OBJECTIVE

To identify genetic variants in linkage disequilibrium with those conferring diabetes susceptibility, a genome-wide association study for young-onset diabetes was conducted in an American-Indian population.

RESEARCH DESIGN AND METHODS

Data come from 300 case subjects with type 2 diabetes with age of onset <25 years and 334 nondiabetic control subjects aged >or=45 years. To provide for tests of within-family association, 121 nondiabetic siblings of case subjects were included along with 140 diabetic siblings of control subjects (172 sibships). Individuals were genotyped on the Affymetrix 100K array, resulting in 80,044 usable single nucleotide polymorphisms (SNPs). SNPs were analyzed for within-family association and for general association in case and control subjects, and these tests were combined by Fisher's method, with priority given to the within-family test.

RESULTS

There were more SNPs with low P values than expected theoretically under the global null hypothesis of no association, and 128 SNPs had evidence for association at P < 0.001. The association of these SNPs with diabetes was further investigated in 1,207 diabetic and 1,627 nondiabetic individuals from the population study who were not included in the genome-wide study. SNPs from 10 genomic regions showed evidence for replication at P < 0.05. These included SNPs on chromosome 3 near ZNF659, chromosome 11 near FANCF, chromosome 11 near ZBTB15, and chromosome 12 near SENP1.

CONCLUSIONS

These studies suggest several regions where marker alleles are potentially in linkage disequilibrium with variants that confer susceptibility to young-onset type 2 diabetes in American Indians.