Abstract Background Whilst endurance athletes are known to have an overall higher prevalence of atrial fibrillation (AF), the prevalence of other non-sustained and sustained atrial arrhythmias (AAs) and ventricular arrhythmias (VAs) remains poorly defined. Purpose We aimed to define the prevalence of non-sustained and sustained AAs (including AF) and VAs in a healthy cohort of young, lifelong and retired elite endurance athletes. We also sought to assess associations between arrhythmias, fitness and measures of cardiac structure. Methods Holter monitors in healthy endurance athletes (n=288) from an international cohort of elite endurance athletes (the Pro@Heart consortium) were analysed for the presence of sustained and non-sustained AAs and VAs. Athletes were sub-divided into young elite athletes (n=176) and former elite athletes (n=112) of which 60 were lifelong (≥2.5 hours/week) and 52 retired (<2.5 hours/week) athletes (determined by current high-intensity exercise activity). Each athlete underwent cardio-pulmonary exercise testing, echocardiography and contrast-enhanced cardiac MRI. Participants with known structural or electrical cardiac disease were excluded. Independent T-test / One-Way ANOVA (parametric data) or Mann-Whitney U / Kruskal-Wallis Tests (non-parametric data) assessed for differences between the groups. Results Young athletes (n=176, 70.5% male) were significantly fitter with a greater degree of cardiac remodelling as compared with lifelong (n=60, 63.3% male) and retired (n=52, 82.7% male) athletes (Figure 1). The prevalence of non-sustained AAs (young: 4.0% vs lifelong: 66.7% vs retired: 61.5%, p<0.001), sustained AAs (young: 0.6% vs lifelong: 8.3% vs retired: 15.4%, p<0.001) and non-sustained VAs (young: 1.7% vs lifelong: 18.3% vs retired: 23.1%, p<0.001) was higher in lifelong and retired athletes compared with young athletes (Figure 2). Non-sustained VAs were predominately monomorphic (85%), short and moderate rate (median duration 2.8 seconds and heart rate 154 beats per minute). No athletes developed sustained VAs and/or sudden cardiac arrest. Young athletes had less VAs that were multifocal in origin compared with lifelong and retired athletes (young: 29.5%, lifelong: 37.3%, retired: 53.5%, p=0.027). There was no difference in prevalence of any arrhythmia between lifelong and retired athletes. There were no significant correlations between arrhythmias and fitness or cardiac imaging parameters. Conclusion Healthy athletes have a significant prevalence of non-sustained atrial and ventricular arrhythmias and sustained atrial arrhythmias that is more prevalent in older athletes. In older athletes, the prevalence of arrhythmias is similar regardless of whether athletes are actively engaged in training or retired suggesting that pro-arrhythmic remodelling in athletes is sustained, not immediately reversible and may be minimally responsive to detraining.Arrhythmias and cardiac imagingArrhythmias amongst athlete groups