Abstract
BACKGROUND: Gliomas, particularly high-grade gliomas such as glioblastoma, represent a major challenge due to their poor prognosis. While dietary factors have been proposed as potential modulators of glioma risk, causal inference has been hindered by confounding and reverse causality in observational studies. This study employs Mendelian randomization to investigate the causal relationship between dietary factors and glioma risk. METHODS: A two-sample MR framework was applied, utilizing genome-wide association study data for 22 dietary exposures and glioma risks, including both GBM and non-GBM subtypes. Instrumental variables (genetic variants) were identified for each dietary factor to address confounding and pleiotropy. Causal inference was conducted using inverse-variance weighted regression, complemented by MR-Egger and MR-PRESSO analyses to assess and correct for potential pleiotropy. RESULTS: A positive causal association was observed between the intake of cooked vegetables and the GBM risk (OR = 6.55, 95% CI: 1.86-23.12, p = 0.00350). While alcohol intake demonstrated a protective effect for non-GBM risk (OR = 0.770, 95% CI: 0.61-0.97, p = 0.029), beer was substantially linked to an increased risk of non-GBM gliomas (OR = 4.82, 95% CI: 1.84-12.59, p = 0.0014). Other dietary factors did not exhibit significant causal associations. CONCLUSIONS: These findings suggest that certain dietary factors, including cooked vegetable intake, beer consumption, and alcohol intake, may exert a causal influence on glioma risk. This study provides new insights into the potential dietary determinants of glioma and underscores the need for further investigation into modifiable risk factors for glioma prevention.