Child and Adolescent Bipolar Disorder: a Review of the Past 10 Years
January 1st, 1997
by Barbara Geller, M.D. and Joan Luby, M.D.
J Am Acad Child Adoles Psychiatry 36:1168-1176, 1997
Objective: To provide a review of the epidemiology, phenomenology, natural course, comorbidity, neurobiology, and treatment of child and adolescent bipolar disorder (BP) for the past 10 years. This review is provided to prepare applicants for recertification by the American Board of Psychiatry and Neurology. Method: Literature from Medline and other searches for the past 10 years, earlier relevant articles, and the authors' experience and ongoing National Institute of Mental Health-funded project "Phenomenology and Course of Pediatric Bipolarity" were used. Results: Age-specific,developmental (child, adolescent, and adult) DSM-IV criteria manifestations; comorbidity and differential diagnoses; and episode and course features are provided. Included are age-specific examples of childhood grandiosity, hypersexuality, and delusions. Differential diagnoses (e.g. specific language disorders, sexual abuse, conduct disorder [CD], schizophrenia, substance abuse), suicidality, and BP-II are discussed. Conclusion: Available data strongly suggest that prepubertal onset BP is a nonepisodic, chronic, rapid cycling, mixed manic state that may be comorbid with attention-deficit hyperactivity disorder(ADHD) and CD or have features of ADHD and/or CD as initial manifestations. Systematic research on pediatric BP is in its infancy and will require ongoing and future studies to provide developmentally relevant diagnostic methods and treatment. J Am Acad Child AdolesPsychiatry, 1997, 36(9):1168-1176. Key Words: child, adolescent, bipolar disorder, mania, hypomania, delusions, grandiosity, hypersexuality
As noted in a recent letter by Schneider etal. (1996), if one looks to fit children and adolescents into adult criteria for manic-depressive illness, it will be difficult except for those adolescents who have adult-type onset, i.e., individuals with good functioning until the abrupt onset of marked manic symptomatology that often requires hospitalization, is responsive to treatment, and is succeeded by inter episode well-being (McGlashan,1988). Thus, a developmental, age-specific viewpoint needs to be considered for pediatric patients who do not have the adult-type onset.
Analogies to two other occurrences in bioscience are useful to understand the developmental perspective. The first occurrence is that different illnesses may have different neurobiological (e.g. genetic,neurotransmitter) mechanisms and thus have differences in severity with earlier age of onset (Childs and Scriver, 1986). A classic example is the comparison of juvenile to adult-onset diabetes in which genetic mechanisms and severity differ. The second situation occurs when the same causative agent can have different clinical manifestations at different times in the life cycle. An example of the second situation is when 6-OH-dopamine is given to infant versus adult rats. Infant rats given this compound develop hyperactivity, whereas geriatric rats develop parkinsonian symptomatology.
On the basis of the occurrence of either (or both) of the neurobiological mechanisms noted above, it is developmentally possible for childhood-onset manic-depressive illness to be more severe; to have a chronic, nonepisodic course; and to have mixed, rapid-cycling features similar to the clinical picture reported for severely ill, treatment-resistant adults (Geller et al., 1995;Himmelhoch and Garfinkel, 1986;Hsu, 1986; Hsu andStarzynski, 1986; Stancer and Persad,1982). A possibility also exists that only the most severe manic-depressive children receive clinical attention because manic episodes that last a few weeks might be tolerated by parents as a phase of growing up, especially if these do not interfere with school performance. Our experience of colleagues requesting "hallway" consultations suggests that this may be the case.
The following review assumes that future data will support continuities across the age span.
As yet, no national or international epidemiological study of bipolar disorder (BP) during the pediatric years is available. However, data fromCarlson and Kashani (1988) andLewinsohn et al. (1995) suggest that prevalence during the adolescent years is at least that of the adult population.
These reports and those below taken together -- i.e. epidemiological data, secular trends, high switch rates, data from inpatient services, and reports from chart reviews -- support that the prevalence of child and adolescent manic-depressive illness is at least that in the adult population and may be increasing. A secular trend, i.e., earlier age of onset of BP with successively later years of birth, has also been reported(Rice et al., 1987).
Underdiagnosis of childhood bipolarity has been noted by several authors who have described a high prevalence on inpatient services (Gammonet al., 1983; Isaac, 1995) and a high prevalence of both diagnosed and undiagnosed cases on chart reviews (Welleret al., 1986; Wozniak et al., 1995). Another source of underdiagnosis during childhood and adolescence is that manyparents who are bipolar, and thus at higher risk of having bipolar offspring, remain underdiagnosed themselves (Geller,1996). These parents may not recognize the pathological implications of their children's manic behaviors.
Literature on adult samples has noted that 20% to 40% of adults report that their onset was during childhood (Joyce,1984; Lish et al., 1994). Adults with childhood onset by history often also report that the initial episode was depressive(Lish et al., 1994). The latter is consistent with the high rate of switching of prepubertal depression to prepubertal mania(32%) reported by Geller et al (1994b) and of depressed adolescents switching to adolescent-onset mania (20%) (Strober and Carlson, 1982). These rates of switching may be conservative because of the probable underdiagnosis of childhood mania discussed above.
At all ages, manic subjects in the cross-section appear to be the happiest of people because of their infectious, amusing, elated affect. This is also true of children, and it can be very misleading to see a happy child laughing in the office in the context of a miserable history(e.g., school suspensions, family fights). This contrasts with sad, depressed children who everyone thinks are ill because it is more difficult to acknowledge conceptually that happy children have serious psychopathology. Thus, it is important to evaluate children's affect in relationship to historical features in exactly the way one evaluates the incongruity between the infectious elation of manic adult patients in th econtext of histories that include loss of family, unemployment, and jail sentences.
Across the life span, grandiose delusions must be judged by failure to follow the laws of logic and by a firm belief (often to an extent thataction is taken). A common presentation for bipolar children is to harass teachers about how to teach the class; this harrassment is often so intense that teachers telephone parents, begging them to ask their children to desist. These children may fail subjects intentionally because they believe the courses are taught incorrectly. Therefore, their thinking bypasses laws of logic (i.e. that children can choose what to fail or pass), and the beliefs are acted upon by purposely failing courses. Another common grandiose manifestation in children as young as seven is to steal expensive items and be impervious to police officers who attempt to make them understand that what they have done is wrong and illegal. Similar to grandiose adults, grandiose children believe that stealing maybe illegal for other people but not for them. Unlike patients with pure conduct disorder, manic children and adolescents, similar to bipolar adults, frequently know that stealing is a bad thing to do, but they believe that they are "above" the law. Common adolescent grandiose delusions are that they will achieve a prominent profession(e.g. lawyer) even though they are failing at school, i.e., the belief that they can have a high attainment when they have failing school grades bypasses the laws of logic. Asked how he or she will become a lawyer, an adolescent will answer is "I just know I will". Similarly, a manic adolescent, even in the absence of musical talent or ability to carry a tune, might practice all day with the belief that he or she can become a rock star.
Dissimilar to depressed patients who have trouble falling asleep and lie in bed brooding, manic children have high activity levels in the bedroom prior to sleep, e.g. rearranging furniture for several hours. Manic adolescents will wait until parents are asleep and then go out "partying,"whereas manic adults will party and work around the clock.
Pressured speech is relatively similar at all ages in that the individual can be difficult or impossible to interrupt. Racing thoughts are frequently described by children and adolescents in very concrete terms. For example, children state that they are not able to get anything done because their thoughts keep interrupting. An adolescent wished she had a button on her forehead to turn off her thoughts. Flight of ideas in children is similar to that in adults except for age-specific content,e.g., "Do you live in Nashville? Some people have hogs for Thanksgiving. Do you have a key to that door?"
Also at all ages, minor perturbations in the environment can produce marked amounts of distractibility. Increased motor activity and goal-directed behaviors in children and adolescents frequently look like normal activities done in a profuse amount. The manic child may in a brief period of time make curtains, begin an illustrated book, rearrange furniture, and make multiple phone calls, compared with the manic adult,who may start many businesses and join many social groups.
Involvement in pleasurable activities with a high level of danger ismanifested in age-specific behaviors. Hypersexuality in children frequently begins when a child brought up in a conservative home without any history of sexual abuse or excessive exposure to sexual situations begins to use profanity and may tell a teacher to "f--- herself"and "gives her the finger." Children may masturbate frequently,initially openly, and then when told not to do it publicly will simply make frequent trips to the bathroom to continue the stimulation. Children will begin to proposition teachers and make overt sexual comments to classmates. Adolescents develop romantic fantasies and delusions about teachers (see vignette in Geller et al. 1995).Older children and adolescents will call the 1-900 sex telephone lines, which the family discovers when the telephone bill arrives. Older adolescents and adults will have multiple partners with unprotected sexual behaviors and frequently will have an urgency to have sex, e.g., an adolescent wrote to her boyfriend, starting the letter with a sentence that said "When are we going to f---?" Adults will have multiple partners; males may be womanizers; and often there are multiple marriages.
Interest in money appears in young children when they start their own businesses in school and when they begin to order multiple items, trips,and plane tickets from advertised 1-800 and 1-900 telephone numbers. Again, the family frequently does not discover this until items arrive at the house and telephone bills arrive. Adults may overdraw on bank accounts and "top out" on multiple credit cards.
Across the age span, taking more dares is common. In older adolescents and adults, this frequently appears as wild driving, eventuating in many speed and "driving under the influence" tickets. In children it manifests as grandiose delusions that they can fly out the window because they believe that they have that ability or in exaggerations of usual childhood hopping around on trees or between roof tops, based on beliefs that they are above the possibility of danger.
To further exemplify pediatric features, characteristic vignettes of children and adolescents with BP can be reviewed in Gelleret al. (1995). Characterization of preschool-age BP is an important avenue for future investigation.
DIFFERENTIAL DIAGNOSIS AND COMORBIDITY
Table 1 provides a list of differential diagnoses and/or comorbid conditions by age group.
Sexual abuse is especially important as a differential diagnosis duringthe childhood years because manic hypersexuality is often manifested in children by self-stimulatory behaviors including frequent masturbation. Thus, it is useful to obtain a careful history of whether the child could have been abused or exposed to adult sexual behaviors.
Specific language disorders need to be differentiated from flight of ideas because children and adolescents with language disabilities cansound as though they have a thought disorder when they partake inconversation without actual comprehension of the content and/or theability to find the appropriate words to use.
At present, data suggest that for some prepubertal-onset bipolar children, hyperactivity manifestations begin at preschool age and arefollowed by a full manic syndrome during the early grade-school years (Geller,1997b). In these children, it is possible that hyperactivity is the first developmentally age-specific manifestation of prepubertal-onset BP. This hypothesis is consistent with the higher prevalence of attention-deficit hyperactivity disorder (ADHD) in prepubertal- versus adolescent-onset BP. For other bipolar children, ADHD and BP may be comorbid, i.e.,hyperactivity is a separate disorder that coexists. Numerous authors (Biedermanet al., 1995; Borchardt and Bernstein, 1995;Fristad et al., 1992; Gelleret al., 1995; Strober et al., 1988;West et al., 1995) have noted the high prevalence of symptoms of hyperactivity among children and adolescents with bipolarity. When subjects are seen initially because of bipolar symptomatology, approximately 90% of prepubertal and 30% of adolescent bipolars have ADHD (Geller et al., 1995).Manifestations of ADHD overlap with those of multiple other DSM-IVdiagnoses (e.g., BP, major depressive disorder [MDD]). Thus, validation of the distinctness of coexistent ADHD versus similar symptom clusters but dissimilar pathogenesis must await future naturalistic course, family genetic, and other neurobiological studies (Biedermanet al. 1991; Geller, 1997b).
Even with the relatively conservative DSM-IV criteria, conduct disorder occurs in approximately 22% of bipolar children and 18% of bipolar adolescents (Geller et al., 1995).Conduct disorder, similar to ADHD, may be an initial manifestation of prepubertal-onset BP (Geller, 1997b;Kovacs and Pollock, 1995). These comorbid conduct disorders appear related to poor judgment and grandiosity. As an example,a 7-year-old child stole a go-cart, an item that costs several hundred dollars, and was completely unfazed when the police appeared and tried to admonish him, thus demonstrating the grandiosity of stealing such a large object and of being impervious to legal intervention. Conduct disorders during adolescence (which may include driving under the influence, running away for sexual adventures, and stealing large amounts of jewelry)frequently lead to placement of these youngsters in juvenile facilities. Adult antisocial equivalents are well known (e.g., buying new television sets for every room in the hospital; obtaining real estate that the individual cannot afford).
During the teenage years, because of greater perceptual distortions seen in bipolar illness during adolescence, schizophrenia is a major differential (Horowitz, 1975). Differentiation is greatly aided by a family history of mania, which is more probable for BP than schizophrenic adolescents (Strober etal., 1988).
Substance abuse begins to be an important comorbid condition during the teenage years and is an important differential (Horowitz,1975; 1977). For example, laughing fits may be due to smoking marijuana as a differential from the laughing fits that occur during the pediatric years as a manifestation of elation. Furthermore, very rapid cycling (Table2) that is a hallmark of child and adolescent bipolarity (Gelleret al., 1995) can easily be mimicked by amphetamine highs followed bywithdrawal "crashes." Hallucinogens can mimic bipolar perceptualdistortions (Horowitz, 1975; 1977).
Similar to the multiple comorbid anxiety conditions seen with MDD,bipolar patients also manifest multiple comorbid anxiety conditions(approximately 33% of bipolar prepubertal patients and 12% of bipolar adolescent patients ) (Geller et al., 1995).
Table 2 provides a comparison between pubertal-onset versus adult-onset episode and course features.
As noted in the beginning of this article, prepubertal onset manic-depressive disorder may not present with the sudden or acute onset and improved interepisode functioning characteristic of the disorder in older adolescents and adults. Rather, it may present with a picture of continuous, mixed manic, rapid cycling of multiple brief episodes described in detail by Geller et al. (1995).Thus, children may be having a laughing fit and happily doing an arts and crafts project when, without any environmental prompt, they will suddenly become miserable and acutely suicidal, talking about wanting to shoot themselves. Parents frequently describe their frustration at not being able to convince practitioners that their children rapidly cycle, sometimes numerous times in each day. Because this history has been given independently by parents (including those from many parts of the United States who have received Dr. Geller's name from the National Institutes of Health) who have no idea that this cycling pattern has been described by other parents, there is no reason to disbelieve these parental observations. Adults with mixed manic, rapid-cycling BP have a poorer prognosis than those with discrete episodes (Kelleret al., 1993). Therefore, future studies of the adult course of BPchildren will be crucial for developing long-term, prophylactic treatmentsfor implementation during the prepubertal years. Naturalistic follow-up ofbipolar adolescent inpatients has evidenced a poor prognosis (Stroberet al., 1995).
One of the issues that arises for child and adolescent manic-depressiveindividuals is whether or not BP-II disorder has the same implications asit does in the adult population (Coryell etal.,1995; Geller et al., 1994b). The switchrate from BP-II to BP-I in adults has been estimated byCoryell et al. (1989, 1995) to be similar to thelow rate reported by Geller et al. (1994b) forswitching from BP-II to BP-I among prepubertal subjects who switchedduring the prepubertal period. However, it remains possible that BP-II inchildren and adolescents may be an age-specific, developmental precursorto BP-I (Geller et al., 1994b). If the latterwere established, then treatment for BP children might differ from thatfor BP adults in whom BP-II is often treated with the same regimen as MDD(Frank and Kupfer, 1985). Treatment for MDD inpotentially BP pediatric patients may be contraindicated because there isevidence, albeit controversial, that antidepressant therapy mayprecipitate or worsen rapid cycling (Akiskal etal., 1985, 1995; Geller et al., 1993;Wehr and Goodwin, 1979; Wehret al., 1988).
Further research will also be needed to provide better differentiationof whether the Akiskal et al. (1995) concept oftemperamental issues (i.e., that there is essentially a constanttemperamental modulation in some patients) is only semantically differentfrom the mixed manic picture of BP-I and BP-II children, adolescents, andadults (Geller et al., 1995; Kelleret al., 1993).
The role of comorbid personality disorders as prognostic and coursefeatures of adolescent BP remains a poorly studied but important areabased on reported interepisode personality trait impairments in BP adults(Solomon et al. 1996). Johnsonet al. (1995) have noted cluster II personality disorders were moreprominent among BP adolescents. Other work in the area of personalitydisorders among pediatric BP individuals is not yet available, in partbecause of the need for further work on instrumentation (Brentet al., 1990).
Data support a higher risk of suicidality among BP adolescents comparedto adolescents with other diagnoses (Brent et al.,1988, 1993). In addition, comorbidity of mood and substance use disordershas been correlated with higher suicide risk in older adolescents andyoung adults (Rich et al., 1986, 1990). The wellknown high comorbidity of substance dependency and BP in adults isespecially notable because data suggest that "secondary"substance use is more amenable to treatment and has a better prognosis (Geller,1997a; Winokur et al. 1995).
In their classic 1986 paper, Childs and Scriverdescribe different genetic mechanisms for medical illnesses that have bothan early- and late-onset form, e.g., diabetes mellitus. In 1988 and 1992,Strober et al. described this phenomenon forpediatric bipolarity, noting that prepubertal-onset bipolarity was morelikely to be associated with early aggressive hyperactivity, lithiumresistance, and greater familial loading. Thus, clinically it is useful toidentify parents who may have undiagnosed bipolarity (Geller,1996). This is best done by asking way-of-life questions (e.g. howrelatives manage money; driving histories; relatives with more than fourmarriages) because patients with undiagnosed mania are unlikely to thinkof themselves as ill. Vignettes of relatives with undiagnosed mania appearin Geller (1996). Also, possible relationships ofgenomic imprinting (preferential maternal or parental transmission) andmitochondrial inheritance (maternal transmission) to pediatric age ofonset of BP remain intriguing issues for future research (Grigoroiu-Serbanescuet al. 1995; McMahon et al. 1995).
Familial aggregation of alcoholism among bipolar adults has been notedto be greater than among subjects with other diagnoses (Winokuret al., 1995, 1996). A similar high prevalence of alcoholism amongfirst-degree relatives of prepubertal and adolescent subjects with mooddisorders has been reported (Geller, 1997a;Geller et al., 1990, 1992; Puig-Antichet al., 1989; Todd et al., 1996). Further researchon prognostic implications of familial alcoholism among pediatric BP casesis warranted. Another promising line of investigation includes geneticallybased malformation syndromes that include BP behavioral manifestations (Papoloset al., 1996).
The few available neurobiological studies include a single case study ofa hypomanic child who had significantly different urinarymethoxyhydroxyphenylglycol level from those of normal controls (McKnewet al., 1974), a report of enlarged ventricles and increased number ofhyperintensities in a small open pilot study of bipolar children andadolescents (Botteron et al., 1995), and areport comparing sleep and neuroendocrine parameters in depressedadolescents with BP outcomes and those who remain depressed (Rao,1994).
Available work on cognitive characteristics of child and adolescentbipolarity is sparse but includes the work of Decinaet al. (1983). That report noted a significant discrepancy between Verbaland Performance IQ scores in offspring of bipolar parents but not in thenormal control group. This is consistent with neurobiological data inadult samples that support right-sided brain impairments in manicindividuals (Sackeim and Decina, 1983). Fristad(personal communication, November, 1993) noted that bipolar subjects hadhigher IQs than an ADHD control group. Also Kutcher(1993) reported a decrease in math performance based on school recordsamong prebipolar adolescents. This finding may be consistent with theDecina et al. (1983) findings on lowerperformance IQ. It is clear that further work on cognitive impairments andtheir prognostic and treatment implications is needed.
Treatment of childhood bipolarity remains a remarkably understudied areain spite of voluminous literature comprising more than 400 case reports,studies with small numbers, and investigations with populations that werenot diagnosed with DSM-III or higher criteria (Botteronand Geller, 1995; Fetner and Geller, 1992;Kafantaris, 1995; Youngermanand Canino, 1978). Thus, unless there is an expectation thatchildhood bipolarity completely mimics the adult treatment considerations,separate study is warranted. Compelling arguments, however, against thesimilarity of treatment of bipolarity across age groups can be constructedby analogy to the treatment differences between childhood and adult MDD (Gelleret al., 1996). Because there is as yet only one completed double-blind,placebo-controlled study of any medication for child or adolescent maniausing rigorous methodology and design (Geller,1997a), the clinician will be tempted to extrapolate from studies ofadults. However, extrapolation from treatment of MDD in adults did notprove useful, i.e., tricyclic antidepressants have never been shown towork better than placebo in any study of a child and adolescent population(Geller et al., 1996).
The pharmacokinetics of lithium in children has been studied (Vitielloet al., 1988), and, as expected, lithium has a shorter half-life inchildren than in adults. The latter is expected because of the moreefficient renal system of children. More recently, in a completeddouble-blind, placebo-controlled study of lithium for adolescents who werebipolar and substance dependent, lithium was significantly more effectivethan placebo by both completer and intent to treat analyses (Geller,1997a). Literature on lithium suggests that it can be given to childrenwith the same safety precautions used in adults and with similarmonitoring at 6-month intervals for renal, thyroid, calcium and phosphorusindices (Fetner and Geller, 1992;Khandelwal et al., 1984). Further, adouble-blind, placebo-controlled study of lithium for aggressive childrenhas highlighted that there may be some children who will develop cognitiveimpairment at low plasma levels (Silva et al.,1992). This was also noted in a double-blind, placebo-controlled study oflithium for depressed children who had predictors of future bipolarity (Gelleret al., 1994a).
The safest, most rapid method of prescribing lithium is to do sopharmacokinetically using a nomogram (Cooper etal., 1973; Fetner and Geller, 1992;Geller and Fetner, 1989). Alternately, ifobtaining a serum lithium level 24 hours after a single dose isimpractical, a 300-mg total daily dose can be administered until steadystate is reached (Fetner and Geller, 1992). Ifthe lithium level at the 300-mg daily dose is not between 0.8 and 1.2mEq/L, then a linear proportion can be made to estimate the dose needed toreach the desired level (Geller and Fetner,1989). Because of genetic variation in rate of elimination of lithium,slow eliminators can develop unacceptably high serum lithium levels andadverse effects if nonpharmacokinetic administration such asmilligram-per-kilogram dosing is used (Hagino etal. 1995). Tactical problems and side effects with lithium are discussedin detail in Fetner and Geller (1992).
Lithium, however, is not a drug that can be given either to chaoticfamilies or families who are unable to keep multiple appointments formonitoring of lithium levels and renal and thyroid functioning. Many youngbipolar patients have at least one bipolar parent and some (Gaensbaueret al., 1984; Grigoroiu-Serbanescu et al.,1989), but not all (Anderson and Hammen, 1993),offspring studies attest to the negative impact bipolar parenting canproduce. Therefore, it is imperative to have choices of medications thatcan safely and effectively be given in chaotic environments. Furthermore,because of the cycling and abrupt onset of suicidality, it is alsoimportant to have medications that would be safer than lithium if taken inoverdose.
There are a few open, uncontrolled studies addressing anticonvulsanttreatment of BP. Papatheodorou and Kutcher(1993) reported that valproate showed promising results.Himmelhoch and Garfinkel (1986) reported onthe use of carbamazepine for lithium resistant adolescents. A report byIsojarvi et al. (1993) in the New EnglandJournal of Medicine showed that polycystic ovarian disease developed in89% of young females receiving valproic acid for epilepsy compared with27% of epileptic females who were not receiving this preparation. In a1996 article, Isojarvi et al. noted thatvalproate was associated with onset of obesity in more than half of thewomen and that these individuals also developed polycystic ovariandisease. Obviously, these would be prohibitive side effects for mostfemale children with manic-depressive illness. Further work on whether ornot this side effect appears only when the medication is given forepilepsy and independent replication of these findings are warranted.Details of valproate and carbamazepine administration are provided inBotteron and Geller (1995). Low dosechlorpromazine may be another alternative (Botteronand Geller, 1995).
Methylphenidate has been reported, in case studies, both to worsen (Koehler-Troyet al., 1986) and to be a first line of medication (Maxet al., 1995) for bipolar children and adolescents. At present, there is aneed to use trial and error to judge which patients benefit and whichmight be made worse by stimulant medication.
Due to the chronic course of childhood manic-depressive illness andbecause of rapid cycling, mixed features which are known to predict poorresponse in older populations (Geller et al.,1995; Himmelhoch and Garfinkel, 1986;Hsu, 1986; Keller et al.,1993), duration of antimanic treatments is complex. Further, literature onadults suggest that intermittent lithium therapy is worse for outcome thancontinuous, noninterrupted therapy and that it can be difficult torestabilize patients on lithium after interruptions (Ahrenset al., 1995; Muller-Oerlinghausen et al., 1992,1994; Schou, 1995; Schouet al., 1989). Of note, Strober et al. (1990,1995) keep adolescents on antimanic treatments throughout the teenageyears. Strober et al. (1990) have also reportedan open, uncontrolled naturalistic follow-up study of adolescents onlithium. These data strongly support long term maintenance lithium becausesubjects who discontinued lithium had a significantly higher relapse rate.
As yet, this area has not been investigated for children and adolescentswith BP. It may, however, be especially important because of the knownincreased significance of nonshared environmental factors during the earlychildhood years (Pike and Plomin, 1996). Studiesshowing the relationship of negative expressed emotion to poorer outcomeamong bipolar adults argue for similar investigation of nonsharedenvironmental factors among childhood populations (Miklowitzet al., 1988).
Among adults, impairment in psychosocial functioning between BP episodeshas been reported (Coryell et al., 1993;Gitlin et al., 1995). The latter is relevant toan ongoing controlled study of adults who, after stabilization onmedication, are randomly assigned to either a family-focused or a combinedinterpersonal and social rhythm (e.g., sleep) intervention (Miklowitzet al., 1996). It is clear that similar studies of psychosocial therapiesamong younger populations will be warranted when medication maintenancestudies for pediatric BP become available.
J, Wozniak J, Kiely K et al.(1995), CBCL clinical scales discriminate prepubertal children withstructured interview-derived diagnosis of mania from those with ADHD. JAm Acad Child Adolesc Psychiatry 34:464-471
E, Kupfer D (1985), Maintenance treatmentof recurrent unipolar depression: Pharmacology and psychotherapy. In: ChronicTreatments in Neuropsychiatry, Kemali D, Racagni G, eds. New York:Raven, pp 139-151
GD, John K, Rothblum ED, Mullen K,Tischler GL, Weissman MM (1983), Use of a structured diagnostic interviewto identify bipolar disorder in adolescent inpatients: frequency andmanifestations of the disorder. Am J Psychiatry 140:543-547
B (1997a), Double-blindplacebo-controlled study of lithium for adolescents with comorbid bipolarand substance dependency disorders. Presented at the Annual Mid-YearInstitute of the American Academy of Child and Adolescent Psychiatry,March 19-21, 1997, Hamilton, Bermuda.
B, Cooper TB, Graham DL, Fetner HH,Marsteller FA, Wells JM (1992), Pharmacokinetically designed double-blindplacebo-controlled study of nortriptyline in 6-12 year olds with majordepressive disorder. J Am Acad Child Adolesc Psychiatry 31:34-44
B, Cooper TB, Graham DL, Marsteller FA,Bryant DM (1990), Double-blind placebo-controlled study of nortriptylinein depressed adolescents using a "fixed plasma level" design.Psychopharmacol Bull 26:85-90
B, Cooper TB, Zimerman B, Sun K,Williams M, Frazier J (1994a), Double-blind placebo-controlled study oflithium for depressed children with bipolar family histories. Neuropsychopharmacology10(S-122):541S
-Serbanescu M, Christodorescu D,Jipescu I, Totoescu A, Marinescu E, Ardelean V (1989), Psychopathology inchildren aged 10-17 of bipolar parents: psychopathology rate andcorrelates of the severity of the psychopathology. J Affect Disord16:167-79
-Oerlinghausen B, Ahrens B, Grof E et al.(1992), The effect of long-term lithium treatment on the mortality ofpatients with manic-depressive and schizoaffective illness. ActaPsychiatr Scand 86:218-222
-Oerlinghausen B, Wolf T, Ahrens B et al.(1994), Mortality during initial and during later lithium treatment. acollaborative study by the International Group for the Study ofLithium-Treated Patients. Acta Psychiatr Scand 90:295-297
DF, Faedda GL, Veit S et al. (1996),Bipolar spectrum disorders in patients diagnosed with velo-cardio-facial syndrome: Does a hemizygous deletion of chromosome 22q11 result in bipolar affective disorder? Am J Psychiatry 153:1541-1547
U, (1994), Outcomes of unipolar and bipolar depression: Clinical and biological predictors. Presentation at the AnnualMeeting of the American Academy of Child and Adolescent Psychiatry,October 25-30, 1994, New York
HA, Decina P (1983), Lateralizedneuropsychological abnormalities in children of bipolar probands. In: Lateralityand Psychopathology, Flor-Henry P, Gruzelier J, eds., New York:Elsevier Science Publishers
M, Carlson G (1982), Bipolar illness in adolescents with major depression: clinical, genetic, and psychopharmacologic predictors in a three- to four-year prospective follow-up investigation. Arch Gen Psychiatry 39:549-555
M, Morrell W, Burroughs J, Lampert C,Danforth H, Freeman R (1988), A family study of bipolar I disorder in adolescence: early onset of symptoms linked to increased familial loading and lithium resistance. J Affect Disord 15:255-268
M, Schmidt-Lackner S, Freeman R, BowerS, Lampert C, DeAntonio M (1995), Recovery and relapse in adolescents with bipolar affective illness: a five-year naturalistic, prospective follow-up. J Am Acad Child Adolesc Psychiatry 34:724-731
J, Biederman J, Kiely K, Ablon JS,Faraone SV, Mundy E, Mennin D (1995), Mania-like symptoms suggestive ofchildhood-onset bipolar disorder in clinically referred children. J AmAcad Child Adolesc Psychiatry 34:867-876
ChildAdolescentAdultSpecific language disordersXAttention-deficit hyperactivity disorderXXOppositional defiant disorderXXConduct disorderXXSexual abuseXXSchizophreniaXXSubstance abuseXXAntisocial personalityX
Prepubertal and Young AdolescentOlder Adeolescent and AdultInitial episodeMajor depressive disorderManiaEpisode typeRapid-cycling, mixedDiscrete with sudden onsets and clear offsetsDurationChronic, continuous cyclingWeeksInterepisode functioningNonepisodicImproved functioning
© 1997 by the American Academy of Child and Adolescent Psychiatry. This article is reproduced here with permission of the Academy.
Dr. Geller is Professor of Psychiatry and Dr. Luby is Assistant Professor of Psychiatry, Department of Psychiatry, Washington University School of Medicine, St. Louis.
Reprint requests to Dr. Geller
Washington University School of Medicine
4940 Children's Place
St. Louis, MO 63110
This work was supported by NIMH Grant R01 MH53063 "Phenomenology and Course of Pediatric Bipolarity" to Dr. Geller.
- barbara geller