Cognitive function in euthymic bipolar patients, stabilizedschizophrenic patients, and healthy controls

Alessandro Rossi a,*, Luca Arduini b, Enrico Daneluzzo a, Massimiliano Bustini a,Pierluigi Prosperini b, Paolo Stratta c

aDepartment of Clinical Psychology at ``Villa Serena Medical Center'', Viale L.Petruzzi, 19, CittaÁ S. Angelo, 65013,

Pescara, ItalybInstitute of Experimental Medicine, University of L'Aquila, ItalycDepartment of Psychiatry, S. Salvatore Hospital, L'Aquila, Italy

Received 14 February 2000; received in revised form 28 May 2000; accepted 29 June 2000

Abstract

Studies on cognitive function in bipolar disorder have led to contrasting results and few data are available on a�ected subjectsduring the euthymic phase. In the present study we investigated the cognitive function of a cohort of bipolar (n=40) and schizo-

phrenic (n=66) patients compared to healthy controls (n=64). Patients were evaluated in the outpatient setting over at least 3months using a computerized version of Wisconsin Card Sorting Test. Schizophrenic patients showed the worst performance whilethat of the bipolar patients was somewhere between schizophrenic and controls. A discriminant analysis was able to classify cor-

rectly 60.59% of the subjects (schizophrenics 48.5%, bipolars 40%; healthy controls 85.9%). The scores of the Wisconsin CardSorting Test were entered into a principal component analysis, which yielded a 2-factor solution. Even in that analysis bipolarpatients showed intermediate features in comparison with the other groups. These data indicate that bipolar patients have subtleneurocognitive de®cits even after the resolution of an a�ective disorder. As well as observing quantitative di�erences between

groups, the results show di�erent dimensions of cognitive performance within groups suggesting that the de®cit of euthymic bipo-lars could be a dishomogeneous entity, probably more heterogeneous than that in schizophrenia. Studies administering a morecomplete neuropsychological battery could further clarify the nature and meaning of the cognitive de®cits in schizophrenia and

bipolar disorder. # 2000 Elsevier Science Ltd. All rights reserved.

Keywords: Cognitive impairment; Context; Wisconsin Card Sorting Test; Factor analysis; Bipolar disorder; Schizophrenia

1. Introduction

Several lines of evidence demonstrate that schizo-phrenia is an illness characterized by multiple defects infundamental cognitive processes (Gold et. al., 1991;Goldberg and Gold, 1995) while less attention has beenpaid to these aspects in bipolar disorders (Altshuler1993; Goldberg 1999).Neurocognitive research in schizophrenia has broadly

evidenced defects in fundamental cognitive processesincluding attention, language processing and problemsolving. It has been established that de®ciencies in cog-nitive functions could be related to several aspects of

symptomatology and prevent patients from attaining anoptimal adaptation in their daily lives (Green, 1996).Among the variety of these impairments, those in`executive processes' play a relevant role in restrictingpatients ability to retain, acquire, or relearn skills thatare needed for real-world functioning (Keefe, 1995).The Wisconsin Card Sorting Test (WCST) is the mostcommonly used measure of executive functions in termsof concept formation and cognitive ¯exibility and alarge body of literature has consistently shown impair-ment on this task signi®cantly associated with resultantdisability (Jaeger and Berns, 1999).Bipolar disorder is a recurrent illness with signi®cant

disability and heterogeneous outcome (Goodwin andJamison 1990). Several studies have reported the out-come for patients with bipolar disorder to be generallygood, but a sub-population of approximately 5±34%

0022-3956/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.

PI I : S0022-3956(00 )00025-X

Journal of Psychiatric Research 34 (2000) 333±339

www.elsevier.com/locate/jpsychires

* Corresponding author. Tel.: +39-085-9590201; fax: +39-085-

9590400.

E-mail address: a.rossi@villaserena.it (A. Rossi).

has consistently been described as having poor socialoutcome or poor response to treatment (Winokur et al.,1969; Carlson et al., 1974; Johnstone et al., 1985; Har-row et al., 1990). Moreover, evidence has been accumu-lated suggesting that a cognitive de®cit can be identi®edin some bipolar patients: persistent cognitive de®citshave been reported in up to 32% of bipolar patients(Astrup et al., 1959; Bratfos and Haugh, 1968; Dhingraand Rabins, 1991; Stratta et al. 1995), and Ho� et al.(1990) failed to di�erentiate bipolar disorder from schi-zophrenia on several measures of neuropsychologicalfunction.Despite numerous studies demonstrating cognitive

impairment in symptomatic a�ective patients, only afew investigations have been conducted in the euthymicstate (Frangou and Bettany 1999). In these studies cog-nitive impairment has been observed to persist despiteimprovement in clinical state even after an a�ectiveepisode resolves (Friedman et al., 1977; Savard et al.,1980; Sapin et al., 1987; Waddington et al. 1989; Co�-man et al., 1990; Morice, 1990; van Gorp et al., 1998;McGrath et al., 1997). Friedman et al. (1977) found thaton the Halstead Reitan battery, four out of ®ve older(ages 59±68) euthymic bipolar patients manifested cog-nitive impairment that was greater than age alone couldexplain. Savard et al. (1980) found that bipolar patientswere impaired in the Category Test relative to othersubjects. Sapin et al. (1987) demonstrated in 20 euthy-mic medication free bipolar patients no signi®cant dif-ferences in information processing compared withcontrols. Co�man et al. (1990) reported that bipolarpatients showed signi®cant levels of di�usely cognitiveimpairment when compared with controls. Dhingra andRabins (1991) followed 25 bipolar patients 60 years orolder for 5±7 years: they found that 20% of the cohorthad become cognitively or functionally impaired,requiring permanent nursing home placement; 32% ofpatients developed clinically signi®cant cognitive dis-orders. Atre-Vaidya et al. (1998) con®rmed these ®nd-ings and reported that memory de®cit was associatedwith poor psychosocial functioning. Ferrier et al. (1999)reported an impairment of executive function in a groupof 41 euthymic bipolar patients compared to 20 con-trols.Cognitive de®cits may worsen with time and a�ective

episodes. Kessing (1998) in a mixed population of uni-polar and bipolar euthymic patients reported that thosewith recurrent episodes were signi®cantly more cogni-tively impaired than those with a single episode andmore impaired than controls. There was no di�erence inthe severity of the dysfunction between unipolar andbipolar patients. Tham et al. (1997) reported similar®ndings suggesting that a subgroup of patients withrecurring mood disorder and more relapses and epi-sodes of hospitalization were characterized by a certaindegree of cognitive dysfunction.

Despite numerous reports of morphological brainabnormalities in bipolar disorders, the structural corre-lates of these cognitive disturbances remain to be eluci-dated (Pearlson et al., 1984; Hauser et al., 1989;Nasrallah et al., 1989; Rossi et al., 1991; Jurius et al.,1993; Videbech, 1997; Ste�ens and Krishnan 1998;Strakowsky et al., 1999).Because the presence of cognitive impairment has

been held to be uncommon in bipolar disorder andmore likely in schizophrenia, the cognitive function wasassessed, using the Wisconsin Card Sorting Test(WCST), in a sample of bipolar and schizophrenicpatients comparing their performance with that of con-trols.

2. Methods and materials

2.1. Subjects

The subjects were 66 patients (38 men and 28 women)who meet the DSM-III-R criteria for schizophrenia, and40 bipolar patients (26 men and 14 women). A seniorpsychiatrist (AR) who personally interviewed thepatients according to the Structured Clinical Interviewfor DSM-III-R (Spitzer et al., 1987) made diagnoses.The mean age of the schizophrenic patients was 33.14

years (S.D. 8.67), and educational level was 10.05 years(S.D. 3.14). The age at onset of symptoms was 21.09years (S.D. 5.6) with a mean duration of illness of 17.15years (S.D. 15.14). All schizophrenic patients were tak-ing classical antipsychotics, and the mean chlorproma-zine-equivalent dose (Kessler and Waletzky, 1981) was545.13 (S.D. 415.12) at the time of the evaluation.The mean age of the bipolar patients was 35.7 years

(S.D. 11.67), and educational level was 11.48 years (S.D.3.37). The age at onset of symptoms was 24.66 years(S.D. 8.2) with a mean duration of illness of 14.15 (S.D.13.12). They were taking classical neuroleptics (all sub-jects), lithium (n.15), carbamazepine (n.7) and valproicacid (n.20). The mean chlorpromazine-equivalent dosewas 250.12 (S.D. 105.80) at the time of the evaluation.The healthy controls were 64 subjects (30 men and 34

women) chosen among employees and relatives oracquaintances of the hospital sta� working in clinical andadministrative areas. The mean age was 26.4 years (S.D.5.44), and educational level was 14.69 years (S.D. 2.99).All the subjects were screened for history of head

injury, alcohol abuse or serious neurological or physicaldisease. Healthy control subjects were screened for anycurrent or past history of psychiatric disorder. Theywere also screened for absence of family history of psy-choses. None of these subjects showed any psychoticdisturbance or any personality disorders. All the sub-jects were right-handed, according to the EdinburghInventory (Old®eld, 1971).

334 A. Rossi et al. / Journal of Psychiatric Research 34 (2000) 333±339

All the subjects provided written informed consentafter complete description of the study, in accordancewith the university institutional review board.

2.2. Procedure

The WCST was administered to all enrolled subjectswith the standard instructions by Heaton (1981) using acomputerized version also implemented with MEL(Micro Experimental Laboratory) (Schneider, 1989).The computer automatically scored performances, fol-lowing Heaton's rules, with the exception that the ®rstunambiguous error repeating the previously correctprinciple was not scored as perseveration because thesubject had not received feedback indicating that thepreviously correct principle was incorrect. This scoringrule resulted in negligible changes in the perseverativeerror score. The performance on the WCST was basedon ``number of categories achieved'', ``number of totalerrors'', ``numbers of perseverative errors'' and ``uniqueerrors''. Subjects with schizophrenic and bipolar dis-orders were observed in the outpatient setting for atleast 3 months to ensure stable state of illness andeuthymia before the experimental procedure was admi-nistered.

2.3. Statistical analysis

The WCST scores of the three groups where analyzedwith a MANOVA. Because of the signi®cant di�erencebetween patients and controls in age and educationallevel (F=17.5 and 26.08, respectively, df 2.169, P<0.001) such variables were used as covariates in all thebetween group comparisons. Planned univariate con-trasts among groups were also performed. A dis-criminant analysis (DA) was conducted to evaluate thepercentage of cases correctly classi®ed by the WCSTindexes, age and educational level. A factor analysis(FA), principal component with Varimax rotation, wasused to identify the major WCST dimensions along eachgroup (Norusis, 1992).

We used Pearson's correlation to investigate theassociation between the indexes of the cognitive task,illness duration and antipsychotic equivalents. All ana-lyses yielding a P value of 0.05 were considered sig-ni®cant.

3. Results

Table 1 shows WCST ®ndings in the three groups.When age and educational level were covariated, schi-zophrenics did di�er from controls on categories, totalerrors, and unique errors. Bipolar patients did not sta-tistically di�er from controls but di�ered from schizo-phrenic patients for achieved categories and total errors.No statistically signi®cant di�erences in perseverativeerrors were seen in all contrasts. Table 2 shows theresults of the discriminant analysis on WCST, age andeducational level. Discriminant analysis produced asatisfactory degree of separation as indicated by ®nalWilks' lambda (0.9554). The analysis yielded 60.59% ofgrouped cases correctly classi®ed, higher for controls(85.9%), lower for bipolars (40%) and intermediate forschizophrenics (48.5%).Table 3 shows the FA performed on the WCST scores

for each group. Using varimax rotation, two factorsemerged with Eigenvalues >1. A ®rst factor accountedfor 56.4, 59.4 and 62.3% of the original variance amongschizophrenics, bipolars and controls, respectively. Per-severative and total errors highly loaded on this factorin the three groups while in the control group only theachieved categories also loaded.The second factor accounted for 28.2, 24.7 and 23.1%

of the original variance among schizophrenics, bipolarsand controls, respectively.Unique errors characterized this factor in all three

groups while achieved categories showed a di�erentpattern: high loading on the second factor in schizo-phrenics, high loading on the ®rst factor in controls andoverlapping loading on both factors in bipolars. Nostatistically signi®cant correlations between the indexes

Table 1

MANOVA and univariate results for WCST indices of schizophrenics, bipolars and controls with age and educational level as covariatesa

Raw scores (means�S.D.) Adjusted means Contrasts

Scz (n=66)b Bip (n=40)c Con(n=64)d Scz Bip Con Scz vs Con Bip vs Con Bip vs Scz

WCST t P< t P< t P<

Categories 3.27�2.34 3.85�2.23 5.33�1.49 3.41 4.06 4.98 3.29 0.001 0.33 NS 3.19 0.001

Perseverative errors 17.54�11.37 15.47�11.02 8.81�8.69 15.98 13.92 11.93 1.84 NS 0.02 NS 1.58 NS

Total errors 34.04�16.00 29.00�16.30 17.81�14.17 31.91 26.99 21.95 2.98 0.005 0.02 NS 2.61 0.01

Unique errors 3.95�6.42 2.42�3.78 1.34�2.43 3.59 2.06 2.07 1.96 0.05 0.87 NS 0.85 NS

a Hotellings=0.11, P<0.03.b Scz = schizophrenics.c Bip=bipolars.d Con = controls.

A. Rossi et al. / Journal of Psychiatric Research 34 (2000) 333±339 335

of the cognitive task, illness duration, age at onset andcurrent antipsychotic medication equivalents wereobserved.

4. Discussion

Because of the widely reported ®nding of WCSTabnormalities in schizophrenia (Goldberg and Gold,1995), we restrict our discussion to the ®nding of cog-nition in bipolar disorder.We approached the issue of cognitive dysfunction in

bipolar disorders comparing these patients with a groupof schizophrenic patients for whom there is a wide con-sensus of the presence of cognitive dysfunction, and agroup of healthy controls. We looked for quantitativedi�erences between groups, and looked for identifyingdi�erent dimensions of cognitive performance withingroups. With this aim we adopted the WCST that is acomplex problem-solving task that involves multiplecognitive processes rather than a single function(Anderson et al., 1991; Dehaene and Changeux 1991).Its scoring system in fact does not simply measure aunitary function but o�ers opportunity to di�erentiatedissociable cognitive processes. As the literature sug-gests, the indexes we used, achieved categories, perse-verative errors and unique responses could identifyseveral constructs within the WCST performance (Ros-sel and David, 1998).These indexes did not di�erentiate bipolars from

controls, while they did schizophrenics versus controlswith the exception of perseverative errors. The compar-

isons between bipolars and schizophrenics showed anintermediate case where perseverative and unique errorsdid not di�er between the two groups. This ®nding canbe further discussed in the light of the FA we report. Inthe three samples we found a similar two-factor struc-ture. In the ®rst factor, perseverative and total errorswere included, while unique errors were heavily loadedon the second factor. The load of the number ofachieved categories was quite di�erent in the threegroups. When the healthy control sample was con-sidered, the number of achieved categories were loadedon the ®rst factor together with perseverative and totalerrors while in the schizophrenic sample the number ofachieved categories were loaded on the second factortogether with unique errors. When bipolar patients wereconsidered, achieved categories were loaded, instead,almost similarly on both the factors. Therefore,depending upon the sample that was considered, thenumber of achieved categories seems to be a nonspeci®cindex that assumes a di�erent meaning. The ®rst factorcan be interpreted as ``perseveration'' in agreement withSullivan et al. (1993). The second factor is heavily loa-ded on unique responses, and it is likely to be con-sidered more than a residual factor (i.e. a non-perseverative factor) as by Sullivan et al. (1993).Sullivan et al. (1993) in a factor analysis study of

WCST scores from a composite sample of schizo-phrenics, alcoholics and healthy controls, found a three-factor solution. The ®rst factor called ``perseveration''included the number of achieved categories with perse-verative errors, the second factor de®ned ``ine�cientsorting'', included chie¯y failure to maintain set; the

Table 2

Classi®cation results of discriminant analysisa

Actual group Predicted group

Schizophrenics Bipolars Controls

Schizophrenics (n=66) 32 48.5% 24 36.3% 10 15.2%

Bipolars (n=40) 12 30% 16 40% 12 30%

Controls (n=64) 5 7.8% 4 6.3% 55 85.9%

a Percent of ``grouped'' cases correctly classi®ed=60.59%.

Table 3

Factor analysis with Varimax rotation of the WCST conventional scores

Schizophrenics (n=66) Bipolars (n=40) Controls (n=64)

Factor 1 Factor 2 Factor 1 Factor 2 Factor 1 Factor 2

Variance proportion 56.4 28.2 59.4 24.7 62.3 23.1

WCST scores Factor loadings

Categories ÿ0.379 ÿ0.723* ÿ0.564 ÿ0.525 ÿ0.793* ÿ0.012Perseverative errors 0.969* ÿ0.065 0.963* ÿ0.049 0.92* 0.083

Total errors 0.78* 0.576 0.933* 0.275 0.932* 0.263

Unique errors ÿ0.108 0.905* 0.038 0.945* 0.108 0.991

*scores with high and relatively non overlapping loading on the factor.

336 A. Rossi et al. / Journal of Psychiatric Research 34 (2000) 333±339

third factor was considered as a ``non perseverative fac-tor'', which included unique responses.Along with the Heaton (1981) de®nition, unique

errors are the number of sorts to stimulus cards that donot match on any of the three major dimensions ofcolor, form and number established by the task. Uniqueresponses are those out of the context given by the set ofinstructions of the test. Taking into account the proces-sing of ``context information'' theory of Cohen andServan-Schreiber (1992), our second factor can beinterpreted as a ``context factor''. According to thehypothesis of these authors, the ``context information''is an internal representation of information that has tobe constructed and maintained in the mind in a way thatit can be used to mediate an appropriate behavioralresponse. In other words, the subject who is performinga cognitive task must keep in mind a set of necessaryinstructions, i.e. the subject has to construct an internalrepresentation of the task context in order to identifythe relevant stimuli to be taken into account to drive thebehavior through an appropriate response.The factor pattern of the control subjects is similar to

that reported by Sullivan et al. (1993). This ``persevera-tion factor'' can be linked to the Dorso-Lateral-Pre-frontal Cortex (DLPFC) function. WCST is the onlytest known from lesion studies to be sensitive speci®-cally to the DLPFC function. Milner (1963) reportedthat the index of the DLPFC dysfunction was a lowernumber of achieved categories and an increase in perse-verative errors than controls. When the DLPFC isdamaged perseveration become the central character-istic of the pathology and the number of achieved cate-gories is limited from this condition (Lezak, 1995). Onthe other hand, when the DLPFC is well functioning,the successful performance in terms of achieved cate-gories and perseverative errors is linked to a unitaryfunction.The factorial pattern of the schizophrenic patients

could be due to a di�erent pathological process: a pro-cess involving brain regions other than the DLPFCconvexity, particularly their connections. This processwould lead to the load of the index-achieved categorieson the factor we interpreted as context factor. This fac-tor could be considered the crucial element determiningthe schizophrenic cognitive de®cit (Cohen and Servan-Schreiber 1992; Servan-Schreiber et al, 1996; Stratta etal, 1998).Bipolar patients show an intermediate condition

between the schizophrenic and control patterns: thenumber of achieved categories loads almost in the sameway on both factors. This observation can be consideredsu�cient for an indication of the DLPFC dysfunction inbipolar patients? Our results do not permit an answer tothis question. However, these patients show a di�erentpattern from that of schizophrenics and controls. Theimplication of brain regions other than the dorsolateral

frontal convexity or its connections could be speculatedupon. Flor-Henry and Yeudall (1979) o�ered persuasive®ndings supporting the hypothesis that neuropsycholo-gical de®cit in schizophrenia could be related to leftprefrontal dysfunction while bipolar illness showedbilateral or predominantly right prefrontal dysfunction.Interestingly studies based on known lesions and usingWCST failed to di�erentiate between left- and right-sided lesions (Milner, 1963).In this light we could interpret the widespread dis-

tribution of bipolar patients in the ``predicted groups''of the DA putatively re¯ecting a more di�use or di�er-ently lateralized, and perhaps less severe, pre-frontaldysfunction. It is tempting to speculate that the di�erentfactorial pattern we found could indirectly support theFlor-Henry and Yeudall (1979) suggestions, at leastinsofar as the cause of the WCST dysfunction may dif-fer between schizophrenic and bipolar patients.A plausible alternative or complementary hypothesis

has to be considered.The WCST is a neuropsychological device that is

considered quite sensitive to a prefrontal de®cit (Milner,1963; Weinberger et al., 1986), but the speci®city of sucha de®cit has not uniformly replicated (Robinson et al.,1980; Anderson et al., 1991). It is probable that complexhigher cognitive functions, evaluated by WCST, involvea network of neural interconnections (Frith and Done,1988; Weinberger, 1993; Goldberg et al., 1994). A defectin one part of this network could result in a dysfunc-tional reverberation throughout the network. It couldtherefore be hypothesized that the cognitive de®cit pat-tern, as assessed by WCST, represents a ®nal commonpathway disorder in the two groups (Zihl et al., 1998),not a characteristic restricted to schizophrenia only.This hypothesis could be supported by the DA functionwe reported that is not su�cient for exhaustively classi-fying patients and controls as well as in agreement withevidences of similar cognitive de®cits in schizophreniaand a�ective disorders (Jeste et al., 1996).However, some limitations could a�ect the results of

this study. First, even though some studies reported thatmedication at time of testing did not in¯uence the cog-nitive evaluation (Jo�e et al. 1988; Ho� et al., 1990) wecannot exclude that mood stabilizing and/or anti-psychotics could have in¯uenced the cognitive testing ofthe patient groups. Furthermore we cannot exclude thatthe premorbid intelligence could have had a di�erentimpact on the WCST in the two groups or that the twogroups could have di�erent results in other neu-ropsychological testing.Another limitation of the study is the fact that we

used WCST only to assess cognition in patients. Studiesusing a more complete neuropsychological battery eval-uating other cognitive domains, such as working mem-ory, attentional capacities and other central executivefunctions are, therefore, needed.

A. Rossi et al. / Journal of Psychiatric Research 34 (2000) 333±339 337

Bearing in mind these limitations our ®ndings maysuggest that cognition in euthymic mania is not ahomogeneous entity; although speculatively, it could beconsidered even more heterogeneous than cognition inschizophrenia. Further and more re®ned studies couldclarify the meaning and the correlates of this type ofcognitive dysfunction in bipolar disorder and their clin-ical implication as well.

References

Altshuler LL. Bipolar disorder: are repeated episodes associated with

neuroanatomic and cognitive changes? Biological Psychiatry

1993;33:563±5.

Anderson SW, Damasio H. Jones RD, Tranel D. Wisconsin Card

Sorting Test performance as a measure of frontal lobe damage.

Journal of Clinical and Experimental Neuropsychology

1991;13:909±22.

Astrup C, Fossum A, Holmboe RA. Follow-up study of 270 patients

with acute a�ective psychoses.. Acta Psychiatrica Scandinavica

1959;34(Suppl. 135):1±65.

Atre-Vaidya N, Taylor MA, Seidenberg M, Reed R, Perrine A, Glick-

Oberwise F. Cognitive de®cits, psychopathology, and psychosocial

functioning in bipolar mood disorder. Neuropsychiatry Neu-

ropsychology and Behavioral Neurology 1998;11:120±6.

Bratfos O, Haug JO. The course of manic-depressive psychosis. Acta

Psychiatrica Scandinavica 1968;44:89±112.

Carlson GA, Kotin J, Davenport Y, Adland M. Follow-up of 53

bipolar manic patients. British Journal of Psychiatry 1974;124:134±

9.

Co�man JA, Bornstein RA, Olson SC, Schwarzkopf SB, Nasrallah

HA. Cognitive impairment and cerebral structure by MRI in bipo-

lar disorder. Biological Psychiatry 1990;27:1188±96.

Cohen JD, Servan-Schreiber D. Context, cortex and dopamine: a

connectionist approach to behavior and biology in schizophrenia.

Psychological Review 1992;99:45±77.

Dehaene S, Changeux JP. The Wisconsin Card Sorting Test: theore-

tical analysis and modeling in a neural network. Cerebral Cortex

1991;1:62±79.

Dhingra U, Rabins PV. Mania in the elderly: a 5±7 year follow-up.

Journal of American Geriatric Society 1991;39:581±3.

Ferrier IN, Stanton BR, Kelly TP, Scott J. Neuropsychological func-

tion in euthymic patients with bipolar disorder. British Journal of

Psychiatry 1999;175:246±51.

Flor-Henry P, Yeudall IT. Neuropsychological investigation of schi-

zophrenic and manic-depressive psychoses. In: Gruzelier J, Floor-

Henry P, editors. Hemisphere asymmetries of function in psycho-

pathology. Amsterdam: Elsevier, 1979. p. 341±62.

Frangou S, Bettany D. The neuropsychology of bipolar a�ective dis-

order. Bipolar Disorder 1999;(Suppl. 1):1±30.

FriedmanM, Culver C, Ferrell R. On the safety of long-term treatment

with lithium. American Journal of Psychiatry 1977;134:1123±6.

Frith CD, Done DJ. Towards a neuropsychology of schizophrenia.

British Journal of Psychiatry 1988;153:437±43.

Gold JM, Goldberg TE, Kleinman JE, Weinberger DR. The impact of

symptomatic state and pharmacological treatment on cognitive func-

tioning of patients with schizophrenia and mood disorders.. In: Mohr

E, Brower P, editors. Handbook of clinical trials: the neurobehavioral

and injury. Amsterdam: Swetz and Zeitlinger, 1991. p. 185±214.

Goldberg TE. Some fairly obvious distinctions between schizophrenia

and bipolar disorder. Schizophrenia Research 1999;39:127±32.

Goldberg TE, Gold JM. Neurocognitive de®cits in schizophrenia. In:

Hirsch SR, Weinberger DR, editors. Schizophrenia. Oxford: Black-

well Science, 1995. p. 146±62.

Goldberg TE, Torrey EF, Berman KF, Weinberger DR. Relations

between neuropsychological performance and brain morphological

and psychological measures in monozygotic twins discordant for

schizophrenia. Psychiatry Research Neuroimaging 1994;55:51±61.

Goodwin FS, Jamison KR. Manic-depressive illness. New York:

Oxford University Press, 1990.

Green MF. What are the functional consequences of neurocognitive

de®cits in schizophrenia? American Journal of Psychiatry

1996;153:321±30.

Harrow M, Goldberg JF, Grossman LS, Meltzer HY. Outcome in

manic disorders:a naturalistic follow-up study. Archives of General

Psychiatry 1990;47:665±71.

Hauser P, Dauphinais ID, Berrettini W, DeLisi LE, Gelernter J, Post

RM. Corpus callosum dimensions measured by magnetic resonance

imaging in bipolar a�ective disorder and schizophrenia. Biological

Psychiatry 1989;26:659±68.

Heaton RK. Wisconsin Card Sorting Test Manual. Odessa: Psycho-

logical Assessment Resources, 1981.

Ho� AL, Shukla S, Aronson T, Cook B, Ollo C, Baruch S, et al.

Failure to di�erentiate bipolar disorder from schizophrenia on

measures of neuropsychological function. Schizophrenia Research

1990;3:253±60.

Jaeger J, Berns S. Neuropsychological management, treatment and

rehabilitation of psychiatric patients. In: Calev A, editor. Assessment

of neuropsychological functions in psychiatric disorders. Washing-

ton, DC, London, England: American Psychiatric Press, 1999:447±80.

Jeste DV, Heaton SC, Paulsen JS, Ercoli L, Harris J, Heaton RK.

Clinical and neuropsychological comparison of psychotic depres-

sion with nonpsychotic depression and schizophrenia. American

Journal of Psychiatry 1996;153:490±6.

Jo�e RT, MacDonald C, Kutcher SP. Lack of di�erential cognitive

e�ects of lithium and carbamazepine in bipolar a�ective disorder.

Journal of Clinical Psychopharmacology 1988;8:425±8428.

Johnstone EC, Owens DGC, Frith CD, Calvert LM. Institutionaliza-

tion and the outcome of functional psychoses. British Journal of

Psychiatry 1985;146:36±44.

Jurjus GJ, Nasrallah HA, Brogan M, Olson SC. Developmental brain

anomalies in schizophrenia and bipolar disorder: a controlled MRI

study. Journal of Neuropsychiatry and Clinical Neuroscience

1993;5:375±8.

Keefe RSE. The contribution of neuropsychology to psychiatry.

American Journal of Psychiatry 1995;152:6±15.

Kessing LV. Cognitive impairment in the euthymic phase of a�ective

disorder. Psychological Medicine 1998;28:1027±38.

Kessler KA, Waletzky JP. Clinical use of the antipsychotics. American

Journal of Psychiatry 1981;138:202±9.

Lezak MD. Neuropsychological assessment. 3th ed. Oxford University

Press: 1995, 1995.

McGrath J, Scheldt S, Welham J, Clair A. Performance on tests sen-

sitive to impaired executive ability in schizophrenia, mania and well

controls: acute and subacute phases. Schizophrenia Research

1997;26:127±37.

Milner B. E�ects of di�erent brain lesions on card sorting. Archives of

Neurology 1963;9:90±100.

Morice R. Cognitive in¯exibility and pre-frontal dysfunction in schi-

zophrenia and mania. British Journal of Psychiatry 1990;157:50±4.

Nasrallah HA, Co�man JA, Olson SC. Structural brain-imaging ®nd-

ings in a�ective disorders: an overview. Journal of Neuropsychiatry

and Clinical Neuroscience 1989;1:21±6.

Norusis MJ. SPSS for Windows:base system user's guide. Release 5.0.

Chicago (IL): SPSS Inc. 1992.

Old®eld RC. The assessment and analysis of handedness: the Edin-

burgh Inventory. Neuropsychologia 1971;9:97±113.

Pearlson GD, Garbacz DJ, Tompkins RH, Ahn HS, Gutterman DF,

Vero� AE, et al. Clinical correlates of lateral ventricular enlarge-

ment in bipolar a�ective disorder. American Journal of Psychiatry

1984;141:253±6.

338 A. Rossi et al. / Journal of Psychiatric Research 34 (2000) 333±339

Robinson AL, Heaton RK, Lehman RAW, Stilson DW. The utility of

the Wisconsin Card Sorting Test in detecting and localizing frontal

lobe lesions Journal of Consulting Clinical Psychology 1980;48:605±

14.

Rossel SL, David AS. The factor structure of the WCST: di�erence

between schizophrenic patients and controls. Schizophrenia

Research 1998;29(1):29±44.

Rossi A, Stratta P, Di Michele V, Gallucci M, Splendiani A, De

Cataldo S, et al. Temporal lobe structure by magnetic resonance in

bipolar a�ective disorder and schizophrenia. Journal of A�ective

Disorders 1991;21:19±22.

Sapin LR, Berrettini WH, Nurnberger JI, Rothblta LA. Medicational

factors underlying cognitive changes and laterality in a�ective ill-

ness. Biological Psychiatry 1987;22:979±86.

Savard RJ, Rey AC, Post RM. Halstead-Reitan category test in bipo-

lar and unipolar a�ective disorders. Relationship to age and phase

of illness. Journal of Nervous and Mental Disease 1980;168:297±

304.

Schneider W. Enhancing a standard experimental delivery system

(MEL) for advanced psychological experimentation. Behavior

Research Methods, Instruments, & Computers, 1989.

Servan-Schreiber D, Cohen JD, Steingard S. Schizophrenic de®cits in

the processing of context. Archives of General Psychiatry

1996;53:1105±12.

Spitzer RL, Williams JBW, Gibbon M, First MB. Structured clinical

interview for DSM-III R (SCID). New York: New York State Psy-

chiatric Institute, Biometrics Research, 1987.

Ste�ens DC, Krishnan KR. Structural neuroimaging and mood dis-

orders: recent ®ndings, implications for classi®cation, and future

directions. Biological Psychiatry 1998;15:705±12.

Strakowski SM, DelBello MP, Sax KW, Zimmerman ME, Shear PK,

Hawkins JM. Brain magnetic resonance imaging of structural

abnormalities in bipolar disorder. Archives of General Psychiatry

1999;56:254±60.

Stratta P, Daneluzzo E, Mattei P, Rossi A. Phasic asymmetries in

phasic a�ective disorders. Biological Psychiatry 1995;38:131±6.

Stratta P, Daneluzzo E, Bustini M, Casacchia M, Rossi A. Schizo-

phrenic de®cit in the processing of context. Archives of General

Psychiatry 1998;55:186±7.

Sullivan EV, Mathalon DH, Zipursky RB, Kersteen-Tucker Z, Knight

RT, Pfe�erbaum A. Factors of Wisconsin card sorting test as mea-

sures of frontal-lobe function in schizophrenia and in chronic alco-

holism. Psychiatry Research 1993;46:175±99.

Tham A, Engelbrektson K, Mathe AA, Johnson L, Olsson E, Aberg-

Wistedt A. Impaired neuropsychological performance in euthymic

patients with recurring mood disorders. Journal of Clinical Psy-

chiatry 1997;58:6±29.

van Gorp WG, Altshuler L, Theberge DC, Wilkins J, Dixon W. Cog-

nitive impairment in euthymic bipolar patients with and without

prior alcohol dependence. A preliminary study. Archives of General

Psychiatry 1998;55:41±6.

Videbech P. MRI ®ndings in patients with a�ective disorder: a meta-

analysis. Acta Psychiatrica Scandinavica 1997;96:157±68.

Waddington JL, Brown K, O'Neul J, Mc Keon P, Kinsella A. Cogni-

tive impairment, clinical course and treatment history in out-

patients with bipolar a�ective disorder: relationship to tardive dys-

kinesia. Psychological Medicine 1989;19:897±902.

Weinberger DR, Berman KF, Zec RF. Physiologic dysfunction of dor-

solateral prefrontal cortex in schizophrenia: I. Regional cerebral

blood ¯ow evidence . Archives of General Psychiatry 1986;43:114±

25.

Weinberger DR. A connectionist approach to the prefrontal cortex.

Journal of Neuropsychiatry 1993;5:241±53.

Winokur G, Clayton PJ, Reich TJ. Manic depressive illness. St Louis:

Mosby, 1969.

Zihl J, Gron G, Brunnauer A. Cognitive de®cits in schizophrenia and

a�ective disorders: evidence for a ®nal common pathway disorder.

Acta Psichiatrica Scandinavica 1998;97:351±7.

A. Rossi et al. / Journal of Psychiatric Research 34 (2000) 333±339 339