Genome-wide analysis in over 1 million individuals reveals over 2,000 independent genetic signals for blood pressure

doi:10.21203/rs.3.rs-1409164/v1

Download PDF

Article

Genome-wide analysis in over 1 million individuals reveals over 2,000 independent genetic signals for blood pressure

https://doi.org/10.21203/rs.3.rs-1409164/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 30 Apr, 2024

Read the published version in Nature Genetics →

Version 1

posted

You are reading this latest preprint version

Hypertension is a leading cause of premature death affecting more than a billion individuals worldwide. Here we report on the genetic determinants of blood pressure (BP) traits (systolic, diastolic, and pulse pressure) in the largest single-stage genome-wide analysis to date (N = 1,028,980 European-descent individuals). We identified 2,103 independent genetic signals (P < 5x10^− 8) for BP traits, including 113 novel loci. These associations explain ~ 40% of common SNP heritability of systolic and diastolic BP. Comparison of top versus bottom deciles of polygenic risk scores (PRS) based on these results reveal clinically meaningful differences in BP (12.9 mm Hg for systolic BP, 95% CI 11.5–14.2 mm Hg, p = 9.08×10^− 73) and hypertension risk (OR 5.41; 95% CI 4.12 to 7.10; P = 9.71×10^− 33) in an independent dataset. Compared with the area under the curve (AUC) for hypertension discrimination for a model with sex, age, BMI, and genetic ancestry, adding systolic and diastolic BP PRS increased discrimination from 0.791 (95% CI = 0.781–0.801) to 0.814 (95% CI = 0.805–0.824, ∆AUC = 0.023, P = 2.27x10^− 22). Our transcriptome-wide association study detected 2,793 BP colocalized associations with genetically-predicted expression of 1,070 genes in five cardiovascular tissues, of which 500 are previously unreported for BP traits. These findings represent an advance in our understanding of hypertension and highlight the role of increasingly large genomic studies for development of more accurate PRS, which may inform precision health research.

Over thirty percent of adults worldwide have hypertension, which represents a leading modifiable risk factor for cardiovascular disease and death^1–3. Blood pressure (BP) is highly heritable and multiple genome-wide association studies (GWAS) have highlighted its complex, polygenic architecture^4–9.

Two recent large-scale GWAS meta-analyses with over 750,000 participants of European descent^4,5 incorporating available data from biobanks and consortia such as the UK Biobank (UKB), the International Consortium for Blood Pressure (ICBP) and the Million Veteran Program (MVP), identified more than one thousand independent loci associated with systolic BP (SBP), diastolic BP (DBP), and pulse pressure (PP). Both studies revealed new biology and highlighted new pathways for BP regulation or confirmed pathways that have been implicated previously in the genetics of BP. Experience from these and prior BP GWAS reveals that an increase in sample size results in an enriched catalog of BP-associated genetic loci as well as an increase in the proportion of interindividual variation in BP explained by the lead variants.

In this study we conducted a single-stage GWAS meta-analysis combining all available genetic data from UKB, ICBP, MVP, and Vanderbilt University's biorepository of DNA linked to de-identified medical records (BioVU)¹⁰. We accumulated over 1 million individuals of European descent, the largest sample size to date in a genome-wide single-stage association study for BP traits. The analysis was performed using ~ 7.5 million imputed single nucleotide polymorphisms (SNPs) with a minor allele frequency (MAF) of one percent or more.

Our goal was to map the genetic architecture of BP by identifying novel variants and to evaluate our GWAS results in the context of previously published BP loci. Herein we report the discovery of 113 novel loci for BP traits including 35 loci with p-values less than 5x10^− 9 and 78 additional loci with p-values less than 5x10^− 8. Leveraging the large sample size of our meta-analyses and current statistical techniques, we developed genome-wide BP polygenic risk scores (PRS) and tested prediction of clinically relevant differences in BP traits and hypertension risk in independent samples. The large sample size also led to an observation of increased SNP-based heritability (h_SNP²) of BP traits explained by GWAS variants.

We also applied methods that leverage the statistical precision of the GWAS and independent reference data to infer relationships between BP traits and gene expression, and observed evidence of association with BP biology of 500 previously unreported genes. Many of these genes are located in previously mapped regions of the genome, but were not identified by nearest gene annotations in the literature, allowing the scientific yield from BP genetic studies to advance from lists of loci to lists of genes. These analyses provide insights into both the extent to which regulatory effects mediate genetic associations with BP traits, as well as a principled data-driven mapping of associated loci with linked biology. This knowledge can be used to identify potential drug targets, develop testable hypotheses in model systems, and advance understanding of BP regulation at the level of tissues and systems.

Our one-stage meta-analysis study of 7,584,058 SNPs in up to 1,028,980 individuals identified 1,495, 1,504, and 1,318 significant loci for SBP, DBP, and PP, respectively (LD r² <0.1and 1Mb distance; Supplementary Figure 11). After excluding all known loci and their linked variants (LD r² >0.1 at +/- 5Mb) and applying clumping and LD-pruning methods to remaining SNPs to identify independent loci at least 1 Mb apart and not in strong LD (r²<0.1), we detected sentinel SNPs indexing 35 novel loci with a consistent direction of effect in all available studies and that reached a stringent one-stage significance threshold of p<5x10^-9 in at least one of the three continuous BP traits (Figure 1, Table 1). Sentinel SNPs indexing an additional 78 novel loci reached the traditional genome-wide significance threshold (p<5x10^-8) and are reported in Table 2. Of all 113 novel loci (Supplementary Figures 12-13), 44, 45, and 33 sentinel SNPs reached genome-wide significance (p<5x10^-8) for SBP, DBP, and PP, respectively. As in prior studies, the newly discovered loci had smaller effect sizes than previously reported SNPs, owing to the larger sample size and increased power to detect common variants with smaller effect size (Supplementary Figure 14).

LDSR-intercepts

In our overall meta-analyses, genomic inflation factors were calculated and λ_GC values were 1.82, 1.76, and 1.70 for SBP, DBP, and PP, respectively. We calculated the LDSR-intercepts in our overall meta-analyses as well as in the portion of our data novel for BP genetics to evaluate whether inflation of our test statistics was due to polygenicity or residual population substructure (Supplementary Table 7). Attenuation ratios in overall analyses were 0.0884, 0.0844, and 0.0794, while attenuation ratios in the novel partition of our results were 0.0996, 0.0722, and 0.1085 for SBP, DBP, and PP, respectively. These ratios indicate that LD scores from the reference population are well-matched to our study population. LDSR-intercepts in overall analyses were 1.2254, 1.2037, and 1.1756, while intercepts in the novel partition of our results were 1.0931, 1.0624, and 1.0806 for SBP, DBP, and PP, respectively. These intercepts suggest that any observed inflation in our data is due primarily to polygenicity.

Known Loci

Using our data to assign all 3,800 SNPs previously reported for BP traits into loci, we identified 1,165 independent loci at least 1 Mb apart and not in strong LD (r²<0.1) with other loci (Supplementary Table 8). A total of 953 of the 1,165 loci reached genome-wide significance (p<5x10^-8) in our meta-analyses. LD-pruning resulted in 1,723 pairwise-independent genetic signals from known SNPs. Trait-specific effect estimates of these SNPs were used to weight the GRS used in our percent variance explained analyses (Supplementary Table 5).

Results of our meta-analyses provide strong support for previously reported BP SNPs (Supplementary Tables 3 and 4). Of the 3,800 previously reported SNPs, including the exact SNP, an LD proxy if the exact SNP was unavailable, or a distance proxy if neither were available, 82.6% (N=3,137) reached genome-wide significance (p<5x10^-8) in our results while 95.5% (N=3,630) showed evidence of association (p<1x10^-4). Of 298 previously reported SNPs unavailable in our data 227 (76%) were identified in rare-variant, non-European ancestry, and/or gene-environment interaction analyses. MAF and effect sizes of previously reported SNPs in our meta-analyses are concordant with published results (Supplementary Figures 15-16).

Conditional Analysis

Genome-wide conditional analysis of SBP, DBP, and PP meta-analyses identified a total of 267 additional independent significant secondary SNPs reaching a significance threshold of p<5x10^-8 in the conditional joint model (Supplementary Table 9). Of the 267 SNPs, 203 secondary SNPs also reached p<5x10^-8 in our primary meta-analyses, and 23 mapped to one of our 113 novel BP loci.

Summary of GWAS Results

In summary, we detected 1,723 pairwise independent genetic signals among SNPs previously reported for BP traits, 113 genome-wide significant novel loci for SBP, DBP, and PP in our meta-analyses, and 267 additional independent significant secondary SNPs in conditional analysis for a total of 2103 independent genetic signals across all three BP traits.

Variance Explained: GRS and PRS results

Table 3 shows the variance explained by all four GRS and the full PRS in the independent sample of 10,210 Lifelines participants. The GRS of our 113 novel loci explained a small, but significant proportion of BP variance: 0.06%, 0.08% and 0.02% for SBP, DBP, and PP, respectively. Our novel findings contributed a small gain in percentage of variance explained for SBP, DBP, and PP. For example, for SBP, percent variance explained by GRS increased from 6.77% for the 1,723 previously published SNPs to 6.80% after adding the 113 novel sentinel SNPs, and to 6.93% for all 2103 independent BP genetic signals after also adding 267 independent secondary SNPs. Furthermore, the full optimal PRS (P-threshold=1x10^-3, 0.01, 0.01 for SBP, DBP, and PP, respectively) constructed for each BP trait captured a total of 7.17%, 7.83% and 4.53% of the variance in SBP, DBP, and PP, respectively. Supplementary Figure 17 shows variance explained by the three BP PRS at different P-value thresholds.

Decile analyses of full BP PRS for SBP, DBP, PP, and Hypertension in Lifelines

The optimal PRS showed sex-adjusted differences between top and bottom deciles of the PRS distribution of 12.9 mm Hg for SBP (95% CI 11.5 to 14.2 mm Hg, p=9.08×10^-73), 8.8 mm Hg for DBP (95% CI 8.0 to 9.7 mm Hg, p=3.88×10^-93), and 8.3 mm Hg for PP (95% CI 7.4 to 9.3 mm Hg, p=2.16×10^-64) in 10,210 Lifelines participants (Figure 2a, Supplementary Figure 18, Supplementary Table 10). In addition, we observed an over five-fold higher sex-adjusted odds of hypertension (OR 5.41; 95% CI 4.12 to 7.10; p=9.71×10^-33) between the top and bottom deciles of the PRS in Lifelines when modeling both the SBP and DBP PRSs (Figure 2b, Supplementary Table 10).

Hypertension model performance and calibration assessment in Lifelines

AUROC for model 1 including only covariates was 0.79 (95% CI = 0.78-0.80) and 0.81 (95% CI = 0.81-0.82) for model 2 including covariates as well as optimal SBP and DBP PRSs, a difference of 0.02 (p=2.27x10^-22, Supplementary Figure 19). Brier scores for model 1 (0.14) and model 2 (0.14) indicate that our models were reasonably well-calibrated. The Youden indices for model 1 and model 2 were 1.43 and 1.47, respectively, and correspond to the 58^th percentile of the total sample. Hypertension prevalence in Lifelines was 23.6%. Summary statistics for these analyses are provided in Supplementary Table 11.

Decile analyses of full BP PRS for hypertension in UKB African-ancestry individuals

We observed two-fold higher sex-adjusted odds of hypertension (OR 2.40; 95% CI 1.61 to 3.57; p=1.65×10^-5) between the top and bottom deciles of the PRS in 3,341 UKB African-ancestry participants (Supplementary Figure 20, Supplementary Table 12).

Comparison of REML methods to calculate heritability

We calculated h_SNP²of SBP within the UKB BP-GWAS dataset using GCTA-GREML⁴⁵ to compare to previously published results and to check the consistency across different software. We performed an analysis of the SNP-based heritability of SBP in a small subset of the UKB data (N=19,410). This analysis yielded an h_SNP² estimate of 22.8% which is consistent with the estimate of 21.3% reported by Evangelou et al. using BOLT-REML, demonstrating that the GCTA-GREML approach is also therefore appropriate to use for calculation of heritability within our other smaller Lifelines cohort.

Heritability results in Lifelines

The GCTA-GREML⁴⁵ h_SNP² estimates in Lifelines data (N=10,210) were 17.4%, 18.8% and 16.1% for SBP, DBP, and PP, respectively. These GCTA-GREML⁴⁵ h_SNP² estimates were used in the denominator of %VE/h_SNP² calculations as both %VE and h_SNP² were derived from exactly the same dataset. Hence the total proportions of common SNP heritability that our GWAS explained, either for all 2,103 independent BP genetic signals combined or for the full PRS capturing all genome-wide common SNP variation, were 6.93/17.4% = 39.8% and 7.17/17.4% = 41.2%, respectively for SBP, 6.92/18.8% = 36.8% and 7.83/18.8% = 41.6%, respectively for DBP, and 4.47/16.1% = 27.8% and 4.53/16.1% = 28.1%, respectively for PP.

Variant functions of novel loci

More than 90% of the novel loci lie within non-coding regions (Supplementary Table 13). One novel sentinel SNP (rs855791), in addition to seven highly linked (r² >0.8) SNPs are non-synonymous variants in genes at six novel loci: TMPRSS6, GLRX2, RLF, HELQ, ZNF235, and UNC13B; three of these non-synonymous SNPs reside in UNC13B (Supplementary Table 14).

Overlap of novel loci across three BP traits and with other traits

There were concordant associations across BP traits for 113 novel loci (Supplementary Figures 21-22), among which nine are genome-wide significant for a second BP trait. The most similarity in effect directions was between SBP and DBP (r² = 0.68), and the least was between PP and DBP (r² = 0.13), which is consistent with previous observations^4,5,7.

Shared associations with at least one other disease trait reported within the GWAS Catalog or PhenoScanner database were observed for 41 out of the 113 novel loci, i.e., sentinel SNPs and all SNPs in high LD (r² >0.8). The proportion of novel SNPs (36%) exhibiting shared associations was higher than that reported by Evangelou et al (77 out of 535, 14%)⁵.

The novel locus with the most shared associations was MCHR2-AS1, which has significant associations with seven disease/trait categories: anthropometric, reproductive, lipids, thyroid, hemodynamics, neurological, and metabolic. Other loci showed associations with hematological traits (e.g., hemoglobin, red blood cell count, white blood cell count, etc.), immune system (e.g., inflammation, allergy, autoimmune, etc.), respiratory traits (e.g., vital capacity, expiratory volume, expiratory flow, etc.), and minerals (e.g., iron metabolism) (Supplementary Figure 23).

Transcriptome-wide association study and colocalization analysis

With S-PrediXcan analysis, we identified 5,538 statistically significant gene-tissue combinations genetically predictive of BP traits (Supplementary Table 15, Supplementary Figure 24). These combinations correspond to 1,873 unique genes, of which 545 (29%) have been identified by nearest gene mapping of previously reported BP SNPs or novel sentinel SNPs identified in our meta-analyses. A total of 468 (25%) unique genes were previously identified in the BP TWAS by Giri et al⁴. We identified 1,029 (55%) unique genes in this analysis that have not previously been reported in BP GWAS (Supplementary Table 15). The majority of associations were observed in arterial tissues (N=1,503 for tibial artery; N=1,205 for aorta). Associations were evenly distributed across all three BP traits (N=1,851 for SBP; N=1,962 for DBP; N=1,725 for PP).

Additionally, we used COLOC to identify the subset of significant genes for which there was a high posterior probability that a SNP in the S-PrediXcan model for each gene exhibited colocalized association with both gene expression and changes in quantitative measures of BP traits. This analysis refined our S-PrediXcan analysis by characterizing the contribution of underlying expression quantitative trait loci (eQTLs) within our gene models to the observed S-PrediXcan associations. We detected 2,793 gene-tissue pairs where there was a statistically significant S-PrediXcan association with at least one BP trait and high posterior probability (PP.H4>0.9) of colocalization, corresponding to a total of 1,070 distinct genes. Of these 1,070 genes, 500 (47%) have not been previously annotated for SNP associations with BP traits.

Druggable targets from TWAS/COLOC results intersecting with novel and secondary signals

We collated evidence for genes that 1) mapped to our novel sentinel SNPs or 2) mapped to our secondary SNPs but did not map from our primary GWAS or previous GWAS. We then found the intersection with genes that were significant in our TWAS, and highlighted noteworthy examples (Table 4). We identified 38 genes satisfying this criterion, including an established drug target for BP medications (ADRA1A) and five genes targeted by other approved drugs (Supplementary Table 16).

In the largest single-stage GWAS of BP traits to date including over one million European-ancestry adults, we identified over 2,000 independent BP signals from known and novel loci as well as new secondary signals. The richness of results permitted the creation of PRS that captured substantial interindividual variation in BP traits. These full PRS have been made publicly accessible and can now be used by the global research community to explore the contributions of BP to a variety of health outcomes such as cardiovascular disease (myocardial infarction, stroke, heart failure), kidney disease, and cognitive impairment and dementia.

The results of this analysis provide additional insights into the genetic architecture of BP and suggest that expansions of statistical power will continue to yield discovery of additional loci primarily harboring common variants with smaller effect sizes. In addition to mapping genomic locations, this analysis also demonstrates that many loci are associated with BP traits through regulatory effects on gene expression. We identified significant colocalized associations between BP traits and genetically-predicted gene expression of 1,070 genes, 500 of which have not been identified in prior GWAS.

These new gene observations can provide opportunities for further experimentation in model systems, elucidate candidate targets for drug development or repurposing, and potentially provide leverage to understand ancestry-related disparities in BP traits through evaluation in non-European ancestry populations. These findings further demonstrate that the biology of BP is highly complex and polygenic, and may be influenced by thousands of SNPS with extremely subtle effect sizes. In aggregate these associations explain large differences in average BP and have a very strong influence on risk of hypertension. Understanding the heritable influences on BP has the potential to provide foreknowledge of severe hypertension and its sequelae^49,50. This study is therefore another key step towards understanding one of the most complex and highly regulated biological systems in humans that has significant implications for health and for disease treatment and prevention.

Application of the PRS in an external independent study (Lifelines) demonstrated large BP differences between the top and bottom deciles of the PRS distribution of 12.9, 8.8, and 8.3 mm Hg for SBP, DBP, and PP, respectively. Furthermore, the top deciles of the optimal PRS for SBP and DBP exhibited 5.4-fold increased odds of hypertension compared with the bottom deciles. AUROC analyses indicated significant improvement in discrimination and calibration when the optimal PRS for SBP and DBP were included in the predictive model for hypertension. The observed negative predictive value of 91.1% for the full model Youden-index cutoff demonstrates accurate discrimination of false negatives, an important goal in classification of hypertension susceptibility. In top versus bottom decile analysis of hypertension odds in UKB African-ancestry participants, we observed an attenuated odds ratio (OR = 2.40; 95% CI = 1.61 to 3.57; p = 1.65×10^− 5) compared to European-ancestry Lifelines participants. This attenuation is a common limitation of the portability of PRS to other genetic ancestries and is likely due to genetic drift⁵⁹. The improved performance of our PRS over previously reported hypertension loci may allow for the identification of causal contributions of BP for many hypertension-related diseases. These results justify the expansion of BP GWAS in non-European ancestry populations.

The current study explained 41.2% of the common SNP heritability of SBP, using the full PRS. This constitutes a substantial increase of 53.7% compared to 26.8% of the SBP h_SNP² explained and reported by Evangelou et al⁵. Even though we demonstrate that a large proportion of the genetic variance in BP is discoverable by GWAS, another gap still remains between the common variant-based heritability and the total pedigree-based h² estimates that were recently reported to range from 25–30% for SBP, DBP, and PP⁶⁰. This gap is likely attributable to rare variants as has been reported recently for height and BMI on the basis of whole genome sequencing data⁶¹.

Among novel loci, TMPRSS6 (rs855791, PP p = 3.20x10^− 8) is a promising candidate as a potential drug target. This gene, encoding transmembrane serine protease 6, has been implicated in attenuation of dietary iron overload in heart tissue leading to cardioprotective effects^51,52. Genetic variation at TMPRSS6 is also associated with biomarkers of iron overload⁵³. SMAD7 (rs72917789, SBP p = 1.14x10^− 8) has been shown to modulate expression of hepcidin, a key regulator of intestinal iron absorption^54,55. Additionally, GSMT1 (rs36209093, DBP p = 9.94x10^− 15), encoding glutathione S-transferase Mu 1, has been implicated in cardiomyopathy due to iron overload^56,57. These results suggest that altered iron metabolism may play an important role in BP regulation and hypertension-related cardiovascular disease and is consistent with previous studies linking high iron stores to cardiovascular disease⁵⁸.

Evaluation of the intersection of TWAS/COLOC results with novel and secondary loci highlights several genes targeted by approved medications or with compelling biological evidence supporting their role in BP physiology. ADRA1A, encoding the α-1-adrenergic receptor 1A, the product of which is a well-known target for medications treating both hypertension and hypotension,⁶² was previously unreported in GWAS of BP traits. Considering our conditional analysis and TWAS associations at this locus, cis-regulatory variants for ADRA1 may affect efficacy of targeted medications. ABCC8, an established diabetes GWAS locus⁶³ targeted by sulfonylurea medications^64,65, harbors rare variants contributing to pulmonary arterial hypertension^66–68. FGFR2, targeted by anti-angiogenesis medications in the treatment of cancer⁶⁹, is involved in sexual dimorphism of the baroreflex afferent function on BP regulation in rats⁷⁰ and has been implicated in parenchymal and vascular remodeling in pulmonary arterial hypertension⁷¹. These findings are biologically plausible, and the ADRA1A receptor protein is targeted to manipulate BP, demonstrating that our approach detects genes with biological and pharmacological impact. This suggests that additional genes from our analysis may be viable options for drug targeting and further study.

This study has several limitations. Due to the large sample size for this study, independent study samples to replicate our findings in a more traditional 2-stage design are not readily available. We have attempted to address this limitation by implementing more stringent and robust reporting criteria, for example stricter post-QC filtering of the meta-analysis data, applying a higher 5x10^− 9 significance threshold for results, and requiring full concordance in the direction of novel SNP effects across all four datasets in the meta-analysis. Our study examines the genetic determinants of BP traits in non-Hispanic white participants. Investigations including non-European ancestry populations, which may yield greater insight into the genetic architecture of BP traits, are currently underway.

Our study results suggest that effort should still continue for future studies of BP traits to leverage large-scale biobank resources and cohort studies in order to expand the sample size further. The benefits of this approach may include improved homogeneity of associations if the data are collected under uniform conditions, as in UKB. Future studies should also continue to evaluate associations with genetically-predicted gene expression to stimulate other avenues of investigation. Substantial studies including individuals of non-European ancestry should be conducted in order to provide comparisons and insight into ancestrally-related disparities in BP traits. These goals, if accomplished, will provide researchers with translational knowledge to mitigate disparities and reduce the global impact of health outcomes for which hypertension is a highly common risk factor.

Mills, K.T., Bundy, J.D., Kelly, T.N., Reed, J.E., Kearney, P.M., Reynolds, K., Chen, J., and He, J. (2016). Global Disparities of Hypertension Prevalence and Control: A Systematic Analysis of Population-Based Studies From 90 Countries. Circulation 134, 441–450.
GBD 2017 Causes of Death Collaborators (2018). Global, regional, and national age-sex-specific mortality for 282 causes of death in 195 countries and territories, 1980–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet 392, 1736–1788.
GBD 2017 Risk Factor Collaborators (2018). Global, regional, and national comparative risk assessment of 84 behavioural, environmental and occupational, and metabolic risks or clusters of risks for 195 countries and territories, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet 392, 1923–1994.
Giri, A., Hellwege, J.N., Keaton, J.M., Park, J., Qiu, C., Warren, H.R., Torstenson, E.S., Kovesdy, C.P., Sun, Y.V., Wilson, O.D., et al. (2019). Trans-ethnic association study of blood pressure determinants in over 750,000 individuals. Nat. Genet. 51, 51–62.
Evangelou, E., Warren, H.R., Mosen-Ansorena, D., Mifsud, B., Pazoki, R., Gao, H., Ntritsos, G., Dimou, N., Cabrera, C.P., Karaman, I., et al. (2018). Genetic analysis of over 1 million people identifies 535 new loci associated with blood pressure traits. Nat. Genet. 50, 1412–1425.
Wain, L.V., Vaez, A., Jansen, R., Joehanes, R., van der Most, P.J., Erzurumluoglu, A.M., O’Reilly, P.F., Cabrera, C.P., Warren, H.R., Rose, L.M., et al. (2017). Novel Blood Pressure Locus and Gene Discovery Using Genome-Wide Association Study and Expression Data Sets From Blood and the Kidney. Hypertension.
Warren, H.R., Evangelou, E., Cabrera, C.P., Gao, H., Ren, M., Mifsud, B., Ntalla, I., Surendran, P., Liu, C., Cook, J.P., et al. (2017). Genome-wide association analysis identifies novel blood pressure loci and offers biological insights into cardiovascular risk. Nat Genet 49, 403–415.
Hoffmann, T.J., Ehret, G.B., Nandakumar, P., Ranatunga, D., Schaefer, C., Kwok, P.-Y., Iribarren, C., Chakravarti, A., and Risch, N. (2017). Genome-wide association analyses using electronic health records identify new loci influencing blood pressure variation. Nat Genet 49, 54–64.
Ehret, G.B., Ferreira, T., Chasman, D.I., Jackson, A.U., Schmidt, E.M., Johnson, T., Thorleifsson, G., Luan, J., Donnelly, L.A., Kanoni, S., et al. (2016). The genetics of blood pressure regulation and its target organs from association studies in 342,415 individuals. Nat Genet 48, 1171–1184.
Roden, D.M., Pulley, J.M., Basford, M.A., Bernard, G.R., Clayton, E.W., Balser, J.R., and Masys, D.R. (2008). Development of a large-scale de-identified DNA biobank to enable personalized medicine. Clin. Pharmacol. Ther. 84, 362–369.
Bycroft, C., Freeman, C., Petkova, D., Band, G., Elliott, L.T., Sharp, K., Motyer, A., Vukcevic, D., Delaneau, O., O’Connell, J., et al. (2018). The UK Biobank resource with deep phenotyping and genomic data. Nature 562, 203–209.
McCarthy, S., Das, S., Kretzschmar, W., Delaneau, O., Wood, A.R., Teumer, A., Kang, H.M., Fuchsberger, C., Danecek, P., Sharp, K., et al. (2016). A reference panel of 64,976 haplotypes for genotype imputation. Nat Genet 48, 1279–1283.
International Consortium for Blood Pressure Genome-Wide Association Studies, Ehret, G.B., Munroe, P.B., Rice, K.M., Bochud, M., Johnson, A.D., Chasman, D.I., Smith, A.V., Tobin, M.D., Verwoert, G.C., et al. (2011). Genetic variants in novel pathways influence blood pressure and cardiovascular disease risk. Nature 478, 103–109.
Winkler, T.W., Day, F.R., Croteau-Chonka, D.C., Wood, A.R., Locke, A.E., Mägi, R., Ferreira, T., Fall, T., Graff, M., Justice, A.E., et al. (2014). Quality control and conduct of genome-wide association meta-analyses. Nat Protoc 9, 1192–1212.
Willer, C.J., Li, Y., and Abecasis, G.R. (2010). METAL: fast and efficient meta-analysis of genomewide association scans. Bioinformatics 26, 2190–2191.
Loh, P.-R., Danecek, P., Palamara, P.F., Fuchsberger, C., A Reshef, Y., K Finucane, H., Schoenherr, S., Forer, L., McCarthy, S., Abecasis, G.R., et al. (2016). Reference-based phasing using the Haplotype Reference Consortium panel. Nat. Genet. 48, 1443–1448.
Das, S., Forer, L., Schönherr, S., Sidore, C., Locke, A.E., Kwong, A., Vrieze, S.I., Chew, E.Y., Levy, S., McGue, M., et al. (2016). Next-generation genotype imputation service and methods. Nat. Genet. 48, 1284–1287.
Marchini, J., Howie, B., Myers, S., McVean, G., and Donnelly, P. (2007). A new multipoint method for genome-wide association studies by imputation of genotypes. Nature Genetics 39, 906–913.
Ani, A., van der Most, P.J., Snieder, H., Vaez, A., and Nolte, I.M. (2021). GWASinspector: comprehensive quality control of genome-wide association study results. Bioinformatics 37, 129–130.
1000 Genomes Project Consortium, Auton, A., Brooks, L.D., Durbin, R.M., Garrison, E.P., Kang, H.M., Korbel, J.O., Marchini, J.L., McCarthy, S., McVean, G.A., et al. (2015). A global reference for human genetic variation. Nature 526, 68–74.
Bulik-Sullivan, B.K., Loh, P.-R., Finucane, H., Ripke, S., Yang, J., Patterson, N., Daly, M.J., Price, A.L., and Neale, B.M. (2015). LD Score Regression Distinguishes Confounding from Polygenicity in Genome-Wide Association Studies. Nat Genet 47, 291–295.
Altshuler, D.M., Gibbs, R.A., Peltonen, L., Dermitzakis, E., Schaffner, S.F., Yu, F., Bonnen, P.E., de Bakker, P.I.W., Deloukas, P., Gabriel, S.B., et al. (2010). Integrating common and rare genetic variation in diverse human populations. Nature 467, 52–58.
Surendran, P., Drenos, F., Young, R., Warren, H., Cook, J.P., Manning, A.K., Grarup, N., Sim, X., Barnes, D.R., Witkowska, K., et al. (2016). Trans-ancestry meta-analyses identify rare and common variants associated with blood pressure and hypertension. Nat Genet 48, 1151–1161.
Surendran, P., Feofanova, E.V., Lahrouchi, N., Ntalla, I., Karthikeyan, S., Cook, J., Chen, L., Mifsud, B., Yao, C., Kraja, A.T., et al. (2020). Discovery of rare variants associated with blood pressure regulation through meta-analysis of 1.3 million individuals. Nat Genet 52, 1314–1332.
Feitosa, M.F., Kraja, A.T., Chasman, D.I., Sung, Y.J., Winkler, T.W., Ntalla, I., Guo, X., Franceschini, N., Cheng, C.-Y., Sim, X., et al. (2018). Novel genetic associations for blood pressure identified via gene-alcohol interaction in up to 570K individuals across multiple ancestries. PLoS One 13, e0198166.
Takeuchi, F., Akiyama, M., Matoba, N., Katsuya, T., Nakatochi, M., Tabara, Y., Narita, A., Saw, W.-Y., Moon, S., Spracklen, C.N., et al. (2018). Interethnic analyses of blood pressure loci in populations of East Asian and European descent. Nat Commun 9, 5052.
de Las Fuentes, L., Sung, Y.J., Noordam, R., Winkler, T., Feitosa, M.F., Schwander, K., Bentley, A.R., Brown, M.R., Guo, X., Manning, A., et al. (2021). Gene-educational attainment interactions in a multi-ancestry genome-wide meta-analysis identify novel blood pressure loci. Mol Psychiatry 26, 2111–2125.
Sung, Y.J., de Las Fuentes, L., Winkler, T.W., Chasman, D.I., Bentley, A.R., Kraja, A.T., Ntalla, I., Warren, H.R., Guo, X., Schwander, K., et al. (2019). A multi-ancestry genome-wide study incorporating gene-smoking interactions identifies multiple new loci for pulse pressure and mean arterial pressure. Hum Mol Genet 28, 2615–2633.
Sung, Y.J., Winkler, T.W., de Las Fuentes, L., Bentley, A.R., Brown, M.R., Kraja, A.T., Schwander, K., Ntalla, I., Guo, X., Franceschini, N., et al. (2018). A Large-Scale Multi-ancestry Genome-wide Study Accounting for Smoking Behavior Identifies Multiple Significant Loci for Blood Pressure. Am J Hum Genet 102, 375–400.
Kichaev, G., Bhatia, G., Loh, P.-R., Gazal, S., Burch, K., Freund, M.K., Schoech, A., Pasaniuc, B., and Price, A.L. (2019). Leveraging Polygenic Functional Enrichment to Improve GWAS Power. Am J Hum Genet 104, 65–75.
Buniello, A., MacArthur, J.A.L., Cerezo, M., Harris, L.W., Hayhurst, J., Malangone, C., McMahon, A., Morales, J., Mountjoy, E., Sollis, E., et al. (2019). The NHGRI-EBI GWAS Catalog of published genome-wide association studies, targeted arrays and summary statistics 2019. Nucleic Acids Res 47, D1005–D1012.
Staley, J.R., Blackshaw, J., Kamat, M.A., Ellis, S., Surendran, P., Sun, B.B., Paul, D.S., Freitag, D., Burgess, S., Danesh, J., et al. (2016). PhenoScanner: a database of human genotype-phenotype associations. Bioinformatics 32, 3207–3209.
Purcell, S., Neale, B., Todd-Brown, K., Thomas, L., Ferreira, M.A.R., Bender, D., Maller, J., Sklar, P., de Bakker, P.I.W., Daly, M.J., et al. (2007). PLINK: a tool set for whole-genome association and population-based linkage analyses. Am. J. Hum. Genet. 81, 559–575.
Vaez, A., Jansen, R., Prins, B.P., Hottenga, J.-J., de Geus, E.J.C., Boomsma, D.I., Penninx, B.W.J.H., Nolte, I.M., Snieder, H., and Alizadeh, B.Z. (2015). In Silico Post Genome-Wide Association Studies Analysis of C-Reactive Protein Loci Suggests an Important Role for Interferons. Circ Cardiovasc Genet 8, 487–497.
Wang, K., Li, M., and Hakonarson, H. (2010). ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res 38, e164.
Yang, J., Ferreira, T., Morris, A.P., Medland, S.E., Genetic Investigation of ANthropometric Traits (GIANT) Consortium, DIAbetes Genetics Replication And Meta-analysis (DIAGRAM) Consortium, Madden, P.A.F., Heath, A.C., Martin, N.G., Montgomery, G.W., et al. (2012). Conditional and joint multiple-SNP analysis of GWAS summary statistics identifies additional variants influencing complex traits. Nat. Genet. 44, 369–375, S1-3.
Scholtens, S., Smidt, N., Swertz, M.A., Bakker, S.J.L., Dotinga, A., Vonk, J.M., van Dijk, F., van Zon, S.K.R., Wijmenga, C., Wolffenbuttel, B.H.R., et al. (2015). Cohort Profile: LifeLines, a three-generation cohort study and biobank. Int J Epidemiol 44, 1172–1180.
International Schizophrenia Consortium, Purcell, S.M., Wray, N.R., Stone, J.L., Visscher, P.M., O’Donovan, M.C., Sullivan, P.F., and Sklar, P. (2009). Common polygenic variation contributes to risk of schizophrenia and bipolar disorder. Nature 460, 748–752.
Bilimoria, K.Y., Liu, Y., Paruch, J.L., Zhou, L., Kmiecik, T.E., Ko, C.Y., and Cohen, M.E. (2013). Development and evaluation of the universal ACS NSQIP surgical risk calculator: a decision aid and informed consent tool for patients and surgeons. J Am Coll Surg 217, 833–842.e1-3.
Pencina, M.J., and D’Agostino, R.B. (2015). Evaluating Discrimination of Risk Prediction Models: The C Statistic. JAMA 314, 1063–1064.
Steyerberg, E.W., Vickers, A.J., Cook, N.R., Gerds, T., Gonen, M., Obuchowski, N., Pencina, M.J., and Kattan, M.W. (2010). Assessing the performance of prediction models: a framework for traditional and novel measures. Epidemiology 21, 128–138.
Arkes, H.R., Dawson, N.V., Speroff, T., Harrell, F.E., Alzola, C., Phillips, R., Desbiens, N., Oye, R.K., Knaus, W., and Connors, A.F. (1995). The covariance decomposition of the probability score and its use in evaluating prognostic estimates. SUPPORT Investigators. Med Decis Making 15, 120–131.
Robin, X., Turck, N., Hainard, A., Tiberti, N., Lisacek, F., Sanchez, J.-C., and Müller, M. (2011). pROC: an open-source package for R and S + to analyze and compare ROC curves. BMC Bioinformatics 12, 77.
Loh, P.-R., Kichaev, G., Gazal, S., Schoech, A.P., and Price, A.L. (2018). Mixed-model association for biobank-scale datasets. Nat Genet 50, 906–908.
Yang, J., Lee, S.H., Goddard, M.E., and Visscher, P.M. (2011). GCTA: a tool for genome-wide complex trait analysis. Am. J. Hum. Genet. 88, 76–82.
Barbeira, A.N., Dickinson, S.P., Bonazzola, R., Zheng, J., Wheeler, H.E., Torres, J.M., Torstenson, E.S., Shah, K.P., Garcia, T., Edwards, T.L., et al. (2018). Exploring the phenotypic consequences of tissue specific gene expression variation inferred from GWAS summary statistics. Nat Commun 9, 1825.
GTEx Consortium (2015). Human genomics. The Genotype-Tissue Expression (GTEx) pilot analysis: multitissue gene regulation in humans. Science 348, 648–660.
Giambartolomei, C., Vukcevic, D., Schadt, E.E., Franke, L., Hingorani, A.D., Wallace, C., and Plagnol, V. (2014). Bayesian test for colocalisation between pairs of genetic association studies using summary statistics. PLoS Genet. 10, e1004383.
Sakaue, S., Kanai, M., Karjalainen, J., Akiyama, M., Kurki, M., Matoba, N., Takahashi, A., Hirata, M., Kubo, M., Matsuda, K., et al. (2020). Trans-biobank analysis with 676,000 individuals elucidates the association of polygenic risk scores of complex traits with human lifespan. Nat Med 26, 542–548.
Vaura, F., Kauko, A., Suvila, K., Havulinna, A.S., Mars, N., Salomaa, V., FinnGen, null, Cheng, S., and Niiranen, T. (2021). Polygenic Risk Scores Predict Hypertension Onset and Cardiovascular Risk. Hypertension 77, 1119–1127.
Du, X., She, E., Gelbart, T., Truksa, J., Lee, P., Xia, Y., Khovananth, K., Mudd, S., Mann, N., Moresco, E.M.Y., et al. (2008). The serine protease TMPRSS6 is required to sense iron deficiency. Science 320, 1088–1092.
Truksa, J., Gelbart, T., Peng, H., Beutler, E., Beutler, B., and Lee, P. (2009). Suppression of the hepcidin-encoding gene Hamp permits iron overload in mice lacking both hemojuvelin and matriptase-2/TMPRSS6. Br J Haematol 147, 571–581.
Benyamin, B., Esko, T., Ried, J.S., Radhakrishnan, A., Vermeulen, S.H., Traglia, M., Gögele, M., Anderson, D., Broer, L., Podmore, C., et al. (2014). Novel loci affecting iron homeostasis and their effects in individuals at risk for hemochromatosis. Nat Commun 5, 4926.
Charlebois, E., and Pantopoulos, K. (2021). Iron overload inhibits BMP/SMAD and IL-6/STAT3 signaling to hepcidin in cultured hepatocytes. PLoS One 16, e0253475.
Kautz, L., Meynard, D., Monnier, A., Darnaud, V., Bouvet, R., Wang, R.-H., Deng, C., Vaulont, S., Mosser, J., Coppin, H., et al. (2008). Iron regulates phosphorylation of Smad1/5/8 and gene expression of Bmp6, Smad7, Id1, and Atoh8 in the mouse liver. Blood 112, 1503–1509.
Singh, M.M., Kumar, R., Tewari, S., and Agarwal, S. (2019). Association of GSTT1/GSTM1 and ApoE variants with left ventricular diastolic dysfunction in thalassaemia major patients. Hematology 24, 20–25.
Wu, K.-H., Chang, J.-G., Ho, Y.-J., Wu, S.-F., and Peng, C.-T. (2006). Glutathione S-transferase M1 gene polymorphisms are associated with cardiac iron deposition in patients with beta-thalassemia major. Hemoglobin 30, 251–256.
Salonen, J.T., Nyyssönen, K., Korpela, H., Tuomilehto, J., Seppänen, R., and Salonen, R. (1992). High stored iron levels are associated with excess risk of myocardial infarction in eastern Finnish men. Circulation 86, 803–811.
Martin, A.R., Gignoux, C.R., Walters, R.K., Wojcik, G.L., Neale, B.M., Gravel, S., Daly, M.J., Bustamante, C.D., and Kenny, E.E. (2017). Human Demographic History Impacts Genetic Risk Prediction across Diverse Populations. Am J Hum Genet 100, 635–649.
Tegegne, B.S., Man, T., van Roon, A.M., Asefa, N.G., Riese, H., Nolte, I., and Snieder, H. (2020). Heritability and the Genetic Correlation of Heart Rate Variability and Blood Pressure in > 29 000 Families: The Lifelines Cohort Study. Hypertension 76, 1256–1262.
Wainschtein, P., Jain, D.P., Yengo, L., Zheng, Z., TOPMed Anthropometry Working Group, T.-O. for P.M.C., Cupples, L.A., Shadyab, A.H., McKnight, B., Shoemaker, B.M., Mitchell, B.D., et al. (2019). Recovery of trait heritability from whole genome sequence data. BioRxiv 588020.
Martínez-Salas, S.G., Campos-Peralta, J.M., Pares-Hipolito, J., Gallardo-Ortíz, I.A., Ibarra, M., and Villalobos-Molina, R. (2007). Alpha1A-adrenoceptors predominate in the control of blood pressure in mouse mesenteric vascular bed. Auton Autacoid Pharmacol 27, 137–142.
Mahajan, A., Taliun, D., Thurner, M., Robertson, N.R., Torres, J.M., Rayner, N.W., Payne, A.J., Steinthorsdottir, V., Scott, R.A., Grarup, N., et al. (2018). Fine-mapping type 2 diabetes loci to single-variant resolution using high-density imputation and islet-specific epigenome maps. Nat Genet 50, 1505–1513.
Hambrock, A., Löffler-Walz, C., and Quast, U. (2002). Glibenclamide binding to sulphonylurea receptor subtypes: dependence on adenine nucleotides. Br J Pharmacol 136, 995–1004.
Qin, X., Zhong, J., and Lan, D. (2020). The use of glimepiride for the treatment of neonatal diabetes mellitus caused by a novel mutation of the ABCC8 gene. J Pediatr Endocrinol Metab 33, 1605–1608.
Lago-Docampo, M., Tenorio, J., Hernández-González, I., Pérez-Olivares, C., Escribano-Subías, P., Pousada, G., Baloira, A., Arenas, M., Lapunzina, P., and Valverde, D. (2020). Characterization of rare ABCC8 variants identified in Spanish pulmonary arterial hypertension patients. Sci Rep 10, 15135.
Le Ribeuz, H., Capuano, V., Girerd, B., Humbert, M., Montani, D., and Antigny, F. (2020). Implication of Potassium Channels in the Pathophysiology of Pulmonary Arterial Hypertension. Biomolecules 10,.
Southgate, L., Machado, R.D., Gräf, S., and Morrell, N.W. (2020). Molecular genetic framework underlying pulmonary arterial hypertension. Nat Rev Cardiol 17, 85–95.
Eichholz, A., Merchant, S., and Gaya, A.M. (2010). Anti-angiogenesis therapies: their potential in cancer management. Onco Targets Ther 3, 69–82.
Chen, P., Xu, B., Feng, Y., Li, K.-X., Liu, Z., Sun, X., Lu, X.-L., Wang, L.-Q., Chen, Y.-W., Fan, X.-X., et al. (2020). FGF-21 ameliorates essential hypertension of SHR via baroreflex afferent function. Brain Res Bull 154, 9–20.
El Agha, E., Schwind, F., Ruppert, C., Günther, A., Bellusci, S., Schermuly, R.T., and Kosanovic, D. (2018). Is the fibroblast growth factor signaling pathway a victim of receptor tyrosine kinase inhibition in pulmonary parenchymal and vascular remodeling? Am J Physiol Lung Cell Mol Physiol 315, L248–L252.

Acknowledgements:

Author Support

J.N.H. is supported by K12HD04348 (PI K.E. Hartmann). T.E. and A.M. were supported by EU CoE grant No. 2014-2020.4.01.15-0012 and Estonian Research Council grant PRG1291.

MVP

This research is based in part on data from the Million Veteran Program, Office of Research and Development, Veterans Health Administration. This publication does not represent the views of the Department of Veterans Affairs or the United States Government.

UKB

This research has been conducted using the UK Biobank Resource under Application Numbers 236 and 10035. This research was supported by the British Heart Foundation (grant SP/13/2/30111). Large-scale comprehensive genotyping of UK Biobank for cardiometabolic traits and diseases: UK CardioMetabolic Consortium (UKCMC).

BioVU

BioVU is supported by institutional funding, the 1S10RR025141-01 instrumentation award, and by the Vanderbilt CTSA grant UL1TR000445 from NCATS/NIH. This work was conducted in part using the resources of the Advanced Computing Center for Research and Education at Vanderbilt University, Nashville, TN, supported in part by an S10 instrumentation award (1S10OD023680-01).

Lifelines Cohort Study

The Lifelines Biobank initiative has been made possible by funding from the Dutch Ministry of Health, Welfare and Sport, the Dutch Ministry of Economic Affairs, the University Medical Center Groningen (UMCG the Netherlands), University of Groningen and the Northern Provinces of the Netherlands. The generation and management of GWAS genotype data for the Lifelines Cohort Study is supported by the UMCG Genetics Lifelines Initiative (UGLI).

The authors wish to acknowledge the services of the Lifelines Cohort Study, the contributing research centers delivering data to Lifelines, and all the study participants.

Lifelines Cohort Study group authors:

Raul Aguirre-Gamboa (1), Patrick Deelen (1), Lude Franke (1), Jan A Kuivenhoven (2), Esteban A Lopera Maya (1), Ilja M Nolte (3), Serena Sanna (1), Harold Snieder (3), Morris A Swertz (1), Judith M Vonk (3), Cisca Wijmenga (1)

Department of Genetics, University of Groningen, University Medical Center Groningen, The Netherlands
Department of Pediatrics, University of Groningen, University Medical Center Groningen, The Netherlands
Department of Epidemiology, University of Groningen, University Medical Center Groningen, The Netherlands

KORA Study

The KORA study was initiated and financed by the Helmholtz Zentrum München –German Research Center for Environmental Health, which is funded by the German Federal Ministry of Education and Research (BMBF) and by the State of Bavaria. Furthermore, KORA research was supported within the Munich Center of Health Sciences (MC-Health), Ludwig-Maximilians-Universität, as part of LMUinnovativ and funded by the Bavarian State Ministry of Health and Care through the research project DigiMed Bayern (www.digimed-bayern.de). Group authors include Annette Peters (PI) and Christian Gieger (CoPI) and Melanie Waldenberger. The informed consents given by KORA study participants do not cover data posting in public databases. However, data are available upon request from KORA Project Application Self-Service Tool (https://epi.helmholtz-muenchen.de/) Data requests can be submitted online and are subject to approval by the KORA Board.

Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) Consortium

The CHARGE cohorts were supported in part by HL105756.

Cardiovascular Health Study (CHS)

This CHS research was supported by NHLBI contracts HHSN268201200036C, HHSN268200800007C, HHSN268201800001C, N01HC55222, N01HC85079, N01HC85080, N01HC85081, N01HC85082, N01HC85083, N01HC85086, 75N92021D00006; and NHLBI grants U01HL080295, R01HL087652, R01HL105756, R01HL103612, R01HL120393, and U01HL130114 with additional contribution from the National Institute of Neurological Disorders and Stroke (NINDS). Additional support was provided through R01AG023629 from the National Institute on Aging (NIA). A full list of principal CHS investigators and institutions can be found at CHS-NHLBI.org. The provision of genotyping data was supported in part by the National Center for Advancing Translational Sciences, CTSI grant UL1TR001881, and the National Institute of Diabetes and Digestive and Kidney Disease Diabetes Research Center (DRC) grant DK063491 to the Southern California Diabetes Endocrinology Research Center.

MESA

MESA and the MESA SHARe projects are conducted and supported by the National Heart, Lung, and Blood Institute (NHLBI) in collaboration with MESA investigators. Support for MESA is provided by contracts 75N92020D00001, HHSN268201500003I, N01-HC-95159, 75N92020D00005, N01-HC-95160, 75N92020D00002, N01-HC-95161, 75N92020D00003, N01-HC-95162, 75N92020D00006, N01-HC-95163, 75N92020D00004, N01-HC-95164, 75N92020D00007, N01-HC-95165, N01-HC-95166, N01-HC-95167, N01-HC-95168, N01-HC-95169, UL1-TR-000040, UL1-TR-001079, UL1-TR-001420. Funding for SHARe genotyping was provided by NHLBI Contract N02-HL-64278. Genotyping was performed at Affymetrix (Santa Clara, California, USA) and the Broad Institute of Harvard and MIT (Boston, Massachusetts, USA) using the Affymetrix Genome-Wide Human SNP Array 6.0. Also supported in part by NHLBI CHARGE Consortium Contract HL105756. The provision of genotyping data was supported in part by the National Center for Advancing Translational Sciences, CTSI grant UL1TR001881, and the National Institute of Diabetes and Digestive and Kidney Disease Diabetes Research Center (DRC) grant DK063491 to the Southern California Diabetes Endocrinology Research Center. Infrastructure for the CHARGE Consortium is supported in part by the National Heart, Lung, and Blood Institute (NHLBI) grant R01HL105756.

National Institutes of Health (NIH)

This research was supported in part by the Intramural Research Program of the NIH, National Human Genome Research Institute (NHGRI; project number Z01 HG200417), National Institute on Aging (NIA; project number Z01 AG000535), National Institutes of Health, Department of Health and Human Services, as well as the National Institute of Neurological Disorders and Stroke.

CROATIA Study

The CROATIA cohorts were funded by grants from the MRC (United Kingdom), European Commission Framework 6 project EUROSPAN (contract number LSHG-CT-2006-018947), Croatian Science Foundation (grant 8875), the Republic of Croatia Ministry of Science, Education and Sports (216-1080315-0302) and the Croatian National Center of Research Excellence in Personalized Healthcare (grant number KK.01.1.1.01.0010) and the Center of Competence in Molecular Diagnostics (KK.01.2.2.03.0006).

Young Finns Study

The Young Finns Study has been financially supported by the Academy of Finland: grants 322098, 286284, 134309 (Eye), 126925, 121584, 124282, 255381, 256474, 283115, 319060, 320297, 314389, 338395, 330809, 104821, 129378 (Salve), 117797 (Gendi), and 141071 (Skidi); the Social Insurance Institution of Finland; Competitive State Research Financing of the Expert Responsibility area of Kuopio, Tampere and Turku University Hospitals (grant X51001); Juho Vainio Foundation; Paavo Nurmi Foundation; Finnish Foundation for Cardiovascular Research; Finnish Cultural Foundation; The Sigrid Juselius Foundation; Tampere Tuberculosis Foundation; Emil Aaltonen Foundation; Yrjö Jahnsson Foundation; Signe and Ane Gyllenberg Foundation; Diabetes Research Foundation of Finnish Diabetes Association; EU Horizon 2020 (grant 755320 for TAXINOMISIS and grant 848146 for To Aition); European Research Council (grant 742927 for MULTIEPIGEN project); Tampere University Hospital Supporting Foundation, Finnish Society of Clinical Chemistry and the Cancer Foundation Finland.

HTO

Collection of the HTO cohort was funded by the Wellcome Trust. Genotyping was funded by the British Heart Foundation (BHF). BK holds a BHF Personal Chair (CH/13/2/30154).

Women’s Genome Health Study (WGHS)

The WGHS is supported by the National Heart, Lung, and Blood Institute (HL043851 and HL080467) and the National Cancer Institute (CA047988 and UM1CA182913) with funding for genotyping provided by Amgen.

Justification for the Use of Statins in Prevention: an Intervention Trial Evaluating Rosuvastatin (JUPITER)

JUPITER and the genotyping in the JUPITER study population were funded by AstraZeneca.

Conflicts of Interest:

M.A.N.’s participation in this project was part of a competitive contract awarded to Data Tecnica International LLC by the National Institutes of Health to support open science research, he also currently serves on the scientific advisory board for Clover Therapeutics and is an advisor to Neuron23 Inc as a data science fellow. B.M.P. serves on the Steering Committee of the Yale Open Data Access Project funded by Johnson & Johnson. P.V. received an unrestricted grant from GlaxoSmithKline to build the CoLaus study (2003). V.S. has received honoraria for consulting from Novo Nordisk and Sanofi and has ongoing research collaboration with Bayer Ltd. (all unrelated to this project). R.L. is a part-time consultant of Metabolon, Inc. M.J.C. is Chief Scientist for Genomics England, a UK Government company. J.M.M.H. is a full-time employee of Novo Nordisk Ltd. C.J.O. is currently employed by Novartis Institutes for Biomedical Research (unrelated to this project) and remains credentialed as a without compensation researcher with the Veterans Administration. T.S. is co-founder of Zoe Ltd. A.S.B. reports institutional grants from AstraZeneca, Bayer, Biogen, BioMarin, Bioverativ, Novartis, Regeneron and Sanofi. J.N.D. reports grants, personal fees and non-financial support from Merck Sharp & Dohme (MSD), grants, personal fees and non-financial support from Novartis, grants from Pfizer and grants from AstraZeneca outside the submitted work. J.N.D. sits on the International Cardiovascular and Metabolic Advisory Board for Novartis (since 2010); the Steering Committee of UK Biobank (since 2011); the MRC International Advisory Group (ING) member, London (since 2013); the MRC High Throughput Science 'Omics Panel Member, London (since 2013); the Scientific Advisory Committee for Sanofi (since 2013); the International Cardiovascular and Metabolism Research and Development Portfolio Committee for Novartis; and the Astra Zeneca Genomics Advisory Board (2018). E.E is co-founder has received consultation fees from Open DNA Ltd (unrelated to this project).

Author Contributions:

Data analysis

JMK, ZK, AV, TX, AA, EE, JNH, HRW

Manuscript drafting

JMK, ZK, AV, TX, AA, EE, JNH, JCD, DL, TLE, PBM, HS, HRW

Manuscript editing

JMK, ZK, AV, TX, AA, EE, JNH, LY, WJY, MT, AG, PMV, DIC, APM, MJC, SH, JSK, DC, JRA, ACM, RJL, KK, RS, AAH, PPP, CPN, NJS, LR, UG, OM, HR, JFW, HC, BMP, YL, JIR, XG, KMR, PV, JS, CL, MDT, VG, JL, JT, ZK, SR, VS, GG, ST, JWJ, PvdH, PMR, FG, VV, AG, HW, SEH, IJD, PJvdM, AJO, BDK, CH, AC, MB, LJS, TB, CM, MJ, AP, CG, EGL, FC, JH, PK, SE, MHD, OP, MPC, EC, MC, RL, EH, HS, BS, MW, DPS, ML, AT, MD, VG, JPC, DR, IK, EB, MT, TL, OTR, ADJ, CN, MJB, AFD, PJS, NP, JCC, RE, DS, TE, AM, RJS, ML, AH, JH, EdG, ADM, CNP, IMN, YM, JM, AW, EZ, JMH, CJO, TS, MAN, EMS, YL, CMvD, ASB, JND, CM, NJW, KK, JCD, DL, TLE, PBM, HS, HRW

Table 1. 35 Novel loci (p<5e-9) with concordant direction of effect in all available studies after distance-based (+/-500 kb) and LD (r2 >0.1) pruning

SNP	CHR:BP	Trait	Nearest Gene	Effect Allele	Other Allele	EAF	Effect	SE	P-value	N_eff	P_het
rs36209093	1:110229787	DBP	GSTM1	T	C	0.688	0.170	0.022	9.94E-15	566609	0.640
rs6669446	1:118223275	DBP	TENT5C	C	T	0.421	-0.089	0.015	4.29E-09	1016030	0.876
rs817140	1:193271526	SBP	LINC01031	C	T	0.276	0.161	0.027	2.52E-09	995667	0.558
rs300753	2:209622	PP	SH3YL1	T	C	0.548	0.106	0.017	1.06E-09	975060	0.487
rs57503539	2:9803203	DBP	YWHAQ	A	G	0.210	-0.118	0.019	3.42E-10	968278	0.988
rs538180	3:16363689	SBP	OXNAD1,RFTN1	A	T	0.417	-0.151	0.025	8.47E-10	989900	0.833
rs3821817	3:187456904	PP	BCL6	G	C	0.178	-0.135	0.023	4.59E-09	944591	0.274
rs10018970	4:84452950	SBP	GPAT3	A	G	0.501	-0.142	0.024	4.24E-09	995118	0.322
rs7671332	4:152163489	DBP	SH3D19	C	T	0.039	0.233	0.040	4.27E-09	969793	0.803
rs62370646	5:42515027	DBP	GHR	C	A	0.188	-0.119	0.019	7.97E-10	1008790	0.701
rs57989773	6:100629078	DBP	MCHR2-AS1	C	T	0.245	-0.123	0.018	2.49E-11	909846	0.141
rs2286130	7:156990554	SBP	UBE3C	T	C	0.241	-0.167	0.028	2.82E-09	1004680	0.142
rs10087280	8:49391836	DBP	LOC101929268	G	A	0.171	-0.127	0.020	1.86E-10	1011420	0.079
rs11988716	8:57153503	SBP	CHCHD7	G	A	0.134	0.211	0.036	3.62E-09	975452	0.840
rs2978398	8:146130326	SBP	ZNF250	A	G	0.423	-0.152	0.025	9.06E-10	964640	0.645
rs34361301	9:14535119	PP	NFIB	C	T	0.266	0.122	0.020	6.76E-10	974146	0.322
rs61241090	9:35191014	PP	UNC13B	C	T	0.236	-0.123	0.020	1.40E-09	991997	0.658
rs10991952	9:94252964	SBP	NFIL3	G	A	0.299	0.169	0.027	1.98E-10	978737	0.641
rs10819246	9:129643296	DBP	ZBTB34	T	G	0.099	0.166	0.025	5.91E-11	995493	0.341
rs2987903	9:133711263	PP	ABL1	A	G	0.128	-0.154	0.026	2.49E-09	995802	0.631
rs10904910	10:17266389	SBP	VIM-AS1	A	C	0.310	0.155	0.026	2.56E-09	996726	0.481
rs61890399	11:66325484	SBP	ACTN3	C	T	0.104	-0.244	0.042	4.02E-09	927273	0.896
rs4944038	11:73783478	PP	C2CD3	T	A	0.477	-0.101	0.017	4.30E-09	1006450	0.330
rs61909958	11:96151677	DBP	JRKL-AS1	G	C	0.188	-0.123	0.020	6.11E-10	941830	0.165
rs11604175	11:124619407	DBP	VSIG2	T	C	0.256	0.104	0.017	1.99E-09	1000810	0.470
rs3765618	11:128769876	DBP	C11orf45	G	C	0.088	-0.180	0.027	3.87E-11	974839	0.266
rs76637716	12:51355243	DBP	HIGD1C	A	G	0.056	-0.244	0.035	1.62E-12	922798	0.230
rs36563	14:71352648	SBP	PCNX1	G	T	0.845	0.206	0.033	6.21E-10	1001700	0.908
rs28490942	15:51559845	DBP	MIR4713HG	C	G	0.449	-0.089	0.015	3.25E-09	1015690	0.524
rs116643984	15:101791212	PP	CHSY1	A	C	0.163	-0.143	0.024	2.45E-09	954926	0.806
rs12919839	16:56859216	DBP	NUP93	T	C	0.286	-0.099	0.017	2.15E-09	1013420	0.471
rs8056413	16:84082650	DBP	MBTPS1	T	G	0.599	-0.093	0.016	1.75E-09	989746	0.718
rs34139656	16:88534923	PP	ZFPM1	G	A	0.327	-0.117	0.019	7.30E-10	939836	0.819
rs880132	18:7131618	SBP	LAMA1	C	T	0.573	-0.182	0.027	1.04E-11	846466	0.347
rs117777118	18:77161324	DBP	NFATC1	A	G	0.040	-0.358	0.049	2.40E-13	636875	0.446

SNPs are ordered by chromosome and position. SNP = dbSNP accession number; CHR:BP = chromosome and build 37 position; Trait = primary BP trait for which the most significant association was observed and for which summary statistics are provided in subsequent columns; Nearest Gene = most proximal gene within 250kb of sentinel SNP; Effect allele = allele corresponding to measured effect on the outcome; Other allele = allele not corresponding to measured effect on the outcome; EAF = effect allele frequency in the meta-analysis; Effect = measured effect in the meta-analysis (mmHg); SE = standard error of the measured effect in the meta-analysis; P-value = association p-value for the measured effect in the meta-analysis; N_eff = effective number of subjects in the GWAS meta-analysis (calculated at study-level as N * SNP imputation quality INFO); P_het = value for Cochran’s Q test of statistical heterogeneity in the GWAS meta-analysis.

Table 2. 78 Additional novel (p<5e-8) loci with concordant direction of effect in all available studies after distance-based (+/-500 kb) and LD (r2 >0.1) pruning

SNP	CHR:BP	Trait	Nearest Gene	Effect Allele	Other Allele	EAF	Effect	SE	P-value	N_eff	P_het
rs2320590	1:21155195	DBP	EIF4G3	T	C	0.550	0.085	0.015	1.38E-08	1021190	0.559
rs12134085	1:40763095	PP	COL9A2	T	C	0.198	-0.133	0.023	1.05E-08	864823	0.003
rs2092867	1:61877445	SBP	NFIA	A	C	0.647	0.145	0.025	7.40E-09	997972	0.928
rs10889711	1:68143195	DBP	GADD45A	C	T	0.631	-0.091	0.016	6.57E-09	1000190	0.628
rs12084868	1:72229240	PP	NEGR1	A	G	0.028	0.302	0.055	4.67E-08	898662	0.546
rs12145044	1:93524045	SBP	MTF2	G	T	0.047	-0.328	0.060	4.10E-08	931542	0.068
rs565522	1:112261533	DBP	RAP1A	C	T	0.435	-0.086	0.015	1.74E-08	986922	0.857
rs71664847	1:115019239	PP	TRIM33	T	A	0.190	0.126	0.022	9.99E-09	993341	0.633
rs4573493	1:166023209	SBP	FAM78B	C	T	0.491	-0.138	0.024	1.42E-08	976312	0.204
rs6723772	2:12994692	SBP	TRIB2	T	C	0.105	-0.227	0.040	1.55E-08	965181	0.383
rs10172510	2:32620888	SBP	BIRC6	A	G	0.439	0.134	0.024	2.85E-08	1002860	0.389
rs56350535	2:39061959	SBP	DHX57	A	G	0.123	-0.207	0.038	3.85E-08	955006	0.953
rs75243511	2:54738168	SBP	SPTBN1	C	T	0.045	0.338	0.060	1.42E-08	959440	0.343
rs6729623	2:105205551	SBP	LINC01102	G	A	0.496	-0.141	0.024	5.88E-09	984559	0.319
rs11123059	2:125429006	SBP	CNTNAP5	A	G	0.568	0.134	0.024	3.47E-08	994840	0.954
rs11690153	2:127839534	SBP	BIN1	C	T	0.194	0.174	0.032	4.48E-08	925702	0.381
rs13022015	2:128822702	SBP	UGGT1	C	A	0.185	-0.173	0.031	3.08E-08	981303	0.417
rs10208493	2:196590414	PP	SLC39A10	T	C	0.571	-0.099	0.017	1.36E-08	986727	0.186
rs2306623	3:25424929	DBP	RARB-AS1	C	T	0.670	0.093	0.016	5.22E-09	1013160	0.202
rs9877020	3:43992455	SBP	MIR138-1	T	C	0.161	0.184	0.033	2.57E-08	988044	0.688
rs62253186	3:69919744	PP	MITF	G	C	0.061	0.217	0.038	7.14E-09	920630	0.940
rs844218	3:71607861	SBP	FOXP1	G	A	0.688	-0.143	0.026	4.03E-08	989767	0.465
rs6805393	3:117492152	DBP	LINC02024	A	G	0.508	-0.083	0.015	3.27E-08	1021160	0.244
rs73231988	3:136692308	DBP	IL20RB	A	G	0.116	0.135	0.024	1.45E-08	982317	0.956
rs6822301	4:72002332	DBP	SLC4A4	G	A	0.198	0.108	0.019	1.73E-08	978454	0.493
rs7665985	4:153006312	SBP	LINC02273	C	T	0.641	-0.141	0.026	3.85E-08	954739	0.167
rs9685837	4:187818466	DBP	FAT1	A	G	0.307	-0.092	0.016	1.98E-08	990892	0.731
rs172906	5:38616887	DBP	LIFR-AS1	C	A	0.558	0.095	0.016	7.13E-09	853173	0.837
rs13162174	5:39444718	SBP	DAB2	T	G	0.601	-0.143	0.025	6.45E-09	998700	0.863
rs10061553	5:58352210	DBP	PDE4D	T	C	0.312	-0.089	0.016	4.60E-08	1002160	0.953
rs34237622	5:76884661	DBP	OTP	A	G	0.164	-0.114	0.021	3.72E-08	974348	0.349
rs9477605	6:10034452	DBP	TFAP2A	A	G	0.353	0.087	0.016	3.22E-08	1013520	0.917
rs9370995	6:17477425	DBP	CAP2	G	C	0.536	0.084	0.015	3.31E-08	993051	0.460
rs4053778	6:85988429	PP	LINC02535	G	A	0.395	-0.103	0.018	8.67E-09	968001	0.332
rs72943226	6:99548729	PP	MIR548AI	A	G	0.319	0.102	0.018	3.12E-08	999371	0.245
rs1546722	6:109625797	DBP	CCDC162P	G	A	0.517	-0.087	0.015	7.46E-09	1017590	0.350
rs9320778	6:121258543	PP	TBC1D32	T	C	0.750	0.115	0.020	1.05E-08	973244	0.750
rs67615620	7:15421023	PP	AGMO	C	T	0.199	0.122	0.022	2.32E-08	973513	0.463
rs75177877	7:16117030	PP	CRPPA	T	C	0.172	0.136	0.024	7.02E-09	941294	0.169
rs11136373	8:1212030	SBP	DLGAP2	G	C	0.633	0.141	0.026	4.40E-08	943472	0.370
rs2953937	8:34164285	PP	LINC01288	C	A	0.133	0.143	0.026	1.76E-08	993180	0.135
rs36036692	8:108319395	PP	ANGPT1	G	C	0.374	0.100	0.018	1.89E-08	997711	0.167
rs6982341	8:134229535	DBP	CCN4	G	A	0.581	0.085	0.015	1.74E-08	1026530	0.203
rs9886857	9:27230388	SBP	TEK	A	G	0.151	-0.187	0.034	3.13E-08	993289	0.944
rs112324977	9:80751434	PP	CEP78	A	T	0.131	-0.147	0.026	1.07E-08	992833	0.763
rs2224858	9:83432105	SBP	TLE1	G	A	0.815	0.179	0.031	7.37E-09	1006540	0.092
rs72751391	9:122890934	PP	MIR147A	T	C	0.125	0.156	0.027	1.23E-08	907738	0.942
rs3861882	9:132465304	SBP	PRRX2	C	T	0.281	-0.147	0.027	4.79E-08	994844	0.137
rs11022023	11:11793978	PP	MIR8070	A	G	0.083	-0.174	0.032	4.80E-08	964778	0.539
rs1319701	11:19736996	SBP	NAV2	G	T	0.497	0.134	0.024	3.19E-08	984376	0.499
rs11212666	11:108350451	DBP	POGLUT3	T	A	0.413	0.085	0.016	4.39E-08	966213	0.971
rs77759442	11:110657616	PP	ARHGAP20	T	C	0.132	0.150	0.026	5.91E-09	960047	0.358
rs17542254	11:113655696	SBP	CLDN25	G	A	0.278	0.148	0.027	3.98E-08	991891	0.368
rs1062298	12:12045264	PP	ETV6	T	G	0.421	0.097	0.018	3.24E-08	977558	0.084
rs12828693	12:46385848	PP	SCAF11	T	C	0.205	0.125	0.022	8.20E-09	970335	0.806
rs1732235	12:52418075	DBP	NR4A1	C	T	0.498	-0.087	0.015	8.18E-09	1010320	0.293
rs278123	12:120124578	SBP	CIT	A	G	0.318	0.142	0.026	4.37E-08	996302	0.300
rs56312513	13:38249726	DBP	TRPC4	A	C	0.261	0.098	0.017	1.12E-08	1007860	0.013
rs190533862	13:40671137	SBP	LINC00332	A	T	0.064	0.288	0.052	2.75E-08	913864	0.154
rs9596839	13:54264395	SBP	LINC00558	A	G	0.291	-0.155	0.027	5.87E-09	985600	0.471
rs4517643	13:94417873	PP	GPC6	C	A	0.566	0.101	0.018	7.47E-09	990764	0.633
rs9554446	13:98859019	PP	FARP1	A	T	0.095	0.166	0.030	2.32E-08	974068	0.586
rs7350752	14:21841154	DBP	SUPT16H	A	G	0.123	-0.147	0.026	1.89E-08	782069	0.270
rs2774052	14:59900020	DBP	GPR135	G	A	0.543	0.087	0.015	1.07E-08	1007010	0.517
rs2041330	14:71874638	DBP	SIPA1L1	G	A	0.440	0.084	0.015	3.42E-08	1002690	0.698
rs12883344	14:84911548	DBP	SNORD3P3	A	C	0.399	0.083	0.015	4.94E-08	1019700	0.507
rs7160184	14:88825415	SBP	SPATA7	T	C	0.094	-0.241	0.042	6.22E-09	993970	0.175
rs9671694	14:103330144	PP	TRAF3	G	C	0.337	0.104	0.019	2.40E-08	961925	0.743
rs2034879	15:72429989	DBP	SENP8	A	G	0.737	0.097	0.018	4.39E-08	946840	0.980
rs983353	15:82186535	DBP	MEX3B	G	A	0.302	0.091	0.017	3.65E-08	995889	0.217
rs7174977	15:94214587	DBP	LINC02207	T	A	0.637	0.091	0.016	8.15E-09	993989	0.598
rs12943001	17:78238645	PP	RNF213	C	T	0.641	-0.111	0.020	2.29E-08	830757	0.064
rs9675039	17:81036344	SBP	METRNL	A	G	0.367	0.146	0.026	1.07E-08	953745	0.372
rs17766830	18:44040660	SBP	RNF165	C	T	0.263	0.165	0.029	1.16E-08	919040	0.980
rs72917789	18:46461487	SBP	SMAD7	T	C	0.069	-0.277	0.049	1.14E-08	973426	0.329
rs2125578	19:44746657	DBP	ZNF227	T	C	0.539	-0.083	0.015	2.70E-08	1022260	0.556
rs146827176	20:35169916	DBP	DLGAP4-AS1	T	C	0.048	-0.205	0.037	2.75E-08	940533	0.727
rs855791	22:37462936	PP	TMPRSS6	G	A	0.563	-0.096	0.017	3.24E-08	996826	0.861

SNPs are ordered by chromosome and position. SNP = dbSNP accession number; CHR:BP = chromosome and build 37 position; Trait = primary BP trait for which the most significant association was observed and for which summary statistics are provided in subsequent columns. Nearest Gene = most proximal gene within 250kb of sentinel SNP; Effect allele = allele corresponding to measured effect on the outcome; Other allele = allele not corresponding to measured effect on the outcome; EAF = effect allele frequency in the meta-analysis; Effect = measured effect in the meta-analysis (mmHg); SE = standard error of the measured effect in the meta-analysis; P-value = association p-value for the measured effect in the meta-analysis; N_eff = effective number of subjects in the GWAS meta-analysis (calculated at study-level as N * SNP imputation quality INFO); P_het = value for Cochran’s Q test of statistical heterogeneity in the GWAS meta-analysis.

Table 3. Variance explained in SBP, DBP and PP for all four GRSs and the full PRS analysed in independent Lifelines data.

GRS	SBP		DBP		PP
GRS	VE (%)	P-value	VE (%)	P-value	VE (%)	P-value
1) all 1723 known SNP	6.77	3.60×10^-158	6.77	8.52×10^-158	4.29	5.96×10^-100
2) 113 novel sentinel SNPs	0.06	0.00927	0.08	0.00298	0.02	0.0741
3) 1723 known + 113 sentinel SNPs	6.80	6.67×10^-159	6.83	3.48×10^-159	4.29	7.05×10^-100
4) 1723 known + 113 sentinel SNPs + 267 secondary SNPs	6.93	4.97×10^-162	6.92	2.55×10^-161	4.47	3.73×10^-104
5) optimal full PRS	7.17	7.25×10^-168	7.83	3.63×10^-183	4.53	1.60×10^-105

GRS = genetic risk score; SBP = systolic blood pressure; DBP = diastolic blood pressure; PP = pulse pressure; VE = variance explained by the risk score for the respective blood pressure trait expressed as a percentage; P-value = association p-value for the risk score with the respective blood pressure trait; PRS = polygenic risk score.

Table 4. Prioritized genes through converging evidence across analyses

Gene	SNP	GWAS P_min	GWAS Trait_min	Prior TWAS	^aTWAS_SBP	^aTWAS_DBP	^aTWAS_PP	^bCOLOC_SBP	^bCOLOC_DBP	^bCOLOC_PP	Count/Drug
GSTM2	rs36209093	9.94E-15	DBP	FALSE	-----	----↓	-----	-----	-----	-----	-
PCNX	rs36563	6.21E-10	SBP	FALSE	↑↑↑--	-----	-----	FTT--	-----	-----	-
UBE3C	rs2286130	2.82E-09	SBP	FALSE	-↓---	-----	-↓---	-T---	-----	-T---	-
COL9A2	rs12134085	1.05E-08	PP	TRUE	-----	-----	↓↓---	-----	-----	TT---	-
BIN1	rs11690153	4.48E-08	SBP	FALSE	↑---↑	-----	-----	T---T	-----	-----	-
TRIM33	rs71664847	9.99E-09	PP	FALSE	-----	-----	-↑---	-----	-----	-T---	-
SCAF11	rs12828693	8.20E-09	PP	FALSE	-----	-----	----↓	-----	-----	----F	-
GPC6	rs4517643	7.47E-09	PP	FALSE	-----	-----	↑↑↑--	-----	-----	TTT--	-
SLC39A10	rs10208493	1.36E-08	PP	FALSE	-----	-----	-↓↓↓↓	-----	-----	-TTTT	-
IL20RB	rs73231988	1.45E-08	DBP	FALSE	↓↓---	↓----	-----	FF---	F----	-----	-
PRRX2	rs3861882	4.79E-08	SBP	FALSE	↑----	-----	↑↑---	T----	-----	TT---	-
TMEM51	rs7553381	6.48E-09	SBP	FALSE	---↑-	-----	-----	---F-	-----	-----	-
FUBP1	rs750720	9.01E-10	PP	FALSE	-↑---	↑↑↑--	-----	-F---	FFF--	-----	-
CASQ2	rs4073778	5.00E-13	PP	TRUE	-----	-----	-↑↑↑-	-----	-----	-TTT-	-
MEF2D	rs1185700	9.68E-12	PP	FALSE	-↑---	-----	-↑---	-T---	-----	-T---	-
FGFR2	rs12255289	2.97E-09	DBP	FALSE	-----	↑↑---	-----	-----	TF---	-----	NINTEDANIB, BRIVANIB, PALIFERMIN, THALIDOMIDE, FOSTAMATINIB
ABCC8	rs77889556	2.91E-08	PP	TRUE	-----	-----	-↓---	-----	-----	-T---	12
FOXN3	rs7151849	1.77E-10	PP	FALSE							-
CKB	rs8017780	1.40E-10	PP	FALSE	-----	↓----	-----	-----	F----	-----	-
ARID3B	rs74781061	1.81E-09	DBP	FALSE	-----	-↓---	-----	-----	-F---	-----	-
NAGLU	rs86312	4.94E-08	SBP	FALSE	-----	-----	↑↑---	-----	-----	TT---	N-ACETYLGLUCOSAMINE
MYL12A	rs7811	1.14E-10	PP	FALSE	-----	-----	↓↓↓--	-----	-----	TTT--	-
ACTN4	rs2303040	2.00E-10	PP	FALSE	-----	-----	↑-↑↑-	-----	-----	T-TT-	-
CCDC97	rs56254331	1.18E-10	DBP	FALSE	-----	---↑-	-----	-----	---F-	-----	-
KLHL23	rs78843689	1.08E-08	DBP	FALSE	-↓---	-----	-----	-F---	-----	-----	-
RP11-460N16.1	rs9868203	1.28E-08	PP	FALSE	-----	-----	-↓---	-----	-----	-T---	-
SLC15A2	rs9842387	3.20E-08	SBP	FALSE	↑-↑↑-	-----	-----	T-TT-	-----	-----	BENZYLPENICILLIN
DNAJC13	rs2369796	3.32E-08	DBP	FALSE	-----	↑---↑	-----	-----	F---F	-----	-
ANKH	rs2921604	4.17E-08	DBP	FALSE	----↓	----↓	-----	----F	----F	-----	-
LNPEP	rs114772891	5.72E-09	SBP	FALSE	-↑---	-↑--↑	-----	-T---	-T--T	-----	-
BTN2A1	rs2893856	3.00E-11	DBP	FALSE	----↑	----↑	-----	----F	----F	-----	-
NOTCH4	rs2849017	1.78E-09	DBP	TRUE	↓↓↓--	-----	↓↓↓↓-	TFF--	-----	TTFT-	-
AMZ1	rs798538	3.50E-10	DBP	FALSE	-----	--↑--	-----	-----	--F--	-----	-
GRB10	rs79617314	1.36E-09	PP	FALSE	-----	↓----	↑----	-----	F----	T----	-
CLIP2	rs229872	2.08E-08	DBP	FALSE	-----	↑----	-----	-----	F----	-----	-
GTF2IRD1	rs37613	3.08E-08	SBP	FALSE	↑----	-----	-----	T----	-----	-----	-
ADRA1A	rs58623861	9.07E-10	DBP	FALSE	-↑---	-↑---	-----	-F---	-F---	-----	86

Selection criteria: evidence from TWAS and nearest-gene mapping of sentinel SNPs from GWAS meta-analysis; Gene = Gene was significant in genetically predicted gene expression analysis using S-PrediXcan for aorta, tibial artery, left ventricle, atrial appendage, and whole blood tissues and was annotated using ANNOVAR as the gene nearest the sentinel SNP at that locus; SNP = Sentinel SNP from GWAS meta-analyses for each independent locus; GWAS P_min = minimum p-value across all GWAS meta-analyses; GWAS Trait_min = BP trait corresponding to the GWAS P_min; Prior TWAS = Indicates if the association was replicated in the previous TWAS performed by Giri et al; TWAS_SBP = indicates the direction of effect for significant associations in the SBP TWAS in aorta, tibial artery, left ventricle, atrial appendage, and whole blood tissues, respectively; TWAS_DBP = indicates the direction of effect for significant associations in the DBP TWAS in aorta, tibial artery, left ventricle, atrial appendage, and whole blood tissues, respectively; TWAS_PP = indicates the direction of effect for significant associations in the PP TWAS in aorta, tibial artery, left ventricle, atrial appendage, and whole blood tissues, respectively; COLOC_SBP = indicates if the gene met the posterior probability threshold of ≥90% for colocalization of SBP association and gene expression in aorta, tibial artery, left ventricle, atrial appendage, and whole blood tissues, respectively; COLOC_DBP = indicates if the gene met the posterior probability threshold of ≥90% for colocalization of DBP association and gene expression in aorta, tibial artery, left ventricle, atrial appendage, and whole blood tissues, respectively; COLOC_PP = indicates if the gene met the posterior probability threshold of ≥90% for colocalization of PP association and gene expression in aorta, tibial artery, left ventricle, atrial appendage, and whole blood tissues, respectively; Count/Drug = Column summarizes drugs targeting genes when the number of drugs is ≤5 or the total number of drugs targeting the gene when the number of drugs is >5 which were identified to interrelate with the gene from the following databases: Guide to Pharmacology Interactions, DTC, DrugBank, JAX-CKB, My Cancer Genome, PharmGKB, Clearity Foundation Clinical Trials, TDG Clinical Trials, TALC, TTD, TEND, and/or ChEMBL Interactions.

^aIndicates whether gene expression was positively associated (↑), negatively associated (↓), or non-significant (-) in S-PrediXcan analyses
^bIndicates whether posterior probability was ≥90% (T), <90% (F), or unavailable (-) in COLOC analyses

We conducted a single-stage BP GWAS meta-analysis of individuals of European ancestry evaluating common SNPs. SBP, DBP, and PP GWAS summary statistics from each study were from linear regression models analyzing SNP associations adjusted for age at BP measurement, age², sex, BMI, and the top 10 genetic principal components. Inferences were limited to SNPs with imputation quality (INFO) scores of 0.1 or higher, Hardy-Weinberg equilibrium p-values greater than or equal to 1x10^-6, and MAF greater than or equal to 1%. PP was calculated in each study as the difference between SBP and DBP.

Study Populations

The total sample size for this investigation was up to 1,028,980 adults from the meta-analysis of four BP GWAS datasets: UKB, ICBP, MVP, and BioVU. Characteristics of these studies are presented in Supplementary Table 1. ICBP is a large meta-analysis of 77 studies, therefore descriptive characteristics were not available.

UKB

UKB includes ~500,000 volunteers aged 40-69 years of age ascertained through NHS registers. Following informed consent participants completed a standardized questionnaire on life course exposures, medical history and treatments and underwent a standardized portfolio of phenotypic tests including two BP measurements taken seated after two minutes rest using an appropriate cuff and an Omron HEM-7015IT digital BP monitor. A manual sphygmomanometer was used if the standard automated device could not be employed. Body mass index (BMI) was calculated as weight (kg) divided by height squared (m²) with weight measured using an electronic weighing scale (Tanita BC-418). The participants undergo longitudinal life course linkage to electronic health data including Hospital Episode Statistics and Office for National Statistics cause of death data.

DNA extraction and genotyping for UKB has been previously described¹¹. Briefly, UKB genetic data includes genotypes for 488,377 individuals. DNA was extracted from stored blood samples and genotyping was carried out by Affymetrix Research Services Laboratory. 49,950 participants involved in the UK Biobank Lung Exome Variant Evaluation (UK BiLEVE) study were genotyped at 807,411 markers using the Affymetrix UK BiLEVE Axiom Array and 438,427 participants were genotyped using the Affymetrix UK Biobank Axiom Array (825,927 markers), which shares 95% of marker content with the UK BiLEVE Axiom Array.

Variants were imputed centrally by UK Biobank using a reference panel that merged the UK10K and 1000 Genomes Phase 3 panel as well as the Haplotype Reference Consortium (HRC) panel¹². For current analysis only SNPs imputed from the HRC panel were analyzed (N=39,235,157), of which ~7.1 million SNPs with minor allele frequency (MAF) >1% and imputation quality INFO >0.1 are analyzed here for GWAS.

For the UKB GWAS, we calculated the mean SBP and DBP values from two automated (N=418,755) or two manual (N=25,888) BP measurements. For individuals with one manual and one automated BP measurement (N=13,521), we used the mean of these two values. For individuals with only one available BP measurement (N=413), we used this single value. Following both genetic and phenotypic data QC and by excluding pregnant women (n=372) and those individuals who had withdrawn consent (N=36), the sample size for analysis therefore included N=458,577 and N=458,575 self-reported European-ancestry individuals for SBP and DBP, respectively. For measures taken while a patient was on an antihypertensive medication, we added 15 mm Hg to SBP and 10 mm Hg to DBP. We performed linear mixed model (LMM) association testing under an additive genetic model of the three (untransformed) continuous, medication-adjusted BP traits (SBP, DBP, PP) for all measured and imputed genetic variants in dosage format using the BOLT-LMM (v2.3) software.

ICBP

ICBP GWAS is an international consortium to investigate BP genetics and has been previously described elsewhere^6,13. All study participants were of European descent and were imputed to either the 1000 Genomes Project Phase 1 integrated release version 3 [March 2012] all ancestry reference panel or the HRC panel. The final ICBP GWAS dataset included 77 studies comprising data from 299,024 individuals. Three quantitative BP traits were analyzed: SBP, DBP, and PP. Within each study, BP measures were adjusted for medication use by adding 15 and 10 mm Hg to SBP and DBP, respectively.

Prior to meta-analysis of all 77 ICBP GWAS studies, we undertook central quality control checks across all studies. This included checks to ensure allele frequency consistency (across studies and with reference populations), checks of effect size and standard error distributions (i.e., to highlight phenotype issues) and generation of quantile-quantile (QQ) plots and genomic inflation factor lambdas to check for over- or under-inflation of test statistics. Genomic control was applied (if lambda>1) at study-level. Variants with imputation quality <0.1 were excluded prior to meta-analysis. EasyQC was used for the quality control process¹⁴. Finally, data were filtered to SNPs with MAF ≥1% and effective sample size (reflecting the quality of genotype imputation) >60% of the total effective sample size. Meta-analysis was performed using METAL software employing inverse variance weighted fixed-effects models¹⁵. Between-study heterogeneity was assessed using the Cochran’s Q statistic and we performed additional filtering removing heterogeneous variants with Cochran’s Q p<1x10^-4.

MVP Study

The MVP study is a large cohort of fully consented participants who were recruited from the patient populations of 63 Department of Veterans Affairs (VA) medical facilities. Summary statistics from the analysis of 220,501 self-reported non-Hispanic white participants were included in our meta-analysis. These results have been previously reported by Giri et al⁴. Briefly, DNA was extracted from whole blood and genotyped using a custom Affymetrix array (Axiom Biobank; Thermo Fischer Scientific Inc, Waltham, MA, USA). Genotype calling and QC were performed centrally and genotypes were phased using EAGLE v2¹⁶ and imputed from the 1000 Genomes Project phase 3 version 5 reference panel using Minimac3¹⁷ software. Participants included adults (age ≥18 years) with non-Emergency Department outpatient SBP and DBP measures available in their electronic health record. For individuals with greater than or equal to three measures available, median SBP and corresponding DBP were used in analysis. For rare cases where fewer than three measures were available, the lowest available SBP and corresponding DBP were used. We observed an average of 220 measures across individuals. In individuals in whom the median SBP value was observed at multiple clinical encounters on distinct dates, we used the earliest of those measures to identify the DBP, age, BMI, and anti-hypertensive treatment status of the individual at that time. Measures were ineligible if they occurred at or after an International Classification of Diseases Ninth Revision (ICD-9) code from the groups 585 (chronic kidney disease), 405 (secondary hypertension), or 428 (heart failure). If pain scores were available, BP measures taken during encounters when a pain score ≥5 was recorded were also ineligible. BP measures were adjusted for medication use by adding 15 and 10 mm Hg to SBP and DBP, respectively. Linear regression association tests were conducted using additive models for untransformed medication-adjusted BP traits (SBP, DBP, PP) using SNPTEST-v2.5.4-beta¹⁸.

BioVU

The BioVU DNA Repository is a deidentified database of electronic health records that are linked to patient DNA samples at Vanderbilt University Medical Center. Summary statistics from the analysis of 50,649 self-reported non-Hispanic white participants were included in our meta-analysis. A detailed description of the database and how it is maintained has been published elsewhere¹⁰. BioVU participant DNA samples were genotyped on a custom Illumina Multi-Ethnic Genotyping Array (MEGA-ex; Illumina Inc., San Diego, CA, USA). Quality control (QC) was conducted, excluding samples or variants with missingness rates above 2%. Samples were also excluded if consent had been revoked, sample was duplicated, or failed sex concordance checks. Imputation was performed on the Michigan Imputation Server v1.2.4¹⁷ using Minimac4 and the Haplotype Reference Consortium panel v1.1¹².

Among BioVU participants, we selected unrelated self-reported adults of European ancestry (age ≥ 18 years) and used the earliest median eligible non-Emergency Department outpatient measured SBP in the electronic health record, and the corresponding DBP. For individuals with fewer than three measurements available (N=2,933), the lowest available SBP and corresponding DBP were used. On average, there were 69 SBP measures per individual. Measures were considered ineligible if they occurred at or after an ICD-9/10 billing code from the groups 585/N18 (chronic kidney disease), 405/I15 (secondary hypertension), or 428/I50 (heart failure). For measures taken while a patient was on an antihypertensive medication, we added 15 mm Hg to SBP and 10 mm Hg to DBP. We performed linear regression association tests with additive models for untransformed medication-adjusted BP traits (SBP, DBP, PP) using SNPTEST-v2.5.4-beta¹⁸.

Study-level QC

We applied a harmonized QC procedure for each BP trait in all four studies (i.e., 12 GWAS datasets in total) using the GWASInspector R package¹⁹. The 1000 Genomes Project reference panel²⁰, supplemented with the Haplotype Reference Consortium data panel¹², was used as the reference dataset for appropriate flipping and/or switching of the alleles, checking for allele frequency concordance with the 1000 Genomes reference, annotating dbSNP rs accession numbers, and constructing harmonized identifiers for meta-analyses. SNP effect sizes from ICBP were considered as the reference to validate the reported effect sizes from the other three GWAS datasets (Supplementary Figures 1-3)⁷.

The following criteria were then used for filtering the GWAS datasets: i) SNPs only (i.e., no insertions/deletions, copy number variants, etc.); ii) MAF greater than or equal to 1%; iii) imputation quality (INFO) greater than 0.1; iv) HWE p-value greater than or equal to 1x10^-6. Effective sample size (N_EFFECTIVE) was calculated as the product of total sample size and INFO for each SNP.

Meta-analysis

We initially applied LD Score Regression²¹ (LDSR) to the summary statistics for three of our four component datasets (UKB, MVP, and BioVU) to calculate the LDSR intercepts that were used to correct for pre-meta-analysis genomic inflation. ICBP summary statistics, as a meta-analysis of 77 independent cohorts, were previously corrected for genomic inflation⁵. HapMap3²² SNP alleles and pre-calculated LD scores from 1000 Genomes Project²⁰ European reference data supplied with the package were used to calculate LDSR intercepts. Observed LDSR intercepts for each dataset were as follows: 1.2177, 1.2195, and 1.1851 for UKB, 1.0530, 1.0247, and 1.0413 for MVP, and 1.0288, 1.0127, and 1.0207 for BioVU, for SBP, DBP and PP respectively. Inverse-variance weighted fixed-effects meta-analysis of common (MAF≥0.01) bi-allelic SNPS with imputation quality (INFO) greater than or equal to 0.1 across our four studies was performed using METAL¹⁵ software. No further GC correction was applied to the meta-analysis results which combined our four datasets together.

QC of the meta-analysis results

Similar to study-level QC, we used the GWASInspector R package¹⁹ to ensure standardization and perform QC of post-meta-analysis summary statistics. Analyses included: i) checks of allele frequency concordance with the 1000 Genomes reference and concordance of effect sizes with ICBP (Supplementary Figure 4); ii) evaluation of Q-Q plots and genomic inflation factors (Supplementary Figure 4); and iii) evaluation of bivariate scatterplots of key summary statistics to identify patterns indicating the presence of low-quality SNPs (Supplementary Figure 5).

These analyses revealed the presence of SNPs in our data with low effective sample sizes and large standard errors, as well as a sub-peak of SNPs with higher effective sample sizes and large standard errors. Based on these observations, we applied a filtering threshold for SNPs that were present in at least three of our four studies or SNPs that reached an effective sample size greater than or equal to 60% of the maximum (Supplementary Figures 6- 8). Application of these criteria to achieve an optimal balance between quality of retained SNPs and sample size resulted in 7,584,058 SNPs available for analysis.

Distinguishing known from novel loci

Published BP SNPs

We collated published BP GWAS and compiled all 3,800 unique BP SNPs reported to date (Supplementary Tables 2-3). In many BP-GWAS papers the list of previously reported BP variants has focused on the lead sentinel variant and with validated evidence from independent replication. To expand to a fully comprehensive list of known variants, we curated a list of all published common and rare variants, including results from studies conducted in non-European ancestries, all types of methodological analyses including interaction analyses, results from both one-stage and two-stage study designs, and secondary variants reported from conditional or fine-mapping analyses. We began with the list of all 984 SNPs from the total of 901 previously known and novel loci reported from Evangelou et al⁵, then added to these: i) any secondary SNPs reported from conditional analyses in publications up to 2018^5,7,9,23; ii) SNPs reported from a large one-stage discovery analysis prior to 2018⁸; iii) SNPs reported by Giri et al 2019⁴; and all other SNPs from GWAS published between 2018 and the end of 2020^24–30. We removed duplicated SNPs to generate a unique set of ~3,800 SNPs. Subsequent checks of our results in GWAS Catalog³¹ and PhenoScanner³² confirmed that all published BP variants had been successfully captured. For QC purposes, we compared the allele frequencies and the resulting effect estimates of these published SNPs in our GWAS meta-analysis data with the published data.

Linkage disequilibrium analyses

Linkage disequilibrium (LD) was calculated using PLINK³³ with 1000 Genomes Project²⁰ phase 3 version 5 European reference genotypes. LD proxies were captured for the ~3,800 previously reported BP SNPs at an r² threshold >0.8 and a maximum distance of 500 kb. Further, we identified the most strongly associated SNP within 500 kb of each known SNP regardless of LD (i.e., “distance proxies”). The strongest trait-specific associations of these previously reported SNPs, their best LD proxies, and best distance proxies in our meta-analyses are presented in Supplementary Table 4.

We partitioned our data into known and unknown subsets. To identify the “unknown” portion of our GWAS results, we removed previously reported SNPs, SNPs within 500 kb of previously reported SNPs, LD proxies for previously reported SNPs at an r² threshold >0.1 and a maximum distance of 5 Mb, and SNPs within the HLA region of chromosome 6 (25-34 MB) from each of our meta-analyses. QQ plots of all SNPs versus unknown SNPs are shown in Supplementary Figure 9.

Reporting criteria for novel loci

All remaining SNPs reaching genome-wide significance (p<5x10^-8) and consistent direction of effect in all available studies were clumped into 1 Mb regions and the most significant SNP for any trait was selected from each region as a sentinel variant for the locus. Novel sentinel SNPs were checked for pairwise-LD against all other novel sentinel SNPs at an r² >0.1 to confirm independence. Considering our one-stage study design, a significance threshold of p<5x10^-9 was imposed for primary reporting of novel sentinel SNPs, with additional loci subsequently reported at the traditional p<5x10^-8 significance threshold.

Categorizing known variants into independent loci

Similarly, previously reported SNPs, their best LD proxy if the SNP was unavailable in our data, or the best distance proxy if neither was available, were clumped into 1 Mb regions and the most significant SNP for any trait was selected. Selected SNPs were then checked for pairwise-LD against all other selected SNPs at an r² >0.1 to confirm independence. The most significant SNP for any trait was selected within each LD block, and these independent SNPs were designated as known sentinel SNPs.

LDSR approach for determination of polygenicity

We applied LDSR to each of our three meta-analyses (SBP, DBP, and PP), as well as the novel proportion of each meta-analysis, and compared these values with genomic inflation factors to determine if inflation of our test statistics was due to population substructure or polygenicity.

Annotation of variant functions and shared associations for novel signals

Novel signals were extended to their linked variants (r² >0.5) using an in silico sequencing approach³⁴. PLINK³³ was used for LD calculations and ANNOVAR³⁵ software to annotate the nearest genes for novel signals and to annotate variant functions. Then the extended loci (r² >0.8) were used to search the GWAS Catalog³¹ as well as PhenoScanner³² for shared associations (p<5x10^-8).

Conditional Analysis

Genome-wide joint conditional analysis was performed using GCTA-COJO³⁶ specifying a five Mb LD window and a genome-wide significance threshold of 5x10^-8, and using UKB European-ancestry sample genotypes as the LD reference. For each of our three BP traits, summary statistics were analyzed by chromosome to build a stepwise joint conditional model that selected independently associated SNPs. Pairwise-LD was calculated in both the 1000 Genomes Project²⁰ phase 3 version 5 European reference genotypes and UKB European-ancestry sample genotypes. SNPs in LD (r²>0.1 in either UKB or 1000 Genomes reference at ±5 Mb) with known or novel sentinel SNPs from our primary analysis or in LD with known SNPs not available in our data were excluded. Among SNPs identified in conditional analysis, the most significant SNP for any trait was selected within each LD block, and these independent SNPs were designated as secondary SNPs. Secondary SNPs were further evaluated to determine if they fell within the novel portion of our data.

Lifelines Cohort Study genotype and phenotype data

For our study, genetic risk score (GRS) is defined as a risk score comprised of SNPs reaching genome-wide significance (p<5x10^-8) in our analyses or in previously published studies and polygenic risk score (PRS) as a full genome-wide risk score optimized at a selected p-value threshold that explains the maximum trait variance. We calculated GRS and PRS and assessed variance explained in Lifelines data (Supplementary Figure 10). Both GRS and PRS were calculated as the sum of an individual's risk alleles, weighted by BP trait-specific risk allele effect sizes.

The Lifelines cohort is a large prospective population-based cohort study performed in 167,729 individuals living in the North of the Netherlands with a unique three generation design, aiming at investigating risk factors for multifactorial diseases³⁷. It was approved by the medical ethics committee of the University Medical Center Groningen and conducted in accordance with Helsinki Declaration Guidelines. All participants signed an informed consent form prior to enrollment.

A subset of 38,030 volunteers were genotyped using the Infinium Global Screening Array MultiEthnic Disease Version, according to manufacturer’s instructions, at the Human Genomics Facility of the Erasmus Medical Center, Rotterdam and the Department of Genetics, University Medical Center Groningen. Standard QC was performed on both samples and markers. Samples with a genotyping call rate<99%, outliers for heterozygosity and sex mismatches were excluded, as well as samples that did not show consistent information between reported familial information and observed identity-by-descent sharing with family members, and between genotypes available from this and previous studies. Variants with a genotyping call rate <99%, Hardy-Weinberg equilibrium P <1×10⁻⁶ or excess of Mendelian errors in families (>1% of the parent-offspring pairs) were removed. A total of 36,339 samples and 571,420 autosomal and X-chromosome markers passed quality checks. The genotyping dataset was then imputed at the Sanger imputation server1 using the HRC panel v1.1.

From the set of 36,339 samples, we selected 10,782 unrelated individuals who are also independent from Lifelines samples that were included in a previous ICBP meta-GWAS⁵. After excluding 552 children (age<18 years), 12 pregnant women, five individuals without SBP or DBP, and three individuals without BMI, a final total number of 10,210 individuals were included for analyses.

In Lifelines, BP was measured every minute during a period of ten minutes using an automated DINAMAP Monitor (GE Healthcare) and the average of the final three readings was recorded for SBP and DBP. Participants with a measured BP ≥140/90 mm Hg irrespective of treatment and those taking antihypertensive medication (ATC codes C02, C03, C07, C08, C09) irrespective of BP were defined as having hypertension. In continuous trait analyses, 15 mm Hg was added to SBP and 10 mm Hg was added to DBP for 1,236 individuals who were taking antihypertensive medication. PP was calculated using these medication-adjusted BP values.

GRS and PRS construction and percentage of variance explained

To calculate the percentage of BP variance explained by genetic variants in an independent dataset, we generated the residuals from a regression of each BP trait against sex, age, age² and body mass index in 10,210 Lifelines individuals. We then fit a second linear model for the trait residuals with the top ten principal components and a third linear model for the trait residuals with ten principal components plus GRS. The difference in the adjusted R² between the third and the second model is the estimation of the percentage of variance of the dependent (BP) variable explained by the GRS. To evaluate the contribution of previously reported BP loci, as well as novel and secondary loci detected in our analyses, to observed variance in BP traits, and to test the predictive value of our genome-wide results, we constructed four different GRS and a PRS: (i) GRS of 1,723 pairwise-independent (LD-pruned with r²<0.1) SNPs from published known loci; (ii) GRS of 113 sentinel SNPs at genome-wide significant (p<5x10^-8) novel loci; (iii) GRS of 1,723 known SNPs plus 113 sentinel SNPs at genome-wide significant novel loci; (iv) GRS of 1,723 known SNPs plus 113 SNPs from novel loci plus 267 secondary SNPs, and; (v) full PRS at optimally selected p-value threshold (1x10^-3, 0.01, 0.01 for SBP, DBP, and PP, respectively) that maximized variance explained in Lifelines data.

We generated GRS and PRS by multiplying the risk allele dosages for each SNP by its respective effect size as weight, and then summed all SNPs in the score. The four different GRS included the same set of SNPs for all three BP traits (SBP, DBP, and PP), but were weighted by the trait-specific beta coefficients from the GWAS results for SBP, DBP, and PP. Summary statistics for all SNPs in the GRS are displayed in Supplementary Table 5.

For each BP trait, we calculated full PRS by the clumping and thresholding approach³⁸. Summary statistics of final GWAS results for each trait and the LD reference panel of 503 European ancestry samples from 1000 Genomes phase 3²⁰ were used. SNPs with ambiguous strands (A/T or C/G) were removed for the score derivation. An LD-driven clumping procedure was then performed by PLINK version 1.90 (r²<0.1, 1000 kb window). Finally, the PRS were generated at 17 selected P-value thresholds (1x10^-8, 5x10^-8, 1x10^-7, 5x10^-7, 1x10^-6, 5x10^-6, 1x10^-5, 5x10^-5, 1x10^-4, 5x10^-4, 1x10^-3, 5x10^-3, 0.01, 0.05, 0.1, 0.5, 1) and the optimal PRS with the maximum of variance explained in Lifelines data were selected. Summary statistics for all SNPs in optimal PRS of SBP, DBP, and PP are displayed in Supplementary Tables 6a-c.

Decile analyses of full BP PRS for SBP, DBP, PP, and hypertension in Lifelines

To evaluate to what extent BP PRS were predictive for SBP, DBP, PP, and hypertension, we selected the optimal PRS of SBP, DBP, and PP for decile analyses of their respective traits and modeled the joint effect of the optimal PRS for SBP and DBP for hypertension analyses. Then we applied linear and logistic regression with adjustment for sex to compare BP levels and risk of hypertension, respectively, in all deciles versus the bottom decile of the PRS distribution of 10,210 Lifelines individuals. P-values were calculated from the normal distribution for BP traits and from a chi-square distribution with two degrees of freedom for hypertension.

Hypertension model performance and calibration assessment in Lifelines

Hypertension model discrimination probability and calibration were examined by calculating the area under the receiver operating ‎characteristics curve (AUROC)^39,40 and Brier score^41,42, respectively. These analyses were implemented using the pROC R-package⁴³ with 10-fold cross-validation. An AUROC value of 0.5 indicates no discriminative probability, while a value of 1 is a perfect discrimination probability. The Brier score is the average squared difference between predicted probability and observed outcome with values approaching zero indicating high calibration. The cut-off value of hypertension odds to predict high risk were identified using the Youden index, the point on the AUROC where sensitivity and specificity are maximized. Statistics were calculated for two models: 1) a model including covariates used in GWAS-meta-analyses (sex, age, age², BMI; model 1), and 2) a model including covariates and optimized PRS for SBP and DBP (model 2).

Decile analyses of full BP PRS for hypertension in UKB African-ancestry individuals

To evaluate to what extent BP PRS were predictive for hypertension in non-European ancestry individuals, we modeled the joint effect of the optimal PRS for SBP and DBP on odds of hypertension in UKB African-ancestry individuals.

SNPs were imputed centrally by UKB using a reference panel that merged the UK10K and 1000 Genomes Phase 3 panel as well as the Haplotype Reference Consortium (HRC) panel. For our analysis, only SNPs imputed from the HRC panel were considered. Data were limited to genotyped and imputed variants with imputation INFO scores of 0.4 or higher, HWE p-values >5x10^-8, and MAF >0.01.The hypertension phenotype was provided by UKB.

We performed our analysis using sex-adjusted logistic regression to compare risk of hypertension in all deciles versus the bottom decile of the distribution of 3,341 UKB self-reported “Black” individuals. P-values were calculated from a chi-square distribution with two degrees of freedom.

Comparison of REML methods to calculate heritability

The h_SNP² of BP traits has previously been calculated within the N~457k UKB cohort GWAS dataset using the restricted maximum likelihood (REML) method BOLT-REML⁴⁴, e.g. with h_SNP²=21.3% for SBP⁵. To check the consistency across different software, and to compare to previously published results, we calculated h_SNP²of SBP within the UKB BP-GWAS dataset using GCTA-GREML³⁶. The full imputed genetic data was converted from BGEN dosage format into hard-call genotyped PLINK format. SNPs were filtered according to MAF >1% and high imputation quality with INFO ≥0.9 from the central UKB QC, and then restricted to only the set of SNPs present in our full meta-analysis dataset. Due to the high memory RAM that GCTA software requires, we selected a representative subset from UKB for our analysis. We calculated percentiles of principal components PC1 & PC2 of all individuals from the centrally provided UKB QC data and extracted the most homogeneous subset of individuals centered around the median data-points with both PC1 and PC2 within the 40-60^th percentile range, resulting in a subset sample size of N=19,410. Within GCTA the genetic relatedness matrix (GRM) was generated for each autosome separately, then merged together and filtered for relatedness according to a 0.2 cut-off to remove any first and second degree relatives. Then h_SNP² for SBP was calculated with adjustment of the same covariates applied to the UKB BP-GWAS, namely: sex, age, age², BMI, genotyping chip array and the top ten PCs. One-tailed p-values were calculated according to the h_SNP²and SE results in base R.

Heritability analyses in Lifelines data

We used GCTA-GREML⁴⁵ to calculate h_SNP² for BP in exactly the same Lifelines dataset as in the variance explained analyses (N=10,210). SNPs in Lifelines were restricted to the same list of SNPs used in the UKB GCTA-GREML⁴⁵ analyses. Then h_SNP² for SBP, DBP, and PP was calculated with adjustment of sex, age, age², BMI, and ten PCs.

In Silico Transcriptome-wide association study

Genetically Predicted Gene Expression Analysis

Our in silico transcriptome-wide association study (TWAS) was performed using S-PrediXcan⁴⁶, an approach that imputes genetically predicted gene expression in a given tissue and tests predicted expression for association with a GWAS outcome using SNP-level summary statistics. For this study, input included summary statistics from each of the meta-analyses (SBP, DBP, and PP) and gene-expression references for five tissues from GTEx⁴⁷ v7 including aorta, tibial artery, left ventricle, atrial appendage, and whole blood. Our analyses incorporated covariance matrices based on 1000 Genomes²⁰ European populations to account for LD structure. Bonferroni-corrected significance threshold was 1.55x10^-6 to account for the total number of gene models assessed across all tissues in these analyses.

Colocalization analysis

The hypothesis that a single variant underlies GWAS and eQTL associations at a given locus (i.e. colocalization) was tested using coloc⁴⁸, a Bayesian gene-level test that evaluates GWAS and eQTL association summary statistics at each SNP at the locus and provides gene- and SNP-level posterior probabilities for colocalization. For this analysis, input included results for common variants in our study and eQTL summary statistics corresponding to the gene-expression references used in S-PrediXcan analysis, restricting to only variants included in the S-PrediXcan models. Output includes posterior probabilities for the null hypothesis (PP.H0) that SNPs at the locus are associated with neither gene expression nor the outcome (i.e. SBP, DBP or PP), the first alternative hypothesis (PP.H1) that SNPs are associated with expression but not the outcome, the second alternative hypothesis (PP.H2) that SNPs are associated with the outcome but not expression, the third alternative hypothesis (PP.H3) that SNPs are associated with both expression and the outcome but not colocalized, and the fourth alternative hypothesis (PP.H4) that SNPs associated with both expression and the outcome are colocalized. Also included are annotations of the SNP with the highest PP.H4 at each locus and the corresponding posterior probability. A PP.H4 greater than 90% was considered evidence of colocalization.

Data availability statement

Full summary statistics of our analyses are available from the study authors upon request. Summary statistics for sentinel SNPs for each BP-trait, as well as optimized PRS, are available in Supplementary Tables. Statistically significant reports for S-PrediXcan results for all 5 tissues for all BP-traits evaluated are also made available in the Supplementary Tables.

Ethics statement

Our study is based on meta-analysis of previously published, publicly available data for which appropriate site-specific Institutional Review Boards and ethical review at local institutions have previously approved use of this data.

Yes there is potential Competing Interest. M.A.N.’s participation in this project was part of a competitive contract awarded to Data Tecnica International LLC by the National Institutes of Health to support open science research, he also currently serves on the scientific advisory board for Clover Therapeutics and is an advisor to Neuron23 Inc as a data science fellow. B.M.P. serves on the Steering Committee of the Yale Open Data Access Project funded by Johnson & Johnson. P.V. received an unrestricted grant from GlaxoSmithKline to build the CoLaus study (2003). V.S. has received honoraria for consulting from Novo Nordisk and Sanofi and has ongoing research collaboration with Bayer Ltd. (all unrelated to this project). R.L. is a part-time consultant of Metabolon, Inc. M.J.C. is Chief Scientist for Genomics England, a UK Government company. J.M.M.H. is a full-time employee of Novo Nordisk Ltd. C.J.O. is currently employed by Novartis Institutes for Biomedical Research (unrelated to this project) and remains credentialed as a without compensation researcher with the Veterans Administration. T.S. is co-founder of Zoe Ltd. A.S.B. reports institutional grants from AstraZeneca, Bayer, Biogen, BioMarin, Bioverativ, Novartis, Regeneron and Sanofi. J.N.D. reports grants, personal fees and non-financial support from Merck Sharp & Dohme (MSD), grants, personal fees and non-financial support from Novartis, grants from Pfizer and grants from AstraZeneca outside the submitted work. J.N.D. sits on the International Cardiovascular and Metabolic Advisory Board for Novartis (since 2010); the Steering Committee of UK Biobank (since 2011); the MRC International Advisory Group (ING) member, London (since 2013); the MRC High Throughput Science 'Omics Panel Member, London (since 2013); the Scientific Advisory Committee for Sanofi (since 2013); the International Cardiovascular and Metabolism Research and Development Portfolio Committee for Novartis; and the Astra Zeneca Genomics Advisory Board (2018). E.E is co-founder has received consultation fees from Open DNA Ltd (unrelated to this project).

Download PDF

Journal Publication

published 30 Apr, 2024

Read the published version in Nature Genetics →

Version 1

posted

You are reading this latest preprint version

Genome-wide analysis in over 1 million individuals reveals over 2,000 independent genetic signals for blood pressure

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Results

Discussion

References

Declarations

Tables

methods

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1