The effect of alcohol consumption on substrate remodeling and ablation outcome of paroxysmal atrial fibrillation (PAF) remains unknown.

We performed circumferential pulmonary vein isolation (CPVI) and voltage mapping of left atrium (LA) during sinus rhythm in 122 consecutive patients with symptomatic PAF (age, 55.4±9.4 years; 73.8% men). Low‐voltage zones (LVZs) were semiquantitatively estimated and presented as low‐voltage index (LVI). Each patient's daily alcohol consumption history was recorded at baseline and classified into alcohol abstainers, moderate drinkers, and heavy drinkers based on the National Institute on Alcohol Abuse and Alcoholism definition. Follow‐up was ≥12 months for AF recurrence. Alcohol abstainers and moderate and heavy drinkers were 70 (57.4%), 13 (10.6%), and 39 (32.0%), respectively. In total, LVZs were observed in 44 patients (36.1%). Daily alcohol consumption independently predicted presence of LVZs (odds ratio [OR], 1.097; 95% confidence interval [CI], 1.001–1.203; P=0.047). During mean follow‐up of 20.9±5.9 months, 40 patients (35.1%) experienced AF recurrence. Success rate was 81.3%, 69.2%, and 35.1% in alcohol abstainers, moderate drinkers, and heavy drinkers, respectively (overall log rank, P<0.001). Multivariate analysis showed that both alcohol consumption and LVI were independent predictors of AF recurrence (hazard ratio [HR], 1.579; 95% CI, 1.085–2.298; P=0.017; HR, 2.188; 95% CI, 1.582–3.026; P<0.001, respectively). Furthermore, mediation analysis revealed that LVZs acted as a partial mediator in effect of alcohol consumption on AF ablation outcomes.

Daily alcohol consumption was associated with atrial remodelling, and heavy drinkers have substantial risk for AF recurrence after CPVI.