Inheritance of susceptibility to juvenile onset diabetes.

Regardless of the well-documented population associations between juvenile-onset diabetes (JOD) and certain HLA types, whatever haplotypes are segregating in JOD families may be followed to provide information on mode of inheritance of the disorder. It is essential to group together for analysis families with the same number of affected sibs. We assume a single locus determining susceptibility, closely linked to the HLA region, and ignore recombination, expected to be rare. Our first approach also assumes that the frequency of the susceptible genotype is so small that affected individuals may be considered to arise from only one mating type; the particular mating type depends on the mode of inheritance of susceptibility. Our sample is the result of pooling our own data with published studies of HLA haplotype segregation in families with two or more JOD offspring. Given the assumptions, we find that the data are more plausibly explained by a one-dose than by a two-dose or "recessive" hypothesis. We then develop the analysis further by adding a crude but explicit estimate of the frequency of the susceptible genotype, based on disease prevalence and penetrance of the genotype. The one-dose hypothesis is strongly supported by this analysis as well. We also consider some problems of ascertainment arising from heterogeneity of the disorder and selection against diabetics. Studies involving unaffected relatives of diabetics are suggested which might test further the conclusions drawn here.