Abstract Body: INTRODUCTION
Elevated blood pressure (BP) is a leading contributor to global morbidity and mortality, with notable ancestral and racial differences in BP traits in the United States and globally. To date, most BP genome-wide association studies (GWAS) are predominately European (EUR), with the current largest study being EUR only.

METHODS
We combined newly conducted GWAS with publicly available summary statistics to study three BP traits: systolic [SBP], diastolic [DBP], and pulse pressure [PP]. BP traits were regressed onto additively encoded genotypes, adjusted for age, age-squared, sex, body mass index, and the first 10 principal components. Using METAL, we performed fixed-effects inverse-variance-weighted meta-analyses including over 2.5M individuals (35% non-EUR), within and across ancestry groups.

RESULTS
Multi-ancestry meta-analyses of the autosomes identified 1,351 significant independent loci (GCTA software, P < 5x10^-8, minor allele frequency > 0.01) for SBP, 1,206 for DBP, and 1,201 for PP. Among these, we detected 629 (SBP), 532 (DBP), and 540 (PP) previously unreported signals, which corresponded to a total of 1,521 unique novel loci across BP traits. Ancestry stratified analyses revealed 43 unique loci in Admixed American, 135 in African, 343 in Asian, and 3,737 in EUR populations across all three BP traits. A single loci was common to every ancestry. Sex stratified analyses across all BP traits revealed 274 significant independent loci in females and 403 in males, of which 52 and 181 were unique to females and males, respectively. Chromosome X meta-analyses identified 97 significant loci for SBP, 525 for DBP, and 110 for PP. The summary statistics from the multi-ancestry and stratified meta-analysis were used to estimate genetically predicted gene expression (GPGE) with S-PrediXcan. For all three BP traits across cardiovascular relevant tissues, we identified a total of 4,886 significant (Bonferroni P < 5.77x10^-7) gene-tissue pairs. Significant GPGE was associated with SBP at 633 unique loci (52 novel), DBP at 745 loci (48 novel), and PP at 630 loci (225 novel).

CONCLUSIONS
This enrichment of detected signals is made possible through increased ancestral heterogeneity and sample size, which more than doubles the largest GWAS to date. This study provides power to discover and replicate variants contributing to population differences in BP traits and builds a foundation for precision medicine opportunities among traditionally understudied populations.