Combined positivity for HLA DQ2/DQ8 and IA-2 antibodies defines population at high risk of developing type 1 diabetes

Abstract

Aims/hypothesis: Prevention trials in first-degree relatives of type 1 diabetic patients are hampered by large interindividual differences in progression rate to diabetes. We investigated whether specific combinations of immune and genetic markers can identify subgroups with more homogeneous progression to clinical onset. Methods: Antibodies against islet cell cytoplasm (ICA), insulin (IAA), glutamate decarboxylase (GADA) and IA-2 protein (IA-2A) were measured in 790 non-diabetic control subjects and 4,589 first-degree relatives under age 40. Results: On first sampling, 11.1% of the siblings presented at least one antibody type (p<0.001 vs other relatives). During follow-up (median 52 months) 43 subjects developed type 1 diabetes (31 siblings, ten offspring of a diabetic father, two offspring of a diabetic mother). Using Kaplan-Meier survival analysis and Cox regression, IA-2A conferred the highest 5-year diabetes risk (>50%) irrespective of the number of antibodies present. In initially IA-2A-positive relatives (n=58) progression to hyperglycaemia depended more on HLA DQ status than on type of kinship (84% progression in the presence of DQ2/DQ8 vs 32% in its absence; p<0.003). In IA-2A-negative relatives (n=4,531) 5-year progression to diabetes increased with the number of other antibodies (ICA, GADA and/or IAA) (p<0.001) but overall did not exceed 10% even for two or more antibodies. Among relatives initially positive for one or more antibody type other than IA-2A (n=315), there was significantly more progression to diabetes (overall still <10%) in carriers of DQ2 (p<0.001 vs no DQ2), regardless of DQ8 status. Conclusions/interpretation: These observations suggest that the HLA-DQ-inferred risk of diabetes can proceed through two distinct pathways distinguished by IA-2A status. Combined positivity for DQ2/DQ8 and IA-2A defines a more homogeneous high-risk population for prevention trials than those used so far.

Adiponectin levels do not predict clinical onset of type 1 diabetes in antibody-positive relatives

“…In the DASP 2005 programme, sensitivity and specificity were respectively 38 and 98% for IAA, 82 and 93% for GADA and 76 and 98% for IA-2A. HLA-DQ polymorphisms, proinsulin and C-peptide were determined as previously described [6,7]. Random serum adiponectin levels were measured by the Linco ELISA method (Linco Research, St Charles, MO, USA), in which mouse immunoglobulin is added to avoid interference of heterophile antibodies.…”

“…Stepwise multiple regression analysis was performed to determine the independent contribution of variables on log 10 -transformed adiponectin levels. Diabetes-free survival was studied by Kaplan-Meier analysis and a log-rank test [6]. A Cox proportional hazards model, performed by the forward stepwise method, was used to investigate the independent contributions of the strongest risk factors identified by univariate analysis (p≤0.001) [6].…”

“…A Cox proportional hazards model, performed by the forward stepwise method, was used to investigate the independent contributions of the strongest risk factors identified by univariate analysis (p≤0.001) [6]. Statistical tests were performed two-tailed by SPSS for Windows 13.0 (SPSS, Chicago, IL, USA) and considered significant whenever p<0.05 or p<0.05/k in the case of k comparisons (Bonferroni adjustment) [6]. (Fig.…”

“…Since low levels of adiponectin, an insulin-sensitising adipokine, have been associated with obesity, insulin resistance and risk of type 2 diabetes [4,5], we investigated whether adiponectin can serve as predictive marker for type 1 diabetes in first-degree relatives of known patients. In this group, islet antibodies, especially insulinoma-associated protein-2 antibodies (IA-2A), combined with genetic or beta cell functional markers confer high risk but identify individuals with variable progression rates towards diabetes [6,7]. We assessed whether adiponectin levels allow identification of relatives at risk of rapid progression to type 1 diabetes.…”

“…Classification as type 1 diabetes was on the basis of centrally determined autoantibodies. BMI was expressed as a standard deviation score (BMI-SDS) by comparison with an age-and sex-matched reference cohort [6].…”