Associations between multiple food intake and glioblastoma risk

Abstract

Emerging evidence suggests that food intake is possibly associated with glioblastoma (GBM), potential causal links between food intake and GBM risk remain unclear. Two-sample Mendelian randomization and multivariable Mendelian randomization analyses were conducted using pooled statistics from genome-wide association studies for GBM (European ancestry, N = 218,792) and 6 exposure factors: bread intake (N = 452,236) and bacon intake (N = 64,949) from European populations, and Danish pastry, oat crunch, latte, and chocolate bar intake (each N ≈ 1469) from the UK Biobank South Asian ancestry cohorts. The analyses were based on inverse variance weighted (IVW), MR-Egger, weighted median, simple models and weighted models. Results were based primarily on IVW using random effects and then sensitivity analyses, including heterogeneity detected through IVW and MR-Egger, horizontal pleiotropy analyzed by MR-Egger, and the leave-one-out method used to identify instrumental variables with promising effects, were performed. In the multivariable Mendelian randomization analysis, the protective effects of bread intake (odds ratio [OR] = 0.0046, 95% confidence interval [CI] = 0.0001–0.1, P = .022) and bacon intake (OR = 0.0003, 95% CI = 0.0001–00.1, P = .007) remained robust after mutual adjustment. For South Asian cohorts, oat crunch (OR = 1.679, 95% CI = 1.102–2.557, P = .016) and latte intake (OR = 1.340, 95% CI = 1.032–1.741, P = .028) persisted as independent risk factors for GBM. Conversely, the association for chocolate bar intake was attenuated to non-significance (OR = 1.51, 95% CI = 1.038–2.213, P = .031) after adjusting for other dietary components. Genetically predicted bread and bacon intake (based on European genome-wide association studies data) were associated with a decreased risk of GBM in European populations. Conversely, genetic liability to oat crunch and latte intake (based on South Asian ancestry data) showed potential positive associations with GBM risk, while Danish pastry intake suggested a protective effect. These findings highlight the potential impact of long-term dietary habits on GBM susceptibility, though the magnitude of certain associations warrants cautious interpretation due to population-specific genetic architectures.