Childhood subclinical phenotypes have been informative for etiological research and as a target for preventative interventions. Using a prospective longitudinal general population cohort we investigated whether childhood manic symptoms predicted a diagnosis of bipolar disorder (BD) or other psychiatric disorders by early adulthood.

Subthreshold manic symptoms at age 11 years (n = 1907) and clinical outcomes by age 19 years (n = 1584) were ascertained in the TRacking Adolescents’ Individual Lives Survey (TRAILS), a prospective Dutch community cohort. We used latent class analysis to stratify TRAILS participants at age 11 years into distinct classes based on the pattern and severity of childhood manic symptoms. We then determined the association between class membership and clinical diagnoses by age 19 years.

At age 11 years, we identified a normative class with negligible symptoms (n = 862), a mildly symptomatic (n = 846) and a highly symptomatic class (n = 199). The risk of BD was moderately increased in individuals in the mildly symptomatic class (OR = 2.65, 95% CI 1.41–5.01), and substantially increased in the highly symptomatic class (OR = 7.08, 95% CI = 3.32–15.11). Children in the highly symptomatic class were additionally characterized by lower IQ and socioeconomic status, greater family dysfunction and increased rates of parental psychiatric morbidity. Class membership did not show significant associations with depressive, anxiety and substance abuse disorders by age 19 years.

The results provide support to developmental models of BD, and suggest that manic symptoms in childhood may be a marker for adult disorders and therefore potentially useful for early identification of at risk individuals.

The author has not supplied his declaration of competing interest.