Some combinations of HLA-A,-B and -DR antigens occur more frequently t
han would be expected from their gene frequencies in the population. T
his phenomenon, referred to as Linkage Disequilibrium (LD) has been th
e origin of many speculations. One hypothesis to explain LD is that so
me haplotypes are protected from recombination. A second hypothesis is
that these HLA antigens preferentially recombine after cross-over, to
create an LD haplotype. We tested these 2 hypotheses: from a pool of
over 10,000 families typed in our department, we analyzed 126 families
in which HLA-A:B or B:DR recombinant offspring was documented. To ove
rcome a possible bias in our material, we used the nonrecombined haplo
types from the same 126 families as a control group. Our results show
that the number of cross-overs through LD haplotypes is not significan
tly lower then would be expected if recombination occurred randomly. A
lso the number of LD haplotypes created upon recombination was not sig
nificantly increased.