Coronary artery disease, genetic risk and the metabolome in young individuals

Abstract

Background: Genome-wide association studies have identified genetic variants associated with coronary artery disease (CAD) in adults - the leading cause of death worldwide. It often occurs later in life, but variants may impact CAD-relevant phenotypes early and throughout the life-course. Cohorts with longitudinal and genetic data on thousands of individuals are letting us explore the antecedents of this adult disease. Methods: 148 metabolites, with a focus on the lipidome, measured using nuclear magnetic resonance ( ¹H-NMR) spectroscopy, and genotype data were available from 5,907 individuals at ages 7, 15, and 17 years from the Avon Longitudinal Study of Parents and Children (ALSPAC) cohort. Linear regression was used to assess the association between the metabolites and an adult-derived genetic risk score (GRS) of CAD comprising 146 variants. Individual variant-metabolite associations were also examined. Results: The CAD-GRS associated with 118 of 148 metabolites (false discovery rate [FDR] < 0.05), the strongest associations being with low-density lipoprotein (LDL) and atherogenic non-LDL subgroups. Nine of 146 variants in the GRS associated with one or more metabolites (FDR < 0.05). Seven of these are within lipid loci: rs11591147 PCSK9, rs12149545 HERPUD1-CETP, rs17091891 LPL, rs515135 APOB, rs602633 CELSR2-PSRC1, rs651821 APOA5, rs7412 APOE-APOC1. All associated with metabolites in the LDL or atherogenic non-LDL subgroups or both including aggregate cholesterol measures. The other two variants identified were rs112635299 SERPINA1 and rs2519093 ABO. Conclusions: Genetic variants that influence CAD risk in adults are associated with large perturbations in metabolite levels in individuals as young as seven. The variants identified are mostly within lipid-related loci and the metabolites they associated with are primarily linked to lipoproteins. Along with further research, this knowledge could allow for preventative measures, such as increased monitoring of at-risk individuals and perhaps treatment earlier in life, to be taken years before any symptoms of the disease arise.

Keywords: ALSPAC; Coronary artery disease; childhood and adolescence; genetics; metabolomics.

Figure 1.. The distribution of the coronary artery disease genetic risk score amongst the individuals (N = 5,907) in the study.

The score was made from 146 common genetic variants that associated with coronary artery disease in adults. Each variant was weighted by the effect size of its association with the disease.

Figure 2.. Association between 148 metabolites and a coronary artery disease genetic risk score.

QQ-plot where each dot represents an association between one of the 148 metabolites and the genetic risk score comprised of 146 common variants. 98 of the 148 metabolites were lipoproteins and put into six groups based on size and density ( Supplementary Table 3). The “other” group contains the rest of the 50 metabolites. LDL = low-density lipoprotein, VLDL = very low-density lipoprotein, HDL = high-density lipoprotein, Other = non-lipoprotein metabolites.

Figure 3.. The association between 98 lipoprotein measures split into six subgroups and a coronary artery disease genetic risk score.

The lipoproteins were organised into six groups based on size and density ( Supplementary Table 3). LDL = low-density lipoprotein, VLDL = very low-density lipoprotein, HDL = high-density lipoprotein.

Figure 4.. The association between all SNPs and all metabolites.

Each metabolite was regressed against each SNP and the P values from these analyses are presented here. Grey indicates a high P value and red a low one. The lipoproteins were grouped as previously ( Supplementary Table 3).