(P-006) Stroke Risk Associated with Incident Antidepressant Use: Delineated with Age

Background: Most patients who start antidepressants treatment are at young to middle-age. Although the association between antidepressant use and stroke risk has been investigated in past decades, its detailed information remained insufficient such as risk variations among different age groups and antidepressant categories.

Objectives: To assess the stroke risks associated with antidepressants among different age groups and antidepressant categories.

Methods: 1. Design and setting: We enrolled the incident antidepressants users (≥18 years) for the cohort study, who were identified from the Taiwan National Health and Welfare Insurance database from 1 July 2010 to 30 June 2015. The incident users were those who did not receive any antidepressants during the 36 months prior to their first antidepressants prescription. Propensity score was calculated according to age, gender, comorbidities, and co-medications for each patient, and subsequently utilized to match the antidepressant users with their control at a ratio of 1:1. Adult patients who received antidepressants were grouped into young adults ( <45 years old), adults (45~64 years old), and elderly (≥ 65 years old).
2. Statistical analysis: In the main analysis, the observation period was set as a 180-day time from initial antidepressants treatment, and Cox proportional hazards regression model was used to explore the association between antidepressant use and stroke risk at different age groups.

Results: We included 1,072,922 antidepressants users and 1,072,922-matched non-users, both had similar baseline characteristics after propensity score match method. Among young adults, the stroke risk was 3.68 times higher (95% CI, 2.56 to 5.29) in antidepressants users than non-users, and the risks of adults and elderly groups reduced to 3.08 times (95% CI, 2.65 to 5.38) and 2.42 times (95% CI, 2.22 to 2.63), respectively. Moreover, in young adults, the risk among TCA users was 4.14 times (95% CI, 2.39-7.15) higher than non-users, while no differences were seen with SSRI and SNRI. The stroke risks associated with SNRI were relatively high among adults (HR, 4.14; 95% CI, 2.24-7.67) and elderly (HR, 2.90; 95% CI, 1.89-4.47) groups.

Conclusions: For the initial prescribing of antidepressants, physicians are suggested to consider risk differences among age groups and various antidepressant categories. Meanwhile, close monitoring of stroke risk is warranted.