Background The association between adiposity and increased liver cancer risk is well-recognized, yet underlying metabolic mechanisms require elucidation. This study aimed to identify metabolic mediators linking adiposity markers to liver cancer and assess their potential causality using two-sample Mendelian randomization (MR) analysis. Methods and findings We conducted a 1:1 matched nested case-control study within a population-based and prospective cohort study—the Shanghai Men’s Health Study (SMHS). The SMHS was initiated in 2002–2006, including 61,469 Chinese men aged 40–74 years, and has been followed up for over 20 years. Targeted metabolomic profiling was performed on baseline plasma samples. Associations between seven anthropometric measurements (body mass index BMI, waist circumference, waist-to-hip ratio, waist-to-height ratio, a body shape index, hip circumference, and adult weight gain), 186 circulating metabolites, and liver cancer risk were assessed. Linear and conditional logistic regression model adjusted for multiple confounders (including smoking, alcohol drinking, physical activity, chronic hepatitis and cirrhosis, diabetes, etc.) were used. Pathway analysis and network analysis were conducted to explore the biological functions of these metabolites. Parallel mediation analysis was employed to quantify the mediating effects through metabolites. Subsequently, MR analysis was performed to investigate potential causal relationships. This study incorporated 322 incident liver cancer cases and 322 cancer-free controls. Participants diagnosed with liver cancer had higher proportions of seropositive hepatitis B surface antigen (63.7%) compared to their matched controls (6.2%). We identified 27 intermediate metabolites associated with both adiposity markers and liver cancer risk, which formed an interconnected functional network. Pyroglutamic acid demonstrated the most robust consistency, being significantly associated with seven anthropometric measurements (β per doubling with BMI = 0.17; 95% confidence interval CI: 0.09, 0.24) and liver cancer (odds ratio per doubling = 1.56; 95% CI: 1.13, 2.15). Pathway analysis highlighted significant alterations in energy, lipid, and amino acid metabolism. Specifically, Phenylalanine, tyrosine, and tryptophan biosynthesis showed the highest impact, suggesting a key role for aromatic amino acid metabolism. Parallel mediation analysis demonstrated significant indirect effects via intermediate metabolites for six of the seven anthropometric measurements, with the proportion mediated by the identified metabolite clusters reaching 0.16 (95% CI: 0.05, 0.29) for BMI. MR analysis provided evidence supporting potential causality for 23 of 108 initially observed associations. The strongest association was observed between WC and oxoglutaric acid (β IVW per standard deviation = 0.31; 95% CI: 0.17, 0.43). Notably, while the observational analysis suggested a broad metabolic mediation of the adiposity marker-liver cancer association, the MR findings pinpointed a more specific and limited set of causal metabolic mediators. The main limitation of this study was the population mismatch between the observational (Chinese men) and the MR (European ancestry) analyses, which may limit the generalizability of the findings to other populations. Conclusions Integrating prospective observational and genetic evidence, we identified specific metabolic mediators linking adiposity to liver cancer, particularly involving amino acid, lipid and energy metabolism. These findings enhanced molecular understanding of adiposity-driven hepatocarcinogenesis and provided potential metabolic targets for future primary prevention strategies.
Li et al. (Mon,) studied this question.