Polygenic risk for white matter hyperintensities is associated with early cerebrovascular events partly through hemodynamic measures in cognitively unimpaired middle-aged and older adults with low cardiovascular risk.

Abstract

White matter hyperintensities (WMH) are a hallmark of cerebrovascular disease. They are often found in middle-aged individuals and are associated with a greater risk of stroke and vascular dementia. Although traditional cardiovascular risk factors are linked to WMH, some individuals with low vascular risk according to conventional scales still show WMH burden, suggesting an increased vulnerability. This study aimed to elucidate the biological mechanisms underlying the presence of WMH in cognitively unimpaired (CU) middle-aged and older individuals with low cardiovascular risk. We included 1,072 CU participants from the ALFA study with a low cardiovascular risk profile for late-life dementia (CAIDE-I score 9). We defined a multi-stage exploratory study design to reveal the potential biological pathways driving WMH in the study sample. First, we estimated the genetic predisposition to WMH using polygenic scoring (PRS) and used this score as a predictor of: (a) WMHV as a subclinical quantitative measure of global and regional WMH burden and (b) pathological WMH levels (pathological: Fazekas score 2), as a qualitative measure of clinically relevant WMH. Covariate-adjusted Spearman's rank correlation tests evaluated the association between the PRS and regional and global WMH volumes (WMHV), while a logistic regression model was performed to explore the association with pathological WMH. Second, group-stratified partial correlations (CAIDE-specific factors) were explored to identify homogeneous groups with persistent genetic associations with WMH, beyond the presence of cardiovascular risk factors. Third, an enrichment analysis of the PRS-annotated genes unveiled the biological mechanisms leading to WMH burden. Finally, based on the enrichment analysis, we examined the role of cardiometabolic traits as biomarkers of WMHV. Genetic predisposition to WMH was associated with larger global and regional WMHV after adjusting for age and sex, specifically in frontal areas. In this group, larger WMHV were associated with poorer executive function. Group-stratified analyses showed significant correlations particularly among older participants, those with hypercholesterolemia and those with lower educational attainment. Gene-set enrichment involved vascular, neuronal and cellular pathways, and blood pressure measurements partially mediated the association between the genetic risk for WMH and the actual WMHV. These findings support a polygenic contribution to cerebrovascular burden and nominate cardiac function as a biological link along the heart-brain axis. While the PRS is not yet clinically actionable, our results propose and prioritize hemodynamic monitoring as an early, testable intervention in genetically susceptible individuals, to help prevent cerebrovascular damage and downstream cognitive impairment in healthy participants with low cardiovascular risk profile.