When I first encountered the NOVA classification framework, what struck me was its departure from nutrient-based thinking. For decades, nutrition science focused on macronutrients and micronutrients in isolation. NOVA, developed by Carlos Monteiro and colleagues at the University of São Paulo, organizes food by the degree and purpose of industrial processing rather than by nutrient content. This reframing turns out to have substantial predictive value for health outcomes.
The NOVA Framework
Monteiro et al. (2019, Public Health Nutrition) formalized the NOVA classification into four groups. Group 1 is unprocessed or minimally processed foods: whole fruits, vegetables, legumes, eggs, meat, fish, plain milk, and plain yogurt — consumed in forms close to their natural state. Group 2 is processed culinary ingredients: oils, butter, sugar, flour, and salt — substances extracted from Group 1 foods and used in cooking but not typically eaten alone. Group 3 covers processed foods: canned fish, simple cheeses, cured meats, and fermented vegetables — foods made by adding salt, sugar, oil, or vinegar to Group 1 foods for preservation or palatability, with short recognizable ingredient lists.
Group 4 — ultra-processed food and drink products — is where the classification becomes most clinically significant. These are industrial formulations made largely from substances derived from foods (hydrolyzed proteins, modified starches, hydrogenated fats) combined with additives that have no culinary equivalent: emulsifiers, flavor enhancers, colorants, artificial sweeteners, and stabilizers. Examples include packaged snack foods, breakfast cereals with long ingredient lists, flavored yogurts, reconstituted meat products, carbonated soft drinks, and most packaged breads containing more than five ingredients. The key indicator is not a single ingredient but the combination of industrial processes and additive classes not found in home cooking.
The Hall et al. RCT
The most compelling experimental evidence for the dietary impact of ultra-processed foods came from Kevin Hall and colleagues at the NIH, published in Cell Metabolism in 2019. Twenty adults were admitted to a clinical research unit and randomly assigned in a crossover design to two 2-week dietary conditions: one consisting of ultra-processed foods, one of unprocessed foods. Both diets were matched for total calories, sugar, fat, fiber, and macronutrient ratios on paper. Participants were told to eat as much or as little as they wanted.
During the ultra-processed phase, participants consumed on average approximately 508 more calories per day and gained roughly 0.9 kg. During the unprocessed phase, they spontaneously reduced intake and lost weight. Eating rate was measurably faster during the ultra-processed condition, suggesting these foods do not activate satiety signaling normally. What I find significant about this study is that it was an inpatient RCT — not a dietary recall survey — which eliminates many of the confounders that plague observational nutrition research.
Why Ultra-Processed Food Is Difficult to Stop Eating
Several mechanisms have been proposed. First, hyper-palatable formulation: food scientists engineer combinations of fat, sugar, and salt in ratios not found in nature, which appear to engage the brain’s reward circuitry in ways whole foods do not replicate. Second, fiber displacement: UPFs typically contain little intact fiber, so the gastric emptying-slowing and satiety hormone-stimulating (GLP-1, PYY) effects of fiber are absent. Third, faster eating rate, as documented in the Hall study, means calories arrive in circulation before satiety signals can catch up. Fourth, emerging animal model and preliminary human data suggests certain emulsifiers — including carboxymethylcellulose and polysorbate 80 — may alter gut mucosal integrity and microbiome composition, though this work requires replication in human intervention trials before drawing strong conclusions.
Practical Identification and Reduction
Identifying ultra-processed foods requires reading ingredient lists rather than relying on front-of-package claims. Reliable indicators of Group 4 classification include: modified starches, hydrolyzed vegetable protein, maltodextrin, high-fructose corn syrup, artificial or natural flavors (listed as such), emulsifiers (lecithin, mono- and diglycerides, carrageenan, guar gum in combination with other additives), artificial colorants, and synthetic preservatives like sodium benzoate or BHT. A working heuristic: if the ingredient list contains substances you would not find in a home kitchen, the product is likely Group 4.
Practical reduction does not require eliminating all packaged food. Cook from whole ingredients most of the time; when buying packaged items, prioritize those with five or fewer recognizable ingredients; replace ultra-processed snacks with whole-food equivalents such as whole fruit, unsalted nuts, or plain yogurt; and reduce reliance on packaged breads, flavored cereals, and sweetened beverages. The dose-response relationship between UPF consumption and adverse health outcomes is still being quantified, but the direction of evidence is consistent.
Not medical advice. Content is informational only. Consult a qualified healthcare provider before making changes to your health regimen.
Leave a Reply