Cardiovascular exercise and burden of arrhythmia in patients with atrial fibrillation - A randomized controlled trial



Physical activity at moderate-high intensity is recommended to prevent lifestyle diseases. Patients with atrial fibrillation are at risk of a sedentary lifestyle due to fear of exercise-induced episodes of atrial fibrillation. The burden of arrhythmia can be reduced by physical exercise. The effect of exercise intensity on burden of atrial fibrillation needs to be studied further.

Methods and results

In a 12-week randomized controlled trial, 76 patients with paroxysmal/persistent atrial fibrillation were allocated to perform exercise at either low intensity or high intensity (50% and 80% of maximal perceived exertion, respectively). Primary outcome was burden of AF measured by daily electrocardiography-reporting during 12 weeks. Secondarily, change in maximal oxygen uptake (peak VO2) and 1-year hospitalization was compared between low and high intensity exercise. Sixty-three patients completed the follow-up. In the intention-to-treat analysis, we found no statistical difference in burden of atrial fibrillation between low and high intensity exercise (incidence rate ratio 0.742, 95% CI 0.29–1.91, P = 0.538). No serious adverse events were reported and there was no difference in hospitalization between the two exercise groups. Both exercise groups improved significantly in peak VO2 (low intensity: 3.62 mL O2/kg/min, SD 3.77; high intensity: 2.87 mL O2/kg/min, SD 4.98), with no statistical difference between-groups (mean difference: 0.76 mL O2/kg/min, 95% CI -3.22–1.7).


High intensity physical exercise was not superior to low intensity physical exercise in reducing burden of atrial fibrillation. HI exercise was well tolerated; no evidence of an increased risk was found for HI compared to LI exercise. Larger studies are required to further prove our findings.

Trial registration NCT01817998 


Autor / Fonte:Ane Katrine Skielboe, Thomas Quaade Bandholm, Stine Hakmann, Malene Mourier, Thomas Kallemose, Ulrik Dixen PloS One 2017, 12 (2): e0170060