Ultra-processed food consumption and atherosclerosis risk: a prospective cohort study of UK Biobank

Abstract

Background Ultra-processed foods (UPFs) have become dominant in modern diets, yet their relationship with atherosclerosis—the primary pathological basis of cardiovascular disease—remains inadequately characterized. We investigated the association between UPF consumption and incident atherosclerosis risk and explored underlying metabolomic and proteomic molecular signatures. Methods We conducted a prospective analysis of 133,797 UK Biobank participants free of atherosclerosis at baseline. Dietary intake was assessed using 24-h dietary recalls, and UPF consumption was categorized according to the NOVA classification system. Incident atherosclerosis cases were identified through hospital admission records and death registries. Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs). In subset analyses, we performed nuclear magnetic resonance metabolomic profiling and plasma proteomic analysis using the Olink platform to identify molecular signatures associated with UPF-related atherosclerosis risk. Results During a median follow-up of 10.93 years (interquartile range, 10.83–11.40 years), 654 incident cases of atherosclerosis were documented. After multivariable adjustment, participants in the highest quartile (Q4) of servings/day showed an elevated risk (HR 1.40, 95% CI 1.12–1.75; p = 0.003) compared to the lowest quartile (Q1). Similarly, participants in Q4 of UPF consumption (g/day) had a 29% increased risk of atherosclerosis compared to those in Q1 (HR 1.29, 95% CI 1.03–1.60; p = 0.024). Restricted cubic spline analysis revealed higher UPF intake was associated with higher risk. Sparse LASSO identified UPF-related signatures comprising 30 metabolites and 25 proteins, predominantly involving lipoprotein, energy-metabolic, inflammatory, and endocrine pathways. Several biomarkers remained associated with incident atherosclerosis after multivariable adjustment and multiple-testing correction. Conclusion Higher UPF consumption is associated with increased atherosclerosis risk. Distinct metabolomic and proteomic signatures characterized by systemic inflammation, lipid dysregulation, and enzymatic alterations may reflect potential pathways. These findings support dietary recommendations emphasizing minimally processed foods for cardiovascular disease prevention and highlight candidate molecular correlates requiring independent validation.