Neuromuscular impairments alter energetic cost landscape curvature and stride speed variability in post-stroke locomotion

Neuromuscular impairments induce compensatory effects which alter the dynamics of human movement, but the mechanism linking specific impairments to post-stroke locomotion remains poorly understood. Here, we combine a predictive neuromusculoskeletal simulation framework with experimental gait observations in stroke survivors to test how two hemiparetic impairments, reduced muscle strength and increased baseline muscle activity, reshape the energetic cost landscape. We then evaluate whether impairment-dependent changes in the cost landscape curvature are associated with stride speed variability, which is experimentally observed to be higher after stroke. Using neuromusculoskeletal simulations, we show that increased paretic muscle activity reduces local curvature near the cost-minimized speed more than reduced paretic muscle strength and find that this predicts increases in stride speed variability observed in hemiparetic locomotion. These results support a mechanistic hypothesis that flatter cost landscapes reduce the relative cost of suboptimal behavior and, therefore, may contribute to increased motor variability after stroke.