Effects of exercise interventions on 24-h continuous glucose profiles in non-diabetic populations: A systematic...

Emerging evidence suggests that glycemic variability may be relevant to metabolic health, even in people without diabetes. This study aims to assess the effects of exercise on 24-h continuous glucose profiles in the non-diabetic population.

Adhering to PRISMA guidelines, PubMed, Web of Science, Embase, and CNKI databases were searched up to August 25, 2025. The meta-analysis was performed on the 24-h mean blood glucose (MBG), mean amplitude of glycemic excursion (MAGE), peak blood glucose (MAX), 4-6-h incremental area under the glucose curve (iAUC), and time in range (TIR). Risk of bias was assessed using RoB2 and ROBINS-I V2. Subgroup analyses considered population, age, study design, and exercise type.

Twenty-three studies (18 RCTs, five non-RCTs; 408 participants) were included. Exercise significantly reduced MBG (weighted mean difference [WMD] = -0.17 mmol/L), MAGE (WMD = -0.27 mmol/L), MAX (WMD = -0.63 mmol/L), and 4-6-h iAUC (standardized mean difference [SMD] = -0.60), while increasing TIR (WMD = 6.22%), compared to non-exercise or usual activity controls. Other outcomes showed no significant changes. Subgroup analyses indicated that age influenced MBG, while population and exercise type affected MAGE. Risk of bias showed 38.9% RCTs as low risk, 61.1% RCTs with some concerns, and all non-RCTs at moderate risk.

Acute exercise produced minor, parameter-specific improvements in certain CGM metrics (MBG, MAGE, MAX, TIR, and 4-6-h iAUC). These modest changes indicate subtle physiological adaptations. Larger, longer, and better-controlled trials are needed to determine their clinical relevance.