Prevalence and predictors of type 2 diabetes mellitus in people with bipolar disorder: a systematic review and meta-analysis.

Abstract

This systematic review and meta-analysis assessed the prevalence and predictors of type 2 diabetes mellitus (T2DM) in people with bipolar disorder. We also compared the prevalence of T2DM in people with bipolar disorder versus age- and gender-matched healthy controls. PubMed, EMBASE, PsycARTICLES, and CINAHL were searched from inception till October 23, 2014 using the medical subject headings terms bipolar disorder AND diabetes OR glucose. There was no language restriction. Observational studies including retrospective, cross-sectional, and prospective designs were eligible if they included participants with bipolar disorder diagnosed according to recognized diagnostic criteria (DSM or ICD). Nineteen studies were included (n = 18,060; 54.8% male). Two independent authors extracted data in accordance with the meta-analysis of observational studies in epidemiology guidelines and PRISMA statement. A random effects meta-analysis was utilized. The overall prevalence of T2DM was 9.4% (95% CI, 6.5%-12.7%). Compared with age- and gender-matched controls (n = 783,049; 48.7% male), people with bipolar disorder (n = 6,595; 48.6% male) had double the risk of T2DM (relative risk = 1.98; 95% CI, 1.6-2.4, P < .001). No significant moderators were found. In an exploratory regression analysis, the variance in T2DM prevalence in the background population was associated with the variance in T2DM prevalence in people with bipolar disorder (5 studies, n = 4,983) (r(2) = 0.85, t = 4.09, P = .03). T2DM is significantly more common in people with bipolar disorder than in healthy controls of similar age and sex. The current meta-analysis furthermore indicates that changes in food, built, and social environments are needed in order to curb the diabetes epidemic in this high-risk population.