Transient impaired glucose tolerance in Pima Indians: is it important?

As part of a continuing epidemiological study of non-insulin dependent diabetes among Pima Indians 154 subjects who had had a transient impairment of glucose tolerance were followed up for 1.2-16.9 (median 5.8) years after their glucose tolerance had returned to normal. Of these, 49 subsequently developed diabetes; 26 subsequently developed impaired glucose tolerance; and 79 had normal glucose tolerance at the last examination. The cumulative incidence of diabetes was 16% and 48% at five and 10 years of follow up respectively, compared with 3% and 8% for a control group of 1245 members of the same population. After adjustment for age, sex, body mass index, and plasma glucose concentration two hours after glucose loading the incidence of diabetes among the subjects who had had transient impaired glucose tolerance was 3.0 times that among the controls (95% confidence interval 2.1 to 4.3). Proportional hazards function analysis indicated that obesity was the most important predictor of subsequent development of diabetes. The results suggest that transient impairment of glucose tolerance indicates, at least in some subjects, a predisposition to diabetes and should not be considered clinically unimportant.