Trajectory change of left ventricular ejection fraction after rhythm control for atrial fibrillation in heart failure

Abstract Aims Rhythm control therapy has shown great benefits for patients with atrial fibrillation (AF) and heart failure (HF). However, few studies have evaluated the effects of rhythm control on left ventricular ejection fraction (LVEF) trajectory across the whole HF spectrum. Our study explored the prevalence and predictors of LVEF trajectory changes and their prognostic implications following rhythm control. Methods and results Depending on the treatment strategy, the cohort was classified into rhythm and rate control groups. Alterations in HF types and LVEF trajectory were recorded. The observational endpoints were all‐cause mortality and HF‐related admission. Predictors of LVEF trajectory improvement in the rhythm control group were evaluated. After matching, the two groups had similar age [mean age (years): rhythm/rate control: 63.96/65.13] and gender [male: rhythm/rate control: n = 228 (55.6%)/233 (56.8%)]. Based on baseline LVEF measurement, the post‐matched cohort had 490 HF with preserved ejection fraction (rhythm/rate control: n = 260/230; median LVEF: 58.00%/57.00%), 99 HF with mildly reduced ejection fraction (rhythm/rate control: n = 50/49; median LVEF: 45.00%/46.00%), and 231 HF with reduced ejection fraction (rhythm/rate control: n = 100/131; median LVEF: 32.50%/33.00%). Trajectory analysis found that the rhythm control group had a greater percentage of LVEF trajectory improvement than the rate control group [80 (53.3%) vs. 71 (39.4%), P = 0.012]. Cox regression analysis also showed that the rhythm control group was more likely to have improved LVEF trajectory compared with the rate control group {hazard ratio [HR] 1.671 [95% confidence interval (CI) 1.196–2.335], P = 0.003}. In the survival analysis, the rhythm control group experienced significant lower risks of all‐cause mortality [HR 0.600 (95% CI 0.366–0.983), P = 0.043] and HF‐related admission [HR 0.611 (95% CI 0.496–0.753), P