Ultra-Processed Foods: Why the NOVA Classification Matters

In 2009, a Brazilian nutritionist named Carlos Monteiro proposed a radical idea: the degree of processing matters more than individual nutrients. It was not a fringe theory published in an obscure journal. Monteiro, a professor of nutrition and public health at the University of Sao Paulo, laid out a classification system that would fundamentally reshape how researchers, governments, and eventually consumers think about what they eat.

He called it NOVA. Not an acronym — just a name meaning “new” in Portuguese. And the core insight was deceptively simple: we had spent decades obsessing over fat, sugar, sodium, and calories while ignoring the elephant in the room. The industrial transformation of food itself — the processing — was driving the obesity and chronic disease epidemics in ways that nutrient panels could not capture.

Seventeen years later, NOVA has been adopted by the United Nations Food and Agriculture Organization, cited in over 14,000 peer-reviewed papers, and used as the basis for national dietary guidelines in multiple countries. The food industry, predictably, hates it.

The Four NOVA Groups

NOVA classifies all foods into four groups based on the extent and purpose of their processing.

Group 1: Unprocessed or minimally processed foods. These are foods altered only by processes like drying, pasteurizing, freezing, or fermenting — things that do not add new substances. Fresh fruit. Eggs. Plain milk. Dried beans. Frozen vegetables. Cuts of meat. Coffee beans. The food is still recognizable as food.

Group 2: Processed culinary ingredients. These are substances extracted from Group 1 foods and used in cooking: oils pressed from seeds or olives, butter churned from cream, sugar milled from cane, salt mined from the earth, flour ground from wheat. You would not eat these on their own, but you use them to prepare Group 1 foods. Your grandmother had all of these in her kitchen.

Group 3: Processed foods. Group 1 foods modified by Group 2 ingredients using simple methods like canning, bottling, or baking. Canned vegetables in brine. Cheese. Freshly baked bread. Salted or smoked meats. Canned fish in oil. The food is still recognizable, but it has been deliberately altered for preservation or taste. These typically have two or three ingredients, and you could make them at home.

Group 4: Ultra-processed foods. This is where things change. Ultra-processed foods (UPFs) are industrial formulations made mostly or entirely from substances derived from foods and additives. They contain ingredients you would never find in a home kitchen: hydrogenated oils, modified starches, hydrolyzed protein, maltodextrin, invert sugar, high-fructose corn syrup. They use cosmetic additives — flavor enhancers, emulsifiers, humectants, sequestrants, glazing agents — designed to make the product hyper-palatable and shelf-stable.

Think soft drinks, packaged snacks, instant noodles, reconstituted meat products, frozen pizza, sweetened breakfast cereals, and most fast food. The distinguishing feature is not any single ingredient but the overall formulation: these products are engineered to be convenient, cheap, highly palatable, and profitable. They are designed to be eaten quickly and in large quantities.

The key question Monteiro raised was whether this industrial formulation — independent of the nutrient profile — was itself a driver of poor health outcomes. In 2019, a researcher at the National Institutes of Health answered that question definitively.

The Hall Study: 20 People Who Changed the Debate

Kevin Hall is a senior investigator at the National Institute of Diabetes and Digestive and Kidney Diseases. In May 2019, he published a study in Cell Metabolism that became the single most important piece of evidence in the ultra-processed food debate.

The design was elegant in its simplicity. Twenty adults were admitted to the NIH Clinical Center’s metabolic ward — a controlled environment where every bite of food could be measured. Each participant spent two weeks eating an ultra-processed diet, and two weeks eating an unprocessed diet, in randomized order. Both diets were matched for total calories presented, macronutrient composition, sugar, sodium, fat, and fiber. Participants were told to eat as much or as little as they wanted.

The results were striking. On the ultra-processed diet, participants consumed an average of 508 more calories per day. They ate faster. They gained about 0.9 kg (roughly 2 lbs) over the two-week period. On the unprocessed diet, they lost about the same amount. The differences appeared within the first few days and persisted throughout.

This was the first randomized controlled trial to demonstrate that ultra-processed foods drive overconsumption independent of the nutritional composition. It was not the sugar. It was not the fat. It was not the salt. Something about the ultra-processed formulation itself — the texture, the speed at which it could be eaten, the combination of additives — caused people to eat more.

Hall’s study did not prove why UPFs drove overconsumption, but it ruled out the most common defense the food industry offers: that their products are fine as long as the nutrient numbers look acceptable on the label.

60% of American Calories

The Hall study was a controlled experiment. The population-level data tells an even more alarming story.

According to analyses of NHANES (National Health and Nutrition Examination Survey) data, ultra-processed foods account for approximately 58-60% of total calorie intake among US adults. For adolescents, the figure is even higher — approaching 67%. The United Kingdom is not far behind, with estimates around 57% of calories coming from UPFs.

Brazil, ironically the country that gave us NOVA, has significantly lower UPF consumption at roughly 20% of calories. This is not a coincidence. Brazil was the first country to incorporate UPF warnings into its national dietary guidelines, back in 2014. Its population still relies heavily on rice, beans, and fresh produce prepared at home.

The epidemiological evidence linking high UPF consumption to adverse health outcomes has grown enormous. A 2024 umbrella review published in The BMJ, led by researchers at Deakin University and Johns Hopkins, pooled 45 meta-analyses encompassing millions of participants. The findings were consistent across studies and continents: higher UPF consumption was associated with 32 adverse health outcomes, including increased risk of cardiovascular disease, type 2 diabetes, anxiety, depression, all-cause mortality, and certain cancers. The associations were dose-dependent — more UPF, more risk.

France, Belgium, and Australia have all initiated major research programs examining UPF consumption in their populations. The concern is no longer limited to nutrition academics. It has become a mainstream public health issue.

The Additive Question

One of the most debated questions in UPF research is why these foods are harmful. Is it the additives? The disrupted food matrix? The speed of consumption? The answer, increasingly, appears to be all of the above.

On the additive front, specific ingredients common in UPFs have shown concerning effects in controlled studies. A 2015 study published in Nature by Chassaing et al. found that two common emulsifiers — polysorbate 80 and carboxymethylcellulose — altered the gut microbiome composition in mice, promoted intestinal inflammation, and increased metabolic syndrome markers. These emulsifiers are in hundreds of everyday products: ice cream, salad dressings, plant-based milks, baked goods.

Artificial sweeteners, another UPF hallmark, have been shown to alter gut bacteria in human trials. A 2022 study in Cell by Suez et al. demonstrated that saccharin, sucralose, aspartame, and stevia all affected the human gut microbiome, with saccharin and sucralose impairing glycemic responses.

But additives are only part of the picture. UPFs are also characterized by what researchers call food matrix disruption — the physical structure of the food has been broken down and reconstituted in ways that change how the body processes it. Whole grains are ground into fine flour. Intact cells are ruptured. Fiber is removed and sometimes added back as isolated powders. The result is food that requires less chewing, passes through the stomach faster, and triggers weaker satiety signals.

This is where the Hall study becomes especially relevant. Even when the nutrients were matched, the ultra-processed meals were eaten faster — about 17 grams per minute compared to 10 grams per minute for unprocessed meals. The soft, pre-chewed texture of UPFs may bypass the body’s natural calorie-regulation mechanisms. You finish eating before your gut has had time to tell your brain you are full.

The additives and the processing are not separate problems. They are intertwined features of the same industrial food system.

Policy Responses Around the World

Governments have responded to the UPF evidence at very different speeds.

Brazil (2014) was the pioneer. Its national dietary guidelines, updated with direct input from Monteiro’s research group, include the recommendation: “Avoid ultra-processed foods.” The guidelines emphasize cooking at home with fresh ingredients and treating processed foods as small additions to meals based on unprocessed foods. It remains the gold standard for food policy that acknowledges processing as a risk factor.

France has been exploring whether its Nutri-Score front-of-pack labeling system should incorporate a UPF adjustment. Currently, Nutri-Score evaluates nutrients per 100g but does not penalize for ultra-processing. A reformulated Nutri-Score that accounts for NOVA classification has been proposed but not yet adopted.

The United Kingdom commissioned its Scientific Advisory Committee on Nutrition (SACN) to conduct a formal review of UPF evidence, which reported in 2023 that the evidence was “broadly consistent” in linking higher UPF consumption to adverse health outcomes, while noting methodological limitations.

Colombia and Israel have implemented front-of-pack warning labels — black octagonal stamps alerting consumers when products are high in sugar, sodium, or saturated fat. While not NOVA-based specifically, these labels disproportionately flag ultra-processed products and have been shown to shift purchasing behavior.

The United States has no UPF-specific policy. The Dietary Guidelines for Americans do not reference NOVA or ultra-processing. The FDA’s regulatory framework evaluates individual additives through the GRAS (Generally Recognized as Safe) system but does not consider the cumulative effect of multiple additives in a single ultra-processed product, nor the processing itself.

What NOVA Gets Right — and Wrong

NOVA is not a perfect system. Critics raise legitimate points. Under strict NOVA classification, artisan sourdough bread made with flour, water, salt, and a natural starter is a Group 3 processed food, while mass-produced white bread with added emulsifiers and preservatives is Group 4 ultra-processed — even though a consumer might reasonably consider them equivalent. Plant-based meat alternatives, which many people choose for environmental reasons, are almost universally Group 4. Infant formula is ultra-processed by definition, yet it is a lifesaving product for millions of babies.

These edge cases matter. NOVA was designed as an epidemiological research tool, not a consumer-facing scoring system. It captures broad patterns about industrial food production, not nuances about individual products.

But what NOVA gets right is more important than what it gets wrong. It captures something that nutrient panels fundamentally miss: the industrial transformation of food into products engineered to override our natural appetite regulation. It explains why two foods with identical calorie counts can have dramatically different effects on how much we eat and how our bodies respond. And it has generated a body of evidence — the Hall RCT, the BMJ umbrella review, thousands of cohort studies — that is increasingly difficult to dismiss.

The next time you pick up a product with a long ingredient list full of substances you do not recognize, you are looking at NOVA Group 4. Understanding what that means — and what the research says about eating a diet dominated by those products — is one of the most impactful things you can do for your long-term health.

GradeMyLabel scores the individual additives that make a product ultra-processed — the emulsifiers, the modified starches, the artificial sweeteners, the flavor enhancers — giving you a concrete, ingredient-level view of what NOVA captures at the category level. Because knowing a food is ultra-processed is the first step. Knowing why it is ultra-processed, and which specific additives carry the most evidence of concern, is where actionable decisions begin.