Cancer incidence and development are strongly correlated with obesity, however there is insufficient data to support a causal relationship between intentional weight loss and the prevention or promotion of cancer. We investigated the causal relationship between weight loss following Roux-en-Y gastric bypass (RYGB) and the incidence of 18 cancers using Mendelian randomization (MR). A genome-wide association studies (GWAS) data related to weight loss following RYGB from the GWAS catalog database were used as exposure, and GWAS data related to 18 cancers from the Medical Research Council integrative epidemiology unit open GWAS project were used as outcomes. In order to investigate the causal relationship between exposure and results, we used a two-sample MR approach. The primary analysis technique was inverse variance weighting, with weighted median, and MR-Egger regression utilized as supplemental techniques to confirm the findings. Heterogeneity and horizontal pleiotropy were investigated using a variety of sensitivity studies, including the Cochran Q test, MR-Egger regression pleiotropy test, MR pleiotropy residual sum and outlier, and leave-one-out analysis. We included a total of 4 single-nucleotide polymorphisms as instrumental variables through rigorous quality control screening. Under the limitations of Bonferroni correction threshold (P < 2.78 × 10-3), our results suggest that the weight loss following RYGB has a significant causal relationship with a reduced risk of breast (odds ratio [OR]: 0.784; 95% confidence interval [CI]: 0.762-0.808; P = 2.167e-58) and lung cancer (OR: 0.992; 95% CI: 0.987-0.997; P = .0023), and a potential causal relationship with a decreased risk of hematological cancer (OR: 0.9998462; 95% CI: 0.9997088-0.9999836; P = .028) and an increased risk of cervical cancer (OR: 1.000123; 95% CI: 1.0000313-1.000215; P = .009). Sensitivity analysis confirms the robustness of our analysis results. Genetically predicted weight loss following RYGB has significant causal effects in reducing the risk of breast and lung cancer. It also has potential benefits in lowering the risk of hemotological cancers and increasing the risk of cervical cancer. Considering the limitations of our study, the reliability of its results and the underlying mechanisms require further investigation.
Read full abstract