Swapping foods like premade lasagna for hand-made spaghetti Bolognese made a difference [arstechnica.com]:
In a small randomized controlled trial, people lost twice as much weight when their diet was limited to minimally processed food compared to when they switched to a diet that included ultraprocessed versions of foods but was otherwise nutritionally matched.
The trial, published in Nature Medicine [nature.com] by researchers at University College London, adds to a growing body of evidence that food processing, in addition to simple nutrition content, influences our weight and health. Ultraprocessed foods have already been vilified for their link to obesity—largely through weaker observational studies—but researchers have struggled to shore up the connection with high-quality studies and understand their impact on health.
The ultraprocessed foods researchers provided in the new trial were relatively healthy ones—as ultraprocessed foods go. They included things like multigrain breakfast cereal, packaged granola bars, flavored yogurt cups, fruit snacks, commercially premade chicken sandwiches, instant noodles, and ready-made lasagna. But, in the minimally processed trial diet, participants received meals from a caterer rather than ones from a grocery store aisle. The diet included overnight oats with fresh fruit, plain yogurt with toasted oats and fruit, handmade fruit and nut bars, freshly made chicken salad, and from-scratch stir fry and spaghetti bolognese.
While the level of processing differed between the diets, the large-scale nutrition content—fat, protein, carbohydrates, fiber—were similar, as was the proportions of fruits, vegetables, dairy, and starchy food. Overall, both diets adhered to the dietary guidance from the UK government, called the Eatwell Guide (EWG).
[...] On both diets, participants lost weight. But, they lost more by ditching the ultraprocessed foods. At the end of eight weeks, participants lost about 2 percent of their weight on the MPF diet, while they lost 1 percent on the UPF diet. The numbers are small, but the authors note that the trial period is short. Modeled over a full year, the researchers estimated that people who stuck to the MPF diet would lose between 9 percent and 13 percent of their weight, while those who stuck to the UPF diet would lose 4 percent to 5 percent.
In addition to more weight loss, the MPF diet is linked to more fat mass loss, fewer cravings, and lower triglycerides, a factor in cardiovascular health. On the other hand, participants had lower LDL (bad cholesterol) on the UPF diet. The researchers suggested that it would require longer periods on the diets to sort out the effects on cardiovascular health.
The study has several limitations, most notably its small size and brevity. However, it still gives researchers a lot to unpack, including why people lost more weight on the MPF diet. The authors suggest it could be because people simply end up eating more on the UPF diet; ultraprocessed foods are both "hyperpalatable" and crammed with nutrients, i.e., "energy dense." High density, fast eating, and less chewing might mean more intake, the authors speculate. In contrast, the MPF diet scored lower on the taste and flavor ratings, suggesting people may simply eat less.
[...] "The best advice to people would be to stick as closely to nutritional guidelines as they can by moderating overall energy intake, limiting intake of salt, sugar and saturated fat, and prioritizing high-fiber foods such as fruits, vegetables, pulses and nuts," Batterham said in a statement. But, she added, "choosing less processed options such as whole foods and cooking from scratch, rather than ultra-processed, packaged foods or ready meals, is likely to offer additional benefits in terms of body weight, body composition, and overall health."
Journal Reference:Dicken, S.J., Jassil, F.C., Brown, A. et al. Ultraprocessed or minimally processed diets following healthy dietary guidelines on weight and cardiometabolic health: a randomized, crossover trial. Nat Med (2025). https://doi.org/10.1038/s41591-025-03842-0 [doi.org]