Effectiveness and safety of blood flow restriction training for patients with type 2 diabetes mellitus: a systematic review with meta analysis

Blood flow restriction training (BFRT) enables low-load exercise to elicit adaptations comparable to high-load training, offering a potentially viable option for individuals with type 2 diabetes mellitus (T2DM) who have limited tolerance to conventional exercise. Evidence regarding its incremental effectiveness and safety in this population remains incomplete. Systematic review and meta-analysis following PRISMA 2020 guidelines. Eight databases were searched from inception to April 2026. Studies enrolling adults with T2DM were eligible. A two-stream framework was applied: long-term studies (≥ 8 weeks) informed effectiveness and acute single-session studies informed safety, analysed independently. The primary meta-analysis was restricted to active exercise comparators. Hedges’ g was calculated using a random-effects model, with subgroup analyses by training modality and GRADE certainty assessment. Fourteen articles representing 12 independent studies ( n = 555) were included. BFRT produced a significant small reduction in total cholesterol compared with active exercise controls (g = -0.30, 95% CI -0.58 to -0.02; p = 0.039). Overall effects on HbA1c and fasting plasma glucose were non-significant, though exploratory subgroup analyses suggested training modality as a possible moderator. The low-intensity aerobic BFRT yielded moderate effects on HbA1c (g = -0.55; p = 0.003) and fasting plasma glucose (g = -0.67; p = 0.009), whereas low-load resistance BFRT showed no meaningful glycaemic advantage. Given the few studies per subgroup, these modality-specific findings require cautious interpretation. No serious adverse events were recorded and acute haemodynamic responses remained within safety thresholds. Certainty of evidence ranged from low to moderate. BFRT may confer modest incremental metabolic benefits over active exercise in T2DM, with a significant effect observed for total cholesterol and modality-dependent effects on glycaemic outcomes. No serious adverse events were observed under supervised conditions, although the limited evidence base precludes definitive safety characterisation and larger trials with standardised protocols and extended follow-up are required.