Simultaneous assessment of genetic and epigenetic contributions to plasma lipid levels with respect to cardiovascular risk

Abstract

BackgroundThis study aims to develop a model for simultaneously assessing genetic and epigenetic contributions to plasma lipid levels. MethodsThe predictive model was developed using two cardiovascular risk groups, i.e., individuals with high low-density lipoprotein cholesterol (LDL-C) levels ([≥]160 mg/dl, N = 296) and coronary artery disease (CAD) (N = 315), in contrast to reference (max N = 3,801) and non-CAD individuals (N = 164). For genetic predisposition, rare pathological variants in five target genes related to familial hypercholesterolemia (FH) were screened, while common variants were characterized to calculate a polygenic risk score (PRS). The methylation risk score (MRS) was also calculated for epigenetic profiles based on DNA methylation levels at 13 CpG sites. A relationship between these variables and lipid levels was analyzed in regression and quantile models. ResultsA total of 17 rare FH-related gene variants were identified in patients with high LDL-C or CAD, significantly more prevalent than in the general Japanese population (2.8% vs. 0.2%, P <1x10-15). For the rare variants plus PRS, the predictability of individual LDL-C increased (correlation coefficient between predicted and measured values, r = 0.261, P = 1.7x10-11) compared to PRS alone (r = 0.151, P = 1.2x10-4). PRS and MRS had the most significant impact on high-density lipoprotein cholesterol and triglycerides, respectively. The two risk scores had additive effects on these traits. ConclusionsOur results provide proof-of-concept that assessing the relative contribution of genetic predisposition and DNA methylation levels (reflecting past environmental exposures) may help individuals refine their dyslipidemia treatment.