I RECENTLY scoured my kitchen looking for trouble, and I found plenty. There was a packet of instant noodles in a cupboard. Tins of baked beans and a box of muesli. In the fridge, a Jamaican patty, ketchup, hummus and probiotic yoghurts. Over in the bread bin, a loaf. I didn鈥檛 dare peek in the freezer.
These foods are part of my normal diet, which I don鈥檛 think is especially unhealthy. But by eating them, I may be opening myself up to obesity, heart disease, a fatty liver, cancer and more. That鈥檚 if you believe the increasing worries over ultra-processed foods (UPFs) and how bad they supposedly are for our health.
Advertisement
But amid the warnings, there are still many open questions. Are UPFs really bad for you? If so, why? In fact, what exactly are ultra-processed foods anyway? Sprinkle in the myriad social and economic issues intimately associated with the purchase of said foods (see 鈥淯ltra-processed do鈥檚 and don鈥檛鈥檚鈥�, below), and it is no wonder everyone is so confused.
In an attempt to get some clarity on the matter, I have spoken to researchers at the forefront of the debate. And while there are no clear answers on UPFs, it is possible to navigate this nutritional quagmire.
Humans have been to make it tastier, more digestible, more resistant to decay and more convenient. Salting, drying, fermenting, pickling and smoking were invented to preserve foods; milling produced flour to bake bread. Cooking turned unpromising or toxic raw ingredients into tasty, safe and nutritious meals.
During the industrial revolution, however, mechanisation entered the food system. In 1802, the first sugar beet refinery opened in what is now Poland. Ten years later came the first canning factory, in London (sadly not in Canning Town). In 1864, Louis Pasteur invented his eponymous processing technology in which foods are treated with mild heat to eliminate pathogens and extend shelf life. In 1869, margarine was invented. Processed food was on the shelves, and its share of the market has been growing ever since. Around 1950, .
The term 鈥渦ltra-processed food鈥� appeared in the scientific literature in the late 2000s and subsequent studies have linked it to various health conditions. But there was an obvious problem: what exactly are UPFs?
The science of these foods took hold in 2009 when at the University of S茫o Paulo in Brazil and his colleagues introduced a new . This was in response to the growing interest in processed foods and the absence of a clear definition of what they were.
Which foods are ultra-processed?
NOVA (which is Portuguese for 鈥渘ew鈥�) divides foods into four groups: unprocessed and minimally processed; processed culinary ingredients, such as fats, sugars and salt; processed foods, which are created by adding group 2 ingredients to group 1 foods and/or using preservation methods, and, finally, UPFs, making up group 4.
Monteiro defined the last of these as created by the fractioning of whole foods into their constituents, such as sugars, fats and fibres, before these substances are chemically modified and industrially assembled into products, frequently with additives. 鈥淚t looks a recipe for chronic disease,鈥� he says 鈥� and increasingly the science suggests he is right.
Foods in this category include breakfast cereals, instant soups, ready-made pizzas and pasta, biscuits, fish fingers, or fish sticks, reconstituted meat products, ice cream, cakes, packaged breads, carbonated drinks and many more. Basically, a lot of what is in my cupboard.
The system broke new and controversial ground in nutrition science by ignoring the nutrients completely and focusing on the degree of processing. The rationale for this, says Monteiro, was because the extent and purpose of processing of foods is now more useful than the nutrient content in assessing their impact on health.
The UPF industry鈥檚 goal, says Monteiro, is to maximise profits by cajoling consumers to abandon freshly prepared food. 鈥淭hey have affordable prices because of low-cost ingredients, they are convenient, they have long durations, they are engineered to have craving-like palatability and they are aggressively marketed,鈥� he says. In the US, Canada and UK, UPFs have long provided around half of all calories consumed, though their rise seems to have plateaued in the past decade. Consumption is still increasing in as well as middle and low-income ones.
Monteiro says there are easy ways to tell if a product is ultra-processed: it contains at least one food substance rarely used in home cooking 鈥� such as high-fructose corn syrup, chemically modified oils or hydrolysed proteins 鈥� or additives such as flavourings, colourings, glazing agents or thickeners.
Why are ultra-processed foods bad for you?
NOVA was quickly adopted by researchers and numerous papers came out on the correlation between category 4 foods and ill health. But the big breakthrough came in 2019, when at the US National Institutes of Health (NIH) and his colleagues ran a randomised-controlled trial 鈥� the gold standard of evidence gathering 鈥� to .
They recruited 10 men and 10 women with a body mass index of around 27, just slightly over what is deemed 鈥渉ealthy鈥� 鈥� and housed them at the NIH Clinical Center in Bethesda, Maryland, for 28 days. For two weeks, the volunteers were fed a diet high in either UPFs or unprocessed food. In the following two weeks, they received the opposite diet. The regimens were equal in terms of overall calorie content, energy density, protein, carbohydrates, fat, fibre, sugars and sodium, but differed widely in the amount of calories from UPFs. The unprocessed diet delivered 88 per cent of its calories via unprocessed foods, mainly whole fruits and vegetables and some meat, fish, eggs and yoghurt. The UPF diet delivered 81 per cent of its calories in UPF form, such as burgers, fries, canned ravioli and hot dogs.
The volunteers were given three meals a day plus unlimited access to diet-appropriate snacks and 鈥� crucially 鈥� were told they could eat as much or as little as they wanted.
The results were striking. On the UPF diet, volunteers ate around 500 more calories a day than on the unprocessed diet, and after two weeks had gained 0.9 kilograms on average. On the unprocessed diet, they lost 0.9 kilograms, regardless of whether they were given it in the first two weeks or the second. The weight changes were exactly in line with the differences in calorie consumption, says Hall. 鈥淭here鈥檚 no magic here, the changes in body weight are directly related to the changes in calorie intake.鈥� The conclusion is obvious: UPFs 鈥渓ead people to overeat calories鈥�, he says.
When they dug deeper, the researchers discovered that when people were on the UPF diet, they generally ate bigger meals and consumed more fat and carbohydrates.
Since the work, which remains the only published randomised-controlled trial on UPFs, dozens of other studies have found 鈥� not just obesity, but also cardiovascular disease, asthma, cancer, depression, inflammatory bowel disease, fatty liver disease, kidney disease, type 2 diabetes and all-cause mortality. For the latter, , which followed almost 20,000 people between 1999 and 2014, than the lowest consumers.
According to Monteiro, more than 70 high-quality studies have now been published showing a relationship between consumption of UPFs and 20 illnesses.
That may sound like clinching evidence, but there is plenty of wiggle room. Except for Hall鈥檚 randomised-controlled trial, all the data comes from observational studies, where subjects are monitored in their normal life. These produce a lower standard of evidence than a randomised-controlled trial and .
This uncertainty has led some critics to downplay the link between UPFs and ill health. One recent concluded that 鈥渢o date, there is insufficient documentation of a causal role played [in obesity] by processing鈥� and criticised Hall鈥檚 randomised-controlled trial as 鈥渓imited鈥�, in part because the 20 subjects didn鈥檛 find the UPF diet more palatable than the unprocessed one. The lead author of the review disclosed that he is a member of the scientific advisory boards of and the , although no funding was received for the study.
Monteiro dismisses these critics. But he accepts that the causal mechanisms remain unclear, beyond the obvious one of UPFs delivering poorer nutrition via high calorie and salt loads and reduced fibre, vitamins, plant chemicals and micronutrients. 鈥淚 suspect there鈥檚 more than one mechanism linking UPFs with different diseases,鈥� he says.
Last year, a team led by at the University of Paris, France, that could contribute to the unhealthiness of a high-UPF diet. One is that extreme processing makes the foods softer and easier to consume, digest and absorb, and hence promotes overeating. Indeed, Hall found in the randomised-controlled trial that when people were on the UPF diet, they ate faster, at a rate of 50 calories per minute versus 30 on the unprocessed diet.
This is perhaps the most pressing question surrounding UPFs: is the processing itself partly responsible for their health effects? Monteiro thinks so and some others agree. In April, doctor and academic , who is author of Ultra-Processed People: Why do we all eat stuff that isn鈥檛 food鈥� and why can鈥檛 we stop?, told 快猫短视频: 鈥淲hen you separate [whole foods] into their molecular components and chemically modify them, they seem to interact with the body in a very different way.鈥�
But others aren鈥檛 swallowing it. 鈥淲e think it鈥檚 unclear, due to limitations in the available evidence, whether the associations are due to the nutritional characteristics of the food or whether there鈥檚 any independent effect of the processing,鈥� says at Queen鈥檚 University Belfast, UK, speaking in his capacity as chair of the UK鈥檚 . 鈥淲e believe it鈥檚 mostly due to the nutritive and sensory properties of those foods, not a specific characteristic of their degree of processing,鈥� says , who was involved in Hall鈥檚 randomised-controlled trial and is working with him on another to probe further.
Another of Srour鈥檚 proposed mechanisms is that UPFs could be contaminated with toxins leached from packaging or created during processing and heating, such as acrolein, acrylamide, trans fats and . These are currently a major focus of dietary epidemiology, according to , also at Wageningen University.
Additives could be a culprit, too: there are and most aren鈥檛 under suspicion, but there is . But Feskens is sceptical. 鈥淚t鈥檚 difficult to imagine how a small amount of emulsifier can have such an impact,鈥� she says.
Ultra-processed foods and your microbiome
UPFs may also disrupt the microbiome. Human guts contain trillions of microbes that play a crucial role in digestion, metabolism and immunity. Studies in mice and humans have shown that .
One stinging criticism of the whole debate over UPFs is that this is just old science with a new label 鈥� that NOVA is simply a new way to categorise foods that are high in salt, sugar, fat and refined carbohydrates, which are already known to be unhealthy in large quantities. 鈥淚s this really anything new?鈥� asks Hall.
Another criticism of NOVA category 4 is that it captures foods that are actually pretty healthy. A supermarket wholemeal loaf is classed as ultra-processed because it contains additives, according to Feskens. An equivalent loaf from an artisan baker would feature in category 3 despite being nutritionally almost indistinguishable, she says. In fact, it may be nutritionally inferior because supermarket bread is often fortified with nutrients. Around 55 per cent of Dutch fibre consumption comes from UPFs, largely bread, she says, so in that respect the NOVA category 4 is unhelpful. 鈥淣ot all UPFs are equally unhealthy,鈥� says Feskens.
According to Srour, of more than 220,000 group 4 products available in France, according to . Some researchers, including Feskens, think that NOVA鈥檚 category 4 could be split in two to differentiate between 鈥済ood鈥� and 鈥渂ad鈥� UPFs.
NOVA is also subjective, with experts sometimes disagreeing wildly on the category foods should be in. A asked more than 300 nutrition professionals to categorise 231 different foods; there was unanimous agreement on only four. This calls into question NOVA鈥檚 usefulness as a tool to make dietary recommendations or policy, says Forde.
Nonetheless, Monteiro stands by his classification and says there is now enough evidence to take public health action. His strategy is based on efforts in tobacco control: provide reliable information in dietary guidelines, restrict or prohibit advertising, put warnings on packaging, ban sales in schools and tax the products heavily while subsidising unprocessed and minimally-processed foods. Indeed, his native Brazil has .
That, however, may be counterproductive. Demonising category 4 foods or restricting access to them will deprive a lot of people of nutritious, cheap and convenient foods. It may also make the industry less inclined to develop better processing techniques to enhance the nutritional quality of foods, says at the Quadram Institute in Norwich, UK. 鈥淭here is a real risk of demonising foods that are otherwise nutritionally adequate or even beneficial,鈥� says Forde.
How much ultra-processed food should you eat?
So where does that leave us? 鈥淢ost people would benefit from a reduction in their intake of UPFs, purely based on their nutritional content,鈥� says Young. But how much is too much? 鈥淭hat鈥檚 a big question,鈥� says Monteiro. 鈥淲e鈥檝e been looking for this threshold. The data says that the mechanisms are dose dependent. Five to 10 per cent of calories [from UPFs] won鈥檛 have a big effect on health, but at 20 to 30, the problems start to appear.鈥�
Ultimately, for most consumers with busy lives and finite food budgets, UPFs are hard to avoid altogether and rather handy to have. My store cupboards tell a story: I try to cook from scratch, but often don鈥檛 or can鈥檛. I confronted the freezer and found a ready-made pizza. I will definitely eat it. With a salad.
ULTRA-PROCESSED DO'S AND DON'T'S
DO cut down
Keep your calories from ultra-processed foods under 20 to 30 per cent of your total calorie consumption. This is the point at which studies see a link between UPFs and ill health (see main story).
DO diversify
Alongside any UPFs, eat 30 whole foods, such as pulses, grains and vegetables, per week. This is linked to reduced inflammation associated with better physical and mental health.
DO be discerning
Not all ultra-processed foods are as unhealthy as each other. Seeded, wholegrain bread from a supermarket may be fortified with essential nutrients and be nutritionally more beneficial than a home-made white loaf.
DON'T panic
Studies linking additives such as acrylamide or emulsifiers to cancer and other conditions haven't been substantiated in human trials.
DON'T judge
Ultra-processed foods are safe, convenient and cheap, and can often supply much-needed nutrition to families. Replacing UPFs with home-cooked meals where you can is a more sustainable and healthy change for most people.
Graham Lawton is a聽features writer at 快猫短视频
