Abstract Number: Esoc2026a1525 Prognostic Value of Stress Hyperglycaemia Metrics in Acute Ischaemic Stroke

Abstract Background and aims Stress hyperglycaemia is frequent in acute ischaemic stroke (AIS) and has been associated with poor outcomes. We aimed to compare Stress hyperglycaemia ratio (SHR) metrics and evaluate their prognostic value afer AIS. Methods Retrospective study including patients with AIS admitted to a stroke unit (2021-2023). SHR was assessed using multiple metrics: SHR0, admission plasma glucose/(28.7 × HbA1c) − 46.7; SHR1, fasting plasma glucose (FPG)/HbA1c; SHR2, admission random plasma glucose (RPG)/HbA1c; SHR3, FPG/(1.59 × HbA1c) − 2.59; and SHR4, admission RPG/(1.59 × HbA1c) − 2.59. The glycaemic gap (admission RPG − (1.59 × HbA1c) − 2.59 ) and hyperglycaemia during the first 48 hours (any glucose ≥155 mg/dL or ≥8.5 mmol/L) were recorded. Primary outcome was poor outcome at 3 months (mRS ≥3). Variable selection was performed using LASSO regression, followed by multivariate regression models. Discriminative performance was evaluated using receiver operating characteristic curves and area under the curve (AUC). Results 559 patients were included, 34.3% with poor outcomes. All SHR metrics were significantly associated with poor outcome (p 0.001), with modest discrimination (AUC 0.56–0.64). SHR1 and SHR4 showed best performance. In multivariable models, SHR1 and SHR4 remained independent predictors (adjusted OR ≈6; p 0.001), with early hyperglycaemia. Overall model discrimination was good (AUC ≈0.83), with stronger associations in non-diabetic patients. Conclusions Stress hyperglycaemia is independently associated with poor outcomes after AIS. SHR1 and SHR4 provide the most robust prognostic information, particularly in patients without diabetes, and improve risk stratification when incorporated into multivariable models. Conflict of interest All authors: nothing to disclose.