Reliability of different models to assess heart rate recovery after submaximal bicycle exercise

Objectives: Different models to assess HR recovery have been developed but knowledge of their reliability is poor at different submaximal exercise intensities and recovery durations. Our aim was to determine the reliability of HR recovery after a test on a cycle ergometer. Design: Twenty-one healthy individuals performed a submaximal exercise at 65% and 80% HRmax followed by passive recovery. The exercise was repeated (retest) within 2 weeks to assess reliability. Method: HR recovery was assessed by 8 models, based on monoexponential kinetics or absolute recovery (recovered HR at fix time points). Intraclass correlation coefficient (ICC) and standard error of measurement (SEM, %SEM) were used to address reliability of measurements. Results: We found that: (1) assessment of HR recovery after 80% HRmax exercise leads to more reliable values than after 65% HRmax exercise (mean ICC: 0.827 vs. 0.747); (2) a longer recovery time increases the absolute consistency of the measurement (%SEM: 26.7 at 60 s, 19.5 at 120 s and 16.3 at 180 s, irrespective of model or exercise intensity); (3) T30min is more reliable than T30 (ICC: 0.691 vs. 0.528; %SEM: 28.5 vs. 70.8) for the calculation of the fast component of HR recovery (HR kinetics calculated over ≤60 s) after exercise. Conclusions: A good sensitivity of measurement – large ICC and small SEM – is obtained for analysis of HR recovery after submaximal exercises on the cycle ergometer, especially for internal workloads of 80% HRmax. At this workload, consistency of results increases for recoveries of longer duration.