Task-oriented treadmill training with versus without posterior elastic resistance in chronic stroke survivors with limited walking function: a randomised controlled trial with exploratory strength-GRF-speed associations

Abstract Background Reduced gait speed, impaired ground reaction force (GRF) generation, and lower-limb weakness are common in chronic stroke survivors and contribute to limited functional mobility. Although treadmill training can improve walking capacity, it remains unclear whether adding posterior elastic resistance provides incremental benefits in propulsion-related kinetics and in the functional linkage among strength, GRF, and walking performance, particularly in individuals with limited walking function. This study compared intensity-matched task-oriented treadmill training with versus without posterior resistance and secondarily explored changes in strength-GRF-speed relationships. Methods Forty chronic stroke participants were randomly assigned to conventional treadmill training (CON) or treadmill training with posterior resistance (EXP). Both groups received training for 30 min/session, twice weekly for eight weeks. Primary outcomes included preferred and maximum gait speed, bilateral vertical and anterior-posterior GRFs during walking, and lower-limb muscle strength. Secondary outcomes included walking energy expenditure (V̇O₂). Multivariate repeated-measures ANOVA was used to assess training effects. Pre-post associations among gait speed, GRFs, and strength were evaluated as secondary exploratory analyses. Results In the primary unadjusted analyses, both groups demonstrated significant time effects for preferred and maximum gait speed, walking energy expenditure, and affected-side knee flexor, knee extensor, and ankle plantarflexor strength (all p < 0.05). However, sensitivity analyses adjusted for stroke onset duration showed that several time effects were attenuated. A significant between-group difference was observed only for the second vertical GRF peak on the affected side at maximum speed, which was higher in EXP than CON and remained significant after adjustment ( p = 0.049). No broad group × time interaction effects were observed. Exploratory analyses showed that post-training associations among gait speed, affected-side strength, and anterior-posterior GRF appeared more consistent. Conclusions In chronic stroke survivors with limited walking function, intensity-matched treadmill-based gait training improved walking speed, oxygen uptake, and lower-limb strength, although several time effects were attenuated after adjusting for stroke onset duration. Adding posterior resistance showed a condition-specific between-group difference in paretic late-stance vertical loading at maximum speed but did not result in broad superiority in functional outcomes. Posterior elastic resistance may serve as a feasible adjunct to augment paretic-limb loading during high-demand walking. Trial registration ClinicalTrials.gov Protocol Registration: trial number NCT04974840.