Is juvenile diabetes determined by a single gene closely linked to HLA?

The transmission behavior of insulin-dependent juvenile diabetes mellitus (JDM) has been studied with respect to its frequency in the relatives of JDM probands and its possible linkage to the HLA complex. Mathematical analysis shows that under a single locus hypothesis a very restricted range of incidence rates is possible in the full siblings of probands once the concordance rate in monozygotic (MZ) twins is specified. Specifically, for a given population prevalence of the disease, high concordance rates in MZ twins require high incidence rates in siblings, and low rates require low incidence rates, if a single locus model is th be valid. Moreover, if these rates do conform to a single locus model, then they give additional information about possible linkage between the purported JDM susceptibility gene and the HLA complex. By using observations on the identity by descent scores at the HLA locus of sibling pairs, both of whom are affected with JDM, it is shown that tight linkage of a disease susceptibility locus is possible only when the MZ twin and sibling incidence rates are low, whereas high rates support loose linkage. If the single locus model is rejected, then an alternative hypothesis, involving epistasis between a JDM susceptibility locus and genes in (or close to) the HLA complex can be suggested as a mechanism whereby JDM would appear to be linked to HLA within families while maintaining an association with HLA at the population level.