At the start of his latest clinical trial in 2018, National Institutes of Health researcher Kevin Hall was sure he wouldn’t see a difference.
His study, intended to monitor caloric intake and weight gain, offered its participants one of two nearly identical menus. Both contained the same number of calories, and comparable amounts of carbohydrates, fats, and proteins. Even the diets’ fiber, sugar, and sodium contents were matched. Nutrient-wise, they were about as similar as two meal plans could get.
But as the days ticked by, Hall quickly began to see how wrong his initial hunch had been. Despite the superficial similarities, one group was eating much more of the food they were offered. And by the end of two weeks, the members of that same group had gained an average of two pounds, while their counterparts had lost two pounds.
The only explanation was the one factor Hall had thought would have no effect at all: While one menu was made up mostly of whole, unprocessed foods, the other—the one tied to weight gain—was composed almost entirely of ultra-processed foods.
Compared to unprocessed foods like fresh fruits and nuts, ultra-processed foods like cookies and chips tend to have more calories, sugar, fat, and salt, all of which have been linked to putting on weight. But the findings from Hall’s team, published today in the journal Cell Metabolism, are the first to show there’s something inherent to ultra-processed foods, independent of nutritional makeup, that seems to encourage overeating.
“This is really important work,” says Dana Small, a psychologist and neuroscientist studying food choice at Yale University who was not involved in the study. “This study produces a definitive answer to a question we did not have a definitive answer to.”
Though it’s not yet clear why ultra-processed foods have this effect, the results underscore the importance of an issue that goes beyond effective dieting. With their cheapness, convenience, and long shelf life, ultra-processed foods now make up more than half the calories Americans eat. These numbers tick even further upwards for underrepresented minorities, as well as in lower income populations.
“This is not about willpower—we’re living in a manipulated food environment,” says Ashley Gearhardt, a psychologist studying food addiction at the University of Michigan who was not involved in the study. “Ultra-processed foods are unique in ways that we are only just starting to understand.”
Technically, any food that’s been mechanically or chemically altered from its original state can be considered “processed.” That label applies to milk, tofu, frozen spinach, and countless other foods that appear in our diets, and doesn’t automatically designate a product as unhealthy.
Ultra-processed foods, on the other hand, take things one step further by including ingredients that provide cheap, “industrial” sources of dietary energy and nutrients—like added sugars, fats, and chemical preservatives—that enhance an item’s flavor, texture, or shelf life. Some offenders in the ultra-processed arena are familiar, like candy and chicken nuggets; others, like sweetened yogurts, reduced-fat salad dressing (or reduced-fat anything, for that matter), and packaged soup, might be a little more surprising.
For years, scientists have been linking ultra-processed foods to a variety of poor health outcomes, including cancer, obesity, and even an increased risk of death. Most of these studies, however, have been limited to questionnaires and diet records that rely on people to accurately report what they’ve eaten, and can’t establish direct cause and effect.
So Hall and his team decided to do what no other group had done before: Round up 20 people, house them at the National Institutes of Health Clinical Center, and prepare, serve, and track every single morsel of food they ate for a month. Each person was randomly assigned to either an ultra-processed or unprocessed menu for the first two weeks, then switched. Both diets consisted of three meals and a glut of snacks, providing almost 4,000 calories each day, and participants were told to eat as much or as little as they wanted.
In terms of their nutritional composition, the two diets were equivalent on almost every front, down to the average number of calories per gram of food. But while ultra-processed foods had no part in the unprocessed menu, they contributed more than 80 percent of the calories in the ultra-processed diet.
It took some serious finagling to get the menus to match, while keeping the ratio of carbohydrate to fat to protein within a healthy range, Hall says. The team also had to take great pains to keep the study’s results mum as they unfolded, even outfitting participants in loose-fitting clothing to mask any weight gain or loss.
But the work paid off. In the end, the only real difference between the groups was the proportion of ultra-processed foods in their diets.
“I can’t think of another study that has been this well controlled for so long,” Gearhardt says. “That allows us to make much more confident interpretations of what these foods are really doing.” If Hall and his team saw clear differences in outcome, there would be a pretty clear culprit.
But even Hall was surprised to discover how quickly changes in eating behavior unfolded. When put on the ultra-processed diet, participants started eating an average of 500 extra calories a day, resulting in several people gaining weight and body fat over the two-week stint. The difference had nothing to do with the amount of food they’d been offered, or even how good it tasted (when asked, participants reported the two menus were equally appetizing and satisfying). But the inclusion of ultra-processed food had triggered a subtle, and likely subconscious, shift in behavior.
The study was brief, and there’s no telling if these results will hold true on a longer time scale, says Vasanti Malik, a nutrition researcher at the Harvard School of Public Health who was not involved in the study. To that point, the difference in the number of calories consumed on each diet actually decreased slightly as time went on. But that doesn’t make the upshot any less concerning: Something in this ultra-processed diet was making people eat more, and it didn’t take long to see the effects.
What exactly that something was, though, is still an open question. Finding the answer will require a lot more research, but Hall has a few theories.
Although both diets were similar in energy density, or the number of calories per gram, these calculations also counted drinks, including juices and lemonade that acted as vehicles for fiber supplements in the ultra-processed diet. But beverages might not make people feel full in the same way that solid food does—and when the researchers took liquids out of the equation, the solid foods on the ultra-processed menu packed in more calories per bite. This might have made it easy for people on this diet to scarf down a lot of calories, Hall says.
People also ate much faster when put on the ultra-processed diet, consuming 17 more calories per minute compared to the unprocessed diet. It takes a while for our brains to register the feeling of fullness, and this lag gives our mouths plenty of time to overeat—an easy thing to do with ultra-processed foods, which are often softer and easier to chew and swallow, Hall says.
The researchers are currently gearing up for a repeat study that might put these theories to the test. One critical change will involve revamping the ultra-processed diet to include more stews and gumbos. Hopefully, Hall says, this swap will both dilute the energy density of the solid food portion of the meal and encourage people to eat more slowly.
There’s plenty to grapple with, but one needn’t see directly into the belly of the beast to know to avoid its bite. The study’s results are a resounding confirmation of years of nutritional counsel—and if anything, the problem is only getting worse.
As food production becomes increasingly industrialized, ultra-processed foods have taken over the diets of modern Americans. Even when put on diets at two extremes of a spectrum, study participants experienced a bigger change in their hormone levels when shifting to the unprocessed menu, suggesting that their baseline was more aligned with an ultra-processed diet.
That’s not terribly surprising, given that these products tend to be rich in fat and sugars, which set off the brain’s pleasure system, Gearhardt says. Over time, our bodies get used to the reward, and crave it in even higher quantities. This vicious cycle of addiction makes ultra-processed foods a tough habit to break.
But weaning ourselves off ultra-processed foods entirely isn’t a practical goal. The fact remains that ultra-processed foods require less time, money, and effort to purchase and consume—and they’re effectively marketed as such. In the study, the ingredients that went into the unprocessed meal plan cost nearly 50 percent more than those for the ultra-processed menu. Tacking on the hours, skills, equipment, and energy invested in the storage and preparation of perishable whole foods, it’s no wonder ultra-processed foods have become a mealtime fixture over the past 30 years, Hall says.
What’s more, because cost is such a big incentive, the burden of ultra-processed foods is disproportionately shouldered by people who make less money—those with the fewest means to avoid them. This trend threatens to exacerbate existing health disparities between socioeconomic brackets.
There’s no quick fix to this problem. But the path forward is with studies like these, Small says. By pinpointing the mystery factor in ultra-processed foods—whatever it is that’s causing us to eat more and gain weight—researchers might be able to partner with the food industry to cook up cheap, convenient foods that can still confer some health benefits.
In the end, “just giving people nutrition advice won’t be enough,” Gearhardt says. “We need to advocate for policy initiatives and support people’s needs. This is a social justice issue. And we’re living it right now.”