What Is Ultra-Processed Food? A Plain-English Guide, and What the School Lunch Figures Show

Ultra-processed food (UPF) is in the news a lot. Ask a few people what it means, though, and you get a few different answers. Some think it means anything from a packet. Some think it means anything unhealthy. Neither is quite right.

This is a plain-English guide to what the term means, how to spot it, what the research shows and where it stops, and what the figures say about school lunches.

Where does the term come from?

It comes from a food classification system called NOVA, developed by Carlos Monteiro and colleagues at the University of São Paulo in Brazil and first proposed in 2009. NOVA sorts food by how much industrial processing it has had, not by how much sugar, salt or fat it contains. It has four groups.

The four NOVA groups

  1. Unprocessed or minimally processed foods. The edible parts of plants and animals, as they come, or with minimal processing such as washing, drying, freezing, boiling, grinding or pasteurising, with no salt, sugar, oil or additives added. Fresh and frozen fruit and vegetables, whole grains, dried lentils and beans, eggs, fresh meat and fish, milk and plain yoghurt.
  2. Processed culinary ingredients. Things made from group 1 foods or taken from nature, used to season and cook. Oils, butter, salt, sugar, vinegar, honey.
  3. Processed foods. Simple products made by adding group 2 ingredients to group 1 foods. Cheese, tinned vegetables, salted nuts, fruit in syrup.
  4. Ultra-processed foods. Industrial formulations made from several ingredients, including substances with little or no culinary use and additives that make the product more palatable or longer-lasting. Whole foods are absent or make up only a small share.

How do you spot one?

The simplest test is the ingredients list. A long list that includes things you would not find in a home kitchen is the usual sign. Consumer group Which? lists the typical markers as emulsifiers, preservatives, acidity regulators, thickeners and stabilisers, colourings, flavourings, artificial sweeteners, modified starch and glucose syrup.

Examples Which? gives include fizzy drinks, biscuits and cakes, flavoured yoghurts, instant soups, ready meals, fish fingers and jarred pasta sauce. We looked at one real label in our garlic bread post, and at the jar-sauce habit in I Thought the Jar Was the Healthy Choice.

What the school lunch figures show

A pooled analysis of the UK National Diet and Nutrition Survey, using food diaries from 2008 to 2017, looked at what children ate at lunch. It found that the median share of lunchtime calories from ultra-processed food was about 73% for primary school pupils and about 78% for secondary school pupils. That covers school meals and packed lunches together. The analysis is a conference abstract.

A separate paper from Imperial College London researchers, published in the journal Nutrients in 2022 and reported by Imperial, split the same survey years by lunch type:

  • School meals: 64% of calories from ultra-processed food (61% for primary pupils, 70% for secondary pupils).
  • Packed lunches: 82% of calories.

So in this data, school meals contained less ultra-processed food than packed lunches did, and both contained a lot.

Three things to keep in mind. The data runs to 2017, so it may not match lunches today. Secondary pupils reported their own diets, which the researchers say probably means their intake is underestimated. And the study describes what children ate. It does not test what that food does.

What does the research say about health?

The UK's Scientific Advisory Committee on Nutrition (SACN) reviewed the evidence in 2023. It found that reviews had consistently reported higher consumption of ultra-processed foods was associated with a higher risk of adverse health outcomes. It also said that the studies are almost all observational, that eating a lot of ultra-processed food may simply be a sign of other unhealthy eating patterns, and that the evidence "needs to be treated with caution". It called for better-quality trials and for a classification that can be reliably applied in the UK.

One small trial is often quoted. In a US National Institutes of Health study, 20 adults spent two weeks on an ultra-processed diet and two weeks on an unprocessed one, with meals matched for calories, sugar, fibre, fat and carbohydrate. On the ultra-processed diet they ate about 500 more calories a day and gained about 0.9kg. It was a small study, small protein differences might explain up to half of the calorie gap, and the researchers say the specific cause is unknown.

It is not a simple good-and-bad list

NOVA is broad. SACN noted that its categories group together foods with very different nutrition, and Which? points out that some ultra-processed foods, such as wholemeal bread or plain cereals, can still provide fibre and protein. Most of the concern is about products high in fat, sugar and salt. Looking at the ingredients list tells you more than the label on the front of the pack.

Our own position is a cautious one: the research is not settled, so we do not claim ultra-processed food causes any particular disease. But we do not need proof to stop risking it. Where we can, we choose the shorter ingredients list, and we are not perfect at it. A partial swap is still a real change.

Where do you start?

Look at the flavour base of the meals you cook most often. A jar of sauce, a packet mix or a stock cube is often where the longest ingredients list sits, and it is the part you can learn to make yourself. That is what our recipes are for.

Questions people ask

Is all ultra-processed food bad?

Nobody can say that. NOVA is a broad category, the research is mostly observational, and SACN says the evidence should be treated with caution. Some ultra-processed foods are higher in fat, sugar and salt than others, and the ingredients list helps you tell the difference.

Is a packed lunch healthier than a school meal?

In the data above, packed lunches had a higher share of ultra-processed food (82% of calories, against 64% for school meals). That is an average across thousands of children, not a verdict on any one lunch box or school.

Is home cooking always free of ultra-processed food?

No. Home cooking can use a jar, a packet or a stock cube, which is how most of us start. The point is how much of the dish came out of a factory.

Sources: NOVA classification (Monteiro et al., University of São Paulo); Imperial College London, July 2022; Parnham et al., Nutrients, 2022; pooled analysis of the National Diet and Nutrition Survey 2008 to 2017; SACN statement on processed foods and health, 2023; National Institutes of Health, 2019; Which?.

Back to blog