Several studies have demonstrated sex differences in risk of thromboembolism and death among patients with atrial fibrillation, but it is unclear to what extent these associations relate to actual physiological differences. To date, no study has investigated sex differences with concomitant control for lifestyle related factors known to influence stroke risk. We used data from the Danish Diet, Cancer and Health study, including 57,053 participants (52% female) aged 50-64 years. The study population for this study included the 2,895 patients (36% female) with incident atrial fibrillation after inclusion. Data were linked to outcomes identified using nationwide registries. Risk of thromboembolism and death according to female sex were analysed using Cox proportional hazards models. After a median follow-up of 5.0 years, 137 men and 62 women suffered a thromboembolic event, and 349 men and 151 women died. In a crude analysis, female sex was associated with a non-significant lower risk of thromboembolism (hazard ratio [HR] 0.82, 95% confidence interval [CI] 0.61-1.11). Adjustment for differences in antithrombotic therapy, relevant comorbidities and lifestyle did not change this association (HR 0.77, 95% CI 0.55-1.13). In the final model, female sex was associated with a lower risk of death (HR 0.65, 95% CI 0.51-0.84). The associations were similar in a sensitivity analysis of women not taking hormone replacement therapy, and the effect of hormone replacement therapy use within females was non-significant for both endpoints of thromboembolism and death. In conclusion, in a relatively young population of patients with atrial fibrillation, female sex was associated with a lower risk of thromboembolism and death.
Keywords: Atrial fibrillation; death; epidemiology; sex differences; stroke; thromboembolism.