Higher SHR associated with increased mortality in critically ill patients with AKI: BMC

A new study published in the journal of BMC Nephrology revealed that a U-shaped link between high stress hyperglycemia ratio (SHR) and death in critically sick patients with AKI.

As a prognostic indicator in critical illness, the stress hyperglycemia ratio, which measures acute glucose rise compared to baseline, has drawn interest. Elevated SHR may be associated with increased all-cause mortality in patients with acute kidney injury (AKI), indicating its potential use in predicting risk and directing glycemic control during acute treatment. The prognostic accuracy of SHR for in-hospital and intensive care unit (ICU) mortality in patients with acute kidney injury was examined in this research.

From the Medical Information Mart for Intensive Care (MIMIC-IV) database, this research examined 4,418 AKI patients. SHR was computed by dividing entry blood glucose (mg/dl) by (28.7 × HbA1c(%) −46.7). Based on SHR levels, patients were divided into 4 groups. Kaplan-Meier survival analysis was used to examine mortality rates stratified by SHR groups, and restricted cubic splines (RCS) and Cox proportional hazards models were used to evaluate the relationship between SHR and mortality. The consistency of these correlations across clinically relevant groups was assessed by subgroup analysis. Predictive models were constructed using machine learning methods (logistic regression, Boruta feature selection, random forest).

The in-hospital and intensive care unit death rates among 4,418 patients (36.03% female) were 6.88% and 4.48%, respectively. SHR and mortality were shown to have a U-shaped connection (nonlinear p = 0.015 for in-hospital mortality, p = 0.048 for ICU mortality). Significant survival disparities across SHR groups were found by Kaplan-Meier analysis (log-rank p < 0.01).

According to Cox regression, there was a 23% increase in in-hospital mortality (HR = 1.23, 95% CI: 1.09–1.40) and a 47% increase in ICU mortality (HR = 1.47, 95% CI: 1.25–1.74) for every unit increase in SHR. AUCs for mortality prediction using machine learning models ranged from 0.76 to 0.77.

Overall, even after multivariable adjustment, SHR demonstrated a statistically significant correlation with in-hospital and ICU mortality in this sizable cohort of critically ill AKI patients. SHR is still a cheap potential variable to supplement current ICU mortality models as it may be computed from regular entry glucose and HbA1c.

Source:

Zou, D., Yin, J., Liang, Y., Li, Z., Liu, J., Ou, W., & Liu, B. (2025). Association between stress hyperglycemia ratio and all-cause mortality in patients with acute kidney injury: a retrospective analysis of the MIMIC-IV database using machine learning. BMC Nephrology, 26(1), 598. https://doi.org/10.1186/s12882-025-04523-3