Comparison of models for the physiological estimation of internal mechanical power in cycling

Abstract

Physiological models for estimating internal mechanical power (IP) generally share a common basis: the sum of IP and the external mechanical power (EP) is equal to the total mechanical power, as represented by net VO₂. While a biomechanical perspective of IP argues against this simple summation, physiological models serve a valuable purpose because they account for the total flow of energy through the system. Despite their common foundation, estimates of IP using various physiological models have not been consistent. Two pre-existing models that use a physiological approach, as well as variations of them and a simple body mass-cadence relationship, were applied to submaximal data from eight well-trained male cyclists. Three incremental cycling tests were performed at cadences of 50-55, 80-85 and 110-115 rev • min⁻¹. Differences in the mean and limits of agreement were used to show that values of IP calculated using two previously described models were not similar at any of the cadences tested. It was also shown that using relevant energy equivalents for converting VO₂ into metabolic power produced smaller values for IP than when using a generic energy equivalent. Differences in values for IP in the published literature, therefore, might not necessarily be caused by differences in participant characteristics, but rather differences in the accuracy of the variables that are input into the IP models.