Frontiers in Public Health · Nov 10, 2026
Structured exercise and glycemic state transitions in adults with prediabetes: a systematic review and meta-analysis of randomized controlled trials
By Zixia Liu, Jiangping Wu, Zhou Zongkai et al.
Background and objective: Prediabetes is an intermediate hyperglycemic state between normal glucose regulation (NGR) and type 2 diabetes mellitus (T2DM) in which bidirectional state transitions can occur. Previous exercise reviews have mostly focused on continuous metabolic markers such as glycated hemoglobin and fasting plasma glucose, whereas diabetes-prevention reviews have mostly evaluated multicomponent lifestyle interventions jointly comprising diet, exercise, and weight management. The effects of structured exercise on return to NGR and progression from prediabetes to T2DM remain unclear. Methods: We included randomized controlled trials (RCTs) comparing structured exercise with eligible control conditions in adults with prediabetes. The primary outcomes were return to NGR and progression to T2DM. Continuous metabolic markers were assessed as supportive outcomes. We searched nine bibliographic databases and two trial registries through 21 August 2026. When appropriate, random-effects meta-analyses summarized dichotomous outcomes as risk ratios (RRs) and continuous outcomes as mean differences (MDs), with 95% confidence intervals (CIs). Time-to-event hazard ratios (HRs) were reported separately. Risk of bias and evidence certainty were assessed using RoB 2 and GRADE, respectively. Results: Nine independent randomized trials, represented by 23 reports, were included. Eight trials contributed to at least one quantitative synthesis, and one was synthesized narratively only. For return from prediabetes to NGR, four trials involving 222 participants yielded a pooled RR of 2.04 (95% CI 0.95–4.35). The confidence interval included the null, and the certainty of evidence was very low. For progression from prediabetes to T2DM, four trials involving 1,616 participants yielded a pooled RR of 0.58 (95% CI 0.42–0.79), with low-certainty evidence. Three HRs reported by two trials were below 1 and directionally consistent with the fixed-follow-up RRs but were not pooled. The certainty of evidence was very low for FPG, 2hPG, fasting insulin, and HOMA2-IR. HbA1c was reported by only one study and was not included in the final formal GRADE Profile. Conclusion: The effect of structured exercise on return to NGR remains uncertain because the evidence is of very low certainty. Low-certainty evidence suggests that structured exercise may reduce progression from prediabetes to T2DM.