Chronic autoimmunity of type I diabetes

A considerable body of data supports the hypothesis that type I diabetes is a chronic progressive autoimmune disorder. Individuals with very high probability of progressing to diabetes can now be readily identified. Assays for autoantibodies reacting with insulin (IAA), glutamic acid decarboxylase (GAD65AA), and the neuroendocrine tyrosine phosphatase ICA512/IA-2 (ICA512AA) allow for the identification of more than 95% of individuals developing type I diabetes. The expression of a single autoantibody does not indicate high risk for diabetes and in general, prediabetic individuals express a series of biochemically defined autoantibodies. Levels of such autoantibodies are usually stable over years of follow-up. Unusual variants of autoantibody expression (e.g. GAD-ICA with high titers of GAD65 autoantibodies as the sole autoantibody) have low prognostic significance. Given the presence of multiple autoantibodies, low first phase insulin secretion (following intravenous glucose) is the best predictor of time to diabetes onset. Measurement of autoantibodies can now be automated and applied to large populations such that screening and prediction in the general population is now feasible. We favor the hypothesis that insulin may be the primary autoantigen for type I diabetes, and therapies which after the immune response to insulin may lead to safe and effective preventive modalities.