Abstract Number: Esoc2026a2319 Multifactorial Risk Models for Ischemic Stroke Subtypes in Men and Women

Abstract Background and aims Individual risk factors for ischemic stroke (IS) are known, but their cumulative, gender-specific impact on IS subtypes is unclear. We developed multifactorial models to identify key risk combinations for atherothrombotic, cardioembolic, and lacunar subtypes in men and women aged 46-75. Methods This retrospective study compared IS patients (n = 384) with a comparison group of patients with chronic cerebral ischemia (n = 289), all aged 46-75. Subtypes were classified per TOAST criteria. Multifactorial logistic regression with stepwise exclusion built models, with quality assessed by ROC curve analysis. Results Models revealed distinct gender-specific risk combinations. For men, atherothrombotic IS was associated with carotid atherosclerosis, a history of Transient Ischemic Attack (TIA), and high-grade arterial hypertension (AH); cardioembolic IS with ischemic heart disease, atrial fibrillation (AF), TIA, and high-grade AH; and lacunar IS with previous IS and high-grade AH. For women, Diabetes Mellitus (DM) was a key differentiator for atherothrombotic (TIA, carotid atherosclerosis, DM, high-grade AH) and lacunar (DM, previous IS, high-grade AH) subtypes. Cardioembolic IS in women was linked to a history of myocardial infarction, AF, TIA, and high-grade AH. The models demonstrated high predictive accuracy (AUC: men = 0.922; women = 0.927). Conclusions This study identifies distinct, gender-specific risk factor combinations for IS subtypes. These non-invasive models provide a valuable tool for stratifying patient risk, enabling personalized primary and secondary stroke prevention. Conflict of interest Kostiantyn Stepanchenko and Alla Zhidkova nothing to disclose