Abstract
The causal effect of lower plasma sclerostin on cardiovascular disease (CVD) risk has previously been examined with the aim of investigating potential side effects of pharmacological sclerostin inhibition for treatment of osteoporosis. We explored the relationship between plasma sclerostin levels and CVDs and bone phenotypes using Mendelian randomization (MR) and correlation between plasma sclerostin levels and these outcomes. We used variants identified in genome-wide association studies of plasma sclerostin levels in large proteomic datasets from the UK Biobank (Olink) and Iceland (SomaScan) as instruments in two separate MR analyses. These analyses did not provide evidence of association between the effects of sequence variants on plasma sclerostin levels and their effects on CVDs and CVD risk factors (P > 0.05). Several of the instruments had heterogenic effects on bone phenotypes and causal estimates in MR were non-significant (P > 0.05/8). Plasma sclerostin levels correlated positively with coronary artery disease, myocardial infarction and CVD risk factors. Our results do not provide evidence supporting the hypothesis that lower plasma sclerostin levels increase CVD risk and suggest that plasma sclerostin levels are not a good surrogate for pharmacological inhibition.