Why ultra-processed food makes you eat more

Ultra-processed food tends to be soft, dense and quick to eat, so more of it goes down before your body registers that you’re full. On a US government research ward in 2019, people ate 508 more calories a day on ultra-processed meals than on unprocessed ones, with sugar, fat, fiber and salt matched on paper.

Twenty people, two menus

Twenty adults, 10 men and 10 women, checked into the NIH Clinical Center in Bethesda, Maryland, for four weeks, in a study run by Kevin Hall’s team at the National Institutes of Health. Each person spent two weeks on an ultra-processed menu and two weeks on an unprocessed one, in random order.

The team designed the two menus to offer the same calories, the same calories per gram, and the same protein, carbs, fat, sugar, fiber and salt. Then they let people eat as much or as little as they wanted, with up to an hour for each meal, and measured what they ate.

On the ultra-processed weeks people ate 508 calories a day more, about 280 of them from carbs and 230 from fat, while protein didn’t change. They gained about 2 pounds (0.9 kg). On the unprocessed weeks they lost about 2 pounds. The 508 calories weren’t explained by taste, since people rated the two menus about equally pleasant and equally familiar, and they didn’t report feeling any hungrier on one menu than the other.

They did eat faster. The ultra-processed meals went down 17 calories a minute faster, and 7.4 g of food a minute faster.

How it works

Think of an apple and a glass of apple juice. The apple takes a few minutes of biting and chewing, and by the time it’s gone your stomach and gut have started telling your brain that food has arrived. The juice is gone in seconds, before any of that has started. Ultra-processed food tends to sit closer to the juice end of that line, even when it’s solid.

1. It’s soft. Hall’s team suggested the ultra-processed foods were softer and easier to chew and swallow. In the Tokyo trial, men chewed less for every calorie on their ultra-processed week.
2. Soft food goes down faster. Less chewing means more food per minute. On the NIH ward that was 17 calories a minute more, and in a Singapore lunch study, harder meals were eaten more slowly, with 26% fewer calories eaten.
3. It’s dense. Food with less water and fiber in it carries more calories in each bite. On the NIH menus, the solid food on the ultra-processed side had about 85% more calories per gram than the solid food on the unprocessed side.
4. Fullness lags. The signals that you’ve eaten come from your stomach stretching and from hormones your gut releases as food arrives. On the unprocessed menu, Hall’s team found more PYY, a gut hormone that tells your brain you’ve eaten, and less ghrelin, the hormone that makes you hungry.
5. The calories pile up before the signals arrive. Across 2,733 meals on the NIH ward, the meals that were denser, eaten faster or built from very tasty food combinations were the ones people ate more of.

Each of these goes with eating more, but the trials changed all of them at once, so nobody can say yet how much each one adds.

The proof

We’re confident that people eat more on ultra-processed menus in the short run. Three controlled trials found it, in three countries, with very different people and foods. The evidence for why is good but thinner. Eating speed and calories per bite have the most behind them, and the authors of a 2025 Lancet review name high calorie density, very tasty food, soft texture and broken-down food structure as the drivers of overeating.

The best case against it

Critics make three points.

The NIH menus weren’t matched the way the headline says. The two menus offered the same calories per gram only when drinks were counted. The ultra-processed meals were low in fiber, so the team dissolved fiber supplements in drinks to match it, and the solid food on that side was about 85% denser. On this view, density did the work, and “ultra-processed” is just a label on dense food. But the low fiber that forced the supplements came from the ultra-processed foods themselves, so the density was part of what the menu was testing. And the England trial, with both diets built to the same healthy eating guide, went the same way.

It’s just that the food tastes better. That would make sense, but the NIH volunteers rated the two menus about equally pleasant. The 2023 analysis of ward meals did find that very tasty food combinations went with eating more, so flavor is likely part of it, alongside speed.

The trials are small and short. That’s true. The biggest had 55 people, and the longest ran 8 weeks on each diet. Nobody has run a trial long enough to show the gap lasts for years. Hall’s team also suggested that people may keep eating until they’ve had enough protein, which could explain part of the gap.

Who it doesn’t apply to

If you’re trying to eat more, the same features help. During long endurance sessions, a gel or a sports drink is group 4, and it does a job that a chewy meal can’t. If a small appetite makes it hard to eat enough, food that goes down easily isn’t a problem to fix, and if you have a condition that limits what you can eat, keep to the plan your doctor or dietitian gave you.

Test it on yourself

1. Find your soft staples. Take the quiz or pick your usual day in your day’s share, and note which foods carry the most ultra-processed calories.
2. Swap three of them for food you chew. Oatmeal for cereal, bakery sourdough or a corn tortilla for packaged bread, air-popped popcorn or almonds for chips. The swap ladder has more.
3. Eat as much as you want for two weeks. Don’t count anything. That’s how the trials ran.
4. Weigh yourself on the same morning each week. On the NIH ward people lost about 2 pounds in two weeks on the unprocessed menu, and one weigh-in can swing more than that, so compare the averages of a few mornings.

The three trials side by side

How the extra calories add up is on bodyfat.guide, in is a calorie just a calorie? and how to eat less without being hungry.