Pan-cancer analysis of the impact of fatty acids: a two-sample and multivariable Mendelian randomization study

This study used Mendelian randomization (MR) to investigate the causal association between circulating fatty acids (FAs) and the incidence of 10 human cancer types.

Single nucleotide polymorphisms (SNPs) significantly associated with circulating FAs were selected from 500,000 UK Biobank blood samples and 114,999 plasma samples from a large GWAS. Summary-level data on 174,006 cancer patients were obtained from the FinnGen biobank. The inverse variance weighted method was used for causal estimation, with additional analyses including MR-Egger regression, weighted median, weighted mode, Cochrane Q test, MR-PRESSO global test, and leave-one-out analyses. Multivariable Mendelian randomization (MVMR) was applied to adjust for potential confounders.

This study found a genetically causal effect of polyunsaturated FAs on kidney cancer incidence (OR: 1.528; 95% CI 1.164-2.266; P = 0.001), which remained significant after Bonferroni adjustment. The causal impact of omega-6, omega-3, and linoleic acid on kidney cancer risk was also observed (omega-6: OR = 1.586, P = 0.002; omega-3: OR = 1.311, P = 0.014; linoleic acid: OR = 1.527, P = 0.007). MVMR confirmed the consistent causal relationship (OR = 1.553, P = 0.0047) after adjusting for multiple variables. Results were validated in a larger cohort.

Higher circulating polyunsaturated FAs, especially omega-6, were associated with an increased risk of kidney cancer. Suggestive associations were found in small cell lung, rectal, bladder, pancreatic, and esophageal cancer.