High-intensity interval training after stroke: a three-level random-effects meta-analysis with cluster-robust
Xuanzi Zhang1, Jie Yang1, Juanjuan Hu1
1Department of Physical Education, Chengdu College of University of Electronic Science and Technology of China, Chengdu, China.
Objective:
To estimate the effects of high-intensity interval training (HIIT) on balance, walking outcomes, and physiological endpoints after stroke, and to generate hypotheses about whether training-load parameters may explain variability in intervention effects.
Methods:
We searched PubMed, Web of Science, Embase, Scopus, and the Cochrane Library from inception to December 31, 2025, for English-language randomized controlled trials of HIIT in post-stroke populations. Two reviewers independently screened records and extracted data. Risk of bias was assessed using RoB 2, and certainty of evidence was assessed using GRADE. Effect sizes were calculated as Hedges' g using between-group change scores. When change-score standard deviations were unavailable, they were imputed using a pre-post correlation of r = 0.5, with sensitivity analyses varying r. Effects were pooled using three-level random-effects models to accommodate dependent effect sizes. Statistical inference, including 95% confidence intervals and p-values, was based on cluster-robust variance estimation with small-sample correction. Meta-regression and subgroup analyses were conducted as exploratory, hypothesis-generating analyses of heterogeneity rather than confirmatory tests of training-load effects. The protocol was registered in PROSPERO (CRD42027809778).
Results:
Fourteen trials involving 717 participants were included. Pooled estimates suggested that HIIT may improve balance, as measured by the Berg Balance Scale, although the effect was small and the certainty of evidence was low (ES = 0.20, 95% CI 0.01 to 0.39, p = 0.039). HIIT may also improve walking endurance, as measured by the 6-min walk test, but the certainty of evidence was very low (ES = 0.41, 95% CI 0.22 to 0.61, p < 0.001). No statistically significant effect was observed for 10-meter walk test time (ES = 0.06, 95% CI-0.16 to 0.29, p = 0.579). Pooled estimates suggested a possible improvement in cardiorespiratory fitness, although the certainty of evidence was very low (ES = 0.36, 95% CI 0.05 to 0.66, p = 0.021). Effects on systolic blood pressure (ES = 0.05, 95% CI-0.25 to 0.36, p = 0.722) and diastolic blood pressure (ES = 0.25, 95% CI-0.05 to 0.56, p = 0.099) were not statistically significant. Evidence for stroke severity, assessed using the Scandinavian Stroke Scale, was sparse and uncertain (ES = 0.29, 95% CI-0.04 to 0.62, p = 0.084). Exploratory meta-regression and subgroup analyses identified preliminary signals of between-study variability, but these findings should be interpreted strictly as hypothesis-generating because of the small number of trials, limited outcome-specific effect sizes, and multiple comparisons. Overall, the certainty of evidence was low to very low across outcomes.
Conclusions:
Low- to very-low-certainty evidence suggests that HIIT may improve balance, walking endurance, and cardiorespiratory fitness after stroke, whereas effects on short-distance walking speed or gait control, blood pressure, and stroke severity remain uncertain. Apparent associations between training-load parameters and outcomes should not be interpreted as evidence of optimal HIIT prescriptions. These exploratory findings require confirmation in adequately powered randomized trials with standardized HIIT definitions, rigorous reporting of achieved intensity, and longer follow-up.
Systematic Review Registration:
https://www.crd.york.ac.uk/prospero/, identifier: CRD42027809778.