Ultra-Processed People

Ultra-Processed People begins with a deliberately unsettling reframing: perhaps the central question is not whether an individual has enough discipline, but what kind of substances and systems now make up an ordinary diet. Chris van Tulleken calls ultra-processed food, or UPF, a novel category of industrial products that has become normal in many countries. His book is forceful, personal and openly political. But its most useful invitation is simpler: stop treating food choice as something made in a vacuum. Van Tulleken’s argument is not that all processing is bad. Cooking, freezing, fermenting, pasteurising, canning and milling have long made food safer, more available and sometimes more nutritious. The distinction he explores comes from the NOVA classification, which focuses on the nature and purpose of processing, not merely on a food’s calories or nutrients. At the far end are formulations assembled from refined substances, extracts and additives, often engineered to be convenient, stable, cheap and highly appealing. The book asks us to take that category seriously without turning it into a purity test. A packaged food is not automatically UPF; an unfamiliar ingredient is not automatic proof of danger; and a person’s diet cannot be judged from one item in a shopping basket. Still, van Tulleken argues that when a large share of a population’s food comes from products designed primarily for profit, scale and repeat purchase, the effects may reach beyond nutrition labels. That is why this is a book about science and power at once. It examines eating behaviour, food texture, additives and health research, but also advertising, affordability, retail and inequality. Its moral target is not the person grabbing a quick meal after work. It is a food environment that makes the most heavily marketed option feel like the easiest, cheapest and sometimes only realistic option. The aim of this summary is to retain that empathy while being careful about what the evidence can—and cannot—yet establish.

NOVA sorts foods into four broad groups. Group one includes unprocessed or minimally processed foods: fruit, vegetables, beans, eggs, plain yoghurt, meat, grains and foods altered mainly by cutting, chilling, drying or similar practical steps. Group two covers culinary ingredients extracted from foods or nature, such as oil, butter, sugar and salt. Group three includes processed foods made by combining foods with those ingredients, such as many breads, cheeses, canned vegetables or preserved fish. Group four, ultra-processed foods, includes industrial formulations that often contain substances rarely used in home kitchens and additives intended to alter flavour, colour, texture or shelf life. The key idea is purpose. In the NOVA account, UPFs are not defined simply by being high in sugar, salt, fat or calories. Some clearly are. But the category also asks whether the product is a reconstruction: a food-like item made from separated components and designed to be ready-to-consume, intensely convenient and commercially durable. Soft drinks, many confectionery products, some packaged snacks, instant meals and certain breakfast cereals are common examples. Yet classification at the boundaries can be difficult. Products vary by country and recipe, and reasonable people can disagree about particular breads, plant-based foods or fortified products. That ambiguity matters. NOVA is a research and public-health framework, not a perfect chemical detector of harm. It can capture features that nutrient-by-nutrient thinking misses, such as food structure, the replacement of meals by snacks, and the industrial strategies behind product design. But it should not lead us to assume that every product in the category has identical effects, or that minimally processed automatically means nutritionally ideal. Van Tulleken uses the framework because he believes the usual language of ‘healthy choices’ conceals the transformation of food into a marketable formula. His sharper claim is that UPF is not merely food with a few undesirable nutrients added; it may be a different kind of eating experience. The rest of the book investigates possible reasons: how these products are physically built, how quickly they can be eaten, how they interact with appetite, and how their dominance is maintained.

To make the issue concrete, van Tulleken undertakes a self-experiment, moving to a diet in which ultra-processed products supply most of his calories. He records changes in appetite, mood, weight and eating behaviour. The experience gives the book momentum: he describes feeling pulled toward more food and becoming aware of how little effort some products require to consume. It also helps him show that a UPF-heavy diet can look ordinary rather than grotesque. It resembles what many people encounter every day. But a self-experiment is storytelling and hypothesis generation, not controlled proof. One person’s experience cannot identify a universal cause, separate every dietary change or predict what will happen to someone else. Van Tulleken generally uses his experiment as a window into the lived reality of eating these products; listeners should resist treating it as clinical evidence. The book’s most important controlled evidence comes from a small NIH inpatient randomised crossover trial led by Kevin Hall and colleagues. Twenty adults spent two weeks on an ultra-processed diet and two weeks on an unprocessed diet, in random order. The menus were designed to be broadly matched for presented calories, macronutrients, sugar, sodium and fibre. Participants could eat as much as they wished. During the UPF phase, they consumed roughly 500 additional calories per day on average and gained weight; during the unprocessed phase, they lost weight. That result is powerful because participants lived in a controlled setting and each person experienced both diets. It supports the proposition that, under those particular conditions, the UPF menu led people to eat more. But it was short, involved only twenty adults and tested whole diet patterns rather than isolating one culprit. It cannot tell us whether texture, energy density, speed, reward, additives, food appearance or other differences did the work. The study does not prove that every UPF causes overeating. It tells us that processing-related features deserve serious investigation rather than dismissal.

Van Tulleken places special emphasis on food structure—what researchers often call the food matrix. A whole apple, apple purée and apple-flavoured drink can contain related ingredients, yet they ask different things of the body and of the eater. Chewing takes time. Intact structures occupy volume. A meal with varied textures may slow us down and make stopping points more noticeable. By contrast, many ultra-processed products are soft, crisp, melt-in-the-mouth or easy to swallow rapidly. Their physical design can make large amounts effortless to eat. Eating speed is a plausible part of the puzzle. In the NIH trial, participants ate the UPF meals faster. That observation does not establish why they consumed more, but it fits a sensible mechanism: fullness signals take time to develop, so rapid eating can allow intake to outrun satiety. Food texture is not a trivial decorative feature. It may be an important part of how a product functions. Energy density is another piece. Foods high in calories per gram can deliver substantial energy before volume, chewing and fullness have had much chance to register. UPFs are often energy-dense, though not uniquely so. A diet’s energy density can be changed by water, fibre, fat content and food structure; a rich homemade dessert can be energy-dense too. The point is not that energy density alone explains modern eating, but that it works alongside convenience, palatability and speed. Van Tulleken’s broader claim is that products are engineered as systems. A snack can be designed for a particular crunch, flavour release, dissolve rate, portion format and repeatability. This shifts the question from ‘Why did I lack restraint?’ to ‘What properties made restraint unusually difficult?’ It is a compassionate question, but it should not erase human variety. People respond differently, meals have social meaning, and the evidence has not reduced appetite to one universal mechanism. Still, form matters, and nutrition labels alone do not fully describe it.

The book turns next to additives: emulsifiers, flavourings, sweeteners, colours, thickeners and other substances that help turn separated ingredients into products with reliable texture, appearance and shelf life. Van Tulleken is especially concerned that some additives may influence the gut microbiome, inflammation or appetite in ways conventional nutrition science has underexamined. This is an important research question, but it requires disciplined language. Safety approval for a specific additive does not mean that every combination, cumulative exposure pattern or dietary context has been tested for every long-term outcome. Equally, an ingredient with a technical name is not evidence of harm. Experimental research, including animal and limited human studies, has raised hypotheses about some emulsifiers and other additives. Those findings cannot simply be converted into a claim that additives as a class are causing disease in everyday diets. Mechanistic research is evolving, and effects may differ by substance, dose, person and food matrix. The appetite argument is broader than additives. Van Tulleken suggests that UPFs can be unusually rewarding because they combine intense flavours, easy consumption and convenient availability. He is drawn to addiction-like language because some people report cravings, loss of control and persistent eating despite regret. The analogy can illuminate how commercial products exploit reward systems, but it should not become a diagnosis applied to everyone who enjoys a biscuit or buys frozen pizza. Scientific debates continue over how best to define food addiction and distinguish it from ordinary appetite, habit and deprivation. The practical insight is not to fear every ingredient panel. It is to notice patterns. Does a product make eating fast, automatic and detached from hunger? Does it encourage grazing, distraction or repeated buying? Is its appeal based mostly on an engineered sensory hit rather than a meal that satisfies? Those questions do not replace medical or nutritional guidance. They simply help reveal why ‘just eat less’ can be an inadequate response to foods designed to make more eating easy.

For van Tulleken, UPF is not an accidental collection of convenient products. It is the outcome of an economic model. Industrial formulations can use cheap commodity inputs, travel long distances, survive storage, generate recognisable brands and deliver high margins. Their repeatability is commercially valuable: a product can taste the same everywhere, every time. Marketing then does more than announce an existing need. It helps create habits, identities and occasions to eat. The book argues that companies compete not principally to nourish people but to sell more product. That is not a conspiracy claim; it is a description of how profit-seeking firms generally operate. What matters is the mismatch between private incentives and public health. If revenue rises when people eat frequently, if children become lifelong customers, and if product reformulation can preserve sales while projecting healthfulness, then voluntary restraint by individual companies has limits. Marketing is especially relevant because it can blur the line between information and persuasion. Packaging and health halos may foreground protein, vitamins, plant ingredients or a single reduced nutrient while hiding the product’s overall character. Van Tulleken is sceptical of the idea that reformulated UPFs can solve the problem if the commercial goal remains maximum consumption. That is his argument, not a settled conclusion for every product. Reformulation that reduces trans fats, sodium or added sugars can have genuine public-health value. Yet it may not address texture, speed, displacement of meals or aggressive promotion. His proposed shift is from consumer responsibility to commercial accountability. People deserve clear information, protection from misleading claims and environments where nourishing options are genuinely accessible. This is not anti-convenience. Convenience can be a social good, particularly for people balancing work, care, disability, transport and limited cooking facilities. The question is who gets to define convenient food, who bears the health risks, and whether market success should be allowed to depend on making the least health-supporting option the default.

A major ethical strength of the book is its refusal to make UPF consumption a marker of personal virtue. Many people rely on packaged foods because they are affordable, nearby, familiar, shelf-stable and quick. A person working irregular shifts, living far from a full supermarket, cooking without reliable equipment or feeding a family on a tight budget is not facing the same menu of choices as a person with money, time and a well-stocked kitchen. Advice that ignores this can become another form of blame. Van Tulleken argues that food environments are patterned by inequality. In many places, heavily promoted products are ubiquitous while fresh ingredients and time to prepare them are scarce. Children encounter advertising early. Schools, hospitals, transport hubs and neighbourhood stores can normalise a narrow range of options. These conditions are not created by any one shopper, so they cannot be fixed solely at the checkout. His language of rights follows from this. He calls for a right to know what foods are, a right to protection from exploitative marketing and a right to good, affordable food. Rights-based responses might include stronger standards for marketing to children, clearer front-of-pack information, healthier public procurement, support for cooking and food access, and fiscal or regulatory measures shaped carefully to avoid making hardship worse. The exact mix is a democratic and policy question, not something science alone can settle. This framework also changes what practical action looks like. For an individual household, the goal need not be perfect avoidance. Where possible, it may mean increasing meals and snacks based on recognisable ingredients, choosing options that feel satisfying rather than merely restrictive, and making the easier choice a little less automatic. For institutions, it means changing availability and price. The important message is collective: asking for better food conditions is not a luxury preference. It is a reasonable demand for an environment that does not profit from undermining people’s capacity to eat well.

The health evidence is concerning, but it is not all the same kind of evidence. Large observational studies repeatedly find that people who report eating more UPF tend, on average, to have higher risks of several adverse outcomes. A 2024 BMJ umbrella review combined existing meta-analyses and found associations across many outcomes, with particularly strong evidence classifications for some links, including cardiovascular mortality, common mental disorders and type 2 diabetes. These findings justify attention, further research and policy discussion. Yet association is not proof that UPF itself caused every observed outcome. People with higher UPF intake may also differ in income, stress, sleep, smoking, access to care, overall dietary pattern and many other factors. Researchers adjust for known confounders, but observational studies cannot eliminate all of them. Reverse causation can also complicate interpretation: health conditions may influence what people eat. The consistency of the associations is meaningful, but it does not settle every causal question. The Hall randomised feeding trial provides a more direct causal signal for short-term calorie intake and weight change in its controlled setting. It does not demonstrate long-term risks of cancer, depression, cardiovascular disease or diabetes. Nor does it pinpoint whether additives, texture, energy density, eating rate or another feature is responsible. A coherent scientific position can therefore hold two ideas at once: the existing evidence is serious enough to warrant action and precaution, while mechanisms and effects of particular products still need better trials. This is also where uncertainty should protect against two errors. One is complacency: demanding impossible certainty before improving food environments. The other is overstatement: treating all industrially made food as toxic or claiming a particular packaged product explains a person’s illness. WHO’s ongoing work toward evidence-informed guidance on UPF shows that this remains a live scientific and policy field. The most responsible conclusion is concern without panic, curiosity without credulity, and policy designed to learn as it acts.

Ultra-Processed People asks us to look beyond nutrients and individual choice toward the industrial design of food. Chris van Tulleken’s central claim is that many modern products are built to be highly convenient, profitable and difficult to stop eating, and that this design may contribute to widespread health harms. He links the issue to texture, speed of eating, energy density, additives, reward, marketing and the unequal environments in which people make choices. The evidence supports parts of that picture more strongly than others. The small NIH randomised trial showed that participants ate substantially more and gained weight on an ultra-processed menu than on an unprocessed one during short, controlled inpatient periods. Observational research consistently links higher UPF exposure with poorer health outcomes, but these studies cannot prove every association is causal. Research on additives and precise biological mechanisms remains important but incomplete. So the book is most useful not as a catalogue of forbidden foods, but as a challenge to an old story: that poor diets are simply failures of knowledge or character. Its practical philosophy is humane. Do not turn NOVA into a test of purity. Do not assume every processed food is harmful, every unfamiliar ingredient is dangerous or every household can cook from scratch. Instead, notice the broad pattern of what makes up a diet and what surrounds it. Where choices exist, favour food that retains recognisable ingredients, structure and meal-like satisfaction. Where choices do not exist, the failure is not personal. Finally, van Tulleken wants food to become a civic issue. Better labels, limits on manipulative marketing, fairer retail environments, public investment and affordable access to good food are not substitutes for personal agency; they are conditions that make agency more real. The lasting question is not, ‘Why can’t people control themselves?’ It is, ‘Why have we built a food system that asks them to fight so hard, so often, just to eat ordinarily?’