High density GWAS for LDL cholesterol in African Americans using electronic medical records reveals a strong protective variant in APOE

Abstract

Only one low-density lipoprotein cholesterol (LDL-C) genome-wide association study (GWAS) has been previously reported in -African Americans. We performed a GWAS of LDL-C in African Americans using data extracted from electronic medical records (EMR) in the eMERGE network. African Americans were genotyped on the Illumina 1M chip. All LDL-C measurements, prescriptions, and diagnoses of concomitant disease were extracted from EMR. We created two analytic datasets; one dataset having median LDL-C calculated after the exclusion of some lab values based on comorbidities and medication (n= 618) and another dataset having median LDL-C calculated without any exclusions (n= 1,249). SNP rs7412 in APOE was strongly associated with LDL-C in both datasets (p < 5 × 10(-8) ). In the dataset with exclusions, a decrease of 20.0 mg/dL per minor allele was observed. The effect size was attenuated (12.3 mg/dL) in the dataset without any lab values excluded. Although other signals in APOE have been detected in previous GWAS, this large and important SNP association has not been well detected in large GWAS because rs7412 was not included on many genotyping arrays. Use of median LDL-C extracted from EMR after exclusions for medications and comorbidities increased the percentage of trait variance explained by genetic variation.

Figure 1

EMR extraction of LDL‐C values and exclusion dates and creation of the analytic datasets. Rx = prescription, Dx = prevalent disease. This flowchart depicts the information extracted from the EMR, and the phenotypes and analytic dataset created from the extracted EMR information. The start of the process is depicted at the top of the figure. Diamonds represent questions asked to extract information or restrict the analytic dataset, while rectangles represent data collected. The top third of the figure depicts how the dataset with no lab values excluded was assembled, while the bottom two thirds show the additional steps required to assemble the more restrictive dataset after exclusions. The figure demonstrates that median LDL‐C was calculated at two separate points for the two dataset; for this reason, an individual present in both datasets may have different LDL‐C values in each dataset.

Figure 2

(A) GWAS of LDL‐C in eMERGE African Americans with exclusions. (B) GWAS of LDL‐C in eMERGE African Americans without any lab values excluded. Manhattan plots display the p values for the association of approximately 1 million SNPs with LDL‐C; each circle represents a single SNP association plotted by chromosome position (x axis) and log(10) transformed p value (y axis). The p values are from regression equations with SNPs modeled additively, adjusted for age2, sex, eMERGE site, genotyping batch, and the first three principal components. In both plots, only the association of LDL‐C and SNP rs7412 in APOE exceeded the genome‐wide threshold for significance.