A systematic evaluation of 151 candidate genes for their association with osteoporosis and osteoporotic fracture in a meta-analysis of genome-wide association data

Abstract

Background: Osteoporosis has a strong heritable component and many candidate genes have been proposed to be involved in the regulation of bone mineral density (BMD) but few of these genes demonstrated replicated associations in independent studies. Methods: To assess the relationship between all common polymorphisms, represented by single nucleotide polymorphisms (SNPs), in previously proposed candidate osteoporosis genes with BMD and fracture we performed a large-scale multicentre metaanalysis of genome-wide association data. Cohorts included 19,195 participants (14,277 women) from five European populations for BMD. Data on fracture were available from a prospective cohort (n=5974) from the Netherlands. We identified through the HuGENet Navigator 151 genes that had been proposed in the literature as potentially associated with osteoporosis. We evaluated data on 34,697 common HapMap single nucleotide polymorphisms (SNPs) across all these genes. BMD at femoral neck and lumbar spine was measured by DXA. Fractures were defined as clinically apparent, site-specific, validated fractures. Results: 179 tagging SNPs achieved statistical significance and stemmed from 9 separate gene loci (ESR1, F2, ITGA1, LRP5, SOST, SPP1, TNFRSF11A, TNFRSF11B and TNFSF11) and were associated with either BMD site. The SNP most strongly associated with spine BMD was rs6469804 from the TNFRSF11B locus (p=1.4null10E-14). The top SNP for femoral neck BMD was rs2070852, which is in the F2 gene (p=4.0null10E-9). For most candidate genes no SNP achieved statistical significance. For statistically significant SNPs, the effect sizes were typically small; the effect ranged from 0.04 to 0.18 SD change in BMD per allele. 33 of the 179 SNPs significantly influenced risk of osteoporotic fracture and arose from the SPP1, SOST and LRP5 loci; the effects of each SNP on risk of fracture were modest. Conclusions: Nine genes proposed as being important for osteoporosis are indeed associated with regulation of BMD and 3 of those also affect risk of fracture in this very large collaborative genome-wide meta-analysis. However, the majority of previously described candidate genes had no consistent association with BMD.