ABSTRACT NUMBER: ESOC2026A433 MODIFIABLE RISK FACTORS INFLUENCE BRAIN HEALTH AT ADMISSION IN ACUTE ISCHEMIC STROKE

Abstract Background and aims Brain health impacts resilience to neurological diseases and is associated with favorable functional outcomes after ischemic stroke (IS). However, the relationship between modifiable vascular risk factors and structural brain health at IS onset has not been explored. Methods We analyzed patients surviving IS with available admission MRI from the GASROS cohort. We quantified brain health using the MRI-based biomarker effective Reserve (eR), a latent variable derived from age, white matter hyperintensity, and normal-appearing brain volume, which is automatically calculated from admission T2-FLAIR sequences. Data on modifiable risk factors were recorded at acute hospital admission. In multivariable regression models, we investigated the associations between brain health and (1) modifiable risk factors; (2) Life’s Essential 8 (LE8) sub scores, and (3) LE8 total score, adjusting for sex. Results We analyzed 511 patients surviving IS (median age 66 [IQR 55-76] years, 34% female). Baseline characteristics included diagnoses of diabetes (20%), hypertension (63%), history of smoking (65%), body mass index (BMI, median: 27 [IQR 24-30]) and NIHSS (median: 2 [IQR 1-6]). Hypertension (aOR [95% confidence interval], 0.50 [0.40-0.64]), lower eGFR (per 10 mL/min/1.73m2 increase, 1.30 [1.24-1.37]), and lower BMI (1.05 [1.03-1.08]) were associated with worse brain health at admission (Figure 1). Calculating LE8 sub scores, higher blood pressure, higher blood glucose, and lower BMI were associated with worse brain health. We observed no association between the total LE8 score and brain health. Conclusions Modifiable vascular risk factors are associated with MRI markers of structural brain health at admission in patients with ischemic stroke. Conflict of interest EL is supported by Swedish Research Council (2023-06531), the Swedish Heart Lung Foundation (no. 20230904, 2024135125). AC: nothing to disclose. LA: nothing to disclose. KA: nothing to disclose. NSR is supported by NINDS U19NS115388 and the MGB AI Cures Grant Award. MDS is supported by AI Cures Grant Award, Heitman Stroke Foundation, and NIA R21AG083559. Figure 1 - belongs to Results