Ultra-processed food (UPF) is a category of food that’s charged with invading our diets and ruining our health. But is it as bad as it’s made out to be? Hatty Willmoth writes.

How do we decide that a food is ‘unhealthy’? Is it by counting how many vitamins and minerals it contains? Should we measure healthiness by calories? Or should we follow in the footsteps of food packaging labels and assess fat, sugar and salt instead?

All those judgements are sometimes described as ‘blunt instruments’, telling us some of the story, but leaving out a lot of crucial information about how a food will affect our bodies.

But increasing attention is being directed towards the NOVA system, whereby food is categorised into four camps according to its processing.

Is NOVA the best way to judge healthiness? That’s up for debate.

The NOVA system

The four NOVA categories are: unprocessed or minimally processed food; processed culinary ingredients; processed food; and ultra-processed food.

Group 1 includes ‘whole’ or ‘real’ food, defined by NOVA as plant and animal foods separated from nature. They may have been slightly altered (peeled, chilled, packaged, butchered, etc.) but remain recognisable as what they are and largely intact. Examples include fresh fruits, vegetables, grains, meat, fish, pulses, legumes, nuts and seeds.

Group 2 is made up of substances that are primarily used to prepare, season and cook Group 1 foods. Called ‘processed culinary ingredients’, they enable home cooking but are not generally consumed by themselves. Examples include olive oil, butter, demerara sugar and table salt.

Group 3 is processed food: the results of combining foods from Group 1 and Group 2. This might include homemade stew, restaurant spaghetti bolognese, artisan bread, sundried tomatoes, tinned fruit in syrup, or cheese. Group 3 foods usually contain a few ingredients that could all be found in a domestic kitchen.

Then there’s Group 4: ultra-processed food, or UPF. The name may be an oxymoron, because some argue that these are not actually foods, but “industrial formulations manufactured from substances derived from foods or synthesized from other organic sources” (Food systems and diets, 2016).

UPF is made in factories using cheap refined ingredients and complex industrial processes that are not replicable in a domestic setting. Convenient, branded and aggressively marketed, UPF is designed primarily for profit.

When trying to spot UPF, look at a food’s ingredients list. If it contains emulsifiers, sweeteners, thickeners, colours, flavour enhancers, bulking agents, humectants, protein isolates, modified starch, maltodextrin, high-fructose corn syrup, or anything similar, then it’s UPF.

It’s not difficult to find. UPF occupies most of the space on our supermarket shelves and somewhere between 51% and 68% of the average UK diet.

To learn more about ultra-processed food, read this article which explains the food group in more depth.

The origins of NOVA

The NOVA system was first developed in 2009 by a team of Brazilian scientists at Sao Paulo University. Ever since, the team have issued stark warnings against eating UPF.

A healthy diet should mostly comprise of Group 1 foods (unprocessed or minimally processed), they say, while Group 4 foods (ultra-processed) should be avoided. 

And now, momentum is building for UPF to be recognised as harmful. In recent years, study after study has found an association between a higher consumption of UPF and adverse health outcomes.

More specifically, UPF has been linked with a higher prevalence of overweight and obesity; type 2 diabetes, hypertension, and cardiovascular disease; depression; all-cause mortality risk; poor maternal and child health; and more.

Today, NOVA is the basis of 10 countries’ food guidelines, including Brazil, France, Israel and Mexico – but countries such as the UK seem reluctant to follow suit.

The big question is whether there is enough evidence to label the industrial processing of food as inherently problematic, or whether ultra-processed foods are just more likely to be unhealthy.

Dr Chris van Tulleken's month of ultra-processed food

In 2021, public awareness of UPF began to swell after celebrity doctor and television presenter Chris van Tulleken starred in a documentary called ‘What are we feeding our kids?’, switching up his diet to test the effects of UPF personally.

Accustomed to eating a diet that was about 30% UPF, van Tulleken took on an 80%-UPF diet for four weeks, to test the hypothesis that UPF was itself the problem.

By the end of the study, van Tulleken had gained almost 7kg. He reported feeling unhappy, anxious, sluggish and old on the diet, and said he experienced heartburn, he struggled to sleep, his libido dropped, and he developed piles from constipation.

Not only that, but scientists monitoring van Tulleken noticed that his hunger hormone (ghrelin) increased by 30%, and his brain exhibited changes comparable to those of someone taking an addictive substance such as alcohol, nicotine or illicit drugs.

So, on top of making van Tulleken feel awful, UPF was addictive, moreish, tasty, and ultimately unsatisfying, so he felt hungry soon after eating substantial quantities of food.

Still unconvinced: the correlation argument

However, not everybody agrees that UPF is the problem, as was clear from a statement released by the British Nutrition Foundation (BNF) in response to van Tulleken’s documentary.

BNF said: “The diet Dr van Tulleken followed was, overall, a very unhealthy diet – with many foods that are calorie (energy) dense and likely to be high in salt, sugar and saturated fats … with very few fruits and vegetables and very little fibre. Thus, it is not surprising that he put on weight and did not feel well.

“Alongside all the unhealthy products consumed in the programme, the ultra-processed definition also includes foods that can be included [as] part of a healthy diet such as wholemeal sliced bread, low fat, lower sugar yoghurts, wholegrain breakfast cereals, fish fingers, or baked beans that can contribute important nutrients to the diet.”

Ultra-processed food is generally unhealthy

In the UK, food regulations define ‘unhealthy’ food as that which is high in fat, salt or sugar (HFSS), irrespective of processing.

There is a great deal of overlap between HFSS food and UPF, such as most ‘junk’ food: shop-bought biscuits, crisps, sweets, fast food burgers, drinks, tubs of ice cream, and so on.

There are exceptions, such as sliced bread, low-fat yoghurts, high-fibre breakfast cereals, and protein snack bars, which are all usually UPF but marketed as healthful.

In general, though, UPF tends to contain a great deal of processed fats, sugars, starches and fats.

In fact, the Brazilian NOVA group have also investigated the link between UPF and excessive sugar intake in the UK, in a 2019 paper.

Using data from the UK National Diet and Nutrition Survey (NDNS) – a large, nationally representative database of food diaries taken from 2008 until 2014 – the scientists estimated the extent to which our national sugar problem was implicated in our love of UPF.

And they concluded that eliminating UPF from our diets would remove about 47% of our excessive free sugar consumption. In contrast, getting rid of domestic table sugar would reduce the problem by just 9.4%.

It makes sense that UPF would be sweet, fatty and salty; manufacturers use cheap oils, syrups, starches, and flavourings to make their products appealing, cheaply, for maximum profit.

As a result, it is commonly argued that the link between UPF and poor health is just a strong correlation, and that processing isn’t the issue.

A randomised controlled trial

But in 2019, the first randomised controlled trial was set up to specifically test for a causal relationship between UPF and adverse health effects.

Over the course of four weeks, 20 healthy-weight adults were monitored in a controlled clinical setting while eating carefully devised diets.

For two weeks, subjects received one diet – either mostly ultra-processed or minimally processed food – and then, for another two weeks, they switched to the other diet.

Meals were designed to be as similar as possible across the diets in terms of calories, energy density, macronutrients, sugar, salt and fibre, and subjects were instructed to eat as much or as little as they wanted at each meal.

The results were decisive. Participants ate an average of an extra 508 calories per day when eating a mostly ultra-processed diet compared with minimally processed one.

They also gained an average of 1kg of weight while eating UPF and lost about 1kg on a minimally processed diet.

Up until this study, evidence about UPF had overwhelmingly been observational, based on pre-existing data that wasn’t originally intended to measure UPF or draw conclusions about its harmfulness.

But this randomised controlled trial – the gold-standard of nutritional evidence – indicates that processing itself might be the issue, even when nutritional composition is similar.

The 'why' behind ultra-processed food

Some companies are now calling for UPF to be directly tackled by UK food regulations. In June 2023, the First Steps Nutrition Trust (FSNT) published a paper which looked at the harm caused by UPF through the lens of early-years health.

It estimated that children in the UK between the ages of two and five get an average of 61% of their total energy from UPF – more than their counterparts in Australia (47%) and the US (58%) – and cautioned that a young child’s diet could shape taste preferences, dietary habits and health over the course of their lifetime.

The paper also set out a number of reasons why UPF might contribute towards ill-health, citing its impact on hunger hormones; how easy it is to overeat; the fact that it encourages snacking and eating on the go; and its use of potentially harmful additives that might disrupt gut health, such as emulsifiers.

Plus, the authors argued that national efforts to reduce our consumption of HFSS foods inadvertently promoted UPF, as manufacturers could boast that their products were ‘low fat’, ‘low calorie’, ‘reduced salt’, and so on, while selling foods that were nevertheless highly processed and unhealthy.

Van Tulleken wrote in the paper’s foreword that the evidence linking UPF to ill-health was “robust”, but that anti-UPF recommendations would be difficult to implement due to inevitable pushback from the food industry, the perception of UPF as cheap and convenient, and the seminal place UPF now occupies in Western food culture.

“We need a revolution in terms of what our children consume,” he wrote, “and this can only be achieved when we disentangle ourselves from the industry that profits from harming them.”

Is change coming?

But a month after the FSNT released their paper (11 July 2023), a report was published by the Scientific Advisory Committee on Nutrition (SACN) – an independent advisory body to the UK Government – which concluded that there was still not enough evidence to definitively conclude that UPF is harmful by definition.

It acknowledged that systematic reviews have “consistently reported” that eating more UPF was “associated with increased risk of adverse health outcomes” but cautioned that “there are uncertainties around the quality of evidence available”.

Studies of UPF, the report stated, were “almost exclusively observational” and “confounding factors … may not be adequately accounted for”.

Is NOVA the answer?

So, is UPF, as defined by the NOVA system, the enemy of good health? Should we pressure governments and food organisations to act against it?

Heather Rosa, Dean of the Institute for Optimum Nutrition, says the answer is complicated.

“There is some value in [the NOVA system],” she says. “Most people understand that the more processed food is, generally the worse it is for you.

“The more you fraction food into smaller and smaller particles, the more you change the impact it has on the body.”

Many of these foods are also addictive, she explains: “If you eat a lot of refined sugar and fat at the same time, with that textural feeling in the mouth, you trigger the bliss point in the brain. With salt too, you’ve got the ultimate taste explosion, so it’s very difficult to not want more. The brain has been lit up; it’s addictive food.”

On top of that, the low nutritional value of most UPF encourages us to overeat it.

“Most ultra-processed foods are low in protein,” Rosa says, “but the body is seeking the 15% protein that it needs for survival and will keep you eating until you reach it.”

However, processing itself is not necessarily bad, she says. “You can’t feed a large nation unless you have some form of industrial processing.

“We’ve got too many people on the planet now who can’t access local food at the point of production, so you do need to have a safe system of delivering food to people. And that means some form of processing to enhance storage and prevent spoilage.”

She adds: “Not all ultra-processed foods are problematic. Baby formula is NOVA category four; it’s an essential food. For specialist groups, you need a level of processing to make foods non-allergenic.”

Much like other systems of judging and categorising food, the NOVA system seems to be yet another blunt instrument – at least according to Rosa.

“Classifying food is probably one of the most difficult things we attempt to do in the nutrition space,” she says. “The issue is the nuance is lost when people try to find a simple approach.

“Ultra-processed foods contain many ingredients that have different effects in different areas of the body. We don’t fully understand the impact of single foods on function, let alone foods when they’ve been altered and then amalgamated back together in multi-ingredient products.

“I don’t think we need another scoring system. When you start to categorise things, you get into all sorts of mess.”

Instead, Rosa’s key advice for healthy eating is: “Eat whole food and not too much of it. Avoid refined vegetable oils and avoid sugar.”

Enjoyed this article?

Read about sustainable eating

Check out our CPD course on ultra-processed food

For articles and recipes, subscribe to the Optimum Nutrition newsletter

Discover our courses in nutrition