Predictive simulation of diabetic gait: Individual contribution of ankle stiffness and muscle weakening.

Diabetic neuropathic individuals present massive muscle strength reduction at the ankle plantar- and dorsiflexors and increased joint stiffness. Our aim is to investigate the adaptation strategies to these musculoskeletal alterations during walking by means of predictive simulations. We used a seven segment planar musculoskeletal model actuated by eight Hill-type muscles in each leg. The effect of all passive tissue in muscles and other joint structures was modeled by net passive joint moment curves. The predictive simulations were generated by solving an optimal control problem that minimized a cost function, including effort and tracking terms, using direct collocation and a commercial optimal control package. We simulate four conditions to represent the weakening of the distal muscles triceps sural (TS) and tibialis anterior (TA), and five conditions to represent the effect of increasing nonlinear ankle stiffness in flexion. The weakening of the distal muscles leads to a delayed action of the TS and a progressive decrease of the gastrocnemius peak force in the push-off phase. This distal deficit is compensated by a larger hip flexion moment resulting from an increase in the iliopsoas muscle force in this phase, known as the hip strategy. The adaptation mechanisms observed in response to an increase in ankle stiffness include the hip strategy and the exploitation of the passive joint structures as springs, which store energy during midstance and release it during push-off, reducing TS force and power in this phase and leading to a consistent decrease in the overall muscle force levels.