Background: In patients with coronary artery disease, the target intensity-level of exercise training is usually based on a training heart rate that aims to be close to the upper level of metabolic aerobic exercise.
Aim: We intended to evaluate whether a training heart rate calculated with the Karvonen formula after a conventional exercise test is comparable with the heart rate at the anaerobic threshold in patients after myocardial infarction treated with beta-blockers and if not to propose a new formula.
Methods and results: In this multicenter prospective study, 115 consecutive beta-blocked patients recovering from myocardial infarction performed a cardiopulmonary exercise test to determine the anaerobic threshold. The training heart rate determined by the Karvonen formula was compared with the heart rate at the anaerobic threshold in a derivation sample (n=58) and a validation sample (n=57) of patients. The Karvonen training heart rate was significantly lower than the heart rate at the anaerobic threshold (91+/-5 versus 102+/-17 bpm, P<0.0001) in the first sample of patients and this difference was clinically relevant in 40% of patients. Thus, a 'modified Karvonen training heart rate', equal to 0.8xx(maximum heart rate-resting heart rate)+resting heart rate, was calculated by linear regression in the derivation sample and prospectively assessed in the validation sample. The modified Karvonen training heart rate was closer to the heart rate at the anaerobic threshold than the Karvonen training heart rate, and the difference between the modified Karvonen training heart rate and the heart rate at the anaerobic threshold was clinically relevant in only 5% of patients.
Conclusion: The Karvonen formula underestimates the heart rate at the anaerobic threshold in beta-blocked patients, which may lead to undertraining of patients with coronary artery disease; we propose another formula more adapted to these patients.