Fea. Mcguigan et al., Susceptibility to osteoporotic fracture is determined by allelic variationat the Sp1 site, rather than other polymorphic sites at the COL1A1 locus, OSTEOPOR IN, 11(4), 2000, pp. 338-343
Previous studies have identified an association between osteoporotic fractu
re and a polymorphism affecting a Spl binding site in the first intron of t
he collagen type I alpha 1 gene (COL1A1). It is currently unclear, however,
whether this association is direct or the result of linkage disequilibrium
with other polymorphisms situated nearby. In this study we analyzed the re
lationship between four well-characterized single-nucleotide polymorphisms
at the COL1A1 locus and osteoporotic fracture in 93 patients with vertebral
fracture and 88 age-matched controls randomly selected from the community.
We studied a MspI polymorphism 26 kb upstream of the COL1A1 gene, the Spl
binding site polymorphism in intron I, a RsaI polymorphism in intron 5 and
a MnlI polymorphism in exon 52. The Sp1 and RsaI polymorphisms were in stro
ng linkage disequilibrium (chi(2)=77.87, p<0.001) and weaker linkage disequ
ilibrium was detected between the Sp1 and MnlI polymorphisms (chi(2)=5.54,
p<0.025). There was a significant association between COL1A1 haplotypes tha
t included the Sp1 and RsaI polymorphisms and fracture (p<0.05-0.001), but
no association with haplotypes that included only the MspI and MnlI polymor
phisms. This association with fracture was strongest when haplotypes were g
rouped by Spl alleles (chi(2)=11.15, d.f. = 1; p = 0.001). Furthermore, log
istic regression analysis showed that of all the polymorphisms tested, only
the Sp1 binding site polymorphism acted as an independent predictor of fra
cture: odds ratio [95% CI] = 2.26 [1.09-4.69]. These data suggest that it i
s the Sp1 polymorphism rather than other polymorphisms at the COL1A1 locus
which act as a marker for osteoporotic fractures.