Risk of stroke among individuals on selective serotonin reuptake inhibitors (SSRIs) versus non-SSRI antidepressants: a population-based retrospective cohort study and meta-analysis

Despite the presence of studies indicating a potential elevated risk of stroke associated with selective serotonin reuptake inhibitors (SSRIs), current evidence is inconclusive. This study aims to evaluate the stroke risk associated with SSRI use using non-SSRI antidepressants as comparators through retrospective cohort study and meta-analysis of observational studies. We extracted data from a territory-wide public healthcare database in Hong Kong to conduct a retrospective cohort study of patients aged 18+ years who initiated SSRI or non-SSRI antidepressants between January 2018 and April 2024. We estimated incidence rate ratios (IRRs) using weighted Poisson regression with robust variance estimation and hazard ratios (HRs) using weighted Cox proportional hazards models, comparing SSRI versus non-SSRI initiators. We subsequently conducted a systematic review and meta-analysis. Quality of studies was assessed using the Newcastle-Ottawa Scale. 124,854 individuals were included in the cohort study, among which 56,279 were SSRI users. SSRI users had an adjusted HR of 0.95 (95% CI 0.77–1.17) for stroke compared to non-SSRI users, suggesting a non-significant lower risk of stroke. Findings were consistent across subgroups by stroke types (i.e. ischemic stroke and hemorrhagic stroke). The result of our cohort study was aggregated with 5 other observational studies, and a pooled RR was estimated (RR 0.94, 95% CI 0.83–1.06). Our findings suggested that compared with non-SSRI antidepressants, SSRIs are not associated with a higher risk of stroke based on all available observational data. • Population-based cohort compared SSRIs with non-SSRI antidepressants using an active comparator design. • SSRIs were not associated with an increased risk of overall stroke versus non-SSRI antidepressants. • Findings were consistent across ischemic and hemorrhagic stroke subtypes. • Results were robust after inverse probability weighting and competing risk analyses. • Meta-analysis of all eligible observational studies confirmed the null association.