Prevalence of DSM-5 mixed features: A meta-analysis and systematic review

The definition of mixed features by the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) remains controversal; however, there has been no systematic review of the prevalence of DSM-5 mixed features. We conducted a meta-analysis and systematic review to examine the prevalence of DSM-5-defined mixed features in major depressive episodes (MDE) and manic/hypomanic episodes.

We systematically searched all literature types (i.e., observational, cross-sectional, cohort, retrospective chart review, and post-hoc analysis) in electronic databases including MEDLINE, Embase, CINAHL, PsycINFO, and Web of Science from 2013 to 2020.

A total of 17 studies with 20 samples were selected. The pooled prevalences of the mixed features in MDE and manic/hypomanic episodes were 11.6% (95% confidence interval [CI] = 7.9-16.7%) and 26.8 (95% CI = 17.0-39.5%), respectively. The prevalence of mixed features during major depressive disorder in East Asian countries was the lowest, which ranged from 0-2.2%. The subgroup analysis did not identify any influential factors for substantial heterogeneity. Most of the individual studies demonstrated moderate to high risk of bias.

Despite the increasing attention and controversy surrounding DSM-5-defined mixed features, few studies have systematically estimated the prevalence. Future studies with appropriate design and sample sizes should measure the prevalence of mixed features during MDE and manic/hypomanic episodes.