Assessment of 3-D joint contact load predictions during postural/stretching exercises in aged females.

Abstract

Our goal was to estimate knee and hip joint contact forces during a variety of unconstrained "stretching" exercises that are often recommended for people with arthritis. The population of interest was aged females with and without significant osteoarthritis (OA). Three-dimensional (3-D) kinematic, force platform, and selected electromyographic (EMG) data were measured. A relatively standard and reasonably efficient technique was used for all subjects and tasks: inverse dynamics to estimate joint reaction forces and net moments, followed by heuristic reductionist techniques for predicting muscle load-sharing. Muscle force predictions were compatible with measured EMG activity for some tasks but less so for others, especially those with significant axial and medio-lateral movement components or high co-contraction. Such results suggest several improvements, which come at computational cost: 3-D joint models which better document muscle and passive tissue loading-sharing in abduction and axial torsion, and dynamic or novel static optimization approaches that can inherently predict muscle co-contraction. Nonetheless, the predicted joint contact loadings provide estimates that are essentially lower bounds on the likely joint contact loads. Based on our results we suggest that the side-kick and back-kick tasks be modified because the loading levels on the support leg are potentially excessive for the aged arthritic population.