Heritable factors appear to account for much of the risk for Hodgkin diseas
e (HD). There is evidence for an HLA-linked gene, but other predisposing lo
ci remain unaccounted for. The observation of a family coinheriting both HD
and Leri-Weill dyschondrosteosis (LWD) suggests that a gene conferring ris
k for HD resides adjacent to the LWD locus. The gene responsible for LWD, S
HOX, localizes to the short-arm pseudoautosomal region (PAR) of the X and Y
chromosomes. A unique segregation pattern for PAR-linked genes has been pr
edicted-that affected sibs will tend to be same sex. An excess of sex-conco
rdant affected sib pairs with HD has been noted but has been attributed to
an environmental etiology. These two observations-sex concordance in sib pa
irs with HD and cosegregation of HD and LWD-impelled a test of the hypothes
is that there is a PAR-localized gene for HD. By first scoring recombinatio
ns dissociating sex from phenotype in individuals from pedigrees with LWD,
we determined a male maximum recombination frequency (theta(max)) of .405.
This places SHOX near the short-arm telomeres of the sex chromosome and sup
ports the prediction that PAR recombination is obligatory for spermatogenes
is. By inferring recombinations between HD and sexual phenotype in sib pair
s, we predict, for the postulated HD gene, a male theta(max) as high as .25
4, which places it in proximity to SHOX. Morton's nonparametric affected-si
b-pair "beta" model was used in the evaluation of linkage between HD and ph
enotypic sex and gave a LOD score of 2.41. Using this approach, we reevalua
ted evidence for HLA linkage in HD in haplotyped sib pairs and found a LOD
score of 2.00. The resulting beta values indicate that the putative PAR- an
d HLA-linked loci account for 29% and 40%, respectively, of the heritabilit
y of HD in an American population.