The ultra-processed food problem

TL;DR. Ultra-processed food (UPF) is a precise category called NOVA Group 4. Carlos Monteiro set the rules in 2009. The test is simple: does the food contain industrial ingredients no home cook owns? Kevin Hall ran the proof in 2019 at the NIH. He gave 20 adults two diets with matched calories, sugar, salt, fat, and fiber. On the UPF diet, they ate 500 extra calories a day and gained about a pound a week. Around 250 large studies show the same pattern. Every 10% jump in UPF intake raises disease risk by 10 to 25%. UPF works through four levers. It destroys the food matrix. It stacks salt, sugar, and fat to bypass fullness signals. It uses fake sweeteners that confuse the body. And it damages the gut microbiome with emulsifiers. Reformulation cannot fix this. "Lower sugar Cheerios" is still NOVA 4 because the industrial ingredients stay. Your goal: cut UPF from 50 to 60% of your diet down to under 30%. Start with sweet drinks, snacks, and sauces.

What you'll learn

• The test that separates NOVA Group 3 from Group 4.
• What Hall's 2019 trial proved and why it matters so much.
• The 250 studies showing dose-dependent harm.
• The four ways UPF tricks your appetite.
• Why reformulation cannot fix the problem.
• How to drop UPF below 30% of your diet.

NOVA in depth

Carlos Monteiro published the NOVA system in 2009 in Public Health Nutrition (DOI 10.1017/S1368980009005291). He sorted food by how it gets made, not by its nutrients. This is why a homemade pizza and a frozen pizza can look the same on a label and act very different in your body. NOVA has four groups.

• Group 1, unprocessed or minimally processed. Whole foods you can eat as is. Apples, brown rice, lentils, eggs, plain yogurt, frozen broccoli, ground beef, rolled oats.
• Group 2, cooking ingredients. Stuff you pull from Group 1 or nature to cook with. Oils, butter, salt, sugar, vinegar, honey.
• Group 3, processed foods. Group 1 foods preserved with Group 2 ingredients using methods a home cook knows. Real bread, tinned tomatoes, smoked fish, aged cheese, cured ham, sauerkraut, olives in brine.
• Group 4, ultra-processed food. Industrial mixes built from extracted substances (RBD seed oils, modified starches, hydrolyzed proteins, glucose syrups, maltodextrin, invert sugars) or lab chemicals (artificial flavors, low-calorie sweeteners). They use additives a home cook never buys: emulsifiers, gums, stabilizers, flavorings, colors.

The line between Group 3 and Group 4 is the line that counts. A real sourdough and a "soft whole grain" supermarket loaf have similar macros. But the supermarket loaf has DATEM, mono- and diglycerides, modified wheat starch, dough conditioners, and added gluten. You cannot buy those at a grocery store.

Chris van Tulleken sums up the test this way. If it has an ingredient you would not keep in your kitchen, it is UPF. That means modified starches, RBD oils, protein isolates, emulsifiers, gums, flavorings, or low-calorie sweeteners. One such ingredient is enough. Organic Heinz baked beans are NOVA 3. The standard kind is NOVA 4 because of modified cornflour and spice extracts.

Robert Lustig reaches the same line from biochemistry in Metabolical. UPF means missing fiber and added sugar, plus seed oils, emulsifiers, nitrates, and toxins from high-heat cooking. Real food protects your liver and feeds your gut. UPF does neither. Marion Nestle calls UPF "the right unit of analysis." Research and policy should track UPF, not chase one nutrient at a time.

NOVA gets fuzzy at the edges. That is fine. So does the line between "alcoholic" and "social drinker." A model is useful when it predicts outcomes, and NOVA does. It predicts deaths, heart disease, cancer, type 2 diabetes, dementia, and IBD across many cohorts, even after you adjust for nutrients.

The Hall 2019 trial

The strongest cause-and-effect study in nutrition is Kevin Hall's 2019 trial at the NIH Clinical Center, published in Cell Metabolism (DOI 10.1016/j.cmet.2019.05.008). Hall is a physicist. He was skeptical of NOVA and built the trial to disprove it.

He used 20 adults on the NIH metabolic ward for 4 weeks. Each person ate two diets in random order, 2 weeks each. One was 80% UPF by calories. The other was 80% NOVA Group 1 to 3. He matched the diets on calories offered, sugar, salt, fat, fiber, and macros. Subjects ate as much as they wanted, with 3 meals and snacks each day. They had about twice the food they could finish.

On the UPF arm, subjects ate around 500 extra calories per day. They gained about a pound a week. On the whole-food arm, the same people lost weight. The gap showed up by day 3. Hunger and fullness scores were similar on both diets. Subjects did not feel hungrier. They ate faster and ate more without noticing.

This rules out the usual pushback. The same people ate both diets, so income and lifestyle do not explain it. The diets matched on sugar, salt, fat, and fiber. If those were the key variables, intake should have matched too. It did not. Processing itself drove the gap. Hall now treats UPF as the main variable in his lab.

The four engineering levers

Hall's trial proves that UPF drives overeating. The next question is how. Van Tulleken lays out four levers. No single lever does all the work. All four together make a modern UPF product.

Lever 1, destroyed food matrix. Anthony Fardet's work on the food matrix shows that physical structure matters as much as nutrients. The way food is built shapes how your gut and brain respond to it. A 1977 study compared whole apples, apple puree, and apple juice. The whole apple gave a slow blood-sugar rise and lasting fullness. The juice and puree caused spikes and crashes. UPF grinds the matrix down and rebuilds it shelf-stable. The result is soft, dry, and packed with calories. On Hall's UPF arm, subjects ate about 17 more calories per minute than on the whole-food arm.

Lever 2, salt, sugar, and fat stacked together. Howard Moskowitz's 1970s industry work found the "bliss point." That is the ratio that makes a food taste best. Real foods rarely stack sugar, salt, and fat at high levels. Steak gives you salt and fat. An apple gives you sugar and fiber. UPF stacks them. Pringles deliver salt, fat, modified starch, and umami boosters like glutamate, guanylate, and inosinate. Those signal aged protein the chip does not contain. Cola hides 9 teaspoons of sugar behind sour phosphoric acid, bitter caffeine, cold, and fizz. Warm flat cola tastes awful. That is your appetite control working as it should.

Lever 3, mixed-up taste signals. Dana Small at Yale showed that you learn to crave flavors paired with calories. Your body builds expectations from what your tongue tastes. Low-calorie sweeteners break that contract. They give sweetness with no energy. Allison Sylvetsky found that kids drinking diet sodas eat more total calories than kids drinking water. A 2014 Nature paper by Suez and colleagues showed that sucralose, aspartame, and saccharin changed the mouse gut microbiome and raised blood sugar. The effect even transferred through fecal transplant. Sweeteners are not free.

Lever 4, gut damage. Benoit Chassaing and Andrew Gewirtz published a 2015 paper in Nature (DOI 10.1038/nature14232). At doses humans eat, two common emulsifiers (carboxymethylcellulose, E466, and polysorbate 80, E433) thinned the gut's protective mucus. Bacteria pushed up against the gut lining. The mice got inflammation and metabolic syndrome. Some got colitis. Tim Spector ran a 10-day fast-food test on his son Tom. Tom lost around 40% of his detectable gut microbial species. The loss stuck for years. Xanthan gum has spread through the guts of billions of UPF eaters and brought a new species absent from hunter-gatherers (Ostrowski, Michigan). Trehalose is linked to bad C. difficile outbreaks. Your gut microbes make 50,000 chemicals and run much of your immune system. UPF rewires that system in weeks.

The epidemiology

Around 250 large studies now track UPF intake against health outcomes. The wider literature has 2,000+ related papers. The pattern is dose-dependent. The more UPF you eat, the worse your odds. Sam Dicken's review (from Rachel Batterham's group at UCL) lists numbers like these:

• All-cause death: +10% UPF → about +15% risk
• Breast cancer: +10% UPF → about +10% risk
• Type 2 diabetes: +10% UPF → about +15% risk
• Dementia: +10% UPF → about +25% risk
• Heart disease, depression, IBD, and frailty: all rise the same way

These effects hold up after adjusting for nutrients, dietary patterns, smoking, drinking, exercise, income, and education. They are real, not an artifact of UPF eaters being poorer or less healthy. The harm is also not just about getting fat. Heart disease, dementia, IBD, and depression rise with UPF intake even in people who stay at the same weight. UPF hurts you through inflammation, gut damage, and metabolic problems on its own. Even if you stay thin, UPF is still doing damage.

Why reformulation doesn't fix it

For 40 years, the food industry's go-to response has been reformulation. Lower the sugar. Lower the sodium. Add fiber. Add protein. Add vitamins. Lower-sugar Cheerios. "Healthier" Coca-Cola with stevia. Plant-based meat with pea protein isolate.

This cannot fix the category, because NOVA Group 4 is set by ingredients, not macros.

A cereal of whole oats, water, and salt is NOVA 1. Drop the sugar 30% and add maltodextrin, soy protein isolate, natural flavorings, modified corn starch, tocopherols, BHT, and "color from vegetable juice." Now it is still NOVA 4. The emulsifiers and flavorings damage the matrix and microbiome no matter what the macros say. Hall's result holds for "healthier" UPF too. Matrix destroyed. Calorie density high. Flavors mismatched. Gut-active additives present.

Reformulation usually piles on more processing. Cut the sugar and you add fake sweeteners. Cut the fat and you add gums and emulsifiers to fake the mouthfeel. "Plant-based" meat is soy protein isolate, pea protein, methylcellulose, leghemoglobin, coconut oil, and flavorings. That is extreme NOVA 4. Better climate footprint than beef, maybe. Closer to food, no.

Nestle puts it this way: reformulation legitimizes UPF and crowds out the alternative. Lower-sugar Cheerios is not progress toward food. It is the same product with better marketing. The lesson at the shelf is simple. The absence of bad nutrients tells you very little. The presence of industrial ingredients tells you a lot.

Practical action

The U.S. and U.K. pull 50 to 60% of their calories from UPF. Portugal sits at 10%. A reasonable target is under 30%. That is where the risk curve flattens. It rises fastest above 50% and gentler below 30%. You do not need to quit UPF entirely. That usually backfires. The lever is your UPF share.

Where to cut first, in order of impact:

• Sweet drinks. Soda, energy drinks, vitamin water, sports drinks, sweetened plant milks, and coffee drinks. The single biggest win. Swap in water, plain coffee, plain tea, or plain sparkling water.
• Packaged snacks. Chips, cookies, crackers, granola bars, "protein bars," candy. Lever 1 and lever 2 at their most extreme.
• Sauces and dressings. Industrial ketchup, dressings, and marinades pack in HFCS, modified starches, emulsifiers, and flavorings. Make your own with oil, vinegar, salt, mustard, and herbs.
• Breakfast cereals. Swap in rolled oats, plain yogurt with fruit, or eggs.
• Ultra-processed bread. Swap in real bread (sourdough or NOVA 3 bakery loaves) or skip it.

Keep these. They are NOVA 3 or below. Frozen vegetables and fruit (often more nutritious than out-of-season "fresh"). Tinned tomatoes, beans, and sardines. Real bread, plain whole-milk yogurt, real cheese, plain pasta, and dried legumes. Pasture-raised eggs, meat, and fish. Olive oil, vinegar, salt, and honey.

Lustig sums it up in 6 words: protect the liver, feed the gut. Real food does both. UPF does neither.

FAQ

Is sourdough UPF? A real sourdough loaf is NOVA 3. It needs flour, water, salt, starter, and time. A supermarket "sourdough" with industrial yeast, dough conditioners (DATEM, mono- and diglycerides), modified starches, and sourdough flavoring is NOVA 4. The ingredient list decides, not the label on the front.

Are plant milks UPF? Almost all store-bought plant milks are NOVA 4. They have emulsifiers (gellan gum, sunflower lecithin), stabilizers, added sugars, refined oils, and fortified vitamins. Plain 2-ingredient soy milk exists but is rare. Plant milks may suit lactose-intolerant or vegan eaters. They are still not minimally processed and not equal to dairy.

Are protein powders UPF? Yes, by definition. Whey, soy, and pea protein isolates are extracted fractions. That is exactly what NOVA 4 is built from. Most also have flavorings, sweeteners, gums, and lecithin. A scoop now and then is fine. Three shakes a day is a high-UPF diet in fitness clothing.

What about "minimally processed" claims? That phrase has no enforced meaning in the U.S. or U.K. Pret a Manger built a brand on "natural" while selling bread with DATEM, L-cysteine hydrochloride, glycerol, and ascorbic acid. Read the ingredient list, not the front of the pack.

Is bagged salad UPF? Plain bagged leaves are NOVA 1. A chlorine wash and gas flush count as preservation, not formulation. A salad kit with dressing and crispy toppings in sachets is NOVA 4 because of the dressing and toppings. Eat the leaves with your own dressing.

Beyond Burger and Impossible Burger? Extreme NOVA 4. Soy protein concentrate, coconut oil, methylcellulose, pea protein isolate, soy leghemoglobin, and natural flavorings. Their insulin spike is similar to a beef burger. They have little fiber. They may shrink a meat-eater's climate footprint. They are still not health foods. Beans, lentils, tofu, and tempeh are NOVA 1 to 3 and act very different in your body.

Why isn't NOVA in U.S. or U.K. dietary guidelines? Industry lobbying. The 2025 U.S. DGAC said no to cutting UPF because the RCTs were "too short." They do not apply that standard to other guidelines. Brazil, France, Israel, Belgium, and Uruguay use NOVA. The U.S. and U.K. do not. Adopting it would name specific products and companies.

Sources

• Monteiro, C. A. (2009). Nutrition and health. The issue is not food, nor nutrients, so much as processing. Public Health Nutrition, 12(5), 729–731. DOI: 10.1017/S1368980009005291.
• Hall, K. D., Ayuketah, A., Brychta, R., et al. (2019). Ultra-processed diets cause excess calorie intake and weight gain: an inpatient randomized controlled trial of ad libitum food intake. Cell Metabolism, 30(1), 67–77. DOI: 10.1016/j.cmet.2019.05.008.
• Chassaing, B., Koren, O., Goodrich, J. K., et al. (2015). Dietary emulsifiers impact the mouse gut microbiota promoting colitis and metabolic syndrome. Nature, 519, 92–96. DOI: 10.1038/nature14232.
• van Tulleken, C. (2023). Ultra-Processed People: The Science Behind the Food That Isn't Food. Knopf. Summary of ~250 dose-dependent epidemiology papers and the four-lever mechanism framework.
• Lustig, R. H. (2021). Metabolical: The Lure and the Lies of Processed Food, Nutrition, and Modern Medicine. Harper Wave. Operational definition of UPF and the protect-the-liver/feed-the-gut frame.
• Nestle, M. (2025). What to Eat Now. Basic Books. UPF as the right unit of analysis; reformulation critique.
• Spector, T. (2020). Spoon-Fed: Why Almost Everything We've Been Told About Food Is Wrong. Jonathan Cape. Tom Spector's ten-day fast-food experiment and the 40% microbial-diversity loss.

Related glossary

• NOVA classification
• NOVA Group 4
• Ultra-processed food
• Food matrix
• Emulsifier
• Modified starch
• Industrial seed oils