitude of the asym-metry favouring the left hemisphere found for anxious participants isconsistent with the studies reviewed earlier that reported or inferredrelatively greater left than right anterior activity in anxious subjectsclassireg ed on the basis of worry or trait anxiety (eg Carter et al 1986Tucker et al 1978 Tyler amp Tucker 1982) Our frontal data suggest that itwill be important to investigate whether anxious apprehension is bestcharacterised by the magnitude of the asymmetry favouring the left hemi-sphere or by the level of left hemisphere activity in isolation
In contrast the parietal data supported our prediction regarding theassociation of anxious arousal with greater right hemisphere activityWhen listening to the fear and sad narratives anxious participants showeda selective increase in right parietal activity that was not demonstrated bycontrols The right parietal effect however was not obtained for neutraland happy narratives that had previously received pleasant valence and lowarousal ratings These reg ndings are consistent with those studies that havereported or inferred greater right hemisphere activity during anxious statescharacterised by panic and stress-inducing conditions (eg Reiman et al1984 Stewart et al 1988 Tucker et al 1977)
An important question here has to do with the discrepancies in frontalasymmetries that have been reported Like others (eg Baxter et al 1987Gur et al 1988 Wu et al 1991) we did not reg nd heightened right anterioractivity in anxiety However a number of other studies have (eg Davidsonet al 1996 Reivich et al 1983 Tomarken amp Davidson 1994) Asdiscussed earlier we have previously argued that anterior regions areuniquely involved in the valence dimension of emotion with pleasantaffect associated with greater left activity and unpleasant affect associatedwith greater right activity (eg Heller 1990) Therefore we would expectheightened right anterior activity in an anxious group only if they differedfrom controls on measures indexing affective valence In our recent studythe nondepressed anxious group did not differ from controls on either theGBI Dysthymia scale or the PANAS PA and NA scales
It may also be important to measure the level of anxious apprehension inthe sample studied For example Tomarken and Davidson (1994) found
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 435
that with Marloweplusmn Crowne scores controlled trait anxiety failed to predictfrontal asymmetry However their groups were divided according to amedian split
In summary our results directly support the proposed distinctionbetween anxious apprehension and anxious arousal as types of anxietythat may account for signireg cant variance in patterns of brain activityThey suggest furthermore that Hellerrsquo s model should be rereg ned to reflectthe likelihood that parietal activation will covary with anxious arousal butnot with anxious apprehension If these distinctions are accurate we wouldexpect to observe increases in right parietotemporal activity only underconditions when anxiety is accompanied by increased arousal
Thus in homodynamic and EEG studies a failure to reg nd increased rightposterior activity in anxiety is not surprising as variations in anxiousarousal are likely to be large Some individuals may experience anxiousarousal frequently especially those with disorders more characteristic ofanxious arousal such as panic disorder phobias or post-traumatic stressdisorder Others although apprehensive may rarely experience anxiousarousal In addition the situation may vary in the degree to which it elicitsanxious arousal Being forced to give a speech is likely to induce a greatdeal of anxious arousal in social phobics thus we would have predicted asignireg cant increase in right parietal activity precisely as found by David-son et al (submitted) Note that this effect was only found during theAnticipation condition not during baseline Planning or after thespeech During the Anticipation condition subjects listened while a voiceannounced every 30 seconds for three minutes the time remaining beforethey would be given a topic and allowed to plan for their speech Pre-sumably this period is the most stressful and the most likely to result inanxious arousal Indeed the correlation between heart rate acceleration tothe Anticipation condition and the increase in state anxiety was signireg cant
More recently Nitschke Heller Imig and Miller (submitted) haveattempted to identify anxious arousal and anxious apprehension usingquestionnaires Data from 783 students for a battery of anxiety and depres-sion questionnaires provided strong support for the anxiety distinctionproposed here at least on a psychological level of analysis The Moodand Anxiety Symptom Questionnaire-Anxious Arousal scale (MASQWatson et al 1995ab) was chosen as a measure of anxious arousalwhereas the Penn State Worry Questionnaire (PSWQ Meyer MillerMetzger amp Borkovec 1990 Molina amp Borkovec 1994) was included asa measure of anxious apprehension The correlation between the two scales(r = 20) was much lower than those involving the other more generalscales of anxiety included in the study (MASQ-General Anxiety scaleBeck Anxiety Inventory and the State-Trait Anxiety Inventory-Traitscale ranging from r = 47 to r = 65) These results further support the
436 HELLER AND NITSCHKE
argument that anxious arousal and anxious apprehension are distinctanxiety states
However new questions are highlighted by this distinction betweenanxious arousal and anxious apprehension particularly in the light ofHellerrsquo s model If anxious apprehension is associated with left anterioractivity the overall pattern of regional brain activity appears to be identicalto the brain activity associated with pleasant emotional states (ie more leftthan right anterior activity) However anxious apprehension might beexpected to fall in the portion of the circumplex model of emotionassociated with activated unpleasant affect (as suggested by adjectivessuch as fearful nervous jittery anxious) Below we suggest a numberof theoretical possibilities for resolving this apparent discrepancy all ofwhich remain to be examined empirically
In general it will be important to clarify the degree to which anxiousapprehension is in fact characterised by unpleasant versus pleasant valenceas well as high versus low arousal In addition it may be helpful toconsider these anterior asymmetries from the approach-withdrawal frame-work employed by Davidson (1992ab Davidson amp Tomarken 1989)Under conditions where anxious apprehension leads to approach tenden-cies one might expect an asymmetry in favour of the left anterior region Incontrast when anxious apprehension is associated with withdrawal tenden-cies (perhaps as anxious arousal increases) an asymmetry may emerge infavour of the right anterior region Conceivably the degree to whichanxious apprehension is characterised by approach versus withdrawaltendencies would depend on the degree of stress and discomfortinvolved the personrsquo s coping style and other factors With regard tovalence we would predict that anxious apprehension characterised byapproach tendencies would be accompanied by relatively pleasant affectAnxious apprehension accompanied by withdrawal tendencies in contrastwould be accompanied by relatively unpleasant affect
It is also possible that instead of a direct link between anxious appre-hension and approach behaviour they share more fundamental character-istics associated with left anterior functions For example one might arguethat they both involve sequential thinking and the development of hier-archically structured plans focusing on future concerns (for a review ofrelevant left anterior functions see Tomarken amp Keener this Issue)
Another important issue has to do with the nature of the asymmetricpattern of activity As reviewed earlier we found that the asymmetry infavour of the left hemisphere for participants classireg ed on the basis of traitanxiety was due to a relative decrease in right hemisphere activity Thusthe left frontal region of the anxious participants was not signireg cantly moreactive than that of controls It will therefore be important to examine thepossibility that absolute activity of the left anterior region plays a more
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 437
important role in the modulation of valence (eg Davidson et al 1995)than of anxious apprehension
Another possibility is that independent left anterior systems are asso-ciated with anxious apprehension versus pleasant affect Liotti and Tucker(1995) review theories that there may be multiple functionally segregatedparallel frontal cortico-striato-thalamic circuits some associated withmotor functions some with set shifting and behavioural inhibition andsome with emotional information processing and hedonic responsivity tostimuli Possibly one of these circuits remacr ects activation associated withpleasant affect and another remacr ects activation associated with anxiousapprehension andor approach Different neurotransmitter systems couldbe involved More precise methods of localisation would therefore benecessary to distinguish the unique brain activity associated with each ofthese psychologicalbehavioural phenomena
Comorbidity of Depression and Anxiety
In considering the issue of decomposing emotional states with regard to thesymptomatology in affective and anxiety disorders we have previouslyargued that issues of comorbidity of depression and anxiety are crucial(Heller et al 1995) A review of the literature reveals that rates of comor-bidity are extremely high (Alloy Kelly Mineka amp Clements 1990 Helleret al 1995 Hiller et al 1989 Katon amp Roy-Byrne 1991) Even in our veryrecent work (Nitschke et al submitted) using the MASQ a measuredesigned by Watson and colleagues to avoid symptom overlap in themeasurement of anxiety and depression (Watson et al 1995ab) comor-bidity was fairly pronounced in a large college student sample The MASQAnxious Arousal (AA) and Anhedonic Depression (AD) scales were used asmeasures of those constructs and the PSWQ provided the measure ofanxious apprehension Of the 125 (out of 783) participants scoring abovethe 80th percentile on at least one of the three scales of interest and belowthe 50th percentile on any where the 80th percentile cut-off was not metfully 50 demonstrated comorbidity by scoring above the 80th percentileon at least two of the scales This rate of comorbidity emerged despite thelow correlations found between the three scales (MASQ AA and MASQAD r = 20 MASQ AA and PSWQ r = 20 MASQ AD and PSWQ r =36) relative to the correlations found for the other more general measuresof depression including the MASQ General Mixed General Depression andGeneral Anxiety scales Beck Depression Inventory Beck Anxiety Inven-tory General Behaviour Inventory-Dysthymia scale and the State-TraitAnxiety Inventory-Trait scale (ranging from r = 47 to r = 76)
Because at least some studies have found opposite patterns of activityfor the right posterior region during depression versus during anxiety (ie
438 HELLER AND NITSCHKE
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
account for the reg ndings of bilateral increases or no asymmetries In themeantime we are lacking a model to explain these discrepancies
ACCOUNTING FOR INCONSISTENCIES INBRAIN ACTIVITY
We are thus confronted with inconsistent results with regard to regionalbrain activity in both depression and anxiety In previous work we haveargued that in order to resolve these discrepancies it is helpful to turn tothe psychological literature on symptomatology in depression and anxietyIn the sections that follow we make some general observations that applyto both depression and anxiety before focusing on each more exclusively
In addressing the inconsistencies for the blood macr ow reg ndings severalconcerns about the literature are worthy of note First several of the mostfrequently cited studies in this literature sampled a restricted set of brainregions (eg Reiman et al 1984 Reivich et al 1983) Furthermore theregions of interest were not consistent from one study to the nextSecond statistical procedures employing multiple t-tests rather thanrepeated-measures ANOVAs are questionable especially with regard todrawing conclusions about hemispheric differences Third although somestudies examined selected diagnostic groups (eg major depressive dis-order OCD panic disorder GAD) most treated depression or anxiety asa unitary construct and did not consider that different types of each mightbe associated with different patterns of brain activity Moreover with theexception of recent OCD studies most did not take into account thedegree of comorbidity with depression or anxiety which may be criticalin determining the patterns of brain activity that occur (Heller et al1995 Tucker 1988)
As emphasised by Davidson Tomarken and colleagues (Davidson1993 Tomarken Davidson Wheeler amp Doss 1992 Tomarken David-son Wheeler amp Kinney 1992) there are a number of methodogicalapproaches that need to be considered when asymmetric patterns ofEEG activity do not emerge Some of these considerations are relevantto the design of blood macr ow studies as well One issue is that of stability(ie the extent to which brain activity remacr ects a stable individual pattern)This can be addressed by obtaining a sufreg cient number of baseline samplesas well as test-retest data For example Tomarken et al (1992) found thatonly participants with a stable pattern of anterior asymmetry across twomeasurement occasions showed an association between left frontalhypoactivation and more dispositional negative affect as measured bythe PANAS
It is also important to note that for the most part EEG studies haveexamined resting or baseline activity Davidson (1993) has pointed out
430 HELLER AND NITSCHKE
that baseline levels of frontal activation asymmetry may remacr ect individualtendencies toward a valenced affective response but this relationship maynot be observed in the absence of a specireg c environmental elicitor Thushe emphasises the need to examine the relationship between individualdifferences in patterns of asymmetry and affective response to emotionallychallenging tasks As applied to the pattern of posterior reg ndings theseconsiderations suggest the possibility that more robust differences inposterior brain activity might be yielded by studies measuring brainactivity while participants perform tasks for which depressed peopleshow decrements Additional methodological approaches highlighted byDavidson and colleagues such as the use of extreme groups and classifica-tion of participants according to the affect measures as opposed to the brainasymmetry measures would also be useful in future studies of brainactivity
A number of other possible explanations for the failure to reg nd con-sistent decreases in parieto-temporal activity during depression have beenproposed (Heller 1990 1993b) One of these explanations focused on theseverity of the depression This supposition was based on the observationthat posterior asymmetries were more likely to be reported when depres-sion was relatively severe Heller suggested that under less extremeaffective conditions task-induced activity predominates masking theaffect-induced activity More recently Heller et al (1995) suggestedthat severity per se is not the issue rather because the comorbiditywith anxiety decreases as severity of depression increases (Hiller Zau-dig amp Rose 1989) the opposing effects of anxiety on posterior brainactivity diminishes
Heller (1993b) also suggested that the inconsistencies in the depressionand anxiety literatures may be due to signireg cant differences in regionalbrain activity that covary more strongly with dispositional characteristicsassociated with emotional functioning (eg extraversion) than with situa-tionally induced emotions Several studies have reported perceptual(Berenbaum amp Williams 1994 Heller amp Nisenson 1993) and EEG(Nitschke et al 1994) asymmetries for extraversion Nitschke et al alsofound that extraversion moderated the relationship between anxiety andEEG activity across both frontal central and parietal regions For anxiousparticipants high extraversion was associated with less activity (morealpha) whereas for nonanxious (and nondepressed) controls high extra-version was associated with more activity (less alpha) These two sets ofreg ndings albeit preliminary provide evidence for the need to acknowledgeregional differences in brain function for personality characteristics suchas extraversion (for a review see Nitschke Heller amp Miller in prep) andpoint to the possible signireg cance of personality in understanding regionalbrain activity in affective and anxiety disorders
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 431
Subtypes of Depression and Anxiety
Another promising avenue for elucidating the discrepancies reported in theliterature for both depression and anxiety is the analysis of subtypes
Depression Many distinctions for depression have been proposed overthe years including most notably bipolar versus unipolar endogenousversus exogenous and neurotic versus reactive The current Diagnostic andstatistical manual of mental disordersplusmn fourth edition (DSMIV APA 1994)subdivides unipolar depression into major depressive disorder and dysthy-mic disorder and provides further specireg city with criteria for severity andfor psychotic melancholic catatonic or atypical features
Perhaps the most useful distinction in physiological research on depres-sion has been the comparison of melancholic to nonmelancholic depres-sion The dereg ning feature of melancholic depression is anhedonia which isthe inability to experience pleasure Current consensus as indicated by theDSMIV (APA 1994) suggests that melancholia is most strongly charac-terised by a loss of pleasure in all activities and a lack of responsivity topleasurable stimuli Many other symptoms may be shared by both melan-cholic and nonmelancholic depression including depressed mood distur-bances in eating and sleeping patterns low self-esteem inability toconcentrate psychomotor retardation fatigue indecisiveness feelings ofguilt and thoughts about death There is current controversy over whethermelancholic depression is qualitatively different from nonmelancholicdepression with some authors arguing that the primary distinction is oneof severity (Zimmerman Coryell amp Pfohl 1986) Nonetheless neuropsy-chological and neurophysiological studies have been consistent in demon-strating differences between melancholic and nonmelancholic depression
In a comprehensive series of studies Bruder has shown that perceptualasymmetries differ for different subtypes of depression (for a review seeBruder 1995) On a perceptual asymmetry task melancholic patientsshowed an abnormally large right ear (left hemisphere) advantage for adichotic syllables task but no left ear (right hemisphere) advantage for acomplex tone task (Bruder 1995) The lack of a left ear advantage was dueto poor left ear accuracy consistent with other studies demonstrating righthemisphere dereg cits in depressed people However nonmelancholicdepressed patients did not differ from normal controls on either test Alongthese lines in a PET study of melancholic depression Bench et al (1992)reported less activity in the left dorsolateral prefrontal cortex compared tonondepressed controls Note that these data could be consistent with eithera qualitative or a quantitative distinction between melancholic and non-melancholic depression However other data may be more consistent witha qualitative difference between the two types of depression For example
432 HELLER AND NITSCHKE
Davidson Larson and Abercrombie (1995) reported that depression sever-ity in melancholic subjects was predicted by the extent to which activationwas decreased in the left lateral prefrontal region In contrast amongnonmelancholics depression severity was predicted by the extent to whichactivation was increased in the right medial prefrontal area
Davidson et alrsquo s (1995) data have implications for theories regardingthe relationship between pleasantunpleasant affect and anterior regions ofthe brain Research on the psychological characteristics of depression andanxiety has suggested that lack of positive affect (or more preciselyactivated pleasant affect Larsen amp Diener 1992) is unique to depression(Clark amp Watson 1991) Indeed in designing a questionnaire to measureboth general and specireg c components of depression and anxiety Watsonand colleagues (Watson 1995ab) developed a specireg c depression scalethat primarily measures anhedonia Because anhedonia is a fundamentalcharacteristic of melancholia it seems reasonable to assume that decreasedpleasant affect would accompany this type of depression If so Hellerrsquo smodel which predicts less left than right anterior activity for decreasedpleasant affect is completely consistent with the available data
However Davidson et alrsquo s (1995) data suggests the possibility thataffective valence may depend less on the relative balance of activationbetween the left and right anterior regions and more on specireg c systemsassociated with each hemisphere Whereas melancholic depression ischaracterised by decreased pleasant affect depression without melan-cholic features may be better characterised by increased unpleasantaffect Following this line of reasoning Hellerrsquo s model which in itscurrent form argues for a link between anterior asymmetry and the valencedimension would not distinguish between melancholic and nonmelan-cholic depression If affective valence depends more on specireg c systemsassociated with each hemisphere as Davidson et alrsquo s data suggest Hellerrsquo smodel may need to be expanded to remacr ect the unique contributions of theleft and right hemispheres to pleasantunpleasant affect For example itmight be more accurate to link the left anterior region with the activatedpleasant affect (ie positive affect) octant of the circumplex model ofemotion and the right anterior region with the activated unpleasant affect(ie negative affect) octant (see Larson amp Diener 1992)
Because few studies have specireg cally examined the full range of poten-tial affective descriptors in relation to patterns of brain activity theseissues remain to be resolved In most research only isolated portions ofthe circumplex have been sampled For example because positive andnegative affect as measured by the PANAS provide information regardingonly two of the four quadrants in the circumplex model studies that do notcollect data regarding the other two quadrants cannot shed light on howwell they might have accounted for patterns of brain activity
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 433
Anxiety As reviewed earlier studies that examine regional brainactivity in anxiety are even more plagued by inconsistencies than studiesof depression In keeping with the theme of this article we have argued thatby decomposing anxiety into subtypes on the basis of the psychologicalliterature we can potentially resolve some of these discrepancies
In previous work we have suggested that it may be important toconsider the type of anxiety that is being examined (Heller et al 19951997a) In particular we argued that regional brain activity might bedifferent for anxious arousal (eg panic) and anxious apprehension (egworry) types of anxiety that appear to differ signireg cantly in psychologicaland physiological characteristics (eg Barlow 1988 1991 Eysenck 1957Klein 1987 Watson et al 1995b) Anxious arousal has been described byWatson et al as being distinguished by symptoms of physiological hyper-arousal and somatic tension Panic attacks and high-stress states would fallinto this category In contrast anxious apprehension involves worry and ischaracterised by verbal rumination typically about future events This typeof anxiety would be characteristic of obsessive-compulsiveness general-ised anxiety states and trait anxiety (as identireg ed by self-reports of anxiousapprehension and worry on various questionnaires) As noted by Heller etal (1997a) the two kinds of anxiety are not mutually exclusive Extremedegrees of anxious apprehension (eg being worried or fearful about thefuture) may prompt experiences of anxious arousal (eg somatic symptomsand exaggerated physiological responses to stressful events)
We have argued that when the literature on brain activity in anxiety isreviewed in the light of this distinction between anxious arousal andanxious apprehension most of the studies reporting greater right hemi-sphere activity have either studied panic attacks (eg Reiman et al 1984)or high-stress situations likely to entail anxious arousal (eg Tucker et al1977) On the other hand studies indicating greater left hemisphere activity(primarily anterior) have examined OCD (eg Baxter et al 1987) GAD(Wu et al 1991) or questionnaire-dereg ned trait anxiety indexing anxiousapprehension (eg Tucker et al 1978) Indeed as mentioned earlierstudies have highlighted various regions in both hemispheres As aresult it is difreg cult to pinpoint a particular region that is clearly a focalpoint in anxiety Nonetheless the data do suggest that right hemisphereregions emerge as particularly salient in some types of anxiety (ie anxiousarousal) whereas left hemisphere regions emerge as particularly salient inother types of anxiety (ie anxious apprehension)
In a direct test of this hypothesis (Heller et al 1997a) we selectedparticipants on the basis of self-reported trait anxiety using a measureindexing anxious apprehension in a between-subjects design We thenmanipulated anxious arousal on a within-subjects basis Specireg cally wecontrasted brain activity (EEG alpha) during rest periods to brain activity
434 HELLER AND NITSCHKE
during a task in which participants listened to emotional narrativesdesigned to elicit anxious arousal (ie fear and sad narratives that hadpreviously received unpleasant valence and high arousal ratings)
For frontal activity group differences were unaffected by the narrativesAnxious participants showed a larger asymmetry in favour of the lefthemisphere than controls during the rest and listen periods of all narrativetypes as well as for eight minutes of resting baseline preceding thenarrative task Examination of the data indicated that the increased asym-metry was primarily due to a decrease in right hemisphere activity for theanxious participants Nonetheless the increase in magnitude of the asym-metry favouring the left hemisphere found for anxious participants isconsistent with the studies reviewed earlier that reported or inferredrelatively greater left than right anterior activity in anxious subjectsclassireg ed on the basis of worry or trait anxiety (eg Carter et al 1986Tucker et al 1978 Tyler amp Tucker 1982) Our frontal data suggest that itwill be important to investigate whether anxious apprehension is bestcharacterised by the magnitude of the asymmetry favouring the left hemi-sphere or by the level of left hemisphere activity in isolation
In contrast the parietal data supported our prediction regarding theassociation of anxious arousal with greater right hemisphere activityWhen listening to the fear and sad narratives anxious participants showeda selective increase in right parietal activity that was not demonstrated bycontrols The right parietal effect however was not obtained for neutraland happy narratives that had previously received pleasant valence and lowarousal ratings These reg ndings are consistent with those studies that havereported or inferred greater right hemisphere activity during anxious statescharacterised by panic and stress-inducing conditions (eg Reiman et al1984 Stewart et al 1988 Tucker et al 1977)
An important question here has to do with the discrepancies in frontalasymmetries that have been reported Like others (eg Baxter et al 1987Gur et al 1988 Wu et al 1991) we did not reg nd heightened right anterioractivity in anxiety However a number of other studies have (eg Davidsonet al 1996 Reivich et al 1983 Tomarken amp Davidson 1994) Asdiscussed earlier we have previously argued that anterior regions areuniquely involved in the valence dimension of emotion with pleasantaffect associated with greater left activity and unpleasant affect associatedwith greater right activity (eg Heller 1990) Therefore we would expectheightened right anterior activity in an anxious group only if they differedfrom controls on measures indexing affective valence In our recent studythe nondepressed anxious group did not differ from controls on either theGBI Dysthymia scale or the PANAS PA and NA scales
It may also be important to measure the level of anxious apprehension inthe sample studied For example Tomarken and Davidson (1994) found
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 435
that with Marloweplusmn Crowne scores controlled trait anxiety failed to predictfrontal asymmetry However their groups were divided according to amedian split
In summary our results directly support the proposed distinctionbetween anxious apprehension and anxious arousal as types of anxietythat may account for signireg cant variance in patterns of brain activityThey suggest furthermore that Hellerrsquo s model should be rereg ned to reflectthe likelihood that parietal activation will covary with anxious arousal butnot with anxious apprehension If these distinctions are accurate we wouldexpect to observe increases in right parietotemporal activity only underconditions when anxiety is accompanied by increased arousal
Thus in homodynamic and EEG studies a failure to reg nd increased rightposterior activity in anxiety is not surprising as variations in anxiousarousal are likely to be large Some individuals may experience anxiousarousal frequently especially those with disorders more characteristic ofanxious arousal such as panic disorder phobias or post-traumatic stressdisorder Others although apprehensive may rarely experience anxiousarousal In addition the situation may vary in the degree to which it elicitsanxious arousal Being forced to give a speech is likely to induce a greatdeal of anxious arousal in social phobics thus we would have predicted asignireg cant increase in right parietal activity precisely as found by David-son et al (submitted) Note that this effect was only found during theAnticipation condition not during baseline Planning or after thespeech During the Anticipation condition subjects listened while a voiceannounced every 30 seconds for three minutes the time remaining beforethey would be given a topic and allowed to plan for their speech Pre-sumably this period is the most stressful and the most likely to result inanxious arousal Indeed the correlation between heart rate acceleration tothe Anticipation condition and the increase in state anxiety was signireg cant
More recently Nitschke Heller Imig and Miller (submitted) haveattempted to identify anxious arousal and anxious apprehension usingquestionnaires Data from 783 students for a battery of anxiety and depres-sion questionnaires provided strong support for the anxiety distinctionproposed here at least on a psychological level of analysis The Moodand Anxiety Symptom Questionnaire-Anxious Arousal scale (MASQWatson et al 1995ab) was chosen as a measure of anxious arousalwhereas the Penn State Worry Questionnaire (PSWQ Meyer MillerMetzger amp Borkovec 1990 Molina amp Borkovec 1994) was included asa measure of anxious apprehension The correlation between the two scales(r = 20) was much lower than those involving the other more generalscales of anxiety included in the study (MASQ-General Anxiety scaleBeck Anxiety Inventory and the State-Trait Anxiety Inventory-Traitscale ranging from r = 47 to r = 65) These results further support the
436 HELLER AND NITSCHKE
argument that anxious arousal and anxious apprehension are distinctanxiety states
However new questions are highlighted by this distinction betweenanxious arousal and anxious apprehension particularly in the light ofHellerrsquo s model If anxious apprehension is associated with left anterioractivity the overall pattern of regional brain activity appears to be identicalto the brain activity associated with pleasant emotional states (ie more leftthan right anterior activity) However anxious apprehension might beexpected to fall in the portion of the circumplex model of emotionassociated with activated unpleasant affect (as suggested by adjectivessuch as fearful nervous jittery anxious) Below we suggest a numberof theoretical possibilities for resolving this apparent discrepancy all ofwhich remain to be examined empirically
In general it will be important to clarify the degree to which anxiousapprehension is in fact characterised by unpleasant versus pleasant valenceas well as high versus low arousal In addition it may be helpful toconsider these anterior asymmetries from the approach-withdrawal frame-work employed by Davidson (1992ab Davidson amp Tomarken 1989)Under conditions where anxious apprehension leads to approach tenden-cies one might expect an asymmetry in favour of the left anterior region Incontrast when anxious apprehension is associated with withdrawal tenden-cies (perhaps as anxious arousal increases) an asymmetry may emerge infavour of the right anterior region Conceivably the degree to whichanxious apprehension is characterised by approach versus withdrawaltendencies would depend on the degree of stress and discomfortinvolved the personrsquo s coping style and other factors With regard tovalence we would predict that anxious apprehension characterised byapproach tendencies would be accompanied by relatively pleasant affectAnxious apprehension accompanied by withdrawal tendencies in contrastwould be accompanied by relatively unpleasant affect
It is also possible that instead of a direct link between anxious appre-hension and approach behaviour they share more fundamental character-istics associated with left anterior functions For example one might arguethat they both involve sequential thinking and the development of hier-archically structured plans focusing on future concerns (for a review ofrelevant left anterior functions see Tomarken amp Keener this Issue)
Another important issue has to do with the nature of the asymmetricpattern of activity As reviewed earlier we found that the asymmetry infavour of the left hemisphere for participants classireg ed on the basis of traitanxiety was due to a relative decrease in right hemisphere activity Thusthe left frontal region of the anxious participants was not signireg cantly moreactive than that of controls It will therefore be important to examine thepossibility that absolute activity of the left anterior region plays a more
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 437
important role in the modulation of valence (eg Davidson et al 1995)than of anxious apprehension
Another possibility is that independent left anterior systems are asso-ciated with anxious apprehension versus pleasant affect Liotti and Tucker(1995) review theories that there may be multiple functionally segregatedparallel frontal cortico-striato-thalamic circuits some associated withmotor functions some with set shifting and behavioural inhibition andsome with emotional information processing and hedonic responsivity tostimuli Possibly one of these circuits remacr ects activation associated withpleasant affect and another remacr ects activation associated with anxiousapprehension andor approach Different neurotransmitter systems couldbe involved More precise methods of localisation would therefore benecessary to distinguish the unique brain activity associated with each ofthese psychologicalbehavioural phenomena
Comorbidity of Depression and Anxiety
In considering the issue of decomposing emotional states with regard to thesymptomatology in affective and anxiety disorders we have previouslyargued that issues of comorbidity of depression and anxiety are crucial(Heller et al 1995) A review of the literature reveals that rates of comor-bidity are extremely high (Alloy Kelly Mineka amp Clements 1990 Helleret al 1995 Hiller et al 1989 Katon amp Roy-Byrne 1991) Even in our veryrecent work (Nitschke et al submitted) using the MASQ a measuredesigned by Watson and colleagues to avoid symptom overlap in themeasurement of anxiety and depression (Watson et al 1995ab) comor-bidity was fairly pronounced in a large college student sample The MASQAnxious Arousal (AA) and Anhedonic Depression (AD) scales were used asmeasures of those constructs and the PSWQ provided the measure ofanxious apprehension Of the 125 (out of 783) participants scoring abovethe 80th percentile on at least one of the three scales of interest and belowthe 50th percentile on any where the 80th percentile cut-off was not metfully 50 demonstrated comorbidity by scoring above the 80th percentileon at least two of the scales This rate of comorbidity emerged despite thelow correlations found between the three scales (MASQ AA and MASQAD r = 20 MASQ AA and PSWQ r = 20 MASQ AD and PSWQ r =36) relative to the correlations found for the other more general measuresof depression including the MASQ General Mixed General Depression andGeneral Anxiety scales Beck Depression Inventory Beck Anxiety Inven-tory General Behaviour Inventory-Dysthymia scale and the State-TraitAnxiety Inventory-Trait scale (ranging from r = 47 to r = 76)
Because at least some studies have found opposite patterns of activityfor the right posterior region during depression versus during anxiety (ie
438 HELLER AND NITSCHKE
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
that baseline levels of frontal activation asymmetry may remacr ect individualtendencies toward a valenced affective response but this relationship maynot be observed in the absence of a specireg c environmental elicitor Thushe emphasises the need to examine the relationship between individualdifferences in patterns of asymmetry and affective response to emotionallychallenging tasks As applied to the pattern of posterior reg ndings theseconsiderations suggest the possibility that more robust differences inposterior brain activity might be yielded by studies measuring brainactivity while participants perform tasks for which depressed peopleshow decrements Additional methodological approaches highlighted byDavidson and colleagues such as the use of extreme groups and classifica-tion of participants according to the affect measures as opposed to the brainasymmetry measures would also be useful in future studies of brainactivity
A number of other possible explanations for the failure to reg nd con-sistent decreases in parieto-temporal activity during depression have beenproposed (Heller 1990 1993b) One of these explanations focused on theseverity of the depression This supposition was based on the observationthat posterior asymmetries were more likely to be reported when depres-sion was relatively severe Heller suggested that under less extremeaffective conditions task-induced activity predominates masking theaffect-induced activity More recently Heller et al (1995) suggestedthat severity per se is not the issue rather because the comorbiditywith anxiety decreases as severity of depression increases (Hiller Zau-dig amp Rose 1989) the opposing effects of anxiety on posterior brainactivity diminishes
Heller (1993b) also suggested that the inconsistencies in the depressionand anxiety literatures may be due to signireg cant differences in regionalbrain activity that covary more strongly with dispositional characteristicsassociated with emotional functioning (eg extraversion) than with situa-tionally induced emotions Several studies have reported perceptual(Berenbaum amp Williams 1994 Heller amp Nisenson 1993) and EEG(Nitschke et al 1994) asymmetries for extraversion Nitschke et al alsofound that extraversion moderated the relationship between anxiety andEEG activity across both frontal central and parietal regions For anxiousparticipants high extraversion was associated with less activity (morealpha) whereas for nonanxious (and nondepressed) controls high extra-version was associated with more activity (less alpha) These two sets ofreg ndings albeit preliminary provide evidence for the need to acknowledgeregional differences in brain function for personality characteristics suchas extraversion (for a review see Nitschke Heller amp Miller in prep) andpoint to the possible signireg cance of personality in understanding regionalbrain activity in affective and anxiety disorders
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 431
Subtypes of Depression and Anxiety
Another promising avenue for elucidating the discrepancies reported in theliterature for both depression and anxiety is the analysis of subtypes
Depression Many distinctions for depression have been proposed overthe years including most notably bipolar versus unipolar endogenousversus exogenous and neurotic versus reactive The current Diagnostic andstatistical manual of mental disordersplusmn fourth edition (DSMIV APA 1994)subdivides unipolar depression into major depressive disorder and dysthy-mic disorder and provides further specireg city with criteria for severity andfor psychotic melancholic catatonic or atypical features
Perhaps the most useful distinction in physiological research on depres-sion has been the comparison of melancholic to nonmelancholic depres-sion The dereg ning feature of melancholic depression is anhedonia which isthe inability to experience pleasure Current consensus as indicated by theDSMIV (APA 1994) suggests that melancholia is most strongly charac-terised by a loss of pleasure in all activities and a lack of responsivity topleasurable stimuli Many other symptoms may be shared by both melan-cholic and nonmelancholic depression including depressed mood distur-bances in eating and sleeping patterns low self-esteem inability toconcentrate psychomotor retardation fatigue indecisiveness feelings ofguilt and thoughts about death There is current controversy over whethermelancholic depression is qualitatively different from nonmelancholicdepression with some authors arguing that the primary distinction is oneof severity (Zimmerman Coryell amp Pfohl 1986) Nonetheless neuropsy-chological and neurophysiological studies have been consistent in demon-strating differences between melancholic and nonmelancholic depression
In a comprehensive series of studies Bruder has shown that perceptualasymmetries differ for different subtypes of depression (for a review seeBruder 1995) On a perceptual asymmetry task melancholic patientsshowed an abnormally large right ear (left hemisphere) advantage for adichotic syllables task but no left ear (right hemisphere) advantage for acomplex tone task (Bruder 1995) The lack of a left ear advantage was dueto poor left ear accuracy consistent with other studies demonstrating righthemisphere dereg cits in depressed people However nonmelancholicdepressed patients did not differ from normal controls on either test Alongthese lines in a PET study of melancholic depression Bench et al (1992)reported less activity in the left dorsolateral prefrontal cortex compared tonondepressed controls Note that these data could be consistent with eithera qualitative or a quantitative distinction between melancholic and non-melancholic depression However other data may be more consistent witha qualitative difference between the two types of depression For example
432 HELLER AND NITSCHKE
Davidson Larson and Abercrombie (1995) reported that depression sever-ity in melancholic subjects was predicted by the extent to which activationwas decreased in the left lateral prefrontal region In contrast amongnonmelancholics depression severity was predicted by the extent to whichactivation was increased in the right medial prefrontal area
Davidson et alrsquo s (1995) data have implications for theories regardingthe relationship between pleasantunpleasant affect and anterior regions ofthe brain Research on the psychological characteristics of depression andanxiety has suggested that lack of positive affect (or more preciselyactivated pleasant affect Larsen amp Diener 1992) is unique to depression(Clark amp Watson 1991) Indeed in designing a questionnaire to measureboth general and specireg c components of depression and anxiety Watsonand colleagues (Watson 1995ab) developed a specireg c depression scalethat primarily measures anhedonia Because anhedonia is a fundamentalcharacteristic of melancholia it seems reasonable to assume that decreasedpleasant affect would accompany this type of depression If so Hellerrsquo smodel which predicts less left than right anterior activity for decreasedpleasant affect is completely consistent with the available data
However Davidson et alrsquo s (1995) data suggests the possibility thataffective valence may depend less on the relative balance of activationbetween the left and right anterior regions and more on specireg c systemsassociated with each hemisphere Whereas melancholic depression ischaracterised by decreased pleasant affect depression without melan-cholic features may be better characterised by increased unpleasantaffect Following this line of reasoning Hellerrsquo s model which in itscurrent form argues for a link between anterior asymmetry and the valencedimension would not distinguish between melancholic and nonmelan-cholic depression If affective valence depends more on specireg c systemsassociated with each hemisphere as Davidson et alrsquo s data suggest Hellerrsquo smodel may need to be expanded to remacr ect the unique contributions of theleft and right hemispheres to pleasantunpleasant affect For example itmight be more accurate to link the left anterior region with the activatedpleasant affect (ie positive affect) octant of the circumplex model ofemotion and the right anterior region with the activated unpleasant affect(ie negative affect) octant (see Larson amp Diener 1992)
Because few studies have specireg cally examined the full range of poten-tial affective descriptors in relation to patterns of brain activity theseissues remain to be resolved In most research only isolated portions ofthe circumplex have been sampled For example because positive andnegative affect as measured by the PANAS provide information regardingonly two of the four quadrants in the circumplex model studies that do notcollect data regarding the other two quadrants cannot shed light on howwell they might have accounted for patterns of brain activity
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 433
Anxiety As reviewed earlier studies that examine regional brainactivity in anxiety are even more plagued by inconsistencies than studiesof depression In keeping with the theme of this article we have argued thatby decomposing anxiety into subtypes on the basis of the psychologicalliterature we can potentially resolve some of these discrepancies
In previous work we have suggested that it may be important toconsider the type of anxiety that is being examined (Heller et al 19951997a) In particular we argued that regional brain activity might bedifferent for anxious arousal (eg panic) and anxious apprehension (egworry) types of anxiety that appear to differ signireg cantly in psychologicaland physiological characteristics (eg Barlow 1988 1991 Eysenck 1957Klein 1987 Watson et al 1995b) Anxious arousal has been described byWatson et al as being distinguished by symptoms of physiological hyper-arousal and somatic tension Panic attacks and high-stress states would fallinto this category In contrast anxious apprehension involves worry and ischaracterised by verbal rumination typically about future events This typeof anxiety would be characteristic of obsessive-compulsiveness general-ised anxiety states and trait anxiety (as identireg ed by self-reports of anxiousapprehension and worry on various questionnaires) As noted by Heller etal (1997a) the two kinds of anxiety are not mutually exclusive Extremedegrees of anxious apprehension (eg being worried or fearful about thefuture) may prompt experiences of anxious arousal (eg somatic symptomsand exaggerated physiological responses to stressful events)
We have argued that when the literature on brain activity in anxiety isreviewed in the light of this distinction between anxious arousal andanxious apprehension most of the studies reporting greater right hemi-sphere activity have either studied panic attacks (eg Reiman et al 1984)or high-stress situations likely to entail anxious arousal (eg Tucker et al1977) On the other hand studies indicating greater left hemisphere activity(primarily anterior) have examined OCD (eg Baxter et al 1987) GAD(Wu et al 1991) or questionnaire-dereg ned trait anxiety indexing anxiousapprehension (eg Tucker et al 1978) Indeed as mentioned earlierstudies have highlighted various regions in both hemispheres As aresult it is difreg cult to pinpoint a particular region that is clearly a focalpoint in anxiety Nonetheless the data do suggest that right hemisphereregions emerge as particularly salient in some types of anxiety (ie anxiousarousal) whereas left hemisphere regions emerge as particularly salient inother types of anxiety (ie anxious apprehension)
In a direct test of this hypothesis (Heller et al 1997a) we selectedparticipants on the basis of self-reported trait anxiety using a measureindexing anxious apprehension in a between-subjects design We thenmanipulated anxious arousal on a within-subjects basis Specireg cally wecontrasted brain activity (EEG alpha) during rest periods to brain activity
434 HELLER AND NITSCHKE
during a task in which participants listened to emotional narrativesdesigned to elicit anxious arousal (ie fear and sad narratives that hadpreviously received unpleasant valence and high arousal ratings)
For frontal activity group differences were unaffected by the narrativesAnxious participants showed a larger asymmetry in favour of the lefthemisphere than controls during the rest and listen periods of all narrativetypes as well as for eight minutes of resting baseline preceding thenarrative task Examination of the data indicated that the increased asym-metry was primarily due to a decrease in right hemisphere activity for theanxious participants Nonetheless the increase in magnitude of the asym-metry favouring the left hemisphere found for anxious participants isconsistent with the studies reviewed earlier that reported or inferredrelatively greater left than right anterior activity in anxious subjectsclassireg ed on the basis of worry or trait anxiety (eg Carter et al 1986Tucker et al 1978 Tyler amp Tucker 1982) Our frontal data suggest that itwill be important to investigate whether anxious apprehension is bestcharacterised by the magnitude of the asymmetry favouring the left hemi-sphere or by the level of left hemisphere activity in isolation
In contrast the parietal data supported our prediction regarding theassociation of anxious arousal with greater right hemisphere activityWhen listening to the fear and sad narratives anxious participants showeda selective increase in right parietal activity that was not demonstrated bycontrols The right parietal effect however was not obtained for neutraland happy narratives that had previously received pleasant valence and lowarousal ratings These reg ndings are consistent with those studies that havereported or inferred greater right hemisphere activity during anxious statescharacterised by panic and stress-inducing conditions (eg Reiman et al1984 Stewart et al 1988 Tucker et al 1977)
An important question here has to do with the discrepancies in frontalasymmetries that have been reported Like others (eg Baxter et al 1987Gur et al 1988 Wu et al 1991) we did not reg nd heightened right anterioractivity in anxiety However a number of other studies have (eg Davidsonet al 1996 Reivich et al 1983 Tomarken amp Davidson 1994) Asdiscussed earlier we have previously argued that anterior regions areuniquely involved in the valence dimension of emotion with pleasantaffect associated with greater left activity and unpleasant affect associatedwith greater right activity (eg Heller 1990) Therefore we would expectheightened right anterior activity in an anxious group only if they differedfrom controls on measures indexing affective valence In our recent studythe nondepressed anxious group did not differ from controls on either theGBI Dysthymia scale or the PANAS PA and NA scales
It may also be important to measure the level of anxious apprehension inthe sample studied For example Tomarken and Davidson (1994) found
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 435
that with Marloweplusmn Crowne scores controlled trait anxiety failed to predictfrontal asymmetry However their groups were divided according to amedian split
In summary our results directly support the proposed distinctionbetween anxious apprehension and anxious arousal as types of anxietythat may account for signireg cant variance in patterns of brain activityThey suggest furthermore that Hellerrsquo s model should be rereg ned to reflectthe likelihood that parietal activation will covary with anxious arousal butnot with anxious apprehension If these distinctions are accurate we wouldexpect to observe increases in right parietotemporal activity only underconditions when anxiety is accompanied by increased arousal
Thus in homodynamic and EEG studies a failure to reg nd increased rightposterior activity in anxiety is not surprising as variations in anxiousarousal are likely to be large Some individuals may experience anxiousarousal frequently especially those with disorders more characteristic ofanxious arousal such as panic disorder phobias or post-traumatic stressdisorder Others although apprehensive may rarely experience anxiousarousal In addition the situation may vary in the degree to which it elicitsanxious arousal Being forced to give a speech is likely to induce a greatdeal of anxious arousal in social phobics thus we would have predicted asignireg cant increase in right parietal activity precisely as found by David-son et al (submitted) Note that this effect was only found during theAnticipation condition not during baseline Planning or after thespeech During the Anticipation condition subjects listened while a voiceannounced every 30 seconds for three minutes the time remaining beforethey would be given a topic and allowed to plan for their speech Pre-sumably this period is the most stressful and the most likely to result inanxious arousal Indeed the correlation between heart rate acceleration tothe Anticipation condition and the increase in state anxiety was signireg cant
More recently Nitschke Heller Imig and Miller (submitted) haveattempted to identify anxious arousal and anxious apprehension usingquestionnaires Data from 783 students for a battery of anxiety and depres-sion questionnaires provided strong support for the anxiety distinctionproposed here at least on a psychological level of analysis The Moodand Anxiety Symptom Questionnaire-Anxious Arousal scale (MASQWatson et al 1995ab) was chosen as a measure of anxious arousalwhereas the Penn State Worry Questionnaire (PSWQ Meyer MillerMetzger amp Borkovec 1990 Molina amp Borkovec 1994) was included asa measure of anxious apprehension The correlation between the two scales(r = 20) was much lower than those involving the other more generalscales of anxiety included in the study (MASQ-General Anxiety scaleBeck Anxiety Inventory and the State-Trait Anxiety Inventory-Traitscale ranging from r = 47 to r = 65) These results further support the
436 HELLER AND NITSCHKE
argument that anxious arousal and anxious apprehension are distinctanxiety states
However new questions are highlighted by this distinction betweenanxious arousal and anxious apprehension particularly in the light ofHellerrsquo s model If anxious apprehension is associated with left anterioractivity the overall pattern of regional brain activity appears to be identicalto the brain activity associated with pleasant emotional states (ie more leftthan right anterior activity) However anxious apprehension might beexpected to fall in the portion of the circumplex model of emotionassociated with activated unpleasant affect (as suggested by adjectivessuch as fearful nervous jittery anxious) Below we suggest a numberof theoretical possibilities for resolving this apparent discrepancy all ofwhich remain to be examined empirically
In general it will be important to clarify the degree to which anxiousapprehension is in fact characterised by unpleasant versus pleasant valenceas well as high versus low arousal In addition it may be helpful toconsider these anterior asymmetries from the approach-withdrawal frame-work employed by Davidson (1992ab Davidson amp Tomarken 1989)Under conditions where anxious apprehension leads to approach tenden-cies one might expect an asymmetry in favour of the left anterior region Incontrast when anxious apprehension is associated with withdrawal tenden-cies (perhaps as anxious arousal increases) an asymmetry may emerge infavour of the right anterior region Conceivably the degree to whichanxious apprehension is characterised by approach versus withdrawaltendencies would depend on the degree of stress and discomfortinvolved the personrsquo s coping style and other factors With regard tovalence we would predict that anxious apprehension characterised byapproach tendencies would be accompanied by relatively pleasant affectAnxious apprehension accompanied by withdrawal tendencies in contrastwould be accompanied by relatively unpleasant affect
It is also possible that instead of a direct link between anxious appre-hension and approach behaviour they share more fundamental character-istics associated with left anterior functions For example one might arguethat they both involve sequential thinking and the development of hier-archically structured plans focusing on future concerns (for a review ofrelevant left anterior functions see Tomarken amp Keener this Issue)
Another important issue has to do with the nature of the asymmetricpattern of activity As reviewed earlier we found that the asymmetry infavour of the left hemisphere for participants classireg ed on the basis of traitanxiety was due to a relative decrease in right hemisphere activity Thusthe left frontal region of the anxious participants was not signireg cantly moreactive than that of controls It will therefore be important to examine thepossibility that absolute activity of the left anterior region plays a more
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 437
important role in the modulation of valence (eg Davidson et al 1995)than of anxious apprehension
Another possibility is that independent left anterior systems are asso-ciated with anxious apprehension versus pleasant affect Liotti and Tucker(1995) review theories that there may be multiple functionally segregatedparallel frontal cortico-striato-thalamic circuits some associated withmotor functions some with set shifting and behavioural inhibition andsome with emotional information processing and hedonic responsivity tostimuli Possibly one of these circuits remacr ects activation associated withpleasant affect and another remacr ects activation associated with anxiousapprehension andor approach Different neurotransmitter systems couldbe involved More precise methods of localisation would therefore benecessary to distinguish the unique brain activity associated with each ofthese psychologicalbehavioural phenomena
Comorbidity of Depression and Anxiety
In considering the issue of decomposing emotional states with regard to thesymptomatology in affective and anxiety disorders we have previouslyargued that issues of comorbidity of depression and anxiety are crucial(Heller et al 1995) A review of the literature reveals that rates of comor-bidity are extremely high (Alloy Kelly Mineka amp Clements 1990 Helleret al 1995 Hiller et al 1989 Katon amp Roy-Byrne 1991) Even in our veryrecent work (Nitschke et al submitted) using the MASQ a measuredesigned by Watson and colleagues to avoid symptom overlap in themeasurement of anxiety and depression (Watson et al 1995ab) comor-bidity was fairly pronounced in a large college student sample The MASQAnxious Arousal (AA) and Anhedonic Depression (AD) scales were used asmeasures of those constructs and the PSWQ provided the measure ofanxious apprehension Of the 125 (out of 783) participants scoring abovethe 80th percentile on at least one of the three scales of interest and belowthe 50th percentile on any where the 80th percentile cut-off was not metfully 50 demonstrated comorbidity by scoring above the 80th percentileon at least two of the scales This rate of comorbidity emerged despite thelow correlations found between the three scales (MASQ AA and MASQAD r = 20 MASQ AA and PSWQ r = 20 MASQ AD and PSWQ r =36) relative to the correlations found for the other more general measuresof depression including the MASQ General Mixed General Depression andGeneral Anxiety scales Beck Depression Inventory Beck Anxiety Inven-tory General Behaviour Inventory-Dysthymia scale and the State-TraitAnxiety Inventory-Trait scale (ranging from r = 47 to r = 76)
Because at least some studies have found opposite patterns of activityfor the right posterior region during depression versus during anxiety (ie
438 HELLER AND NITSCHKE
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
Subtypes of Depression and Anxiety
Another promising avenue for elucidating the discrepancies reported in theliterature for both depression and anxiety is the analysis of subtypes
Depression Many distinctions for depression have been proposed overthe years including most notably bipolar versus unipolar endogenousversus exogenous and neurotic versus reactive The current Diagnostic andstatistical manual of mental disordersplusmn fourth edition (DSMIV APA 1994)subdivides unipolar depression into major depressive disorder and dysthy-mic disorder and provides further specireg city with criteria for severity andfor psychotic melancholic catatonic or atypical features
Perhaps the most useful distinction in physiological research on depres-sion has been the comparison of melancholic to nonmelancholic depres-sion The dereg ning feature of melancholic depression is anhedonia which isthe inability to experience pleasure Current consensus as indicated by theDSMIV (APA 1994) suggests that melancholia is most strongly charac-terised by a loss of pleasure in all activities and a lack of responsivity topleasurable stimuli Many other symptoms may be shared by both melan-cholic and nonmelancholic depression including depressed mood distur-bances in eating and sleeping patterns low self-esteem inability toconcentrate psychomotor retardation fatigue indecisiveness feelings ofguilt and thoughts about death There is current controversy over whethermelancholic depression is qualitatively different from nonmelancholicdepression with some authors arguing that the primary distinction is oneof severity (Zimmerman Coryell amp Pfohl 1986) Nonetheless neuropsy-chological and neurophysiological studies have been consistent in demon-strating differences between melancholic and nonmelancholic depression
In a comprehensive series of studies Bruder has shown that perceptualasymmetries differ for different subtypes of depression (for a review seeBruder 1995) On a perceptual asymmetry task melancholic patientsshowed an abnormally large right ear (left hemisphere) advantage for adichotic syllables task but no left ear (right hemisphere) advantage for acomplex tone task (Bruder 1995) The lack of a left ear advantage was dueto poor left ear accuracy consistent with other studies demonstrating righthemisphere dereg cits in depressed people However nonmelancholicdepressed patients did not differ from normal controls on either test Alongthese lines in a PET study of melancholic depression Bench et al (1992)reported less activity in the left dorsolateral prefrontal cortex compared tonondepressed controls Note that these data could be consistent with eithera qualitative or a quantitative distinction between melancholic and non-melancholic depression However other data may be more consistent witha qualitative difference between the two types of depression For example
432 HELLER AND NITSCHKE
Davidson Larson and Abercrombie (1995) reported that depression sever-ity in melancholic subjects was predicted by the extent to which activationwas decreased in the left lateral prefrontal region In contrast amongnonmelancholics depression severity was predicted by the extent to whichactivation was increased in the right medial prefrontal area
Davidson et alrsquo s (1995) data have implications for theories regardingthe relationship between pleasantunpleasant affect and anterior regions ofthe brain Research on the psychological characteristics of depression andanxiety has suggested that lack of positive affect (or more preciselyactivated pleasant affect Larsen amp Diener 1992) is unique to depression(Clark amp Watson 1991) Indeed in designing a questionnaire to measureboth general and specireg c components of depression and anxiety Watsonand colleagues (Watson 1995ab) developed a specireg c depression scalethat primarily measures anhedonia Because anhedonia is a fundamentalcharacteristic of melancholia it seems reasonable to assume that decreasedpleasant affect would accompany this type of depression If so Hellerrsquo smodel which predicts less left than right anterior activity for decreasedpleasant affect is completely consistent with the available data
However Davidson et alrsquo s (1995) data suggests the possibility thataffective valence may depend less on the relative balance of activationbetween the left and right anterior regions and more on specireg c systemsassociated with each hemisphere Whereas melancholic depression ischaracterised by decreased pleasant affect depression without melan-cholic features may be better characterised by increased unpleasantaffect Following this line of reasoning Hellerrsquo s model which in itscurrent form argues for a link between anterior asymmetry and the valencedimension would not distinguish between melancholic and nonmelan-cholic depression If affective valence depends more on specireg c systemsassociated with each hemisphere as Davidson et alrsquo s data suggest Hellerrsquo smodel may need to be expanded to remacr ect the unique contributions of theleft and right hemispheres to pleasantunpleasant affect For example itmight be more accurate to link the left anterior region with the activatedpleasant affect (ie positive affect) octant of the circumplex model ofemotion and the right anterior region with the activated unpleasant affect(ie negative affect) octant (see Larson amp Diener 1992)
Because few studies have specireg cally examined the full range of poten-tial affective descriptors in relation to patterns of brain activity theseissues remain to be resolved In most research only isolated portions ofthe circumplex have been sampled For example because positive andnegative affect as measured by the PANAS provide information regardingonly two of the four quadrants in the circumplex model studies that do notcollect data regarding the other two quadrants cannot shed light on howwell they might have accounted for patterns of brain activity
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 433
Anxiety As reviewed earlier studies that examine regional brainactivity in anxiety are even more plagued by inconsistencies than studiesof depression In keeping with the theme of this article we have argued thatby decomposing anxiety into subtypes on the basis of the psychologicalliterature we can potentially resolve some of these discrepancies
In previous work we have suggested that it may be important toconsider the type of anxiety that is being examined (Heller et al 19951997a) In particular we argued that regional brain activity might bedifferent for anxious arousal (eg panic) and anxious apprehension (egworry) types of anxiety that appear to differ signireg cantly in psychologicaland physiological characteristics (eg Barlow 1988 1991 Eysenck 1957Klein 1987 Watson et al 1995b) Anxious arousal has been described byWatson et al as being distinguished by symptoms of physiological hyper-arousal and somatic tension Panic attacks and high-stress states would fallinto this category In contrast anxious apprehension involves worry and ischaracterised by verbal rumination typically about future events This typeof anxiety would be characteristic of obsessive-compulsiveness general-ised anxiety states and trait anxiety (as identireg ed by self-reports of anxiousapprehension and worry on various questionnaires) As noted by Heller etal (1997a) the two kinds of anxiety are not mutually exclusive Extremedegrees of anxious apprehension (eg being worried or fearful about thefuture) may prompt experiences of anxious arousal (eg somatic symptomsand exaggerated physiological responses to stressful events)
We have argued that when the literature on brain activity in anxiety isreviewed in the light of this distinction between anxious arousal andanxious apprehension most of the studies reporting greater right hemi-sphere activity have either studied panic attacks (eg Reiman et al 1984)or high-stress situations likely to entail anxious arousal (eg Tucker et al1977) On the other hand studies indicating greater left hemisphere activity(primarily anterior) have examined OCD (eg Baxter et al 1987) GAD(Wu et al 1991) or questionnaire-dereg ned trait anxiety indexing anxiousapprehension (eg Tucker et al 1978) Indeed as mentioned earlierstudies have highlighted various regions in both hemispheres As aresult it is difreg cult to pinpoint a particular region that is clearly a focalpoint in anxiety Nonetheless the data do suggest that right hemisphereregions emerge as particularly salient in some types of anxiety (ie anxiousarousal) whereas left hemisphere regions emerge as particularly salient inother types of anxiety (ie anxious apprehension)
In a direct test of this hypothesis (Heller et al 1997a) we selectedparticipants on the basis of self-reported trait anxiety using a measureindexing anxious apprehension in a between-subjects design We thenmanipulated anxious arousal on a within-subjects basis Specireg cally wecontrasted brain activity (EEG alpha) during rest periods to brain activity
434 HELLER AND NITSCHKE
during a task in which participants listened to emotional narrativesdesigned to elicit anxious arousal (ie fear and sad narratives that hadpreviously received unpleasant valence and high arousal ratings)
For frontal activity group differences were unaffected by the narrativesAnxious participants showed a larger asymmetry in favour of the lefthemisphere than controls during the rest and listen periods of all narrativetypes as well as for eight minutes of resting baseline preceding thenarrative task Examination of the data indicated that the increased asym-metry was primarily due to a decrease in right hemisphere activity for theanxious participants Nonetheless the increase in magnitude of the asym-metry favouring the left hemisphere found for anxious participants isconsistent with the studies reviewed earlier that reported or inferredrelatively greater left than right anterior activity in anxious subjectsclassireg ed on the basis of worry or trait anxiety (eg Carter et al 1986Tucker et al 1978 Tyler amp Tucker 1982) Our frontal data suggest that itwill be important to investigate whether anxious apprehension is bestcharacterised by the magnitude of the asymmetry favouring the left hemi-sphere or by the level of left hemisphere activity in isolation
In contrast the parietal data supported our prediction regarding theassociation of anxious arousal with greater right hemisphere activityWhen listening to the fear and sad narratives anxious participants showeda selective increase in right parietal activity that was not demonstrated bycontrols The right parietal effect however was not obtained for neutraland happy narratives that had previously received pleasant valence and lowarousal ratings These reg ndings are consistent with those studies that havereported or inferred greater right hemisphere activity during anxious statescharacterised by panic and stress-inducing conditions (eg Reiman et al1984 Stewart et al 1988 Tucker et al 1977)
An important question here has to do with the discrepancies in frontalasymmetries that have been reported Like others (eg Baxter et al 1987Gur et al 1988 Wu et al 1991) we did not reg nd heightened right anterioractivity in anxiety However a number of other studies have (eg Davidsonet al 1996 Reivich et al 1983 Tomarken amp Davidson 1994) Asdiscussed earlier we have previously argued that anterior regions areuniquely involved in the valence dimension of emotion with pleasantaffect associated with greater left activity and unpleasant affect associatedwith greater right activity (eg Heller 1990) Therefore we would expectheightened right anterior activity in an anxious group only if they differedfrom controls on measures indexing affective valence In our recent studythe nondepressed anxious group did not differ from controls on either theGBI Dysthymia scale or the PANAS PA and NA scales
It may also be important to measure the level of anxious apprehension inthe sample studied For example Tomarken and Davidson (1994) found
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 435
that with Marloweplusmn Crowne scores controlled trait anxiety failed to predictfrontal asymmetry However their groups were divided according to amedian split
In summary our results directly support the proposed distinctionbetween anxious apprehension and anxious arousal as types of anxietythat may account for signireg cant variance in patterns of brain activityThey suggest furthermore that Hellerrsquo s model should be rereg ned to reflectthe likelihood that parietal activation will covary with anxious arousal butnot with anxious apprehension If these distinctions are accurate we wouldexpect to observe increases in right parietotemporal activity only underconditions when anxiety is accompanied by increased arousal
Thus in homodynamic and EEG studies a failure to reg nd increased rightposterior activity in anxiety is not surprising as variations in anxiousarousal are likely to be large Some individuals may experience anxiousarousal frequently especially those with disorders more characteristic ofanxious arousal such as panic disorder phobias or post-traumatic stressdisorder Others although apprehensive may rarely experience anxiousarousal In addition the situation may vary in the degree to which it elicitsanxious arousal Being forced to give a speech is likely to induce a greatdeal of anxious arousal in social phobics thus we would have predicted asignireg cant increase in right parietal activity precisely as found by David-son et al (submitted) Note that this effect was only found during theAnticipation condition not during baseline Planning or after thespeech During the Anticipation condition subjects listened while a voiceannounced every 30 seconds for three minutes the time remaining beforethey would be given a topic and allowed to plan for their speech Pre-sumably this period is the most stressful and the most likely to result inanxious arousal Indeed the correlation between heart rate acceleration tothe Anticipation condition and the increase in state anxiety was signireg cant
More recently Nitschke Heller Imig and Miller (submitted) haveattempted to identify anxious arousal and anxious apprehension usingquestionnaires Data from 783 students for a battery of anxiety and depres-sion questionnaires provided strong support for the anxiety distinctionproposed here at least on a psychological level of analysis The Moodand Anxiety Symptom Questionnaire-Anxious Arousal scale (MASQWatson et al 1995ab) was chosen as a measure of anxious arousalwhereas the Penn State Worry Questionnaire (PSWQ Meyer MillerMetzger amp Borkovec 1990 Molina amp Borkovec 1994) was included asa measure of anxious apprehension The correlation between the two scales(r = 20) was much lower than those involving the other more generalscales of anxiety included in the study (MASQ-General Anxiety scaleBeck Anxiety Inventory and the State-Trait Anxiety Inventory-Traitscale ranging from r = 47 to r = 65) These results further support the
436 HELLER AND NITSCHKE
argument that anxious arousal and anxious apprehension are distinctanxiety states
However new questions are highlighted by this distinction betweenanxious arousal and anxious apprehension particularly in the light ofHellerrsquo s model If anxious apprehension is associated with left anterioractivity the overall pattern of regional brain activity appears to be identicalto the brain activity associated with pleasant emotional states (ie more leftthan right anterior activity) However anxious apprehension might beexpected to fall in the portion of the circumplex model of emotionassociated with activated unpleasant affect (as suggested by adjectivessuch as fearful nervous jittery anxious) Below we suggest a numberof theoretical possibilities for resolving this apparent discrepancy all ofwhich remain to be examined empirically
In general it will be important to clarify the degree to which anxiousapprehension is in fact characterised by unpleasant versus pleasant valenceas well as high versus low arousal In addition it may be helpful toconsider these anterior asymmetries from the approach-withdrawal frame-work employed by Davidson (1992ab Davidson amp Tomarken 1989)Under conditions where anxious apprehension leads to approach tenden-cies one might expect an asymmetry in favour of the left anterior region Incontrast when anxious apprehension is associated with withdrawal tenden-cies (perhaps as anxious arousal increases) an asymmetry may emerge infavour of the right anterior region Conceivably the degree to whichanxious apprehension is characterised by approach versus withdrawaltendencies would depend on the degree of stress and discomfortinvolved the personrsquo s coping style and other factors With regard tovalence we would predict that anxious apprehension characterised byapproach tendencies would be accompanied by relatively pleasant affectAnxious apprehension accompanied by withdrawal tendencies in contrastwould be accompanied by relatively unpleasant affect
It is also possible that instead of a direct link between anxious appre-hension and approach behaviour they share more fundamental character-istics associated with left anterior functions For example one might arguethat they both involve sequential thinking and the development of hier-archically structured plans focusing on future concerns (for a review ofrelevant left anterior functions see Tomarken amp Keener this Issue)
Another important issue has to do with the nature of the asymmetricpattern of activity As reviewed earlier we found that the asymmetry infavour of the left hemisphere for participants classireg ed on the basis of traitanxiety was due to a relative decrease in right hemisphere activity Thusthe left frontal region of the anxious participants was not signireg cantly moreactive than that of controls It will therefore be important to examine thepossibility that absolute activity of the left anterior region plays a more
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 437
important role in the modulation of valence (eg Davidson et al 1995)than of anxious apprehension
Another possibility is that independent left anterior systems are asso-ciated with anxious apprehension versus pleasant affect Liotti and Tucker(1995) review theories that there may be multiple functionally segregatedparallel frontal cortico-striato-thalamic circuits some associated withmotor functions some with set shifting and behavioural inhibition andsome with emotional information processing and hedonic responsivity tostimuli Possibly one of these circuits remacr ects activation associated withpleasant affect and another remacr ects activation associated with anxiousapprehension andor approach Different neurotransmitter systems couldbe involved More precise methods of localisation would therefore benecessary to distinguish the unique brain activity associated with each ofthese psychologicalbehavioural phenomena
Comorbidity of Depression and Anxiety
In considering the issue of decomposing emotional states with regard to thesymptomatology in affective and anxiety disorders we have previouslyargued that issues of comorbidity of depression and anxiety are crucial(Heller et al 1995) A review of the literature reveals that rates of comor-bidity are extremely high (Alloy Kelly Mineka amp Clements 1990 Helleret al 1995 Hiller et al 1989 Katon amp Roy-Byrne 1991) Even in our veryrecent work (Nitschke et al submitted) using the MASQ a measuredesigned by Watson and colleagues to avoid symptom overlap in themeasurement of anxiety and depression (Watson et al 1995ab) comor-bidity was fairly pronounced in a large college student sample The MASQAnxious Arousal (AA) and Anhedonic Depression (AD) scales were used asmeasures of those constructs and the PSWQ provided the measure ofanxious apprehension Of the 125 (out of 783) participants scoring abovethe 80th percentile on at least one of the three scales of interest and belowthe 50th percentile on any where the 80th percentile cut-off was not metfully 50 demonstrated comorbidity by scoring above the 80th percentileon at least two of the scales This rate of comorbidity emerged despite thelow correlations found between the three scales (MASQ AA and MASQAD r = 20 MASQ AA and PSWQ r = 20 MASQ AD and PSWQ r =36) relative to the correlations found for the other more general measuresof depression including the MASQ General Mixed General Depression andGeneral Anxiety scales Beck Depression Inventory Beck Anxiety Inven-tory General Behaviour Inventory-Dysthymia scale and the State-TraitAnxiety Inventory-Trait scale (ranging from r = 47 to r = 76)
Because at least some studies have found opposite patterns of activityfor the right posterior region during depression versus during anxiety (ie
438 HELLER AND NITSCHKE
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
Davidson Larson and Abercrombie (1995) reported that depression sever-ity in melancholic subjects was predicted by the extent to which activationwas decreased in the left lateral prefrontal region In contrast amongnonmelancholics depression severity was predicted by the extent to whichactivation was increased in the right medial prefrontal area
Davidson et alrsquo s (1995) data have implications for theories regardingthe relationship between pleasantunpleasant affect and anterior regions ofthe brain Research on the psychological characteristics of depression andanxiety has suggested that lack of positive affect (or more preciselyactivated pleasant affect Larsen amp Diener 1992) is unique to depression(Clark amp Watson 1991) Indeed in designing a questionnaire to measureboth general and specireg c components of depression and anxiety Watsonand colleagues (Watson 1995ab) developed a specireg c depression scalethat primarily measures anhedonia Because anhedonia is a fundamentalcharacteristic of melancholia it seems reasonable to assume that decreasedpleasant affect would accompany this type of depression If so Hellerrsquo smodel which predicts less left than right anterior activity for decreasedpleasant affect is completely consistent with the available data
However Davidson et alrsquo s (1995) data suggests the possibility thataffective valence may depend less on the relative balance of activationbetween the left and right anterior regions and more on specireg c systemsassociated with each hemisphere Whereas melancholic depression ischaracterised by decreased pleasant affect depression without melan-cholic features may be better characterised by increased unpleasantaffect Following this line of reasoning Hellerrsquo s model which in itscurrent form argues for a link between anterior asymmetry and the valencedimension would not distinguish between melancholic and nonmelan-cholic depression If affective valence depends more on specireg c systemsassociated with each hemisphere as Davidson et alrsquo s data suggest Hellerrsquo smodel may need to be expanded to remacr ect the unique contributions of theleft and right hemispheres to pleasantunpleasant affect For example itmight be more accurate to link the left anterior region with the activatedpleasant affect (ie positive affect) octant of the circumplex model ofemotion and the right anterior region with the activated unpleasant affect(ie negative affect) octant (see Larson amp Diener 1992)
Because few studies have specireg cally examined the full range of poten-tial affective descriptors in relation to patterns of brain activity theseissues remain to be resolved In most research only isolated portions ofthe circumplex have been sampled For example because positive andnegative affect as measured by the PANAS provide information regardingonly two of the four quadrants in the circumplex model studies that do notcollect data regarding the other two quadrants cannot shed light on howwell they might have accounted for patterns of brain activity
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 433
Anxiety As reviewed earlier studies that examine regional brainactivity in anxiety are even more plagued by inconsistencies than studiesof depression In keeping with the theme of this article we have argued thatby decomposing anxiety into subtypes on the basis of the psychologicalliterature we can potentially resolve some of these discrepancies
In previous work we have suggested that it may be important toconsider the type of anxiety that is being examined (Heller et al 19951997a) In particular we argued that regional brain activity might bedifferent for anxious arousal (eg panic) and anxious apprehension (egworry) types of anxiety that appear to differ signireg cantly in psychologicaland physiological characteristics (eg Barlow 1988 1991 Eysenck 1957Klein 1987 Watson et al 1995b) Anxious arousal has been described byWatson et al as being distinguished by symptoms of physiological hyper-arousal and somatic tension Panic attacks and high-stress states would fallinto this category In contrast anxious apprehension involves worry and ischaracterised by verbal rumination typically about future events This typeof anxiety would be characteristic of obsessive-compulsiveness general-ised anxiety states and trait anxiety (as identireg ed by self-reports of anxiousapprehension and worry on various questionnaires) As noted by Heller etal (1997a) the two kinds of anxiety are not mutually exclusive Extremedegrees of anxious apprehension (eg being worried or fearful about thefuture) may prompt experiences of anxious arousal (eg somatic symptomsand exaggerated physiological responses to stressful events)
We have argued that when the literature on brain activity in anxiety isreviewed in the light of this distinction between anxious arousal andanxious apprehension most of the studies reporting greater right hemi-sphere activity have either studied panic attacks (eg Reiman et al 1984)or high-stress situations likely to entail anxious arousal (eg Tucker et al1977) On the other hand studies indicating greater left hemisphere activity(primarily anterior) have examined OCD (eg Baxter et al 1987) GAD(Wu et al 1991) or questionnaire-dereg ned trait anxiety indexing anxiousapprehension (eg Tucker et al 1978) Indeed as mentioned earlierstudies have highlighted various regions in both hemispheres As aresult it is difreg cult to pinpoint a particular region that is clearly a focalpoint in anxiety Nonetheless the data do suggest that right hemisphereregions emerge as particularly salient in some types of anxiety (ie anxiousarousal) whereas left hemisphere regions emerge as particularly salient inother types of anxiety (ie anxious apprehension)
In a direct test of this hypothesis (Heller et al 1997a) we selectedparticipants on the basis of self-reported trait anxiety using a measureindexing anxious apprehension in a between-subjects design We thenmanipulated anxious arousal on a within-subjects basis Specireg cally wecontrasted brain activity (EEG alpha) during rest periods to brain activity
434 HELLER AND NITSCHKE
during a task in which participants listened to emotional narrativesdesigned to elicit anxious arousal (ie fear and sad narratives that hadpreviously received unpleasant valence and high arousal ratings)
For frontal activity group differences were unaffected by the narrativesAnxious participants showed a larger asymmetry in favour of the lefthemisphere than controls during the rest and listen periods of all narrativetypes as well as for eight minutes of resting baseline preceding thenarrative task Examination of the data indicated that the increased asym-metry was primarily due to a decrease in right hemisphere activity for theanxious participants Nonetheless the increase in magnitude of the asym-metry favouring the left hemisphere found for anxious participants isconsistent with the studies reviewed earlier that reported or inferredrelatively greater left than right anterior activity in anxious subjectsclassireg ed on the basis of worry or trait anxiety (eg Carter et al 1986Tucker et al 1978 Tyler amp Tucker 1982) Our frontal data suggest that itwill be important to investigate whether anxious apprehension is bestcharacterised by the magnitude of the asymmetry favouring the left hemi-sphere or by the level of left hemisphere activity in isolation
In contrast the parietal data supported our prediction regarding theassociation of anxious arousal with greater right hemisphere activityWhen listening to the fear and sad narratives anxious participants showeda selective increase in right parietal activity that was not demonstrated bycontrols The right parietal effect however was not obtained for neutraland happy narratives that had previously received pleasant valence and lowarousal ratings These reg ndings are consistent with those studies that havereported or inferred greater right hemisphere activity during anxious statescharacterised by panic and stress-inducing conditions (eg Reiman et al1984 Stewart et al 1988 Tucker et al 1977)
An important question here has to do with the discrepancies in frontalasymmetries that have been reported Like others (eg Baxter et al 1987Gur et al 1988 Wu et al 1991) we did not reg nd heightened right anterioractivity in anxiety However a number of other studies have (eg Davidsonet al 1996 Reivich et al 1983 Tomarken amp Davidson 1994) Asdiscussed earlier we have previously argued that anterior regions areuniquely involved in the valence dimension of emotion with pleasantaffect associated with greater left activity and unpleasant affect associatedwith greater right activity (eg Heller 1990) Therefore we would expectheightened right anterior activity in an anxious group only if they differedfrom controls on measures indexing affective valence In our recent studythe nondepressed anxious group did not differ from controls on either theGBI Dysthymia scale or the PANAS PA and NA scales
It may also be important to measure the level of anxious apprehension inthe sample studied For example Tomarken and Davidson (1994) found
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 435
that with Marloweplusmn Crowne scores controlled trait anxiety failed to predictfrontal asymmetry However their groups were divided according to amedian split
In summary our results directly support the proposed distinctionbetween anxious apprehension and anxious arousal as types of anxietythat may account for signireg cant variance in patterns of brain activityThey suggest furthermore that Hellerrsquo s model should be rereg ned to reflectthe likelihood that parietal activation will covary with anxious arousal butnot with anxious apprehension If these distinctions are accurate we wouldexpect to observe increases in right parietotemporal activity only underconditions when anxiety is accompanied by increased arousal
Thus in homodynamic and EEG studies a failure to reg nd increased rightposterior activity in anxiety is not surprising as variations in anxiousarousal are likely to be large Some individuals may experience anxiousarousal frequently especially those with disorders more characteristic ofanxious arousal such as panic disorder phobias or post-traumatic stressdisorder Others although apprehensive may rarely experience anxiousarousal In addition the situation may vary in the degree to which it elicitsanxious arousal Being forced to give a speech is likely to induce a greatdeal of anxious arousal in social phobics thus we would have predicted asignireg cant increase in right parietal activity precisely as found by David-son et al (submitted) Note that this effect was only found during theAnticipation condition not during baseline Planning or after thespeech During the Anticipation condition subjects listened while a voiceannounced every 30 seconds for three minutes the time remaining beforethey would be given a topic and allowed to plan for their speech Pre-sumably this period is the most stressful and the most likely to result inanxious arousal Indeed the correlation between heart rate acceleration tothe Anticipation condition and the increase in state anxiety was signireg cant
More recently Nitschke Heller Imig and Miller (submitted) haveattempted to identify anxious arousal and anxious apprehension usingquestionnaires Data from 783 students for a battery of anxiety and depres-sion questionnaires provided strong support for the anxiety distinctionproposed here at least on a psychological level of analysis The Moodand Anxiety Symptom Questionnaire-Anxious Arousal scale (MASQWatson et al 1995ab) was chosen as a measure of anxious arousalwhereas the Penn State Worry Questionnaire (PSWQ Meyer MillerMetzger amp Borkovec 1990 Molina amp Borkovec 1994) was included asa measure of anxious apprehension The correlation between the two scales(r = 20) was much lower than those involving the other more generalscales of anxiety included in the study (MASQ-General Anxiety scaleBeck Anxiety Inventory and the State-Trait Anxiety Inventory-Traitscale ranging from r = 47 to r = 65) These results further support the
436 HELLER AND NITSCHKE
argument that anxious arousal and anxious apprehension are distinctanxiety states
However new questions are highlighted by this distinction betweenanxious arousal and anxious apprehension particularly in the light ofHellerrsquo s model If anxious apprehension is associated with left anterioractivity the overall pattern of regional brain activity appears to be identicalto the brain activity associated with pleasant emotional states (ie more leftthan right anterior activity) However anxious apprehension might beexpected to fall in the portion of the circumplex model of emotionassociated with activated unpleasant affect (as suggested by adjectivessuch as fearful nervous jittery anxious) Below we suggest a numberof theoretical possibilities for resolving this apparent discrepancy all ofwhich remain to be examined empirically
In general it will be important to clarify the degree to which anxiousapprehension is in fact characterised by unpleasant versus pleasant valenceas well as high versus low arousal In addition it may be helpful toconsider these anterior asymmetries from the approach-withdrawal frame-work employed by Davidson (1992ab Davidson amp Tomarken 1989)Under conditions where anxious apprehension leads to approach tenden-cies one might expect an asymmetry in favour of the left anterior region Incontrast when anxious apprehension is associated with withdrawal tenden-cies (perhaps as anxious arousal increases) an asymmetry may emerge infavour of the right anterior region Conceivably the degree to whichanxious apprehension is characterised by approach versus withdrawaltendencies would depend on the degree of stress and discomfortinvolved the personrsquo s coping style and other factors With regard tovalence we would predict that anxious apprehension characterised byapproach tendencies would be accompanied by relatively pleasant affectAnxious apprehension accompanied by withdrawal tendencies in contrastwould be accompanied by relatively unpleasant affect
It is also possible that instead of a direct link between anxious appre-hension and approach behaviour they share more fundamental character-istics associated with left anterior functions For example one might arguethat they both involve sequential thinking and the development of hier-archically structured plans focusing on future concerns (for a review ofrelevant left anterior functions see Tomarken amp Keener this Issue)
Another important issue has to do with the nature of the asymmetricpattern of activity As reviewed earlier we found that the asymmetry infavour of the left hemisphere for participants classireg ed on the basis of traitanxiety was due to a relative decrease in right hemisphere activity Thusthe left frontal region of the anxious participants was not signireg cantly moreactive than that of controls It will therefore be important to examine thepossibility that absolute activity of the left anterior region plays a more
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 437
important role in the modulation of valence (eg Davidson et al 1995)than of anxious apprehension
Another possibility is that independent left anterior systems are asso-ciated with anxious apprehension versus pleasant affect Liotti and Tucker(1995) review theories that there may be multiple functionally segregatedparallel frontal cortico-striato-thalamic circuits some associated withmotor functions some with set shifting and behavioural inhibition andsome with emotional information processing and hedonic responsivity tostimuli Possibly one of these circuits remacr ects activation associated withpleasant affect and another remacr ects activation associated with anxiousapprehension andor approach Different neurotransmitter systems couldbe involved More precise methods of localisation would therefore benecessary to distinguish the unique brain activity associated with each ofthese psychologicalbehavioural phenomena
Comorbidity of Depression and Anxiety
In considering the issue of decomposing emotional states with regard to thesymptomatology in affective and anxiety disorders we have previouslyargued that issues of comorbidity of depression and anxiety are crucial(Heller et al 1995) A review of the literature reveals that rates of comor-bidity are extremely high (Alloy Kelly Mineka amp Clements 1990 Helleret al 1995 Hiller et al 1989 Katon amp Roy-Byrne 1991) Even in our veryrecent work (Nitschke et al submitted) using the MASQ a measuredesigned by Watson and colleagues to avoid symptom overlap in themeasurement of anxiety and depression (Watson et al 1995ab) comor-bidity was fairly pronounced in a large college student sample The MASQAnxious Arousal (AA) and Anhedonic Depression (AD) scales were used asmeasures of those constructs and the PSWQ provided the measure ofanxious apprehension Of the 125 (out of 783) participants scoring abovethe 80th percentile on at least one of the three scales of interest and belowthe 50th percentile on any where the 80th percentile cut-off was not metfully 50 demonstrated comorbidity by scoring above the 80th percentileon at least two of the scales This rate of comorbidity emerged despite thelow correlations found between the three scales (MASQ AA and MASQAD r = 20 MASQ AA and PSWQ r = 20 MASQ AD and PSWQ r =36) relative to the correlations found for the other more general measuresof depression including the MASQ General Mixed General Depression andGeneral Anxiety scales Beck Depression Inventory Beck Anxiety Inven-tory General Behaviour Inventory-Dysthymia scale and the State-TraitAnxiety Inventory-Trait scale (ranging from r = 47 to r = 76)
Because at least some studies have found opposite patterns of activityfor the right posterior region during depression versus during anxiety (ie
438 HELLER AND NITSCHKE
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
Anxiety As reviewed earlier studies that examine regional brainactivity in anxiety are even more plagued by inconsistencies than studiesof depression In keeping with the theme of this article we have argued thatby decomposing anxiety into subtypes on the basis of the psychologicalliterature we can potentially resolve some of these discrepancies
In previous work we have suggested that it may be important toconsider the type of anxiety that is being examined (Heller et al 19951997a) In particular we argued that regional brain activity might bedifferent for anxious arousal (eg panic) and anxious apprehension (egworry) types of anxiety that appear to differ signireg cantly in psychologicaland physiological characteristics (eg Barlow 1988 1991 Eysenck 1957Klein 1987 Watson et al 1995b) Anxious arousal has been described byWatson et al as being distinguished by symptoms of physiological hyper-arousal and somatic tension Panic attacks and high-stress states would fallinto this category In contrast anxious apprehension involves worry and ischaracterised by verbal rumination typically about future events This typeof anxiety would be characteristic of obsessive-compulsiveness general-ised anxiety states and trait anxiety (as identireg ed by self-reports of anxiousapprehension and worry on various questionnaires) As noted by Heller etal (1997a) the two kinds of anxiety are not mutually exclusive Extremedegrees of anxious apprehension (eg being worried or fearful about thefuture) may prompt experiences of anxious arousal (eg somatic symptomsand exaggerated physiological responses to stressful events)
We have argued that when the literature on brain activity in anxiety isreviewed in the light of this distinction between anxious arousal andanxious apprehension most of the studies reporting greater right hemi-sphere activity have either studied panic attacks (eg Reiman et al 1984)or high-stress situations likely to entail anxious arousal (eg Tucker et al1977) On the other hand studies indicating greater left hemisphere activity(primarily anterior) have examined OCD (eg Baxter et al 1987) GAD(Wu et al 1991) or questionnaire-dereg ned trait anxiety indexing anxiousapprehension (eg Tucker et al 1978) Indeed as mentioned earlierstudies have highlighted various regions in both hemispheres As aresult it is difreg cult to pinpoint a particular region that is clearly a focalpoint in anxiety Nonetheless the data do suggest that right hemisphereregions emerge as particularly salient in some types of anxiety (ie anxiousarousal) whereas left hemisphere regions emerge as particularly salient inother types of anxiety (ie anxious apprehension)
In a direct test of this hypothesis (Heller et al 1997a) we selectedparticipants on the basis of self-reported trait anxiety using a measureindexing anxious apprehension in a between-subjects design We thenmanipulated anxious arousal on a within-subjects basis Specireg cally wecontrasted brain activity (EEG alpha) during rest periods to brain activity
434 HELLER AND NITSCHKE
during a task in which participants listened to emotional narrativesdesigned to elicit anxious arousal (ie fear and sad narratives that hadpreviously received unpleasant valence and high arousal ratings)
For frontal activity group differences were unaffected by the narrativesAnxious participants showed a larger asymmetry in favour of the lefthemisphere than controls during the rest and listen periods of all narrativetypes as well as for eight minutes of resting baseline preceding thenarrative task Examination of the data indicated that the increased asym-metry was primarily due to a decrease in right hemisphere activity for theanxious participants Nonetheless the increase in magnitude of the asym-metry favouring the left hemisphere found for anxious participants isconsistent with the studies reviewed earlier that reported or inferredrelatively greater left than right anterior activity in anxious subjectsclassireg ed on the basis of worry or trait anxiety (eg Carter et al 1986Tucker et al 1978 Tyler amp Tucker 1982) Our frontal data suggest that itwill be important to investigate whether anxious apprehension is bestcharacterised by the magnitude of the asymmetry favouring the left hemi-sphere or by the level of left hemisphere activity in isolation
In contrast the parietal data supported our prediction regarding theassociation of anxious arousal with greater right hemisphere activityWhen listening to the fear and sad narratives anxious participants showeda selective increase in right parietal activity that was not demonstrated bycontrols The right parietal effect however was not obtained for neutraland happy narratives that had previously received pleasant valence and lowarousal ratings These reg ndings are consistent with those studies that havereported or inferred greater right hemisphere activity during anxious statescharacterised by panic and stress-inducing conditions (eg Reiman et al1984 Stewart et al 1988 Tucker et al 1977)
An important question here has to do with the discrepancies in frontalasymmetries that have been reported Like others (eg Baxter et al 1987Gur et al 1988 Wu et al 1991) we did not reg nd heightened right anterioractivity in anxiety However a number of other studies have (eg Davidsonet al 1996 Reivich et al 1983 Tomarken amp Davidson 1994) Asdiscussed earlier we have previously argued that anterior regions areuniquely involved in the valence dimension of emotion with pleasantaffect associated with greater left activity and unpleasant affect associatedwith greater right activity (eg Heller 1990) Therefore we would expectheightened right anterior activity in an anxious group only if they differedfrom controls on measures indexing affective valence In our recent studythe nondepressed anxious group did not differ from controls on either theGBI Dysthymia scale or the PANAS PA and NA scales
It may also be important to measure the level of anxious apprehension inthe sample studied For example Tomarken and Davidson (1994) found
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 435
that with Marloweplusmn Crowne scores controlled trait anxiety failed to predictfrontal asymmetry However their groups were divided according to amedian split
In summary our results directly support the proposed distinctionbetween anxious apprehension and anxious arousal as types of anxietythat may account for signireg cant variance in patterns of brain activityThey suggest furthermore that Hellerrsquo s model should be rereg ned to reflectthe likelihood that parietal activation will covary with anxious arousal butnot with anxious apprehension If these distinctions are accurate we wouldexpect to observe increases in right parietotemporal activity only underconditions when anxiety is accompanied by increased arousal
Thus in homodynamic and EEG studies a failure to reg nd increased rightposterior activity in anxiety is not surprising as variations in anxiousarousal are likely to be large Some individuals may experience anxiousarousal frequently especially those with disorders more characteristic ofanxious arousal such as panic disorder phobias or post-traumatic stressdisorder Others although apprehensive may rarely experience anxiousarousal In addition the situation may vary in the degree to which it elicitsanxious arousal Being forced to give a speech is likely to induce a greatdeal of anxious arousal in social phobics thus we would have predicted asignireg cant increase in right parietal activity precisely as found by David-son et al (submitted) Note that this effect was only found during theAnticipation condition not during baseline Planning or after thespeech During the Anticipation condition subjects listened while a voiceannounced every 30 seconds for three minutes the time remaining beforethey would be given a topic and allowed to plan for their speech Pre-sumably this period is the most stressful and the most likely to result inanxious arousal Indeed the correlation between heart rate acceleration tothe Anticipation condition and the increase in state anxiety was signireg cant
More recently Nitschke Heller Imig and Miller (submitted) haveattempted to identify anxious arousal and anxious apprehension usingquestionnaires Data from 783 students for a battery of anxiety and depres-sion questionnaires provided strong support for the anxiety distinctionproposed here at least on a psychological level of analysis The Moodand Anxiety Symptom Questionnaire-Anxious Arousal scale (MASQWatson et al 1995ab) was chosen as a measure of anxious arousalwhereas the Penn State Worry Questionnaire (PSWQ Meyer MillerMetzger amp Borkovec 1990 Molina amp Borkovec 1994) was included asa measure of anxious apprehension The correlation between the two scales(r = 20) was much lower than those involving the other more generalscales of anxiety included in the study (MASQ-General Anxiety scaleBeck Anxiety Inventory and the State-Trait Anxiety Inventory-Traitscale ranging from r = 47 to r = 65) These results further support the
436 HELLER AND NITSCHKE
argument that anxious arousal and anxious apprehension are distinctanxiety states
However new questions are highlighted by this distinction betweenanxious arousal and anxious apprehension particularly in the light ofHellerrsquo s model If anxious apprehension is associated with left anterioractivity the overall pattern of regional brain activity appears to be identicalto the brain activity associated with pleasant emotional states (ie more leftthan right anterior activity) However anxious apprehension might beexpected to fall in the portion of the circumplex model of emotionassociated with activated unpleasant affect (as suggested by adjectivessuch as fearful nervous jittery anxious) Below we suggest a numberof theoretical possibilities for resolving this apparent discrepancy all ofwhich remain to be examined empirically
In general it will be important to clarify the degree to which anxiousapprehension is in fact characterised by unpleasant versus pleasant valenceas well as high versus low arousal In addition it may be helpful toconsider these anterior asymmetries from the approach-withdrawal frame-work employed by Davidson (1992ab Davidson amp Tomarken 1989)Under conditions where anxious apprehension leads to approach tenden-cies one might expect an asymmetry in favour of the left anterior region Incontrast when anxious apprehension is associated with withdrawal tenden-cies (perhaps as anxious arousal increases) an asymmetry may emerge infavour of the right anterior region Conceivably the degree to whichanxious apprehension is characterised by approach versus withdrawaltendencies would depend on the degree of stress and discomfortinvolved the personrsquo s coping style and other factors With regard tovalence we would predict that anxious apprehension characterised byapproach tendencies would be accompanied by relatively pleasant affectAnxious apprehension accompanied by withdrawal tendencies in contrastwould be accompanied by relatively unpleasant affect
It is also possible that instead of a direct link between anxious appre-hension and approach behaviour they share more fundamental character-istics associated with left anterior functions For example one might arguethat they both involve sequential thinking and the development of hier-archically structured plans focusing on future concerns (for a review ofrelevant left anterior functions see Tomarken amp Keener this Issue)
Another important issue has to do with the nature of the asymmetricpattern of activity As reviewed earlier we found that the asymmetry infavour of the left hemisphere for participants classireg ed on the basis of traitanxiety was due to a relative decrease in right hemisphere activity Thusthe left frontal region of the anxious participants was not signireg cantly moreactive than that of controls It will therefore be important to examine thepossibility that absolute activity of the left anterior region plays a more
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 437
important role in the modulation of valence (eg Davidson et al 1995)than of anxious apprehension
Another possibility is that independent left anterior systems are asso-ciated with anxious apprehension versus pleasant affect Liotti and Tucker(1995) review theories that there may be multiple functionally segregatedparallel frontal cortico-striato-thalamic circuits some associated withmotor functions some with set shifting and behavioural inhibition andsome with emotional information processing and hedonic responsivity tostimuli Possibly one of these circuits remacr ects activation associated withpleasant affect and another remacr ects activation associated with anxiousapprehension andor approach Different neurotransmitter systems couldbe involved More precise methods of localisation would therefore benecessary to distinguish the unique brain activity associated with each ofthese psychologicalbehavioural phenomena
Comorbidity of Depression and Anxiety
In considering the issue of decomposing emotional states with regard to thesymptomatology in affective and anxiety disorders we have previouslyargued that issues of comorbidity of depression and anxiety are crucial(Heller et al 1995) A review of the literature reveals that rates of comor-bidity are extremely high (Alloy Kelly Mineka amp Clements 1990 Helleret al 1995 Hiller et al 1989 Katon amp Roy-Byrne 1991) Even in our veryrecent work (Nitschke et al submitted) using the MASQ a measuredesigned by Watson and colleagues to avoid symptom overlap in themeasurement of anxiety and depression (Watson et al 1995ab) comor-bidity was fairly pronounced in a large college student sample The MASQAnxious Arousal (AA) and Anhedonic Depression (AD) scales were used asmeasures of those constructs and the PSWQ provided the measure ofanxious apprehension Of the 125 (out of 783) participants scoring abovethe 80th percentile on at least one of the three scales of interest and belowthe 50th percentile on any where the 80th percentile cut-off was not metfully 50 demonstrated comorbidity by scoring above the 80th percentileon at least two of the scales This rate of comorbidity emerged despite thelow correlations found between the three scales (MASQ AA and MASQAD r = 20 MASQ AA and PSWQ r = 20 MASQ AD and PSWQ r =36) relative to the correlations found for the other more general measuresof depression including the MASQ General Mixed General Depression andGeneral Anxiety scales Beck Depression Inventory Beck Anxiety Inven-tory General Behaviour Inventory-Dysthymia scale and the State-TraitAnxiety Inventory-Trait scale (ranging from r = 47 to r = 76)
Because at least some studies have found opposite patterns of activityfor the right posterior region during depression versus during anxiety (ie
438 HELLER AND NITSCHKE
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
during a task in which participants listened to emotional narrativesdesigned to elicit anxious arousal (ie fear and sad narratives that hadpreviously received unpleasant valence and high arousal ratings)
For frontal activity group differences were unaffected by the narrativesAnxious participants showed a larger asymmetry in favour of the lefthemisphere than controls during the rest and listen periods of all narrativetypes as well as for eight minutes of resting baseline preceding thenarrative task Examination of the data indicated that the increased asym-metry was primarily due to a decrease in right hemisphere activity for theanxious participants Nonetheless the increase in magnitude of the asym-metry favouring the left hemisphere found for anxious participants isconsistent with the studies reviewed earlier that reported or inferredrelatively greater left than right anterior activity in anxious subjectsclassireg ed on the basis of worry or trait anxiety (eg Carter et al 1986Tucker et al 1978 Tyler amp Tucker 1982) Our frontal data suggest that itwill be important to investigate whether anxious apprehension is bestcharacterised by the magnitude of the asymmetry favouring the left hemi-sphere or by the level of left hemisphere activity in isolation
In contrast the parietal data supported our prediction regarding theassociation of anxious arousal with greater right hemisphere activityWhen listening to the fear and sad narratives anxious participants showeda selective increase in right parietal activity that was not demonstrated bycontrols The right parietal effect however was not obtained for neutraland happy narratives that had previously received pleasant valence and lowarousal ratings These reg ndings are consistent with those studies that havereported or inferred greater right hemisphere activity during anxious statescharacterised by panic and stress-inducing conditions (eg Reiman et al1984 Stewart et al 1988 Tucker et al 1977)
An important question here has to do with the discrepancies in frontalasymmetries that have been reported Like others (eg Baxter et al 1987Gur et al 1988 Wu et al 1991) we did not reg nd heightened right anterioractivity in anxiety However a number of other studies have (eg Davidsonet al 1996 Reivich et al 1983 Tomarken amp Davidson 1994) Asdiscussed earlier we have previously argued that anterior regions areuniquely involved in the valence dimension of emotion with pleasantaffect associated with greater left activity and unpleasant affect associatedwith greater right activity (eg Heller 1990) Therefore we would expectheightened right anterior activity in an anxious group only if they differedfrom controls on measures indexing affective valence In our recent studythe nondepressed anxious group did not differ from controls on either theGBI Dysthymia scale or the PANAS PA and NA scales
It may also be important to measure the level of anxious apprehension inthe sample studied For example Tomarken and Davidson (1994) found
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 435
that with Marloweplusmn Crowne scores controlled trait anxiety failed to predictfrontal asymmetry However their groups were divided according to amedian split
In summary our results directly support the proposed distinctionbetween anxious apprehension and anxious arousal as types of anxietythat may account for signireg cant variance in patterns of brain activityThey suggest furthermore that Hellerrsquo s model should be rereg ned to reflectthe likelihood that parietal activation will covary with anxious arousal butnot with anxious apprehension If these distinctions are accurate we wouldexpect to observe increases in right parietotemporal activity only underconditions when anxiety is accompanied by increased arousal
Thus in homodynamic and EEG studies a failure to reg nd increased rightposterior activity in anxiety is not surprising as variations in anxiousarousal are likely to be large Some individuals may experience anxiousarousal frequently especially those with disorders more characteristic ofanxious arousal such as panic disorder phobias or post-traumatic stressdisorder Others although apprehensive may rarely experience anxiousarousal In addition the situation may vary in the degree to which it elicitsanxious arousal Being forced to give a speech is likely to induce a greatdeal of anxious arousal in social phobics thus we would have predicted asignireg cant increase in right parietal activity precisely as found by David-son et al (submitted) Note that this effect was only found during theAnticipation condition not during baseline Planning or after thespeech During the Anticipation condition subjects listened while a voiceannounced every 30 seconds for three minutes the time remaining beforethey would be given a topic and allowed to plan for their speech Pre-sumably this period is the most stressful and the most likely to result inanxious arousal Indeed the correlation between heart rate acceleration tothe Anticipation condition and the increase in state anxiety was signireg cant
More recently Nitschke Heller Imig and Miller (submitted) haveattempted to identify anxious arousal and anxious apprehension usingquestionnaires Data from 783 students for a battery of anxiety and depres-sion questionnaires provided strong support for the anxiety distinctionproposed here at least on a psychological level of analysis The Moodand Anxiety Symptom Questionnaire-Anxious Arousal scale (MASQWatson et al 1995ab) was chosen as a measure of anxious arousalwhereas the Penn State Worry Questionnaire (PSWQ Meyer MillerMetzger amp Borkovec 1990 Molina amp Borkovec 1994) was included asa measure of anxious apprehension The correlation between the two scales(r = 20) was much lower than those involving the other more generalscales of anxiety included in the study (MASQ-General Anxiety scaleBeck Anxiety Inventory and the State-Trait Anxiety Inventory-Traitscale ranging from r = 47 to r = 65) These results further support the
436 HELLER AND NITSCHKE
argument that anxious arousal and anxious apprehension are distinctanxiety states
However new questions are highlighted by this distinction betweenanxious arousal and anxious apprehension particularly in the light ofHellerrsquo s model If anxious apprehension is associated with left anterioractivity the overall pattern of regional brain activity appears to be identicalto the brain activity associated with pleasant emotional states (ie more leftthan right anterior activity) However anxious apprehension might beexpected to fall in the portion of the circumplex model of emotionassociated with activated unpleasant affect (as suggested by adjectivessuch as fearful nervous jittery anxious) Below we suggest a numberof theoretical possibilities for resolving this apparent discrepancy all ofwhich remain to be examined empirically
In general it will be important to clarify the degree to which anxiousapprehension is in fact characterised by unpleasant versus pleasant valenceas well as high versus low arousal In addition it may be helpful toconsider these anterior asymmetries from the approach-withdrawal frame-work employed by Davidson (1992ab Davidson amp Tomarken 1989)Under conditions where anxious apprehension leads to approach tenden-cies one might expect an asymmetry in favour of the left anterior region Incontrast when anxious apprehension is associated with withdrawal tenden-cies (perhaps as anxious arousal increases) an asymmetry may emerge infavour of the right anterior region Conceivably the degree to whichanxious apprehension is characterised by approach versus withdrawaltendencies would depend on the degree of stress and discomfortinvolved the personrsquo s coping style and other factors With regard tovalence we would predict that anxious apprehension characterised byapproach tendencies would be accompanied by relatively pleasant affectAnxious apprehension accompanied by withdrawal tendencies in contrastwould be accompanied by relatively unpleasant affect
It is also possible that instead of a direct link between anxious appre-hension and approach behaviour they share more fundamental character-istics associated with left anterior functions For example one might arguethat they both involve sequential thinking and the development of hier-archically structured plans focusing on future concerns (for a review ofrelevant left anterior functions see Tomarken amp Keener this Issue)
Another important issue has to do with the nature of the asymmetricpattern of activity As reviewed earlier we found that the asymmetry infavour of the left hemisphere for participants classireg ed on the basis of traitanxiety was due to a relative decrease in right hemisphere activity Thusthe left frontal region of the anxious participants was not signireg cantly moreactive than that of controls It will therefore be important to examine thepossibility that absolute activity of the left anterior region plays a more
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 437
important role in the modulation of valence (eg Davidson et al 1995)than of anxious apprehension
Another possibility is that independent left anterior systems are asso-ciated with anxious apprehension versus pleasant affect Liotti and Tucker(1995) review theories that there may be multiple functionally segregatedparallel frontal cortico-striato-thalamic circuits some associated withmotor functions some with set shifting and behavioural inhibition andsome with emotional information processing and hedonic responsivity tostimuli Possibly one of these circuits remacr ects activation associated withpleasant affect and another remacr ects activation associated with anxiousapprehension andor approach Different neurotransmitter systems couldbe involved More precise methods of localisation would therefore benecessary to distinguish the unique brain activity associated with each ofthese psychologicalbehavioural phenomena
Comorbidity of Depression and Anxiety
In considering the issue of decomposing emotional states with regard to thesymptomatology in affective and anxiety disorders we have previouslyargued that issues of comorbidity of depression and anxiety are crucial(Heller et al 1995) A review of the literature reveals that rates of comor-bidity are extremely high (Alloy Kelly Mineka amp Clements 1990 Helleret al 1995 Hiller et al 1989 Katon amp Roy-Byrne 1991) Even in our veryrecent work (Nitschke et al submitted) using the MASQ a measuredesigned by Watson and colleagues to avoid symptom overlap in themeasurement of anxiety and depression (Watson et al 1995ab) comor-bidity was fairly pronounced in a large college student sample The MASQAnxious Arousal (AA) and Anhedonic Depression (AD) scales were used asmeasures of those constructs and the PSWQ provided the measure ofanxious apprehension Of the 125 (out of 783) participants scoring abovethe 80th percentile on at least one of the three scales of interest and belowthe 50th percentile on any where the 80th percentile cut-off was not metfully 50 demonstrated comorbidity by scoring above the 80th percentileon at least two of the scales This rate of comorbidity emerged despite thelow correlations found between the three scales (MASQ AA and MASQAD r = 20 MASQ AA and PSWQ r = 20 MASQ AD and PSWQ r =36) relative to the correlations found for the other more general measuresof depression including the MASQ General Mixed General Depression andGeneral Anxiety scales Beck Depression Inventory Beck Anxiety Inven-tory General Behaviour Inventory-Dysthymia scale and the State-TraitAnxiety Inventory-Trait scale (ranging from r = 47 to r = 76)
Because at least some studies have found opposite patterns of activityfor the right posterior region during depression versus during anxiety (ie
438 HELLER AND NITSCHKE
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
that with Marloweplusmn Crowne scores controlled trait anxiety failed to predictfrontal asymmetry However their groups were divided according to amedian split
In summary our results directly support the proposed distinctionbetween anxious apprehension and anxious arousal as types of anxietythat may account for signireg cant variance in patterns of brain activityThey suggest furthermore that Hellerrsquo s model should be rereg ned to reflectthe likelihood that parietal activation will covary with anxious arousal butnot with anxious apprehension If these distinctions are accurate we wouldexpect to observe increases in right parietotemporal activity only underconditions when anxiety is accompanied by increased arousal
Thus in homodynamic and EEG studies a failure to reg nd increased rightposterior activity in anxiety is not surprising as variations in anxiousarousal are likely to be large Some individuals may experience anxiousarousal frequently especially those with disorders more characteristic ofanxious arousal such as panic disorder phobias or post-traumatic stressdisorder Others although apprehensive may rarely experience anxiousarousal In addition the situation may vary in the degree to which it elicitsanxious arousal Being forced to give a speech is likely to induce a greatdeal of anxious arousal in social phobics thus we would have predicted asignireg cant increase in right parietal activity precisely as found by David-son et al (submitted) Note that this effect was only found during theAnticipation condition not during baseline Planning or after thespeech During the Anticipation condition subjects listened while a voiceannounced every 30 seconds for three minutes the time remaining beforethey would be given a topic and allowed to plan for their speech Pre-sumably this period is the most stressful and the most likely to result inanxious arousal Indeed the correlation between heart rate acceleration tothe Anticipation condition and the increase in state anxiety was signireg cant
More recently Nitschke Heller Imig and Miller (submitted) haveattempted to identify anxious arousal and anxious apprehension usingquestionnaires Data from 783 students for a battery of anxiety and depres-sion questionnaires provided strong support for the anxiety distinctionproposed here at least on a psychological level of analysis The Moodand Anxiety Symptom Questionnaire-Anxious Arousal scale (MASQWatson et al 1995ab) was chosen as a measure of anxious arousalwhereas the Penn State Worry Questionnaire (PSWQ Meyer MillerMetzger amp Borkovec 1990 Molina amp Borkovec 1994) was included asa measure of anxious apprehension The correlation between the two scales(r = 20) was much lower than those involving the other more generalscales of anxiety included in the study (MASQ-General Anxiety scaleBeck Anxiety Inventory and the State-Trait Anxiety Inventory-Traitscale ranging from r = 47 to r = 65) These results further support the
436 HELLER AND NITSCHKE
argument that anxious arousal and anxious apprehension are distinctanxiety states
However new questions are highlighted by this distinction betweenanxious arousal and anxious apprehension particularly in the light ofHellerrsquo s model If anxious apprehension is associated with left anterioractivity the overall pattern of regional brain activity appears to be identicalto the brain activity associated with pleasant emotional states (ie more leftthan right anterior activity) However anxious apprehension might beexpected to fall in the portion of the circumplex model of emotionassociated with activated unpleasant affect (as suggested by adjectivessuch as fearful nervous jittery anxious) Below we suggest a numberof theoretical possibilities for resolving this apparent discrepancy all ofwhich remain to be examined empirically
In general it will be important to clarify the degree to which anxiousapprehension is in fact characterised by unpleasant versus pleasant valenceas well as high versus low arousal In addition it may be helpful toconsider these anterior asymmetries from the approach-withdrawal frame-work employed by Davidson (1992ab Davidson amp Tomarken 1989)Under conditions where anxious apprehension leads to approach tenden-cies one might expect an asymmetry in favour of the left anterior region Incontrast when anxious apprehension is associated with withdrawal tenden-cies (perhaps as anxious arousal increases) an asymmetry may emerge infavour of the right anterior region Conceivably the degree to whichanxious apprehension is characterised by approach versus withdrawaltendencies would depend on the degree of stress and discomfortinvolved the personrsquo s coping style and other factors With regard tovalence we would predict that anxious apprehension characterised byapproach tendencies would be accompanied by relatively pleasant affectAnxious apprehension accompanied by withdrawal tendencies in contrastwould be accompanied by relatively unpleasant affect
It is also possible that instead of a direct link between anxious appre-hension and approach behaviour they share more fundamental character-istics associated with left anterior functions For example one might arguethat they both involve sequential thinking and the development of hier-archically structured plans focusing on future concerns (for a review ofrelevant left anterior functions see Tomarken amp Keener this Issue)
Another important issue has to do with the nature of the asymmetricpattern of activity As reviewed earlier we found that the asymmetry infavour of the left hemisphere for participants classireg ed on the basis of traitanxiety was due to a relative decrease in right hemisphere activity Thusthe left frontal region of the anxious participants was not signireg cantly moreactive than that of controls It will therefore be important to examine thepossibility that absolute activity of the left anterior region plays a more
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 437
important role in the modulation of valence (eg Davidson et al 1995)than of anxious apprehension
Another possibility is that independent left anterior systems are asso-ciated with anxious apprehension versus pleasant affect Liotti and Tucker(1995) review theories that there may be multiple functionally segregatedparallel frontal cortico-striato-thalamic circuits some associated withmotor functions some with set shifting and behavioural inhibition andsome with emotional information processing and hedonic responsivity tostimuli Possibly one of these circuits remacr ects activation associated withpleasant affect and another remacr ects activation associated with anxiousapprehension andor approach Different neurotransmitter systems couldbe involved More precise methods of localisation would therefore benecessary to distinguish the unique brain activity associated with each ofthese psychologicalbehavioural phenomena
Comorbidity of Depression and Anxiety
In considering the issue of decomposing emotional states with regard to thesymptomatology in affective and anxiety disorders we have previouslyargued that issues of comorbidity of depression and anxiety are crucial(Heller et al 1995) A review of the literature reveals that rates of comor-bidity are extremely high (Alloy Kelly Mineka amp Clements 1990 Helleret al 1995 Hiller et al 1989 Katon amp Roy-Byrne 1991) Even in our veryrecent work (Nitschke et al submitted) using the MASQ a measuredesigned by Watson and colleagues to avoid symptom overlap in themeasurement of anxiety and depression (Watson et al 1995ab) comor-bidity was fairly pronounced in a large college student sample The MASQAnxious Arousal (AA) and Anhedonic Depression (AD) scales were used asmeasures of those constructs and the PSWQ provided the measure ofanxious apprehension Of the 125 (out of 783) participants scoring abovethe 80th percentile on at least one of the three scales of interest and belowthe 50th percentile on any where the 80th percentile cut-off was not metfully 50 demonstrated comorbidity by scoring above the 80th percentileon at least two of the scales This rate of comorbidity emerged despite thelow correlations found between the three scales (MASQ AA and MASQAD r = 20 MASQ AA and PSWQ r = 20 MASQ AD and PSWQ r =36) relative to the correlations found for the other more general measuresof depression including the MASQ General Mixed General Depression andGeneral Anxiety scales Beck Depression Inventory Beck Anxiety Inven-tory General Behaviour Inventory-Dysthymia scale and the State-TraitAnxiety Inventory-Trait scale (ranging from r = 47 to r = 76)
Because at least some studies have found opposite patterns of activityfor the right posterior region during depression versus during anxiety (ie
438 HELLER AND NITSCHKE
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
argument that anxious arousal and anxious apprehension are distinctanxiety states
However new questions are highlighted by this distinction betweenanxious arousal and anxious apprehension particularly in the light ofHellerrsquo s model If anxious apprehension is associated with left anterioractivity the overall pattern of regional brain activity appears to be identicalto the brain activity associated with pleasant emotional states (ie more leftthan right anterior activity) However anxious apprehension might beexpected to fall in the portion of the circumplex model of emotionassociated with activated unpleasant affect (as suggested by adjectivessuch as fearful nervous jittery anxious) Below we suggest a numberof theoretical possibilities for resolving this apparent discrepancy all ofwhich remain to be examined empirically
In general it will be important to clarify the degree to which anxiousapprehension is in fact characterised by unpleasant versus pleasant valenceas well as high versus low arousal In addition it may be helpful toconsider these anterior asymmetries from the approach-withdrawal frame-work employed by Davidson (1992ab Davidson amp Tomarken 1989)Under conditions where anxious apprehension leads to approach tenden-cies one might expect an asymmetry in favour of the left anterior region Incontrast when anxious apprehension is associated with withdrawal tenden-cies (perhaps as anxious arousal increases) an asymmetry may emerge infavour of the right anterior region Conceivably the degree to whichanxious apprehension is characterised by approach versus withdrawaltendencies would depend on the degree of stress and discomfortinvolved the personrsquo s coping style and other factors With regard tovalence we would predict that anxious apprehension characterised byapproach tendencies would be accompanied by relatively pleasant affectAnxious apprehension accompanied by withdrawal tendencies in contrastwould be accompanied by relatively unpleasant affect
It is also possible that instead of a direct link between anxious appre-hension and approach behaviour they share more fundamental character-istics associated with left anterior functions For example one might arguethat they both involve sequential thinking and the development of hier-archically structured plans focusing on future concerns (for a review ofrelevant left anterior functions see Tomarken amp Keener this Issue)
Another important issue has to do with the nature of the asymmetricpattern of activity As reviewed earlier we found that the asymmetry infavour of the left hemisphere for participants classireg ed on the basis of traitanxiety was due to a relative decrease in right hemisphere activity Thusthe left frontal region of the anxious participants was not signireg cantly moreactive than that of controls It will therefore be important to examine thepossibility that absolute activity of the left anterior region plays a more
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 437
important role in the modulation of valence (eg Davidson et al 1995)than of anxious apprehension
Another possibility is that independent left anterior systems are asso-ciated with anxious apprehension versus pleasant affect Liotti and Tucker(1995) review theories that there may be multiple functionally segregatedparallel frontal cortico-striato-thalamic circuits some associated withmotor functions some with set shifting and behavioural inhibition andsome with emotional information processing and hedonic responsivity tostimuli Possibly one of these circuits remacr ects activation associated withpleasant affect and another remacr ects activation associated with anxiousapprehension andor approach Different neurotransmitter systems couldbe involved More precise methods of localisation would therefore benecessary to distinguish the unique brain activity associated with each ofthese psychologicalbehavioural phenomena
Comorbidity of Depression and Anxiety
In considering the issue of decomposing emotional states with regard to thesymptomatology in affective and anxiety disorders we have previouslyargued that issues of comorbidity of depression and anxiety are crucial(Heller et al 1995) A review of the literature reveals that rates of comor-bidity are extremely high (Alloy Kelly Mineka amp Clements 1990 Helleret al 1995 Hiller et al 1989 Katon amp Roy-Byrne 1991) Even in our veryrecent work (Nitschke et al submitted) using the MASQ a measuredesigned by Watson and colleagues to avoid symptom overlap in themeasurement of anxiety and depression (Watson et al 1995ab) comor-bidity was fairly pronounced in a large college student sample The MASQAnxious Arousal (AA) and Anhedonic Depression (AD) scales were used asmeasures of those constructs and the PSWQ provided the measure ofanxious apprehension Of the 125 (out of 783) participants scoring abovethe 80th percentile on at least one of the three scales of interest and belowthe 50th percentile on any where the 80th percentile cut-off was not metfully 50 demonstrated comorbidity by scoring above the 80th percentileon at least two of the scales This rate of comorbidity emerged despite thelow correlations found between the three scales (MASQ AA and MASQAD r = 20 MASQ AA and PSWQ r = 20 MASQ AD and PSWQ r =36) relative to the correlations found for the other more general measuresof depression including the MASQ General Mixed General Depression andGeneral Anxiety scales Beck Depression Inventory Beck Anxiety Inven-tory General Behaviour Inventory-Dysthymia scale and the State-TraitAnxiety Inventory-Trait scale (ranging from r = 47 to r = 76)
Because at least some studies have found opposite patterns of activityfor the right posterior region during depression versus during anxiety (ie
438 HELLER AND NITSCHKE
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
important role in the modulation of valence (eg Davidson et al 1995)than of anxious apprehension
Another possibility is that independent left anterior systems are asso-ciated with anxious apprehension versus pleasant affect Liotti and Tucker(1995) review theories that there may be multiple functionally segregatedparallel frontal cortico-striato-thalamic circuits some associated withmotor functions some with set shifting and behavioural inhibition andsome with emotional information processing and hedonic responsivity tostimuli Possibly one of these circuits remacr ects activation associated withpleasant affect and another remacr ects activation associated with anxiousapprehension andor approach Different neurotransmitter systems couldbe involved More precise methods of localisation would therefore benecessary to distinguish the unique brain activity associated with each ofthese psychologicalbehavioural phenomena
Comorbidity of Depression and Anxiety
In considering the issue of decomposing emotional states with regard to thesymptomatology in affective and anxiety disorders we have previouslyargued that issues of comorbidity of depression and anxiety are crucial(Heller et al 1995) A review of the literature reveals that rates of comor-bidity are extremely high (Alloy Kelly Mineka amp Clements 1990 Helleret al 1995 Hiller et al 1989 Katon amp Roy-Byrne 1991) Even in our veryrecent work (Nitschke et al submitted) using the MASQ a measuredesigned by Watson and colleagues to avoid symptom overlap in themeasurement of anxiety and depression (Watson et al 1995ab) comor-bidity was fairly pronounced in a large college student sample The MASQAnxious Arousal (AA) and Anhedonic Depression (AD) scales were used asmeasures of those constructs and the PSWQ provided the measure ofanxious apprehension Of the 125 (out of 783) participants scoring abovethe 80th percentile on at least one of the three scales of interest and belowthe 50th percentile on any where the 80th percentile cut-off was not metfully 50 demonstrated comorbidity by scoring above the 80th percentileon at least two of the scales This rate of comorbidity emerged despite thelow correlations found between the three scales (MASQ AA and MASQAD r = 20 MASQ AA and PSWQ r = 20 MASQ AD and PSWQ r =36) relative to the correlations found for the other more general measuresof depression including the MASQ General Mixed General Depression andGeneral Anxiety scales Beck Depression Inventory Beck Anxiety Inven-tory General Behaviour Inventory-Dysthymia scale and the State-TraitAnxiety Inventory-Trait scale (ranging from r = 47 to r = 76)
Because at least some studies have found opposite patterns of activityfor the right posterior region during depression versus during anxiety (ie
438 HELLER AND NITSCHKE
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
anxious arousal) we reasoned that the failure to distinguish carefullybetween the two could lead in part to the inconsistent reg ndings in theliterature for that region of the brain Furthermore the overlap betweendepression and anxiety decreases as the severity of the disorder (majordepression or panic disorder) increases (Hiller et al 1989) It would thusbe less likely that anxiety and depression would be confounded from aneuropsychological perspective in more severe cases which is whenparieto-temporal differences between depressed and nondepressed peopleare most likely to emerge (Davidson amp Tomarken 1989)
As mentioned earlier Heller et al (1995) examined college studentsclassireg ed as either depressed anxious or both It emerged that depressionas expected was associated with reduced right hemisphere activity asmeasured by the CFT In direct contrast anxiety was associated withincreased right hemisphere activity These results are consistent with thepredictions made by Hellerrsquo s model for depression and anxiety Further-more they highlight the probability that studies failing to distinguishdepression and anxiety particularly in less-distressed populations arelikely to reg nd the divergent patterns of brain activity obscured
The inconsistencies for anterior regions in studies of depression andanxiety might also be explained by an understanding of comorbidity Forexample if a nonmelancholic depressed sample also has high levels ofanxious apprehension the increase in right anterior activity associated withsuch depression might be matched by an increase in left anterior activityassociated with anxious apprehension resulting in a pattern of asymmetrylooking identical to a nondepressed nonanxious control sample Melanch-olia and anxious apprehension appear to be associated with oppositepatterns of activity in the left anterior region which might serve to canceleach other out conceivably resulting in a ` normalrsquo rsquo level of left anterioractivity
CONCLUSION AND QUESTIONS FORFUTURE RESEARCH
We have reviewed the literature on depression and anxiety with an eye tohighlighting the inconsistencies that have been reported in activity andinvolvement of different brain regions We argued that many of theseinconsistencies may be resolved by turning to the psychological literatureon emotion depression and anxiety as a guide to decomposing complexemotional states into subcomponents In the process we considered Hel-lerrsquo s model in the light of recent research and rereg nements in our theoreticalposition We suggested that it is important to be clear that our predictionsfor the right parieto-temporal region are specireg c to anxious arousal Wealso suggested that it may be necessary to modify the model to remacr ect the
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 439
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
potentially dual nature of the anterior systems associated with valence Toaddress this possibility we have recommended that future studies samplethe full range of affective descriptors on the circumplex model of emotionwhen examining the relations between affective states and brain activityand function Despite these rereg nements the fundamental tenets of themodel continue to be consistent with emerging evidence in depressionand anxiety Anterior regions of the brain consistently emerge as impor-tant in affective valence whereas the right parieto-temporal region isimportant in arousal The model therefore remains a useful guide to theoryand research
Two related questions can be raised with regard to the brain mechanismsin emotion A fundamental issue has to do with the question as to why theobserved asymmetries are associated with affective valence In particularwe do not know precisely why higher left than right anterior activation isassociated with pleasant affect or why higher right than left activation isassociated with unpleasant affect As relative activation of various brainregions has emerged as a critical component of different affective statesearlier perspectives have shifted from right versus left hemisphere specia-lisation for affect to differential involvement of the anterior and posteriorhemispheres and corresponding asymmetries in activity (see Kinsbourne1988 and Levy et al 1983 for further discussion of the crucial distinctionbetween specialisation and activation with regard to psychopathology andemotion) Such asymmetries in activity are clearly expected to affect thecognitive and information processing functions of specialised regions of thebrain (a topic of considerable interest at present Heller amp Nitschke 1997Liotti amp Tucker 1995 Tomarken amp Keener this Issue also Tucker 1981)However the reason why anterior asymmetries in brain activity are asso-ciated with affective valence in the ways they are is an unanswered questionof continuing importance Some authors (eg Liotti amp Tucker 1995Robinson Boston Starkstein amp Price 1988) have discussed the possibilitythat different neurotransmitter systems are associated with the behaviouralcharacteristics of different emotional states Nonetheless it appears that ourunderstanding of this aspect of the regulation of affect is rudimentary
Furthermore the emphasis on regional differences in activity leads to aresearch agenda in which we would hope to elucidate the mechanismsunderlying such asymmetries in activity Examples of these could includecortically driven phenomena such as cognitions (eg Clore 1994) sub-cortically driven phenomena such as interactions with the amygdala (egLeDoux 1993) biochemical phenomena such as dopamine and norepi-nephrine pathways (eg Liotti amp Tucker 1995 Tucker amp Williamson1984) externally applied phenomena such as electrical impulses lesionsor even life stress and stable personality characteristics such as extraver-sion (Heller 1993b Nitschke et al in prep)
440 HELLER AND NITSCHKE
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
In summary our approach to understanding brain function in affectiveand anxiety disorders has focused on identifying the components of theemotional phenomena accompanying these disorders In this regard wehave been guided by the circumplex model of emotion as well as bypsychological theories that have attempted to describe the unique charac-teristics of various emotional states We have emphasised the importanceof considering subtypes of depression and anxiety and their comorbiditywhen confronted with conmacr icting results Using this strategy we believethat we can piece together some parts of the puzzle as to which regions ofthe brain are involved in depression and anxiety
Manuscript received 17 June 1997
REFERENCES
Allen JJ Iacono WG Depue RA amp Arbisi X (1993) Regional EEG asymmetries inbipolar seasonal affective disorder before and after phototherapy Biological Psychiatry33 642plusmn 646
Alloy LB Kelly KA Mineka S amp Clements CM (1990) Comorbidity of anxiety anddepressive disorders A helplessness-hopelessness perspective In JD Maser amp CRCloninger (Eds) Comorbidity of mood and anxiety disorders (pp 499plusmn 543) Washing-ton DC American Psychiatric Press
APA (American Psychiatric Association) (1994) Diagnostic and statistical manual ofmental disorders (4th ed) Washington DC Author
Banich MT Stolar N Heller W amp Goldman R (1992) A dereg cit in right-hemisphereperformance after induction of a depressed mood Neuropsychiatry Neuropsychologyand Behavioral Neurology 5 20plusmn 27
Barlow DH (1988) Anxiety and its disorders The nature and treatment of anxiety andpanic New York Guilford Press
Barlow DH (1991) Disorders of emotion Psychological Inquiry 2 58plusmn 71Baxter LR Phelps ME Mazziotta JC Guze BH Schwartz JM amp Selin
CE (1987) Local cerebral glucose metabolic rates in obsessive-compulsive disorderArchives of General Psychiatry 44 211plusmn 218
Baxter LR Phelps ME Mazziotta JC Schwartz JM Berner RH Selin CE ampSumida RM (1985) Cerebral metabolic rates for glucose in mood disorders Studieswith positron emission tomography and macr uorodeoxyglucose F 18 Archives of GeneralPsychiatry 42 441plusmn 447
Baxter LR Schwartz JM Phelps ME et al (1989) Reduction of prefrontal cortexglucose metabolism common to three types of depression Archives of General Psychia-try 46 243plusmn 250
Bench CJ Friston KJ Brown RG Frackowiak RSJ amp Dolan RJ (1993) Regionalcerebral blood macr ow in depression measured by positron emission tomography Therelationship with clinical dimensions Psychological Medicine 23 579plusmn 590
Bench CJ Friston KJ Brown RG Scott LC Frackowiak RSJ amp DolanRJ (1992) The anatomy of melancholia Focal abnormalities of cerebral blood macr owin major depression Psychological Medicine 22 607plusmn 615
Berenbaum H amp Williams M (1994) Extraversion hemispatial bias and eyeblink ratesPersonality and Individual Differences 17 849plusmn 852
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 441
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
Berndt DJ amp Berndt SM (1980) Relationship of mild depression to psychologicaldereg cit in college students Journal of Clinical Psychology 36 868plusmn 874
Borod JC (1993) Cerebral mechanisms underlying facial prosodic and lexical emotionalexpression A review of neuropsychological studies and methodological issues Neuro-psychology 7 445plusmn 463
Bruder GE (1995) Cerebral laterality and psychopathology Perceptual and event-relatedpotential asymmetries in affective and schizophrenic disorders In RJ Davidson amp KHugdahl (Eds) Brain asymmetry (pp 661plusmn 691) Cambridge MA The MIT Press
Bruder GE Quitkin FM Stewart JW Martin C Voglmaier MM amp HarrisonWM (1989) Cerebral laterality and depression Differences in perceptual asymmetryamong diagnostic subtypes Journal of Abnormal Psychology 98 177plusmn 186
Buchsbaum MS Hazlett E Sicotte N Stein M Wu J amp Zetin M (1985)Topographic EEG changes with benzodiazepine administration in generalized anxietydisorder Biological Psychiatry 20 832plusmn 842
Buchsbaum MS Wu J Haier R Harlett E Ball R Katz M Sokolski K Lagunas-Solar M amp Langer D (1987) Positron emission tomography assessment of effects ofbenzodiazepines on regional glucose metabolic rate in patients with anxiety disordersLife Science 40 2393plusmn 2400
Caltagirone C Zoccolotti P Originale G Daniele A amp Mammucari A (1989)Autonomic reactivity and facial expression of emotion in brain-damaged patients InG Gainotti amp C Caltagirone (Eds) Emotions and the dual brain (pp 204plusmn 221) BerlinSpringer-Verlag
Carter WR Johnson MC amp Borkovec TD (1986) Worry An electrocortical analysisAdvances in Behavioral Research and Therapy 8 193plusmn 204
Clark LA amp Watson D (1991) Tripartite model of anxiety and depression Psycho-metric evidence and taxonomic implications Journal of Abnormal Psychology 100316plusmn 336
Clore GL (1994) Why emotions require cognition In P Ekman amp RJ Davidson (Eds)The nature of emotion Fundamental questions (pp 181plusmn 191) New York OxfordUniversity Press
Davidson RJ (1984) Affect cognition and hemispheric specialization In CE Izard JKagan amp R Zajonc (Eds) Emotion cognition and behavior (pp 320plusmn 365) New YorkCambridge University Press
Davidson RJ (1992a) Anterior cerebral asymmetry and the nature of emotion Brain andCognition 20 125plusmn 151
Davidson RJ (1992b) Emotion and affective style Hemispheric substrates PsychologicalScience 3 39plusmn 43
Davidson RJ (1993) Parsing affective space Perspectives from neuropsychology andpsychophysiology Neuropsychology 7 464plusmn 475
Davidson RJ Larson C amp Abercrombie H (1995 October) Prefrontal electrophysio-logical asymmetries differentiate between melancholic and non-melancholic depressionPaper presented at the annual meeting of the Society for Research in PsychopathologyIowa City IO
Davidson RJ Marshall JR Tomarken AJ amp Henriques JB (submitted) While aphobic waits Regional brain electrical and autonomic activity predict anxiety in socialphobics during anticipation of public speaking
Davidson RJ Schaffer CE amp Saron C (1985) Effects of lateralized presentations offaces on self-reports of emotion and EEG asymmetry in depressed and non-depressedsubjects Psychophysiology 22 353plusmn 364
Davidson RJ amp Tomarken AJ (1989) Laterality and emotion An electrophysiologicalapproach In F Boller amp J Grafman (Eds) Handbook of neuropsychology (pp 419plusmn441) Amsterdam Elsevier
442 HELLER AND NITSCHKE
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
Dawson ME amp Schell AM (1982) Electrodermal responses to attended and non-attended stimuli during dichotic listening Journal of Experimental Psychology HumanPerception and Performance 8 82plusmn 86
Drevets WC Videen TO Preskorn SH Price JL Carmichael ST amp RaichleME (1992) A functional anatomical study of unipolar depression Journal of Neu-roscience 12 3628plusmn 3641
Eysenck HJ (1957) The dynamics of anxiety and hysteria New York PraegerFlor-Henry P (1976) Lateralized temporal-limbic dysfunction and psychopathology An-
nals of the New York Academy of Sciences 280 777plusmn 795
Flor-Henry P (1979) On certain aspects of the localization of the cerebral systemsregulating and determining emotion Biological Psychiatry 14 677plusmn 698
Gainotti G (1972) Emotional behavior and hemisphere side of lesion Cortex 8 41plusmn 55Gainotti G Caltagirone C amp Zoccolotti P (1993) Leftright and corticalsubcortical
dichotomies in the neuropsychological study of human emotions Cognition and Emotion7 71plusmn 93
George MS Ketter TA Perekh P et al (1994) Spatial ability in affective illnessDifferences in regional brain activation during a spatial matching task NeuropsychiatryNeuropsychology and Behavioral Neurology 7 143plusmn 153
George MS Ketter TA Perekh PI Horowitz B Herscovitch P amp PostRM (1995) Brain activity during transient sadness and happiness in healthy womenAmerican Journal of Psychiatry 152 341plusmn 351
Goldstein SG Filskov SB Weaver LA amp Ives JO (1977) Neuropsychologicaleffects of electroconvulsive therapy Journal of Clinical Psychology 37 187plusmn 197
Greenwald MK Cook EW amp Lang PJ (1989) Affective judgment and psychophy-siological response Dimensional covariation in the evaluation of pictorial stimuliJournal of Psychophysiology 3 51plusmn 64
Gruzelier J Seymour K Wilson L Jolley A amp Hirsch S (1988) Impairments onneuropsychologic tests of temporohippocampal and frontohippocampal functions andword macr uency in remitting schizophrenia and affective disorders Archives of GeneralPsychiatry 45 623plusmn 629
Gur RC Gur RE Skolnick BE Resnick SM Silver FL Chowluk J Muenz LObrist WD amp Reivich M (1988) Effects of task difreg culty on regional cerebral bloodmacr ow Relationships with anxiety and performance Psychophysiology 25 229plusmn 232
Heilman KM Schwartz HD amp Watson RT (1978) Hypoarousal in patients with theneglect syndrome and emotional indifference Neurology 28 229plusmn 232
Heller W (1986) Cerebral organization of emotional function in children UnpublishedDoctoral Dissertation University of Chicago
Heller W (1990) The neuropsychology of emotion Developmental patterns and implica-tions for psychopathology In N Stein BL Leventhal amp T Trabasso (Eds) Psycho-lolgical and biological approaches to emotion (pp 167plusmn 211) Hillsdale NJ LawrenceErlbaum Associates Inc
Heller W (1993a) Gender differences in depression Perspectives from neuropsychologyJournal of Affective Disorders 29 129plusmn 143
Heller W (1993b) Neuropsychological mechanisms of individual differences in emotionpersonality and arousal Neuropsychology 7 476plusmn 489
Heller W Etienne MA amp Miller GA (1995) Patterns of perceptual asymmetry indepression and anxiety Implications for neuropsychological models of emotion andpsychopathology Journal of Abnormal Psychology 104 327plusmn 333
Heller W amp Nisenson L (1993 March) Individual Differences in Characteristic Percep-tual Asymmetries Predict Inter- versus Intrapersonal Orientation Poster presented at theannual meeting of the New York Academy of Sciences New York City
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 443
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
Heller W amp Nitschke JB (1997) Regional brain activity in emotion A framework forunderstanding cognition in depression Cognition and Emotion 11 637plusmn 661
Heller W Nitschke JB Etienne MA amp Miller GA (1997a) Regional brain activitypatterns differentiate types of anxiety Journal of Abnormal Psychology 106 376plusmn 385
Heller W Nitschke JB amp Lindsay DL (1997b) Neuropsychological correlates ofarousal in self-reported emotion Cognition and Emotion 11 383plusmn 402
Henriques JB amp Davidson RJ (1990) Regional brain electrical asymmetries discrimi-nate between previously depressed and healthy control subjects Journal of AbnormalPsychology 99 22plusmn 31
Henriques JB amp Davidson RJ (1991) Left frontal hypoactivation in depression Jour-nal of Abnormal Psychology 100 535plusmn 545
Hiller W Zaudig M amp Rose M (1989) The overlap between depressionand anxiety ondifferent levels of psychopathology Journal of Affective Disorders 16 223plusmn 231
Hugdahl K Franzon M Andersson B amp Walldebo G (1983) Heart-rate responses(HRR) to lateralizedvisual stimuli Pavlovian Journal of Biological Science 18 186plusmn 198
Jaeger J Borod JC amp Peselow ED (1987) Depressedpatients have atypical hemispacebiases in the perception of emotional chimeric faces Journal of Abnormal Psychology96 321plusmn 324
Johnsen BH amp Hugdahl K (1991) Hemispheric asymmetry in conditioning to facialemotional expressions Psychophysiology 28 154plusmn 162
Johnsen BH amp Hugdahl K (1993) Right hemisphere representation of autonomicconditioning to facial emotional expressions Psychophysiology 30 274plusmn 278
Katon W amp Roy-Byrne PP (1991) Mixed anxiety and depression Journal of AbnormalPsychology 100 337plusmn 345
Kinsbourne M (1988) Hemisphere interactions in depression In M Kinsbourne (Ed)Cerebral hemisphere function in depression (pp 133plusmn 162) Washington DC AmericanPsychiatric Press
Klein DF (1987) Anxiety reconceptualized In DF Klein (Ed) Anxiety (pp 1plusmn 35)Basel Karger
Kronfol Z Hamsher KD Digre K amp Waziri R (1978) Depression and hemisphericfunctions Changes associated with unilateral ECT British Journal of Psychiatry 132560plusmn 567
Ladavas E Nicoletti R Umilta C amp Rizzolatti G (1984) Right hemisphere interfer-ence during negative affect A reaction time study Neuropsychologia 22 479plusmn 485
Lang PJ Greenwald MK Bradley MM amp Hamm AO (1993) Looking at picturesAffective facial visceral and behavioral reactions Psychophysiology 30 261plusmn 273
Larsen RJ amp Diener E (1992) Promises and problems with the circumplex model ofemotion In MS Clark (Ed) Review of personality and social psychology (pp 25plusmn 59)Newbury Park CA Sage
LeDoux JE (1993) Emotional networks in the brain In M Lewis amp JM Haviland (Eds)Handbook of emotions (pp 109plusmn 118) New York Guilford Press
Lee GP Loring DW Meador KJ Flanigin HF amp Brooks BS (1988) Severebehavioral complications following intracarotid sodium amobarbital injection Implica-tions for hemispheric asymmetry of emotion Neurology 38 1233plusmn 6
Levy J Heller W Banich MT amp Burton LA (1983) Are variations among right-handed individuals in perceptual asymmetries caused by characteristic arousal differ-ences between hemispheres Journal of Experimental Psychology Human Perceptionand Performance 9 329plusmn 359
Liotti M amp Tucker DM (1995) Emotions in asymmetric corticolimbic networks In RJDavidson amp KM Hugdahl (Eds) Brain Asymmetry (pp 389plusmn 423) Cambridge MAMIT Press
444 HELLER AND NITSCHKE
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
Liotti M Sava D Rizzolatti G amp Caffarra P (1991) Differential hemispheric asym-metries in depression and anxiety A reaction time study Biological Psychiatry 29887plusmn 899
Martinot JH Allilaire JF Mazolyer BM Hantouche E Huret JD Legaut-DemareFDeslauriersAGHardyPPappataS BaronJCamp SyrotaA (1990) Obsessive-compulsive disorder A clinical neuropsychological and positron emission tomographystudy Acta Psychiatrica Scandinavica 82 233plusmn 242
Mathew RJ Wilson WH amp Daniel DG (1985) The effect of nonsedative doses ofDiazepam on regional cerebral blood macr ow Biological Psychiatry 20 1109plusmn 1116
Meyer TJ Miller ML Metzger RL amp Borkovec TD (1990) Development andvalidation of the Penn State Worry Questionnaire Behaviour Research Therapy 28487plusmn 495
Molina S amp Borkovec TD (1994) The Penn State Worry Questionnaire Psychometricproperties and associated characteristics In GCL Davey amp F Tallis (Eds) WorryingPerspectives on theory assessment and treatment (pp 265plusmn 283) Chichester Wiley
Morrow L Vrtunski PB Kim Y amp Boller F (1981) Arousal responses to emotionalstimuli and laterality of lesions Neuropsychologia 19 65plusmn 71
Mountz JM Modell JG Wilson MW Curtis GC Lee MA Schmaltz S amp KuhlDE (1989) Positron emission tomographic evaluation of cerebral blood macr ow duringstate anxiety in simple phobia Archives of General Psychiatry 46 501plusmn 504
Myslobodsky MS amp Horesh N (1978) Bilateral electrodermal activity in depressivepatients Biological Psychiatry 6 111plusmn 120
Nitschke JB Heller W Etienne MA amp Miller GA (1995) Specireg city of frontal EEGasymmetry in anxiety and depression during emotion processing [Abstract] Psychophy-siology 32 S56
Nitschke JB Heller W Etienne MA Taitano EK Cecola TA amp Miller GA(1994) Moderating effects of extraversion on regional brain activity in anxiety [Ab-stract] Psychophysiology 31 S72
Nitschke JB Heller W Imig J amp Miller GA (submitted) Distinguishing dimen-sions of anxiety and depression
Nitschke JB Heller W amp Miller GA (in prep) Attentional bias in the physiology andpsychology of extraversion A neuropsychological replacement
Nordhal TE Semple WE Gross M Mellman TA Stein MB Goyer P King ACUhlde TW amp Cohen RM (1990) Cerebral glucose metabolic differences in patientswith panic disorder Neuropsychopharmacology 3 261plusmn 272
Post RM DeLisi LE Holcomb HH Uhde TW Cohen R amp BuchsbaumM (1987) Glucose utilization in the temporal cortex of affectively ill patients Positronemission tomography Biological Psychiatry 22 545plusmn 553
Reiman EM Raichle ME Butler FK Herscovitch P amp Robins E (1984) A focalbrain abnormality in panic disorder a severe form of anxiety Nature 310 683plusmn 685
Reivich M Gur R amp Alavi A (1983) Positron emission tomographic studies of sensorystimuli cognitive processes and anxiety Human Neurobiology 2 25plusmn 33
Robinson DL Boston JD Starkstein SE amp Price TR (1988) Comparison of maniaand depression after brain damage Causal factors American Journal of Psychiatry 145142plusmn 148
Sackeim HA Greenberg MS Weiman AL Gur RC Hungerbuhler JP amp Gesch-wind N (1982) Hemispheric asymmetry in the expression of positive and negativeemotions Neurological evidence Archives of Neurology 39 210plusmn 218
Schaffer CE Davidson RJ amp Saron C (1983) Frontal and parietal electroencephalo-gram asymmetry in depressed and nondepressed subjects Biological Psychiatry 18753plusmn 762
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 445
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
Stewart RS Devous MD Rush AJ Lane L amp Bonte FJ (1988) Cerebral bloodmacr ow changes during sodium-lactate-induced panic attacks American Journal of Psychia-try 145 442plusmn 449
Swedo SE Schapiro MB Grady CL Cheslow DL Leonard HL Kumar AFriedland R Rapoport SI amp Rapoport JL (1989) Cerebral glucose metabolismin childhood-onset obsessive-compulsive disorder Archives of General Psychiatry 46518plusmn 523
Tomarken AJ amp Davidson RJ (1994) Frontal brain activation in repressors and non-repressors Journal of Abnormal Psychology 103 339plusmn 349
Tomarken AJ Davidson RJ Wheeler RE amp Doss RC (1992) Individual differencesin anterior brain asymmetry and fundamental dimensions of emotion Journal of Person-ality and Social Psychology 62 676plusmn 687
Tomarken AJ Davidson RJ Wheeler RE amp Kinney L (1992) Psychometric proper-ties of resting anterior EEG asymmetry Temporal stability and internal consistencyPsychophysiology 29 576plusmn 592
Tucker DM (1981) Lateral brain function emotion and conceptualizationPsychologicalBulletin 89 19plusmn 46
Tucker DM (1988) Neuropsychological mechanisms of affective self-regulation In MKinsbourne (Ed) Cerebral hemisphere function in depression (pp 99plusmn 131) Washing-ton DC American Psychiatric Press
Tucker DM Antes JR Stenslie CE amp Barnhardt TM (1978) Anxiety and lateralcerebral function Journal of Abnormal Psychology 87 380plusmn 383
Tucker DM Roth RS Arneson BA amp Buckingham V (1977) Right hemisphereactivation during stress Neuropsychologia 15 697plusmn 700
Tucker DM Stenslie CE Roth RS amp Shearer SL (1981) Right frontal lobe activa-tion and right hemisphere performance Decrement during a depressedmood Archives ofGeneral Psychiatry 38 169plusmn 174
Tucker DM amp Williamson PA (1984) Asymmetric neural control systems in humanself-regulation Psychological Review 91 185plusmn 215
Tyler SK amp Tucker DM (1982) Anxiety and perceptual structure Individual differ-ences in neuropsychological function Journal of Abnormal Psychology 91 210plusmn 220
Uytdenhoef P Portelange P Jacquy J Charles G Linkowski P amp MendlewiczJ (1983) Reigonal cerebral blood macr ow and lateralized hemispheric dysfunction indepression British Journal of Psychiatry 143 128plusmn 132
Valenstein E amp Heilman KM (1984) Emotional disorders resulting from lesions of thecentral nervous system In KM Heilman amp E Valenstein (Eds) Clinical neuropsychol-ogy (pp 413plusmn 438) New York Oxford University Press
Watson D Clark LA Weber K Assenheimer JS Strauss ME amp McCormickRA (1995a) Testing a tripartite model II Exploring the symptom structure of anxietyand depression in student adult and patient samples Journal of Abnormal Psychology104 15plusmn 25
Watson D Weber K Assenheimer JS Clark LA Strauss ME amp McCormickRA (1995b) Testing a tripartite model I Evaluating the convergent and discriminantvalidity of anxiety and depression symptom scales Journal of Abnormal Psychology104 3plusmn 14
Wittling W (1990) Psychophysiological correlates of human brain asymmetry Bloodpressure changes during lateralized presentation of an emotionally laden reg lm Neuro-psychologia 28 457plusmn 470
Wittling W (1995) Brain asymmetry in the control of autonomic-physiologic activity InRJ Davidson amp K Hugdahl (Eds) Brain asymmetry (pp 305plusmn 357) Cambridge MAMIT Press
446 HELLER AND NITSCHKE
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447
Wu JC Buchsbaum MS Hershey TG Hazlett E Sicotte N amp JohnsonJC (1991) PET in generalized anxiety disorder Biological Psychiatry 29 1181plusmn 1199
Yeudall LT Schopmacr ocher D Sussman PS Barabash W Warneke IB Gill D OttoW Howarth B amp Termansen PE (1983) Panic attack syndrome with and withoutagoraphobia Neuropsychological and evoked potential correlates In P Flor-Henry amp JGruzelier (Eds) Laterality and psychopathology (pp 195plusmn 216) New York Elsevier
Zimmerman M Coryell W amp Pfohl B (1986) Melancholic subtyping A qualitative orquantitative distinction American Journal of Psychiatry 143 98plusmn 100
Zoccolotti P Caltagirone C Benedetti N amp Gainotti G (1986) Perturbation desresponses veAcirc geAcirc tatives aux stimuli eAcirc motionnels au cours des leAcirc sions heAcirc mispheAcirc riquesunilaterales EnceAcirc phale (Paris) 12 263plusmn 268
Zoccolotti P Scabini D amp Violani C (1982) Electrodermal responses in patients withunilateral brain damage Journal of Clinical Neuropsychology 4 143plusmn 150
BRAIN ACTIVITY IN DEPRESSION AND ANXIETY 447