Amputee motor control during loaded walking with stiffer and powered prostheses

Background Individuals with transtibial amputation often rely on prosthetic feet with fixed mechanical properties, which may limit distal limb neuromuscular strategies during demanding tasks such as load carriage. Although stiffer and powered prosthetic feet are purported to aid ambulation under load, their influence on motor control during loaded walking remains unclear. This study investigated whether changing prosthetic foot prescription alters muscle activation, and motor module composition and complexity during load carriage. Methods Twelve individuals with unilateral transtibial amputation walked at their preferred speed while carrying no load and a 13.6-kg (30-lb) load in multiple conditions (front, back, intact-side, and prosthetic-side). Subjects used a clinically prescribed prosthetic foot, the same foot but one category stiffer, and, in a subset ( n = 7), a powered ankle-foot prosthesis. Surface electromyography was collected from eight intact-limb muscles and five residual-limb muscles; module analysis primarily focused on the intact limb. Muscle activation and co-contraction were quantified, and a muscle motor module analysis using non-negative matrix factorization was performed to quantify motor control. Linear mixed-effects models tested differences between feet in muscle activation and module complexity, while module similarity and activation timing were evaluated using cosine similarity and one-dimensional statistical parametric mapping, respectively. Results Prosthetic foot types influenced activation of several intact-limb muscles across load conditions. Compared with the prescribed foot, the powered foot increased medial gastrocnemius activation by 26–29% during unloaded and intact-side load walking and rectus femoris activation by 46% during unloaded walking, while the stiffer foot increased vastus medialis activation by 37% during intact-side load walking. Despite these muscle-level changes, module complexity remained largely unchanged, with only limited task-specific differences. Three modules explained > 90% of electromyographic variance in most subjects, with only modest reductions in first-module variance (≈ 2–3%) for the powered foot during back and intact-side-load conditions. Module composition and activation timing were similar across feet. Exploratory analysis of the residual-limb muscles found no significant between-foot differences in module complexity within any load condition. Conclusions Although prosthetic feet with different mechanics alter specific muscle activation patterns during loaded walking, the underlying module composition and activation timing were largely unchanged, while module complexity showed small, task-specific changes that did not produce consistent changes in the number of modules required. These findings suggest that individuals with transtibial amputation accommodate changes in prosthetic foot mechanics and added load primarily through changes in muscle activation rather than reorganization of motor control strategies. Trial Registraion ClinicalTrials.gov, NCT07159490, Registration date: 01 June 2022.