Association of Body Mass Index and the Risk of Gastro-Esophageal Cancer: A Mendelian Randomization Study in a Japanese Population

Zhaoping Zang,Yi Shao,Rena Nakyeyune,Yi Shen,Chen Niu,Lingyan Zhu,Xiaoli Ruan,Tong Wei,Ping Wei,Fen Liu
DOI: https://doi.org/10.1080/01635581.2022.2132266
2022-10-09
Nutrition and Cancer
Abstract:There are growing concerns that body mass index (BMI) is related to cancer risk at various anatomical sites, including the upper gastrointestinal tract, and the existence of a causal relationship remains unclear. The Mendelian randomization (MR) method uses instrumental genetic variables of risk factors to explore whether a causal relationship exists while preventing confounding. In our study, genome-wide association study (GWAS) data from the BioBank Japan (BBJ) project were used. Genetic variants were chosen as instrumental variables using inverse-variance weighting (IVW), MR–Egger regression and weighted-median methods to estimate the causal relationship between BMI and the risk of gastro-esophageal cancer. We found no evidence to support a causal association between BMI and risk of gastric cancer [odds ratio (OR) =0.99 per standard deviation (SD) increase in BMI; 95% confidence interval (CI): (0.76–1.30); P = 0.96] or esophageal cancer [0.78(0.50–1.22); P = 0.28] using the IVW method. Sensitivity analysis did not reveal any sign of horizontal pleiotropy. Additionally, in the gender-stratified analysis, no causal association was found. Findings from this study do not support a causal effect of BMI on gastro-esophageal cancer risk. However, we cannot rule out a modest or nonlinear effect of BMI.
What problem does this paper attempt to address?