CVD outcomes. Genotyping was performed using Illumina Infinium Multi-Ethnic AMR/AFR BeadChip (MEGA) arrays on 7546 Black adults who were hypertensive and randomized to either chlorthalidone or lisinopril. Imputation was performed using version r2 of the NHLBI TOPMed reference panel. Participants were excluded if they failed genotyping, had sex mismatch or genotyping call rate <0.95, or if they were an outlier in the PCA (outside of 6 standard deviations). Since ALLHAT used a fasting glucose ≥ 140 mg/dL for the definition of T2D, we excluded controls that had baseline fasting glucose ≥126 mg/dL (N=201 excluded) or missing a fasting glucose measure altogether (N=1209 excluded). This resulted in 5498 individuals eligible for this study (Table 1). Imputed variants were inspected for their imputation quality scores (R2) and it was noted that more than 99% of the variants with MAF >1% had an imputation quality >0.6. High-quality genotype calls with genotypic probability >0.9 were retained.
