Ultra-processed food

What the term actually means

Processing is not one thing. Freezing peas is processing. Milling flour is processing. The NOVA classification separates these by degree, and the category people argue about is the fourth: industrial formulations made largely from substances extracted from foods — starches, protein isolates, modified oils — plus additives that make the result palatable and stable.

The distinguishing feature is not that they are unhealthy by definition. It is that they are assembled from fractions rather than made from foods, and a fraction is the part of a food someone wanted, separated from the parts they did not.

The displacement problem

Requirements for nutrients are roughly fixed. Appetite is satisfied by energy. Those two facts together mean every calorie from a food carrying little besides energy is a calorie of room that something else could have used.

Milling is the clearest case. Take the germ and bran off a grain and you remove about four-fifths of its magnesium along with most of the fibre and much of the B vitamins. The energy stays. Fortification puts a few of those back — usually iron and some B vitamins, because they are cheap and stable — and does not put back the magnesium, the fibre, or the several dozen compounds nobody is measuring.

This is why the pattern shows up in national data as overweight and deficient at once. The two accounts are separate and only one is in surplus.

The trial that changed the conversation

Most nutrition evidence is observational, and observational data cannot easily separate ultra-processed food from being poorer, more rushed and more stressed.

One controlled inpatient study did separate it. Participants lived in a research facility and were given either ultra-processed or minimally processed meals, matched for calories, sugar, fat, fibre and sodium, and told to eat as much as they wanted. On the ultra-processed diet they ate substantially more — several hundred calories a day — and gained weight. On the other they lost it.

The matching is the important part. Something about the food itself drove the over-eating, independent of its nutrient profile on paper. It is one study with a small number of people, and it is the best evidence there is.

What follows, and what does not

What does not follow is that a category label tells you whether a food is good. Wholemeal supermarket bread is technically ultra-processed and is a perfectly reasonable food. Tinned beans and frozen vegetables are processed and among the best value on any shelf.

What does follow is that the share matters more than the individual item. When more than half your energy comes from formulations, the gaps stop being a rounding error and start being the pattern in the survey data.

The useful question is therefore not "is this food processed" but "what is this food displacing". A biscuit eaten after a meal displaces nothing. Half a day of energy from snacks displaces a great deal.