A recent study published in Plos Medicine shows that molecules in urine and blood can objectively reflect a person's intake of ultra-processed foods, providing a new way to study their association with diseases such as diabetes and cancer.
Ultra-processed foods are industrially produced foods that often contain additives, emulsifiers and other ingredients not commonly found in home cooking. They include sweetened yogurt, pre-made bread, packaged snacks and other types. There have been studies linking high intakes of ultra-processed foods to an increased risk of obesity, cardiovascular disease, diabetes and some cancers, but these studies mostly rely on participants' dietary recalls and have limited accuracy.
A research team from the National Cancer Institute analyzed blood and urine samples from 718 healthy individuals aged 50-74, combined with their dietary records, and used machine learning to estimate the proportion of ultra-processed foods in daily energy intake. Results showed that ultra-processed foods accounted for an average of 50% of participants' energy intake, but individual differences varied significantly (12%-82%). The diets of the high-intake group were typically high in carbohydrates, added sugars, and saturated fat, and low in protein and fiber.
Metabolite analysis found that metabolites related to the risk of type 2 diabetes were more common in samples from people who consumed more ultra-processed foods. Some urine samples also detected molecules related to food packaging, while metabolites from fresh fruits and vegetables were less common. The research team further verified the reliability of the metabolite detection through a controlled experiment: 20 subjects alternated between ultra-processed and non-ultra-processed diets for two weeks, and the metabolite characteristics could accurately distinguish the two dietary patterns.
Experts point out that traditional diet assessment methods have large errors, and this technology is expected to reveal the specific mechanisms by which ultra-processed foods harm health. It is currently difficult to distinguish the metabolic differences between ultra-processed foods and diets high in salt, sugar, and fat, but understanding this could lead to improved formulations by the food industry.