Heart rate response and recovery in bicycle exercise testing - reference values and association with mortality

Abstract Introduction Chronotropic incompetence and impaired heart rate (HR) recovery are related to mortality. Guidelines lack specific recommendations for reference values on HR recovery assessment. Purpose To define normal values for HR response and recovery in individuals referred for a bicycle exercise stress test and to relate these factors to risk of mortality. Methods We included 9,917 subjects (46% females) aged 18–85 years who performed a bicycle exercise test. We defined age- and sex-specific reference equations for peak HR, HR reserve, and HR recovery at 1 and 2 minutes (HRR1 and HRR2) using a subgroup of individuals without apparent cardiovascular (CV) risk nor comorbidities (N=2,774). Associations between blunted HR indices (<5thpercentile) and mortality were analysed using Cox regression in all subjects and in individuals with normal exercise capacity (N=8,070). Results All HR indices were age-dependent. The 5th percentiles of HR reserve, HRR1, and HRR2 correlated weakly with previously reported reference values. During median follow-up of 8.6 years 1,110 (11%) died. HR recovery variables were the strongest predictors of all-cause (HRR1, hazard ratio 1.66 (95% confidence interval, 1.45–1.90)) and CV mortality (HRR2, 1.59 (1.26–2.00)) after adjustments for potential confounders. Blunted HR response (peak HR) increased risk of all-cause (1.32 (1.14–1.53)) and CV mortality (1.41 (1.09–1.83)). Combined blunted HRR1 and HRR2 or HR response and HRR1 further increased risk of all-cause (1.92 (1.64–2.24) and 1.85 (1.55–2.22), respectively) and CV mortality (1.95 (1.50–2.54) and 1.94 (1.42–2.65), respectively). HR recovery variables (HRR1, 1.97 (1.63–2.39) and HRR2, 1.76 (1.46–2.12)) remained stronger predictors of mortality than peak HR (1.38 (1.10–1.72)) in individuals with normal exercise capacity. Conclusions All HR variables were age-dependent and associated with mortality. Blunted HR recovery was more strongly related to mortality than chronotropic incompetence, and their combination added prognostic value. Furthermore, blunted HR recovery variables, rather than blunted HR response, remained strongly associated with mortality even in the presence of normal exercise capacity.