Two-Year Prospective Follow-Up of Children With a Prepubertal and Early Adolescent Bipolar Disorder Phenotype

OBJECTIVE: Longitudinal outcomes of bipolar disorder with onset in the late teenage years or in adulthood have been reported, but little is known about the natural history of childhood-onset mania. This study sought to provide rates and predictors of recovery and relapse in children with a prepubertal and early adolescent bipolar disorder phenotype. METHOD: Eighty-nine consecutively ascertained outpatient subjects (mean age=10.9 years [SD=2.7]) received comprehensive research assessments, including separate interviews of mothers about their children and of children about themselves, at baseline and at 6, 12, 18, and 24 months after baseline. The study phenotype required DSM-IV mania with elation and/or grandiosity as one criterion to distinguish the study phenotype from a diagnosis of mania based on criteria overlapping with those for attention deficit hyperactivity disorder and to ensure that subjects had at least one of the two cardinal features of mania (i.e., elation and/or grandiosity). Subjects were treated by their own community practitioners. RESULTS: The proportions of subjects who recovered from mania and who relapsed after recovery were 65.2% and 55.2%, respectively. The mean time to recovery was 36.0 weeks (SD=25.0). Relapse occurred after a mean of 28.6 weeks (SD=13.2). Living with an intact biological family significantly predicted rate of recovery, and a low level of maternal warmth significantly predicted rate of relapse. CONCLUSIONS: The relatively poor outcomes of these subjects may be related to their phenotypic resemblance to severely ill adults with bipolar disorder who have mixed mania, continuous rapid cycling, psychosis, and treatment-resistant psychopathology. A lower level of effectiveness of mood stabilizers in children cannot be ruled out. Although the significance of maternal warmth as a predictor is consistent with reports in adult mania, the significance of intact family as a predictor may be unique to childhood mania.