BACKGROUND: Epigenetic modifications have been linked to atherosclerotic cardiovascular disease and could constitute therapeutic targets. OBJECTIVES: In this study, we sought to identify differentially methylated CpG sites related to atherosclerosis across vascular beds, and to investigate to what extent these epigenetic signatures reflect cardiovascular risk factors (CVRFs). METHODS: Blood DNA methylation at 767,735 CpG sites was investigated in 3,688 individuals from 2 prospective cohort studies to create a comprehensive atlas of epigenetic modifications in carotid, coronary, and peripheral atherosclerosis. Atherosclerosis was objectively assessed by medical diagnoses, ultrasonography-assessed carotid plaque presence, and ankle-brachial index. Epigenome-wide association studies (EWAS) with 5% false discovery rate correction were performed to explore the relationship between DNA methylation patterns and atherosclerosis, as well as with CVRFs. Methylation scores were trained with the use of ridge regression for each atherosclerosis phenotype in one cohort, and evaluated for discriminatory and prognostic ability in a separate cohort. The influence of CVRFs on atherosclerosis-associated methylation was investigated by: 1) quantification of overlapping CpG sites associated with CVRFs in EWAS; and 2) observing the effect of CVRF trait adjustment on estimates and P values. RESULTS: Totals of 1,687, 3,131, and 5,852 CpG sites were significantly associated with, respectively, carotid, coronary, and peripheral atherosclerosis; 2,155 sites were significantly associated in 2 or more settings. The most strongly associated CpG sites across phenotypes mapped to 4 loci-an intergenic region on chromosome 2 (near ALPP/ALPG), AHRR, PRSS23, and F2RL3-with additional strong signals in ABCG1 and DHCR24 in coronary atherosclerosis. Epigenetic scores were predictive of 3-point major adverse cerebrovascular and cardiovascular events (HRs ranging from 1.23 to 1.39; all P 90% of atherosclerosis-associated CpG sites intersected with CVRF-associated sites, particularly smoking (up to 90%), followed by inflammation (60%) and metabolic traits (44%). Atherosclerosis EWAS adjustment for smoking pack-years alone reduced the median absolute estimate of significant CpG sites by 19.6% to 29.0%, and joint adjustment for all CVRF markers by 25.5% to 32.8%. CONCLUSIONS: We identified novel and validated previously known associations of DNA methylation with 3 subtypes of atherosclerosis. Results suggest that in blood, the epigenetic signature of atherosclerosis largely reflects cumulative exposure, particularly smoking and cardiometabolic dysregulation, rather than strongly distinguishing vasculature-specific biological processes.
“A single blood assay therefore can provide a durable readout of cumulative inflammatory and cardiometabolic burden, useful for exposure assessment and risk stratification.”
Ingold et al. (Mon,) studied this question.