Atrial fibrillation and exercise capacity in patients with and without heart failure during exercise-based cardiac rehabilitation

Abstract Background/Introduction Atrial fibrillation (AF) impairs exercise tolerance and exacerbates hemodynamic limitations in cardiac patients. The coexistence of AF and heart failure (HF) further compromises hemodynamic function during exercise. The impact of AF on changes in peak exercise aerobic power (VO2peak) and aerobic endurance (6 minute walk distance test, 6MWT) after completing an exercise-based cardiac rehabilitation (EBCR) remains unclear. Purpose This study examined the association between AF and VO2peak in patients with and without AF, both overall and within an HF subgroup. Methods This retrospective single-center cohort study included patients who underwent EBCR at a specialized in-patient rehabilitation center between December 2022 and June 2024. Patients were eligible if they completed a 6MWT at both admission and discharge. Estimated VO₂peak (ml/min/kg) was derived from maximal power output and METs obtained during exercise stress test and adjusted using spiroergometry results in a subset of patients. The corresponding regression equation VO₂/kgₑstcor = (0. 0446 * 6MWT) – 1. 8981 was applied for further analyses. HF was defined as a left ventricular ejection fraction (LVEF) 50% at admission. To analyze the association between AF and VO2, we performed linear regression adjusted for age and sex. Results A total of 744 patients were analyzed, including 172 (23%) with a history of AF. The overall median age was 65 years (57–72 years), 78% male, and the median length of stay was 20 days (20–27 days). Median left ventricular ejection fraction (LV-EF) was 55. 0% (47–60%) in patients without AF and 52. 0% (45–59) with AF. Among 203 HF patients, 53/203 (26%) had AF. At admission, estimated VO₂peak tended to be lower among patients with AF (14. 2 ml/kg/min 10-17) compared to those without AF (15. 6 ml/min/kg 12–20) (Figure 1), and a similar trend was seen in HF patients with AF (14. 2 ml/kg/min 9-17) vs those without AF (16. 2 ml/min/kg 11–20). All groups improved VO2peak to a similar extent over the course of CBCR, such that at discharge, estimated VO₂peak in patients without HF and AF was 21. 9 ml/min/kg (18–25; median difference: 5. 9 ml/kg/min 4–8) and 21. 4 ml/min/kg (18–25; median difference: 7. 0 ml/kg/min 5–10) in AF patients. In HF patients, discharge VO₂ was 21. 9 ml/min/kg (17–25; median difference: 5. 5 ml/kg/min 3–8) and 19. 5 ml/min/kg (16–25; median difference: 6. 3 ml/kg/min 4–9) in patients with HF and AF. AF was associated with a significantly lower VO₂ at admission (-0. 97 ml/kg/min, p=0. 039), showed no significant effect at discharge (p=0. 499) and was linked to a greater VO₂ improvement during rehabilitation (+1. 25 ml/kg/min, p0. 001). Conclusion Patients with AF tend to have lower estimated VO₂peak at admission, both in the overall cohort and in HF patients. All groups showed clinically relevant improvements in exercise capacity over the course of EBCR, with slightly greater progress in AF patients. Estimated VO₂peak violin plots