Genetic Architecture of Cerebral White Matter Hyperintensities in Diverse Hispanic/Latino Adults.

Abstract

Background and objectives: Cerebral white matter hyperintensities (WMHs) on MRI are part of the spectrum of age-related brain vascular injury and are associated with increased risk of stroke and dementia. Genome-wide association studies (GWASs) conducted mostly in populations of European ancestry have identified several genetic loci. Although Hispanic/Latino adults have a greater burden of WMHs than their non-Hispanic White counterparts, they are vastly underrepresented in genetic studies. We sought to characterize the genetic architecture of WMHs in a Hispanic/Latino cohort by investigating the transferability of known WMH genetic loci and by leveraging Hispanic/Latino genetic diversity to map novel loci.

Methods: We conducted genome-wide association and admixture mapping analyses of WMH volume in a sample of 2,159 diverse Hispanic/Latino adults (mean age: 62.4 years; 66% female). We investigated associations at 27 previously identified WMH loci. To identify additional loci, we meta-analyzed our genome-wide association results with those of the largest GWASs published to date.

Results: Accounting for population differences in linkage disequilibrium, we found some evidence of transferability of 20 of the 27 known WMH loci. Owing to power limitations, we could not exclude transferability of the remaining loci. Multiancestry meta-analysis combining our Hispanic/Latino genome-wide association results with those from a GWAS of non-Hispanic White (NHW) and African American (AA) populations identified a novel locus on 12q22 (p = 1.8 × 10^-8) near NTN4 and tagged by rs10859915, which was previously associated with blood pressure and is an expression quantitative trait locus of AMDHD1. Admixture mapping identified a novel locus on 14q13.2, where higher counts of European ancestry at that locus were significantly associated with higher WMH volume (p = 4.9 x 10^-7). This locus spans an 800-kilobase region containing RALGAPA1, with known impact on neuronal function and brain development. Aggregated rare coding variants in this gene were associated with WMHs in a previous analysis of 20,719 stroke-free and dementia-free adults.

Discussion: Our study suggests that WMH loci previously identified in NHW and AA individuals are relevant to Hispanic/Latino adults. It demonstrates the power of the diverse Hispanic/Latino population to fine-map known genetic loci and discover novel ones, augmenting our understanding of the genetic architecture of cerebral WMHs.