Gait speed-dependent modulation of paretic versus non-paretic propulsion in persons with chronic stroke

BACKGROUND: Persons with chronic stroke (PwCS) exhibit impaired paretic propulsion generation. Consequently, PwCS walk slower than healthy peers and rely more on their non-paretic leg, leading to propulsion asymmetry. However, it remains unclear how propulsion symmetry is influenced by walking at various gait speeds. This study aimed to investigate the relation between gait speed and propulsion symmetry in PwCS and controls. METHODS: Fifteen PwCS and sixteen healthy controls walked on an instrumented treadmill at randomized speeds, ranging from 0.2 m/s to comfortable walking speeds for PwCS or 0.4 to 1.6 m/s for controls, with 0.2 m/s increments. PwCS continued to their maximum speed with 0.1 m/s increments. Propulsion, derived from the anteroposterior component of the ground reaction force, was defined as propulsion peak and propulsion impulse. The primary outcome was propulsion peak and impulse symmetry (paretic propulsion / total propulsion), with secondary outcomes being propulsion peak and impulse per leg. The relationship between gait speed and propulsion metrics was analyzed using linear mixed models (LMM). RESULTS: PwCS exhibited clear propulsion peak and impulse asymmetry across all gait speeds, while controls maintained symmetrical propulsion. LMMs revealed no change in propulsion peak symmetry with gait speed (β = 0.12, SE = 0.090, p = 0.19), with considerable variability among PwCS. Propulsion impulse symmetry improved with increasing gait speed (β = 0.39, SE = 0.048, p < 0.001), especially in PwCS who had greater asymmetry at comfortable walking speed. Propulsion peak and impulse increased with gait speed in both legs for PwCS and controls. The propulsion peak increase was stronger in the non-paretic compared to the paretic leg (0.16 ± 0.043 vs. 0.12 ± 0.042 N/kg per 0.1 m/s), while the propulsion impulse increase was similar between legs. CONCLUSIONS: PwCS showed reduced paretic leg contribution to forward propulsion across various gait speeds. The relative paretic contribution for propulsion peak remained constant while it increased with gait speed for propulsion impulse, especially in those with greater asymmetry at their comfortable walking speed. Furthermore, all participants were able to increase paretic propulsion peak and impulse above their propulsion at comfortable walking speed, suggesting some residual paretic capacity.