Standard mechanical energy analyses do not correlate with muscle work in cycling.

The goal of this study was to assess the utility of experimental methods to quantify mechanical energy expenditure (MEE) in human movement. To achieve this goal, a theoretical model of steady-state cycling driven by individual muscle actuators was used to produce two distinct pedal simulations. The simulations yielded the same pedaling rate and power output, but one reduced the MEE by avoiding eccentric muscle contractions. Contractile element force and length change in the individual muscles was used to quantify the total positive and negative work produced by the muscles. Three methods using external measurements were applied to the simulated movement. The three methods to quantify MEE were based on: (1) segment kinematic measurements, (2) work done by total joint powers and (3) intercompensated joint powers, i.e. negative work from one joint is transferred to an adjacent joint where energy is being generated (positive work) via biarticular muscles. The results showed that none of the MEE analyses were correlated to the MEE of the individual muscles, with errors reaching 40%. Errors were mainly attributed to the inability of the MEE methods to account for co-contractions of antagonistic muscle groups. This phenomenon occurred primarily when one muscle generated force during activation while the antagonist generated force during deactivation.