Scientists receive funding from the sector to change texture and try to reduce the speed at which these products are consumed
Crunchy at first bite, but soon too soft to require chewing, a vast assortment of ultra-processed foods pass through the digestive tract with almost no resistance. This is, in part, what delays the hormonal signals linked to satiety. Now, a group of scientists in the Netherlands believes it is possible to moderate these features and thus avoid excessive consumption.
The program at Wageningen University and Research (WUR), named Restructure, investigates how food consistency and chewing effort influence eating pace and consequently, how much is consumed. If texture does indeed prove to be decisive in this process, products associated with various health risks may continue to occupy store shelves.
Subscribe to our newsletter in English and receive our stories straight to your inbox
Judging by the article published in the American Journal of Clinical Nutrition, the Sensory Science and Food Behavior group at WUR’s Division of Human Nutrition and Health found exactly what they were looking for.
That was the experiment reported by the researchers: for two weeks, 39 volunteers lived on a diet of ultra-processed foods alone. Meals were offered freely and had similar energy density, taste and level of processing. They were also equivalent in portion size and variety.
The only change was in texture. One range of ultra-processed foods demanded more chewing; the other could be consumed effortlessly. The objective? To see if this difference altered the amount of calories consumed throughout the day.
It did. Participants consumed around 369 fewer calories when meals had textures that encouraged slower chewing. Average consumption dropped from 2,671 to 2,301 calories per day, a reduction of almost 14%.
This is not a case of the first randomized clinical trial (RCT) to compare diferente diets.
The study, published in 2019 and conducted by Kevin Hall at the National Institutes of Health (NIH), the leading US agency supporting biomedical science, yielded unexpected results.
A group of 40 people stayed in hospital for four weeks, receiving all meals under the supervision of scientists. The meals were equivalent in terms of calories and nutrients.
In the ultra-processed diet, participants in the study consumed an average of 508 calories per day, gaining 900 grammes of body weight as a result.
The effect was the opposite when they consumed fresh, minimally processed foods. Instead of gaining weight, they lost almost the same 900 grammes in the same period.
Joio contacted the consortium responsible for the study on mastication on several occasions. Among other points, they questioned whether the result alters or merely complements the evidence linking diets rich in ultra-processed foods to poorer health outcomes. At the time of writing this report, there had been no response.
Misguided take
The idea that the physical shape of ultra-processed foods makes it easier to consume more calories than the body needs has been translated into accessible language in a book called Ultra-processed people: why do we all eat stuff that isn’t food…and why can’t we stop?, by British doctor and researcher Chris van Tulleken.
At one point, the author draws on the work of Anthony Fardet, from the French National Institute for Agriculture, Food and the Environment (Inrae), to introduce the concept of the “food matrix”. This refers to the physical system which keeps fibre, water and nutrients integrated to the natural structure of foods. The human digestive system, he points out, has evolved to break it down gradually. When this matrix is ruptured, the way the body reacts to food also changes.
The apple is classic example: if eaten whole, it causes a gradual increase in blood sugar and maintains satiety for longer. Apple juice, on the other hand, – and even a purée made from the fruit – causes rapid spikes in glucose and insulin, followed by a drop in those levels and a feeling of hunger again. The nutrients are still there; what is lost is the structure that modulates their absorption.
This process is intensified with ultra-processed foods. Plant components and meats are ground, pulverized, extruded and recombined until they lose almost all physical resistence. The result, writes Van Tulleken, are “pre-chewed” products which can be devoured effortlessly.
While part of the literature considers this characteristic as one of the reasons for revising food and nutrition policies, the Big Food lobby treats it as an opportunity for intervention.
The new special issue of The Lancet magazine aligns with the first diagnosis. An infographic representation in the publication shows that improving the diet of populations cannot depend on changes in consumer behavior.
Even when the industry offers reformulated versions that promise to make more moderate choices easier, policies are needed in order to “regulate the marketing and production” of ultra-processed foods that “confront corporate power” and that reorganize food systems to prioritize “health, justice and sustentainability”.
Objectives such as changing the texture of ultra-processed foods do little to advance proposals for alternatives to food systems. Even less so in curbing corporate interference in regaluratory processes.
The fact is that the field operates within limits that lead it to favor solutions compatible with the existing system. This is how specific corrections come to be touted as the statute for major transformation. The very architecture of the project expresses, to some extent, this way of framing the problem.
Rigged game
Restructure’s governance has two levels. The main one is a Steering Committee, described as “fully independent” and responsible for the “most important decisions” about the overall focus of the project. The group is formed by international experts in the fields of eating behavior, energy intake regulation, public health nutrition, food processing and technological innovation communication.
One of them is Remco Havermans, a professor at Maastricht University, whose work focuses on psychology, with an emphasis on taste perception and eating behavior. The core also includes Barbara Rolls, a professor at Pennsylvania State University, renowned for her research on satiety, eating behavior and energy intake regulation, and David J. Mela, a consultant and advisor in nutrition Science and research. This trio is responsible for defining priorities and guiding the direction of the project.
At the second level we find the Advisory Committee, comprising representatives from all parties involved in the consortium, food companies, ingredient companies and industry assotiations. The committee is advisory in nature and its recommendations are not binding.
Dialogue takes place indirectly, mediated by the Next Food Collective organization. The Steering Committee has the final say on what will or will not be incorporated into the program.
The research is co-financed by the Top Consortium for Knowledge and Innovation Agri & Food (TKI Agri & Food), a Dutch program that promotes public-private partnerships. In addition, it receives direct contributions from companies connected to the food and ingredients industry.
Joio questioned the consortium about about each partner’s participation in financing the research, in absolute and proportonal terms. The team did not respond.
Voluntary shortsightedness
It is no surprise that foods with high energy density encourage excessive consumption. Ultra-processed foods, however, affect satiety in other ways. Generally speaking, they are poor in fibre, which speeds up digestion and reduces the feeling of a full stomach. Their consumption can also cause fluctuations in blood sugar levels, stimulating hunger shortly thereafter.
Added to this is the issue of “convenience”, which is almost always presented as a virtue. In actual fact, eating anywhere, without a table, plates or cutlery, makes it difficult to realize what and how much is being eaten. The so-called economy packs, which reduce the price per quantity, also encourage abuse. Not to mention the marketing of these products, which are often more aggressive than that of minimally processed “analogous products”.
These elements are overlooked in the consortium’s analyses. NOVA group 4 is no longer understood as an expression of an industrial food model. Instead, it is treated as a heterogeneous set of adjustable items, whose harmfulness would depend on specific technical attributes.
It is in this sense that Restructure reaches a conclusion consistent with the approach adopted in the analysis. Since diets equally classified as ultra-processed produced significant differences in consumption solely due to variations in texture and pace of intake, they argue that NOVA does not constitute a homogeneous category from a nutritional or behavioral point of view.
When contacted by the report team, epidemiologist Carlos Augusto Monteiro, creator of the NOVA Classification, stated that investigations into specific mechanisms do not invalidate what is already known at a population level. Accumulated evidence shows that diets rich in ultra-processed foods are associated to worse health outcomes, although not all mechanisms are fully described.
“There is absolutely no need to consume ultra-processed foods. Although not everything has been resolved, it is the private sector that needs them”, he states.
Scientific board
The profile of the researchers also helps to understand the path the research ended up taking. Project leader Ciarán Forde has had a series of interactions with organizations from the food and ingredient sector. These include travel reimbursements, participation in scientific commitees and advisory work with companies and institutes connected to the industry. Among them, names such as Kerry Taste & Nutrition, Ajinomoto, Nestlé and PepsiCo. These relationships are disclosed in conflict of interest statements.
A similar trajectory is repeated among other members. Marlou Lasschuijt received fees for lectures and participation in activities sponsored by industry organizations, in addition to being linked to initiatives funded by food companies. Alwine Kardinaal holds a leadership position at Top Institute Food & Nutrition (TiFN), a public-private partnership between government, universities and the food industry which, by definition, assumes the co-production of knowledge for industrial application.
Other members are involved in fields traditionallly guided towards the liaison between science and product. Markus Stieger, Kees de Graaf (who sits on Mars’ global nutrition council), Lise Heuven and Marieke van Bruinessen work in areas such as sensory science, oral processing and ingestive behavior – key disciplines for food reformulation and the development of products adjusted to consumption parameters.
The Project also incorporates researchers who move between science, industry and communication. Nina McGrath Works at the European Food Information Council (Eufic), an entity funded by corporations in the food sector. Zhen Liu is a business developer in the area of food and biobased research, with a declared focus on innovation and liaison between research and the market. Annet Roodenburg’s career path, meanwhile, includes a 16-year background in the food industry before moving on to initiatives related to labeling and “healthier” choices.
These connections are neither hidden away nor illegal. Most of them appear in declarations of interest, institutional profiles and public presentations. Even so, when viewed together, they help to explain why Restructure is organized around questions compatible with reformulation, consumption modulation and product adjustments.
It should be noted that 8 of the 11 members of the consortium signed the letter that challenged Monteiro’s participation on the WHO committee on ultra-processed foods – a story that Joio covered in detail in this article.
International intrigue
On August 28, 2025, Restructure’s work was still months away from being peer-reviewed. That did not prevent the group from making their findings public.
On that day, a scientific symposium inserted the project on the program of the 23rd International Congress of Nutrition (ICN), held at the Palais des Congrès de Paris. Led by Kees de Graaf e Nina McGrath, the session caused visible discomfort among part of the audience.
One of the slides compiled opinions from national technical committees to support the idea that there is still no scientific consensus on the classification of ultra-processed foods.
According to the presentaton, oficial assessments from different countries had pointed out limitations in the use of the NOVA classification. The French health agency, Anses, stated in a 2024 report that “it is not possible to use this classification system”, adding that the proposal “still lacks scientific validation”. Meanwhile, the Nordic countries’ nutrition committee concluded in 2023 that the classification “does not add to existing dietary recommendations and food classifications”.
At a certain point, the speakers stepped on the gas. To the surprise of the audience, an image of Lord Voldemort, the main antagonist in the Harry Potter saga, appeared on the screen. Above the character were the words “UPF (acronym for Ultra-processed Foods), the foods that cannot be named”, in reference to the villain whose name, in the story, no one dares to pronoune.
Next to it was what the authors called the “magic wand”: none other than Restructure, presented as the key to unlock the scientific impasse surrounding ultra-processed foods.
It is alleged that “processed foods” had been turned into the absolute villains of the food debate – a category already being handed down a sentence ahead of any technical distinction being made. The group hedges their bets on there being new studies which would allow for the dismantling of this narrative, repositioning the discussion on more specific and mechanistic grounds.
The conversation later shifted from mockery to warning. Criticism of ultra-processed foods, the researchers said, may be understandable, to a point – after all, it arose as a reaction to a dysfunctional food system. The problem lies in the consequences of this criticism.
On the screen, a series of rhetorical questions: who would buy more expensive, less convenient food, with lower sensory appeal? Who would have time to cook everything from scratch every day? And in low- and middle-income countries, what would happen to small and medium-sized local industries responsible for jobs and income?
Instead of discussing the food model, the conversation started to revolve around its limitations – cost, time, feasibility – and how to preserve it.
The group also maintains thet there is “confusion” between ultra-processed foods and foods classified solely for their high saturated fat, sugar and sodium content (HFSS). Critics of this position, however, point out that recent literature indicates the problem is not limited to nutritional profile; it also involves physical characteristics, level of processing, hyper-palatability and consumption patterns that encourage excessive consumption.
Finally, it is argued that that many recommendations are more “aspirational” than operational. Yet it is these recommendations that have formed the basis for concrete policies – from restrictions on advertising to children to changes in school meals and front-of-pack labeling, adopted in many countries.
The report team requested access to the presentation. The International Union of Nutritional Sciences (IUNS) did not respond.
The deeper layer
The discussion is more technical than it appears. It has the form and vocabulary of academic science, but operates in alignment with private interests. This is not a problem in itself. In some cases, the interest in making a profit ends up, by contingency, converging with public gains.
This was seen, for example, in the expansion of urban sanitation, the spread of refrigeration technologies or improvements in vehicle safety standards. Within these contexts market growth also meant, at least in part, reducing risks and improving living conditions.
This becomes unlikely, however, when funding depends on the sale of products associated with harm to human health and the environment, such as in the alcohol and tobacco markets, gambling and parts of the food sector.
When scientific evidence points to the need for industry regulation, efforts to contain it multiply – sometimes reaching the point of discrediting the source itself. In a webinar promoted by the US Food and Drug Administration (FDA) – the agency that regulates food, drugs, cosmetics and tobacco – in December 2024, one participant suggested that the NOVA classification might not apply equally across all food contexts.
In her view, it must be remembered that the system was developed in Brazil, a country which is said to have a food environment different from that of the United States, marked by more home-cooked meals and greater social interaction at the table.
In the U.S., she noted – correctly – these products dominate everyday diets. The conclusion, therefore, is that the approach conceived in the Brazilian guidelines “may not be the most appropriate for the U.S. or for other countries”.
Not true. The definition of ultra-processed foods does not depend on local culture or specific ingredients, but on the degree and purpose of industrial processing. A soft drink, an extruded snack or a filled cookie are still what they are, whether in the US or China.
That is why studies based on the NOVA classification have been conducted in dozens of countries, including the United States, Canada, United Kingdom, France and Australia. Kevin Hall’s own trial demonstrates how the model remains valid beyond its country of origin.
Nor is this the first time a Brazilian contribution has been slow to translate into broader change. In an interview with Joio, researcher Patricia Jaime, scientific coordinator of Nupens/USP, recalls Josué de Castro, a physician and geographer from Recife who gained international recognition for demonstrating that hunger was neither an act of fate nor the result of natural limitations. In works such as Geography of hunger (1946) and Geopolitics of hunger (1951), he analyzed how historical processes, development models and economic and social structures would produce and perpetuate the problem.
Even so, his ideas took time to resonate more broadly. For decades, United Nations agencies regarded protein deficiency as the most important feature of child malnutrition. This interpretation guided policies and investments for years. In 1974, researcher Donald McLaren acknowledged the mistake in an article published in The Lancet, titled The great protein fiasco.
The lesson is twofold. First, once a diagnosis becomes the oficial lens, it can capture policies and investments even after it begins to falter. Second, correcting a course demands more than new evidence: it depends on dismantling institutional consensus and entrenched interests.
Patricia sees echoes of this when countries in the Global North start to incorporate the topic of processing without adopting the consolidated vocabulary found in the literature. In Canada, for example, the expression highly processed foods is preferred. The new Dietary Guides for Americans 2025-2030 use the same term.
An important step is the fact that the problem is finally beginning to be acknowledged. But the choice reveals something curious: concepts discussed for years are now appearing in official documents under other names, a belated and partial recognition of what had already been pointed out – now simply too difficult to ignore.
Today, according to a survey conducted on the PubMed database, the difference is striking between studies that use the term ultra-processed foods (UPFs) and those that adopt alternative classifications. “The ratio is almost 11 to 1. Over ten thousand studies use ultra-processed foods, while the other systems do not reach one thousand”, Patricia reports. In meta-analyses, she adds, “virtually all studies use the NOVA Classification”.
This does not mean that the search for new systems is illegitimate. As the full professor at the USP School of Public Health recalls, epidemiology operates through the accumulation of evidence, while the investigation of specific mechanisms falls to other disciplines. The problem arises when the absence of such explanations is used to disqualify classifications that are already supported by population data.
“Some causal links will still need to come from basic science”, she concludes. “But that does not invalidate what has already been demonstrated”
The way out for the industry has been to sponsor academics and research centers willing to use their prestigie to help turn a structural problem into a formulation problem.
This is also a segmentaton strategy: companies can “improve” certain products based on texture and speed of consumption, while continuing to sell, on a large scale, cheap ultra-soft versions made for fast consumption. Innovation does not replace the worse products. It simply coexists with them.