Table 3.
Summary of association results of rs9547970 in two studied cohorts
| Either LS or FN BMD | LS BMD | FN BMD | Vertebral fracturea (n = 1,746) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| A1 | A2 | MAF | P value | OR (95% CI)/β | P value | OR (95% CI)/β | P value | OR (95% CI)/β | P value | OR (95% CI) | |
| HKSC extreme cohort (n = 1,572) | G | A | 0.265 | 6.8 × 10−4 | 1.41 (1.16–1.73) | 0.007 | 1.38 (1.09–1.76) | 0.019 | 1.42 (1.06–1.91) | NA | NA |
| HKOS prospective cohort (n = 2,509) | G | A | 0.278 | NA | NA | 0.023a | −0.078b | 0.039a | −0.061b | 0.007 | 1.33 (1.08–1.62) |
| Meta-analysisc (n = 4,081) | NA | NA | 0.003 | NA | 0.010 | NA | NA | NA | |||
The top imputation finding, rs9547970, was validated by direct genotyping in the Hong Kong Southern Chinese (HKSC) extreme cohort and was replicated in the Hong Kong Osteoporosis Study (HKOS) prospective cohort. The results were adjusted for age, height, weight, gender, and LS BMD (vertebral fracture only)
A1 Minor/effect allele, A2 major allele, MAF minor allele frequency, OR odds ratio; OR >1 the effect allele is associated with the higher risk of low BMD or vertebral fracture, NA not available
aIn the replication cohort (HKOS prospective cohort), the listed P values of BMD were one-sided, as they have the same direction of effect to the initial analysis in the HKSC extreme cohort. Other P values were all two sided.
bThe effects were presented as regression coefficient (β) estimated using the linear regression model.
cThe meta-analysis was done using a weighted z-transform test