Observational studies have reported associations between breast cancer and a range of metabolic traits, from circulating metabolites to metabolic disease. Although both breast cancer and metabolic traits exhibit extensive polygenic architecture, it remains unclear whether they share inherited genetic signatures that could inform biological mechanisms or clinical management. The nature and direction of any putative causal relationships also remain underexplored.
To address this, we analysed genetic correlation and causation between 307 metabolic traits and four large breast cancer genome-wide association studies (GWAS): Zhang et al. (2020; 133,384 cases and 113,789 controls) and FinnGen Release 12 (24,270 cases and 222,078 controls), including ER-positive (14,540 cases) and ER-negative (9,724 cases) subtype analyses. These were analysed against GWAS of 249 circulating metabolites, 50 biochemical and blood cell traits from the UK Biobank, and 8 metabolic syndrome traits.
Linkage disequilibrium score regression identified 274 significant genetic correlations (FDR < 0.05), many involving lipoprotein subclasses. Across two complementary causal inference frameworks, latent causal variable modelling and Mendelian randomisation using CAUSE, no metabolic trait showed evidence for a causal effect on breast cancer. Instead, 10 VLDL and HDL related trait pairings demonstrated strong evidence for correlated pleiotropy rather than direct causation, suggesting shared biology.
MAGMA gene-level analyses highlighted MAP3K1 as the most consistent established genic signal across these trait pairings, while FTO and IGF1 appeared repeatedly across VLDL related combinations. Notably, SETD9 and MIER3 were significant across all VLDL pairings and surpassed Bonferroni thresholds. Their consistent association with these traits, despite not being recognised breast cancer susceptibility genes, highlights their potential novelty within the shared inherited architecture linking metabolic traits and breast cancer.
Together, these findings reveal shared common variant signatures between breast cancer susceptibility and lipoprotein traits, highlighting new opportunities to clarify the biological links between metabolism and breast cancer risk.