[Prevalence and predictive value of GAD65 autoantibodies and their correlation with HLA DR-DQ genotypes in children with type-1 diabetes]

Abstract

Introduction: Development of type 1 diabetes is caused by immune mediated destruction of pancreatic beta-cells and is associated with development of islet cell specific antibodies that are molecular markers of the diabetogenic process. Studies on islet cell antibodies will lead to a better understanding of the pathomechanism and improved prediction of the disease. AIMS AND PATIENTS: In present study prevalence of glutamate-decarboxylase (GAD65) antibodies was analysed in a registry based collection of childhood type 1 diabetes cases (n = 122), first degree relatives (n = 164) and ethnically and geographically matched healthy children (n = 2664) in Hungary. Association of GADA with various HLA DQA1-DQB1 genotypes was also studied.

Methods: GADA was determined using a routine radioligand assay and HLA DQA1-DQB1 genotypes were identified with allele-specific polymerase chain reaction method.

Results: GADA was more prevalent in children with type 1 diabetes as compared to first degree relatives or the healthy population (71.3%, 95% CI:63.3-79.3 versus 9.1%, 95% CI: 4.7-13.5 and 1.3%, 95% CI: 0.9-1.7; p < 10(-4) and p < 10(-4), respectively). Girls with diabetes were more often positive for GADA than boys (84.1%, 57.6%, p = 0.003) which tendency was seen also among healthy children (1.8%, 0.8%, p = 0.03). Among parents of diabetic children fathers had higher prevalence of GADA than mothers (16.1%, 3.9%, p = 0.03). Diagnostic sensitivity, specificity and positive predictive value of GADA were as follows: 71.3%, 98.7% and 3.7%, respectively. GADA titer was higher in diabetes patients than in first degree relatives or healthy children (62.4 rU +/- 45.8 rU, 23.6 +/- 14.1 rU; p < 10(-4), 17.2 +/- 9.6; p < 10(-4)). GADA was more prevalent among patients with HLA DR3-DQ2 haplotype while it was less prevalent in cases with HLA DR4-DQ8 haplotype as compared to cases not carrying these haplotypes (p < 10(-4) and p = 0.01, respectively).

Conclusions: GAD65 antibodies are specific disease markers for type 1 diabetes. The increased prevalence of GADA among first degree relatives of type 1 diabetes patients indicates an increased risk for development of diabetes in this population. GADA is associated with HLA DR3-DQ2 haplotype and female sex. Positive predictive value of GADA is considerably lower in the general population than in first degree relatives, consequently, for more accurate diabetes prediction the use of additional immune and genetic markers is necessary.