RESUMEN
BACKGROUND: While factors associated with mixed states have been extensively studied, data are scant regarding the clinical heterogeneity of mixed states. The EMBLEM study was a prospective, observational study on patients with manic and mixed states. We describe and compare baseline characteristics and 24-month clinical course of mixed states with predominant depressive symptoms (MSDS) and mixed states with predominant manic symptoms (MSMS). METHODS: Adult inpatients/outpatients with bipolar disorder were enrolled within the standard course of care if they initiated or changed oral medication for acute mania or mixed states. A logistic regression was used to identify the baseline factors associated with each polarity. Comparisons with mixed episode without symptom predominance (OMS) were performed for informational purpose only. RESULTS: About 573 mixed patients were analyzed (23.7% of the cohort): 59.5% (n=341) had MSMS, 11.9% (n=68) had MSDS, and 28.6% (n=164) had OMS. At baseline, hallucinations/delusions during the index episode, inpatient status, high CGI-BP overall score, and low education level were more often associated with MSMS versus MSDS. Alcohol abuse or dependence and selective serotonin reuptake inhibitor (SSRI) or benzodiazepine use at inclusion were significantly more frequent with MSDS. MSDS had a significantly lower 24-month recurrence rate than MSMS; MSMS experienced more switches to mania whereas MSDS switched more to depression. LIMITATIONS: The post hoc dimensional definitions in the study require caution in the interpretation of the results. CONCLUSION: These results present evidence of clinical heterogeneity within mixed states. Predominant manic or depressive symptoms within mixed episode could influence clinicians' decisions in term of hospitalization, treatment, and perception of bipolar severity.