Abstract
BACKGROUND: Heart rate recovery (HRR) has been observed to be a significant prognostic measure in patients with heart failure (HF). However, the prognostic value of HRR has not been examined in regard to the level of patient effort during exercise testing. Using the peak respiratory exchange ratio (RER) and a large multicenter HF database we examined the prognostic utility of HRR. METHODS: Cardiopulmonary exercise testing (CPX) was performed in 806 HF patients who then underwent an active cool-down of at least 1 min. Peak oxygen consumption (VO2), ventilatory efficiency (VE/VCO2 slope), and peak RER were determined with subjects categorized into subgroups according to peak RER (<1.00, 1.00-1.09, ≥ 1.10). HRR was defined as the difference between heart rate at peak exercise and 1 min following test termination. Patients were followed for major cardiac events for up to four years post-CPX. RESULTS: There were 163 major cardiac events (115 deaths, 20 left ventricular assist device implantations, and 28 transplantations) during the four year tracking period. Univariate Cox regression analysis results identified HRR as a significant (p<0.05) univariate predictor of adverse events regardless of the RER achieved. Multivariate Cox regression analysis in the overall group revealed that the VE/VCO2 slope was the strongest predictor of adverse events (chi-square: 110.9, p<0.001) with both HRR (residual chi-square: 16.7, p<0.001) and peak VO2 (residual chi-square: 10.4, p<0.01) adding significant prognostic value. CONCLUSIONS: HRR after symptom-limited exercise testing performed at sub-maximal efforts using RER to categorize level of effort is as predictive as HRR after maximal effort in HF patients.