Genome-Wide Association with Diabetes-Related Traits in the Framingham Heart Study
Meigs, James B.
Manning, Alisa K.
Fox, Caroline S.
Florez, Jose C.
Cupples, L. Adrienne
MetadataShow full item record
Citation (published version)Meigs, James B, Alisa K Manning, Caroline S Fox, Jose C Florez, Chunyu Liu, L Adrienne Cupples, Josée Dupuis. "Genome-wide association with diabetes-related traits in the Framingham Heart Study" BMC Medical Genetics 8 (Suppl 1): S16. (2007)
BACKGROUND: Susceptibility to type 2 diabetes may be conferred by genetic variants having modest effects on risk. Genome-wide fixed marker arrays offer a novel approach to detect these variants. METHODS: We used the Affymetrix 100K SNP array in 1,087 Framingham Offspring Study family members to examine genetic associations with three diabetes-related quantitative glucose traits (fasting plasma glucose (FPG), hemoglobin A1c, 28-yr time-averaged FPG (tFPG)), three insulin traits (fasting insulin, HOMA-insulin resistance, and 0–120 min insulin sensitivity index); and with risk for diabetes. We used additive generalized estimating equations (GEE) and family-based association test (FBAT) models to test associations of SNP genotypes with sex-age-age2-adjusted residual trait values, and Cox survival models to test incident diabetes. RESULTS: We found 415 SNPs associated (at p < 0.001) with at least one of the six quantitative traits in GEE, 242 in FBAT (18 overlapped with GEE for 639 non-overlapping SNPs), and 128 associated with incident diabetes (31 overlapped with the 639) giving 736 non-overlapping SNPs. Of these 736 SNPs, 439 were within 60 kb of a known gene. Additionally, 53 SNPs (of which 42 had r2 < 0.80 with each other) had p < 0.01 for incident diabetes AND (all 3 glucose traits OR all 3 insulin traits, OR 2 glucose traits and 2 insulin traits); of these, 36 overlapped with the 736 other SNPs. Of 100K SNPs, one (rs7100927) was in moderate LD (r2 = 0.50) with TCF7L2 (rs7903146), and was associated with risk of diabetes (Cox p-value 0.007, additive hazard ratio for diabetes = 1.56) and with tFPG (GEE p-value 0.03). There were no common (MAF > 1%) 100K SNPs in LD (r2 > 0.05) with ABCC8 A1369S (rs757110), KCNJ11 E23K (rs5219), or SNPs in CAPN10 or HNFa. PPARG P12A (rs1801282) was not significantly associated with diabetes or related traits. CONCLUSION: Framingham 100K SNP data is a resource for association tests of known and novel genes with diabetes and related traits posted at. Framingham 100K data replicate the TCF7L2 association with diabetes.
RightsCopyright 2007 Meigs et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution 2.0 License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.