Predicting Early Hematoma Enlargement in Spontaneous Intracerebral Hemorrhage: Hypocalcemia, Hyperglycemia, and CT Signs

Spontaneous intracerebral hemorrhage (sICH) is a severe, stroke subtype, and early hematoma enlargement (HE) strongly predicts mortality and disability. This study aimed to investigate the predictive value of hypocalcemia, hyperglycemia, and computed tomography (CT) imaging signs for early HE in sICH. A total of 238 sICH patients admitted to the First People's Hospital Affiliated to Shanghai Jiao Tong University from November 2021 to November 2024 were retrospectively included. Patients were divided into the HE group (n = 103) and non-HE group (n = 135) based on HE occurrence. Univariate and multivariate logistic regression analyses were performed to identify independent risk factors for HE. Receiver operating characteristic (ROC) curves were used to evaluate the predictive efficiency of biochemical indicators [serum calcium (Ca²⁺), blood glucose (GLU)] and CT signs (blend sign, black hole sign, island sign), and a combined prediction model was constructed. HE patients showed significantly lower Ca²⁺ levels (1.92 mmol/L vs. 2.41 mmol/L, P < 0.001) and higher GLU levels (8.38 mmol/L vs. 6.31 mmol/L, P < 0.001) compared with non-HE patients. Multivariate analysis revealed that hypocalcemia (≤2.24 mmol/L; OR=0.001, P < 0.001), hyperglycemia (≥7.96 mmol/L; OR=1.791, P = 0.006), and expansion-prone hematoma CT signs (OR=2.990, P < 0.001) were independent risk factors for HE. ROC curve analysis indicated that the combined prediction model of Ca²⁺ (AUC=0.853), GLU (AUC=0.752), and CT signs (AUC=0.791) achieved the optimal efficiency for HE prediction (AUC=0.959, sensitivity=98.1 %, specificity=98.4 %). The combination of hypocalcemia (≤2.24 mmol/L), hyperglycemia (≥7.96 mmol/L), and expansion-prone hematoma CT signs significantly improves the accuracy of predicting early HE in sICH, providing a basis for clinical personalized treatment. These findings are of great significance for optimizing risk stratification and intervention strategies in sICH.