Cross-sectional associations between device-measured physical activity and atrial fibrillation

Background Atrial fibrillation (AF) may lead to thrombosis and increased risk of stroke, myocardial infarction, heart failure, and premature death. AF increases substantially with age, and estimates indicate that one in three individuals ≥55 years in Europe will develop AF. Physical activity (PA) may contribute to the prevention of AF, as it improves overall cardiovascular health. However, the positive effect of PA on AF is mainly based on subjective self-reports of PA and the pattern of the association is still unclear. To our knowledge, few studies have so far examined the association between accelerometry-measured PA and AF. Purpose The aim of this study is to explore the association between accelerometry-measured PA and self-reported AF. Methods The seventh survey of a large population-based study was conducted in 2015-2016 and included 21083 men and women aged ≥40 years (65% attendance). Of these, a random sample of 5194 participants wore a triaxial accelerometer for at least ≥10 hours/day for ≥4 days to measure PA, in addition to having valid data on all covariates. Based on validated cut-points, the counts were averaged and categorized into moderate-to-vigorous PA (MVPA, ≥2690 counts/min). Steps were collected from accelerations on the vertical axis. AF was self-reported and dichotomized into no AF or current/previous AF. The associations between PA and AF were estimated by univariable and multivariable logistic regression analyses and are presented as odds ratio (OR) with 95% confidence intervals (CI). The multivariable model was adjusted for age, sex, body mass index, systolic blood pressure, alcohol, heart attack, and heart failure. The present study was approved by the Regional Committee of Health Research Ethics and was carried out in accordance with the Declaration of Helsinki. Results In an unadjusted model, the odds of having AF was reduced with 10% for every 10-minute increase in MVPA/day (OR 0.90, 95% CI 0.87-0.94), and 13% for every 1000 steps/day (OR 0.87, 95% CI 0.84-0.90) in a linear association. After adjustment, there was no significant association between MVPA and AF (OR 0.97, 95% CI 0.93-1.01), whereas the odds of having AF was reduced by 5% for every 1000 steps/day (OR 0.95, 95% CI 0.91-0.998) in a linear association. When stratified by sex, the odds of having AF was reduced by 10% for every 1000 steps/day in females (OR 0.90, 95% CI 0.84-0.97) in a linear association, whereas no statistical association between steps/day and AF was observed in males (OR 0.98, 95% CI 0.93-1.04), or between MVPA and AF in females (OR 0.94, 95% CI 0.87-1.02) or males (OR 0.99, 95% CI 0.94-1.03). Conclusion These results from a large population-based study showed that the odds of having AF decreased as the dose of PA increased, although not significantly in males. Further studies exploring longitudinal associations between PA and AF using objectively measured PA are warranted.