Luuk Vos

Accurate prediction of both mechanical and metabolic behavior of muscle remains an important challenge in biomechanics. The typically employed Hill-type muscle models have shown limited success in this regard. Therefore, Huxley-type models, in which mechanical and metabolic behavior is linked through cross-bridge cycling, have gained renewed attention. Previous studies fitted the cross-bridge cycling rate parameter values of such models on mechanical behavior only. It seems reasonable to assume that accurate predictions of mechanical behavior in a Huxley-type model will also result in accurate predictions of metabolic behavior as both depend on the cross-bridge dynamics. Here, we show that this assumption does not hold. We simulated previously collected mechanical and metabolic data from an experiment where participants performed either isometric or dynamic knee extensions in the gravitational field. We modeled this experiment with a musculoskeletal model consisting of two segments driven by one Huxley-type muscle model. We obtained 10 sets of cross-bridge rate parameter values by systematically varying the value of one of the rate parameters and optimizing the values of the remaining rate parameters with respect to the mechanical behavior. We then compared the predicted mechanical and metabolic behavior between the 10 sets. The predicted mechanical behavior was similar for all 10 sets. However, the accuracy of the predicted metabolic behavior differed substantially between the 10 sets. Our findings illustrate that different sets of cross-bridge rate parameter values may lead to similar mechanical behavior. We conclude that this should be exploited to obtain accurate predictions of mechanical and metabolic behavior simultaneously in Huxley-type muscle models.