Abstract
This study explores the causal relationship between plasma lipidome and 6 common cancers using Mendelian randomization. Genome-wide association study data on 179 lipid species were used as exposures, with genome-wide association study datasets for 6 cancers as outcomes. The primary method was inverse variance weighted analysis, supplemented by Mendelian randomization-Egger regression and weighted median approaches. Sensitivity analyses, including the Steiger test, were conducted to evaluate the directionality and robustness of the causal relationships. Key findings identified phosphatidylethanolamine (16:0_20:4) as a risk factor for hepatic cancer, while sphingomyelin (d40:1) was protective. For lung cancer, sterol ester (27:1/20:4) increased risk, whereas sterol ester (27:1/18:2) reduced it. Sterol ester (27:1/20:3) was a risk factor for colorectal cancer, while phosphatidylcholine (18:2_0:0) was protective. Phosphatidylcholine (16:0_20:4) increased esophageal cancer risk, while phosphatidylcholine (16:0_18:3) offered protection. Phosphatidylinositol (18:0_20:4) increased thyroid cancer risk, while phosphatidylinositol (16:0_18:2) was protective. Diacylglycerol (18:1_18:2) was protective against breast cancer. These findings shed light on the causal link between lipid metabolism and cancer risk, contributing to the understanding of lipid involvement in cancer pathogenesis and offering potential strategies for early detection and treatment.