Islet Autoantibody Patterns in Patients With Type 2 Diabetes Aged 60 and Higher: A Cross-Sectional Study in a Chinese Hospital

Front Endocrinol (Lausanne). 2018 May 25;9:260. doi: 10.3389/fendo.2018.00260. eCollection 2018.


Background: Some elderly citizens with a clinical diagnosis of type 2 diabetes had evidence of positive islet autoantibodies. We aimed to discover their islet autoantibody patterns and independent correlative factors that might lead to a better understanding of significance of islet autoimmunity in the progression of elderly diabetes.

Methods: A total of 541 inpatients of clinically diagnosed type 2 diabetes aged 60 and over were recruited. Three islet autoantibodies including insulin autoantibody (IAA), islet cell antibody (ICA), and glutamic acid decarboxylase antibody (GADA) as well as clinical and biochemical characteristics were tested and collected in Huashan Hospital. Associations between these antibodies and clinical features were analyzed by Spearman correlation and binary logistic analyses.

Results: In our current study, total positive rate of islet autoantibodies (IAA, ICA, and GADA) was 35.67% with 26.62% for individual IAA, 5.55% for ICA, and 5.91% for GADA, in elderly with type 2 diabetes. None of combinations of such autoantibodies were observed, with the exception of IAA + ICA (0.74%, n = 4), IAA + GADA (1.48%, n = 8), and ICA + GADA (0.18%, n = 1). Compared with GADA negative group, patients in positive group tended to have lower level of fasting and postprandial C peptide, fasting blood glucose (FBG), and body mass index (BMI). After adjusted for the BMI, FBG, and postprandial C peptide, fasting C peptide seemed to be an independent factor related to GADA positivity (OR = 0.52, p = 0.02). As for patients with positive IAA, they were more likely to have insulin treatment with longer duration of diabetes, higher level of BMI, and lower level of postprandial C peptide. After adjusted for the duration of diabetes, BMI, and postprandial C peptide, insulin treatment was a significant predictor for IAA positivity (OR = 5.20, p < 0.0001). Furthermore, hs-CRP was positively related to ICA positivity, and hs-CRP appeared to be an independent indicator for ICA (OR = 3.43, p = 0.008).

Conclusion: In elderly with type 2 diabetes, high prevalence rate of IAA was frequently accompanied with insulin treatment, while ICA and GADA were more closely associated with the systemic inflammation and beta-cell failure, respectively.

Keywords: elderly; glutamic acid decarboxylase autoantibodies; insulin autoantibodies; islet autoantibody; islet cell cytoplasmic autoantibodies; type 2 diabetes.