Assess prevalence and clinical relevance of 'with mixed features” using a more liberal (2 opposite pole symptoms) compared to the more conservative DSM-5 (3 opposite pole symptoms) threshold in depressed bipolar disorder (BD) patients.

BD outpatients were assessed with the Systematic Treatment Enhancement Program for BD (STEP-BD) Affective Disorders Evaluation. Prevalence and clinical correlates of baseline depressive episodes 'with mixed features” were compared using a more liberal threshold and the more conservative DSM-5 threshold.

Among 503 BD patients, 151 (30.8%) had baseline syndromal major depressive episodes, among whom 'with mixed features” occurred in 22.5% (34/151) using a more liberal threshold, but in only 9.9% (15/151) using the more conservative DSM-5 threshold. Hence, the rate of 'with mixed features” for depressive episodes using the more liberal compared to the conservative threshold was more than twice (2.3 times) as high (Chi-square=8.8, p=0.004). Moreover, the more liberal threshold yielded more important statistically significant clinical correlates of 'with mixed features” compared to pure depressive episodes, which were not significant using the more conservative DSM-5 threshold. Specifically, using the more liberal threshold, mixed compared to pure depression was associated with more anxiety disorder comorbidity and alcohol use disorder comorbidity, and less antidepressant use.

Further studies are warranted to assess our preliminary observation that a more liberal 'with mixed features” threshold for bipolar depression may be more inclusive and have more clinical relevance.