Effect of genetically predicted sclerostin on cardiovascular biomarkers, risk factors, and disease outcomes

Abstract

Sclerostin inhibitors protect against osteoporotic fractures, but their cardiovascular safety remains unclear. We conducted a cis-Mendelian randomisation analysis to estimate the causal effect of sclerostin levels on cardiovascular risk factors. We meta-analysed three GWAS of sclerostin levels including 49,568 Europeans and selected 2 SNPs to be used as instruments. We included heel bone mineral density and hip fracture risk as positive control outcomes. Public GWAS and UK Biobank patient-level data were used for the study outcomes, which include cardiovascular events, risk factors, and biomarkers. Lower sclerostin levels were associated with higher bone mineral density and 85% reduction in hip fracture risk. However, genetically predicted lower sclerostin levels led to 25-85% excess coronary artery disease risk, 40% to 60% increased risk of type 2 diabetes, and worse cardiovascular biomarkers values, including higher triglycerides, and decreased HDL cholesterol levels. Results also suggest a potential (but borderline) association with increased risk of myocardial infarction. Our study provides genetic evidence of a causal relationship between reduced levels of sclerostin and improved bone health and fracture protection, but increased risk of cardiovascular events and risk factors.

Fig. 1. Forest plot of MR estimates.

Effect sizes were calculated using the generalised inverse variance weighted method. Blue represents the results based on published GWAS summary statistics as the outcome; maroon the results obtained using a linear regression (for continuous outcomes) and logistic regression (for categorical outcomes) on the UK Biobank outcomes; and orange indicates the results using cox regression for UK Biobank survival outcomes. Error bars indicate the 95% confidence interval. A The horizontal axis shows the SD change per 1 SD decrease in sclerostin levels. GWAS results for LDL and HDL-Cholesterol are in mg/dL, mmol/L for GWAS results of fasting glucose, % change for HbA1c and SD increase for the other outcomes. B The horizontal axis shows the odds/hazard ratio per 1 SD decrease in sclerostin levels. Source data are provided as a Source Data file. Note: UKB-LR = UK Biobank logistic regression, UKB-SA = UK Biobank survival analysis. Details about sample sizes used to calculate the MR estimates can be found in Supplementary Table 5, Supplementary Table 7, and Supplementary Table 8.

Fig. 2

Schema of the study design.