Abstract Background Osteoporosis and cardiovascular disease (CVD) are major contributors to global disease burden and health expenditure. There are well established associations between these two conditions, but it is unknown whether such relationships are causal or simply reflect common risk factors. Therefore, a comprehensive two-sample Mendelian Randomisation (MR) analysis was conducted to investigate the causal relationship between BMD and seven CVDs, leveraging summary statistics from large-scale genome-wide association studies (GWAS). Methods Single nucleotide polymorphisms (SNPs) were selected as genetic instruments based on rigorous quality control criteria, including genome-wide significance (p 5 × 10–8) and independence (linkage disequilibrium r2 0.001 within 10,000 kb genomic region). Summary-level GWAS data for total-body BMD were obtained from a European GWAS meta-analysis of 66,628 participants1. The summary statistics for CVDs (atrial fibrillation, angina, ischaemic heart disease, heart failure, hypertension, myocardial infarction and non-ischaemic cardiomyopathy) were extracted from European-descent individuals in UK Biobank, using data from three published GWASs. Independent validation was conducted using data from the FinnGen consortium, comprising 224,737 participants of European ancestry2. Causal estimates were calculated using inverse variance weighted (IVW), complemented by sensitivity analyses weighted median, weighted mode and MR-Egger regression (MR-Egger) and MR pleiotropy residual sum and outlier (MR-PRESSO) to assess pleiotropy and robustness. Statistical power calculations were conducted using the mRnd tool, applying a Bonferroni-corrected significance level of α=0.05/8 and 90% power. Results A total of 85 lead SNPs consistent with the MR correlation hypothesis were included from GWAS data of total-body BMD. No evidence of causality was found between genetically predicted BMD and the seven CVDs investigated, including atrial fibrillation (IVW-estimated β: 0.011, SE: 0.03, p=0.73), angina (β: 0.04, SE: 0.03, p=0.17), ischaemic heart disease (β: 0.009, SE: 0.03, p=0.74), myocardial infarction (β: 0.02, SE: 0.03, p=0.36), heart failure (β: 0.004, SE: 0.04, p=0.91), hypertension (β: -0.01, SE: 0.01, p=0.44) and non-ischaemic cardiomyopathy (β: 0.1, SE=0.08, p=0.20). Results were consistent across sensitivity analyses with no evidence of horizontal pleiotropy or directional bias. Validation analysis using the independent Finnish dataset, yielded null associations across all methods (Figure 1). Conclusion The findings from this analysis do not support a causal relationship between genetically predicted BMD and a broad spectrum of CVDs. These findings suggest that previously observed associations may be attributable to shared risk factors rather than direct causal mechanisms. Public health strategies should focus on modification of shared risk factors to alleviate the burden of both bone and CVDs.
Condurache et al. (2025) studied this question.