The main purpose of this study was to compare the sagittal plane loading ( change in magnitude and rate of change) of the joints of the support leg during the impact phase in running to the vertical ground reaction force (VGRF) and to compare these loading characteristics between subjects using a direct dynamics simulation model of the musculoskeletal system. During development of the impact model, the modelling of force production in skeletal muscle was examined, as well as the influences of muscles, soft tissue movement, and foot inversion-eversion on the impact loading of the body. Evaluation of the model with respect to experimental measurements showed that it was able to predict the peak and average rate of change of the VGRF with mean error and standard deviation of 12.4 ± 7.8% and 16.2 ± 9.3%, respectively, for five subjects. In addition, the kinematics of each subject were well represented.

The results of the study suggested that muscles generally increased the external loading, but not the loading of the joints. Soft tissue movement generally decreased the external loading and, hence, the internal loading. Foot inversion-eversion generally increased the time to the peak VGRF and, as a result, decreased the rate of loading both externally and internally. The precise influences varied between subjects.

Animal experiments in which repetitive loading, similar to the vertical ground reaction force in running, has produced degenerative changes to articular cartilage suggest that runners may be at risk to the development of osteoarthritis. The two results of this study that (a) the loading in the joints of the lower extremities were always less than the external loading and (b) there was considerable variability in joint loading between subjects provide possible explanations why osteoarthritis is not generally observed in runners. A third alternative is also proposed based on adaptation of cartilage to mechanical load.

The model developed in this study has the potential to assist in evaluating each of the above possibilities in runners. However, further evaluation of the assumptions and parameters used in the model is first required.