Beta cell function in autoimmune diabetes in young adults

This thesis presents the results from a study of the first four years of diabetes in 879 patients aged 15-34 years at diagnosis. The aims of this study were to compare clinical classification with the presence of autoantibodies, to study the natural course of beta cell function during the first years of disease and to define predictive factors for the remaining beta cell function after diagnosis. Of all patients, 76% were classified as type 1, 14% as type 2 and 9.3% were not possible to classify. Autoantibodies such as islet cell antibodies (ICA), glutamic acid decarboxylase antibodies (GADA), antibodies against tyrosine phosphatase (IA-2A) and insulin autoantibodies (IAA) were analysed. Frequencies of autoantibodies in type 1 patients with (n=78) and without ketoacidosis (n=517) did not differ: ICA (63% vs 57%), GADA (63% vs 66%), IA-2A (35% vs 44%) and IAA (20% vs 15%). At least one of ICA, GADA or IA-2A were found in 47% of type 2 and in 59% of the unclassifiable patients. After three years, most (93%) of these autoantibody positive patients were on insulin and GADA was the strongest prognostic factor. Non-fasting C-peptide at diagnosis differentiated better than fasting C-peptide between autoimmune and non-autoimmune diabetes. Positivity for GADA gave lower C-peptide during the first three years after diagnosis compared with positivity for ICA and/or IA-2A, but after four years there was no significant difference. Low initial C-peptide and high GADA were significant risk factors for a low C-peptide after two years. We concluded that determination of autoantibodies improved the clinical classification and especially GADA marked a more severe form of beta cell damage and need for insulin.