Clinical correlates of DSM-5 mixed features in bipolar disorder: A meta-analysis

Background: Systematic data on clinical correlates of mixed features in bipolar disorder are not available, so far. We conducted a systematic review and meta-analysis estimating the association between DSM-5 mixed features and candidate characteristics in depressive and manic/hypomanic episodes. Methods: We included observational studies indexed in the main electronic databases. The association between DSM-5 mixed features and relevant correlates was estimated using odds ratio (OR) and standardized mean difference (SMD) with 95% confidence intervals (CIs), for categorical and continuous variables, respectively. Analyses were based on random effects models. Results: Eight studies were included, involving 3070 individuals (1495 with a major depressive episode and 1575 with hypo/manic episode). No clinical characteristics were associated with mixed features in subjects with a depressive episode. Among subjects with a manic/hypomanic episode, those with mixed features were more likely to have a history of suicide attempts (OR: 2.37; 95%CI: 1.42 to 3.94; I-2=39.7%), co-occurring anxiety disorders (OR: 2.67; 95%CI: 1.28 to 5.57; I-2=0%), and a rapid cycling course (OR=4.23; 95%CI: 1.29 to 13.81; I-2=0%), with less severe manic symptoms (SMD=-0.40; 95%CI: -0.65 to -0.16; I-2=0%). Limitations: (1) the heterogeneity of methods across studies and the inconsistency of findings; (2) the limited amount of data on correlates of DSM-5 mixed features; (3) the possible influence of publication bias. Conclusions: Findings of this meta-analysis show that mixed features among individuals with a manic/hypomanic episode may identify a special clinical population, characterized not only by depressive symptoms, but also by anxiety, rapid cycling, and suicidality.