J Diabetes. 2025 Dec;17(12):e70179. doi: 10.1111/1753-0407.70179.
ABSTRACT
BACKGROUND: Blood glucose levels and glycemic variability are associated with poor prognosis in critically ill patients. This study aims to investigate the relationship between average glucose (AG) and glycemic variability (GV) with 28-day mortality in patients with cardiac arrest (CA).
METHODS: This retrospective study extracted glucose measurements during ICU stay from all patients diagnosed with CA in the eICU Collaborative Research Database (version 2.0), the Medical Information Mart for Intensive Care III (version 1.4) and IV (version 3.1), and the Emergency Intensive Care Unit of Beijing Chaoyang Hospital. Multivariable Cox proportional hazards regression models and restricted cubic spline (RCS) analysis were used to explore the associations between AG, GV, and outcomes, with subgroup analyses performed to validate the findings.
RESULTS: In 6110 CA patients, RCS analysis revealed a nonlinear relationship between AG and 28-day mortality (p for nonlinear < 0.001), whereas GV exhibited a linear association (p for nonlinear = 0.058). Multivariable Cox regression analysis demonstrated that AG Q4 (AG ≥ 171.75 mg/dL), GV Q4 (GV ≥ 35.8%), and Group 4 (AG ≥ 139.7 and GV ≥ 25.8%) were associated with increased 28-day death risk (AG Q4: Hazard Ratio [95% Confidence Interval] 1.90 [1.69, 2.14], p < 0.001; GV Q4: 1.45 [1.30, 1.62], p < 0.001; Group 4: 1.80 [1.61, 2.00], p < 0.001), consistent across subgroup analyses.
CONCLUSIONS: Our study showed that higher AG and GV were associated with mortality in CA patients. Furthermore, AG and GV were independent predictors of 28-day mortality in these patients.
PMID:41354643 | DOI:10.1111/1753-0407.70179