Stephen V. Faraone1
In their masterful chapter about gender and age in bipolar disorder, Shul-man et al have provided us with a comprehensive view of the heterogeneity of bipolar disorder (BD). As they describe, age at onset is a dimension of heterogeneity that has received much attention in recent years. With one report finding that 59% of BD patients had symptom onset prior to the age of 20, it would appear that this seriously disabling mental disorder frequently begins in youth.
Despite such strong evidence for youth onset BD, its existence was ignored for many years. Then, in the 1980s, several case reports and series described children presenting with symptoms suggestive of BD, which responded to lithium carbonate [1-4]. Yet many of these children had never been diagnosed with a mood disorder, which suggested that the pediatric form of BD, rather than being rare, might be difficult to diagnose due to atypical clinical features.
These atypical features have been well characterized in the literature [5,6]. Among BD children, the most common mood disturbance is not euphoria [2, 7]. Instead, these children present with severe irritability and prolonged and aggressive outbursts of temper . The irritability of childhood BD is severe, persistent, and often violent . Although these severe outbursts are episodic, these children are chronically irritable or angry [2, 7, 9]. In their literature review, Geller and Luby  concluded that childhood-onset BD is a non-episodic, chronic, rapid-cycling, mixed manic state. Thus, one reason why pediatric BD is difficult to diagnose is because it presents with an atypical picture characterized by predominantly irritable mood, mixed with symptoms of major depression, and chronic course, as opposed to euphoric, biphasic, and episodic course. With some exceptions, the clinical features of BD are similar in childhood and adolescence . Notably, the atypical features of youth BD are similar to the syndrome of mixed BD, which affects about one-fourth of BD adults .
1 Pediatric Psycho-pharmacology Unit, Massachusetts General Hospital and Harvard Medical School Department of Psychiatry, 15 Parkman Street, Boston, MA 02114, USA
As Shulman et a! note, another source of diagnostic dilemma for childhood BD is its frequent comorbidity with attention deficit hyperactivity disorder (ADHD). Most BD youth meet diagnostic criteria for ADHD, with prevalence figures ranging from 60% to 90% [8, 12-14]. Faraone et a!.  found that adolescents with childhood-onset BD had the same rates of comorbid ADHD as manic children (90%) and that both these groups had higher rates of ADHD than adolescents with adolescent-onset BD (60%). Sachs et a!  reported that, among adults with bipolar disorder, a history of comorbid ADHD was only evident in those subjects with onset of BD before 19 years of age. Chang et a!.  reported that the onset of BD in adults with a history of ADHD was 11.3 years of age. These findings suggest that early onset of BD may identify a developmental subtype of the disorder that is frequently comorbid with ADHD [10, 17, 18].
The high levels of comorbidity between BD and ADHD raise a fundamental, nosologic question: do children presenting with BD and ADHD have ADHD, BD, or both? One method to address these uncertainties has been to examine the transmission of comorbid disorders in families . As shown in a meta-analysis by Faraone et a! , studies that examined rates of ADHD (or ADD, hyperactivity) among the offspring of adults with BD found higher rates of ADHD among these children compared with controls. Likewise, the meta-analysis showed that family studies of ADHD children find high rates of BD among relatives.
In three different samples, Faraone et a! [20-22] found that relatives of children with BD were at high risk for ADHD, that was indistinguishable from the risk in relatives of children with ADHD and no BD. However, BD and the comorbid condition of BD plus ADHD selectively aggregated among relatives of manic youth compared with those with ADHD and comparison children. This pattern of transmission in families suggested that BD in children might be a familially distinct subtype of either BD or ADHD, an idea which is consistent with the work of Strober et a! [23, 24] and Todd et a! , who proposed that pediatric BD might be a distinct subtype of BD with a high familial loading.
Like ADHD, conduct disorder (CD) is strongly associated with pediatric BD. This has been seen separately in studies of children with CD, ADHD and BD [8, 26-28]. These reports are consistent with the well-documented comorbidity between CD and major depression , considering that juvenile depression often leads to BD [30, 31].
Biederman et a! [32, 33] investigated the overlap between BD and CD in a consecutive sample of referred youth and in a sample of ADHD subjects to clarify its prevalence and correlates. They found a striking similarity in the features of BD regardless of comorbid CD. Both the comorbid and non-comorbid subjects with BD had high rates of major depression, anxiety disorders, oppositional disorder, and psychosis than CD and ADHD
children [32, 33]. In addition, BD comorbid with CD was associated with poorer functioning and an increased risk for psychiatric hospitalization . Subjects with both CD and BD also had a higher familial and personal risk for mood disorders than other CD subjects, who had a higher personal risk for antisocial personality disorder . These studies suggest that subjects who receive diagnoses of both CD and BD may have both disorders. Although more research is needed to clarify this issue, it raises the hope that some cases of delinquency may respond to mood stabilizers.
Although clinical lore has long attributed bipolar symptoms in children to trauma, there have been few systematic studies of this issue. Kessler et a! reported elevated rates of BD among adults and adolescents with post-traumatic stress disorder (PTSD)  and Helzer et a!  found high rates of BD among adults with PTSD. Neither of these studies determined if the BD was primary or secondary to the trauma. In a longitudinal study, Wozniak et a!  identified pediatric BD as an important antecedent for, rather than consequence of, traumatic life events. Although these findings need independent replication, they suggest that clinicians treating traumatized children should not dismiss severe irritability and mood lability as consequences of the trauma. These may indicate an underlying BD, and thus have implications for treatment.
Let us hope that Shulman et a!'s call for more research is heeded. As they suggest, the field needs more data about youth-onset BD to more firmly establish the fundamental phenomenologic differences with adult-onset BD. Ideally, such work would lead to developmentally sensitive diagnostic criteria which take into account differential symptom expression with age. New criteria are needed to better differentiate typical and atypical BD.
Studies of phenomenology will provide a firm nosologic foundation for further work examining the causes and correlates of BD. Ideally, studies from multiple domains (e.g., neuroimaging, genetics, family environment) will eventually clarify if putative clinical subtypes of BD have different causes or pathophysiologic correlates in the brain.
We also need much more treatment research in youth. The outcome measures used in treatment studies of youth BD should attend to the wide range of symptoms of mood dysregulation, aggression and ADHD symptoms. The high levels of comorbidity between BD and other disorders suggest that multiple treatment strategies may be needed. Given the severity of BD, designing such studies in an ethical manner presents a major challenge for clinical researchers.
Was this article helpful?