Farnflower Newsletter
Eat

Are ultra-processed foods bad for you?

People ate about 500 more calories a day on ultra-processed food in a controlled trial. Large studies add disease links, mostly via drinks and animal products.

The NIH feeding trial

Offered as much as they wanted, people ate more and gained weight on ultra-processed food.

Average weight change after two weeks on each diet, same 20 adults

  1. Ultra-processed diet +0.9 kg
  2. Unprocessed diet −0.9 kg
Farnflower Source: Hall et al., 2019 · 20 adults, NIH Clinical Center

A diet built mostly on them seems to be. In the best-known tightly controlled trial, people ate considerably more and gained weight on ultra-processed food, and large observational studies link high intakes to heart disease, type 2 diabetes and poorer mental health. But the category is broad: in a large European study, ultra-processed animal products and sweetened drinks carried the risk, while ultra-processed bread and cereals did not.

Key takeaways

  • Ultra-processed means made largely from industrial ingredients and additives, not simply “processed” or “packaged”.
  • In the NIH trial, the two diets were matched for sugar, fat, fiber and salt, yet people ate more of the ultra-processed one.
  • Across 45 pooled analyses, higher intake was linked to 32 health problems, with convincing evidence for heart disease deaths, type 2 diabetes and anxiety.
  • Most of this evidence is observational and rated low quality, so it shows a link and not proof.
  • Cooking more from basic ingredients is the simplest change, without needing to read every label.

What counts as ultra-processed food?

Ultra-processed foods are products assembled from substances extracted from foods or modified in a factory, usually with additives that make them tastier or more appealing. Researchers classify them with the Nova system, which sorts foods by how much and why they were processed. The group that created it suggests a simple test: if the ingredient list contains at least one thing you would rarely or never use in a kitchen, such as high-fructose corn syrup, hydrolyzed protein, flavorings, emulsifiers or sweeteners, the product counts as ultra-processed (Monteiro et al., 2019). By that test, frozen vegetables, canned beans and plain yogurt are not ultra-processed, while most soft drinks and many packaged snacks are.

What did the famous NIH trial show?

Kevin Hall and his team at the US National Institutes of Health admitted 20 adults to their clinical center for four weeks. For two weeks the participants got ultra-processed meals, for two weeks unprocessed ones, in random order. Both diets were matched for calories, nutrients, sugar, fiber and salt, and people could eat as much or as little as they liked. On the ultra-processed diet they ate about 500 calories a day more, mostly from carbohydrate and fat, and they gained weight; on the unprocessed diet they lost about the same amount (Hall et al., 2019).

This is the trial everyone cites, and for good reason: it is a randomized experiment, not a survey. Two weeks and 20 people are also its limits. It shows that ultra-processed food makes it easy to overeat. It can’t show what happens over years, or which feature of the food does it, whether that is texture, flavor or something else.

Large studies link ultra-processed food to a long list of health problems. A 2024 umbrella review in The BMJ gathered 45 pooled analyses covering almost ten million people. Higher ultra-processed food intake was linked to 32 of the health outcomes studied. The evidence was rated convincing for a higher risk of dying from cardiovascular disease, for type 2 diabetes, and for anxiety and common mental disorders, and highly suggestive for death from any cause, depression and obesity (Lane et al., 2024).

The quality of much of this evidence was rated low or very low, because it comes from observational studies in which people report their own diets (Lane et al., 2024). People who eat a lot of ultra-processed food also tend to differ in income and daily stress, and no model removes all of that.

Are all ultra-processed foods equally bad?

No, ultra-processed foods are not all equally bad, and this matters for your shopping. The European EPIC study followed 266,666 adults in seven countries for about eleven years and looked at who developed both cancer and a cardiometabolic disease such as heart disease or diabetes. Higher ultra-processed intake went with a higher risk overall. Within the category, though, the link came mainly from ultra-processed animal-based products and from sugar-sweetened and artificially sweetened drinks. Ultra-processed breads and cereals and plant-based alternatives showed no link (Cordova et al., 2023).

My conclusion: treat “ultra-processed” as a useful warning label for a style of eating, not as a verdict on every product with an additive. A packaged whole-grain bread and a sausage roll can both be ultra-processed and are nothing alike.

What I’d do

  • Cut sugary and diet soft drinks first, since sweetened drinks carried much of the risk in the European study. The diet soda article looks at the sweetener side.
  • Swap ultra-processed meat products such as sausages for beans, fish, eggs or fresh meat most days of the week. The red meat article explains why processed meat stands apart.
  • Keep useful packaged basics like canned beans and frozen vegetables.
  • Cook one simple meal from basic ingredients more often than you do now.
  • Eat ultra-processed snacks slowly and from a plate, since the trial showed how easily they lead to overeating.

Frequently asked questions

Is all packaged food ultra-processed? No. Canned beans, frozen spinach or plain oats are processed but not ultra-processed. The ingredient list tells you: one item you wouldn’t use in your own kitchen is the sign (Monteiro et al., 2019).

Are plant-based meat alternatives ultra-processed? Usually, yes. In the European study, though, they showed no link with the diseases measured (Cordova et al., 2023), and they are a small part of most diets.

Is ultra-processed food the same as junk food? No. The Nova system classifies food by how it was made, not by how healthy it looks. A packaged bread can be ultra-processed, and in the European study ultra-processed breads and cereals showed no link with disease (Cordova et al., 2023).

Can a trial with 20 people prove anything? It can show what it measured. The NIH trial was randomized and tightly controlled, so it shows that ultra-processed food makes overeating easy over two weeks. It can’t say what happens over years (Hall et al., 2019).

Sources

  • Cordova, R., Viallon, V., Fontvieille, E., Peruchet-Noray, L., Jansana, A., Wagner, K.-H., et al. (2023). Consumption of ultra-processed foods and risk of multimorbidity of cancer and cardiometabolic diseases: A multinational cohort study. The Lancet Regional Health – Europe, 35, 100771. https://doi.org/10.1016/j.lanepe.2023.100771
  • Hall, K. D., Ayuketah, A., Brychta, R., Cai, H., Cassimatis, T., Chen, K. Y., et al. (2019). Ultra-processed diets cause excess calorie intake and weight gain: An inpatient randomized controlled trial of ad libitum food intake. Cell Metabolism, 30(1), 67–77.e3. https://doi.org/10.1016/j.cmet.2019.05.008
  • Lane, M. M., Gamage, E., Du, S., Ashtree, D. N., McGuinness, A. J., Gauci, S., et al. (2024). Ultra-processed food exposure and adverse health outcomes: Umbrella review of epidemiological meta-analyses. BMJ, 384, e077310. https://doi.org/10.1136/bmj-2023-077310
  • Monteiro, C. A., Cannon, G., Levy, R. B., Moubarac, J.-C., Louzada, M. L., Rauber, F., et al. (2019). Ultra-processed foods: What they are and how to identify them. Public Health Nutrition, 22(5), 936–941. https://doi.org/10.1017/s1368980018003762

One question a month, answered properly