Islet cell antibodies and glutamic acid decarboxylase antibodies, but not the clinical phenotype, help to identify type 1(1/2) diabetes in patients presenting with type 2 diabetes.

This study was undertaken to determine which type 1 diabetes-associated autoantibodies and what clinical characteristics are most useful to identify patients with type 1(1/2) diabetes. We studied 125 patients, recently diagnosed clinically with type 2 diabetes for the presence of islet cell antibodies (ICA), insulin autoantibodies (IAA), antibodies to glutamic acid decarboxylase(GADAb), and IA-2a (IA-2Ab). Patients with a diagnosis of type 2 diabetes who met all of the following criteria at diagnosis were studied: age > or = 30 years, no history of ketonuria or ketoacidosis, and not requiring insulin treatment. Thirty-six patients (29%) were positive for at least 1 antibody. Thirty-two (26%) were ICA positive and 20 (16%) GADAb positive. Insulin autoantibodies and IA-2Ab occurred less frequently in 2 (1.6%) and 8 (6.4%) patients, respectively. There was no significant difference in the ages at diagnosis between the Ab(+) and Ab(-) patients, age in years (range) 47.2 (32 to 64) versus 51.2 (31 to 77), respectively, P =.06. Body mass index (BMI) was different in the 2 groups, with Ab(+) patients being less obese, BMI (range) 28.3 kg/m(2) (17.6 to 54.9) versus 32.0 kg/m(2) (19.2 to 68.8), respectively, P =.01. Clinical presentation of diabetes was more commonly symptomatic with polyuria and polydipsia in Ab(+) patients, while in Ab(-) patients, diagnosis was more often incidental, P =.002. However, more than 95% of patients overlapped in both age and BMI irrespective of antibody status. Similarly, 42% of Ab(+) patients had their diabetes diagnosed incidentally, while 29% of Ab(-) patients presented with polyuria and polydipsia. We therefore conclude that screening with antibodies, mainly ICA and GAD, but not age, BMI, or clinical presentation should be used to identify type 1(1/2) diabetes. Copyright 2001 by W.B. Saunders Company