Abstract 977: Metabolomic and proteomic signatures of ultra-processed foods with liver cancer and liver disease

Abstract Introduction: Higher intake of ultra-processed foods is associated with increased risk of obesity and type 2 diabetes, but evidence is limited on associations of metabolomics and proteomics with ultra-processed foods and liver diseases. We aimed to evaluate associations of metabolomic and proteomic signatures related to ultra-processed food intake with liver cancer and liver disease. Methods: Data from a total of 173, 840 participants aged 40-69 years from the UK Biobank were analyzed. Ultra-processed food intake was assessed by 24-hour dietary recalls. Metabolites were measured by nuclear magnetic resonance spectroscopy and plasma proteome was profiled using the Olink platform. Liver disease outcomes were ascertained using data from the in-hospital, cancer registry, and death registry including metabolic dysfunction-associated steatotic liver disease (MASLD), cirrhosis, liver cancer, and severe liver disease (hepatocellular carcinoma and other severe liver diseases). Elastic net regression with 10-fold cross validation was used to calculate the omics signatures related to ultra-processed food intake. We used Cox proportional hazards regression models to estimate hazard ratios (HR) and 95% confidence intervals (CIs) for ultra-processed food intake and its omics signatures and liver disease outcomes adjusting for multiple potential confounding factors. Results: With a median of 8.9 years follow-up, ultra-processed food intake was positively associated with risk of MASLD (HR Tertile 3 vs. tertile 1 1.42; 95% CI 1.21-1.67), cirrhosis (HR 1.25; 95% CI 1.05-1.48), and severe liver disease (HR 1.52; 95% CI 1.21-1.91) but not with liver cancer (HR 1.30; 95% CI 0.89-1.91). The metabolic signature score of ultra-processed foods was positively associated with risk of MASLD (HR 3.25; 95% CI 2.06-5.14). The proteomic signature score of ultra-processed foods was positively associated with risk of MASLD (HR 1.99; 95% CI 1.04-3.80) and cirrhosis (HR 2.00; 95% CI 1.12-3.60). A total of 34 metabolites and 65 proteins were significantly associated with ultra-processed food intake based on the multivariable-adjusted model and elastic net regression. These metabolites and proteins are linked to biological pathways such as lipids metabolism, oxidative stress, and inflammatory response. Conclusions: Ultra-processed food intake and its metabolic and proteomic signatures are positively associated with risk of adverse liver outcomes. Citation Format: Longgang Zhao, 1 Yun Chen, 1 Alyssa Clay-Gilmour, 2 Jiajia Zhang, 2 Xuehong Zhang, 1 Susan Steck2. Metabolomic and proteomic signatures of ultra-processed foods with liver cancer and liver disease [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 977.