Show summary Hide summary
A new study from researchers at the National Institutes of Health presents the first blood- and urine-based test able to estimate how much ultraprocessed food a person consumes. The finding offers a more objective tool for nutrition science and could sharpen research into diet-related illnesses at a time when processed foods dominate many diets.
How the research worked
Scientists applied machine learning to metabolic data and identified hundreds of small molecules — or metabolites — in blood and urine that track with intake of industrially produced, ready-to-eat foods. From those signals they created a composite biomarker score intended to predict the share of daily calories coming from ultraprocessed foods.
San Francisco restaurants to try now: 38 can’t-miss spots
Spring party appetizers that keep hosting easy: 9 prep-ahead recipes
The work combined two data sources. Researchers first analyzed samples and diet records from 718 older adults to find metabolite patterns tied to processed-food consumption. They then tested the approach in a controlled, short-term feeding trial where 20 adults alternated two-week periods eating either a high-ultraprocessed diet or a diet free of those items.
Key results
Overall, the metabolite-based score distinguished levels of ultraprocessed food intake and reflected changes when participants switched diets. The investigators reported their results in PLOS Medicine, noting the metabolic signals spanned many biological pathways — a clue to how diet may influence health beyond simple calorie counts.
| Item | Details |
|---|---|
| Primary data | 718 participants with blood/urine samples and 12 months of diet records |
| Validation trial | 20 adults; two-week high-UPF period vs. two-week UPF-free period |
| Outcome | Hundreds of correlated metabolites; a predictive biomarker score |
| Publication | PLOS Medicine |
Why this matters now
Large nutritional studies have long relied on self-reported food diaries and questionnaires, which are vulnerable to recall error and misclassification. An objective biomarker could reduce those biases and strengthen links drawn between diet and chronic conditions such as obesity, cardiovascular disease and some cancers.
Researchers said the metabolite signals also reveal the biological complexity of dietary exposures: processed-food intake appears to affect multiple metabolic routes rather than a single pathway, which has implications for how scientists interpret associations between diet and disease.
Limitations and next steps
Investigators caution that the biomarker approach needs broader validation. The initial datasets skewed toward older adults, so performance across younger people, diverse diets and different levels of processed-food consumption remains uncertain.
Refinement will require testing in larger, more varied populations and improving the score’s accuracy across demographic groups. Only after further validation could the method be widely adopted in population studies or clinical research.
Practical takeaways for readers
- Using objective biomarkers could change how researchers measure diet and its health impacts in coming years.
- For individuals aiming to limit processed foods, focusing on nutrition facts—especially added sugars, saturated fat and sodium—remains a practical approach.
- Policy and public-health recommendations may become more targeted as biomarker-based evidence accumulates.
While more work lies ahead, this study marks a step toward more accurate, biology-based measurement of modern dietary patterns and their consequences for public health.
