Genetic Markers Explain Differences in HbA1c Levels between African-American (AA) and Non-Hispanic White (NHW) Participants of the Diabetes Prevention Program (DPP)

Abstract

HbA1c levels are higher in AA than NHW independent of glycemia. We investigated whether genetics could explain observed differences in HbA1c levels between AA and NHW DPP participants. We tested 1) genetic variants known to cause hemoglobinopathies; 2) a genetic risk score (GRS) based on 60 variants associated with HbA1c in the most recent genome-wide association meta-analysis; and 3) Principal Components (PC) factors that are generated by a statistical procedure that uses a linear transformation to convert a set of observations of correlated genetic markers measured across the genome (HumanCore Exome arrays + 1000 Genomes imputation) into a set of values mirroring genetic variation due to continental ancestry to capture substructure admixture. Of 2,658 eligible DPP participants, 537 (20%) self-identified as AA and 1,476 (56%) as NHW. HbA1c was measured with ion-exchange high-performance liquid chromatography (Biorad Variant). Despite comparable fasting and 2h-glucose levels, AA had higher HbA1c (mean±SD= 6.2±0.6%) compared with NHW (5.8±0.4%; P<0.001). In AA, the genetic variant causing sickle cell trait was associated with higher HbA1c (β (SE)= +0.44 (0.08)%-unit per risk allele; P=2.1x10-4). The GRS was associated with HbA1c in both AA (β= +0.017 (0.005); P=0.002) and NHW (β= +0.016 (0.002); P=5.4x10-11). Self-identified AA were distributed along the first PC axis, as expected in mixed ancestry populations. The first PC explained 60% of the 0.4 unit difference in HbA1c between AA and NHW, while the sickle cell variant explained 16%, and GRS explained 14%. In conclusion, genetic variants causing hemoglobinopathies or currently known to influence HbA1c explained a small proportion of HbA1c difference between AA and NHW. A larger proportion of the difference was associated with the first PC, suggesting that other yet unidentified genetic markers influence HbA1c in AA, in addition to non-genetic factors. Disclosure M. Hivert: None. C.A. Christophi: None. K.A. Jablonski: None. S. Edelstein: None. S.E. Kahn: Advisory Panel; Self; Boehringer Ingelheim GmbH, Elcelyx Therapeutics, Inc., Eli Lilly and Company, Intarcia Therapeutics, Inc., Janssen Research & Development, Merck & Co., Inc., Novo Nordisk A/S. S. Golden: None. S. Dagogo-Jack: None. K.J. Mather: Research Support; Self; Merck & Co., Inc., Sanofi-Aventis, Novo Nordisk A/S, Abbott. Advisory Panel; Self; Merck & Co., Inc.. J.A. Luchsinger: None. A. Caballero: None. E. Barrett-Connor: None. W.C. Knowler: None. J.C. Florez: Consultant; Self; Intarcia Therapeutics, Inc.. Consultant; Spouse/Partner; Santen. W.H. Herman: Other Relationship; Self; Merck Sharp & Dohme Corp., Lexicon Pharmaceuticals, Inc.. Consultant; Self; Janssen Scientific Affairs, LLC.. Research Support; Spouse/Partner; Nestlé. Other Relationship; Self; American Diabetes Association. Advisory Panel; Self; National Committee for Quality Assurance (NCQA). D. Research Group: None.