Predicting Exercise Training Outcome From Cardiac Rehabilitation

In Brief PURPOSE: This study was conducted to identify predictors of training effect after 6 months of supervised exercise for cardiac rehabilitation patients. METHODS: Data at baseline and after 6 months of supervised exercise from 60 patients with coronary artery disease were used for this study. All the patients exercised 3 days per week for 5 to 9 months. Attendance exceeded 70%. The training effect was independently evaluated by the change in rate-pressure product at a 5-metabolic equivalent (MET) workload (RPP5), and the change in estimated peak METs (METPK) during treadmill testing. Baseline variables were examined to identify predictors of change in RPP5 and METPK separately using multiple linear regression. RESULTS: A reduction in RPP5 (−23 ± 33 bpm·mm Hg·102; P < .0001) and an increase in METPK (2.1 ± 1.7 METs; P < .0001) were found after training. Baseline fitness was the best predictor of training effect. Inverse relations were found between baseline scores and change scores for RPP5 (r = − 0.68; P < .001) and METpk (r = − 0.39; P = .002). Beta-blocker status also was useful for predicting change in RPP5. No association was found between training effect and age, revascularization status, or exercise-induced ischemia. CONCLUSIONS: Training effect has limited predictability using baseline variables. No baseline patient characteristics other than high initial fitness were related to reduced training effect. Therefore, patients with clinical characteristics similar to those evaluated in this study can achieve a training effect successfully and are eligible for cardiac rehabilitation. The findings of this study support previous research related to this topic. Baseline data from 60 cardiac rehabilitation patients were used to identify predictors of training effect that occurred after 6 months of supervised exercise. Baseline fitness was associated with training effect, while age, revascularization status, exercise-induced ischemia and other clinical variables were not predictive of magnitude of training effect.