A model-based estimation of critical torques reduces the experimental effort compared to conventional testing

Critical torque (CT) is an important fatigue threshold in exercise physiology and can be used to analyze, predict, or optimize performance. The objective of this work is to reduce the experimental effort when estimating CTs for sustained and intermittent isometric contractions using a model-based approach.

We employ a phenomenological model of the time course of maximum voluntary isometric contraction (MVIC) torque and compute the highest sustainable torque output by solving an optimization problem. We then show that our results are consistent with the steady states obtained when simulating periodic maximum loading schemes. These simulations correspond to all-out tests, which are used to estimate CTs in practice. Based on these observations, the estimation of CTs can be formulated mathematically as a parameter estimation problem. To minimize the statistical uncertainty of the parameter estimates and consequently of the estimated CTs, we compute optimized testing sessions. This reduces the experimental effort even further.

We estimate CTs of the elbow flexors for sustained isometric contractions to be 28% of baseline MVIC torque and for intermittent isometric contractions consisting of a 3 s contraction followed by 2 s rest to be 41% of baseline MVIC torque. We show that a single optimized testing session is sufficient when using our approach.

Our approach reduces the experimental effort considerably when estimating CTs for sustained and intermittent isometric contractions.