Towards a precision rehabilitation approach for post-stroke stiff knee gait.

Abstract

Stiff knee gait, characterized by reduced knee flexion during swing, may arise from rectus femoris spasticity or inadequate pre-swing biomechanics. Difficulty in identifying each factor's contribution complicates clinical management. This study aimed to develop a predictive model for determining the contribution of inadequate pre-swing biomechanics to stiff knee gait in people with stroke. This historic cohort study analyzed gait data of 122 people with stroke and 20 healthy controls walking at four speeds. Linear regression analyses examined the relationship between pre-swing biomechanics and peak knee flexion in healthy controls. The pre-swing biomechanical measure explaining most of the variance in peak knee flexion was used in the predictive model, which was then applied to stroke data. Peak knee flexion angles of people with stroke were lower compared to healthy controls (stroke: 41 ± 16°; healthy controls: 61 ± 5°, 52 ± 8°, 57 ± 6°, and 61 ± 4° for self-selected walking speed, 0.4 m/s, 0.8 m/s, and 1.2 m/s, respectively). For healthy controls, peak knee flexion variance was best explained by combined pre-swing ankle and hip work (R² = 0.58). For 65 % of people with stroke, peak knee flexion fell above the lower bound of the regression model's prediction interval, suggesting stiff knee gait may primarily be caused by inadequate ankle push-off and hip pull-off. Our predictive model holds the potential to improve treatment selection by determining the impact of inadequate pre-swing biomechanics on stiff knee gait. In many participants, peak knee flexion was explained by pre-swing biomechanics, highlighting their key role in stiff knee gait.