Prevalence rates and clinical implications of bipolar disorder "with mixed features" as defined by DSM-5

Background We investigated the increase in the prevalence of bipolar disorder with mixed features following the replacement of DSM-IV-TR criteria with DSM-5 criteria. Additionally, we examined the clinical implications of the use of "with mixed features" as a specifier with bipolar disorder.

Method We retrospectively reviewed medical charts from 2003 to 2013. A total of 331 patients diagnosed with bipolar disorder using the DSM-IV TR were enrolled and categorized into four groups: manic/hypomanic with mixed features, manic/hypomanic without mixed features, depressed with mixed features, and depressed without mixed features. These classifications were made in accordance with the DSM-5 definition of bipolar disorder "with mixed features." Changes in the prevalence, demographic and clinical characteristics were compared among the groups.

Results The prevalence rates of mixed features were significantly different when using the DSM-5 criteria vs. the DSM-IV-TR criteria. Patients with mixed features had a younger age of onset, younger age at hospitalization, more frequent hospitalizations for mixed episodes, and greater suicide risk compared with patients without mixed features. Limitations Retrospective study may have resulted in under diagnosis of mixed states.

Conclusions An approximately three-fold greater risk for mixed features was observed in patients with bipolar disorder when using the DSM-5 criteria than when using the DSM-IV-TR criteria. The additional patients may represent patients with sub-syndromal mixed features and could indicate that patients with mixed features are underdiagnosed.