Heart rate reduction and changes in health related quality of life among heart failure patients with prevalent atrial fibrillation: a prospective study
T Juhlin, H Holm Isholth, Z Nezami, P G Platonov, A Jujic, M MagnussonAbstract
Background
Heart rate (HR) control is an important treatment strategy in atrial fibrillation (AF), but it remains uncertain whether reductions in HR translate into meaningful improvements in patient-reported quality of life. The Kansas City Cardiomyopathy Questionnaire (KCCQ) provides a validated measure of health status in cardiovascular disease, but its association to HR changes in AF has not been fully clarified.
Aims
To investigate whether reductions in HR from baseline (hospital admission) to follow-up are associated with improvements in global KCCQ scores in patients with prevalent AF.
Methods
We analysed data from 262 patients with prevalent AF, of whom 114 patients also had been re-examined at 6 months. HR change and KCCQ change were defined as baseline minus follow-up (positive HR change indicates HR reduction; negative KCCQ change indicates improved KCCQ). Associations between HR change and KCCQ change were assessed using Spearman correlation, linear regression (crude and adjusted for age and sex), and a baseline-adjusted change model including baseline KCCQ and baseline HR. Non-linearity was explored using a generalized additive model. For descriptive transition analyses, we examined improvement from HR ≥110 to <110 bpm and from KCCQ ≤50 to >50. Mortality analyses within the same cohort used Kaplan–Meier and Cox regression (crude and adjusted for age and sex), and Cox models with spline terms for baseline HR.
Results
Among 114 patients (mean age 76.9±9.7 years; 28.1% women), mean baseline HR was 89.5±23.9 bpm. A weak inverse association was observed between HR change and KCCQ change (Spearman rho −0.19, p=0.038). In crude linear regression, HR change was not significantly associated with KCCQ change (β −0.14 KCCQ points per 1 bpm HR reduction; p=0.061). In the baseline-adjusted change model (including baseline KCCQ and baseline HR), HR change was not independently associated with KCCQ change (β −0.20 per 1 bpm; p=0.13). Spline modelling did not indicate non-linearity (Figure 1). Transition analyses showed no significant association between HR improvement (≥110 to <110) and KCCQ improvement (≤50 to >50) (p=0.30), although the counts were sparse in the joint improvement category. In survival analyses (75 deaths), baseline HR ≥110 bpm was not associated with mortality in crude Cox regression (HR 0.70, 95%CI 0.39-1.26) or after adjustment for age and sex (HR 0.90, 95%CI 0.50-1.62).
Conclusion
In this HF cohort with prevalent AF and paired baseline/follow-up measurements, HR reduction over 6 months showed only a weak unadjusted correlation with improvement in KCCQ and was not independently associated with KCCQ change after accounting for baseline KCCQ and clinical covariates. Baseline HR category was not associated with mortality in this cohort.Figure 1.For image description, please refer to the figure legend and surrounding text.