Ultra-processed food has gone from a term used mainly by nutrition researchers to a fixture of everyday conversation in a remarkably short time. The science has been moving almost as quickly. A major series in The Lancet last November pulled together the global evidence and called for governments to act. A UCL trial put the question to a direct test, feeding people two diets that matched on nutrients but differed in processing. And this January, the UK’s long-delayed restrictions on junk food advertising finally took legal effect, built on a definition that, as Cambridge researchers have shown, only partly overlaps with what counts as ultra-processed. Together they show both how much stronger the evidence has become and how much it still struggles to fit the way policy is written.

A Diet Built Mostly From Factory Formulations

Ultra-processed foods are defined under the Nova classification, devised in 2009 by Carlos Monteiro of the University of São Paulo, as industrial formulations typically containing ingredients rarely used in home cooking, such as emulsifiers, flavourings and preservatives. In the UK they make up the bulk of what people eat. An analysis of the National Diet and Nutrition Survey for 2008 to 2014 found that ultra-processed foods supplied 56.8 per cent of total energy intake, while unprocessed or minimally processed foods supplied just 30.1 per cent. As the share of ultra-processed food in someone’s diet rose, so did their intake of free sugars, saturated fat and salt, while protein, fibre and potassium fell.

The Lancet’s three-paper series, published online on 18 November 2025 with Monteiro among its lead authors, set out to test whether this dietary pattern is driving the rise in chronic disease. Its authors reviewed the epidemiological evidence linking higher consumption of ultra-processed food with obesity, type 2 diabetes, cardiovascular disease and early death, and according to a published summary of the series, 92 of the 104 long-term studies it examined found an increased risk of chronic disease. Much of that evidence comes from observational studies, which show association rather than proof of cause, but the authors argued that waiting for perfect evidence risks delaying measures already known to improve diets.

Same Nutrients, Different Processing

The UCL trial, published in Nature Medicine in August 2025 and led by Dr Samuel Dicken, was designed to get past the limits of observational research. It enrolled 55 adults and gave each of them two eight-week diets in turn, separated by a four-week break. Both diets were matched to the Eatwell Guide, the government’s official advice on a balanced diet, for fat, saturated fat, protein, carbohydrate, salt, fibre, fruit and vegetables. The only planned difference was processing: overnight oats and homemade spaghetti Bolognese on one diet, breakfast oat bars and a lasagne ready meal on the other. Participants were sent more food than they needed and told to eat as much or as little as they liked.

Both diets led to weight loss, but the minimally processed one produced nearly twice as much, a 2.06 per cent reduction against 1.05 per cent, equivalent to an estimated daily calorie deficit of 290 kilocalories compared with 120. The extra loss came from fat and body water rather than muscle, and participants reported markedly better control over their cravings on the minimally processed diet. “Previous research has linked ultra-processed foods with poor health outcomes. But not all ultra-processed foods are inherently unhealthy based on their nutritional profile,” Dicken said, explaining that the trial was designed to separate processing from nutrition.

The results were not one-sided. Secondary markers such as blood pressure, cholesterol and blood glucose showed no significant harm from the nutritionally balanced ultra-processed diet, and mostly no meaningful difference between the two. The researchers cautioned that longer studies would be needed to judge those measures properly, and the eight-week periods are short compared with the years over which diet shapes health.

A Policy Aimed at a Different Target

On 5 January 2026, new restrictions on advertising “less healthy” food and drink took legal effect in the UK, after a voluntary period that began the previous October. Adverts for identifiable less healthy products can no longer appear on television or on-demand services between 5.30am and 9pm, and paid-for online advertising of them is banned outright. Brand advertising that does not show an identifiable product is exempt. The government estimates the restrictions will prevent around 20,000 cases of childhood obesity.

The rules are built on the UK’s nutrient profiling model, which identifies foods high in fat, salt or sugar, rather than on how heavily a food is processed. Researchers at the MRC Epidemiology Unit in Cambridge examined how far the two definitions overlap, using eleven years of National Diet and Nutrition Survey data published in BMJ Nutrition, Prevention and Health in 2025. They found that 59.8 per cent of energy in the UK diet came from ultra-processed foods and 47.4 per cent from foods high in fat, salt or sugar, but that only 58.7 per cent of the energy from ultra-processed foods also counted as high in fat, salt or sugar. The UK’s profiling model, they concluded, captures at best just over half of the ultra-processed food people eat. The biggest groups of ultra-processed foods that escape it are low-calorie soft drinks and white bread.

Where Processing Fits

The exact share of ultra-processed food in the British diet varies a little depending on the survey years and method used, but every estimate puts it above half. Set side by side, the three strands describe a field moving from association towards mechanism, and a policy system still catching up. The Lancet series makes the population-level case that ultra-processed diets are displacing healthier ones and tracking with chronic disease. The UCL trial suggests that processing may matter for weight even when nutrients are held constant, while also showing that a carefully balanced ultra-processed diet is not automatically harmful over a couple of months. And the advertising rules, the most significant food policy to take effect in years, target fat, salt and sugar rather than processing, leaving close to half of ultra-processed energy outside their reach. Whether future policy should regulate processing directly is now one of the central questions in UK nutrition science, and it is one the evidence is only just beginning to be strong enough to answer.

Sources: The Lancet (Ultra-Processed Foods and Human Health series, 2025); Nature Medicine (Dicken et al, 2025); UCL; Nutrients (Rauber et al, 2018); BMJ Nutrition, Prevention and Health (MRC Epidemiology Unit, 2025); Committees of Advertising Practice; Association for the Study of Obesity.