Different rates of conversion to IDDM in siblings of type 1 diabetic children: the Montreal family study.

ABSTRACT

This study was undertaken to analyze the risk of developing insulin-dependent diabetes (IDDM) in siblings of type 1 diabetic children. Islet-cell antibodies (ICA) were tested in 568 subjects, siblings of type 1 diabetic children. The subjects were followed prospectively for the conversion to clinical diagnosis of IDDM. As a result siblings who were islet cell antibody (ICA)+ positive at the time of diagnosis of the diabetic sibling (index case) had a significantly higher risk of developing IDDM than those who were ICA-. However, of the 19 siblings who developed IDDM, only 10 were ICA+ at the time of the first test but, 17 became ICA+ before diagnosis of IDDM. The interval between a positive test and the clinical diagnosis of IDDM varied between subjects (6-44 months, mean = 18.4 +/- 4.2 S.E. months) but it was less than 1 year in one subject. In addition to the higher risk of developing IDDM when ICA was positive, male sex and younger age of the subjects as well as young age of the index case and multiplex pedigrees were significant predictors of conversion to IDDM. The Cox's regression tree constructed using RECPAM identified three groups of varying rates of conversion to IDDM (1) a group with the slowest progression characterized by ICA- and age of index case > 5 years or female sex (relative hazard = 1); (2) an intermediate progression group consisting of subjects who are ICA- and have both < 5 years of age and male sex (relative hazard = 8.78); (3) a group with the fastest progression consisting of subjects who are ICA+ (relative hazard = 31.45). From these results our data suggest that in addition to ICA, clinical markers such as age, sex and multiple pedigrees are also significant predictors of the rate of conversion to IDDM. Furthermore, screening for ICA in family intervention studies will have to be done frequently, perhaps yearly, and will have to be continued into adult life, particularly in ICA- subjects in order to identify the 85-90% of subjects who become ICA+ at the clinical diagnosis of IDDM.