Abstract
The metabolic state of the body can be a major determinant of bone health. We used a Mendelian Randomization approach to identify metabolites causally associated with bone mass to better understand the biological mechanisms of osteoporosis. We tested bone phenotypes (femoral neck, total hip, and lumbar spine BMD) for association with 280 fasting blood metabolites in 6055 women from TwinsUK cohort with genome-wide genotyping scans. Causal associations between metabolites and bone phenotypes were further assessed in a bidirectional Mendelian Randomization study using genetic markers/scores as instrumental variables. Significant associations were replicated in 624 participants from Hong Kong Osteoporosis Study (HKOS). 15 metabolites showed direct associations with bone phenotypes after adjusting for covariates and multiple testing. Using genetic instruments, 4 of these metabolites were found to be causally associated with hip or spine BMD. These included androsterone sulfate, epiandrosterone sulfate, 5alpha-androstan-3beta17beta-diol disulfate (encoded by CYP3A5), and 4-androsten-3beta17beta-diol disulfate (encoded by SULT2A1). In the HKOS population, all four metabolites showed significant associations with hip and spine BMD in the expected directions. No causal reverse association between BMD and any of the metabolites were found. In the first metabolome-genome-wide Mendelian randomization study of human bone mineral density, we identified 4 novel biomarkers causally associated with BMD. Our findings reveal novel biological pathways involved in the pathogenesis of osteoporosis. This article is protected by copyright. All rights reserved.
| Original language | English |
|---|---|
| Pages (from-to) | 643-650 |
| Journal | Journal of Bone and Mineral Research |
| Volume | 33 |
| Issue number | 4 |
| Early online date | 12 Dec 2017 |
| DOIs | |
| Publication status | Published - Apr 2018 |
Keywords
- Journal Article