Prediction and diagnosis of type 1 diabetes using beta-cell autoantibodies

The clinical manifestation of type 1 diabetes is the endpoint of a long-lasting immune-mediated destruction process of the B-cells. Autoantibodies originating from this process can be applied in the diagnosis and clinical discrimination of autoimmune diabetes as well as in the prediction of this disease. At clinical diagnosis between 80-90% of patients with type 1 diabetes are positive for antibodies to B-cell antigens, such as ICA and antibodies to glutamic acid decarboxylase or IA2. These antibodies can also be detected in the presymptomatic period before onset of the disease, and can thus be used to predict type 1 diabetes. Using a combination of antibodies, diabetes can be predicted in 70-80% of future cases of diabetes, with a positive predictive value between 30-80%, depending on the type of antibody tested for and the population studied. Between 5 and 30% of patients initially diagnosed with type 2 diabetes will show progression to insulin dependency and turn out to have type 1 within three years of diagnosis. It is clinically relevant to identify these patients early in the course of disease, as deterioration of metabolic control results in an increased risk for macro- and micro-vascular complications. Autoantibodies to glutamic acid decarboxylase or ICA are of high diagnostic sensitivity in these cases and are better predictors for future insulin dependency than biochemical or clinical parameters. Increasing knowledge on the applicability of antibodies for diabetes prediction and diagnosis and the development of commercial assays for antibodies to glutamic acid decarboxylase and IA2 antibodies has enabled the implementation of B-cell autoantibodies in routine diagnostic settings.