K.L. Mohlke(1), A.U. Jackson(2), L.J. Scott(2), K.N. Conneely(2), A. Swift(3), K.F. Doheny(4), E.W. Pugh(4), R.M. Watanabe(5), G.R. Abecasis(2), T.T. Valle(6), J. Tuomilehto(6), R.N. Bergman(5), F.S. Collins(3), M. Boehnke(2)

(1) U North Carolina, Chapel Hill, NC; (2) U Michigan, Ann Arbor, MI; (3) NHGRI, Bethesda, MD; (4) CIDR, JHU, Baltimore, MD; (5) U Southern California, Los Angeles, CA; (6) National Public Health Institute, Helsinki, Finland.

Obesity is a major cause of morbidity and mortality in the USA and worldwide. Although obesity shows strong heritability, the underlying genes are not well understood. Waist to hip ratio (WHR) is a measure of fat distribution that strongly predicts cardiovascular disease. In the context of a case-control association study for type 2 diabetes, we performed preliminary association analysis with WHR using the first 869 of a planned 1186 normal glucose tolerant individuals in stage 1 of our 2-stage study. All individuals are participants in the Finland United States Investigation of NIDDM Genetics (FUSION) and/or Finrisk 2002 studies. These 869 individuals have an average WHR of .900 ± .088 (range .67-1.12) and an average BMI of 27.1 ± 3.7 (range 17.5-39.1). Prior to association analysis the trait values were adjusted for age, gender, and birthplace within Finland. 304,968 SNPs genotyped on the Illumina HumanHap300 BeadChip passed initial quality control using criteria of >95%; success, HWE p>.001 and <2 duplicate discrepancies. For each SNP analyzed we calculated p-SNP, an overall p-value for the SNP corrected for 3 genetic models tested. The lowest p-SNP of 2.4x10-6 was observed for rs966935, a SNP on chromosome 10 located 7.5 kb from a putative processed transcript; the median WHR values for 77 GG, 384 AG, and 407 AA individuals were .918, .906 and .890, respectively. Slightly more p-SNP values exceed the significance threshold of .005 than expected by chance (p=.01). These results suggest that the genetic variants responsible for susceptibility to obesity have modest effects not easily distinguished from chance results in 869 samples. Analysis of all stage 1 samples for WHR, waist circumference, and BMI will further clarify the role of these SNPs in obesity susceptibility.