Investigating causality in the associations between cardiometabolic biomarkers and cancer risks

Meta-analyses identified several genome-wide significant associations, including novel variants and loci not previously linked to circulating bile acid levels. Among the main findings, two significant variants mapped to protein-coding genes, PFKP and NAA35, with links to glycolytic and protein regulation. These findings then informed the selection of genetic instruments for downstream MR analyses.<br><br>Two-sample MR analyses provided evidence supporting a positive association between genetically predicted deoxycholic acid (DCA) levels and colorectal cancer (CRC) risk, particularly among males and for rectal cancer.

This is consistent with experimental and epidemiological evidence implicating secondary bile acids in colorectal carcinogenesis. In contrast, MR analyses of genetically predicted plasma fibrinogen and soluble P-selectin did not support causal associations with lung cancer risk, despite strong observational evidence. Bi-directional MR suggested that elevated fibrinogen levels may reflect smoking exposure rather than a direct aetiological role.<br><br>Overall, this thesis demonstrates the value of genetic epidemiology in strengthening causal inference for circulating biomarkers in cancer.

The findings highlight DCA as a potential modifiable risk factor for CRC and suggest a greater role for haemostatic biomarkers in risk prediction than causal intervention.<br><br><i>Thesis is embargoed until 31 July 2029.</i><br>