A method for predicting peak work rate for cycle ergometer and treadmill ramp tests

BACKGROUND: Prediction of peak work rate (WRpeak) for incremental exercise testing (IET) is important to bring subjects to their maximal performance within the recommended 8-12 min. This study developed a novel method for prediction of WRpeak for IET on cycles and treadmills. METHODS: Peak metabolic equivalent of task (METpred) was predicted based on an existing non-exercise prediction formula, and then, predicted peak work rate (WRpred) was derived from separate formulae for the cycle and the treadmill. Eighteen healthy subjects were included. RESULTS: In males, there was no difference between WRpred versus WRpeak for both the cycle ergometer (277·7 versus 275·6 W, P = 0·70) and the treadmill (264·1 versus 260·5, P = 0·58). In females, there was no difference between WRpred versus WRpeak for the cycle ergometer (187·1 versus 188·3 W, P = 0·90), but a significant difference was found between WRpred versus WRpeak on the treadmill (178·6 versus 151·9 W, P<0·05). For males, the mean absolute percentage errors for WRpred versus WRpeak were 4·6% and 5·7% for the cycle and treadmill, respectively. For females, the errors were 12·2% and 20·8%. The algorithm was successful in achieving the required duration of 8-12 min in 33 of 36 cases. CONCLUSIONS: The peak work rate prediction protocol was accurate in male subjects for both the cycle and the treadmill. In female subjects, the method was accurate for the cycle, but systematically overpredicted the peak work rate on the treadmill. The protocol requires further adaptation for females on the treadmill.