US & Europe · Independent Updated August 2026
The Banned in Europe Registry
Explainers

The Hall Trial: The One Study That Actually Caught Ultra-Processed Food in the Act

In a locked metabolic ward, a NIH scientist matched two diets down to the sugar, salt, and fiber. People ate 500 extra calories a day on the processed one. Here's why that experiment matters, and why its own author keeps pushing back on the hype.

By 8 min read Explainers

Most nutrition studies can’t prove anything. That’s not a dig; it’s the nature of the work. You can’t lock people in a room for years and control every bite, so researchers mostly track what people say they eat and watch what happens to their health decades later. Then everyone argues about whether the pattern means anything, because people who eat a lot of processed food also tend to sit more, sleep worse, earn less, and smoke more. Cause and effect get hopelessly tangled.

Which is what makes the Hall trial such a landmark. For once, someone actually pinned the question down.

If you care about why food feels different in Europe, this is one of the load-bearing studies. It’s the closest thing the field has to a clean experiment on whether ultra-processed food does something to us, or whether it’s just a marker for a messier lifestyle. And the person who ran it has spent the years since refusing to let either side oversell it. That combination, a strong result and a scrupulously honest author, is rare enough to be worth walking through carefully.

The experiment

In 2019, Kevin Hall and his team at the US National Institutes of Health published a trial in Cell Metabolism that did something almost nobody bothers to do: it fed people inside a metabolic ward, where every gram in and out could be measured.

Twenty adults, weight-stable, checked in and stayed. Each person spent two weeks eating an ultra-processed diet and two weeks eating an unprocessed one, in random order. The trick was in the matching. The two menus were engineered to be equivalent for the things nutritionists usually blame: total calories offered, energy density, sugar, salt, fat, fiber, and the overall macronutrient split. On paper, they were the same diet wearing different clothes. One version came from packages and industrial ingredients; the other from whole foods cooked on site.

Then the researchers did the important part. They told participants to eat as much or as little as they wanted. Nobody was put on a portion. The question wasn’t “what happens if we feed people more processed food,” it was “what do people choose to eat when the only thing that changes is processing?”

What happened

On the ultra-processed diet, people ate about 500 calories more per day. The exact figure was 508, and the statistics were about as tight as nutrition science gets. The extra intake came from both carbohydrate and fat, and over the two weeks people gained weight. On the unprocessed diet, the same people, given the same freedom, ate less and lost a bit.

Sit with that for a second, because it’s the whole point. Same calories on offer. Same sugar, same salt, same fiber on paper. The only real difference was how industrially the food was made. And that difference alone moved intake by 500 calories a day, which is roughly the gap between weight stability and steady gain for a lot of adults. Nobody was told to overeat. Their appetite did it for them.

This is the sentence to remember: the trial showed that ultra-processing itself, not just the sugar or fat inside processed food, changes how much people eat. Something about the form, softer textures you chew and swallow faster, the speed of eating, the way these foods hit the gut, seems to blunt the signals that tell you to stop. The study didn’t fully nail down the mechanism, and Hall has been careful about that. But it isolated the effect in a way observational data never could.

Why this is the study that matters for Europe

Ultra-processed food isn’t a European regulatory category the way a banned dye is. There’s no EU law that says “this product is too processed to sell.” So this isn’t a story about something being illegal across the Atlantic. It’s a story about how much of the stuff ends up in a shopping basket, and there the two food cultures diverge sharply. Ultra-processed products make up a far larger share of the average American’s calories than the average French or Italian person’s. The Hall trial helps explain why that gap isn’t just cultural preference or willpower. If processing itself nudges appetite upward, then a food supply built on more processing will tend to produce more eating, quietly, across a whole population.

It also reframes a lot of the individual additives this site tracks. When you pick up an American loaf made with potassium bromate or azodicarbonamide, or a bright candy coated in titanium dioxide, the specific additive is usually not the thing that’s going to hurt you at the doses involved. The more honest concern is what those additives signal: a product engineered for shelf life, softness, and crave, the exact profile the Hall trial suggests leads people to eat more. The additive is a fingerprint of the processing, and the processing is what moved the needle. That’s a very different, and more defensible, way to think about a label than “this chemical is poison.” It’s the same thread that runs through why people so often feel better eating in Europe: the answer is rarely one villain molecule, and usually the whole system the molecule lives in.

The 2025 twist that keeps everyone honest

Here’s where the story gets more interesting than the usual “processed food bad” sermon, and where Hall’s own integrity becomes part of the plot.

A popular idea has grown up around ultra-processed food: that it’s literally addictive, that a milkshake lights up the brain the way a drug does. It’s an appealing narrative and it’s been picked up at the highest levels of US health policy. In March 2025, Hall and colleagues published another study in Cell Metabolism testing exactly that claim. They used brain imaging to see whether an ultra-processed milkshake triggered the kind of dopamine surge in the striatum that addictive drugs reliably produce. It didn’t. The responses were small, wildly variable from person to person, below what the scans could reliably detect, and not tied to how much body fat people carried. Whatever ultra-processed food is doing, the data didn’t support “it hijacks your brain like cocaine.”

That result cut against a story that powerful people in Washington were telling. And shortly after, Hall left the NIH, saying aides to the Health and Human Services secretary had blocked him from discussing the milkshake findings with reporters and, in another case, pressured him to change a paper’s content or take his name off it. He resigned in April 2025 citing censorship, and later said he was unlikely to return even after being asked back. Reported by the Washington Post, CNN, and STAT, among others.

You should hold both of Hall’s studies at once, because he does. The 2019 trial is strong evidence that ultra-processed food makes people overeat. The 2025 study is strong evidence that “addictive like a drug” is probably the wrong frame for why. The same researcher produced both, followed the data in each direction even when the second result was politically inconvenient for the people nominally on his side, and paid a professional cost for insisting on the distinction. That’s what good science looks like when it isn’t being conscripted into a slogan.

What to take from it

The Hall trial earns its status. It’s the rare piece of nutrition evidence that actually isolates a cause, and the cause it isolates, ultra-processing, lines up neatly with what people report when they cross the Atlantic and find themselves fuller on less. That’s real, and it’s a decent reason to lean your basket toward foods that look like foods.

But the sober version of the finding is more useful than the scary one. Ultra-processed food seems to loosen the brakes on appetite; it does not appear to seize control of your brain chemistry. You are not fighting an addiction when you skip the packaged snack. You’re making a slightly harder-than-average choice against a food supply engineered to be easy to overeat. Portions, pace, whole ingredients, and the food culture around the meal all still matter, probably more than any single item on the label. The strongest study we have says the processing is the thing to watch. The same study’s author would remind you not to turn that into a horror movie.

Sources