SLEEP DISTURBANCE HAS BEEN INCORPORATED AS AN INCLUSION CRITERION OF AFFECTIVE DISORDER AND IS CLEARLY A DEFINING FEATURE OF BIPOLAR depres-

SLEEP DISTURBANCE HAS BEEN INCORPORATED AS AN INCLUSION CRITERION OF AFFECTIVE DISORDER AND IS CLEARLY A DEFINING FEATURE OF BIPOLAR depression.1 However, current understanding of the relationship between mood and sleep remains quite limited, particularly in the pediatric population.2 In the present study we examine the prevalence and associated sleep characteristics of children with a bipolar mood disturbance behavioral profile. The predominant features of pediatric bipolar disorder (PBD) are elation, grandiosity, flight of ideas or racing thoughts, decreased need for sleep, and hypersexuality during manic states.3-4 The disorder has also been characterized by severe irritability and high levels of hyperactivity.5 Although the presence of mania in the pediatric population has been reported in earlier work,6 more recent evidence suggests that, unlike mania in adults, it is characterized by nonepisodic, chronic, mixed states of ultra rapid cycling or “ultradian” cycling (as many as 300 manic episodes per year or more) with severe irritability.7 Although some controversy remains in the literature regarding the diagnostic validity of PBD, there is increasing evidence supporting its legitimacy as a distinct diagnostic entity. Indeed, consistent with established guidelines,8 uniquely differentiating features, evidence of familiality, specific treatment responsivity, and unique course of the disorder have been identified.9-12 In fact, a National Institute of Mental Health (NIMH) roundtable discussion resulted in the statement that PBD is a valid diagnostic entity and can be diagnosed accurately in prepubertal children13 despite the complexity of symptomatology, differences in adult versus childhood presentation, degree of symptom overlap with other mental disorders, and even in the presence of developmentally appropriate behavior. Two studies have identified sleep characteristics associated with PBD. Rao et al14 used long-term clinical course to separate adolescents who were depressed at the time the sleep study was conducted into those who eventually developed bipolar disorder and those who continued to have unipolar depression. The bipolar group evidenced significantly more Stage 1 sleep than healthy controls. Both the bipolar group and the healthy controls had increased rapid eye movement (REM) latency, lower REM density, and less REM sleep than the unipolar group. Decreased need for sleep is a diagnostic criterion for PBD.1 Consistent with this criterion, Geller et al4 reported that 40% of children with mania, but only 6.2% of children with attention-deficit/hyperactivity disorder (ADHD) and 1.1% of community controls, presented with dramatically decreased need for sleep. Diagnostic uncertainty for PBD is compounded by high comorbidity with ADHD and to a lesser extent with conduct and anxiety disorders.4 These factors, along with the lack of epidemiologic research, lead to a relative paucity of information regarding the prevalence and age of onset for PBD.15 Studies have shown that approximately 7% of children seen at psychiatric facilities meet criteria for PBD.12 Furthermore, Strober16 found that as many as 20% of children diagnosed with unipolar depression went on to develop PBD later in life. Adult epidemiologic data from a recent large community sample of 127,000 people estimated the prevalence of bipolar disease between 3.4% and 3.7% using a mood disorder screening questionnaire, with only 1.4% reporting a formal diagnosis, a finding that is consistent with other studies.17 Evidence from a relatively small cohort of bipolar children examined by Biederman and colleagues9 suggested early onset (6.3 + 4.7 years) for PBD, and NIMH guidelines further suggest Correlates of Sleep and Pediatric Bipolar Disorder