The increased consumption of ultra-processed foods (UPFs), which are made using various extracts and additives, presents growing global health problems. In the United States, UPFs account for more than 50% of the calories consumed by both adults and children. Evidence links these foods to an increased risk of obesity, heart disease, and certain cancers. Traditional methods of assessing people’s diet, such as food frequency questionnaires and dietary recalls, can be inaccurate and are subject to biases, such as recall bias. To address this, scientists at the National Institutes of Health (NIH) developed a new tool called the poly-metabolite score to measure how much UPF people eat impartially. Rather than only relying on self-reported data about diet, the team employed metabolomics, measuring small molecules in blood and urine to identify metabolic markers associated with UPF consumption.
The research involved 718 participants from the Interactive Diet and Activity Tracking in AARP (IDATA) study and 20 participants in a controlled feeding trial. IDATA participants provided digital dietary data and biological samples over 12 months. In the controlled feeding trial, participants alternated between high-UPF diets (approximately 80%) and low-UPF diets (0%) for one week each. Thus, they were able to compare high-UPF and low-UPF consumption.
Most IDATA participants were between the ages of 60 and 69 years, with a balanced gender distribution and a predominantly white demographic. Among them, 43% were overweight and 30% were obese. On average, about half of the calories were derived from UPFs daily. High-UPF diets had lower protein, fiber, and essential nutrients, higher sugars, and carbohydrates. Urine samples from high-UPF consumers had less nitrogen, reflecting reduced protein intake.
Using advanced analytical techniques, including ultra-performance liquid chromatography–tandem mass spectrometry (UPLC-MS/MS) and least absolute shrinkage and selection operator (LASSO) regression, the researchers identified 191 metabolites in blood, 293 in 24-hour urine, and 237 in first-morning urine (FMU) positively associated with UPF intake. Forty-nine metabolites were present in all biological sample types. Participants with higher UPF consumption experienced higher concentrations of harmful metabolites such as N6-carboxymethyllysine (linked with diabetes) and acylcarnitines, along with reduced levels of beneficial compounds such as β-cryptoxanthin, which is commonly found in fruits and vegetables.
The results revealed that ultra-processed diets leave an observable metabolic signature. The poly-metabolite score could be a feasible, practical, and objective metric for future nutrition research. For instance, high levels of levoglucosan in urine were linked to exposure to chemicals from food packaging. Metabolic pathway analysis indicated that UPF intake disrupts xenobiotic metabolism and interferes with amino acid, lipid, carbohydrate, and cellular energy processes, suggesting UPFs cause widespread metabolic problems.
To make these results practical, the researchers calculated poly-metabolite scores using data from 28 serum, 33 24-hour urine, and 23 FMU metabolites. These scores were moderately to strongly correlated with actual levels of UPF in the diet (r ≥ 0.47), and they successfully classified different dietary patterns based on serum, urine tests, and food records, with area under curve (AUC) values ranging from 0.66 to 0.68.Â
Although the social diversity was limited and there were some inconsistencies between dietary recall and biological sample data, it employed rigorous sampling and robust statistical methods. This study represents a significant improvement in nutrition research by providing a reproducible and reliable method of measuring UPF consumption in the general population. Further studies will be necessary to confirm its validity in diverse populations and to explore further the connection between UPF consumption and chronic disease risk.
References:
- National Institutes of Health. NIH researchers develop biomarker scores to predict diets high in ultra-processed foods. Published May 20, 2025. Accessed May 21, 2025. https://www.nih.gov/news-events/news-releases/nih-researchers-develop-biomarker-score-predicting-diets-high-ultra-processed-foods
- Abar L, MartĂnez Steele E, Lee SK, et al. Identification and validation of poly-metabolite scores for diets high in ultra-processed food: An observational study and post-hoc randomized controlled crossover-feeding trial. PLoS Med. 2025;22(5):e1004560. doi:10.1371/journal.pmed.1004560
