Ultra-Processed Foods and the Microbiome: The Strongest Evidence So Far

Ultra-processed foods make up roughly 58% of total caloric intake in the United States and about 57% in the United Kingdom, according to national dietary survey data. In the last decade, a body of research linking UPF consumption to adverse health outcomes has grown from a handful of observational studies to a substantial, multi-national evidence base. The gut microbiome sits at the center of much of this research, and the findings are remarkably consistent across study designs and populations. But consistency of association is not the same as established causation, and the mechanistic questions, the 'why,' remain genuinely open. Here is where the evidence stands.

What is the NOVA classification system and why does it matter?

Before looking at the microbiome data, it helps to understand how researchers define 'ultra-processed.' The NOVA classification system, developed by Carlos Monteiro and colleagues at the University of Sao Paulo, divides all foods into four groups based on the extent and purpose of industrial processing.

The NOVA system has critics. Some nutrition researchers argue that it is too broad, grouping a whole-grain bread with added emulsifiers in the same category as candy. Others contend that it stigmatizes all food processing, which includes beneficial processes like fermentation and pasteurization. These criticisms have merit. But NOVA remains the most widely used framework in UPF research, and most of the major studies discussed here use it.

What did the PREDICT study find about UPFs and gut bacteria?

The PREDICT (Personalised Responses to Dietary Composition Trial) study, led by Tim Spector and colleagues at King's College London and published in Nature Medicine in 2021, is one of the largest studies linking diet to the gut microbiome. It enrolled over 1,100 participants from the U.S. and U.K. and combined detailed dietary data with gut microbiome sequencing.

The study identified 15 gut microbial species associated with favorable health markers (lower inflammation, better metabolic health, better cardiovascular markers) and 15 species associated with unfavorable health markers. The researchers then examined how dietary patterns correlated with the abundance of these species.

Higher intake of minimally processed plant foods was associated with greater abundance of the 'favorable' species, including Prevotella copri and several Firmicutes involved in butyrate production. Higher UPF intake showed the opposite pattern: more of the unfavorable species and fewer of the favorable ones. UPF consumption was also associated with lower overall microbial diversity.

What did the Hall et al. NIH crossover trial show?

The Hall et al. study, published in Cell Metabolism in 2019, is probably the single most important piece of experimental evidence in the UPF debate. Kevin Hall and colleagues at the NIH conducted a randomized, controlled, crossover feeding trial with 20 inpatient participants. Each person spent two weeks on an ultra-processed diet and two weeks on an unprocessed diet, in randomized order.

The diets were carefully matched for total calories, macronutrients (fat, carbohydrate, protein), sugar, sodium, and fiber (fiber was added to beverages in the UPF condition because achieving matched fiber from ultra-processed food sources alone was not possible). Participants could eat as much or as little as they wanted.

The results were striking. On the ultra-processed diet, participants ate approximately 508 more calories per day on average and gained 0.9 kilograms over two weeks. On the unprocessed diet, they lost 0.9 kilograms. The participants reported no differences in hunger, fullness, or satisfaction between the diets. They simply ate more when the food was ultra-processed.

This study did not specifically measure microbiome changes, but it demonstrated something important: UPFs alter eating behavior in ways that calorie and macronutrient matching cannot explain. Something about the food matrix, the texture, the rate of consumption, or the metabolic signaling of UPFs drives excess intake. Whether microbial changes are part of this loop is an active area of investigation.

Do emulsifiers in ultra-processed food directly harm the gut?

This is where the mechanistic research gets specific. Benoit Chassaing and Andrew Gewirtz at Georgia State University have published a series of studies examining the effects of two common food emulsifiers, polysorbate-80 (P80) and carboxymethylcellulose (CMC), on the gut.

Their landmark 2015 study, published in Nature, found that mice consuming P80 or CMC at concentrations reflecting typical human dietary exposure developed low-grade intestinal inflammation, altered microbiome composition, erosion of the protective mucus layer that separates gut bacteria from the intestinal epithelium, and, in genetically susceptible mice, overt colitis. The researchers traced the mechanism to the emulsifiers' ability to thin the mucus layer, which allowed bacteria to make closer contact with the gut lining and trigger immune activation.

A follow-up 2017 study by the same group, published in Gut, used a synthetic human gut microbiome transplanted into germ-free mice and confirmed that the emulsifier-induced changes were microbiome-dependent. The emulsifiers did not just affect the host directly; they changed the microbial community, and the altered microbial community drove the inflammation.

In 2022, Chassaing's group published the first randomized controlled trial of dietary emulsifiers in humans, in Gastroenterology. The study was small (16 participants per group) and short (11 days), but it found that CMC consumption altered the fecal microbiome and metabolome compared to a CMC-free diet. The clinical significance of these short-term changes is not yet clear, but the direction was consistent with the animal findings.

Is it the additives, the low fiber, or the whole dietary pattern?

This is the central unsolved question in UPF-microbiome research. There are at least four plausible mechanisms, and they are not mutually exclusive.

First, specific additives like emulsifiers, artificial sweeteners, and preservatives may directly alter the gut environment. The Chassaing emulsifier work supports this. A 2022 Cell study by Suez et al. showed that saccharin and sucralose altered the gut microbiome in healthy humans within two weeks. These are direct chemical effects on microbial communities.

Second, fiber displacement is probably a major factor. UPFs are typically very low in fiber. The average American consumes only about 16 grams of fiber per day, compared to a recommended 25 to 38 grams. Fiber is the primary substrate for short-chain fatty acid (SCFA) production by colonic bacteria. Without adequate fiber, butyrate-producing bacteria like Faecalibacterium prausnitzii decline, and overall microbial diversity drops. This effect alone could explain much of the observed microbiome changes in high-UPF diets.

Third, the displacement of whole foods matters beyond fiber. Polyphenols from fruits, vegetables, coffee, tea, and dark chocolate are metabolized by gut bacteria and support microbial diversity. A diet high in UPFs is typically low in polyphenols.

Fourth, the food matrix itself may play a role. Ultra-processing changes the physical structure of food in ways that alter how it interacts with the GI tract. Whole grains require more mechanical breakdown than refined flour, which changes where in the gut nutrients become available and which bacteria can access them. The Hall et al. study suggested that something about the UPF food matrix (beyond macronutrient content) drives excess consumption. Whether a similar food-matrix effect exists for the microbiome is an open question.

What does the NutriNet-Sante cohort tell us about UPFs and health outcomes?

The NutriNet-Sante cohort, based in France, is one of the largest ongoing prospective studies examining the relationship between diet and health. With over 100,000 participants and extensive dietary questionnaire data, it has produced several of the most cited UPF studies.

A 2018 study by Fiolet et al. in the BMJ found that a 10% increase in the proportion of UPFs in the diet was associated with a 12% higher risk of overall cancer (hazard ratio 1.12, 95% CI 1.06-1.18). A 2019 study by Schnabel et al. in JAMA Internal Medicine, also from the NutriNet-Sante cohort, found that a 10% increase in UPF consumption was associated with a 14% higher risk of all-cause mortality. Both analyses adjusted for known confounders including caloric intake, smoking, physical activity, and BMI.

These are association studies and cannot prove causation. People who eat more UPFs may differ in unmeasured ways from people who eat fewer. But the consistency of the findings across different outcomes (cancer, mortality, cardiovascular disease, depression, inflammatory bowel disease) and different cohorts in different countries makes the signal hard to dismiss as confounding alone.

What practical steps actually make a difference?

If you are looking at this evidence and wondering what to actually do, here is a proportionate response. You do not need to eliminate all processed food. You do not need to make everything from scratch. You need to shift the ratio.

If you are making dietary changes and want to see whether they correlate with changes in your digestive symptoms, tools like GLP1Gut can help you track food choices alongside symptom patterns over time, giving you personal data rather than one-size-fits-all recommendations.

The bottom line on UPFs and the microbiome

The evidence that high UPF intake is associated with unfavorable microbiome changes is consistent and growing. The mechanistic story, particularly around emulsifiers and fiber displacement, is biologically plausible and supported by animal and limited human data. But we are not yet at the point where we can say 'this specific additive causes this specific gut problem at this specific dose in humans.'

What we can say is that the overall direction of evidence supports eating more whole and minimally processed foods and fewer ultra-processed ones, particularly those containing emulsifiers and artificial sweeteners. This is not a radical position. It is consistent with every major dietary guideline, including the 2025-2030 U.S. Dietary Guidelines, which for the first time explicitly address ultra-processed food as a category to limit.

**Disclaimer:** This article is for informational purposes only and does not constitute medical advice. Always consult with a qualified healthcare provider about your specific health concerns.