Associations of dietary intake with influenza susceptibility: evidence from bidirectional, multivariable, and mediation mendelian randomization analyses

Objectives To investigate the causal associations between dietary intake and influenza susceptibility and severity using two-sample Mendelian randomization (MR), multivariable MR (MVMR), and exploratory mediation MR analyses. Methods Twenty-five dietary factors were obtained from genome-wide association studies (GWAS) of the UK Biobank. Three influenza outcomes derived from the FinnGen consortium were evaluated, including all influenza, regular influenza (influenza without pneumonia), and severe influenza (influenza with pneumonia). The inverse-variance weighted (IVW) method was used as the primary MR approach, followed by sensitivity analyses and MVMR adjusting for smoking, educational attainment, and household income. Benjamini–Hochberg false discovery rate (FDR) correction was applied to both the primary MR and mediation analyses. Results After FDR correction, genetically predicted higher oily fish intake was associated with a lower risk of all influenza, whereas higher fresh fruit and dried fruit intake were associated with a lower risk of severe influenza. MVMR demonstrated that only the protective association for dried fruit intake remained robust after adjustment for potential socioeconomic and lifestyle confounders. Exploratory mediation analyses identified nominal associations for vitamin E and DHA; however, neither remained statistically significant after FDR correction. Conclusions This study provides genetic support for associations between several dietary factors and influenza across different levels of disease severity. Dried fruit intake showed the most robust independent association after multivariable adjustment. Exploratory mediation analyses did not identify statistically significant mediators after correction for multiple testing, highlighting the need for further mechanistic and interventional studies.