Meta-analysis of the interval between the onset and management of bipolar disorder

Objective: To evaluate the length of the interval between the onset and the initial management of bipolar disorder (BD).

Method: We conducted a meta-analysis using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. Systematic searches located studies reporting estimates of the age of onset (AOO) and indicators of the age at initial management of BD. We calculated a pooled estimate of the interval between AOO and age at management. Factors influencing between-study heterogeneity were investigated using sensitivity analyses, meta-regression, and multiple meta-regression.

Results: Twenty-seven studies, reporting 51 samples and a total of 9415 patients, met the inclusion criteria. The pooled estimate for the interval between the onset of BD and its management was 5.8 years (standardized difference,.53; 95% confidence interval,.45 to.62). There was very high between-sample heterogeneity (I ² = 92.6; Q = 672). A longer interval was found in studies that defined the onset according to the first episode (compared to onset of symptoms or illness) and defined management as age at diagnosis (rather than first treatment or first hospitalization). A longer interval was reported among more recently published studies, among studies that used a systematic method to establish the chronology of illness, among studies with a smaller proportion of bipolar I patients, and among studies with an earlier mean AOO.

Conclusions: There is currently little consistency in the way researchers report the AOO and initial management of BD. However, the large interval between onset and management of BD presents an opportunity for earlier intervention.