Abstract
BACKGROUND: This study aimed to investigate the dose-response association between alcohol consumption and atrial fibrillation (AF) risk. METHODS: PubMed, Embase, Cochrane Library, and Web of Science were systematically searched using keywords related to alcohol and AF from the establishment of databases up to 1 March 2021. Prospective studies examining the impact of alcohol on the risk of AF with hazard ratios (HRs) were included. Restricted cubic spline regression was performed to quantify the dose-response relationship between alcohol consumption and AF risk. RESULTS: Thirteen eligible studies were included in the meta-analysis, with a total of 645,826 participants and 23,079 cases of AF. When compared with non-/seldom-drinkers, the pooled adjusted HRs of AF were 1.30 (95% confidence interval [CI]: 1.20-1.41) and 1.00 (95% CI: 0.96-1.05) for high and low alcohol consumption, respectively. Moderate alcohol intake significantly increased the risk of AF in males (HR, 1.21; 95% CI: 1.10-1.33) but not in females (HR, 1.02; 95% CI: 0.91-1.14). The cubic spline regression analysis illustrated that the risk of AF significantly increased with daily alcohol intake in a Non-linear manner (R(2) = 0.64, P = 5.785 × 10(-12)). CONCLUSION: This study revealed a Non-linearly positive association between alcohol intake and the risk of AF. Low alcohol intake was not associated with the development of AF, whereas moderate alcohol intake significantly increased the risk of AF in males but not in females. Our meta-analysis highlighted that alcohol consumption should be restricted to a low level to reduce the risk of AF.