Harder-to-chew ultra-processed food cut daily calories in a 2-week trial

The background: why do ultra-processed diets make people eat more?

“Ultra-processed foods” are, in the definition of the widely used Nova classification, industrial formulations made from food derivatives and chemical substances through a sequence of processes, bearing little resemblance to the original food^[1]. Large observational studies have linked diets high in such foods to a higher body weight^[1].

The most influential experiment on the topic came from the US National Institutes of Health in 2019. Twenty adults lived on a research ward and ate either an ultra-processed or an unprocessed diet for two weeks each, as much as they liked. On the ultra-processed diet they ate about 508 kcal more per day and gained about 0.9 kg, while they lost about the same amount on the unprocessed diet^[2].

That study showed that ultra-processed diets can lead to overeating, but not exactly why. One explanation that researchers have proposed: many ultra-processed foods are soft and easy to swallow, so they are eaten quickly. In the NIH trial, the ultra-processed meals were indeed eaten at a faster rate and contained more calories per gram^[1]. Eating fast is known to go along with eating more: a 2014 analysis of 22 experiments concluded that a slower eating rate leads to lower food intake at a meal^[3]. And in a 2022 study with 50 participants, harder-textured lunches were eaten more slowly and reduced the calories eaten at that meal by 26%, whether the food was ultra-processed or not^[4].

What was missing was evidence that this effect lasts beyond a single meal or day – that people do not simply make up for the missing calories later. The new trial from Wageningen University in the Netherlands was designed to test exactly that^[1].

What the researchers did

The team built two diets from familiar, commercially available Dutch foods, both made up of more than 90% ultra-processed food (measured by calories). One diet used foods with “hard” and “chewy” textures known to slow people down; the other used soft foods that are quick to eat^[1,5]. Both diets offered the same amount of food and calories at each meal – three times a normal Dutch portion, so that people could eat as much as they wanted – with the same calorie density and variety. The two diets were also matched for how much people liked them and how familiar the foods were^[1].

Forty-one healthy volunteers (21 men, 20 women; average age 27, average body mass index 23.4) completed the study. Each ate one diet for 14 days, then returned to their usual food for two weeks, then ate the other diet for 14 days. The order was decided by chance. Because every participant ate both diets, each person served as their own comparison^[1].

On weekdays, participants ate breakfast, lunch and dinner in individual booths at the university's research unit, where every plate was weighed before and after the meal and each meal was filmed to measure eating speed. Snacks were provided to take away. At weekends they ate pre-packed study meals at home and returned the leftovers to be weighed^[1].

To avoid people changing their behaviour because they knew what was being tested, the volunteers were told that the study was about how different protein types are digested. Only 4 of the 41 guessed afterwards that it was really about food texture or eating speed^[1].

What they found

The texture trick worked. All participants ate faster on the soft diet and more slowly on the harder diet. On average, eating speed (measured in grams per minute) differed by 43% between the two diets^[1].

Slower eating meant fewer calories – every day. On the slower diet, participants ate on average 2,301 kcal per day; on the faster diet, 2,671 kcal. The statistical uncertainty is modest: the true average difference is likely to lie between 221 and 517 kcal per day. The gap appeared at breakfast, lunch and dinner and did not shrink over the two weeks, so there was no sign that people compensated by eating more later^[1].

It held for almost everyone. 37 of the 41 participants (90%) ate less on the slower diet, between 80 and 773 kcal per day. In weight, they ate 264 g less food per day^[1].

No extra hunger, no less enjoyment. Participants rated both diets as equally tasty and familiar, and gave both the same overall satisfaction score. Their hunger and fullness ratings before and after meals were practically the same on both diets^[1].

Body weight: too early to tell. Body weight did not change measurably on either diet. Body scans did show a small drop in body fat of 0.43 kg on the slower diet. Over just two weeks, such small changes are hard to interpret, and the authors themselves call for longer studies^[1].

One complication: to keep calorie density identical, the researchers could not match the diets perfectly for nutrients. The faster diet had more fat (33% versus 22% of calories served). However, the authors' additional analyses suggested that eating speed, not fat content, explained the difference in intake: after statistically accounting for the fat served, the gap remained about 310 kcal per day^[1].

What this means

The study adds an important piece to the ultra-processed food debate. Both diets here were ultra-processed, yet they led to clearly different calorie intakes. This suggests that the label “ultra-processed” alone does not determine how much people eat, and that how quickly a food can be eaten is one of the properties that matters. The authors argue that eating speed may even have contributed to the overeating seen in the 2019 NIH trial, where the ultra-processed meals were eaten faster^[1,2]. That is an interpretation, though; the new trial did not include a minimally processed diet, so it cannot directly compare processed with unprocessed food.

Independent experts who commented on the results when they were first presented at a conference in June 2025 saw them as consistent with earlier work. Julian Hamilton-Shield, a professor of diabetes and metabolic endocrinology at the University of Bristol, said the study reinforced earlier data that slowing eating rate reduces calorie intake, and noted that while “it can be difficult to sustain a reduced eating speed”, changing food texture could be an additional tool for weight management at population level^[6].

A practical attraction is that participants ate less without trying and without being told to eat slowly. That could interest food manufacturers who want to reformulate products, and the authors suggest texture could be used in dietary advice^[1]. Amanda Avery, a nutrition and dietetics researcher at the University of Nottingham, saw the results as reinforcing the importance of sitting down and taking time over food, and of choosing foods with more texture that need chewing^[6]. Note, however, that the trial did not test such advice: it tested pre-selected diets in a controlled setting.

It is also worth knowing who funded the work. The trial was part of a public–private research partnership in the Netherlands, and several authors have received speaking fees or travel support from food companies or have advisory roles in the food industry. These ties are declared in the paper^[1]. They do not make the results wrong, but they are a reason to wait for independent replication.

What the study cannot tell us

• Short duration: two weeks per diet. Whether the effect persists for months or years, and whether it leads to weight loss, is unknown; the authors say this needs further research^[1].
• Small, selected group: 41 young, healthy adults with mostly normal weight, who did no more than four hours a week of moderate-to-vigorous exercise and who usually got no more than half of their food from ultra-processed products^[1]. The results may differ in older people or in people with obesity.
• No unprocessed comparison: the trial compared two ultra-processed diets, so it says nothing direct about whether less processed food is better or worse^[1].
• Diets not identical in nutrients: the faster diet contained more fat. The authors' analyses point to eating speed as the main driver, but this cannot be fully separated in a single trial^[1].
• Calories, not health: the main outcome was how much people ate. Other effects of ultra-processed food, such as on blood sugar or the gut, were measured but are not part of this paper^[1].
• Industry links: the research partnership involved private partners, and several authors report food-industry ties^[1].