Abstract
BACKGROUND: Previous studies from Western populations have reported changing temporal trends in ischemic stroke (IS) incidence in females with atrial fibrillation (AF) when compared with males. Nationwide data on such temporal trends in AF-related IS incidence by sex are limited in Asian populations. METHODS: This population-based retrospective cohort study included patients with incident AF diagnosed between 2005 and 2016 from the Korean National Health Insurance Service. Patients with valvular heart disease, prior IS, or anticoagulant use were excluded. Incidence rates (IRs) per 100 person-years and hazard ratios (HRs) for IS were calculated by Fine and Gray competing risk regression. FINDINGS: After exclusions, 290,081 females (mean age: 64.4 years, SD 16.3) and 338,100 males (mean age: 60.1 years, SD 14.9) were included. The mean follow-up duration was 5.7 (SD 4.1) years. At baseline, the CHA(2)DS(2)-VA scores were higher in females than in males (2.0 vs. 1.6, P < 0.0001). IRs for IS declined over time in both sexes (P for trend < 0.0001). The IS incidence in females compared to males was significantly higher in 2005-2006 (1.55 vs. 1.40; HR(unadj): 1.12, 95% confidence interval: 1.06-1.19); however, it was no longer significant in 2015-2016 (1.20 vs. 1.17; HR(unadj): 1.03, 95% confidence interval: 0.99-1.08). The reduction in relative risk primarily originated from the subgroup with CHA(2)DS(2)-VA scores 0-1. Females with CHA(2)DS(2)-VA scores ≥3 consistently showed higher IRs for IS compared to males regardless of adjustment. INTERPRETATION: Sex differences in IS incidence decreased over calendar-year intervals, mainly in low-risk patients with AF. The persistently high IS incidence in high-risk females with AF suggests that sex still remains an important risk modifier. FUNDING: Patient-Centered Clinical Research Coordinating Center, Republic of Korea.