Association of Type of Antidepressant Initiation with Bleeding Risk in Atrial Fibrillation Patients Taking Oral Anticoagulants

Background: Inconsistent evidence suggests that use of certain antidepressants, particularly selective serotonin reuptake inhibitors (SSRIs), in patients using oral anticoagulants (OACs) might be associated with an elevated risk of bleeding. Objective: This study aims to investigate the risk of bleeding associated with initiation of different types of antidepressants among atrial fibrillation (AF) patients on OAC therapy. Patients and methods: A total of 30,336 AF patients (mean age 72.2 years; 54% female) on OAC therapy that started antidepressant treatment were identified from the Truven Health Analytics MarketScan Commercial and Medicare Databases for the period 2007–2015. Exposure was defined as filling a prescription for antidepressant, and categorized as SSRI, serotonin/norepinephrine reuptake inhibitors (SNRIs), serotonin reuptake inhibitors (SRIs), tricyclic antidepressants (TCAs), or other antidepressants. The primary outcome was incident hospitalized bleeding. Associations of antidepressant type with bleeding were assessed calculating hazard ratios (HRs) and 95% confidence intervals (CIs) with adjusted Cox models in pairwise propensity score-matched cohorts. Results: During a mean follow-up of 21 months, we identified 1612 bleeding episodes. In pairwise comparisons, SSRI use was associated with an increased risk of bleeding when compared to most other antidepressants (HR 1.22, 95% CI 0.96–1.54 vs SNRI; HR 1.10, 95% CI 0.90–1.35 vs SRI; HR 1.03, 95% CI 0.82–1.30 vs TCA). SNRI use was associated with the lowest bleeding risk. Results did not differ by OAC type, age, and sex. Conclusions: Among AF patients on OAC initiating antidepressants, risk of bleeding varied across antidepressant type. This information can inform treatment choices among patients receiving OAC.