Using Atrial Cardiomyopathy Metrics to Characterize Atrial Fibrillation in Endurance Athletes

BackgroundAtrial fibrillation (AF) is more prevalent in endurance athletes (EAs) than in the general population, but identifying athletes at greatest risk remains challenging. ObjectivesThis study sought to develop an AF risk assessment tool from associations between AF and electrophysiologic, imaging, and genetic markers of atrial cardiomyopathy in EAs. MethodsEAs and nonathletic control subjects (NAs) underwent comprehensive phenotyping with the use of cardiopulmonary exercise testing, echocardiography, 12-lead electrocardiography (ECG), Holter monitoring for premature atrial contractions (PACs), and biochemical analysis. A genotyping array was used to determine polygenic risk scores for AF (AF-PRS). Logistic regression was used to develop an AF probability model. ResultsEAs (n = 353; median age 41 years, 74% male) had 40% larger LA volumes (42.9 vs 29.7 mL/m2; P < 0.001) than NAs (n = 97; median age 34 years, 64% male). Of the EAs, 96 (27%) had prevalent AF and demonstrated larger LA volumes (46.2 mL/m2 vs 41.9 mL/m2; P < 0.001), reduced LA strain (reservoir 26% vs 32%; P < 0.001), prolonged P-wave duration (124 vs 111 ms; P < 0.001), abnormal P-wave terminal force velocity in lead V1 (26% vs 9%; P < 0.001), and increased PACs (118 vs 15 PACs/24 h; P < 0.001) than EAs without AF. A clinical risk model incorporating these factors demonstrated excellent discrimination (AUC = 0.84). High AF-PRS increased AF odds 2.56-fold but contributed little additional predictive value. Using the model, AF prevalence increased from 4% to 86% in low to very high risk athletes, respectively. ConclusionsA combination of atrial cardiomyopathy markers was associated with high AF prevalence in EAs. A resulting quantitative AF risk identification tool provides promise for interventions aimed at preventing AF and thromboembolic complications in EAs.