Generalization of a model based on biophysical concepts of muscle activation, fatigue and recovery that explains...

To address the limitations of regression-based performance models, the literature describes a fatigue model that reduces the complexities of motor unit activation into a set of first-order differential equations requiring only a few parameters to capture the global effects of activation physiology (M0: maximal force-generating capacity, F: fatigue rate, R: recovery rate). However, there are no solutions to the general form of the equations, which limits its applicability. We formulate an algorithm that allows the equations to be solved if an arbitrary force profile is specified. Furthermore, we support the validity of the model, applying it to exercises found in the literature including quadriceps contractions (M0=954±326 N, F=2.5±0.4%·s(-1), R=0.3±0.3%·s(-1)), cycling (M0=1095±486 W, F=3.5±0.3%·s(-1), R=1.1±0.3%·s(-1)) and running (M0=9.2±1.2 m·s(-1), F=0.9±0.4%·s(-1), R=1.0±0.3%·s(-1)), where effective muscle forces are converted to cycling power and running speed. The model predicts muscle output for 10 maximum efforts and 32 endurance tests, where the coefficient of determination (R2) ranged from 0.81 to 1.00. These results support the hypothesis found in the literature that motor unit activation and fatigue mechanisms lead to a cumulative muscle fatigue effect that can be observed in exercise performance.