Expanded and Novel Loci for A1c Levels Identified Through a Trans-Ethnic Meta-Analysis Approach in European and African American Ancestry Samples. E. Wheeler1, M.-F. Hivert2,3, C.-T. Liu4 on behalf of MAGIC and AAGILE 1) Human Genetics, Wellcome Trust Sanger Institute, Cambridge, United Kingdom; 2) Department of Medicine, Université de Sherbrooke, Sherbrooke, Quebec, Canada; 3) General Medicine Division, Massachusetts General Hospital, Boston, MA, USA; 4) Department of Biostatistics, Boston University, Boston, MA, USA.

   A1c is now recommended for type 2 diabetes (T2D) diagnosis. The Meta-Analyses of Glucose and Insulin-related traits Consortium (MAGIC) previously undertook a genome-wide association study (GWAS) meta-analysis of A1c levels in up to 45,000 non-diabetic European individuals and revealed 10 loci (P<5x10-8). MAGIC has now performed an updated meta-analysis in up to 132,914 European individuals from 28 GWAS and 20 Metabochip cohorts. In parallel, the African American Glucose and Insulin Genetic Epidemiology (AAGILE) Consortium has performed a meta-analysis of GWAS for A1C in up to 7099 non-diabetic individuals from 8 cohorts. Variants were excluded if minor allele frequency (MAF)<0.01, HWE<10-4, call rate<0.95, or failed imputation. In AAGILE, a missense variation in G6PD (Val==>Met) previously associated with glucose-6-phosphatase dehydrogenase deficiency was strongly associated with lower A1c levels (P=6.8x10-122) in line with higher red cell turnover. This variant is monomorphic in populations of European descent but common in African descent populations (YRI hapmap: MAF=0.22). Combining the European and African descent samples using the Meta-ANalysis of Transethnic Association Studies (MANTRA) software, confirmed all 10 previously identified MAGIC loci and revealed an additional 37 loci including 6 loci known to be associated with T2D and/or glucose levels (eg. CDKAL1, log10BF = 9.98 and SLC2A2, log10BF = 10.89) and 7 loci previously found to be associated with red blood cell traits or iron metabolism (eg. BET1L, log10BF = 8.81). Among the loci not previously associated with A1c levels, many are located near genes of high biologic relevance, including a variant in the intronic region of ATAD2B (log10BF = 13.17). Follow-up of these newly identified loci will include fine-mapping, leveraging differences in local linkage disequilibrium structure between the ethnic groups. The newly identified loci improve our understanding of the genetic determinants of A1c, and our findings might influence the use of A1c for T2D diagnosis in specific ethnic backgrounds.

You may contact the first author (during and after the meeting) at