Exploring causal associations between nuclear magnetic resonance biomarkers and colorectal cancer risk.

Background: Emerging evidence shows significant differences in plasma metabolites between colorectal cancer (CRC) patients and healthy controls. However, previous observational studies have been limited by small sample sizes and single sample sources, leading to an incomplete understanding of these metabolites' causal roles in CRC. This study systematically evaluated the causal relationships between 325 nuclear magnetic resonance (NMR) biomarkers and CRC risk using Mendelian randomization (MR), supplemented by colocalization analysis and an independent validation dataset to confirm key biomarkers.

Methods: A genome-wide association study (GWAS) was conducted in a cohort of 250,341 participants from the UK Biobank. MR analysis identified NMR biomarkers with significant causal relationships with CRC. Colocalization analysis was then performed, revealing five biomarkers with high colocalization probabilities (PPH4 > 0.8). These findings were validated in an independent Finnish dataset to confirm the consistency of causal relationships and colocalization signals.

Results: MR analysis identified 28 NMR biomarkers with significant causal associations with CRC risk (P_fdr < 0.05). Colocalization analysis highlighted five biomarkers with strong colocalization signals (PPH4 > 0.8), including Omega-6 fatty acids, Omega-6 to total fatty acids ratio, Omega-3 fatty acids, Linoleic acid to total fatty acids percentage, and Degree of unsaturation. Notably, in the Finnish validation dataset, Linoleic acid to total fatty acids percentage demonstrated a significant causal association with CRC (OR 0.77, 95% CI 0.67-0.87, P = 7.5 × 10^-5, P_fdr = 3.8 × 10^-4) while maintaining a high colocalization probability (PPH4 > 0.8), reinforcing its role as a key causal biomarker.

Conclusions: This study provides the first comprehensive assessment of NMR biomarkers in relation to rectal cancer risk, identifying linoleic acid to total fatty acids percentage as a key causal biomarker. Additionally, omega-6 to omega-3 ratio, omega-6 to polyunsaturated fatty acids percentage, omega-3 to polyunsaturated fatty acids percentage, and degree of unsaturation were also identified, sharing genetic loci with CRC.