Investigating the causal relationship between dietary habits and oral cancer risk: A Mendelian randomization analysis

The main goal of this study was to clarify the causal link between dietary habits and the occurrence of oral cancer via a Mendelian randomization (MR) approach. This method offers valuable insights that can inform the creation of effective management strategies in clinical settings. This investigation involved a comprehensive dataset that included 8,943,774 dietary samples and 3,723,373 oral cancer cases sourced from genome-wide association studies. To explore the possible causal relationship between dietary habits and the occurrence of oral cancer, various statistical methods, such as inverse variance weighted MR analysis, the weighted median approach, and MR-Egger regression analysis, have been employed. The variability in the results was evaluated via the Cochran Q statistic, and to address the issue of multiple comparisons, the MR-Egger and MR-PRESSO methods were utilized. Additionally, sensitivity analysis was conducted via the leave-one-out technique to ensure the reliability of the findings. Among 59 instrumental variables, inverse variance weighted analysis revealed an inverse relationship between oily fish intake and oral cancer risk (odds ratio = 0.998; 95% confidence interval: 0.997-1.000; P = .024). Nonsignificant results emerged from MR-Egger (P = .171) and weighted median (P = .483). Cochran Q testing confirmed no single-nucleotide polymorphism heterogeneity (P > .05), while MR-Egger (intercept = 3.776E-05, P = .389) and MR-PRESSO (P = .534) detected no horizontal pleiotropy. Leave-one-out analysis demonstrated robustness, with no single single-nucleotide polymorphism significantly influencing outcomes. Oily fish intake may reduce the incidence of oral cancer.