Short-term glycemic variability and the risk of adverse ICU outcomes in critically ill patients: a systematic review and meta-analysis

Objective To quantify associations between short-term glycemic variability (GV) metrics and multidimensional adverse outcomes in critically ill patients. Methods We searched PubMed, EMBASE, and Web of Science from inception to August 16, 2025. We included observational studies of adult ICU patients reporting associations between short-term GV and adverse outcomes. Random-effects models were used for all meta-analyses. Where feasible, effect estimates were standardized to a relative risk (RR) comparing the highest versus lowest quartiles of GV. Results We included 36 studies (123,911 patients), and 25 were meta-analyzed. ICU mortality was associated with standard deviation (SD; RR = 2.29, 95% CI 1.71–3.07) and mean absolute glucose change (MAG; RR = 2.24, 95% CI 1.19–4.23). Hospital mortality was associated with coefficient of variation (CV, RR = 1.39, 95% CI 1.05–1.85) and SD (RR = 2.26, 95% CI 1.19–4.30). 28/30-day mortality was associated with CV (RR = 1.34, 95% CI 1.10–1.63) and mean amplitude of glycemic excursions (MAGE; RR = 2.05, 95% CI 1.52–2.77), and MAGE also predicted 90-day mortality (RR = 2.90, 95% CI 1.96–4.30). Furthermore, each unit increase in SD predicted higher infection risk (OR = 1.02, 95% CI 1.01–1.04) but not neurological adverse events (OR = 1.23, 95% CI 0.91–1.66). Conclusion Short-term GV is a robust predictor of mortality across different follow-up windows and clinical settings, as well as infection-related outcomes. The current findings support using short-term GV as a key prognostic marker to complement mean glucose in intensive care. Systematic review registration https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD420251114266 , identifier PROSPERO (CRD420251114266).