Alcohol Consumption and Atrial Fibrillation Risk: An Overview of Systematic Reviews and Network Meta‐Analysis

ISSUES: Although alcohol consumption is linked to atrial fibrillation (AF), the relationship across different intake levels and between sexes remains unclear. This study presents the first network meta-analysis of prospective cohort studies bringing greater precision to these associations. APPROACH: A systematic review identified five meta-analyses on alcohol and AF risk. From these, 13 cohort studies totalling over 80 million person-years were included in a random-effects network meta-analysis, including sex-stratified analyses. KEY FINDINGS: Compared to low-level consumption (< 12 g/day), moderate intake (12-< 24 g/day) slightly increased AF risk (hazard ratio [HR] = 1.07; 95% confidence interval [CI] 1.04-1.10), similar at 24-< 36 g/day (HR = 1.09; 95% CI 1.00-1.20). No significant risk increase was observed for 36-< 60 g/day. Heavy consumption (≥ 60 g/day) showed the highest risk (HR = 2.84; 95% CI 1.57-5.14). Non-drinkers ('Former', 'Never' or 'Occasional') had HRs near 1, except 'None', which showed a slight increase (HR = 1.08; 95% CI 1.04-1.11). IMPLICATIONS: In males, moderate consumption increased AF risk slightly, while heavy intake had a more pronounced effect (HR = 1.49; 95% CI 1.22-1.81). In females, moderate intake had no significant effect, but heavy intake significantly increased risk (HR = 2.53; 95% CI 1.05-6.08). CONCLUSION: This network meta-analysis shows a nonlinear relationship between alcohol consumption and AF risk. Low-level or occasional intake poses the lowest risk. In males, moderate consumption slightly increases AF risk, while in females, risk rises substantially only with heavy intake. These findings support limiting alcohol consumption to reduce AF risk and highlight the need for further sex-stratified studies and consideration of sex-specific recommendations. REGISTRATION: This network meta-analysis is a split review of an overview of reviews examining alcohol-attributable health and social harms, registered under PROSPERO ID442338.