Vibration Therapy for Poststroke Clinical Balance: A Domain-Specific Systematic Review and Multilevel Meta-Analysis

BACKGROUND/OBJECTIVES: Balance impairment is a major target of poststroke rehabilitation, but the effectiveness of mechanical vibration therapy remains uncertain. This systematic review and multilevel meta-analysis evaluated vibration therapy for validated clinical balance outcomes after stroke and explored potential effect modifiers. The review was prospectively registered in PROSPERO (CRD420251110102). METHODS: PubMed, MEDLINE, Embase, the Cochrane Library, and Web of Science were searched from inception to 31 March 2026. Nineteen controlled studies contributed 22 effect-size estimates. The primary analysis used standardized between-group changes from baseline and a three-level random-effects model with study-clustered CR2 inference. Sensitivity analyses, FDR-adjusted moderator analyses, risk-of-bias assessment, and GRADE were performed. RESULTS: Vibration therapy improved clinical balance on average (Hedges' g = 0.593, 95% CI 0.191-0.995; p = 0.006), although the 95% prediction interval crossed the null. Restriction to 16 randomized trials yielded a similar effect (g = 0.507, 95% CI 0.141-0.873; p = 0.010). Excluding studies at high or serious risk of bias attenuated the effect to g = 0.161 (95% CI -0.116-0.439; p = 0.217). None of seven moderators remained significant after FDR correction. The BBS-specific mean difference was 3.326 points (95% CI 1.714-4.938). Certainty was low for randomized and very low for non-randomized evidence. CONCLUSIONS: Vibration therapy may improve poststroke clinical balance on average, but heterogeneity and sensitivity to study quality limit confidence in effect magnitude and reproducibility. Current evidence does not support a specific vibration protocol; rigorous dose-ranging randomized trials are needed.