ABSTRACT
BACKGROUND: The aims of the study were to characterize the relatives at high risk of progression to diabetes and to determine whether rate of progression to diabetes varied according to age, specific combination of antibodies and genetic markers of susceptibility. METHODS: Family members of type 1 diabetic patients were examined through a large medical network for the presence of specific antibodies to beta cell constituents and high risk DQB1 alleles. Antibodies to insulin, GAD and IA-2 as well as ICA were examined in 4,044 family members recruited in a large prospective family study in the Rhone-Alpes region (the GRADI study). Among them, 3,951 non diabetic first degree relatives have been tested on a median of 2.2 occasions and were followed for up to 16 years. RESULTS: Presence of antibodies to GAD (3.6%), IA-2 (4.9%), insulin (2.2%) and ICA at titers equal or above 20JDF units (1.1%) were noticed at the first determination and prevalence increased among ICA positive relatives versus ICA negative relatives. All combinations of markers resulted in specificities above 90%. The positive predictive value of antibodies was dependent on the number of positive antibodies. Combination of antibodies to GAD and IA-2 or GAD and IAA had higher predictive values and sensitivities than ICA titers above 20 JDF units. Additional positivity of IAA increased the predictive value but reduced the sensitivity of the screening procedure. Using a combi GAD/IA-2 assay increased the sensitivity of the screening up to 87.8% but reduced the predictive value to 13.8%. CONCLUSIONS: These data confirm in a large French cohort of first degree relatives of type 1 diabetic patients that combinations of antibodies to beta cell constituents can replace ICA in the first screening procedure. We report that combi GAD/IA-2 assay is well suited for screening purposes. However, time to diabetes in antibody positive relatives appears to be more heterogeneous to what has been described. The importance of genetic markers needs further evaluation.