Sci Rep. 2026 May 28. doi: 10.1038/s41598-026-54404-4. Online ahead of print.
ABSTRACT
We aimed to study the relationship between brain network measures and cognitive performance in this population, focusing on hub regions. A sub-cohort of young stroke survivors (ages 18-49) with confirmed cerebral ischemia from the ODYSSEY study underwent MRI and neuropsychological assessments at baseline (n = 60) and follow-up (n = 46) up to 2 years, the discovery cohort. Additionally, a validation cohort of young stroke survivors with confirmed cerebral ischemia who had baseline standard MRI protocol and neuropsychological assessment (n = 423), as well as follow-up neuropsychological assessment (n = 288), was included for validation analysis. We used diffusion tensor imaging (DTI) based connectivity matrices for graph analysis. Lesion impact scores (combining affected voxel percentage and mean betweenness centrality) and rich club scores (quantifying affected voxels in rich club areas) were calculated using a normative brain atlas derived from DTI data from 23 stroke-free controls. Participants were categorized having no/mild or major vascular cognitive disorder (VCD) and group differences were examined. Among 60 participants (median age: 39.2 years (IQR 27.9-46.2) and 52% women), 20 were classified as having major VCD. The major VCD group exhibited larger lesion volumes (p = 0.01), lower global efficiency (p = 0.03) and local efficiency (p = 0.05) compared to the no/mild VCD group at baseline and follow-up. However, after adjusting for network density in sensitivity analyses, these differences in global and local efficiency were no longer statistically significant. Univariable logistic regression analyses revealed that the Lesion Impact Score were a significant predictor for VCD at follow-up and the Rich Club Score predicted VCD at baseline and follow-up. However, in multivariable logistic regression, both the Lesion Impact Score and the Rich Club Score did not retain predictive significance. Following validation analysis, no predictive values were observed for any of these scores. Our findings indicate a significant association between brain network measures and cognitive function in young stroke survivors, indicating a role of network disruption in post-stroke cognitive impairment. However, our study did not reveal specific associations with hub regions.
PMID:42204300 | DOI:10.1038/s41598-026-54404-4
