Risk of incident atrial fibrillation detected on implantable loop recorder and relationship with physical activity; an individual participant pooled analysis of LOOP and VENTOUX

Background Moderate levels of physical activity have a protective effect against development of atrial fibrillation (AF)(1) but some studies have suggested paradoxically increased risk of AF in those with high levels of physical activity(2). However, previous studies have reported physical activity subjectively and only included participants with symptomatic AF(5). Purpose We aimed to investigate the relationship between physical activity measured by accelerometer and occurrence of symptomatic and asymptomatic AF detected by implantable loop recorder (ILR). Methods This analysis included VENTOUX study participants (male veteran cyclists/triathletes, n=106, age 59±4) and all male LOOP study participants (asymptomatic with one stroke risk factor n=758, age 75±6). Participants of both studies had no history or symptoms of AF and underwent clinical assessment and ILR implantation. The primary endpoint was time to onset of first AF episode lasting ≥6 minutes occurring within the first 3 years. Participants were contacted after detection of AF to assess for the presence of symptoms. Physical activity duration was assessed as an average hours per day via accelerometer data from ILR over the period of follow up. Cox regression analysis was performed with adjustment for CHARGE-AF score. Results Over the follow up period 269/864 (31%) participants developed AF on ILR, 20/269 (7%) of whom were symptomatic. After adjustment for CHARGE-AF score, athletes were not more likely to develop AF (hazard ratio (HR) 1.138, 95% confidence interval (CI) 0.566-2.286, p=0.717). Physical activity (hours/day) was not associated with increased risk of AF (HR 0.937, 95% CI 0.835-1.050, p=0.263). Only lower resting heart rate was associated with AF (HR 0.983, 0.973-0.993, p<0.001). In those who developed AF, factors associated with being symptomatic included increased physical activity, younger age and absence of hypertension. Participants who did 4+ hours per day of physical activity were more likely to be symptomatic from AF compared to individuals who only did 0-1 hours of physical activity (odds ratio 9.29, 95% CI 1.55-55.78, p=0.015). Conclusion After adjustment for baseline risk factors and co-morbidities physical activity was not associated with incident AF detected on ILR. However, individuals with high levels of physical activity when compared to sedentary individuals were more likely to be symptomatic at the onset of AF.