Acute effects of exercise snacks on postprandial glucose and insulin metabolism in adults with obesity: a systematic review and meta-analysis

Yuanbo ChangHai WangXinbi ZhangHaiyuan Liu^*

• Capital University of Physical Education and Sports, Beijing, China

Objective: To quantify the acute effects of brief, frequent interruptions to prolonged sitting ("exercise snacks") on postprandial glucose and insulin in adults with obesity, and to explore potential effect modifiers. Data Sources and Methods: Following PRISMA 2020, seven sources (PubMed, Web of Science, Cochrane Library, Embase, ClinicalTrials.gov, ICTRP, CINAHL) were searched to July 10, 2025. Randomized crossover or parallel trials in adults with obesity comparing activity breaks with uninterrupted sitting were included. Co-primary outcomes were glucose and insulin incremental area under the curve (iAUC); secondary outcomes were total AUC (tAUC) and mean levels. Random-effects meta-analyses synthesized standardized or mean differences (95% CI); heterogeneity was quantified by I². Prespecified subgroup and sensitivity analyses were undertaken; small-study effects were examined when k≥10. Risk of bias was appraised with Cochrane RoB 2.0. (Fixed-effect estimates were additionally inspected when heterogeneity was low to moderate.) Results: Seventeen trials (261 unique participants; predominantly randomized crossover) were included. Versus uninterrupted sitting, activity breaks reduced glucose iAUC (SMD = –0.49, 95% CI –0.85 to –0.14; I² = 76%) and reduced insulin iAUC (SMD = –0.26, 95% CI –0.50 to –0.03; I² = 44%). Glucose tAUC and mean glucose showed non-significant downward trends. Mean insulin decreased (SMD = –0.54, 95% CI –0.97 to –0.10), albeit with high heterogeneity (I² = 76%). Exploratory subgroup analyses suggested larger effects with higher-frequency (≤30-min) and short-bout (≤3-min) interruptions and with walking or simple resistance, although tests for subgroup differences were generally non-significant. Meta-regressions showed age predicted glucose iAUC, BMI and interruption frequency predicted mean insulin, no moderator predicted insulin iAUC, and intervention intensity (daily MET) had minimal, non-significant effects. Findings were robust in leave-one-out and model-assumption sensitivity analyses, with no clear small-study effects for glucose outcomes. Conclusions: In adults with obesity, interrupting sitting about every ≤30 minutes with 2–5 minutes of light-to-moderate walking or simple resistance acutely attenuates postprandial glucose and insulin responses. These findings support exercise "snacks" as a pragmatic behavioral strategy, while longer-term randomized trials are needed to define durability and refine dose parameters. Registration: PROSPERO CRD420251144139.