Examining the causal link between cheese intake and susceptibility to type 2 diabetes through metabolic traits: Insights from Mendelian randomization

Type 2 diabetes (T2D) is a global metabolic disorder influenced by diet. While cheese consumption has been suggested to protect against T2D, the mediating metabolic pathways remain unclear. Using Mendelian randomization (MR), we analyzed genetic variants associated with cheese intake (UK Biobank) and their effects on T2D risk (FinnGen) via 249 metabolites. Instrumental variables were selected under stringent criteria. Causal estimates were derived via inverse-variance weighted method, with sensitivity analyses (MR-Egger method) and mediation testing. Cheese intake was inversely associated with T2D risk (odds ratio [OR] = 0.616). Of 249 metabolites, 26 mediated this relationship, including triglycerides (e.g., total triglycerides levels, 3.55% mediation), free cholesterol (e.g., free cholesterol levels in high-density lipoprotein, 5.84%), glucose (3.43%), and albumin (2.69%). Among the 66 metabolites identified as T2D-associated, notable examples included: AA (Phe OR = 1.14, Leu OR = 1.22, Val OR = 1.13), ketone bodies (acetoacetate OR = 1.62, 3-hydroxybutyrate OR = 1.36), and protective factors (albumin OR = 0.871, apolipoprotein A1 OR = 0.871). Cheese consumption may lower T2D risk primarily by modulating lipid metabolites (triglycerides, cholesterol esters) and albumin. These findings highlight novel metabolic pathways for dietary prevention of T2D and underscore the need for research on cheese-specific components (e.g., fermentation byproducts).