The authors developed a global polygenic risk score (PRS) for white matter hyperintensities (WMH), an MRI marker of cerebral small vessel disease (cSVD), using summary statistics from the largest available European-ancestry WMH genome-wide association study (GWAS; N = 46,944). They optimized the score in 15,320 UK Biobank participants with MRI (UKB-MRI; mean age 67 ± 7.7 years). The best-performing global-PRS showed a strong association with WMH in UKB-MRI (P = 1.78 × 10^-193) and increased the explained variance in WMH (delta-R-squared = +4.1%).
The optimized global-PRS was then tested in independent cohorts spanning the lifespan and clinical settings: young adults from i-Share (N = 1,578; mean age ≈22), older community-dwellers from Three City-Dijon (N = 1,443; mean age ≈73), and memory-clinic patients from Memento (N = 1,831; mean age ≈71). The score predicted WMH in each cohort with reported P-values and incremental R-squared estimates in the original report, demonstrating cross-cohort validity.
To interrogate biological mechanisms, the team constructed 3,794 pathway-specific PRS (ps-PRS). These were defined from canonical pathways annotated in the Molecular Signatures Database (v2023.2). Each ps-PRS aggregated GWAS-derived genetic effects restricted to genes within a given pathway, enabling pathway-level genetic risk profiling for WMH.
Using a permutation-based enrichment approach in the UKB-MRI sample, the investigators identified 127 ps-PRS that were consistently enriched for association with WMH. These 127 pathways grouped into ten biological domains based on clustering analyses. The report highlights enrichment across domains including extracellular matrix, myelination, membrane transport (noted in the GWAS context), and additional pathways later detailed in secondary analyses such as lipid metabolism and ciliogenesis.
Replication was assessed in three external cohorts. Of the 127 enriched ps-PRS, 61 (comprising six clusters and 55 individual pathways) showed enrichment in at least one follow-up cohort: 14 pathways replicated in older community-dwellers, 17 in young adults, and 37 in memory-clinic patients. This differential replication suggests some pathway contributions to WMH are age- or context-specific and that certain mechanisms may be more prominent in clinical populations.
In analyses using up to 355,180 UK Biobank participants, the WMH global-PRS associated with increased risk of incident stroke and dementia outcomes. Specifically, higher global-PRS was associated with incident stroke overall (reported hazard ratio and P-value) and with both ischemic stroke and intracerebral hemorrhage. The global-PRS was also associated with incident all-cause dementia and with vascular or mixed dementia subtype, but it did not show an association with clinically defined Alzheimer’s disease in these analyses.
Secondary pathway-level analyses identified four ps-PRS implicated across age groups and clinical outcomes; these were involved in lipid metabolism, ciliogenesis, and signal transduction, and were enriched in both young and older adults and associated with stroke and dementia. In the memory-clinic cohort, some ps-PRS—most notably pathways related to sphingolipid metabolism—were also associated with dementia.
Furthermore, seven ps-PRS, primarily lipid-related pathways, showed evidence of interaction with hypertension, suggesting that traditional vascular risk factors may modulate the influence of genetic pathway-specific risk on WMH burden.
The study introduces a validated global WMH-PRS that links genetic risk for WMH to clinical complications of cSVD such as stroke and dementia. The pathway-specific framework generates candidate biological targets by identifying pathways consistently enriched for WMH and differentially associated across age groups and clinical settings. These pathway signals—particularly those involving lipid metabolism, ciliogenesis, signal transduction, and sphingolipid pathways—may inform future research into preventive strategies and therapeutic development tailored to mechanistic subtypes of cSVD.
The article is a preprint and has not undergone peer review; the authors caution against using the findings to guide clinical practice pending formal evaluation. Data that support the findings are available on request from the corresponding author but are not publicly available because of participant privacy constraints. The report provides cohort-level sample sizes, cohort descriptions, and statistical outcomes for the principal findings; additional methodological and supplemental materials are referenced in the original preprint.
Overall, this work validates a global genetic risk score for WMH and presents a comprehensive pathway-specific PRS approach that uncovers candidate biological mechanisms underlying cSVD, with translational implications for risk stratification and target discovery.