| Sumario: | Objective: This study aimed to compare the effects of different dual-task training modalities on walking and balance in stroke patients using network meta-analysis. Data sources: Randomized controlled trials were searched in PubMed, Embase, Web of Science, Cochrane Library, China National Knowledge Infrastructure, and Chinese Biomedical Database from inception to September 2025. Review methods: Methodological quality was assessed using the Cochrane risk-of-bias tool. A frequentist approach was applied for network meta-analysis, and evidence quality was evaluated with confidence in the network meta-analysis framework. Results: Thirty-four trials covering 10 dual-task modalities were included. Network meta-analysis results showed that motor-walking dual-task training was the most effective in improving cadence (mean difference = 7.04, 95% CI : 1.95 to 12.12), walking speed (standardized mean difference = 0.67, 95% CI : 0.27 to 1.06), step length (standardized mean difference = 0.56, 95% CI : 0.20 to 0.92), stride (standardized mean difference = 0.76, 95% CI : 0.45 to 1.08), 10-Meter Walk Test (standardized mean difference = 0.73, 95% CI : 0.05 to 1.12) and berg balance scale (mean difference = 5.76, 95% CI : 1.38 to 10.14). Cognitive-balance dual-task training showed the greatest improvement in Timed Up and Go test (mean difference = −4.73, 95% CI : −10.24 to −0.78). The certainty of evidence for most outcomes, assessed using the Grading of Recommendations Assessment, Development and Evaluation framework, ranged from very low to high. Conclusions: Motor-walking dual-task training best improved cadence, walking speed, step length, stride, 10-Meter Walk Test, and Berg Balance Scale. Cognitive-balance dual-task training showed the greatest improvement in Timed Up and Go test. These results inform future guidelines and clinical decisions for stroke patients' walking and balance rehabilitation.
|
|---|