Are trends in diabetes incidence changing for minority children?

PURPOSE: To contrast incidence rates and time trends of typical type 1 diabetes with non-autoimmune insulin-resistant type 2 diabetes in minority children (Early 2).METHODS: A population based registry of all insulin treated African-American and Hispanic children residing in Chicago, aged 0-17 years at onset, diagnosed between 1985-1994 provided data. Medical records were reviewed and abstracted (N = 739), and a subset of patients were interviewed (n = 195). Based on these data sources, cases were assigned to the early type 2 category if one or more of the following variables were present: notation of "atypical" or "possible type 2" diabetes in the medical record; obesity (defined as BMI >/= 27 (kg/m(2))) or acanthosis nigricans noted at diagnosis; a report by the patient of going off insulin six months after initial diagnosis without developing diabetic ketoacidosis; or reporting the use of oral hypoglycemic agents with or without insulin. The Chi-square test for homogeneity was used to determine differences between proportions of groups. Time trend analysis of incidence was done using regression models with the Poisson distribution.RESULTS: The 10-year average annual incidence rate was 10.3 per 100,000 population for typical type 1 diabetes (n = 565) and 3.2 per 100,000 for those with characteristics of early 2 diabetes (n = 174). Over the ten year interval the incidence rate remained static for typical type 1 diabetes. In the early 2 group there was a 9.0% average annual percent increase over the ten year interval (p < 0.01).CONCLUSIONS: The trends in incidence rates vary between typical type and early 2 diabetes. We hypothesize that this is associated with the increasing rate of obesity and a lack of physical activity in young people.