Implications of discoveries from genome-wide association studies in current cardiovascular practice

Abstract

Genome-wide association studies (GWAS) have identified several genetic variants associated with coronary heart disease (CHD), and variations in plasma lipoproteins and blood pressure (BP). Loci corresponding to CDKN2A/CDKN2B/ANRIL, MTHFD1L, CELSR2, PSRC1 and SORT1 genes have been associated with CHD, and TMEM57, DOCK7, CELSR2, APOB, ABCG5, HMGCR, TRIB1, FADS2/S3, LDLR, NCAN and TOMM40-APOE with total cholesterol. Similarly, CELSR2-PSRC1-SORT1, PCSK9, APOB, HMGCR, NCAN-CILP2-PBX4, LDLR, TOMM40-APOE, and APOC1-APOE are associated with variations in low-density lipoprotein cholesterol levels. Altogether, forty, forty three and twenty loci have been associated with high-density lipoprotein cholesterol, triglycerides and BP phenotypes, respectively. Some of these identified loci are common for all the traits, some do not map to functional genes, and some are located in genes that encode for proteins not previously known to be involved in the biological pathway of the trait. GWAS have been successful at identifying new and unexpected genetic loci common to diseases and traits, thus rapidly providing key novel insights into disease biology. Since genotype information is fixed, with minimum biological variability, it is useful in early life risk prediction. However, these variants explain only a small proportion of the observed variance of these traits. Therefore, the utility of genetic determinants in assessing risk at later stages of life has limited immediate clinical impact. The future application of genetic screening will be in identifying risk groups early in life to direct targeted preventive measures.

Key Words: Genome-wide association studies, Cardiovascular disease, Lipids, Blood pressure

INTRODUCTION

Cardiovascular disease (CVD) is the leading cause of morbidity and mortality globally[1,2]. There is a concerted effort to reduce this disease burden, particularly that of coronary heart disease (CHD) and cerebrovascular disease in developed countries[3-5]. These range from primary preventive strategies targeted at risk factors through acute management and secondary prevention strategies[6-8]. Kahn et al[9] estimated that aggressive application of nationally recommended prevention activities for CVD would potentially add approximately 224 million quality adjusted life-years to the US adult population over the next 30 years and improve the average lifespan by at least 1.3 years.

CHD is the result of a combination of genetic and environmental factors. More than 200 risk factors have been associated with CHD and, among these low-density lipoprotein cholesterol (LDL-c) and blood pressure (BP) have been shown through randomized controlled trials to be causally related to CHD. A key factor in reducing the global burden of CVD is early prediction of disease to target preventive interventions. More personalised approaches to CVD prevention are attracting increasing interest. Whilst biomarkers and quantitative traits have been extremely useful in targeting primary prevention, the recent advances in genomics offer a smart option for predicting future risk of disease very early in life using the invariant nature of a genotype throughout an individual’s life-span. For example, Cohen et al[10] demonstrated that a genetic variant resulting in a modest 28% reduction in LDL-c from birth results in an 88% reduction in the risk of CHD. Over the last 5 years, genome-wide association studies (GWAS) have revolutionised the discovery of common genetic variants associated with a range of diseases and traits.

There are three key characteristics of a genetic variant that determine its impact on the phenotype studied - (1) the frequency of the variant; (2) the effect size of the variant on the phenotype; and (3) the number of genetic variants acting on the phenotype. The “common disease common variant” hypothesis (CD:CV) is the model invoked to explain how genes influence common traits such as lipids, coronary artery disease (CAD) and BP[11]. This model proposes, using an evolutionary paradigm, that common disease is due to allelic variants with a frequency greater than 5% in the general population and small individual effect size[12]. The CD:CV framework requires population-wide genotyping of very large numbers of common genetic variants (Single Nucleotide Polymorphisms/SNPs) to determine which variants show significant association with the phenotype studied. Technological advances now allow reliable and high-throughput genotyping of hundreds of thousands of SNPs on a genome-wide scale[13]. Such studies employ large scale association mapping using SNPs, making no assumptions about the genomic location or function of the causal variant, and test the hypothesis that allele frequency differs between individuals with differences in phenotype. In most GWAS, emphasis is given to the “P value” for the association of genotype with disease risk, to reduce the potential for false positive association that arises when the association of hundreds of thousands to millions of markers are tested across the whole genome. The current popular method for multiple-test correction is the frequentist approach of adjusting for a number of independent tests - based on this, a significance level of 5 × 10^-8 is commonly used, in populations of European ancestry for an overall genome-wide significance threshold of 0.05, adjusted for an estimated 1 million independent SNPs in the genome by the Bonferroni method[14]. It should be noted that the Bonferroni method is a fairly conservative correction method that may increase false negative rate. Other corrections like the False Discovery Rate or permutation testing can be used to set a different threshold. In this context, it is pertinent to recognise that the P-value is an index of a true positive signal and does not in any way reflect the predictive potential of the associated variant. The current gold standard of validity is multiple replication in independent samples. We review the implications of positive GWAS findings in current cardiovascular practice.

GWAS AND CHD

We summarise the GWAS results of CHD from nine case-control studies and three cohort studies[15-26] (Figure 1 and Table 1). The effect sizes (OR) of susceptibility alleles were modest and ranged from 1.05-2.0. Common variants in chromosome 9p21 were implicated in nine independent case-control studies[16-23,25] and in two cohort studies[15,25]. The most replicated SNPs at chromosome 9p21 were rs0757278 and rs13333049. The loci corresponding to MTHFD1L, initially identified in the Wellcome Trust Case Control Consortium (WTCCC) study[17], were later replicated in the German Family MI study[18] with genome-wide statistical significance. However, it did not reach genome-wide statistical significance in the combined analysis of ten different data sets in the study by Kathiresan et al[21]. Genetic loci corresponding to CELSR2, PSRC1 and SORT1 on chromosome 1p13.3 are identified in three independent studies[18,20,21].

Table 1 Single Nucleotide Polymorphisms associated with coronary heart disease in genome-wide association studies.

Chromosome	SNP	Position	Sample size	MAF (%)	OR (95% CI)	Pvalue	Proximal gene	Ref.
1	rs646776	109 620 053	9746/9746	81.0	1.17 (1.11-1.24)		CELSR2	[18,20,21]
	rs599839	109 623 689	2801/4582	-	1.29 (1.18-1.40)	4.05 × 10-9	PSRC1
	rs599839	109 623 689	1926/2938	80.8	1.20 (1.10-1.31)	1.30 × 10-5	SORT1
1	rs11206510	55 268 627	25 5381	81.0	1.15 (1.10-1.21)		PCSK9	[21]
1	rs17465637	220 890 152	9746/9746	72.0	1.13 (1.08-1.18)		MIA3	[18,21]
			2801/4582	-	1.20 (1.12-1.30)	1.27 × 10-6
2	rs6725887	203 454 130	9746/9746	14.0	1.17 (1.11-1.23)		WDR12	[21]
2	rs2943634	226 776 324	2801/4582	37/32	1.21 (1.03-1.30)	1.60 × 10-7	Intergenic	[18]
3	rs9818870	139 604 812	19 407/21 366	17.3/15.4	1.15 (1.11-1.19)	7.44 × 10-13	MRAS	[23]
6	rs12526453	13 035 530	25 5381	65.0	1.12 (1.08-1.17)		PHACTR1	[21]
6	rs6922269	151 294 678	2801/4582	30.0/26.0	1.23 (1.15-1.33)	2.90 × 10-8	MTHFD1L	[18]
	rs6922269	151 294 678	1926/2938	29.4/25.3	1.17 (1.04-1.32)	1.50 × 10-5
62	rs2048327	160 783 522	4976/4383	4.1/2.1	1.82 (1.57-2.12)	4.20 × 10-15	SLC22A3	[22]
	rs3127599	160 827 124					LPAL2
	rs7767084	160 882 493					LPA
	rs10755578	160 889 728
9	rs10757278	22 114 477	1607/6728	51.7/45.3	1.28 (1.22-1.35)	3.60 × 10-14	CDKN2A	[15,16,18,19,21,22,23,25]
	rs10757274	22 086 055	-	25.3/20.4	1.33 (1.23-1.47)		CDKN2B
	rs1333049	22 115 503	875/1644	54.0/48.0	1.33 (1.18-1.51)	3.40 × 10-6
	rs1333049	22 115 503	1926/2938	55.4/47.4	1.47 (1.27-1.70)	1.16 × 10-13	MTAP
	rs1333049	22 115 503	12 004/28 949	-	1.24 (1.20-1.28)
	-		9746/9746	56.0	1.28 (1.23-1.33)
	rs4977574	22 088 574	-	-	-
	-		19 407/21 366	-	-
	-		33 282	-	1.20 (1.08-1.34)3
	rs1333049	22 115 503
10	rs1746048	44 095 830	9746/9746	84.0	1.14 (1.08-1.21)		CXCL12	[18,21]
	rs501120	44 073 873	2801/4582	-/-	1.33 (1.20-1.48)	9.46 × 10-8
12	rs2259816	119 919 970	19 407/21 366	37.4/35.8	1.08 (1.05-1.11)		HNF1A-C12 or f43	[23]
16	rs4329913	55 462 933	18 245	-	1.29 (1.02-1.63)3		CETP	[26]
	rs7202364	55 342 891			0.76 (0.59-0.99)3
19	rs1122608	11 024 601	25 5381	75.0	1.15 (1.10-1.20)		LDLR	[20,21]
	rs6511720	11 063 306	1926/2938	90.2	1.29 (1.10-1.52)	6.70 × 10-4
19	rs4420638	50 114 786	1926/2938	20.9	1.17 (1.08-1.28)	1.00 × 10-4	APOE/C1/C4	[20,21]
	rs4420638	50 114 786	14 365/30 576
21	rs9982601	34 520 998	25 5381	13.0	1.28 (1.23-1.33)		MRPS6	[21]
							SLC5A3
							KCNE2

¹WTCC and GerMIFS I and GerMIFS II were added to the total sample; ²Haplotype CCTC; ³Hazard ratio per allele after adjustment for age and multiple risk factors. CELSR2: Cadherin EGF LAG seven-pass G-type receptor 2; PSRC1: Proline/serine-rich coiled-coil 1; SORT1: Sortilin 1; PCSK9: Proprotein convertase subtilisin/kexin type 9; MIA3: Melanoma inhibitory activity family member 3; WDR12: WD repeat protein 12; MRAS: Ras-related protein M-Ras; PHACTR1: Phosphatase and actin regulator 1; MTHFD1L: Methylenetetrahydrofolate dehydrogenase (NADP+ dependent) 1-like; SLC22A3: Solute carrier family 22 (extraneuronal monoamine transporter), member 3; LPAL2: Lipoprotein; Lp(a)-like 2 pseudogene; LPA: Lipoprotein Lp(a); CDKN2A: Cyclin-dependent kinase inhibitor 2A; CDKN2B: Cyclin-dependent kinase inhibitor 2B; MTAP: Methylthioadenosine phosphorylase; CXCL12: Chemokine (C-X-C motif) ligand 12; HNF1A-C12: Hepatocyte nuclear factor-1 homeobox A; CETP: Cholesteryl ester transfer protein plasma; LDLR: Low density lipoprotein receptor; APOE/C1/C4: Apolipoprotein; MRPS6: Mitochondrial ribosomal protein S6; SLC5A3: Solute carrier family 5 (sodium/myo-inositol cotransporter) member 3; KCNE2: Potassium voltage-gated channel subfamily E member 2; SNP: Single nucleotide polymorphisms; MAF: Minor allele frequency; OR: Odds ratio.

Figure 1 Significant genome-wide association study findings in coronary heart disease. CELSR2: Cadherin EGF LAG seven-pass G-type receptor 2; PSRC1: Proline/serine-rich coiled-coil 1; SORT1: Sortilin 1; PCSK9: Proprotein convertase subtilisin/kexin type 9; MRAS: Ras-related protein M-Ras; MTHFD1L: Methylenetetrahydrofolate dehydrogenase (NADP+ dependent) 1-like; SLC22A3: Solute carrier family 22 (extraneuronal monoamine transporter), member 3; LPAL2: Lipoprotein, Lp(a)-like 2 pseudogene; LPA: Lipoprotein Lp(a); CDKN2A: Cyclin-dependent kinase inhibitor 2A; CDKN2B: Cyclin-dependent kinase inhibitor 2B; MTAP: Methylthioadenosine phosphorylase; CXCL12: Chemokine (C-X-C motif) ligand 12.

GWAS AND LIPIDS

Aulchenko et al[27] studied total cholesterol (TC)-associated genetic markers and identified 11 loci significantly associated with the trait (Figure 2 and Table 2): these corresponded to TMEM57, DOCK7, CELSR2, APOB, ABCG5, HMGCR, TRIB1, FADS2/S3, LDLR, NCAN and TOMM40-APOE. Many of these genes are also implicated in other lipid traits. After screening the genome for common variants associated with plasma lipids in > 100 000 individuals of European ancestry, Teslovich et al[28] identified 39 novel loci associated with TC and replicated several other loci found to be associated with lipid traits in the previous GWAS.

Table 2 Single nucleotide polymorphisms associated with total cholesterol identified through genome-wide association studies.

Chromosome	Strongest SNP	Chromosome position	Sample size	MAF (average)	β	Pvalue	Proximal gene	Ref.
1	rs10903129	25 641 524	22 550	54	0.061	5.4 × 10-10	TMEM57	[27]
1	rs1167998	62 704 220	17 346	32	-0.073	6.4 × 10-10	DOCK7	[27]
	rs108889353		19 099	32	-0.079	3.7 × 10-12		[27]
1	rs646776	109 620 053	17 441	22	-0.128	8.5 × 10-22	CELSR2	[27]
1	rs12027135	25 648 320	> 100 000	45	-1.22	4.0 × 10-11	LDLRAP1	[28]
1	rs7515577	92 782 026	> 100 000	21	-1.18	3.0 × 10-8	EVI5	[28]
1	rs2642442	219 040 186	> 100 000	48	-1.36	5.0 × 10-14	IRF2BP2	[28]
2	rs693	21 085 700	22 500	52	-0.096	8.7 × 10-23	APOB	[27]
2	rs6756629	43 918 594	17 472	92	0.145	1.5 × 10-11	ABCG5	[27]
2	rs7570971	135 554 376	> 100 000	34	1.25	2.0 × 10-8	RAB3GAP1	[28]
3	rs2290159	12 603 920	> 100 000	22	-1.42	4.0 × 10-9	RAF1	[28]
5	rs384662	35 421 429	20 873	44	0.092	2.5 × 10-19	HMGCR	[27]
	rs12916	74 692 295	> 100 000	39	2.84	9.0 × 10-47		[28]
5	rs6882076	156 322 875	> 100 000	35	-1.98	7.0 × 10-28	TIMD4	[28]
6	rs3177928	32 520 413	> 100 000	16	2.31	4.0 × 10-19	HLA	[28]
6	rs2814982	34 654 538	> 100 000	11	-1.86	5.0 × 10-11	C6orf106	[28]
6	rs9488822	116 419 586	> 100 000	35	-1.18	2.0 × 10-10	FRK	[28]
7	rs12670798	21 573 877	> 100 000	23	1.43	9.0 × 10-10	DNAH11	[28]
7	rs2072183	44 545 705	> 100 000	25	2.01	3.0 × 10-11	NPC1L1	[28]
8	rs6987702	126 573 908	17 413	29	0.073	3.3 × 10-9	TRIB1	[27]
8	rs2081687	59 551 119	> 100 000	35	1.23	2.0 × 10-12	CYP7A1	[28]
8	rs2737229	116 717 740	> 100 000	30	-1.11	2.0 × 10-8	TRPS1	[28]
10	rs2255141	113 923 876	> 100 000	30	1.14	2.0 × 10-10	GPAM	[28]
11	rs174570	61 353 788	20 916	83	0.088	1.5 × 10-10	FADS2/3	[27]
11	rs10128711	18 589 560	> 100 000	28	-1.04	3.0 × 10-8	SPTY2D1	[28]
11	rs7941030	122 027 585	> 100 000	38	0.97	2.0 × 10-10	UBASH3B	[28]
12	rs11065987	110 556 807	> 100 000	42	-0.96	7.0 × 10-12	BRAP	[28]
12	rs1169288	119 901 033	> 100 000	33	1.42	1.0 × 10-14	HNF1A	[28]
16	rs2000999	70 665 594	> 100 000	20	2.34	3.0 × 10-24	HPR	[28]
19	rs2228671	11 071 912	20 910	88	0.158	9.3 × 10-24	LDLR	[27]
19	rs2304130	19 650 528	20 914	7	-0.153	2.0 × 10-15	NCAN	[27]
19	rs2075650	50 087 459	17 463	15	0.138	2.9 × 10-19	TOMM40-APOE	[27]
	rs157580	50 087 106	20 903	33	-0.09	5.1 × 10-17		[27]
19	rs10401969	19 268 718	> 100 000	7	-4.74	3.0 × 10-38	CILP2	[28]
19	rs492602	53 898 229	> 100 000	49	1.27	2.0 × 10-10	FLJ36070	[28]
20	rs2277862		> 100 000	15	-1.19	4.0 × 10-10	ERGIC3	[28]
20	rs2902940	38 524 901	> 100 000	29	-1.38	6.0 × 10-11	MAFB	[28]

TMEM57: Transmembrane protein 57; DOCK7: Dedicator of cytokinesis 7; CELSR2: Cadherin, EGF LAG seven-pass G-type receptor 2; LDLRAP1: Low density lipoprotein receptor adaptor protein 1; EVI5: Ecotropic viral integration site 5; IRF2BP2: Interferon regulatory factor 2 binding protein 2; APOB: Apolipoprotein B; ABCG5: ATP-binding cassette sub-family G member 5; RAB3GAP1: RAB3 GTPase activating protein subunit 1 (catalytic); RAF1: v-raf-1 murine leukemia viral oncogene homolog 1; HMGCR: 3-hydroxy-3-methylglutaryl-CoA reductase; TIMD4: T-cell immunoglobulin and mucin domain containing 4; HLA: Human leukocyte antigen (HLA) complex; C6orf106: Chromosome 6 open reading frame 106; FRK: Fyn-related kinase; DNAH11: Dynein, axonemal, heavy chain 11; NPC1L1: NPC1 (Niemann-Pick disease, type C1, gene)-like 1; TRIB1: Tribbles Homolog-1 (Trib1); CYP7A1: Cytochrome P450, family 7, subfamily A, polypeptide 1; TRPS1: Trichorhinophalangeal syndrome 1; GPAM: Glycerol-3-phosphate acyltransferase; mitochondrial; FADS: Fatty acid desaturase; SPTY2D1: Suppressor of Ty, domain containing 1 (S. cerevisiae); UBASH3B: Ubiquitin associated and SH3 domain containing B; BRAP: BRCA1 associated protein; HNF1A: Hepatocyte nuclear factor-1 α; HPR: Haptoglobin-related protein; LDLR: Low density lipoprotein receptor; NCAN: Neurocan; TOMM40: Translocase of outer mitochondrial membrane 40 homolog; CILP2: Cartilage intermediate layer protein 2; ERGIC3: Endoplasmic reticulum-Golgi intermediate compartment protein 3; MAFB: V-maf musculoaponeurotic fibrosarcoma oncogene homolog B; SNP: Single nucleotide polymorphisms; MAF: Minor allele frequency.

Figure 2 Significant genome-wide association study findings in total cholesterol. TMEM57: Transmembrane protein 57; DOCK7: Dedicator of cytokinesis 7; CELSR2: Cadherin, EGF LAG seven-pass G-type receptor 2; LDLRAP1: Low-density lipoprotein receptor adaptor protein 1; EVI5: Ecotropic viral integration site 5; IRF2BP2: Interferon regulatory factor 2 binding protein 2; APOB: Apolipoprotein B; ABCG5: ATP-binding cassette sub-family G member 5; RAB3GAP1: RAB3 GTPase activating protein subunit 1 (catalytic); RAF1: V-raf-1 murine leukemia viral oncogene homolog 1; HMGCR: 3-hydroxy-3-methylglutaryl-CoA reductase; TIMD4: T-cell immunoglobulin and mucin domain containing 4; HLA: Human leukocyte antigen (HLA) complex; C6orf106: Chromosome 6 open reading frame 106; FRK: Fyn-related kinase; DNAH11: Dynein, axonemal, heavy chain 11; NPC1L1: NPC1 (Niemann-Pick disease; type C1, gene)-like 1; TRIB1: Tribbles Homolog-1 (Trib1); CYP7A1: Cytochrome P450, family 7, subfamily A, polypeptide 1; TRPS1: Trichorhinophalangeal syndrome 1; GPAM: Glycerol-3-phosphate acyltransferase, mitochondrial; FADS: Fatty acid desaturase; SPTY2D1: Suppressor of Ty, domain containing 1 (S. cerevisiae); UBASH3B: Ubiquitin associated and SH3 domain containing B; BRAP: BRCA1 associated protein; HNF1A: Hepatocyte nuclear factor-1 α; HPR: Haptoglobin-related protein; LDLR: Low-density lipoprotein receptor; NCAN: Neurocan; TOMM40: Translocase of outer mitochondrial membrane 40 homolog; CILP2: Cartilage intermediate layer protein 2; ERGIC3: Endoplasmic reticulum-Golgi intermediate compartment protein 3; MAFB: V-maf musculoaponeurotic fibrosarcoma oncogene homolog B.

Prior to the publication of the meta-analysis of blood lipids conducted by Teslovich et al[28], 29 loci had been found to be associated with variation in high-density lipoprotein cholesterol (HDL-c) levels[20,27-39]. Teslovich et al[28] identified 31 novel loci associated with HDL-c with genome-wide significance. The most commonly-replicated loci are LPL, LIPC, CETP, ABCA1, LIPG, APOA1/C3/A4/A5 and GALNT2 (Figure 3 and Table 3). The LIPC locus has a set of common variants nearly 50 kb upstream of the gene, strongly associated with HDL-c and appearing to be independent of previously described variants that overlap the transcribed sequence of the gene. SNPs close to the mevalonate kinase-methylmalonic aciduria cblB type (MMAB) locus were found to be associated with HDL-c initially by Willer et al[20] and later confirmed by Kathiresan et al[29].

Table 3 Single nucleotide polymorphisms associated with high-density lipoprotein cholesterol identified through genome-wide association studie.

Chromosome	Strongest SNP	Chromosome position	Sample size	MAF (average)	Change in HDLc/β	Pvalue	Proximal gene	Ref.
1	rs2144300	228 361 539	8656	40	-	6.6 × 10-7	GALNT2	[20]
	rs4846914	228 362 314	19 794	40	-0.05 SD	4.0 × 10-8		[30]
1	rs4660293	39 800 767	> 100 000	23	-0.48	4.0 × 10-10	PABPC4	[28]
1	rs1689800	180 435 508	> 100 000	35	-0.47	3.0 × 10-10	ZNF648	[28]
2	rs1260326	27 584 444	16 682	41	0.93%	< 5 × 10-8	GCKR	[31]
2	rs6754295	21 059 688	17 915	25	2.63 (z-sc)1	4.4 × 10-8	APOB	[27]
2	rs2972146	226 808 942	> 100 000	37	0.46	3.0 × 10-9	IRS1	[28]
2	rs10490964	51 926 908	18 245	12	1.35 mg/dL	3.9 × 10-9	COBLL1, GRB14	[26]
	rs12328675	165 249 046	> 100 000	13	0.68	3.0 × 10-10	COBLL1	[28]
4	rs13107325	103 407 732	> 100 000	7	-0.84	7.0 × 10-11	SLC39A8	[28]
5	rs6450176	53 333 782	> 100 000	26	-0.49	5.0 × 10-8	ARL15	[28]
6	rs2814944	34 660 775	> 100 000	16	-0.49	4.0 × 10-9	C6orf106	[28]
6	rs605066	139 871 359	> 100 000	42	-0.39	3.0 × 10-8	CITED2	[28]
6	rs1084651	161 009 807	> 100 000	16	1.95	3.0 × 10-8	LPA	[28]
7	rs4731702	130 083 924	> 100 000	48	0.59	1.0 × 10-15	KLF14	[28]
8	rs2083637	19 909 455	17 922	26	4.14 (z-sc)1	5.5 × 10-18	LPL	[27]
	rs10503669	19 891 970	8656	10		3.2 × 10-10		[20]
	rs331	19 864 685	6382	28	1.5 mg/dL	9.1 × 10-7		[32]
	rs17482753	19 876 926	8180	-		2.8 × 10-11		[33]
	rs326	19 863 719	10 536	22-30		1.8 × 10-8		[35]
	rs331	19 864 685	6382	28	1.5 mg/dL	9.1 × 10-7		[32]
	rs12678919	19 888 502	19 794	10	0.23 SD	2.0 × 10-34		[30]
	rs301	19 861 214	5592	25	0.04	9.3 × 10-11		[36]
8	rs3916027	19 869 148	5592	27	0.04	5.4 × 10-10	SLC18A1	[36]
8	rs331	19 864 685	16 809	27	0.43%	< 5 × 10-8	Intergenic, PPP1R3B, LPL	[31]
	rs9987289	9 220 768	> 100 000	9	-1.21	6.0 × 10-25	PPP1R3B	[25]
8	rs2293889	116 668 374	> 100 000	0.41	-0.44	6.0 × 10-11	TRPS1	[28]
9	r1323432	103 402 758	8656	12	1.93 mg/dL	2.5 × 10-8	GRIN3A	[20]
9	rs3905000	106 696 891	17 913	14	-4.37 (z-sc)1	8.6 × 10-13	ABCA1	[27]
	rs4149268	106 687 041	8656	36		3.3 × 10-7		[20]
	rs3890182	106 687 476	21 312	-		3.0 × 10-10		[29]
	rs9282541	106 660 656	10 536	0-9		4.8 × 10-8		[35]
	rs2515614	106 724 139	16 798	34	0.20%	< 5 × 10-8		[31]
	rs1883025	106 704 122	19 371	26	-0.08 SD	1.0 × 10-9		[30]
9	rs471364	15 279 578	40 414	12	-0.08 SD	3.0 × 10-10	TTC39B	[29]
	rs581080	15 295 378	> 100 000	18	-0.65	3.0 × 10-12		[28]
9	rs1883025	106 704 122	> 100 000	25	-0.94	2.0 × 10-33	ABCA1	[28]
11	rs12225230	116 233 840	6382	18	1.5 mg/dL	5.3 × 10-5	APOA1/C3/A4/A5	[32]
	rs618923	116 159 369	12 111	25	0.30%	< 5 × 10-8		[31]
	rs964184	116 154 127	19 794	14	-0.17 SD	1.0 × 10-12		[30
	rs7350481	116 091 493	8993	28	0.62%	8.8 × 10-10		[36]
	rs7350481	116 091 493	18 245			2.8 × 10-12		[26]
11	rs2923084	10 345 358	> 100 000	17	-0.41	5.0 × 10-8	AMPD3	[28]
11	rs3136441	46 699 823	> 100 000	15	0.78	3.0 × 10-18	LRP4	[28]
11	rs7395662		17 917	39	2.82 (z-sc)1	6.0 × 10-11	MADD-FOLH1	[27]
11	rs174547	61 327 359	40 330	33	-0.09 SD	2.0 × 10-12	FADS1-S3	[30]
11	rs6589565	116 145 447	5592		-0.05	4.4 × 10-7	BUD13	[36]
11	rs2075290	116 158 506	5592	7	-0.05	4.2 × 10-7	ZNF259	[36]
12	rs2338104	108 379 551	8656	45		1.9 × 10-6	MVK/MMAB	[20]
	rs2338104	108 379 551	19 793	45	-0.07 SD	1.0 × 10-10		[30]
	rs7134594	108 484 574	> 100 000	47	-0.44	7.0 × 10-15		[28]
12	rs7134375	20 365 025	> 100 000	42	0.40	4.0 × 10-8	PDE3A	[28]
12	rs4759375	122 362 191	> 100 000	6	0.86	7.0 × 10-9	SBNO1	[28]
12	rs4765127	123 026 120	> 100 000	34	0.44	3.0 × 10-10	ZNF664	[28]
12	rs838880	123 827 546	> 100 000	31	0.61	3.0 × 10-14	SCARB1	[28]
12	rs1818702	102 047 685	16 844	29	0.22%	< 5 × 10-8	Intergenic, ASCL1, PAH	[31]
15	rs1532085	56 470 658	19 736	41	5.03 (z-sc)1	9.7 × 10-36	LIPC	[27]
	rs4115041	121 186 681	8656	33		2.8 × 10-9		[20]
	rs1532085	56 470 658	6382	37	1.8 mg/dL	1.3 × 10-10		[32]
	rs1800588	56 510 967	21 312	-		2.0 × 10-32		[29]
	rs11858164	56 530 023	10 536	27-55		7.0 × 10-8		[35]
	rs1532085	56 470 658	6382	37	1.8 mg/dL	1.3 × 10-10		[32]
	rs1800588	56 510 967	16 811	22	0.60%	< 5 × 10-8		[31]
	rs10468017	56 465 804	19 794	30	0.10 SD	8.0 × 10-23		[30]
	rs1077834	56 510 771	5987	49	1.00%	1.3 × 10-14		[36]
	rs1077834	56 510 771	18 245			1.4 × 10-23		[26]
	rs261342	56 518 445	5592	22	0.03	6.3 × 10-8		[36]
	rs1532085	56 470 658	> 100 000	39	1.45	3.0 × 10-96		[28]
15	rs2652834	61 183 920	> 100 000	20	-0.39	9.0 × 10-9	LACTB	[28]
16	rs1800775	55 552 737	2623	47		2.5 × 10-13	CETP	[28]
	rs1532624	55 562 980	19 674	43	8.24 (z-sc)1	9.4 × 10-94		[27]
	rs3764261	55 550 825	8656	31	2.42 mg/dL	2.8 × 10-19		[20]
	rs3764261	55 550 825	6382	31	4.0 mg/dL	1.0 × 10-41		[32]
	rs1800775	55 552 737	2758	49	2.6 mg/dL	3.0 × 10-13		[29]
	rs1800775	55 552 737	1643	47	3.99 mg/dL	6.1 × 10-15		[33]
	rs9989419	55 542 640	8216	-		8.5 × 10-27		[33]
	rs7205804	55 562 390	10 536	37-50		4.7 × 10-47		[36]
	rs3764261	55 550 825	6382	31	4.0 mg/dL	1.0 × 10-41		[32]
	rs1800775	55 552 737	16 779	49	2.50%	< 5 × 10-8		[31]
	rs3764261	55 550 825	322	-	6.2 mg/dL	3.4 × 10-12		[39]
	rs3764261	55 550 825	18 245			3.7 × 10-93		[26]
	rs173539	55 545 545	19 794	32	0.25 SD	4.0 × 10-75		[30]
	rs3764261	55 550 825	5987	21	2.11%	4.8 × 10-29		[37]
	rs3764261	55 550 825	18 245	30-48		3.7 × 10-93		[27]
	rs17231506	55 552 029	5592	32	0.07	2.3 × 10-36		[36]
	rs3764261	55 550 825	> 100 000	32	3.39	7.0 × 10-380		[28]
16	rs255052	66 582 496	8656	17		1.5 × 10-6	LCAT	[20]
	rs255052	66 582 496	8656 + 4534	-		1.2 × 10-7		[20]
	rs2271293	66 459 571	31 946	11	0.07 SD	9.0 × 10-13		[30]
	rs16942887	66 485 543	> 100 000	12	1.27	8.0 × 10-33		[28]
16	rs2271293	66 459 571	17 910	13	4.99 (z-sc)1	8.3 × 10-16	CTCF-PRMT8	[27]
16	rs289743	55 575 29	5592	31	0.03	8.6 × 10-9	NLRC5	[36]
16	rs2925979	80 092 291	> 100 000	30	-0.45	2.0 × 10-11	CMIP	[28]
17	rs11869286	35 067 382	> 100 000	34	-0.48	1.0 × 10-13	STARD3	[28]
17	rs4148008	64 386 889	> 100 000	32	-0.42	2.0 × 10-10	ABCA8	[28]
17	rs4129767	73 915 579	> 100 000	49	-0.39	8.0 × 10-9	PGS1	[28]
18	rs4939883	45 421 212	16 258	17	-3.98 (z-sc)1	1.6 × 10-11	LIPG	[27]
	rs2156552	45 435 666	8656	16		8.4 × 10-7		[20]
	rs2156552	45 435 666	21 312	-		2.0 × 10-7		[29]
	rs4939883	45 421 212	16 648	16	0.22%	< 5 × 10-8		[31]
	rs4939883	45 421 212	19 785	17	-0.14 SD	7.0 × 10-15		[30]
	rs4939883	45 421 212	18 245			1.4 × 10-9		[26]
	rs7241918	45 414 951	> 100 000	17	-1.31	3.0 × 10-49		[28]
18	rs12967135	56 000 003	> 100 000	23	-0.42	7.0 × 10-9	MC4R	[28]
19	rs769449	50 101 842	16 728	12	0.30%	< 5 × 10-8	APOC1-APOE	[31]
			18 245			2.6 × 10-11		[26]
19	rs2967605	8 375 738	35 151	16	-0.12 SD	1.0 × 10-8	ANGPTL3	[30]
	rs7255436	8 339 196	> 100 000	47	-0.45	3.0 × 10-8		[28]
19	rs737337	11 208 493	> 100 000	8	-0.64	3.0 × 10-9	LOC55908	[28]
19	rs386000	59 484 573	> 100 000	20	0.83	4.0 × 10-16	LILRA3	[28]
20	rs6065906	43 987 422	16 810	48	0.40%	< 5 × 10-8	PLTP	[31]
			18 245			1.9 × 10-14		[26]
20	rs1800961	42 475 778	30 714	3	-0.19 SD	8.0 × 10-10	HNF4A	[30]
	rs1800961	42 475 778	> 100 000	3	-1.88	1.0 × 10-15		[28]
20	rs7679	44 009 909	40 248	19	-0.07 SD	4.0 × 10-9	PLTP	[30]
	rs6065906	43 987 422	> 100 000	18	-0.93	2.0 × 10-22		[28]
22	rs181362	20 262 068	> 100 000	20	-0.46	1.0 × 10-8	UBE2L3	[28]

¹z-sc: the ENGAGE consortium provided the effect size on the z-scale. GALNT2: N-acetylgalactosaminyltransferase 2; PABPC4: Poly(A) binding protein, cytoplasmic 4 (inducible form); ZNF648: Zinc finger protein 648; GCKR: Glucokinase (hexokinase 4) regulator; APOB: Apolipoprotein B; IRS1: Insulin receptor substrate 1; COBLL1: COBL-like 1; GRB14: Growth factor receptor-bound protein 14; SLC39A8: Solute carrier family 39 (zinc transporter) member 8; ARL15: ADP-ribosylation factor-like 15; C6orf106: Chromosome 6 open reading frame 106; CITED2: Cbp/p300-interacting transactivator, with Glu/Asp-rich carboxy-terminal domain 2; LPA: Lipoprotein, Lp(a); KLF14: Kruppel-like factor 14; LPL: Lipoprotein lipase; SLC18A1: Solute carrier family 18 (vesicular monoamine) member 1; PPP1R3B: Protein phosphatase 1, regulatory (inhibitor) subunit 3B; TRPS1: Trichorhinophalangeal syndrome 1; GRIN3A: Glutamate receptor, ionotropic, N-methyl-D-aspartate 3A; ABCA1: ATP-binding cassette, sub-family A (ABC1) member 1; TTC39B: Tetratricopeptide repeat domain 39B; ABCA1: ATP-binding cassette, sub-family A (ABC1) member 1; APOA1: Apolipoprotein A-I; AMPD3: Adenosine monophosphate deaminase 3; LRP4: Low density lipoprotein receptor-related protein 4; MADD-FOLH1: MAP-kinase activating death domain- folate hydrolase (prostate-specific membrane antigen) 1; FADS1-S3: Fatty acid desaturase 1; BUD13: BUD13 homolog; ZNF259: Zinc finger protein 259; MVK: Mevalonate kinase; MMAB: Methylmalonic aciduria (cobalamin deficiency) cblB type; PDE3A: Phosphodiesterase 3A; SBNO1: Strawberry notch homolog 1; ZNF664: Zinc finger protein 664; SCARB1: Scavenger receptor class B member 1; ASCL1: Achaete-scute complex homolog 1; PAH: Phenylalanine hydroxylase; LIPC: Hepatic lipase; LACTB: Lactamase β; CETP: Cholesteryl ester transfer protein plasma; LCAT: Lecithin-cholesterol acyltransferase; CTCF: CCCTC-binding factor (zinc finger protein); PRMT8: Protein arginine methyltransferase 8; NLRC5: NLR family CARD domain containing 5; STARD3: StAR-related lipid transfer (START) domain containing 3; ABCA8: ATP-binding cassette; sub-family A (ABC1) member 8; PGS1: Phosphatidylglycerophosphate synthase 1; LIPG: Lipase endothelial; MC4R: Melanocortin 4 receptor; APOC1: Apolipoprotein C-I; APOE: Apolipoprotien E; ANGPTL3: Angiopoietin-like 3; LILRA3: Leukocyte immunoglobulin-like receptor, subfamily A (without TM domain) member 3; PLTP: Phospholipid transfer protein; HNF4A: Hepatocyte nuclear factor 4 α; PLTP: Phospholipid transfer protein; UBE2L3: Ubiquitin-conjugating enzyme E2L 3; SNP: Single nucleotide polymorphisms; MAF: Minor allele frequency. HDLc: high-density lipoprotein cholesterol.

Figure 3 Significant genome-wide association study findings in high-density lipoprotein cholesterol. GALNT2: N-acetylgalactosaminyltransferase 2; PABPC4: Poly(A) binding protein; cytoplasmic 4 (inducible form); ZNF648: Zinc finger protein 648; GCKR: Glucokinase (hexokinase 4) regulator; APOB: Apolipoprotein B; IRS1: Insulin receptor substrate 1; COBLL1: COBL-like 1; GRB14: Growth factor receptor-bound protein 14; SLC39A8: Solute carrier family 39 (zinc transporter) member 8; ARL15: ADP-ribosylation factor-like 15; C6orf106: Chromosome 6 open reading frame 106; CITED2: Cbp/p300-interacting transactivator, with Glu/Asp-rich carboxy-terminal domain 2; LPA: Lipoprotein, Lp(a); KLF14: Kruppel-like factor 14; LPL: Lipoprotein lipase; SLC18A1: Solute carrier family 18 (vesicular monoamine) member 1; PPP1R3B: Protein phosphatase 1, regulatory (inhibitor) subunit 3B; TRPS1: Trichorhinophalangeal syndrome 1; GRIN3A: Glutamate receptor, ionotropic, N-methyl-D-aspartate 3A; ABCA1: ATP-binding cassette; sub-family A (ABC1) member 1; TTC39B: Tetratricopeptide repeat domain 39B; ABCA1: ATP-binding cassette, sub-family A (ABC1) member 1; APOA1: Apolipoprotein A-I; AMPD3: Adenosine monophosphate deaminase 3; LRP4: Low-density lipoprotein receptor-related protein 4; MADD-FOLH1: MAP-kinase activating death domain- folate hydrolase (prostate-specific membrane antigen) 1; FADS1-S3: Fatty acid desaturase 1; BUD13: BUD13 homolog; ZNF259: Zinc finger protein 259; MVK: Mevalonate kinase; MMAB: Methylmalonic aciduria (cobalamin deficiency) cblB type; PDE3A: Phosphodiesterase 3A; SBNO1: Strawberry notch homolog 1; ZNF664: Zinc finger protein 664; SCARB1: Scavenger receptor class B member 1; ASCL1: Achaete-scute complex homolog 1; PAH: Phenylalanine hydroxylase; LIPC: Hepatic lipase; LACTB: Lactamase β; CETP: Cholesteryl ester transfer protein plasma; LCAT: Lecithin-cholesterol acyltransferase; CTCF: CCCTC-binding factor (zinc finger protein); PRMT8: Protein arginine methyltransferase 8; NLRC5: NLR family CARD domain containing 5; STARD3: StAR-related lipid transfer (START) domain containing 3; ABCA8: ATP-binding cassette; sub-family A (ABC1) member 8; PGS1: Phosphatidylglycerophosphate synthase 1; LIPG: Lipase endothelial; MC4R: Melanocortin 4 receptor; APOC1: Apolipoprotein C-I; APOE: Apolipoprotien E; ANGPTL3: Angiopoietin-like 3; LILRA3: Leukocyte immunoglobulin-like receptor, subfamily A (without TM domain) member 3; PLTP: Phospholipid transfer protein; HNF4A: Hepatocyte nuclear factor 4 α; PLTP: Phospholipid transfer protein; UBE2L3: Ubiquitin-conjugating enzyme E2L 3.

GWAS have identified several genetic loci associated with LDL-c (Figure 4 and Table 4)[20,27-32,34,36,40], such as the study by Teslovich et al[28] which identified 22 novel and 25 previously implicated loci. CELSR2-PSRC1-SORT1 and PCSK9 loci on chromosome 1, APOB, HMGCR, NCAN-CILP2-PBX4, LDLR, TOMM40-APOE, and APOC1-APOE were the most commonly-replicated loci in LDL-c. Several of these loci were also associated with CHD in the WTCCC study[17].

Table 4 Single nucleotide polymorphisms associated with low-density lipoprotein cholesterol identified through genome-wide association studies.

Chromosome	Strongest SNP	Chromosome position	Sample size	MAF (average)	β	Pvalue	Proximal gene	Ref.
1	rs1167998	62 704 220	12 685	22	-0.155	7.8 × 10-23	PSRC1, CELSR2, SORT1	[27]
	rs646776	109 620 053	6382	22	-	4.9 × 10-19		[32]
	rs599839	109 623 689	1636	24	-	1.1 × 10-7		[34]
	rs602633	109 623 034	8589	20	-	4.8 × 10-14		[20]
	rs646776	109 620 053	16 791	22	-0.04	< 5 × 10-8	CELSR2	[31]
	rs646776	109 620 053	21 312	24	-0.16	5 × 10-42	PSRC1	[29]
	rs12740374	109 619 113	19 648	21	-0.23	2.0 × 10-42		[30]
	rs599839	109 623 689	11 685	21	-0.05	1.7 × 10-15		[40]
	rs646776	109 620 053	4337	21	-0.16	4.3 × 10-9		[40]
	rs12740374	109 619 113	5592	21	-0.15	1.8 × 10-9	SORT1	[36]
	rs646776	109 620 053	5592	22	-0.14	3.8 × 10-8		[36]
	rs629301	109 619 829	> 100 000	22	-5.65	1.0 × 10-170		[28]
1	rs11591147	55 278 235	16 826	2	-0.12	< 5 × 10-8	PCSK9	[31]
	rs11591147	55 278 235	12 167	2	-0.13	< 5 × 10-8		[31]
	rs11591147	55 278 235	21 312	1	-0.26	2.0 × 10-24		[30]
	rs11206510	55 268 627	19394			3.5 × 10-11		[20]
	rs11206510	55 268 627	19 629	19	-0.09	4.0 × 10-8		[30]
	rs11591147	55 278 235	5592		-0.55	9.3 × 10-12		[36]
	rs2479409	55 277 238	> 100 000	30	2.01	2.0 × 10-28		[28]
2	rs693	21 085 700	2601	22		7.1 × 10-7	APOB	[38]
	rs562338	21 141 826	1636 + 2631	17		8.6 × 10-13		[34]
	rs515135	21 139 562	8589	83		3.1 × 10-14		[20]
	rs562338	21 141 826	8589 + 10 849			5.6 × 10-22		[20]
	rs693	21 085 700	16 112	52	-0.098	3.6 × 10-17		[28]
	rs506585	21 250 687	6382	20	-4.6	9.3 × 10-9		[32]
	rs506585	21 250 687	16 842	20	-0.04	< 5 × 10-8		[31]
	rs693	21 085 700	21 312	48	0.12	1.0 × 10-21		[29]
	rs515135	21 139 562	19 648	20	-0.16	5.0 × 10-29		[30]
	rs562338	21 141 826	11 685	20	-0.04	1.4 × 10-9		[40]
	rs1713222	21 124 828	4337	16	-0.17	1.0 × 10-8		[40]
	rs562338	21 141 826	5592	18	-0.18	1.2 × 10-11		[36]
	rs1367117	21 117 405	> 100 000	30	4.05	4.0 × 10-114		[28]
2	rs6756629	43 918 594	12 706	92	0.157	2.6 × 10-10	ABCG5	[27]
2	rs6544713	43 927 385	23 456	32	0.15	2 × 10-20	ABCG8	[30]
	rs4299376	43 926 080	> 100 000	30	2.75	2.0 × 10-47	ABCG5/8	[28]
2	rs780094	27 594 741	16 841	40	0.03	< 5 × 10-8	GCKR	[31]
5	rs1501908	156 330 747	27 280	37	-0.07	1 × 10-11	TIMD4-HAVCR1	[30]
5	rs3846662	74 686 840	16 135	44	0.079	1.5 × 10-11	HMGCR	[27]
	rs12654264	74 684 359	2758 + 18 554	39	0.1	1.0 × 10-20		[29]
	rs3846663	74 691 482	19 648	38	0.07	8.0 × 10-12		[30]
	rs12654264	74 684 359	5592	38	0.11	5.8 × 10-8		[36]
6	rs3757354	88 570 980	> 100 000	22	-1.43	1.0 × 10-11	MYLIP	[28]
6	rs1800562	26 201 120	> 100 000	6	-2.22	6.0 × 10-10	HFE	[28]
6	rs1564348	160 498 850	> 100 000	17	-0.56	2.0 × 10-17	LPA	[28]
7	rs12670798	21 573 877	12 695	24	0.089	6.1 × 10-9	DNAH11	[27]
7	rs4731702	130 083 924	16 747	49	-0.02	< 5 × 10-8	KLF14	[31]
8	rs6982636	126 548 497	16 798	47	-0.02	< 5 × 10-8	TRIB1	[31]
8	rs11136341	145 115 531	> 100 000	40	1.4	4.0 × 10-13	PLEC1	[28]
9	rs9411489	135 144 821	> 100 000	20	2.24	6.0 × 10-13	ABO	[28]
11	rs174570	61 353 788	16 153	83	0.11	4.4 × 10-13	FADS2/3	[31]
11	rs3135506	116 167 617	16 837	6	-0.13	< 5 × 10-8	APOA1-A5	[31]
	rs2072560	116 167 036	5592	6	0.22	2.4 × 10-7		[36]
11	rs11220462	125 749 162	> 100 000	14	1.95	1.0 × 10-15	ST3GAL4	[28]
12	rs7307277	123 041 109	16 804	34	-0.02	< 5 × 10-8	CCDC92/DNAH10/ZNF664	[31]
12	rs2650000	119 873 345	39 340	36	0.07	2.0 × 10-8	HNF1A	[30]
14	rs8017377	51 667 587	> 100 000	47	1.14	5.0 × 10-1	NYNRIN	[28]
16	rs708272	55 553 789	16 843	43	-0.04	< 5 × 10-8	CETP	[31]
	rs17231506	55 552 029	5592	32	-0.11	5.0 × 10-7		[36]
17	rs7206971	42 780 114	> 100 000	49	0.78	2.0 × 10-8	OSBPL7	[28]
19	rs1699614	19 519 472	21 312	10	-0.1	3 × 10-8	NCAN, CILP2, PBX4	[29]
	rs2228603	19 190 924	8589	7		1.8 × 10-7		[20]
	rs16996148	19 519 472	19 394			2.7 × 10-9		[20]
	rs10401969	19 268 718	19 648	6	-0.05	2.0 × 10-8		[30]
19	rs688	11 088 602	4267	45		7.3 × 10-7	LDLR	[34]
	rs6511720	11 063 306	8589	9		6.8 × 10-10		[20]
	rs6511720	11 063 306	19 394			4.2 × 10-23		[20]
	rs2228671	11 071 912	16 148	82	0.136	4.2 × 10-14		[27]
	rs6511720	11 063 306	6382	12	-7.7	5.2 × 10-15		[32]
	rs6511720	11 063 306	16 843	12	-0.04	< 5 × 10-8		[31]
	rs6511720	11 063 306	21 312	10	-0.26	2 × 10-51		[29]
	rs6511720	11 063 306	19 648	10	-0.26	2.0 × 10-26		[30]
	rs2228671	11 071 912	4337	12	-0.18	1.1 × 10-8		[40]
	rs17248720	11 059 187	5592	13	-0.31	7.8 × 10-25		[36]
	rs6511720	11 063 306	> 100 000	11	6.99	4.0 × 10-117		[28]
19	rs2075650	50 087 459	12 697	15	0.16	9.3 × 10-19	TOMM40-APOE	[27]
	rs157580	50 087 106	16 160	68	-0.111	2.1 × 10-19		[27]
	rs2075650	50 087 459	4337	13	0.23	7.1 × 10-14		[40]
	rs2075650	50 087 459	5592	14	0.23	1.1 × 10-14		[36]
19	rs4420638	50 114 786	2601	22		3.4 × 10-13	APOC1-APOE	[38]
	rs4420638	50 114 786	4267	19		8.3 × 10-14		[34]
	rs4420638	50 114 786	8589	12		1.5 × 10-21		[20]
	rs4420638	50 114 786	19 394			3.0 × 10-43		[20]
	rs4803750	49 939 467	6382	7	-9.6	3.6 × 10-14		[32]
	rs4803750	49 939 467	16 616	7	-9.3	< 5 × 10-8		[32]
	rs4420638	50 114 786	21 312	20	0.19	1.0 × 10-60		[29]
	rs4420638	50 114 786	11 881	16	0.29	4.0 × 10-27		[30]
	rs4420638	50 114 786	11 685	18	0.06	1.2 × 10-20	APOC2	[40]
	rs12721046	50 113 094	5592	15	0.21	7.6 × 10-14		[36]
	rs12721109	50 139 061	5592	2	-0.54	5.1 × 10-14		[36]
	rs4420638	50 114 786	> 100 000	17	7.14	9.0 × 10-147	APOE	[28]
19	rs10402271	50 021 054	11 685	33	0.03	4.1 × 10-8	BCAM	[40]
	rs4605275	50 030 333	4337	31	-0.13	4.7 × 10-8		[40]
19	rs4803750	49 939 467	4337	7	-0.28	2.4 × 10-11	BCL3	[40]
	rs1531517	49 934 013	5592	7	-0.22	5.3 × 10-8		[36]
19	rs10402271	50 021 054	5592	33	0.15	2.1 × 10-12	PVRL2	[36]
20	rs6065906	43 987 422	16 843	48	0.02	< 5 × 10-8	PLPT	[31]
20	rs6102059	38 662 198	28 895	32	-0.06	4.0 × 10-9	MAFB	[30]
20	rs6029526	39 106 032	> 100 000	47	1.39	4.0 × 10-19	TOP1	[28]

β: Estimated mean; CELSR2: Cadherin EGF LAG seven-pass G-type receptor 2; PSRC1: Proline/serine-rich coiled-coil 1; SORT1: Sortilin 1; PCSK9: Proprotein convertase subtilisin/kexin type 9; APOB: Apolipoprotein B; ABCG: ATP-binding cassette sub-family G; GCKR: Glucokinase (hexokinase 4) regulator; TIMD4: T-cell immunoglobulin and mucin domain containing 4; HAVCR1: Hepatitis A virus cellular receptor 1; HMGCR: 3-hydroxy-3-methylglutaryl-CoA reductase; MYLIP: Myosin regulatory light chain interacting protein; HFE: Human hemochromatosis; LPA: Lipoprotein, Lp(a); DNAH11: Dynein axonemal heavy chain 11; KLF14: Kruppel-like factor 14; TRIB1: Tribbles homolog 1; PLEC1: Plectin; ABO: ABO blood group (transferase A, α 1-3-N-acetylgalactosaminyltransferase, transferase B, α 1-3-galactosyltransferase); FADS: Fatty acid desaturase; APOA1: Apolipoprotien A1; ST3GAL4: ST3 β-galactoside α-2,3-sialyltransferase 4; CCDC92: Coiled-coil domain containing 92; DNAH10: Dynein axonemal heavy chain 10; ZNF664: Zinc finger protein 664; HNF1A: HNF1 homeobox A; NYNRIN: NYN domain and retroviral integrase containing; CETP: Cholesteryl ester transfer protein plasma; OSBPL7: Oxysterol binding protein-like 7; NCAN: Nucleoporin 214 kDa; CILP2: Cartilage intermediate layer protein 2; PBX4: Pre-B-cell leukemia homeobox 4; LDLR: Low density lipoprotein receptor; TOMM40: Translocase of outer mitochondrial membrane 40 homolog; APOE: Apolipoprotien E; APOC2: Apolipoprotein C-II; APOE: Apolipoprotien E; BCAM: Basal cell adhesion molecule; BCL3: B-cell CLL/lymphoma 3; PVRL2: Poliovirus receptor-related 2 (herpesvirus entry mediator B); PLPT: Proteolipid protein; MAFB: V-maf musculoaponeurotic fibrosarcoma oncogene homolog B; TOP1: Topoisomerase (DNA) 1; SNP: Single nucleotide polymorphisms; MAF: Minor allele frequency.

Figure 4 Significant genome-wide association study findings in low-density lipoprotein cholesterol. CELSR2: Cadherin EGF LAG seven-pass G-type receptor 2; PSRC1: Proline/serine-rich coiled-coil 1; SORT1: Sortilin 1; PCSK9: Proprotein convertase subtilisin/kexin type 9; APOB: Apolipoprotein B; ABCG: ATP-binding cassette sub-family G; GCKR: Glucokinase (hexokinase 4) regulator; TIMD4: T-cell immunoglobulin and mucin domain containing 4; HAVCR1: Hepatitis A virus cellular receptor 1; HMGCR: 3-hydroxy-3-methylglutaryl-CoA reductase; MYLIP: Myosin regulatory light chain interacting protein; HFE: Human hemochromatosis; LPA: Lipoprotein, Lp(a); DNAH11: Dynein axonemal heavy chain 11; KLF14: Kruppel-like factor 14; TRIB1: Tribbles homolog 1; PLEC1: Plectin; ABO: ABO blood group (transferase A, α 1-3-N-acetylgalactosaminyltransferase, transferase B, α 1-3-galactosyltransferase); FADS: Fatty acid desaturase; APOA1: Apolipoprotien A1; ST3GAL4: ST3 β-galactoside α-2;3-sialyltransferase 4; CCDC92: Coiled-coil domain containing 92; DNAH10: Dynein axonemal heavy chain 10; ZNF664: Zinc finger protein 664; HNF1A: HNF1 homeobox A; NYNRIN: NYN domain and retroviral integrase containing; CETP: Cholesteryl ester transfer protein plasma; OSBPL7: Oxysterol binding protein-like 7; NCAN: Nucleoporin 214kDa; CILP2: Cartilage intermediate layer protein 2; PBX4: Pre-B-cell leukemia homeobox 4; LDLR: Low-density lipoprotein receptor; TOMM40: Translocase of outer mitochondrial membrane 40 homolog; APOE: Apolipoprotien E; APOC2: Apolipoprotein C-II; APOE: Apolipoprotien E; BCAM: Basal cell adhesion molecule; BCL3: B-cell CLL/lymphoma 3; PVRL2: Poliovirus receptor-related 2 (herpesvirus entry mediator B); PLPT: Proteolipid protein; MAFB: V-maf musculoaponeurotic fibrosarcoma oncogene homolog B; TOP1: Topoisomerase (DNA) 1.

In total, 43 different loci have been found to be associated with triglycerides (TAG) in GWAS (Figure 5 and Table 5). SNPs in proximity to ANGPTL3, APOB, GCKR, MLXIPL, LPL, TRIB1, APOA1/A4/A5/C3, and NCAN-CILP2-PBX4 have been associated with TAG in several GWAS.

Table 5 Single nucleotide polymorphisms associated with triglycerides identified through genome-wide association studies.

Chromosome	Strongest SNP	Studies	Sample size	MAF (average)	β	Pvalue	Proximal gene	Ref.
1	rs1167998	62 704 220	14 268	32	-0.091	2.0 × 10-12	DOCK7	[27]
	rs10889353	62 890 784	14 337	32	-0.085	8.2 × 10-11		[27]
1	rs11591147	55 278 235	16 826	2	-0.09	< 5 × 10-8	PCSK9	[31]
1	rs12042319	62 822 407	4267	34		3.2 × 10-7	ANGPTL3	[34]
	rs10889353	62 890 784	16 831	33	-0.03	< 5 × 10-8		[31]
	rs10889353	62 890 784	8993	14	-0.13	2.0 × 10-9		[37]
	rs12130333	62 964 365	21 312	22	-0.11	2.0 × 10-8		[29]
	rs10889353	62 890 784	19 834	33	-0.05	3.0 × 10-7		[30]
	rs1748195	62 822 181	18 243			1.7 × 10-10		[20]
	rs2131925	62 798 530	> 100 000	32	-4.94	9.0 × 10-43		[28]
1	rs4846914	228 362 314	21 312	40	0.08	7.0 × 10-15	GALNT2	[28]
2	rs6754295	21 059 688	14 338	25	-0.077	2.5 × 10-8	APOB	[27]
	rs673548	21 091 049	12 694	76	0.086	1.1 × 10-8		[27]
	rs673548	21 091 049	16 797	21	-0.04	< 5 × 10-8		[31]
	rs693	21 085 700	21 312	48	0.12	1.0 × 10-21		[29]
	rs7557067	21 061 717	19 840	22	-0.08	9.0 × 10-12		[30]
	rs1042034	21 078 786	> 100 000	22	-5.99	1.0 × 10-45		[28]
2	rs780094	27 594 741	2659	35		3.7 × 10-8	GCKR	[38]
	rs780094	27 594 741	4267	39		8.1 × 10-14		[34]
	rs1260326	27 584 444	8684	40		1.5 × 10-15		[20]
	rs780094	27 594 741	18 243			6.1 × 10-32		[20]
	rs780094	27 594 741	17 790	63	-0.103	3.1 × 10-20		[27]
	rs1260326	27 584 444	6382	41	0.07	1.3 × 10-16		[32]
	rs1260326	27 584 444	16 650	41	0.07	< 5 × 10-8		[31]
	rs1260326	27 584 444	8993	45	-0.101	1.1 × 10-11		[37]
	rs780094	27 594 741	21 312	34	0.13	3.0 × 10-14		[29]
	rs1260326	27 584 444	19 840	45	0.12	2.0 × 10-31		[30]
	rs1260326	27 584 444	5592	40	0.06	1.8 × 10-7		[36]
	rs1260326	27 584 444	> 100 000	41	8.76	6.0 × 10-133		[28]
2	rs10195252	165 221 337	> 100 000	40	-2.01	2.0 × 10-10	COBLL1	[28]
3	rs645040	137 409 312	> 100 000	22	-2.22	3.0 × 10-8	MSL2L1	[28]
4	rs442177	88 249 285	> 100 000	41	-2.25	9.0 × 10-12	KLHL8	[28]
5	rs9686661	55 897 543	> 100 000	20	2.57	1.0 × 10-10	MAP3K1	[28]
6	rs2076530	32 471 794	16 829	43	0.03	< 5 × 10-8	BTNL2	[31]
6	rs2247056	31 373 469	> 100 000	25	-2.99	2.0 × 10-15	HLA	[28]
7	rs13238203	71 767 603	> 100 000	4	-7.91	1.0 × 10-9	TYW1B	[28]
7	rs17145738	72 620 810	2758 + 18 554	13	-0.14	7.0 × 10-22	BCL7B, TBL2, MLXIPL	[29]
							TBL2
	rs11974409	72 627 326	5592	20	-0.08	5.7 × 10-9	MLXIPL	[36]
	rs10551921	107 998 852	5592	20	-0.08	1.3 × 10-8		[36]
7	rs2240466	72 494 205	12 680	87	0.137	1.1 × 10-12	MLXIPL	[27]
	rs11974409	72 627 326	16 839	19	-0.04	< 5 × 10-8		[31]
	rs714052	72 502 805	19 840	12	-0.16	3.0 × 10-15		[30]
	rs17145738	72 620 810	18 243			2.0 × 10-12		[20]
	rs17145738	72 620 810	> 100 000	12	-9.32	6.0 × 10-58		[28]
7	rs4731702	130 083 924	16 714	49	-0.03	< 5 × 10-8	KLF14	[31]
7	rs17145713	72 542 746	5592	20	-0.09	5.3 × 10-10	BAZ1B	[36]
8	rs11776767	10 721 339	> 100 000	37	2.01	1.0 × 10-8	PINX1	[28]
8	rs1495741	18 317 161	> 100 000	22	2.85	5.0 × 10-14	NAT2	[28]
8	rs2083637	19 909 455	14 344	26	-0.107	1.0 × 10-14	LPL	[27]
	rs10096633	19 875 201	12 708	88	0.174	1.9 × 10-18		[27]
	rs12678919	19 888 502	19 840	10	-0.25	2.0 × 10-41		[30]
	rs10096633	19 875 201	8993	12	-0.169	9.3 × 10-14		[37]
	rs331	19 864 685	5592	25	-0.08	1.7 × 10-11		[36]
	rs12678919	19 888 502	> 100 000	12	-13.64	2.0 × 10-115		[28]
8	rs17482753	19 876 926	2652	11		4.9 × 10-7	LPL	[38]
	rs17482753	19 876 926	1636	10		1.2 × 10-9		[34]
	rs17482753	19 876 926	1636 + 2631			5.2 × 10-15		[34]
	rs6993414	19 947 198	8684	46		1.4 × 10-13		[20]
	rs10503669	19 891 970	4267			3.9 × 10-22		[20]
	rs328	19 864 004	6382	11	-0.09	4.7 × 10-11	Intergenic, PPP1R3B, LPL	[32]
	rs331	19 864 685	6382	28	-0.06	1.7 × 10-9		[32]
	rs328	19 864 685	16 812	11	-0.09	< 5 × 10-8	LPL	[31]
	rs328	19 864 004	21 242	9	-0.19	2.0 × 10-28		[29]
8	rs6982636	126 548 497	16 765	47	-0.03	< 5 × 10-8	TRIB1	[31]
	rs17321515	12 655 591	21 242	49	-0.08	4.0 × 10-17		[29]
	rs2954029	126 560 154	8684	56		2.8 × 10-8		[20]
	rs17321515	12 655 591	14 176			7.0 × 10-13		[20]
	rs2954029	126 560 154	19 840	44	-0.11	3.0 × 10-19		[30]
	rs2954029	126 560 154	> 100 000	47	-5.64	3.0 × 10-55		[28]
8	rs3916027	19 869 148	5592	27	-0.08	1.0 × 10-10	SLC18A1	[36]
8	rs7819412	11 082 571	33 336	48	-0.04	3.0 × 10-8	XKR6-AMAC1L2	[30]
10	rs10761731	64 697 616	> 100 000	43	-2.38	3.0 × 10-12	JMJD1C	[28]
10	rs2068888	94 829 632	> 100 000	46	-2.28	2.0 × 10-8	CYP26A1	[28]
11	rs12272004	116 108 934	12 622	7	-0.181	5.4 × 10-13	APO (A1/A4/A5/C3)	[27]
	rs6589566	116 157 633	1636	6		1.5 × 10-11		[34]
	rs6589566	116 157 633	1636 + 2631			3.7 × 10-12		[34]
	rs964184	116 154 127	8684	12		1.5 × 10-16		[20]
	rs12286037	116 157 417	18 422			1.0 × 10-26		[20]
	rs3135506	116 167 617	6382	6	0.13	5.5 × 10-12		[32]
	rs662799	116 168 917	6382	6	0.14	2.9 × 10-15		[32]
	rs3135506	116 167 617	16 804	6	0.14	< 5 × 10-8		[31]
	rs7350481	116 091 493	8993	43	0.24	1.4 × 10-49		[37]
	rs28927680	116 124 283	21 312	7	0.26	2.0 × 10-17		[29]
	rs964184	116 154 127	19 840	14	0.3	4.0 × 10-62		[30]
	rs651821	116 167 789	5592	6	0.21	8.8 × 10-21	APOA1	[36]
	rs964184	116 154 127	> 100 000	13	16.95	7.0 × 10-240		[28]
11	rs174547	61 327 359	38 846	33	0.06	2.0 × 10-14	FADS1-S3	[30]
	rs174546	61 326 406	> 100 000	34	3.82	5.0 × 10-24		[28]
11	rs6589565	116 145 447	5592	7	0.19	4.5 × 10-20	BUD13	[36]
11	rs2075290	116 158 506	5592	7	0.19	6.6 × 10-20	ZNF259	[36]
12	rs7307277	123 041 109	16 771	34	-0.04	< 5 × 10-8	CCDC92/DNAH10/	[31]
							ZNF664
12	rs11613352	-	> 100 000	23	-2.7	4.0 × 10-10	LRP1	[28]
15	rs2412710	40 471 079	> 100 000	2	7	2.0 × 10-8	CAPN3	[28]
15	rs2929282	42 033 223	> 100 000	5	5.13	2.0 × 10-11	FRMD5	[28]
15	rs4775041	56 461 987	8684	67		7.3 × 10-5	LIPC	[20]
	rs4775041	56 461 987	17 104			1.6 × 10-8		[20]
16	Rs11649653	30 825 988	> 100 000	40	-2.13	3.0 × 10-8	CTF1	[28]
16	rs1800775	55 552 737	16 779	49	-0.03	< 5 × 10-8	CETP	[31]
19	rs157580	50 087 106	16 160	33	-0.069	1.2 × 10-8	TOMM40-APOE	[27]
	rs439401	50 106 291	11 885	68	0.086	1.8 × 10-9		[27]
19	rs16996148	19 519 472	21 312	10	-0.1	4.0 × 10-9	NCAN, CILP2, PBX4	[29]
	rs10401969	19 268 718	8684	8		2.3 × 10-7		[20]
	rs16996148	19 519 472	18 391			2.5 × 10-9		[20]
	rs17216525	46 471 516	19 840	7	-0.11	4.0 × 10-11		[30]
	rs12610185	19 582 722	5592	9	-0.1	5.6 × 10-7		[36]
19	rs439401	50 106 291	16 638	35	-0.04	< 5 × 10-8	APOC1-APOE	[31]
	rs439401	50 106 291	> 100 000	36	-5.5	1.0 × 10-30	APOE	[28]
19	rs2304128	19 607 151	5592	9	-0.1	3.2 × 10-7	GMI	[36]
20	rs6065906	43 987 422	16 810	48	0.04	< 5 × 10-8	PLPT	[31]
	rs7679	44 009 909	38 561	19	0.07	7.0 × 10-11		[30]
22	rs5756931	36 875 979	> 100 000	40	-1.54	4.0 × 10-8	PLA2G6	[28]

β: Estimated mean; DOCK7: Dedicator of cytokinesis 7; PCSK9: Proprotein convertase subtilisin/kexin type 9; GALNT2: N-acetylgalactosaminyltransferase 2; ANGPTL3: Angiopoietin-like 3; APOB: Apolipoprotien B; GCKR: Glucokinase (hexokinase 4) regulator; COBLL1: COBL-like 1; MSL2L1: Male-specific lethal 2 homolog; KLHL8: Kelch-like 8; MAP3K1: Mitogen-activated protein kinase kinase kinase 1; BTNL2: Butyrophilin-like 2 (MHC class II associated); HLA: Major histocompatibility complex; TYW1B: tRNA-yW synthesizing protein 1 homolog B; TBL2: Transducin (β)-like 2; BCL7B: B-cell CLL/lymphoma 7B; TBL2: Transducin (β)-like 2; MLXIPL: MLX interacting protein-like; KLF14: Kruppel-like factor 14; BAZ1B: Bromodomain adjacent to zinc finger domain 1B; PINX1: PIN2/TERF1 interacting, telomerase inhibitor 1; NAT2: N-acetyltransferase 2 (arylamine N-acetyltransferase); LPL: Lipoprotein lipase; PP1R3B: Protein phosphatase 1, regulatory (inhibitor) subunit 3B; TRIB1: Tribbles homolog 1; SLC18A1: Solute carrier family 18 (vesicular monoamine) member 1; XKR6: XK Kell blood group complex subunit-related family member 6; AMAC1L2: Acyl-malonyl condensing enzyme 1-like 2; JMJD1C: Jumonji domain containing 1C; CYP26A1: Cytochrome P450 family 26 subfamily A polypeptide 1; APOA1: Apolipoprotein A-I; FADS: Fatty acid desaturase; CCDC92: Coiled-coil domain containing 92; DNAH10: Dynein axonemal heavy chain 10; ZNF664: Zinc finger protein 664; LRP1: Low density lipoprotein receptor-related protein 1; CAPN3: Calpain 3, (p94); FRMD5: FERM domain containing 5; LIPC: Hepatic lipase; CTF1: Cardiotrophin 1; CETP: Cholesteryl ester transfer protein plasma; TOMM40: Translocase of outer mitochondrial membrane 40 homolog; APOE: Apolipoprotien E; NCAN: Nucleoporin 214 kDa; CILP2: Cartilage intermediate layer protein 2; PBX4: Pre-B-cell leukemia homeobox 4; GMIP: GEM interacting protein; PLPT: Palmitoyl-protein thioesterase 1; PLA2G6: Phospholipase A2, group VI (cytosolic, calcium-independent); SNP: Single nucleotide polymorphisms; MAF: Minor allele frequency.

Figure 5 Significant genome-wide association study findings in triglycerides. DOCK7: Dedicator of cytokinesis 7; PCSK9: Proprotein convertase subtilisin/kexin type 9; GALNT2: N-acetylgalactosaminyltransferase 2; ANGPTL3: Angiopoietin-like 3; APOB: Apolipoprotien B; GCKR: Glucokinase (hexokinase 4) regulator; COBLL1: COBL-like 1; MSL2L1: Male-specific lethal 2 homolog; KLHL8: Kelch-like 8; MAP3K1: Mitogen-activated protein kinase kinase kinase 1; BTNL2: Butyrophilin-like 2 (MHC class II associated); HLA: Major histocompatibility complex; TYW1B: tRNA-yW synthesizing protein 1 homolog B; TBL2: Transducin (β)-like 2; BCL7B: B-cell CLL/lymphoma 7B; TBL2: Transducin (β)-like 2; MLXIPL: MLX interacting protein-like; KLF14: Kruppel-like factor 14; BAZ1B: Bromodomain adjacent to zinc finger domain 1B; PINX1: PIN2/TERF1 interacting, telomerase inhibitor 1; NAT2: N-acetyltransferase 2 (arylamine N-acetyltransferase); LPL: Lipoprotein lipase; PP1R3B: Protein phosphatase 1, regulatory (inhibitor) subunit 3B; TRIB1: Tribbles homolog 1; SLC18A1: Solute carrier family 18 (vesicular monoamine) member 1; XKR6: XK Kell blood group complex subunit-related family member 6; AMAC1L2: Acyl-malonyl condensing enzyme 1-like 2; JMJD1C: Jumonji domain containing 1C; CYP26A1: Cytochrome P450 family 26 subfamily A polypeptide 1; APOA1: Apolipoprotein A-I; FADS: Fatty acid desaturase; CCDC92: Coiled-coil domain containing 92; DNAH10: Dynein axonemal heavy chain 10; ZNF664: Zinc finger protein 664; LRP1: Low density lipoprotein receptor-related protein 1; CAPN3: Calpain 3, (p94); FRMD5: FERM domain containing 5; LIPC: Hepatic lipase; CTF1: Cardiotrophin 1; CETP: Cholesteryl ester transfer protein plasma; TOMM40: Translocase of outer mitochondrial membrane 40 homolog; APOE: Apolipoprotien E; NCAN: Nucleoporin 214kDa; CILP2: Cartilage intermediate layer protein 2; PBX4: Pre-B-cell leukemia homeobox 4; GMIP: GEM interacting protein; PLPT: Palmitoyl-protein thioesterase 1; PLA2G6: Phospholipase A2, group VI (cytosolic; calcium-independent).

GWAS AND BP

In 2007, the Framingham Heart Study[41] reported on 1327 individuals whose BP had been sampled longitudinally in the Framingham Community project. In the same year, the WTCCC[17] reported results from 2000 Northern European subjects with HTN. Although a few SNPs did reach a statistical significance of P < 10^-5, none of them achieved genome-wide significance (P < 5 × 10^-8). The most significant GWAS findings in blood pressure are summarized in Table 6 and Figure 6 [42-50].

The global BPgen consortium[42] studied 34 433 subjects of European ancestry, subsequently followed up the findings with direct genotyping of 71 225 individuals of European ancestry and 12 889 individuals of Indian Asian ancestry and conducted a joint analysis. They identified an association between systolic or diastolic BP (SBP/DBP) and common variants in eight regions near the CYP17A1 (intergenic CNNM2/NT5C2), CYP1A2 (intron CSK), FGF5, SH2B3 (intron ATXN2), MTHFR, c10orf107, ZNF652 and intron PLCD3. Furthermore, three of these common variants (MTHFR, CYP17A1 and CYP17A2 or CSK) were associated with HTN (P < 5 × 10^-8). The CHARGE consortium study (n = 29 136) identified 13, 20 and 10 SNPs for SBP, DBP and HTN respectively[43].

In a joint meta-analysis of CHARGE consortium data with BPgen consortium data (n = 34 433)[43], four CHARGE loci attained genome-wide significance for SBP (ATP2B1, CYP17A1, PLEKHA7, SH2B3), six for DBP (ATP2B1, CACNB2, CSK-ULK3, SH2B3, TBX3-TBX5, ULK4) and one for HTN (ATP2B1). The KORA study by Org et al[48] in a South German Cohort identified a SNP upstream of T-cadherin adhesion molecule (CDH13) gene on chromosome 16 (rs11646213) as significantly associated with HTN at a genome-wide level. Finally, in a population of African origin, Adeyemo et al[44] identified four common variants (MYLIP, chr 6; YWHAZ, chr 8; IPO7, chr 11 and SLC24A4, chr 14) associated with SBP with genome-wide significance.

Wang et al[47] identified STK39, SPAK (STE20/SPS1-related proline and alanine rich kinase; a serine/threonine kinase) with a P value of 1.6 × 10^-7 in an Amish cohort. Several other studies also identified potentially important genetic loci associated with BP traits with borderline genome-wide significance. These include ATP2B1[43,51] (ATPase, Ca⁺⁺ transporting, plasma membrane 1) on chromosome 12, FOXD3[41] (fork head box D3) on chromosome 1, CCNG1 (cyclin G1)[48] on chromosome 5, BCAT1 (branched chain aminotransferase 1, cytosolic)[17] on chromosome 12, ATXN2 (ataxin 2)[42,43] on chromosome 12 and TBX3 (T-box3)[43] on chromosome 12 (Figure 6 and Table 6). However, none of these loci were replicated in other studies. Using an extreme case-control design, Padmanabhan et al[50] identified a novel HTN locus on chromosome 16 in the promoter region of uromodulin (UMOD; rs13333226, combined P value 3.6 × 10^-11). The minor G allele of this SNP is associated with a lower risk of HTN [OR (95% CI): 0.87 (0.84-0.91)], reduced urinary UMOD excretion and increased estimated glomerular filtration rate (3.6 mL/min per minor-allele, P = 0.012), and borderline association with renal sodium balance.

Table 6 Single nucleotide polymorphisms associated with hypertension and blood pressure in genome-wide association studies.

Chr	SNP	Position	Ancestry	N (discovery)	Phenotype	Risk allele	Risk allele frequency	OR/β	P	Nearest gene	Ref.
1	rs17367504	11 785 365	E	34 433	SBP	G	0.14	-0.85	2 × 10-13	MTHFR, CLCN6, NPPA, NPPB, AGTRAP	[42,43]
2	rs6749447	168 749 632	E	542	SBP	G	0.28	1.90	8 × 10-5	STK39	[47]
3	rs9815354	41 887 655	E	29 136	DBP	A	0.17	0.49	3 × 10-9	ULK4	[42,43]
4	rs16998073	81 403 365	E	34 433	DBP	T	0.21	0.50	1 × 10-21	FGF5, PRDM8, C4orf22	[42,43]
4	rs991316	100 541 468	AA	1017	SBP	T	0.45	1.62	5 × 10-6	ADH7	[44]
10	rs11014166	18 748 804	E	29 136	DBP	A	0.66	0.37	1 × 10-8	CACNB2	[42,43]
10	rs1530440	63 194 597	E	34 433	DBP	T	0.19	-0.39	1 × 10-9	C10orf107, TMEM26, RTKN2, RHOBTB1, ARID5B, CYP17A1	[42,43]
10	rs1004467	104 584 497	E	29 136	SBP	A	0.90	1.05	1 × 10-10	TMEM26, RTKN2, RHOBTB1, ARID5B, CYP17A1	[42,43]
10	rs11191548	104 836 168	E	34 433	SBP	T	0.91	1.16	3 × 10-7	CYP17A1, AS3MT, CNNM2, NT5C2	[42,43]
11	rs381815	16 858 844	E	29 136	SBP	T	0.26	0.65	2 × 10-9	PLEKHA7	[42,43]
12	rs17249754	88584	EA	8842	SBP, DBP	A	0.37	1.06	9 × 10-7	ATP2B1	[49]
12	rs2681472	88 533 090	E	29 136	SBP, DBP, HTN	A	0.83	0.50	2 × 10-9	ATP2B1	[42,43]
12	rs2681492,	88 537 220	E	29 136	SBP, DBP, HTN	T	0.80	0.85	4 × 10-11	ATP2B1	[42,43]
12	rs3184504	110 368 991	E	29 136	SBP, DBP	T	0.49	0.48	3 × 10-14	ATXN2, SH2B3	[42,43]
12	rs653178	110 492 139	E	34 433	DBP	T	0.53	-0.46	3 × 10-18	ATXN2, SH2B3	[42,43]
12	rs2384550	113 837 114	E	29 136	DBP	A	0.35	0.43	4 × 10-8	TBX3, TBX5	[42,43]
15	rs1550576	56 000 706	AA	1017	SBP	C	0.86	1.92	3 × 10-6	ALDH1A2	[44]
15	rs1378942	72 865 396	E	34 433	DBP	C	0.36	0.43	1 × 10-23	CSK, CYP1A1, CYP1A2, LMAN1L, CPLX3, ARID3B, ULK3	[42,43]
15	rs6495122	72 912 698	E	29 136	DBP	A	0.42	0.40	2 × 10-10	CSK, CYP1A1, CYP1A2, LMAN1L, CPLX3, ARID3B, ULK3	[42,43]
16	rs13333226	20 273 155	E	3320	HTN	A	0.81	1.15	4 × 10-11	UMOD	[50]
16	rs11646213	81 200 152	E	1977	HTN	T	0.60	1.28	8 × 10-6	CDH13	[48]
17	rs12946454	40 563 647	E	34 433	SBP	T	0.28	0.57	1 × 10-8	PLCD3, ACBD4, HEXIM1, HEXIM2	[42,43]
17	rs16948048	44 795 465	E	34 433	DBP	G	0.39	0.31	5 × 10-9	ZNF652, PHB	[42,43]

E: European; AA: African American; EA: East Asians; SBP: Systolic blood pressure; DBP: Diastolic blood pressure; HTN: Hypertension; ACBD4: Acyl-CoA binding domain containing 4; ADH7: Alcohol dehydrogenase 7 (class IV), mu or sigma polypeptide; AGTRAP: Angiotensin II receptor-associated protein; ALDH1A2: Aldehyde dehydrogenase 1 family, member A2; ARID5B: AT rich interactive domain 5B (MRF1-like); AS3MT: Arsenic (+3 oxidation state) methyltransferase; ATP2B1: ATPase; Ca⁺⁺ transporting; plasma membrane 1; ATXN2: Ataxin 2; C10orf107: Chromosome 10 open reading frame 107; C4orf22: Chromosome 4 open reading frame 22; CACNB2: Calcium channel, voltage-dependent, β 2 subunit; CDH13: Cadherin 13, H-cadherin (heart); CLCN6: Chloride channel 6; CNNM2: Cyclin M2; CPLX3: Complexin 3; CSK: C-src tyrosine kinase; CYP17A1: Cytochrome P450, family 17, subfamily A, polypeptide 1; CYP1A1: Cytochrome P450; family 1, subfamily A, polypeptide 1; CYP1A2: Cytochrome P450, family 1, subfamily A, polypeptide 2; FGF5: Fibroblast growth factor 5; HEXIM1: Hexamethylene bis-acetamide inducible 1; HEXIM2: Hexamthylene bis-acetamide inducible 2; LMAN1L: Lectin, mannose-binding, 1 like; MTHFR: Methylenetetrahydrofolate reductase (NAD(P)H); NPPA: Natriuretic peptide A; NPPB: Natriuretic peptide B; NT5C2: 5'-nucleotidase, cytosolic II; PHB: Prohibitin; PLCD3: Phospholipase C, Δ 3; PLEKHA7: Pleckstrin homology domain containing, family A member 7; PRDM8: PR domain containing 8; RHOBTB1: Rho-related BTB domain containing 1; RTKN2: Rhotekin 2; SH2B3: SH2B adaptor protein 3; STK39: Serine threonine kinase 39; TBX3: T-box 3; TBX5: T-box 5; TMEM26: Transmembrane protein 26; ULK3: Unc-51-like kinase 3 (C. elegans); ULK4: Unc-51-like kinase 4 (C. elegans); UMOD: Uromodulin, ZNF652: Zinc finger protein 652; SNP: Single nucleotide polymorphisms; OR: Odds ratio.

Figure 6 Significant genome-wide association study findings in blood pressure. ACBD4: Acyl-CoA binding domain containing 4; ADH7: Alcohol dehydrogenase 7 (class IV), mu or sigma polypeptide; AGTRAP: Angiotensin II receptor-associated protein; ALDH1A2: Aldehyde dehydrogenase 1 family, member A2; ARID5B: AT rich interactive domain 5B (MRF1-like); AS3MT: Arsenic (+3 oxidation state) methyltransferase; ATP2B1: ATPase, Ca⁺⁺ transporting; plasma membrane 1; ATXN2: Ataxin 2; C10orf107: Chromosome 10 open reading frame 107; C4orf22: Chromosome 4 open reading frame 22; CACNB2: Calcium channel, voltage-dependent, β 2 subunit; CDH13: Cadherin 13, H-cadherin (heart); CLCN6: Chloride channel 6; CNNM2: Cyclin M2; CPLX3: Complexin 3; CSK: C-src tyrosine kinase; CYP17A1: Cytochrome P450, family 17, subfamily A; polypeptide 1; CYP1A1: Cytochrome P450, family 1, subfamily A, polypeptide 1; CYP1A2: Cytochrome P450, family 1, subfamily A, polypeptide 2; FGF5: Fibroblast growth factor 5; HEXIM1: Hexamethylene bis-acetamide inducible 1; HEXIM2: Hexamthylene bis-acetamide inducible 2; LMAN1L: Lectin, mannose-binding, 1 like; MTHFR: Methylenetetrahydrofolate reductase (NAD(P)H); NPPA: Natriuretic peptide A; NPPB: Natriuretic peptide B; NT5C2: 5'-nucleotidase, cytosolic II; PHB: Prohibitin; PLCD3: Phospholipase C, Δ 3; PLEKHA7: Pleckstrin homology domain containing, family A member 7; PRDM8: PR domain containing 8; RHOBTB1: Rho-related BTB domain containing 1; RTKN2: Rhotekin 2; SH2B3: SH2B adaptor protein 3; STK39: Serine threonine kinase 39; TBX3: T-box 3; TBX5: T-box 5; TMEM26: Transmembrane protein 26; ULK3: Unc-51-like kinase 3 (C. elegans); ULK4: Unc-51-like kinase 4 (C. elegans); UMOD: Uromodulin; ZNF652: Zinc finger protein 652.

CLINICAL IMPLICATIONS

GWAS are a useful tool in the identification of new and unexpected genetic loci of common diseases and traits, thus providing key novel insights into disease biology. But the clinical utility of these discoveries is negligible at this stage. The comparatively small numbers of variants which have been successfully replicated in several independent studies explain only a small proportion of the observed variation of these traits and explain in aggregate less than 20% of disease heritability. For example, the loci underpinning LDL-C levels[28] and BP account for < 20% of the variance of these quantitative traits. The variants associated with CHD increase disease risk by up to 20% per allele[51,52]. Next generation sequencing is now used to study low-frequency and rare variants that may potentially explain some of the missing heritabilities; however it is likely that studies designed to test for gene-environment interactions and gene-gene interactions may hold the answer. There were attempts to develop genetic profiles using the results from GWAS studies, but these have very limited value in personalised risk prediction as the genotype-phenotype effect sizes are very small. In the few studies that have evaluated the ability of a panel of genetic markers to discriminate CHD cases, the area under the receiver operating characteristic curve has been small indicating that conventional risk factors and family history are better at predicting risk and the incremental advantage of adding genetic markers is negligible. A few studies have attempted reclassification based on incorporation of SNPs from GWAS of CAD, lipids, etc.[52-58], and while they showed some improvement in net reclassification, the interpretation of these are still controversial and not translatable into general use[59]. Many companies are providing direct-to-consumer genetic tests that provide a “genetic risk profile” for an individual using risk alleles of small-to-moderate effects despite the clinical utility of genetic screening not being established. None of the major healthcare providers in Europe and USA have adopted these tests for CHD risk prediction, and the FDA has advised that direct-to-consumer genetic tests should be considered to be medical devices requiring FDA approval for commercial use. The future application of genetic screening will be in identifying risk groups early in life to direct targeted preventive measures and potentially pharmacogenetic tests to identify those at higher risk for adverse events. While technology is not a barrier to achieving this, the discovery, evaluation and deployment of these tests will require the same standards as non-genetic tests[60].