Mood Stabilizers in Children and Adolescents

NEAL D. RYAN. M.D., VINOD S. BHATARA. M.D., AND JAMES M. PEREL. PH.D.

ABSTRACT

Objective: The efficacy of mood stabilizers in children and adolescents has not been studied adequately. This article will

review existing studies and highlight some important issues in designing future studies on these agents. Method:

Electronic databases including Medline. Psych/nfo. and CRISP were searched for data in children receiving compounds

that have mood-stabilizing properties In adults. Reeults: Some open clinical data and an extremely modest amount of

controlled research dat8 suggestlilhium. carbamazepine, and valproate may be effective mood stabilizers in children and

adolescents. There are no controlled data on other potential mood stabilizers in children. Conclusions: The disorders

that may be responsive to mood stabilizers are among the most morbid in child psychietry. More studies are needed to

clarify the efficacy ollhese compounds in children and adolescents and to provide a rational basis lor choosing among

them. J. Am. Acad. Child Ado/esc. Psychiatry. 1999, 38(5):529-536. Key Worda: psychopharmacology. bipolar disorder.

mood stabilizers.

This review examines what is known about the efficacyof mood stabilizers in children and adolescents. Data onthe use of these compounds in children address theiruse in treating bipolar disorder. conduct disorder andattention-deficit hyperactivity disorder. and violentaggression, A number of these compounds ate anricon­vulsants. but that use is not considered here.

Adolescents suffer from bipolar disorder with a clin­ical picture much like that seen in adults, making diag­nosis relatively straightforward. The course of adolescentbipolar disorder is also similar to that seen with adultbipolar disorder (Carlson et al.. 1977; WeIner et al..1977). The diagnosis of bipolar disorder in prepubertalchildren has received much auention recently and pre­sents a more difficult diagnostic problem (Bowring andKovacs, 1992). Bipolar depression in prepubertal chil­dren is frequently comorbid with externalizing disordersand more often presents in a rapid-cycling or "mixed"picture (Kovacs and Pollock, 1995; Milberger et aI.,1995; Wozniak et aI.• 1995). In the absence of definitivestudies in youth, physicians often base their pharmaco­logical treatmclH dccisions on extrapolation of scielHific

Aee-ptrd N",,,,,,btr II, /998.Dn. Ryall o"d I'rrri urr ,viI" Ih, D'porrmml of PryehitllrJ. U"iI"niry of

Pillsbllrgh. Dr. HI,OIOTII i, Co-Dirrero, of Chiltla"tI Adoln",,, I'~y,.f,jllrry.

U"ivmiry ofSOlllh Doltnro M,di"IIISehool, SiorlX Foils.C'ormpo"dm" 10 Dr. RYIIII. 38/1 O'lUrra Srrrrr, Pimbllrgh. 1\01 15213.0890-85(,7/9'.1138115-052911' 199') b)' th~ Am~rican Ac,demy of Child

and Adol'«:<1l! r,y,hi,rry.

J. AM. ACAIl. CHI!.D ADOUSe. PSYCIIIATRY, .1H,~. MAY 1""')

data from studies in adulrs. Even assuming that bipolardisorder is a single disorder or group of disorders through­out the lifespan, biological variabiliry of the organismand age-varying pharmacokinetics and pharmacody­namics limit the usefulness of such extrapolations fromadult studies. For example. it appears that even thoughchild and adolescent unipolar disorder is very likely thesame underlying disorder as adult depression. tricyclicantidepressants may not be as efficacious in children asin adults though selective serotonin rcuptake inhibitorsmay be equally efficacious throughout the lifespan (re­viewed in Birmaher et al.. 1996), Therefore. even ifjuvenile bipolar disorder is similar to the adult form ofthe disorder, we need controlled studies of mood sta­bilizers in youth and cannot merely extrapolate fromadult studies. Because some compounds show sex- andrace-related kinetic variations. these issues require inves­tigation. The developmental factors that playa role inthese variations need to be studied.

LITHIUM

In general. lithium is rdatively well-tolerated in chil­dren. Side effects have been systcmatically reported inchildren as young as 3 years of age (Hagino et al..1995). Lithium is approved by the Food and DrugAdministration (FDA) for treatment of bipolar disorderin adolescents who are 12 years of age or older. but notin prepubertal children. Examinations of the relation­ship between dosage and plasma have been made

529

RYAN 1''1" AI ..

(Malone et al.. 199'5; Vitiello et al.. 1987: Weller et al..1987). The distribution and eliminarion of lithium hasbeen systematically studied and parallels that seen inadults, with some evidence of shorter elimination half·life and higher total clearance in children. Available sideeffects data are from case reports. from small case series.and from systematic reporting of side effects in small,controlled efficacy studies. Common lithium side effectsin children include nausea. diarrhea. tremor. enuresis.fatigue, ataxia (Silva et al.. 1992). leukocyrosis. andmalaise; less commonly seen are renal, ocular. thyroid,neurological. dermatological. and cardiovascular effects.Changes in weight and growth. diabetes. and hair lossare also seen (Rosenberg et al.. 1994). Children youngerthan age 6 may experience neurological effects relativelyfn:quently (Hagino et aI., 1995), and in general youngerchildren seem to experience more side effects rhan doolder children (Campbell er al.. 1991).

'10 datt:o there is only a single published. methodo­logically sound. double-blind. randomized controlledtrial (ReT) of lithium or ocher mood stabilizers inyouth (Geller et aI., 1(98). In one very small crossoverstudy. lithium was superior ro placebo in 2 of 6 cbildrenwho had lithium-responding bipolar parents (McKnewet al.. 1981). In another brief crossover study, lithiumappeared better than placebo in a sample of 11 pre­viously lithium-responding children with manic symp­romatology (DeLong and Nieman. 1983). In an openstudy of to children. lithium alone appeared efficaciousin prepubertal children with psychotic bipolar disorder(Varanka et aI., 1(88). In another, 2 of 6 children withprepubertal mania or hypomania improved with lithium(Brumback and Weinberg. 1977).

In adolescents. open trials of lithium for bipolar dis­order appear to give response rates similar to that seenin adults (Delong and Aldershof, 1987; Varanka et aI.,1988; Youngerman and Canino. 1978). Strober et a1.(1990) conductl:d a naturalistic prospective follow-upstudy of 37 "dolescenrs treated successfully with lithiumfor bipolar disorder during hospitalization. Those whodiscontinued lithium treatmcnr (against advice) weremuch more likely to relapse than those who continuedthe medication.

A recent RCT by Geller tested short-term treatmentwith lithium versus placebo in adolescents with sub­stance dependency disordets and concomitant bipolardisorder. This study was designed as a 2-week. single·blind, placebo washout phase followed by a IO-week.

530

placebo-controlled. double-blind. short-term treatmentstudy. Twenty-five subjects were enrolled. Those ran­domly assigned to receive lithium showed significantlybetter outcome of both their bipolar disorder and theirsecondary drug dependency than those randomly assignedto placebo (Geller et a1.. 1998).

There are 2 NIMH-funded ongoing controlledstudies of mood stabilizers in adolescents:1. "Lithium in Hospitalized Bipolar Manic Adoles­

cents" (Vivian Kafantaris, principal investigator) is astudy of lithium treatment in the acute manic phaseof adolescent bipolar disorder. Enrollment is on­going. and there have been no examinations of thecontrolled portion of the data from this study.

2. "Lithium Prophylaxis in Adolescents With BipolarIllness" (Marrin Keller. Michael Strober, and NealRyan. principal investigators) is a multisite. placebo­controlled study of lithium discontinuation after 6months of medication stabilization. This study isongoing, and interim analyses have not been per­formed. Because of slower than anticipated recruit·ment. the design has been changed to that of adiscontinuation study from whatever mood stabi­lizer or combination of mood stabilizers on whichthe adolescent has been stabilized.

The strategy of lithium augmentation of tricyclicantidepressants is well-established in adult (nonbipolar)major depression (De Montigny et a1.. 1981). Thisstrategy has not yet been tested in youth by ReI: In 2uncontrolled studies, approximately 40% of youthinadequately tesponding to tricyelics showed a favorableresponse to the combination of tricyclics and lithium(Ryan et aI., 1988; Strober et aI., 1992).

Open clinical studies have suggested that lithium maybe useful in the treatment of aggression in children,especially when seen in those with mental retardation orother neurological disorders (Delong and Aldershof,1987), antisocial personality (Schiff et al.. 1982), andextreme aggressiveness (Vetro et aI., 1985). However.one study was less promising when lithium was used foraggression with hyperactivity (Greenhill et al.. 1973).

The 4 double-blind, placebo-controlled studies thathave investigated antiaggressive effect of lithium in chil.dren and adolescents with conduct disorder havereported mixed results. Campbell and associates (1984,1995a) found their hospitalized prepubertal subjects torespond better to lithium than placebo in 2 studies.Conversely, Rifkin and colleagues (1997) reported a

J. AM. ACAIl. CHII.Il AIlOI.ESLI'SVCHIATRV. .IH:~. MAV I')')')

negative response in hospitalized adolescents. This studyis limited by the very short duration (2 weeks) of thetrial. Silva reported statistically negative results in a verysmall study of outpatients with aggressive conduct dis­order (patients receiving lithium appeared clinicallyimproved. but the difference did not approach signif­icance. perhaps because of small sample size) (reviewedby Campbell et al.. 1995b; Silva et al.. 1991).

CARBAMAZEPINE

Carbamazepine (CBZ) is widely used in the treat­ment of seizures in children, and for this reason irskinetics have been well-studied (Camfield et al.. 1992;Cornaggia et aI., 1993; Eeg-Olofsson et al.. 1990; Liuand Delgado. 1994: Suzuki et aI., 1991; Thakker et aI.,1992; Yukawa and Aoyama, 1996). From the neurolog­ical literature, there are also relatively extensive data onside effects in children and adolescents. The mostcommonly seen side effects with this agent in childrenarc drowsiness. loss of coordination. and vertigo. Moreserious side effects reported to the manufacturer overan II-year period during which 4 million patients weretreated included hematological, dermatological. hepatic.and pancreatic effects; 27 cases of aplastic anemia; and10 cases of agranulocytosis (Pellock. 1987). While CBZhas been used to treat a variety of psychiatric disordersin children and adolescents, it is not labeled by the FDAfor psychiatric indications in any age group.

CBZ has demonstrated efficacy in adult bipolar dis­order (Post et al.. 1996; Stuppaeck et al.. 1990). and thecombination of lithium and CBZ may be superior to

lithium therapy alone (Solomon et al.. 1996). In adults.CBZ may be superior to lithium alone in "mixed" orrapid-cycling mania (Calabrese et aI., 1996a); however,some examinations do not support this (Okuma, 1993).There are no controlled studies and few anecdotal dataon CBZ in children. even though it is certainly used asan adjunct to lithium when lithium treatment alone isineffective in treating childhood bipolar disorder.

As recently reviewed by Silva et al. (1996), there are29 reports in the world literature examining CBZ in thetreatment of behavioral problems or high activity levelsin children. Of these, 3 are double-blind, controlledstudies (Garcia Belmonte and Pugliese, 1970; Grohet al.. 1971; Puente et aI., 1973), all of which appearedin the early 19705. with a total of 53 patients receivingCBZ and 52 receiving placebo (2 of the studies hadcrossover design). The overall response percentage in

J, AM. ACAD. CHILD AIHlLESC. l'SYCIlIATIlY, ,IH:~, MAl' I'P!')

MOOD STARll.IZF.RS

these 3 studies was 71 % to the CBZ and 26% to placebo.Diagnostic schemas have changed since these studieswere completed. and a majority of subjects in these 3trials had "abnormal EEGs."

Despite an earlier promising open pilot study by thesame group (Kafantaris et al.. 1992). in a 6-week, dou­ble-blind srudy of 22 children aged 5 to 12 with con­duct disorder who were hospitalized for aggression.CBZ in doses from 400 to 800 mg with serum levelsfrom 5.0 to 9.1 ~g/mL was not superior to placebo inreducing aggressive behavior (Cueva et al .. 1996).

Side effects may be relatively common with CBZ, incomparison with lithium or placebo (Cueva et aI.,1996), and there are several reports of adverse cognitiveand behavioral effects with this compound in children(Bhatara and Carrera, 1994; Pleak et a1.. 1988).

VALPROATE

Because of the widespread use of valproate in thetreatment of seizures in children, its kinetics in mono­therapy, when combined with ochcr anticonvulsants,and in sustained-release forms is well-studied (Battinoet aI., 1995a; Botha et al.. 1995: Brouwer et aI., 1992;Cloyd et aI., 1993; Kriel et aI., 1986; Sugimoto et 01.1..1996; Yukawa, 1995; Zaccara et 01.1.. 1988).

The common side effects of valproate include seda­tion, nausea, vomiting. appetite/weight gain, tremor,hepatic toxicity, hyperammonemia. blood dyscrasias,alopecia, decreased serum carnitine, neural rube defects,pancreatitis, hyperglycemia. and menstrual changes(Rosenberg et al.. 1994). The hepatic roxicity, whichmay lcad to death, appears to occur almost exclusivelyin relatively young children. especially those youngerthan 2 years (Bryant and Dreifuss. J996; Silberstein andWilmore, 1996).

A specific concern has recently been raised thar val­proate may induce a metabolic syndrome, characterizedby obesity, hyperinsulinemia, lipid abnormalities, pol­ycystic ovaries, and hyperandrogenism, particularly inyounger women. In a cohon of Finnish women takingvaJproate for seizures, 80 of rhe women who srarredtaking valproare before the age of 20 years had pol­ycystic ovaries compared with 43% of all women takingvalproate (Isojarvi et aI., 1993). On replacing valproatewirh lamotrigine in 16 women, Isojarvi et al. (1998)found rhe severity of this metabolic syndrome to be re­duced (suggesting a partial reversibility). The generaliz­ability of their findings to psychiatric populations is

531

RYAN ET AI..

unknown because the reports of this syndrome. so far,arc confined to this single cohort with epilepsy.

Divalproex sodium and valproic acid are now fre­quendy used alone or in combination with lithium inthe treatment of adult bipolar disorder (Bowden et al.,1994; Post et al., 1996; Solomon et al., 1997). Valproatemay be more effective than lithium in adult subjectswho have mixed mania (Bowden, 1995; Calabrese et aI.,1996a; Swann et al., 1997). Bipolar adults with seizuresor other neurological conditions may also preferentiallyrespond to valproate (Stoll et aI., 1994).

Several single case reports and small open seriessuggest efficacy for valproate as a mood stabilizer inadolescents (Whittier et al., 1995). In 3 open studies,addition of valproate to previously ineffective psycho­tropic treatments in hospitalized adolescents resulted insymptomatic improvement (Papatheodorou and Kutcher,1993; West et aI., 1994, 1995). Strober (1997) examinedthe clinical course of the mixed manic state treated withvalproate compared with a historical control grouptreated with lithium who were otherwise similar. Thatstudy showed superiority of valproate over lithium forthe mixed form of mania but no difference in efficacybetween the two for classic mania.

In adults, recent controlled data suggest a relationshipbetween trough levels of valproate and clinical response,and the same data suggest that a rapid dose escalationprotocol may lead ro earlier symptom improvement(Bowden et aI., 1996; Kcck et aI., 1993). There is anopen study examining a rapid dose-loading strategy inadolescents (West et al., 1995) suggesting that responsemay be somewhat slower in adolescents than adultS andthat serum levels in the morning after the loading dosemay be relatively lower in adolescents.

In an open naturalistic examination of divalproexsodium in 10 adolescents with chronic temper outburstsand mood lability, there appeared to be improvement inall 10 subjects and discontinuation of medication ap­peared associated with relapse with subsequent improve­ment after restarting medication in 5 of 6 subjects(Donovan er al., 1997).

NOVEL TREATMENTS

Other novel anticonvulsanrs have been suggested to

have mood-stabilizing uses in adult bipolar disorderincluding lamorrigine (Calabtese et aI., 1996b; Waldenet al .. 1996) and gabapentin (Schaffer and Schaffer,

532

1997; Stanton et a!" 1997). For these agents the pediatricdata on kinetics are modest (Battino et aI., 1995b; Elwesand Binnie, 1996). There are no data on their potentialefficacy as mood stabilizers in children or adolescents.

Calcium antagonists including verapamil (Deicken,1990; Giannini and Loiselle, 1996; Hoschl et al., 1992;Lenzi et aI.. 1995; Ostow, 1987) and nimodipine(Goodnick, 1995; Grunze et al., 1996; Pazzaglia et aI.,1993) have possible efficacy as mood stabilizers. There islittle information on the kinetics of these compounds inchildren (Wagner et aI., 1982). There are few data ontheir potential efficacy as mood stabilizers in children oradolescents.

Neuroleptics may be used as an adjunct to mood sta­bilizers in the short-term treatment of bipolar disorderin children and adolescents. One case series suggeststhat c10zapine may be helpful when other neurolepticshave failed in this situation (Kowatch et al., 1995).There are no controlled studies systematically assessingneuroleptics and their use in treatment algorithms forchild bipolar disorder.

DISCUSSION

Overall, there are a modest number of open clinicaltrials and case reports examining mood stabilizers inchildren and adolescents but extremely few RCTs. Todate, the completed Refs with even marginally adequatesample size include only:• A positive srudy of lithium versus placebo in substance­

abusing bipolar adolescems (Geller et al., 1998).• A positive study of lithium versus haloperidol versus

placebo in treating aggression in undersocializedaggressive conduct disorder (Campbell et al., 1984),which was later replicated in a separate sample by thesame group (Campbell et al., 1995a) and which con­trasts with one negative study (Rifkin et al.. 1997).

• Three positive double-blind, controlled studies fromthe early 1970s of CBZ versus placebo for behavioralproblems and high activity levels (Garcia Belmonte andPugliese, 1970; Groh et aI., 1971; Puente et aI., 1973),which are limited by diagnostic schema and inclusionof large number of subjects with "abnormal EEGs."

• A negative 6-week, double-blind study of CBZ versusplacebo in hospitalized aggressive children with con­duct disorder (Cueva et al., 1996).Recruitment of an adequate sample size has been,

perhaps, the most consistent and single most problem-

J. AM. ACAD. CIIII.D ArIOI.f.SC. l'SYCHIATRY. JH:~. MAY 19')9

atic issue in studies of mood stabilizers in children andadolescents. The total pool of potential patients is muchsmaller than for comparable adult studies. Whilebipolar disorder most often has onset during adoles­cence or early adult life. fewer than half of all subjectswill have onset during adolescence, so many people whowill be candidates for studies of bipolar disorder inadulthood will not have yet had a first episode. A roughcalculation suggests that for adolescent studies, onemight have (40 years of adulthood/4 years of adoles­cence) X (100% expressed/50% of cases expressed byadolescence) = 20 times more subjects from which to

sample for an adult bipolar study than for an adolescentbipolar study. Even meaningful changes in the assump­tions (e.g., factoring in changes in reproductive rates,secular increases, etc.) still result in a 5- to 20-foldgreater number of subjects in the pool for adult studiesof bipolar disorder compared with the pool for child oradolescent studies. Other factors complicating child andadolescent studies include higher rates of refusal for thestudy (because to participate requires consent/assentfrom both parent[sJ and child. so more people can vetoparticipation) and potentially higher rates of non­compliance afrer entry. In addition. some managed carehealth plans do not permit adolescent participation instudies that may result in assignment to a placebo-onlymedication arm.

Open naturalistic studies demonstrating feasibilityand tolerability and suggesting that mood stabilizersmay be effective in children and adolescents for bipolardisorder and for other disorders including aggressionwith conduct disorder/attention-deficit hyperactivitydisorder are a necessary first step. While more open datawould always be helpful, such open studies do exist andhave been published with sufficient numbers of chil­dren treated to pave the way for controlled studies ofthese compounds. The next stage of controlled studiesis likely to require multisite studies for sufficient samplesize; funding by industry. the National Institutes ofHealth. private foundations, or a partnership of federaland corporate or private funding sources; working withmental health advocacy groups to change healthprovider rules and to facilitate recruitment; andguidance by regulators about appropriate study design.

Side effect data are also necessaty for the clinician tomake correct judgments on how to use these com­pounds. Important questions about mood stabilizersthat could be answered by systematic side effect data on

J. AM. ACAD. CHII.D ADOI.f.SC. PSYCHIATRY..1H:~, MAY I???

MOOD STABIl.lZERS

samples of 100 or fewer children include the cognitivesequelae of mood stabilizers in children and replicationand extension of the data on polycystic ovaries with val­proate. Available studies suggest that. in general, val­proate and CBZ are well-tolerated by the pediatricpopulation with seizure disorder (reviewed by AmericanAcademy of Pediatrics Committee on Drugs, 1995),and cognitive effects that interfere with learning atschool are rare (Herranz et al.. 1988). However, efficacystudies will not provide sufficient data to answer manyside effect questions for the following reasons: First,some side effects of great clinical importance may berare enough never to be encountered in any realisticallysized clinical study (e.g., the quesrion of cardiovasculardeath from desipramine). Second, one cannot ade­quately extrapolate from relatively frequent but clinicallyunimportant changes in an organ system to the in­frequent but important changes. Similarly, one cannotextrapolate from the rate of side effects seen with highor toxic doses to the rate seen with therapeutic doses.Very large surveys are needed (0 address importantquestions such as. What are the rates of hematologicalor hepatic adverse effects in older children receivingmood stabilizers? Therefore, in addition to efficacystudies, systematic. large-scale. economical systems tocollect adverse event data in children are important.

In summary. we believe that the following are ofgreatest importance in the study of mood stabilizers inchildren and adolescents:• There are no systematic studies on the efficacy of any

of the mood stabilizers for prepubertal bipolar dis­order. This is a priority area givcn the morbidity andchronicity of this disorder.

• Studies of mood stabilizers for aggression and con­duct disorder are few, are small, and have used heter­ogeneous inclusion/exclusion criteria, making itdifficult to know for which populations this approachis useful. Because such disorders are relatively intrac­table, this area deserves more study.

• The available RCT data and a considerable amountof open clinical data suggest that adolescent bipolardisorder probably responds (0 the same agents asadult bipolar disorder. However, the RCT data to

support this conclusion, as of yet, consist of a singlereported study with lithium in substance-abusingbipolar adolescents. Until there are more consistentdata. this question cannot be considered setded. Inaddition, comparative studies examining the efficacy

533

RYAl'.' ET AI..

of these agents, including time to response, have notbeen undertaken. Given data that valproate may havea quicker onset of action than lirhium in bipolaradults (Bowden et aI., 1994) and that it can be givenin a rapid loading maregy (Bowden er a!., 1996; Kecket aI., 1993), comparison of active treatments inadolescent bipolar disorder might permit morerational treatment strategies.

• Many adolescents and children with bipolar disorderdo not respond to any of the first-line pharmaco­logical treatments. Therefore, studies with novelagents should be extended to this population. Inaddition, physicians will continue to use combinationtherapies in the face of either lack of efficacy ordelayed onset of efficacy of single agents. Therefore,the resultant drug-drug interactions also deservesystematic study.

• Systematic assessment of frequent and infrequent sideeffects of these compounds in children with psychi­atric disorders is needed. Existent data in the neuro­logical literature do not completely address side effectsthat may be more frequent in psychiatric populations.For example, CBZ, a tricyclic agent, may inducemania in susceptible children (Bharara and Carrera,1994; Myers and Carrera, 1989; Pleak et aI., 1988;Reiss and O'Donnell, 1984).

• Adequate kinetic data in children and adolescents areneeded on the newer agents to rationally design phase11 and phase 1Il clinical trials in this population. Kin­etic data need to examine the effects of age, race, sex,hormonal milieu, and nutritional statUS on metabolism.

• Objective assessments of compliance are particularlyimportant with bipolar and aggressive children andshould be included in the design of future Ref studies.

• Development is nor a 3- or 4-step process but ratheran interplay of a number of developmental processesproceeding at different rates and interacting witheach other. Many aspects of development can a prioribe expected to influence kinetics and dynamics ofthese compounds including the sex steroid milieu.hepatic and renal development, changes in bindingproteins, changes in distribution compartments, andbrain maturation. Better attention to the richness ofthe development process is needed.

REFERENCESAmeric~n Academy of Pedi;urics Commincc on Drugs (1995). Schavior:ll

and cngnirivc clT<CIS n( anliconvul..nt lher:lpy. Prdi""i" 96:538-540

534

Bmino D. E.tienne M. Avannni G (1995a), Clinical pharmacokinetics o(;mliepileplic drugs in .,..edimic palienlS. pan I: phenobarbilal. primidone,valproic acid. Clhnsuximide and mesuximide. Clin Pharm,,(olrintt29:257-286

Banino D. ESlienne M, Avanlini G (l995b), Clinical pharmacokinetics ofanlicpileplic drugs in paediatric paricn". pan II: phenyroin, carba­mazepine. suhhiame. lamotriginc. vigab,"in. oxcalbaupinc andfclbamalc. Cli" Pharma(olri"rI29:341-369

Bh.,ara VS. Carrcra J (1994). Medic.,ions (or aggressiven.... ) Am A(adChild Adolnc Prychiatry 33:281-283

Birmaber B. Ry~n ND. William.on DE, Brenl DA. Kau(man J (1996).Childhood and adolescenr deprcssion: • review of lhe pa,' 10 years. PallII.J An, Acad Child Adoln( !'ry(hi4try 35: 1575-1583

BOlh.JH. Gray AL. Miller R (1995). A modcl for cSlim.ting individua.lizcdvalpro." c1..",nce v.lues in children.} Clin Pharma(oI35:1020-1024

Bowden CL (1995). Predictors of response to divalprocx and Iilhium.} GinPs""hidlry 56(suppI3):25-30

Bowden CL, Bruggcr AM. Sw.nn AC ct ~I. (1994). Efficacy of div~lproexvsIilhium .nd placebo in Ihe lre~rmenl of mania: lhc Oepakore ManiaSrudy Group. JAMA 271 :918-924

Bowden CL. Janicak PG. OrsuJak P Cl al. (19%). Relation of serum val­proal< conccnnalion 10 r"ponse in mania. Am} Prychiatry 153:765-770

Bowring MA. Kovaes M (1992). Difficulties in diagnosing manic disordcrsamong children .nd adolcsccnlS. J Am A"ul Child Adolrsr Psychiatry31:611-614

Brouwer OF. Pierc!! MS. Eddbrock PM Cl .1. (1992). Convenlional andconlrolled rclca... valproare in childrcn Wilh epilepsy: • croSS-<Jvcr Sludy[ompOling plasma levds and cognirive perform~nces. Epikpry Rn 13:245-253

Brumback RA. Weinberg WA (19771. Mania in childhood. II: rherapeuticnial of lithium carhonarc and furrher dcscriplion of manic-dcpressive ill­ness in childrcn. AmJ Dis Child 131: 1122-1126

Bryant AE III, Drcifuss FE (1996). Va.lproic acid heparic fala.litics. III: US..periencc sincc 1986. Nrur.1bgy 46:465-469

Calabrese JR, Fatemi SH. Kujawa M. Woyshvilk MJ (19960), Prediclors ofn:sponse to mood Slabilizcrs. J Clin P'yehopharmJZco/16:24S-3IS

Calahrese JR. !':nem; SH. Woyshvillc MJ (199Gb). Anlidcprcssant elfcds o(I.morriginc in rapid <)'Cling hipol.. disorder (lclltl). Am J Ptyrhialry153:1136

Camncld P, Hwang Po Camncld C. Fraser A. ~oldin S. a!·Qu.d.h AK (1992),Thc pharmacology of chew.ble vcrsw regu'., carbamazcpinc in chron­ic~lIy l~alcd childrcn wilh epilepsy. CAn} Nrurol Sri 19:104-207

C.mpbcll M. Adams PB, Small AM et .1. (1995.). Lirhium in hospiralizcdaggrc.<sivc childrcn with conducl disorder: a douhle-blind and placcho­conrmllcd study.} Am A(ad Child Ad.lts< l'tyrhialry 34:445-453 [pub­lishcd crrarum appears in} Am A(ad Child AJoksr Psychi4try 34:694.19951

C.mpbell M. K~(.nlaris V, Cueva JE (1995b), An update on thc use o(lilhium carbonare in aggressive children and adolcsccnlS Wilh conductdisOldcr. Prychopharma(oI8uIl31:93-102

Campbell M, Silva RR, Ka(anl.ri. V Cl al. (1991), Prcdicrors of sidc cITeersassocialed with lithium administration in children. Plychophaffl"".1 Bull27:.~73-380

Campbell M. Small AM. Grecn WH Cl al. (1984). Bchavioral efficacy o(haloperidol and lirhium carbonare: ~ comparison in hospitalizcd aggrcs·si.., children with conduct disorder. Arch Gm Psychiatry 41 :650-656

Carlson GA. Davmporr YD. Jamison K (1977). A comparison of ou[Comc in.dolestcnl- and lattl·onset bipolar manic-depressive illness. Am J"'yehiatry 134:919-922

Cloyd Je. Fischcr JH. Krid Rl. Kraw OM (1993). Valproic acid pharmaco­kinetics in childrcn, IV: cITc", of agc and anticpilepric drugs on prorcinbinding and inninsic dcarancc. C1in Pharmacol Thtr 53:22-19

Comaggia C. Gi~nclli S. Ballino D et .1. (1993), Comparali"" pharmaco­kinetic stUdy o( chcwable and convcnlional carbamazcpinc in childrcn.Epilrpsia34:158-160

Cucva JE, Overall JE. Small AM, Armcnrcros JL. Pcrry R. Campbcll M(1996). Carhamaz.cpinc in aggressivc children Wilh conduct disoldcr: adoublc-blind and placcbo.cnnrrolled ..udy. J Am A(ad Child Adam(Ps.ychiatry 35:480-490

J. AM. ACAD. CHIl.D ADOl.F.SC. PSYCHIATRY..18:5. MAY 1999

Dc MonliGny C, Grunborg F. M.yer A, Deschenn JP (1981), Li,hiuminduces rapid relief of depression in lricyclic antidcpressaOf drug non­responders. 8r} l'Iy<hia"J UK:252-256

Deicken RF (1990). Verapamil lreatmeOf of bipolar depression. } Cli"l'IychopharmaeoIIO: 14M-149

Del.ong GR, Aldershof AI. (1987), Long-term experienco wilh lithium lr..al­l1IellC in childhood: m"da[ion wilb clinical di.sno,is,} Am Amd Chi'"Adolnc Psychialry 26:.i89-394

Del.onG GR: Nien;.n GW (1983). Ulhium-induced beh.vior ch.nges inchildren with 'ymp",ms sn&gesting manic-depressive illness. J>~yrho­

pharmacoIB/l1I19:2S8-265Donovan SJ, Susser ES. Nun.... EV, Srcwm JW, Quirkin FM, Klein OF

(1997), Diy-alproex rr...:ltmeR! of disruplive adolc<cents: a repUll of 10cases.} A", /'Iychialry 58:12-15

Ecg-Olufssnn 0, Nil",,~ HI.. loonby B el al. (1990). Diurnal v.ria,ion orcarbm.zepine .nd carbama7.cpinc-10.II-epoxide in plasma and ..liv. inchildren wirh epilepsy: a comparison betwccn conventional .nd slow­release formulations.} CMU N",roI5:\';9-165

Elwcs RD, Binnie CD (\'.1'.16), Clinical pharmacokinetics of ncwer anli­epileplic druGS: lamorriGine, vigabalrin. gaha\'<'nrin and uxcarbazcpine.Cli" Pharmacokinn 30:4ll.\-415

Garda Belmonle MA. Pugli....e AE (1970). Tra,amienlo del syodrome dehiperquinesia infantil coocarbamacepina. EI Di" M,dico 41 :75'J-7('()

Geller B. Cooper TB, Sun K er al. (1998). Double-blind and placebu con­Irolled slUdy of lithium fnr adnleseent bipolar disorders with secondarysubstance dependency. } Am AC4d O,ildAdolrIC l'rychia"J 37: 171-17K

Gi.nnini AJ, Luiselle RH (1'.1')6). Verapamil mainrenance rherapy in bipolarp.lients (leller).'/ Oi" l'~ycbia", ';7:U6

Goodnick I'J (19'.1';). Nimodipine tr,,.tn'eOf of rapid cyclinG bipol., dis­order (leller). J C/i" l's'«hiill'1 56:.BO

Greenhill 1.1. Rieder RO: Wender PH, Buchsl>aum M. Zhan "1' (197.\).Uthium c.trhonall' in the rrc,llmcnt of hypera(tive children. Arrh G~"

l'syciJialry 28:636-640Groh C. Rllscmayr F, Birhaumer N (1971). l'sychnrrope wirkung von carh.­

m,,-<pin bd nighr-epileprischen kindero. M,d MO/laweiJr l';:.\2'.1-3.HGrunz< H, Walden J. Wolf R, Berger M (\9%). Coml,ined Ire;llmcnl wilh

lilhium anJ nimodipine in a bipolar I m.nic syndrome. !'rog Nruro­pryciJopharnldml Hioll'~ychialry 20:419-426

Hagino OR. Weller EB, Wener RA. Washing D, Frisrad MA. KonrrolS SB(\995), Unlaw.rd eflec" of lirhium marmenl in children agetl lourIhrough six years.] Am Acari Child Adolrsr I'IycJ,ialr/.H: 15114-1';90

Herranz JL. Armijo JA, Ancga R (1988), Clinical side dIem or ph...no­harbiral. primidnnc. phenytoin. carhama'lcpinc. and valpro;u.., durinl;monolherapy wi,h children. Epil'pIia 29:794 -804

Hoschl C. Vackova J. Jand. 1\ (19<)2). Mood slabilizing etTl'Cl uf yenl,.mil.BrotiIll.tIll.iJly <).~:208-209

lsojarvi JI. Laarik;inen TJ, Pakarinen AJ. Juntunen KT, MyUyl. W (1993).Polycysric ovaries and hyperamJrogcnism in wumcn laking valpro;l[l' f(lf

epilep'y. N Engl} M,d .U9: 1383-1.188hojarvi JI1'. Rallya J. Myllyla WeI al. (1998), Valpro.I.... lamotrigine. and

insulin-mcdi31cJ risks in wumen wilh epilepsy. Anll N",roI43:4~Cl-4 51Kafanraris V, Campbell M, Padron-Gayol MV, Sm.1I AM, Locascio JJ,

Rosenberg CR (1992). Carbam.zcpine in hospilalizcd aggressive con­ducr disorder children: an open pilnr scudy. PSlchopharmacol 8/11128:19~-\99

Keck PE Jr, McElroy SL, Tugrul KC, Bennell JA (19'.1.~). Valproale 0,,1 load­inG in the lrealment of acule m.nia.} Cli" /'Iycf,iatry 54:305-308

Kovac< M. Pollock M (\995), Bipolar disorder aod comorhid conduce dis­order in childhood and adolescence.} Am Arad Child Adolrsr I'I'/l-hial'1.H:715-7B . .

Kowarch RA, Suppc< T. Gilfillan SK. Fucotes RM, Grannemann 1\0. Em,lieG) (1995). Clol.apine "e;umenl of ehild",n and adolesccolS wilh hipolardisorder and schiznphrenia: a clinical case series. } Child AdolrIrl'syehopharmacoI5:241-25~

Kricl RI.. Fischer JH, Cloyd )C, Gr....,n KH. Fraser tiL (1986). Valpruic acidpharmacokinerie., in children, 111: very high dosage requiremel1l5. !',diatrNwroI2:202-20R

Lenzi A, Mar31.ziti D, R.1ffaelli S. Cassano GB (1995). Efleclivenm uf thecomhin.uion vcrapnmil and chlorprornllZinfo.' in the ueutlnrnt of s~ve~

J. AM. ACAD. CHILL) ..,1l11I.ESC. I'SYCHIATKY..IH:~. MAY I')')')

~OOD STABII.IZERS

manic nr mixed. paricnu. Prog NnlTopJ.yclJophtlrmacol Bioi PIJclJil1fry19:519-528

Liu H. Uclg.do MR (1994). A comprehensive Sludy of rhe relation bc,weenserum concentrations. concentration TJ(ios. and levd/dose ratios oft:arbamazepine and irs metabolites with age. weight. dose. :md clearancesin epileptic children. £pil'!,Iia 35:1221-1229

M.lune RP. Delancy MA, l.uebbcTl JI'. While MA, Biesecker KA, Couper1'8 (1995), The lilhium leSI dose predictiol1 method in aggressive chil­dren. 1'I.'(I·!Jopbarn",rol Bull 3\ :.179-382

MeKnew DH, Cytryn L. Buch,baum MS el al. (1981). Lithium in childrenof lithium-responding parent'. Psychiatry RrI 4:171-180

Milberger S, Biederman J. Far.one SV, Murphy J, Tmang M1' (1995).Au('nrion ddlcil hyrxracriviry disorder :Lnd comorbid disorders: is.liUC'~ ofoverlapping symploms. A",} 1~.yr!Jia"y 152: 179.1-1799

Myers we, Carrera F 111 (1989), CarhamalCpine·induced mania wilhhypccsexu.liry in a '.I-year-old hoy (leller). Am} l'Iychialry 146:400

Okuma T (1993). Effects of cathamazepine al1d lilhium on atrec[ive dis­mders. Nwropslrbobiology 27:138-145

OSlOW M (1987). V...rapamil and hipolar illness (lmer). } Gin Psyrho­phanlul.-oI7:277

l'apathL'Odomu G. Ku[cher SP (1993). Divalpn,,:x sodium trcalmel1l in laICadolescenl and young .dulr aCU'e mani3. l'sycho!,lliIrmllcoI8111129:213-21'.1

pazzaglia I'J. l'oS! RM. Kerter TA. George MS. Marangell I.B (1993).Preliminary cootrolled trial of nimndipine in ulna-rapid cycling .Recrivedysregulation. l's,yciJialry RrI 49:257-272

I'dlock JM (1987), Carbama,epine ,ide cffeclS in children and aduhs.Epil'pIia 28(suppl 3):S(,4-S70

Pleak RR. Birmaher 1\. Gavrilescu A, Abichandal1i C, Williams DT (l9M8).Mania and neuropsychi."ic excilation following ",rbamazepine. } AmA'-lId Cbild Adolnr I'sy<hialry 27:500-';03

Po" RM, Keller TA. Dmicofr K el al. (19%). The place of a",iconvulsan,rherapy in hipolar iIIne.«. Pryd,oplJ(("n'icnln1..Y 128:115-129

Puente RM. Barrig. F, Morales M1' (1973). Emldio doble-ciego con(i1rbilma7.~pina en un grupo de cscolarcs con d3no cu~hr.al: comu­nieacion preliminar. M,d R", M,x 53:97-10 1

Rei.. AL, O'Donnell DJ (\984), Carbatrl31epine-inJuced mania in rwo chil­dren: case report.} Clin Pryehi""y 45:272-174

Rilkin A, Kara;gi B. Dicker R et al. (19<)7), 1.ithium rrealment of conducrdisorders in ado)escl·nlS. Artl} /'Ilrhialry \54:554-555

Roscl1berg DR. Hoillum J, Gershon S (1994), 'lex/book ofPharmarolh"apljor Child and AdolrIcml I'rychiatrie Disordm. New York: I\runnet/Mazd

Ry.n NO. Meyer V. Dachille S. Mav.ie D, Puig-Amich J (1988), Lilhium"nddL~r~ss:1nt augmrnt;1tion in TeA-refractory dC'pr~sion in adoles­cent'.} Am Arad Child AdolrIC I'~yehi,,'ry 27:~71 376

Schaffer CB, Sch.ffer Le (1')'.17), t.;ah.pcnlill in Ihe rr..Jlment of bipolardisorder. Am} P,ycbialry 154:291-292

Schiff HII. S.bin TO. Gefler A. Alexalldor I., Mark V (1982). Urhium inaggressive behavior. Am} PIyrhiarry 13'.1:1346-1.\48

Silberstein SO, Wilmore L.J (]9%), Divalproex sodium: migraine ,realmentanJ mon itoring. HMdad" 36:239-242

Silva RR. Comphcll M. Golden RR. Small AM, Paraki CS. Rosenberg CR(1992). Side effeClS associa'ed with lithium and placebo adminiSl"I;onin aggre,'ive children. l'rychopbarmawl 8u1l28:31 '.I-.i26

Silvo RR. Gonulcz N, Kafanlari, V, Campbdl M (1991), Long-rerm usc oflilhium in aggres,ive conducl disorder children. Presented ar rhe AnnualMccling of lhe American Academy of Child and Adolescent Psychiatry,S.n Francisco. October

Silva RR. Munoz OM, Alpert M (1'.1')6). Corbamazepine us< in children andadolescents with fl';l(UrC~ or ancnliun-dl·fit.:it hypc:racrivilY Jlsurdcr: ameta-analysis.} Am Acad Child AdolrIr l'Iyrhialry .15:352-358

Solomon £lA, Kcilner GI, Ryan CE. Mille; IW (i9%), polyph.rmacy inbipolar I disorder. I't.ychophnrnrarol HIIII32:579-587

Solomon DA, Ryan CEo Keitner GI el al. (1997), A pilm srudy of li,hiumcarbonate plus divalpTocx sodium for rhe continuation and m::lintcnancctrealment uf palienlS with hipolar I disorder. ./ Clin PlyeMlllry 58:9';-99

Slanton 51'. K<ek I'E Jr, McElroy SL (1997), Treatment of .cUle mania wilh!l"b,\,<,nrin (lwer). Am} Plyrhiatry 154:287

Slall AL, Banov M, Kolhrener M ...t al. (1994). Neurologic facrors I'redicI afavorable valproo[e response in bipolar .nd schizo.tTeerive disorders. }Cli" l~yclJOphllr",aroI14:.111-31j

535

RYAN ET At.

Slrober M (I~~7), The natur.li'lic prospective course of juvenile bipolar ill­ness. Present.d ar rhe Second Int.rn'lional Conference on BipolarDi50rder, Pimhurgh, June 20

Slrober M. fre.man R, Rigali J, Schmidl S, Diamond R (1992). Tb. phar­macoth.rapy of depressive i1lnes.' in adolesc.nce. II: elfccts of lithiumaugmenla<ion in nonr<spond... 10 imipramin•. ] Am AcnJ Child Adolnc"'ychiatry 31: 16-20

Suober M. Morrell W, Lamperr C, Burroughs J (1990), Relap'" followingdi$continU;1don of Hthium mainlC'nilnt::C' therapy in adolcscC'nts withbipolar I illn.ss; a nalur.li"ic Slody. Am} "',dJia,ry 147:457-461

Sluppacek C, Barnas C, Mille, C. Schwi.lCl J, Fleischhacker WW (1990),CarbaOl"'-ep;ne ill ,h. prophylaxis of mood disorders. ] C/i" i'Iycho­pharmaco/lO:3~-~2

Sugimoro T. Murn H, Woo M, Nishida N, Murakami K (I~%), Valproatem..abolites in high-dose valpma« plus phenytoin therapy. Epi/tpsia37: 1200-120.'\

SU7.uki Y, Cox S, Hay., J, Wal,oll PD (I~~I), Carbamncpin. age.dose miorelation.•hip in childr.n. T/'" Drug Mo"i' 13:201-208

Swann AC. Bawdcn CL, Morri, D ct aI. (I~~7), Depression during mania:treatment respon'" 10 lithium or dinJproex. Arrh Gtn I'rychiatry 54:37-42

Thakker KM, Maogat S. Garnet< WR, Levy RH, Kochak GM (1992).Comparative bioavailability and "cady Slat. Auctuations of l<gretolcornmcn:ial and carh3muC'pinc OROS rahl(u in ~dulr and ~diatrie

.pil.ptic pati.nts. Biophar", Drug DiJpos I.l:55l)-.569Varanka TM. Weller RA. Wcller EB. frimd MA (1~88). Lithium lreatmenl

of manic el'isod.. with psychotic fealur.s in p,epubc:nat children. Am}i'Iychiarry 145:1557-1559

Veuo A, Slemisrvanyi 1, rallal\ L, Vargha M. S7jlard J (1985). Therap"oticexp<ri.nce with lithium in childhood aggressiviry. Nturopsychobiology14:121-127

Viliello B. B.har D. Ryan P, Malon. R, Delaney MA (1987), Saliva lithiummonitoring.] Am Acad Child Adolm Prychia,ry 26:812-813

Wagner JG, Rocchini Ap, Vasiliad.s J (1982), Prediction of Sleady-state""rap.milplasma concentrations in children and aduhs. Cli" PharmacolThtr 32;172-181

Walden J. H.sslinger B. van Calker D. Berger M (1996). Addilion of lamo­trigine to vall"o". may .nhance efficacy in the tr....m.nt of bipolarafT.etiv. disorder. Pharmacopsychiftrry 29:193-195

Well.r EB. Weller RA. Fristad MA. Cantwell M, Tucker S (1987). Salivalithium monitoring in prepubertal children.] Am Acad Child AdoltIePrychiarry 26: 173-175

Weiner A, Weiner Z. L.onant MA (1977). Bipolar manic-depressive disord.r:a re""",ment of course and outcom•. Compr I'ryc/'u,try 18:327-332

W.st SA. Keck pE. MeElroy SL er at (1994). Open trial of valproat. in the".atm.nr of adolescent mania. } Child Adol"c Psychopharmacol4:263-267

West SA. Keck pEJ. McElroy SL (1995). Oral loading dos.s in th. valproate"eatm.nt of adolescents with mix.d bipolar disorder. } Child Adol"c"'yc/'opharmaco/5:225-231

Whinier Me. West SA. Galli VB. Raute NJ (1')')51. Valproic acid for dysphoricmania in a mentally l'C<ard.d adolescent. ) Gi" I'rychiarry 56:590--591

W07.niak J. Biederman J. Mundy E. M.nnin D. Faraone SV (1995). A pilotfamily study of childhood-ons.t mania. } Am Acad Child Adolnci'Iychiatry .~4: 1577-1583

Youngerman J, Canino fA (19781. \jthium carbo""" Ute in children andadol.scents: a survey of theli ..rature. Arch G... l'rychiatry 35:216-224

Yukawa E (1995). A feasibility study of ,h. multiple.peak approach forpharmacokin.tic screening: popularion-based investigation of valproicacid relative clearance using routine clinical pharmacokinetic data. ]Pharm Pharmacol47: 1048-1052

Yukawa E. Auyama T (19')6), Detection of carbamaupine drug interactionby multipl. peak approach scrccning using routine clinical pharmaco.kinetic data.} eli" Pharmacol 36:752-759

Zaccara G. Mmori A. Moroni F (1988). Clinical pharmacokinetics of val­proic .cid-1988. Clin Pharmacokind 15:367-389

536

Children's Understanding of Sun Protection Behaviours: A Comparative Analysis. J. Morris. M. Bandaranayake. R. McGee

Ohjwivt: Tu invcsti~are awareness of sun protection behaviours in a sample of primary school children in New ualand.McrhodoiDgy: Information was collect.d from 824 primary school children in New l.<-aland using a drawing and wtiting technique.RLJlJJs: The data rev.aled a bi... towards sunsc,een as a method of sun protcction compared with other methods such as clothingand the usc of shad•. Comparisons betwcc:n results obtained from children resident in Australia and England indicat.d a greater

awaroness of sun protection methods amongst the childr.n from Auslralia and New ualand compared with those children livingin England. Concl"Jions: Children as young as 5 and 6 can describe the consequences of overexposure to the sun. and can iIIus­Ira.. mtthnds of sun protection. Sunscreen is seen as the main method of sun protcction. J Paediatr Child Health 1998;34:254-259

J. AM. ACAD. CHILL> AllOLI-:SC. PSYCHIATRY, .\K;5, MAY 1999