Abstract
Osteoporosis is a serious public health problem, especially in the elderly. The most clinical importance of osteoporosis is its association with bone fragility. Bone geometry is an important risk factor for fracture, independent to BMD, and is genetically determined. However, genetic studies on bone geometry compared with BMD are scarce. Estrogen receptor a (ER-a) gene and its products have been shown to be of functional importance in bone metabolism. Although many studies have been performed to test the associations of BMD and ER-a gene, the results are controversial. In this study, seven single nucleotide polymorphisms (SNPs) in the ER-a gene were studied to examine the importance to the variation of femoral neck geometry. The study group included 1873 subjects from 405 Caucasian nuclear families. I used DXA-derived BMD and bone area to estimate 6 important geometry parameters: cross-sectional area (CSA), endocortical diameter (ED), sectional modulus (Z), cross-sectional moment of inertia (CSMI), cortical thickness (CT), and buckling ratio (BR). Haplotypes were reconstructed based on the half length of linkage disequilibrium (LD). The quantitative transmission disequilibrium test (QTDT), a family based association approach robust to population stratification was employed for analyses. Evidence of within-family association between SNP4 (rsl801132) and ED was detected in single locus analyses (P = 0.008) and verified in haplotype analyses (P = 0.034). The association was strengthened in total association analyses with P values of 0.005 and 0.031 for single locus and haplotype analyses, respectively. I also found weaker association between SNP5 (rs932477) and CT and BR in total association analyses with P values of 0.035 and 0.041 for single locus analyses and 0.010 and 0.004 for haplotype analyses, respectively. These findings suggest that the ER-a gene may be important for femoral neck geometry variation, though the significance is dampened by correction for multiple testing. Denser markers and larger sample may be necessary to finally elucidate the relationship between the ER-a gene and bone geometry.