Data Sheet 1_Dose–response relationships between temporal variables of blood flow restriction exercise and physiological adaptations: a systematic review and meta-analysis.docx

These SMDs were pooled using a three-level random-effects model with robust variance estimation (RVE; clustered by study). A DerSimonian-Laird random-effects model with inverse-variance weighting was used for consistency checks.</p>Results<p>Overall, BFR exercise elicited significant physiological adaptations (SMD = 0.249, 95% CI: 0.095–0.405), with findings remaining consistent in the DerSimonian-Laird model (overall SMD = 0.231, 95% CI: 0.111–0.349).

By intensity, effects were: high-intensity BFR group (SMD = 0.208, 95% CI: 0.163–0.252), moderate-intensity group (SMD = 0.315, 95% CI: −0.465–1.095) and low-intensity group (SMD = 0.293, 95% CI: 0.057–0.529). Meta-regression showed positive associations of effect size with BFR duration (β = 0.0136, p = 0.0011) and stimulus duration (β = 0.0104, p = 0.0347). However, after adjusting for the period, BFR duration was not significant (p = 0.0972) due to collinearity; BFR pressure was not a significant moderator.

Regarding specific physiological responses, only VO₂max showed robust statistical significance (SMD = 0.386, 95% CI: 0.055–0.717).</p>Conclusion<p>The most consistent evidence for BFR-induced improvements relates specifically to VO₂max, whereas effects on other physiological outcomes remain tentative. In the initial RVE, among the specific physiological outcomes, VO₂max, MPO, and RER reached significance, whereas HR and MAP were borderline; upon DL re-verification, only VO₂max retained significance.

Low-intensity BFR showed clearer advantages over conventional low-intensity exercise, whereas benefits were smaller versus conventional high-intensity protocols. Both BFR duration and stimulus duration showed positive dose–response relationships with effect size. However, multivariable models revealed that the apparent effect of BFR duration was confounded by the total intervention period.

In addition, no other variables exhibited significant moderating effects.</p>