Comparative Efficacy of Exercise Modalities for Motor Function Recovery After Acute Ischemic Stroke: A Systematic Review and Bayesian Network Meta-Analysis
TL;DR
This first comprehensive evidence synthesis evaluating the comparative effectiveness of specific exercise modalities across distinct recovery domains after AIS offers an evidence-based framework for clinicians to optimize post-stroke rehabilitation protocols and highlights priority areas for future research on dose–response and long-term outcomes.
Abstract
Background: Exercise rehabilitation (ER) has become an important treatment regimen during the recovery phase of acute ischemic stroke (AIS). Optimal exercise prescriptions for post-stroke motor recovery remain controversial. Methods: PubMed, Embase, the Cochrane Library, and Web of Science were systematically searched through November 2025. The Cochrane risk-of-bias (RoB 1) assessment tool was utilized to assess the risk of bias in the original studies. Results: This study included 66 trials comprising 3675 patients. For balance, whole-body tilting postural training (WTPT), unilateral strength training (UST) and robot-assisted unilateral gait training (RAUGT) showed high SUCRA rankings. Activities of daily living likely improved most with robot-assisted Tai Chi training (RATCT), home rehabilitation guidance (HRG) and robot-assisted gait training (RAGT). Fugl-Meyer Assessment (FMA) gains appeared greatest following robot-assisted hand training with electrical stimulation (RAHT + ES), neural inhibition therapy (NIT) and graded motor imagery with electrical stimulation (GMI + ES). Walking endurance seemed most responsive to visual feedback training (VFT), UST and visual aids training (VAT), while functional mobility showed greatest improvement with tilt sensor-assisted gait training with electrical stimulation (TAGT + ES), gait training (GT) and virtual reality therapy with motor imagery (VRT + MI). Conclusions: This study provides the first comprehensive evidence synthesis evaluating the comparative effectiveness of specific exercise modalities across distinct recovery domains after AIS. Rather than supporting generic interventions, the findings highlight the distinct comparative advantages of various emerging technologies and traditional practices. These results offer an evidence-based framework for clinicians to optimize post-stroke rehabilitation protocols and highlight priority areas for future research on dose–response and long-term outcomes.