Abstract 18777: The Low-Density Lipoprotein Cholesterol and Body Mass Index/Type-2 Diabetes Signals in the HMGCR Region Are Not Explained by a Single Variant
Introduction: Statin use has been linked to risk of type-2 diabetes (T2D). Swerdlow et al. explored this relationship in a genetic analysis published in the Lancet, 2014, using two correlated variants in HMGCR, rs17238484 and rs12916, as proxies for HMGCR inhibition by statins, using rs17238484 for the main analysis. The same alleles associated with lower low-density lipoprotein cholesterol (LDL-C), increased body mass index (BMI) and increased risk of T2D. They concluded that the T2D risk linked to statins is at least partially explained by HMGCR inhibition. Our objective was to explore the methodology of this analysis and confirm the results.
Methods and Results: We assessed a 2Mb region containing HMGCR in 268,952 Icelanders in relation to LDL-C, BMI and T2D, imputing all variants (N=11,269) derived from sequencing the whole genomes of 2,636 Icelanders. We obtained information from the relevant consortia summary statistics (GLGC for LDL-C, GIANT for BMI and DIAGRAM for T2D). See results in Table. Rs12916 is the most significant LDL-C variant in the region. After adjusting for rs12916 in Iceland, there is no additional association to LDL-C. About 350 kb from the rs12916 LDL variant is a published BMI variant, rs2112347, that correlates with rs12916 (r2=0.21) and associates with T2D in DIAGRAM with an effect as expected based on the BMI association (OR=1.04, P=0.0037). The rs2112347-BMI variant associates with LDL-C but this association is explained by rs12916. Rs17238484 correlates with rs12916 (r2=0.37) and rs2112347 (r2=0.49). After accounting for these two variants, there is no association between rs17238484 and either LDL-C or BMI/T2D.
Conclusions: There are two correlated but distinct signals in the region containing HMGCR, the rs12916 LDL-C and rs2112347 BMI/T2D signals. The LDL and BMI (and T2D) signals cannot be explained by a single variant in the region. No conclusions can be drawn about the relationship between statin therapy and risk of T2D based on these data.
Author Disclosures: H. Holm: Employment; Modest; I am an employee of deCODE genetics/Amgen Inc. P. Sulem: Employment; Modest; Employee of deCODE genetics / Amgen Inc. H. Helgason: Employment; Modest; Employee of deCODE genetics / Amgen Inc. G. Thorleifsson: Employment; Modest; Employee of deCODE genetics / Amgen Inc. U. Thorsteinsdottir: Employment; Modest; Employee of deCODE genetics / Amgen Inc. D. Gudbjartsson: Employment; Modest; Employee of deCODE genetics / Amgen Inc. K. Stefansson: Employment; Modest; Employee of deCODE genetics / Amgen Inc..
- © 2016 by American Heart Association, Inc.