Resistance training and diabetes mellitus type 2: An umbrella review of systematic reviews and meta-analyses on glycemic and cardiometabolic outcomes

OBJECTIVES Resistance training (RT) is recommended for type 2 diabetes (T2D), but secondary evidence is voluminous and heterogeneous. We conducted an umbrella review to map, appraise, and grade the certainty of RT effects on glycaemic and cardiometabolic outcomes in adults with T2D. METHODS We searched PubMed, Scopus, and Web of Science for systematic reviews/meta-analyses of RT in adults with T2D. The studies were screened, extracted, mapped primary-study overlap (corrected covered area, CCA), and appraised review quality (AMSTAR-2). Certainty was judged at the overview level using a GRADE-style framework integrating review methods (AMSTAR-2). RESULTS Forty-three reviews met inclusion. Overlap was moderate to very high across outcomes (e.g. , CCA HbA1c 9.8%, fasting glucose 21.3%, blood pressure 33.4%, strength 35.3%, HRQoL 56.5%). RT versus non-exercise was associated with small-to-moderate improvements in HbA1c (~-0.3 to -0.6%-points), fasting glucose (~-0.5-1.4 mmol/L), and systolic blood pressure (~-4-7 mmHg); small reductions in triglycerides and LDL/total cholesterol; mixed HDL findings; and large gains in muscle strength, although the relevant evidence base showed substantial overlap across reviews. Effects on body weight/BMI were small or null; fat% and waist circumference tended to decrease modestly; physical HRQoL showed small benefits. Heterogeneity was substantial for several outcomes. AMSTAR-2 flagged frequent critical issues (e.g., use of quality scales instead of domain-based risk-of-bias; infrequent small-study bias assessment). GRADE certainty was moderate for HbA1c, LDL-C, triglycerides, and systolic BP; low-moderate for fasting glucose, diastolic blood pressure; and low for fasting insulin, HDLC, and mental/overall HRQoL. CONCLUSIONS RT likely confers small-to-moderate metabolic and vascular benefits and large strength improvements in adults with T2D, but certainty is tempered by heterogeneity, trial overlap, and recurrent methodological shortcomings. Well-reported, prospectively registered syntheses and trials with standardized RT dose and domain-based bias assessment are needed.