Comparative Efficacy of Specific Exercise Interventions on Post-Stroke Depression, Anxiety and Cognitive Function in Older Adults: A Systematic Review and Network Meta-Analysis of Randomized Controlled Trials.

BACKGROUND: Post-stroke depression, anxiety, and cognitive impairment are prevalent in older adults and may impede recovery. We evaluated the comparative efficacy of exercise interventions for these outcomes using network meta-analysis. METHODS: We systematically searched five international and two Chinese databases from inception to July 2025 for randomized controlled trials enrolling older adults with stroke receiving exercise-based interventions. Primary outcomes were depression, anxiety, and cognitive function. Treatment effects were estimated as standardized mean differences (SMDs) with 95% credible intervals (CrIs), and interventions were ranked using surface under the cumulative ranking curve (SUCRA). Risk of bias and confidence in estimates were assessed with RoB 2 and CINeMA. RESULTS: Fifty-two trials involving 4,170 participants were included; 11 were at low risk of bias. For depression, dual-task training (DTT), mind-body exercise (MBE), and balance/coordination training (BCT) ranked highest and showed statistically credible benefits versus usual care. For anxiety, MBE, resistance/strength training (RST), and DTT ranked highest, with statistically credible benefits observed for RST and DTT, whereas EFR was associated with worse anxiety outcomes. For cognitive function, MBE, general aerobic exercise (GA), and moderate-intensity continuous training (MICT) ranked highest, whereas BCT was the only intervention showing a statistically credible benefit versus usual care. Meta-regression suggested that longer time since stroke was associated with smaller benefits for anxiety and cognitive outcomes. CONCLUSIONS: Exercise interventions may yield outcome-specific benefits after stroke in older adults. These findings support individualized, phase-sensitive exercise prescription in post-stroke rehabilitation.