Abstract
BACKGROUND: Carotid-intima media-thickness predicts cardiovascular events and informs mechanistic research on cardiovascular diseases (CVD). However, CVD research remains Eurocentric despite etiological differences across ancestries. Incorporating Asian populations- who face substantial CVD burden with distinct etiological landscape- can enhance our understanding of cIMT biology and subclinical processes linked to CVD. This study aimed to elucidate methylation-based mechanisms of cIMT through DNA methylation profiling integrated with multi-omics data and clinically informative cIMT thresholds, leveraging an Asian cohort to enhance discovery. METHODS: We conducted an epigenome-wide association study (EWAS) of cIMT using peripheral blood DNA methylation at ~850,000 CpG sites in the Asian Health for Life in Singapore (HELIOS) cohort (n=1,357), followed by targeted trans-ancestry meta-analysis with European cohorts (overall n=2,765). Causal inference analyses (summary data-based Mendelian Randomisation [SMR] and colocalisation) evaluated methylation-mediated effects on cIMT, CVD and proximal gene expression. We derived a methylation risk score (MRS) and tested its association with cIMT thresholds indicative of elevated cardiovascular risk (≥75(th) percentile for age, sex and ethnicity). RESULTS: Three novel CpG-cIMT associations were identified (P<9.35E-07). Causal analyses supported cg08227773 methylation-mediated effects on both coronary artery disease risk (P(SMR)=2.91E-05, coloc PP.H4 =0.91) and NBEAL2 (Neurobeachin-like 2) expression (P(SMR)=9.13E-08, coloc PP.H4=0.69), a gene implicated in immune dysregulation. MRS of cIMT aggregating the three sentinel CpGs was associated with clinically-informative cIMT elevation (Odds Ratio=2.75 for Q4 vs Q1, 95% CI: 1.47-5.13). CONCLUSIONS: Through Asian-led discovery, this study identifies three novel DNA methylation markers for cIMT that are linked to cIMT elevation above clinically meaningful risk thresholds. Causal inference analyses suggest methylation-mediated CAD risk via NBEAL2 regulation, nominating biologically relevant targets while underscoring the need for larger multi-omics resources to refine mechanisms.