The pooled results showed that BFRT improved scores on the Fugl-Meyer Assessment of the Lower Extremity (FMA-LE), a standard measure of motor recovery, by a mean difference of 2.14 points (95% CI 0.74 to 3.54) across ten studies including 646 participants. Balance, measured with the Berg Balance Scale (BBS), improved by a mean difference of 4.72 points (95% CI 1.62 to 7.82) across eight studies with 581 participants. Walking endurance, measured as six-minute walk distance, increased by 8.01 meters (five studies, 271 participants; 95% CI 2.31 to 13.71).

Walking speed, assessed with the 10-meter walk test, showed a mean difference of 0.14 m/s (95% CI 0.03 to 0.25) across four studies with 322 participants. However, this estimate attenuated when eligible grey literature was included, dropping to 0.09 m/s (95% CI −0.03 to 0.21) across six studies with 410 participants, a range that crosses zero and therefore leaves the effect on walking speed unresolved. No clear pooled effects were observed for the Timed Up and Go test, sit-to-stand performance, cadence, or standard full-scale Barthel Index or Modified Barthel Index scores, meaning the technique did not demonstrably improve these measures of mobility and daily activity in the pooled data.

A key caveat concerns whether the statistically significant improvements translate into benefits patients would notice. The authors report that the effects on FMA-LE and six-minute walk distance fell below reported clinically important thresholds. The Berg Balance Scale estimate was close to an approximately 5-point reference value, but its confidence interval spanned effects both below and above that threshold, so even the balance result cannot be confidently classified as clinically important.

In practical terms, the average improvements were small, and the review does not establish that BFRT produces meaningful functional recovery beyond what its statistical significance might suggest. Certainty in the evidence, as graded by GRADE, was moderate for the six-minute walk test, low for the Fugl-Meyer lower-extremity score, and very low for the Berg Balance Scale and 10-meter walk test walking speed. Among the studies with explicit safety information, no BFRT-related serious adverse events or thrombotic events were reported.

The authors caution, however, that safety reporting across the included trials was limited, so the absence of reported harm should be interpreted carefully rather than as definitive evidence of safety. The review's overall conclusion is measured: BFRT may improve selected lower-limb outcomes after stroke, but the clinical importance of several effects remains uncertain, and the strength of evidence varies from moderate to very low depending on the outcome. For clinicians and researchers, the findings position BFRT as a promising but not yet established low-load rehabilitation strategy after stroke, with the balance-scale result the closest to a potentially clinically meaningful effect and walking speed the least secure finding once grey literature was accounted for.

The authors call for larger, better-designed randomized trials that standardize BFRT protocols and report safety outcomes more completely.

Subject of Research: Systematic review and meta-analysis of blood flow restriction training for lower-limb function and walking recovery after stroke

Article Title: Blood flow restriction training for lower-limb function, walking and functional recovery after stroke: a systematic review and meta-analysis

Article References: Kang, H., Yan, Z., Zhang, Z., Xu, Y., Jiang, W., Yang, Z., Zhang, Q., & Yin, Y. (2026). Blood flow restriction training for lower-limb function, walking and functional recovery after stroke: a systematic review and meta-analysis. Frontiers in Neurology, 17, Article 1987031. https://doi.org/10.3389/fneur.2026.1987031

Image Credits: AI Generated

DOI: 10.3389/fneur.2026.1987031

Keywords: stroke rehabilitation, blood flow restriction training, systematic review, meta-analysis, lower-limb function, balance, walking speed, six-minute walk test, Fugl-Meyer Assessment, Berg Balance Scale, GRADE certainty, randomized controlled trials