Glucose in the first 24 hours of ICU admission and 90-day mortality in patients with acute kidney injury

Introduction: Acute kidney injury in the intensive care unit (ICU) is associated with high mortality, but the prognostic impact of early glucose levels remains unclear. Objective: To analyze the association between early ICU glucose levels and 90-day mortality in patients with acute kidney injury according to diabetic status. Materials and methods: This analytical study used clinical data derived from the Medical Information Mart for Intensive Care (MIMIC) database and included 2011 complete-case patients with acute kidney injury. The exposure was the glucose value available for the first 24 hours after ICU admission, and the outcome was 90-day mortality (n = 601). Survival was assessed using Kaplan-Meier curves and the log-rank test. Adjusted Cox models were fitted, non-linear associations were explored using cubic splines, and interaction with diabetes was evaluated. Results: Glucose showed a non-linear association with mortality in the spline model (global p < 0.001), with significant interaction with diabetes (p < 0.001). The curve was compatible with increased risk at extreme glucose values. Among non-diabetic patients, 300 mg/dL was associated with higher mortality compared with 120 mg/dL (HR 1.41; 95% CI 1.05–1.90). Among diabetic patients, 80 mg/dL was associated with higher risk compared with 120 mg/dL (HR 1.49; 95% CI 1.21–1.84). Kaplan-Meier curves showed significant survival differences across glucose categories (log-rank p < 0.001). Conclusions: In ICU patients with acute kidney injury, extreme early glucose values were associated with higher 90-day mortality, with differences according to diabetic status.