In order to investigate whether or not pediatric bipolar disorder (PBD) is becoming more prevalent and whether or not rates vary by country, we meta-analyzed all published epidemiological studies that report pediatric mania or bipolar disorder.

Prevalence rates were recalculated from reported N of cases with bipolar disorder, then logit transformed before calculating inverse variance-weighted averages.

Eleven studies enrolled 10, 896 youths between the ages of 7 and 21 years during a period from 1985 to 2005. Six of the samples were from the USA; five were from other countries (the Netherlands, Spain, Mexico, Ireland, and New Zealand).

The mean prevalence rate of PBD, using a fixed effects model, was 3.6 per 100 (95% confidence interval: 3.2 to 4.0). Given the significant heterogeneity between studies, Q=130.38, df =10, p< .00005, a random effects model provided a more conservative estimate: 1.9 (95% CI: 1.2 to 3.2) per 100.

Surprisingly, the correlation between year of enrollment and prevalence of PBD was negative (r=-.38) and became nonsignificant when controlling for study design differences, contradicting the hypothesis that the rise in clinical diagnoses is attributable to an increase in general population prevalence. Results also challenge the common belief that rates of bipolar are higher in the USA than in other countries. Higher rates in USA samples can be attributed to a tendency to use broader, non-DSM definitions of PBD, (p< .0003); rates are otherwise homogeneous across countries. PBD appears to affect roughly 2 per 100 youths in nonclinical samples around the world.