The search for type 2 diabetes susceptibility genes using whole-genome scans: an epidemiologist's perspective

Most whole-genome scan studies of type 2 diabetes have ascertained families on affected sibpairs. Although assuring enrollment of multiplex families, this strategy may lead to the enrollment of unrepresentative families, particularly in low-risk populations where relatively few diabetics have an affected sibling. In high-risk populations such as Hispanics and Native Americans, population-based ascertainment is often possible, since the high frequency of the disease favors enrollment of multiplex families even when they are not specifically sought. The results of the first generation of genome scan studies caused concern because of a perceived lack of reproducibility. More recently, however, a number of replications have emerged, specifically on chromosomes 1q, 2q, 3q, 9p, 10q, and 11q. The MODY 1 and 3 regions on chromosome 12q and 20q have also been replicated in multiple studies. The clinical features of diabetes in these latter families, however, suggest the common form of type 2 diabetes, rather than maturity-onset diabetes of the young (MODY). Also, diabetes in these families does not appear to be caused by the classical MODY mutations. Interestingly, functional variants of the MODY 4 gene (insulin promoter factor-1) have been associated with both MODY and the common form of type 2 diabetes. Currently available linkage studies implicate relatively broad chromosomal regions, and the challenge of narrowing these regions to facilitate gene discovery remains formidable. The linkages that have been replicated thus far provide good starting points to search for functional variants in type 2 diabetes susceptibility genes.