Trends in type 1 diabetes in Colorado youth: Role of growth and HLA genotypes

The incidence of type 1 diabetes (T1D) is increasing at a rate of 3% per year worldwide. Many risk factors, such as diet, obesity, infectious diseases, infant diet and feeding patterns and genetic predisposition have been implicated, although not consistently, in contributing to the increased risk of T1D. The primary goal of this study was to assess the trends in incidence of T1D, body size, age at diagnosis and HLA susceptibility genes in Colorado youth under 18 years of age. This study identified youth with T1D through two different diabetes registries: The Colorado IDDM Registry and the Search for Diabetes in Youth Study, Colorado site. Trends in incidence of T1D were assessed for Hispanic and non-Hispanic white (NHW) Colorado youth 0-17 years of age. Trends in age at diagnosis were assessed in youth aged 2-17 years and trends in HLA susceptibility genes were assessed in youth aged 5-17 years. A total of 1281 youth with newly diagnosed T1D aged 0-17 years were ascertained in 1978-1988 and 746 youth were ascertained in 2002-2004 in Colorado. Completeness of ascertainment was estimated 96-97% across the entire time period. Using Poisson regression, trends in incidence of T1D; generalized linear models were used to assess trends in age at diagnosis and body size and logistic regression was used to assess trends in HLA susceptibility genes. The incidence of T1D was 14.8/100,000/year (95% CI 14.0-15.6) in 1978-1988 and 23.9/100,000/year (22.2-25.6) in 2002-2004 for the state of Colorado (p<0.0001). From 1978 to 2004, the incidence of T1D increased by 2.3% (1.6-3.1) per year (p<0.0001). Both age at diagnosis with T1D (p=0.0003) and presence of diabetic ketoacidosis (DKA) at diagnosis (p<0.0001) significantly decreased over time; while mean BMI SDS, weight SDS and height SDS increased (p<0.0001). Adjustment for changes in onset DKA did not explain the BMI trends. Increasing height accounted for 15% of changes in age at diagnosis. The frequency of the high-risk HLA genotype (DRB1*03-DQB1*02/DRB1*04-DQB 1*03) was significantly higher in children diagnosed in 1978-88 (39%) compared to those diagnosed in 2002-04 (28%, p=0.05). In summary, the incidence of T1D has increased 1.6 fold among Colorado youth from 1978-1988 to 2002-2004, and both NHW and Hispanic youth are affected by this trend. This study provided evidence that increased linear growth over time accounted at least in part for the trend toward a younger age of diagnosis of T1D in Colorado children. There was evidence that the high-risk HLA genetic susceptibility to develop T1D has decreased over time, thus resulting in a reduced need for high-risk genetic susceptibility to develop T1D over time.