Non-pharmacological interventions for lower limb rehabilitation after stroke: a systematic review and network meta-analysis

Abstract

Stroke is a leading cause of long-term adult disability, and lower limb dysfunction substantially limits functional independence. This study aimed to systematically compare the effects of different non-pharmacological rehabilitation interventions on lower limb recovery after stroke using a network meta-analysis. This systematic review was conducted in accordance with PRISMA guidelines. Randomized controlled trials evaluating non-pharmacological interventions for post-stroke lower limb rehabilitation were identified from MEDLINE, Embase, Web of Science, and PubMed. Outcomes included balance assessed by the Berg balance scale (BBS), motor function assessed by the Fugl-Meyer assessment (FMA), and gait speed assessed by the 10-meter walk test (10MWT). A random-effects network meta-analysis was performed using R software. Evidence certainty was evaluated using the CINeMA approach based on the GRADE framework. 38 randomized controlled trials involving 1,932 participants were included. Compared with conventional rehabilitation, mirror therapy (MT) yielded the greatest improvement in balance (Berg Balance Scale, MD = 5.296, 95% CI 2.104–8.488); functional electrical stimulation (FES) achieved the optimal improvement in gait speed (10-meter walk test, MD = 0.118, 95% CI 0.056–0.180); and robot-assisted gait training (RAGT) showed the most significant benefit in overall lower limb motor function (Fugl-Meyer Assessment lower extremity subscale, MD = 4.608, 95% CI 0.922–8.294). Ranking probability analyses indicated that MT, FES, and RAGT had the highest likelihood of being the top-performing interventions for their respective outcome domains. The certainty of evidence ranged from very low to moderate per the CINeMA framework. Non-pharmacological rehabilitation interventions exhibit domain-specific efficacy for post-stroke lower limb recovery: MT presents the optimal benefit for balance, FES for gait speed, and RAGT for global lower limb motor function. However, these findings should be interpreted with caution due to the overall low-to-moderate certainty of evidence. High-quality, adequately powered randomized controlled trials with standardized outcome assessment and long-term follow-up are warranted to validate these findings and clarify the long-term clinical effectiveness of these interventions.

The authors' abstract, as published at the source. Journal of NeuroEngineering and Rehabilitation, 2026 · DOI ↗