BACKGROUND: Circulating antibodies to glutamic acid decarboxylase (GADab) and tyrosine phosphatase-like molecule IA-2 (IA-2ab) are major indicators for auto-immune destruction of pancreatic islet cells. They identify a majority of Caucasians with type 1 diabetes and approximately 50% of Asians, providing evidence of an idiopathic aetiology in the latter. The present study investigated these autoantibodies in a mixed ethnic group.
METHODS: Hospital clinic patients with clinically defined type 1 (n = 93) and type 2 (n = 300) diabetes and representing Singapore's major ethnic groups--Chinese, Indians and Malays--were studied. GADab and IA-2ab frequencies, and association of autoimmunity status with clinical and biochemical profiles were analysed.
RESULTS: Radio-immunoprecipitation assays detected either or both antibodies (seropositivity) in 41.9% of subjects with type 1 diabetes. GADab was detected in 36.6% and IA-2ab in 23.7% of type 1 diabetics. Prevalence of IA-2ab showed a reduction in frequency with disease duration (P = 0.026). In clinical type 2 diabetics, seropositivity was 10.0% with higher frequency in Malays (17.5%) than Chinese (9.7%) and Indians (4.5%). Multivariate analysis revealed that low fasting C-peptide was associated with seropositivity (odds ratio (OR) = 0.15; 95% confidence interval (CI) = 0.04-0.58). A significant relationship (OR = 13.5; 95% CI = 5.0-36.7) between insulin requirement and duration (>5 years) was also revealed. In patients with type 2 diabetes there was a trend of gradual progression to insulin dependency. However, there was considerable variation in body mass index between ethnic subgroups of type 2 diabetics, particularly for Chinese (mean (SD) = 26.0 (4.7)) and Malays (mean (SD) = 29.2 (5.9); P < 0.001).
CONCLUSIONS: Presence of both antibodies in our mixed ethnic group of type 1 diabetes patients was much lower than in Caucasians. Significant numbers of patients were seronegative for antibodies. Influences due to ethnicity and adiposity would require further investigations.
* Title and MeSH Headings from MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine.