Effect of exercise training on insulin-stimulated glucose disposal: a systematic review and meta-analysis of randomized controlled trials

Background and objectiveThe effect of exercise training on whole-body insulin sensitivity has not been systematically summarized. We aimed to summarize the data from randomized controlled trials evaluating the effect of exercise training on insulin action, in adults.SubjectsMEDLINE, EMBASE, and CENTRAL databases were searched until January 2021. Randomized controlled trials lasting >= 4 weeks, including adults, and evaluating the effect of exercise on insulin-stimulated glucose disposal measured using the hyperinsulinemic euglycemic clamp, were included.MethodsThree reviewers extracted summary data from published trials. The primary outcome was insulin-stimulated glucose disposal. Standardized weighted mean differences (SMD) in glucose disposal between intervention and control were compared. The PEDro scale was used to assess risk of bias.ResultsWe included 25 trials (36 interventions, N = 851). Exercise increased insulin-stimulated glucose disposal relative to control, SMD = 0.52 (95% confidence interval [CI]: 0.39, 0.65; p < 0.001; I-2 = 47%) without significantly suppressing hepatic glucose production. In trials without isotopic tracers, exercise increased glucose disposal (SMD = 0.63; 95% CI: 0.48, 0.77; p < 0.001, I-2 = 55%). In trials with isotopic tracers, exercise increased glucose disposal only when tracers were added to the exogenous glucose used for clamping (SMD = 0.34; 95% CI: 0.03, 0.66, p = 0.034. I-2 = 0%). In a meta-regression model including aerobic exercise, weight change, and tracer technique, only percent weight change explained between trial heterogeneity (beta = 0.069; 95% CI: 0.005, 0.013). The PEDro rating indicated relatively low risk of bias (5.8 +/- 0.22).ConclusionsExercise training for at least four weeks significantly increases insulin-stimulated glucose disposal. Weight loss maximizes the effect and may be needed to improve hepatic insulin sensitivity. Differences in tracer methodology contribute to divergent outcomes and should be considered when assessing conclusions from research examining the effect of exercise on insulin action.RegistrationPROSPERO (CRD42019124381).