A physiological counterpoint to mechanistic estimates of "internal power" during cycling at different pedal rates.

Reported values of "internal power" (IP) during cycling, generated by the muscles to overcome energy changes of moving body segments, are considerably different for various biomechanical models, reflecting the different criteria for estimation of IP. The present aim was to calculate IP from metabolic variables and to perform a physiological evaluation of five different kinematic models for calculating IP in cycling. Results showed that IP was statistically different between the kinematic models applied. IP based on metabolic variables (IP(met)) was 15, 41, and 91 W at 61, 88, and 115 rpm, respectively, being remarkably close to the kinematic estimate of one model (IP(Willems-COM): 14, 43, and 95 W) and reasonably close to another kinematic estimate (IP(Winter): 8, 29, and 81 W). For all kinematic models there was no significant effect of performing 3-D versus 2-D analyses. IP increased significantly with pedal rate - leg movements accounting for the largest fraction. Further, external power (EP) affected IP significantly such that IP was larger at moderate than at low EP at the majority of the pedal rates applied but on average this difference was only 8%.