The first GWAS on HTN, published on the Framingham Heart Study population in 2007 (Table 15.1), found that there were no SNP associations with systolic BP (SBP) or diastolic BP (DBP) [12] that achieved genome-wide significance (GWS; P ≤ 4.4 × 10⁻⁸) . However, at a more modest level of stringency (P < 10⁻⁵), there were seven associations for SBP or DBP. However, none of these were in previously identified candidate-gene loci. A larger-scale GWAS for HTN was performed by the Wellcome Trust Case Control Consortium (WTCCC) using a case-control design, and this study too failed to identify SNPs with GWS (P < 10⁻⁷) [9]. Even the most strongly associated SNPs did not identify genes from physiological systems previously implicated by clinical or genetic studies in HTN. The authors contend some of these common susceptibility variants of large effect size, e.g., promoter of the WNK lysine-deficient protein kinase 1 (WNK1) gene, [13, 14] were not well tagged by the Affymetrix chip they used. In addition, HTN may have fewer common risk alleles of larger effect sizes than some of the other complex phenotypes , in which case identification of susceptibility variants for HTN will need synthesis of findings from multiple large-scale studies. Further, BP is an imperfect trait and may have resulted in misclassification due to inclusion of hypertensive subjects within the control samples. In a replication study, Ehret et al. attempted to replicate six top-associated SNPs from WTCCC in the Family BP Program cohort [15] with very discrepant results. In a GWAS on subjects with type 2 diabetes mellitus (T2DM) and euglycemic controls, Saxena et al. evaluated 18 traits including BP and found no genome-wide associations for BP [16].

The observation that BP variation, in the general population, is due to multiple variants with small effects led to the formation of large consortia, e.g., Cohorts for Heart and Aging Research in Genome Epidemiology (CHARGE) and Global BPgen consortium to identify common generic variation associated with complex traits (Table 15.1). The CHARGE consortium consisted of a large number of participants of European descent [7] and identified 13 SNPs for SBP, 20 for DBP, and 10 for HTN (P < 4 × 10⁻⁷), some of which were common among different BP traits. Mean BP and prevalence of HTN increased in relation to the number of risk alleles carried. When ten CHARGE SNPs for each trait were included in a joint meta-analysis with the Global BPgen consortium (replication sample, another GWAS consortium of similar size), four CHARGE loci attained GWS (P < 5 × 10⁻⁸) for SBP (ATP2B1, CYP17A1, PLEKHA7, SH2B3; Table 15.1), six for DBP (ATP2B1, CACNB2, CSK-ULK3, SH2B3, TBX3-TBX5, ULK4), and one for HTN (ATP2B1; Table 15.1), with considerable concordance among top loci across all three phenotypes . Of note, rs1004467, a common intronic variant in CYP17A1, a gene associated with a rare Mendelian form of HTN, emerged as a genome-wide significant locus in the meta-analysis of results from both consortia. In a similar large study, Newton-Cheh et al. also tested ~ 100,000 subjects (including replication study) and identified associations between SBP or DBP and common variants in eight regions near the CYP17A1, CYP1A2, FGF5, SH2B3, MTHFR, c10orf107, ZNF652, and PLCD3 (P ≤ 1 × 10⁻⁸) genes [8]. All variants associated with BP were also associated with dichotomous HTN. These landmark studies, which identified 13 novel BP loci, paved the way for a number of studies in different populations; however, each study reported new findings and often failed to confirm previous GWAS .

Using the concept that limited genetic and environmental diversity and reduced allelic heterogeneity observed in isolated founder populations could facilitate discovery of loci contributing to both Mendelian and complex diseases, Wang et al. carried out a GWAS of SBP and DBP in Amish subjects from Amish Family Diabetes Study [17]. Strong association signals with several common variants in a serine/threonine kinase gene (STK39) were found, and they confirmed these associations in an independent Amish and four non-Amish Caucasian samples including the Diabetes Genetics Initiative, Framingham Heart Study, GenNet, and Hutterites (P < 10⁻⁶). Two SNPs (rs6749447 and rs3754777) accounted for an estimated allelic effect size of 2 mmHg SBP and 1 mmHg DBP. In a similar effort, Sabatti et al. conducted GWAS for SBP and DBP in Northern Finland Birth Cohort 1966 (NFBC1966) members, drawn from the most genetically isolated Finnish regions; however, no individual locus achieved GWS [18].

Org et al. used Kooperative Gesundheitsforschung in der Region Augsburg (KORA) S3 cohort (n = 1644) recruited from the general population in southern Germany, [19] and identified an association between BP traits and common variants upstream of the CDH13 gene. The initial associations with HTN and DBP were confirmed in two other European population-based cohorts: KORA S4 (Germans) and HYPEST (Estonians). Carriers of the minor allele A had a decreased risk of HTN. A nonsignificant trend for association was also detected with severe family-based HTN in the BRIGHT sample (British). Using an extreme case-control design, Padmanabhan et al. identified a locus on Uromodulin gene (P = 3.6 × 10⁻¹¹) [10]. The minor G allele was associated with a lower risk of HTN (OR, 95 %CI: 0.87, 0.84–0.91), reduced urinary uromodulin excretion, better renal function; and each copy of the G allele is associated with a 7.7 % reduction in risk of CVD events after adjusting for age, sex, BMI, and smoking status (H.R. = 0.923, 95 %CI: 0.860–0.991; P = 0.027). In another subset of 13,446 subjects, they showed that rs13333226 was independently associated with HTN (OR, CI: 0.890.83–0.96, P = 0.004) [10]. Similarly, using another novel two-marker method, Slavin et al. reexamined WTCCC dataset, and detected SNP pairs in five genes associated with HTN: GPR39, XRCC4, MYO6, ZFAT, and MACROD2 along with four other SNP pair regions that were at least 70 kb from any known gene [20].

In 2011, a large meta-analysis of GWAS of European population, using a multistage design in 200,000 subjects [21], identified 29 independent SNPs at 28 loci which were significantly associated with SBP, DBP, or both (P < 5 × 10⁻⁹). Sixteen of these 29 associations were novel loci: Six of these loci contained genes previously known or suspected to regulate BP (GUCY1A3–GUCY1B3, NPR3–C5orf23, ADM, FURIN–FES, GOSR2, and GNAS–EDN3); the other ten provided new clues to BP physiology [21]. They further evaluated whether the BP variants identified in Europeans were associated with BP in subjects of East Asian (N = 29,719), South Asian (N = 23,977), and African (N = 19,775) ancestries. They found significant associations in subjects of East Asian ancestry for SNPs at nine loci (rs1173771, rs633185, rs2521501, rs1327235, rs381815, rs1458038, rs11191548, rs1378942, and rs17249754) and in subjects of South-Asian ancestry for SNPs at six loci (rs2932538, rs1327235, rs6015450, rs1458038, rs11191548, and rs17249754) . The authors attributed the lack of association of BP with some SNPs to small sample size in non-European cohorts, and created a genetic risk scores for SBP and DBP involving all 29 BP variants weighted according to the effect sizes observed in the European samples. In each non-European ancestry group, risk scores were strongly associated with SBP (P = 1.1 × 10⁻⁴⁰ in East Asian, P = 2.9 × 10⁻¹³ in South Asian, P = 9.8 × 10⁻⁴ in African ancestry subjects) and DBP (P = 2.9 × 10⁻⁴⁸, P = 9.5 × 10⁻¹⁵, and P = 5.3 × 10⁻⁵, respectively). In an independent sample of 23,294 women [22], the authors found one SD increase in the genetic risk score was associated with a 21 % increase in the odds of HTN (95 %CI19 %–28 %). Among subjects in the top decile of the risk score, the prevalence of HTN was 29 % compared with 16 % in the bottom decile (OR: 2.09, 95 %CI: 1.86–2.36). In another independent HTN case-control sample, subjects in the top compared to bottom quintiles of genetic risk score differed by 4.6 mmHg SBP and 3.0 mmHg DBP, differences that approach population-averaged BP treatment effects for a single antihypertensive agent [23]. A risk score derived from 29 variants was also significantly associated with CVD, but not kidney disease .

In a relatively smaller GWAS in four Finnish cohorts consisting of metabolic syndrome (MS) cases and controls, both free of T2DM, Kristiansson et al. identified SMEK2 gene locus SNPs to be associated with SBP (P = 4.02 × 10⁻⁸ and P = 4.25 × 10⁻⁸) [24]. In a similar attempt, Kraja et al. studied subjects of European descent from SNP Typing for Association with Multiple Phenotypes from Existing Epidemiologic Data (STAMPEED) consortium, eight unique SNPs were identified for bivariate traits (BP being one of the traits) based on NCEP definition of MS (Table 15.1). None of these SNPs demonstrated a significant association (P < 0.05) with BP alone [25], although some of the SNPs were associated with MS. Using a slightly different model of considering a bivariate response, Das et al. detected eight SNPs for males and seven for females from Framingham Heart Study which are most significant in controlling BP [26].

Few investigators investigated genetic association with related BP phenotypes including mean arterial pressure (MAP), and pulse pressure (PP). Ganesh et al. investigated genetic associations with SBP, DBP, MAP, and PP among subjects of European ancestry by genotyping 50,000 SNPs in 2100 candidate genes for cardiovascular phenotypes and identified two novel associations for SBP and DBP, and confirmed ten previously known loci (Table 15.1) .

In an attempt to overcome earlier problems with misclassification of cases as controls, HYPERGENES project investigators excluded controls that developed HTN at a later age. These subjects were followed for 5–10 years after DNA collection. In a two-stage study of cases and controls from different European regions [27], SNP rs3918226 was associated with HTN in whites (P = 2.58 × 10⁻¹³ and OR of 1.54; 95 %CI: 1.37–1.73) under an additive model. This SNP mapped to a new HTN susceptibility locus in the promoter region of the endothelial NO synthase gene. This finding was further confirmed in a meta-analysis of genotyping data for 21,714 subjects (Anglo-Scandinavian Cardiac Outcomes Trial/AIBIII/NBS, BRIGHT, EPIC-Turin, HYPEST, and NORDIL/MDC samples), resulting in an overall OR of 1.34 (95 %CI: 1.25–1.44; P = 1.032 × 10⁻¹⁴). The quantitative effect of rs3918226 was also estimated in continuous BP phenotypes, resulting in a β-coefficient of 1.91 for SBP and 1.40 for DBP, despite the low P-values of the regression probably because of the low sample size. This is the first GWAS that points to eNOS regulation, though the authors point to the use of Ilumina 1M array as rs3918226 is not present in other commercial arrays and the high rate of recombination in this region resulting in low linkage disequilibrium . Seven additional SNPs within the eNOS gene showed significant P-values, but did not reach GWS. Previously, candidate-gene studies had inconsistently pointed to the association of eNOS with HTN with positive associations in Asian cohorts [28–30], whereas the majority of studies among whites were negative [31].

African Americans (AA) are disproportionately affected by HTN and associated complications . Adeyemo et al. undertook the first GWAS for BP and HTN among AA from the Washington DC area who were all participants of Howard University Family Study (HUFS) and replicated some of the significant SNPs in a sample of West Africans [32]. They identified multiple SNPs reaching GWS (P ≤ 0.05) for SBP in or near the genes: PMS1, SLC24A4, YWHA7, IPO7, and CACANA1H. No variant reached GWS for association with DBP or with HTN as a binary trait. In addition, they attempted to replicate significant SNPs in STK39 and CDH13 genes identified in Amish and German populations and found that variants in both these genes were also associated with SBP among AA. These findings were not confirmed in a separate AA sample from Milwaukee, WI. A subsequent large GWAS for BP was performed among AA recruited in the Candidate Gene Association Resource (CARe) consortium which failed to identify any major loci associated with HTN [33, 34]. However, in a related meta-analysis, two novel loci were identified that reached statistical significance: rs2258119 on chromosome 21 with SBP and rs10474346 on chromosome 5 with DBP . However, neither of these associations was replicated in independent AA samples, again highlighting the difficulty in extending the findings of GWAS to independent populations [35].

In the first high-density association study of HTN among Japanese subjects [36], investigators observed association with rs3755351 (P = 1.7 × 10⁻⁵) in ADD2. Cho et al. described an intergenic SNP near the ATP2B1 gene (P = 1.3 × 10⁻⁷) with an effect size of − 1.309 ± 0.266 mmHg in GWAS among Korean subjects [37].

To increase the genetic contribution and homogeneity of the study trait, Yang et al. focused on young-onset HTN and performed GWAS. They identified an SNP quartet s9308945-rs6711736-rs6729869-rs10495809 located on chromosome 2p22.3. The quartet was 219, 322, 457, and 495 kb downstream of LOC344371 (hypothetical gene), MYADML (pseudo gene), FAM98 A (hypothetical protein), and RASGRP3, respectively [38]. These genes are novel HTN targets identified in this first GWAS of the Han Chinese population.

The Asian Genetic Epidemiology Network BP (AGEN-BP) working group was established to facilitate the identification of genetic variants influencing BP among populations of East Asian ancestry (including Japanese, Han Chinese, Korean, and Malay) [39]. In combined analyses of a three-stage study, six loci reached GWS (P < 5 × 0⁻⁸; Table 15.1). One SNP rs3544 (located near TBX3-TBX5) showed some evidence of allelic heterogeneity in relation to BP. In addition, rs880315 (CASZ1), previously identified in European population was associated with DBP (P = 3.1 × 10⁻¹⁰). Of the 13 other variants in GWAS meta-analyses in subjects of European descent, 7 of the 13 loci (54 %) showed nominally significant associations of the reported lead SNPs in East Asians. These data suggest that although some interpopulation differences may exist in the pathways involved in the BP elevation (or HTN) between Europeans and East Asians, the majority of pathways are common. In addition, SNP rs3544 and BP were associated with a nonsynonymous SNP (rs671) in ALDH2, which determines an individual’s tolerance to alcohol intake and has pleiotropic effects on other metabolic traits and CAD, highlighting the importance of fine-mapping efforts to pinpoint causal variants and causal genes, thereby providing new insights into the physiology of complex diseases. This association was not noted in populations of European descent and appears to be specific to East Asians .

Among other GWAS findings in Asian populations include AKAP13 gene association in a Korean population [40]; IGF1, SLC4A4, WWOX; SFMBT1 gene associations a Han Chinese population [41]; and FSTL4 in another population of Chinese ancestry [42]. Among the GWAS in which no SNP reached GWS (P < 5 × 10⁻⁷⁾ were studies on genetically isolated founder population of the Pacific island of Kosrae [43], Indian-Asian men [44], and Japanese subjects [45].

Among other studies in Asian populations, one study, prospectively investigated the incidence of HTN in subjects with short sleep duration over a 6-year follow-up period in a GWAS (Table 15.1) [46]. They identified three genetic variants associated with an increased risk of incident HTN only in premenopausal women (adjusted hazard ratio 2.43, 95 % CI = 1.36–4.35). Some investigators have also studied other BP related phenotypes in Asian populations. Wain et al. reported GWAS of PP and MAP in discovery and follow-up studies and identified four new PP loci near CHIC2, near PIK3CG—in NOV—and near ADAMTS8, and two new MAP loci (in MAP4 and near ADRB1) and one locus associated with both of these traits (near FIGN) that has also recently been associated with SBP in East Asians (P < 2.7 × 10⁻⁸; Table 15.2) [47]. Zhang et al. carried out GWAS for pulse pressure on 63 middle-aged dizygotic twin pairs using high-density markers [48] and detected a suggestive association (rs17031508, P < 8.3 × 10⁻⁸; Table 15.2) .