Common variants at 30 loci contribute to polygenic dyslipidemia

Abstract

Blood low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol and triglyceride levels are risk factors for cardiovascular disease. To dissect the polygenic basis of these traits, we conducted genome-wide association screens in 19,840 individuals and replication in up to 20,623 individuals. We identified 30 distinct loci associated with lipoprotein concentrations (each with P < 5 x 10(-8)), including 11 loci that reached genome-wide significance for the first time. The 11 newly defined loci include common variants associated with LDL cholesterol near ABCG8, MAFB, HNF1A and TIMD4; with HDL cholesterol near ANGPTL4, FADS1-FADS2-FADS3, HNF4A, LCAT, PLTP and TTC39B; and with triglycerides near AMAC1L2, FADS1-FADS2-FADS3 and PLTP. The proportion of individuals exceeding clinical cut points for high LDL cholesterol, low HDL cholesterol and high triglycerides varied according to an allelic dosage score (P < 10(-15) for each trend). These results suggest that the cumulative effect of multiple common variants contributes to polygenic dyslipidemia.

Figure 1

Summary of genome-wide association results for LDL cholesterol, HDL cholesterol and triglycerides from stage 1. (a-c) Chromosome number versus -log₁₀ P values for LDL cholesterol (a), HDL cholesterol (b) and triglycerides (c). Green, 11 newly identified loci; blue, previously reported loci; gray, loci not subjected to follow-up; red, loci that did not replicate. (d) Quantile-quantile plot for test statistics, with observed association P values plotted as a function of expected P values. Black line, all test statistics; blue line, 19 previously reported loci excluded; green line, 11 loci confirmed in this study also excluded; gray area, 90% confidence region from a null distribution of P values (generated from 100 simulations). Blue and green lines are superimposed for triglycerides.

Figure 2

Regional plots of 11 confirmed associations. (a-k) Association results for SNPs from stage 1 (-log₁₀ P value) as a function of genomic distance (chromosomal position from National Center for Biotechnology Information build hg17) for ABCG8 (a), TIMD4-HAVCR1 (b), MAFB (c), HNF1A (d), FADS1-FADS2-FADS3 (e), LCAT (f), TTC39B (g), HNF4A (h), PLTP (i), ANGPTL4 (j) and XKR6-AMAC1L2 (k). Top three lines show genomic coverage at the locus, with each vertical tick representing directly genotyped (Illumina (Illu317k) or Affymetrix (Affy500k) platforms) or imputed SNPs. Purple diamonds indicate SNP at each locus with the strongest stage 1 association evidence. Each circle represents a SNP, with the color of the circle indicating the correlation between that SNP and the best stage 1 SNP at the locus (purple diamond). P_combined values indicate association evidence for the SNP based on the combined stage 1 and 2 data. In most cases, the best stage 1 SNP at the locus (purple diamond) was taken forward to stage 2; however, alternate SNPs (red or orange diamonds) were taken forward at two loci (HNF1A and PLTP). Light blue lines indicate estimated recombination hot spots in HapMap. Bottom panel shows genes at each locus as annotated in the UCSC Genome Browser Annotation Database as of August 5, 2007. Gray bar indicates associated interval spanning the SNP taken forward to stage 2. Associated intervals were determined as described in Methods. TG, triglyceride.

Figure 3

Mean lipoprotein concentrations and proportion of individuals with low HDL cholesterol, high LDL cholesterol or high triglycerides, as a function of allelic dosage score for HDL cholesterol, LDL cholesterol and triglycerides, respectively. Deciles of allelic dosage score are plotted on the x axis, and the y axis represents either mean lipoprotein concentrations within that decile (gray bars) or the proportion of individuals within that decile who met clinical criteria for ‘high’ or ‘low’ lipoprotein concentrations (black bars; cut points indicated in keys). Analyses were conducted for the following numbers of FHS second-generation participants: 3,206 for HDL cholesterol, 3,090 for LDL cholesterol and 3,216 for triglycerides. The allelic dosage score comprised 32 SNPs from the 30 loci shown in Table 2. For SNPs associated with HDL cholesterol, we modeled the allele associated with lower HDL cholesterol. For SNPs associated with LDL cholesterol or triglycerides, we modeled the allele associated with higher LDL cholesterol or higher triglycerides. Each of the six trends was highly significant (P < 10^-15).