Untargeted metabolomics-based approach to discover biomarkers of apple intake: A randomized controlled trial.

title: "Urinary Biomarkers of Apple Intake: A Metabolomics Trial Identifies a Four-Metabolite Panel"

description: "A randomized crossover trial in 18 adults identifies urinary metabolites that track apple consumption for up to 24 hours, and a four-metabolite panel that distinguishes recent apple eaters."

date: "2025"

category: "Nutrition Research"

evidenceTier: "B"

studyType: "Randomized controlled trial (crossover)"

TL;DR

A small randomized crossover trial found five urinary metabolites that remain elevated for at least 24 hours after apple intake, and a four-metabolite panel that identified recent apple consumers with moderate accuracy (AUC 0.80). This is a single small study, so the findings are promising for dietary assessment but not yet validated in larger or more diverse populations.

What the study found

Dietary assessment has a long-standing problem: people misremember and misreport what they eat. Food questionnaires and diaries are cheap and common, but they are imprecise. Biomarkers of food intake, which are compounds in blood or urine that appear after a specific food is eaten, offer an objective alternative. For most foods, however, suitable biomarkers are scarce, and existing apple biomarkers tend to clear the body quickly, which limits their usefulness outside a controlled setting.

The research team, publishing in *Food Research International*, set out to find urinary biomarkers of apple intake with a longer detection window and to combine them into a panel that could work in everyday life. The trial enrolled 18 healthy adults in a randomized, controlled, crossover design. After a two-day standardized run-in period, participants went through single-exposure, repeat-exposure, and follow-up phases. Urine was collected at multiple time points, and the team analyzed it with untargeted metabolomics using dual-column ultra-high-performance liquid chromatography-mass spectrometry. The researchers also analyzed the original apple and in vitro digested apple samples to help confirm metabolite origins.

Using a statistical pipeline built on the shape of each metabolite's postprandial response over time, the researchers identified five candidate urinary biomarkers that stayed significantly elevated at 24 hours. A second, complementary screen focused on the first six hours after eating and identified 28 early-responsive metabolites. Among these was phloretin-2'-O-glucuronide, a compound that had been reported as an apple biomarker in earlier work, which lends some internal consistency to the method.

The most practically relevant result concerns the panel. Individual metabolites performed poorly at predicting apple intake under free-living conditions, meaning everyday life outside the lab. When the researchers combined markers, a four-metabolite panel built with XGBoost machine learning models achieved the best discrimination among the panels tested, with an AUC of 0.80 for identifying apple consumption within the prior one to two days. An AUC of 0.80 indicates useful but imperfect discrimination: the panel would correctly rank a recent apple eater above a non-eater about 80 percent of the time, while still producing errors.

These results should be read with the study's limitations in mind. The sample size was 18, the population was healthy adults, and the intervention used a defined quantity of apple under controlled conditions. The sample size was not clearly reported in the metadata, so the figure above reflects the abstract. Validation in larger, more diverse cohorts, and in people eating apples as part of mixed diets, is needed before these biomarkers could be used for routine dietary assessment. The study also does not establish whether the metabolites reflect apple-specific compounds in every case, or whether they could arise from other polyphenol-rich foods.

What this means for you

For most readers, this study does not change any personal behavior, and it is not a reason to change how you eat. Its significance lies in research methods. Objective markers of what people eat could make nutrition science more reliable, because diet is one of the hardest exposures to measure accurately.

If you are interested in apples and general health, the broader evidence on whole fruit intake is more relevant than any single biomarker study. Diets rich in whole plant foods are consistently associated with favorable outcomes in observational research, though observational findings cannot establish cause. You can explore related dietary patterns through high-fiber diet and mindful eating, which address how food choices fit into daily routines.

For researchers and clinicians, the practical takeaway is that a multi-marker approach outperformed single markers here, which suggests that dietary biomarker work will likely rely on panels rather than individual compounds.

Study details

Sources

1. Untargeted metabolomics-based approach to discover biomarkers of apple intake: A randomized controlled trial. *Food Research International*. PubMed. https://pubmed.ncbi.nlm.nih.gov/42914471/