Gait speed-dependent foot placement control in persons with chronic stroke and controls

The foot placement strategy is the primary mechanism for mediolateral (ML) balance control during gait and often impaired in persons with chronic stroke (PwCS). While healthy adults show precise control across most gait speeds, except very slow speeds, this relationship remains unexplored in PwCS.

Participants walked on an instrumented treadmill at randomized speeds (0.2 - 1.6 m/s or maximum speed for PwCS). Foot placement deviation, the RMSE between actual and model-fitted foot placements based on CoM position and velocity at initial contact, quantified foot placement control. Best-fitting mixed-effects models (linear, exponential, or quadratic) assessed foot placement deviation across all speeds and, post-hoc, at lower and higher speeds separately.

Data from sixteen PwCS with self-reported balance impairments and 22 controls were analysed. A quadratic model with random intercepts and slopes showed consistently higher deviation across speeds in PwCS. Below 0.8 m/s, deviation decreased linearly (β = -2.19, p < 0.001), without leg or group differences in linear trend or curvature. Above 0.8 m/s, the paretic leg showed a significantly greater increase and curvature (linear β = 1.70, p < 0.001; quadratic β = 0.80, p < 0.001) compared to controls and the non-paretic leg.

Foot placement control exhibits a curvilinear relationship with gait speed in PwCS and controls, with less control at slow speeds in both groups. While PwCS show consistent foot placement control impairments compared to controls across all speeds, paretic impairments increase at higher speeds. These speed-sensitive paretic impairments likely reflect the inability to adapt to increased sensorimotor demands during faster gait.