In an 11-yr screening program carried out on serum samples sent to an autoimmune serology laboratory, 158 patients with clinical or subclinical autoimmune endocrine manifestations and islet cell antibodies (ICAs) in the absence of overt diabetes were identified and followed for the development of insulin-dependent (type I) diabetes. Twenty-two (13.9%) developed type I diabetes in a follow-up of up to 12 yr (mean ± SE 4.8 ± 3.2 yr). The probability of being free of type I diabetes was 69.8% at 10 yr after the first detection of ICAs. Progression to disease was influenced by 1) the amount of ICAs represented by high titers (63% of those with ICAs ≥20 Juvenile Diabetes Foundation units being free of type I diabetes at 10 yr), ICA persistency (59% being free of type I diabetes; P < 0.02 vs. nonpersistent ICA), and complement-fixing (CF)-ICAs (63% being free of type I diabetes; P < 0.05 vs. non–CF-ICA); 2) the coexistence of insulin autoantibodies (IAAs) (25% being free of type I diabetes; P < 0.005 vs. IAA); and 3) a positive family history (1st-degree relative) for type I diabetes (32% being free of type I diabetes; P < 0.005 vs. no family history). There was a trend for diabetes to develop earlier in males of a younger age. No relationships were found with the number, type, or clinical expression of the associated autoimmunities or with a family history of such disorders. These data confirm the predictive ability of ICAs and lAAs for type I diabetes and indicate that the events that lead to disease in individuals with ICAs are influenced by the level of the humoral response and by a familial association of type I diabetes. This study identified a large cohort of individuals with high ICA levels who have not progressed to type I diabetes. A study of these individuals will be valuable in identifying factors important for the ultimate progression to disease.

This content is only available via PDF.