It is the most consequential food category you have never seen defined on a label. Ultra-processed foods — industrial formulations of refined ingredients, additives and flavour systems, from soft drinks and packaged snacks to reconstituted meats and mass-produced breads — supply more than half of all calories eaten in the United States and United Kingdom, and a fast-rising share everywhere else. They are also, according to a mountain of epidemiology, associated with obesity, diabetes, heart disease, depression and earlier death. The argument is no longer whether the pattern exists. It is what, exactly, it means.
The classification at the centre of the debate, NOVA, sorts foods by degree of processing rather than nutrient content — an idea that outrages some nutritionists and electrifies others. Its critics note the category lumps together the plainly dire and the arguably benign: a cola and a wholemeal packaged bread can share a classification. Its defenders answer that the category works precisely because it captures something nutrient labels miss — industrial formulation itself. Our health desk has read the umbrella reviews so you can argue about them accurately.
The Evidence, Ranked by Strength
At the top sits a landmark controlled feeding trial by the National Institutes of Health, in which volunteers ate freely from ultra-processed or unprocessed menus matched for presented nutrients — and consumed about 500 extra calories a day on the ultra-processed diet, gaining weight in two weeks. Below that: dozens of cohort studies tracking hundreds of thousands of people, consistently linking higher ultra-processed intake to worse outcomes, with the usual caveats about confounding — people who eat this way differ in many ways besides diet.
The mechanisms under investigation are refreshingly concrete. Energy density and soft textures speed eating past the point where satiety signals arrive. Formulations engineered for craveability do what they were designed to do. Some additives — emulsifiers are the current suspects — show effects on gut microbes and inflammation in animal and early human work. None of this requires believing in a conspiracy; it requires believing that food engineered to be overeaten is, in fact, overeaten.
What Follows for Policy and Plates
Several countries have moved from evidence to labels: warning labels on ultra-processed products in Latin America, dietary guidelines in Brazil advising avoidance outright. The food industry has responded with reformulations that game the category boundary — a preview of the classification wars to come. For individuals, the robust advice is boring and powerful: the lower the share of industrial formulations in the diet, the better the outcomes in nearly every study, whatever the precise mechanism.
The Industry’s Answer Is a Product Line
Watch what manufacturers do, not what they say, and the classification debate resolves into strategy documents. The largest food companies now reformulate toward the category’s edges — removing the additives that trigger the ultra-processed label, adding fibre and protein that marketing can celebrate — while lobbying against definitions that would capture their flagship products. Some of this is genuine improvement wearing cynical clothes; some is label engineering. Either way, the NOVA framework has achieved what a generation of nutrient labels did not: it has made processing itself a competitive variable.
For shoppers, the translation problem remains. Nobody can memorise ingredient-system philosophy in a supermarket aisle at 6 p.m. The heuristics that survive contact with real life are the ones dietitians actually offer: cook from ingredients more often than you assemble from products; treat the longest ingredient lists as information; and remember that the category’s worst offenders — drinks with no food in them at all — are also the easiest to surrender.
Science rarely offers a category this big and this contested. The wisest reading treats ultra-processed food the way epidemiology treats smoking’s early decades: the signal is strong, the mechanisms are plausible, and waiting for perfect certainty is itself a decision — one the food environment makes for millions daily. More at newsletterissues.com.
