GWAS of bone size yields twelve loci that also affect height, BMD, osteoarthritis or fractures

Abstract

Bone area is one measure of bone size that is easily derived from dual-energy X-ray absorptiometry (DXA) scans. In a GWA study of DXA bone area of the hip and lumbar spine (N ≥ 28,954), we find thirteen independent association signals at twelve loci that replicate in samples of European and East Asian descent (N = 13,608 - 21,277). Eight DXA area loci associate with osteoarthritis, including rs143384 in GDF5 and a missense variant in COL11A1 (rs3753841). The strongest DXA area association is with rs11614913[T] in the microRNA MIR196A2 gene that associates with lumbar spine area (P = 2.3 × 10^-42, β = -0.090) and confers risk of hip fracture (P = 1.0 × 10^-8, OR = 1.11). We demonstrate that the risk allele is less efficient in repressing miR-196a-5p target genes. We also show that the DXA area measure contributes to the risk of hip fracture independent of bone density.

Fig. 1

Manhattan plots of the genome-wide analysis of the DXA bone area measures. The P values (−log10) are plotted against their respective positions on each chromosome. Results are shown for all variants with significance level P < 0.001 and imputation information greater than 0.8. The red line denotes the significance level of intergenic variants, P ≤ 7.9 × 10–10. a Lumbar spine area (L1–L4) (N = 29,059), (b) Total hip area (N = 28,900), (c) Femoral neck area (N = 28,954), (d) Intertrochanteric/shaft area (N = 28,936), (e) Trochanter area (N = 28,944). Source data are provided as a Source Data file

Fig. 2

Repression of miR-196a-5p target genes by MIR196A2 C-allele or T-allele inserts. a–c Volcano plot shown for all pair-wise comparisons of cells with MIR196A2 C-allele or T-allele inserts and empty vector controls. In each comparison, red colored dots indicate the fourteen miR-196a-5p target genes. a C-allele compared with empty vector insert, (b) T-allele compared with empty vector insert, and (c) C-allele compared with T-allele. Negative log2-fold values reflect downregulation with miRNA insert and in a and b, whereas negative log2-fold values reflect downregulation of the C-allele in c. d Distribution of log2 fold-change values derived from the comparison of C-allele and T-allele transfected cells (negative log2 fold represents downregulation of the C-allele). The red and black curves represents the density distribution for the fourteen high-confidence miR-196a-5p target genes and other non-targets, respectively. The direction of effect is indicated below the x-axis by an arrow. Source data are provided as a Source Data file

Fig. 3

Forest plot showing estimated effect of rs11614913[T] in MIR196A2 on hip fractures. The odds ratio, 95% confidence interval (CI), P value of association and number of cases and controls is shown for each study. Source data are provided as a Source Data file

Fig. 4

Effect on DXA bone area measures, height and bone mineral density. The effect on area measures are plotted on the X-axis and the effect on reported height variants from the GIANT consortium, and reported femoral neck BMD (FN_BMD) and lumbar spine BMD (LS_BMD) variants from the GEFOS consortium are plotted on the Y-axis. Effects are in in standard deviations. The color of the dots/triangles/squares indicate the site of the DXA area measure, and the shape indicates the strength of association with height, and hip and lumbar spine BMD; squares, non-significant, dots, significant after correcting for FDR, and triangles, genome wide significant. Source data are provided as a Source Data file