Studies of the associations of circulating sclerostin levels with a variety of outcomes have been inconclusive, although there have been reports of associations with a variety of conditions. It is unclear with such associations, however, whether these findings represent causal effects of sclerostin, or confounding. Mendelian randomization offers an approach to estimating causal effects in the presence of such confounding.
Once common variants robustly associated with serum sclerostin levels are detected it is possible to apply two-sample Mendelian randomization analyses to predict the effect of influencing serum sclerostin on a very wide range of outcomes, utilising the available GWAS data from a large number of conditions that could be influenced by serum sclerostin modification.
Requirements for a comprehensive investigation applying this approach to sclerostin would be (1) assaying sclerostin on a large sample of GWAS’d study participants to produce reliable data of genetic associations with serum sclerostin, (2) performing two-sample Mendelian randomization analyses predicting the effect of modifying sclerostin on a wide variety of health-related outcomes (see appendix for a summary of accessible resources), (3) within single large samples with particularly informative phenotyping (such as ALSPAC and UK Biobank) performing a detailed analysis of the predicted effects of modifying sclerostin.
These data would then provide powerful GWAS assessment of genetic associations with sclerostin as well as the basis for a comprehensive investigation of the association of sclerostin with a wide range pf phenotypes. In addition, factors (e.g. obesity) that may influence sclerostin can also be investigated within both the conventional and MR frameworks.