Clinical utility of HNF1A genotyping for diabetes in aboriginal Canadians.

Abstract

To determine the diagnostic performance characteristics of HNF1A genotyping for diabetes and impaired glucose tolerance (IGT) in Canadian Oji-Cree Indians. We studied all Oji-Cree subjects > or = 50 years of age (96 subjects) who had participated in a community-wide prevalence survey for type 2 diabetes. Subjects were classified either as having "disease," which included type 2 diabetes and IGT, or not. All subjects were genotyped for the HNF1A G319S mutation. The prevalence of disease in this group was 65.7%, of whom 71.4% had type 2 diabetes. For a carrier of HNF1A S319, the specificity, sensitivity, and positive and negative predictive values were 97.0, 30.1, 95.0, and 42.1%, respectively. When the pretest disease prevalence was accounted for, the probability of disease after a positive test was 97.2%, and the probability of disease after a negative test was 42.2%. The values were very similar for the subgroup of subjects with type 2 diabetes alone. The HNF1A genotype appears to be the most specific genetic test yet reported for the prediction of a common multifactorial disease by applying present-day standards of clinical epidemiology in molecular genetics. A positive test result had particular diagnostic value in the Oji-Cree: a subject with HNF1A S319 was virtually certain of having diabetes or IGT by 50 years of age. In contrast, a subject without HNF1A S319 had a reduced risk compared with the age-specific prevalence but was not totally risk-free. Because HNF1A S319 was not the only predisposing factor for diabetes in the Oji-Cree, subjects without HNF1A S319 were still at some risk for diabetes or IGT.