Ultra-Processed Foods: Why Might They Make Us Eat More?

In 2019, twenty people spent four weeks at a research center in the United States, where every meal they ate was provided by the scientists.

For two weeks, they were given a diet composed largely of ultra-processed foods. For the other two, an unprocessed diet. They could eat as much or as little as they wanted.

On the ultra-processed diet, participants consumed an average of about 500 more kilocalories a day. Over two weeks, they gained about 0.9 kilograms on average. On the unprocessed diet, they lost roughly the same amount.

What makes the study particularly interesting is that the researchers were not simply comparing an “unhealthy” diet with a “healthy” one. The menus were designed to be comparable in terms of the energy provided, overall energy density including beverages, macronutrients, sugars, sodium, and fiber. But the diets were not nutritionally identical: they differed, for example, in non-beverage energy density and in certain features of their fat, carbohydrate, and fiber profiles.

Yet people still ate more on the ultra-processed diet.

Why?

That is where one of the most interesting debates in modern nutrition science begins.

Not Every Processed Food Is Ultra-Processed

Cooking is also a form of food processing.

When we boil a potato, roast vegetables, or grind nuts, we alter the original food. But we are usually still working with a whole food or relatively recognizable parts of one.

Many ultra-processed products are different.

They are not simply whole foods that have been cooked, frozen, or ground. They are often new industrial formulations, built from multiple components derived or extracted from foods and recombined into a new product.

To study this distinction, researchers at the University of São Paulo developed the NOVA system. It divides foods into four groups according to the nature, extent, and purpose of their processing. NOVA is not an acronym; it is simply the name of the classification system.

The final group contains what we call ultra-processed foods.

It includes many sweetened drinks, packaged sweet and savory snacks, certain breakfast cereals, ready-to-eat products, and many other industrial formulations.

The category is broad—and that matters.

NOVA can help us determine how a product was made.

But it does not automatically tell us exactly what that product will do in the human body.

Food Is More Than Its Chemical Composition

When we talk about nutrition, we often reduce food to numbers.

How many calories?

How much protein?

How much fat?

How much carbohydrate?

How much fiber?

Those numbers matter. But a food is more than the sum of the nutrients it contains.

Those nutrients are organized within a physical structure.

A whole apple and apple puree may start with the same fruit. But they do not require the same amount of chewing, are not eaten at the same speed, and no longer have the same physical structure.

The same principle can be even more pronounced in products built from heavily modified and recombined food components.

This raises an important question:

Is knowing a food’s chemical composition enough, or does it also matter how that composition is organized within the product itself?

Hall’s study does not provide a definitive answer.

But it shows why the question is worth asking.

How Fast We Eat

In the experiment, participants did not just consume more energy on the ultra-processed diet.

They also ate faster.

This matters because two foods can contain the same number of calories but take very different amounts of time to eat.

A food that must be bitten and chewed for longer naturally limits the rate at which we can consume energy.

A soft food that breaks down easily and can be eaten quickly allows more energy to be consumed in the same amount of time.

Researchers are therefore looking not only at how many calories a food contains, but also at how many calories we can eat per minute.

That does not prove that eating rate explains the full effect of the ultra-processed diet.

But it is probably part of the story.

How Much Energy Is in a Bite?

Eating rate becomes even more interesting when considered alongside energy density.

This is the amount of energy in a given weight of food.

A hundred grams of tomato and a hundred grams of chocolate, for example, weigh the same but contain very different amounts of energy.

If a food is both energy-dense and easy to eat quickly, we can consume a substantial amount of energy in a short time.

That is why eating rate and energy density are increasingly considered together.

Together, they frame a much more specific biological question than simply: “Is this food processed?”

The Structure You Cannot See on the Label

A nutrition label can tell us how much protein, fat, carbohydrate, and energy a product contains.

It is far harder to tell from the label how those substances are physically organized.

Processes such as grinding, crushing, and extrusion can alter a food’s structure.

That can change how much chewing it requires, how quickly it is eaten, and how it breaks down.

There is no simple boundary here between “good structure” and “bad structure.”

Nor does every form of processing have the same effect.

But this physical dimension of food reminds us of something important:

two foods with similar numbers on the label are not necessarily the same food.

What We See Beyond the Laboratory

Hall’s experiment included only twenty people.

Outside the laboratory, however, we have very large observational studies.

In 2024, an umbrella review in The BMJ brought together 45 pooled analyses covering nearly 9.9 million people.

Higher intake of ultra-processed foods was associated with a range of adverse health outcomes. But the strength of the evidence varied substantially across outcomes.

The word “associated” matters here.

Observational data cannot by themselves prove that ultra-processed foods caused the outcomes observed.

People differ in their diets, physical activity, smoking, social circumstances, and many other respects. Researchers try to account for these differences, but they cannot eliminate every alternative explanation.

That is why large observational studies and small controlled experiments are not rivals.

They provide different pieces of the picture.

A Label Is Not a Mechanism

High intake of ultra-processed foods now warrants serious scientific and public attention.

But that does not mean every product in the category is biologically equivalent.

Nor does it mean scientists have identified a single universal mechanism that explains every observed effect.

NOVA helps us see a pattern.

The next task for science is to understand what lies behind it.

Eating rate, energy density, and physical structure are among the leading candidates.

That brings us back to a larger question.

Is food simply the sum of the chemicals it contains?

No.

Their quantities matter, but so does their structure. These substances are part of a product that we must bite, chew, swallow, and digest.

The way a food is made can change the way we eat it.

That may be part of why, in one carefully controlled experiment, people ate about 500 kilocalories more a day without anyone telling them to.