Antidepressants have complex associations with longitudinal depressive burden in bipolar disorder

Abstract Aims Antidepressants are common in bipolar disorder (BD), but controversial due to questionable efficacy/tolerability. We assessed baseline antidepressant use/depression associations in BD. Methods Stanford BD Clinic outpatients, enrolled during 2000–2011, assessed with the Systematic Treatment Enhancement Program for BD (STEP-BD) Affective Disorders Evaluation, were monitored up to two years with the STEP-BD Clinical Monitoring Form while receiving naturalistic expert treatment. Prevalence/correlates of baseline antidepressant use in recovered (euthymic ≥8 weeks)/depressed patients were assessed. Kaplan–Meier survival analyses assessed times to depressive recurrence/recovery in patients with/without baseline antidepressant use, and Cox Proportional Hazard regression analyses assessed covariate effects. Results Baseline antidepressant use was significantly (albeit without Bonferroni multiple comparison correction) less among 105 recovered (31.4%) versus 153 depressed (44.4%) patients, and among recovered patients (again without Bonferroni correction), associated with Caucasian race, earlier onset, worse Clinical Global Impression scores, and hastened depressive recurrence (only if mood elevation episodes were not censored), driven by lifetime anxiety disorder, and more (even with Bonferroni correction) bipolar II disorder, lifetime anxiety and eating disorders, and core psychotropics. Baseline antidepressant use among depressed patients was associated with significantly (again without Bonferroni correction) older age, female gender, and more (even with Bonferroni correction) anxiolytics/hypnotics, complex pharmacotherapy, and core psychotropics, but no other unfavorable illness characteristic/current mood symptom, and not time to depressive recovery. Limitations Tertiary BD clinic referral sample receiving open naturalistic expert treatment. Analyses without/with Bonferroni correction. Conclusions Additional research is required to assess the complex associations between baseline antidepressant use and longitudinal depressive burden in BD.