Stress hyperglycemia ratio for early risk stratification of hematoma expansion in spontaneous intracerebral hemorrhage: a retrospective cohort analysis

Hematoma expansion (HE) is a major determinant of poor outcomes in spontaneous intracerebral hemorrhage (ICH). However, early and practical biomarkers for predicting HE are lacking in emergency settings. This study evaluates the predictive value of the stress hyperglycemia ratio (SHR), an easily obtainable marker, for early identification of patients at high risk for HE. This study was a retrospective cohort study involving 446 ICH patients admitted within 6 h of symptom onset at a single tertiary care center. SHR was calculated as admission blood glucose (mmol/L)/[1.59 × HbA1c (%)−2.59]. Hematoma expansion was defined as either an absolute increase in hemorrhage volume > 6 mL or > 33% relative increase in hematoma volume on follow-up CT. Feature importance was evaluated using random forest and SHAP analysis. Logistic regression, restricted cubic spline models, and subgroup analyses were used to assess the association between SHR and HE, as well as adverse in-hospital events. Among 446 patients, 51 (11.43%) experienced HE. SHR was significantly higher in the HE group (1.33 ± 0.55 vs. 1.09 ± 0.35, P = 0.004). SHR demonstrated moderate predictive ability for HE (AUC = 0.643), with an optimal cutoff of 1.3335. SHR remained independently associated with HE across multivariate models (OR = 2.80, 95% CI: 1.19–6.56). Threshold analysis showed a nonlinear relationship with an inflection point at SHR = 1.06. Higher SHR levels were also associated with longer hospital stays, increased ICU admissions, and more frequent surgical interventions. Subgroup analyses revealed stronger associations in males, patients aged 60–80, those with GCS ≥ 13, and hypertensive individuals. SHR is an independent and practical biomarker for early risk stratification of hematoma expansion in ICH. Its simplicity and accessibility support its integration into emergency clinical workflows.