What Counts As An Ultra-Processed Food? It's A Surprisingly Tricky Question – And It May Not Even Be The Right One To Ask

Eating these “UPFs” might give you cancer, we’re warned; they could be messing up your heart health; in fact, they’ve been linked to damage in every major organ in the human body. They may make you age faster; they can even make you depressed. 

Even worse, they’re all but unavoidable. More than half of the calories consumed by US adults today – and more than two-thirds of those consumed by kids – come from these ultra-processed foods. It is, it seems, a pretty bleak picture.

Except – well, what if it’s not?

Fudgy definitions

The common consensus is clear: ultra-processed foods should be avoided. But that rather raises the question: what are ultra-processed foods, anyway?

It’s a surprisingly difficult question to answer. “There is no standardized definition of UPFs,” points out Faidon Magkos, a professor in the University of Copenhagen’s Department of Nutrition, Exercise and Sports. “There are different classification systems […] and definitions vary.”

The most common explanation – the one you’ll most likely see as a layperson – is this: that a food is ultra-processed if it was produced in a factory or includes ingredients not normally found in a home kitchen. 

It’s a simplified version of what’s called the Nova classification, which sees ultra-processed foods as the final level of increasingly processed substances.

The idea of dividing foods into simple categories of ‘good’ and ‘bad’ is understandably appealing.

Dr Faidon Magkos

“It's beyond just the chemistry of food,” explains Duane Mellor, a senior specialist dietitian and honorary associate professor at the University of Leicester’s Diabetes Centre. 

“They go through ‘unprocessed’ or ‘minimally processed’ foods – these are things like, you know, fresh fruit, possibly frozen fruit, meat, those sorts of things. Then you've got ‘culinary ingredients’, which is really like sugar and salt.” 

“Then you have ‘processed foods’, which is basically the act of anything that's been prepared using standard ingredients,” he continues. “And then ‘ultra processed’ is where it's prepared in an industrial scale, possibly with ingredients which wouldn't normally be used in a traditional kitchen environment.”

If it sounds simple, that’s by design. That initial intuitiveness, though, kind of breaks down on interrogation – does, for example, your store-bought wholemeal bread count as ultra-processed? 

According to the Nova system, it does: in fact, “most supermarket bread will be ultra processed,” says Mellor. “It doesn't matter if it's white or wholemeal […] any manufactured bread will have certain compounds in there to improve the baking process.”

The same goes for, say, fruit yogurt – we’re assuming you don’t have a ready supply of Lactobacillus bulgaricus and Streptococcus thermophilus in your kitchen cupboards – or any fortified cereal. Hummus and jars of pasta sauce fall similarly foul of the description. Cereals are fortified with vitamins and iron; milk has vitamin D added; chocolate, without the addition of the emulsifier and surfactant soy lecithin, will separate and expire quickly.

The distinctions can run to the absurd: “Technically, beer and wine is a ‘processed food’,” Mellor points out, “[but] spirits like whiskey and vodka are ‘ultra-processed’.”

If you’re wondering whether things are more clear-cut in the clinical world, rest assured: they’re actually even worse. 

Pick any two research papers concerning UPFs, and chances are the definitions used will be at least slightly different from each other – if, indeed, they’re included at all. Sometimes, it’s based on how the food was made; other times, on the ingredients; some studies will even base it, at least to an extent, on the supposed intentions of the food company’s marketing or design team – something that’s both unquantifiable and rather irrelevant to a food’s nutritional value.

“[UPF] is a very neat, simple category, which you can plug into a big data set to see if there's differences in disease effects across type 2 diabetes, mental health, cardiovascular disease, some cancers […] it's very nice and simple,” says Mellor. “The problem is, reality is not that simple.”

A banquet of evidence

From a legislative standpoint, it’s tricky to regulate something without a clear idea of what it is. But simply taking, say, the Nova classification as the definition would inevitably create pushback, and not only from the notoriously litigious food industry.

“Some legislation has been discussed about, ‘oh, we need to start taxing ultra-processed foods’,” says Mellor, “and it’s like, how? You can't define it!”

“UPFs have become a new nutritional ‘boogeyman’,” agrees Magkos. “Energy density, texture, and fiber content are properties of foods themselves, not properties of the processing category. It would […] be incorrect to argue that ultraprocessing inherently causes foods to be softer, more energy-dense, or lower in fiber.”

“You can easily design a diet composed largely of minimally processed foods that is soft in texture, high in energy density, and low in fiber,” he points out. “Equally, you can create a UPF-rich diet that is low in energy density, high in fiber, and requires substantial chewing.”

The risk of this debate is we're focusing on the wrong end of the argument.

Dr Duane Mellor

Indeed, not all studies are agreed that UPFs are universally “bad”. One of the best-quality trials out there right now, the UPDATE trial run by University College London’s Center for Obesity Research, found that a diet high in UPFs actually resulted in significantly lower body weights, BMIs, cholesterol levels, and heart rates among adherents. 

And yes, those reductions were higher for the group given minimally-processed diets – but that UPFs are somehow uniquely and only detrimental to health, this does not suggest.

Meanwhile, many studies that have indicated a serious problem with UPFs are, let’s say, often less watertight than they may seem at first glance. That’s not really surprising: almost by definition, it’s extremely difficult to design two diets identical in every way apart from the amount of processing that went into them. Those studies that get close are often stymied in other ways: they’ll have a very small sample size, for example, or only last a short time, limiting the strength of any results.

What this adds up to, unfortunately, is something of a dilemma for public health and nutrition experts.

“People say, ‘we have whole grain bread […] we have yogurt, and we think of these as healthy foods, but they're also ultra processed. So clearly we don't mean those,” says Jerry Mande, Adjunct Professor of Nutrition at the Harvard T.H. Chan School of Public Health and a public health and food policy advisor for the Obama, Clinton, and Bush administrations. 

“And I'm like, no, we don't ‘clearly not mean those’. We just really don't know.”

“The idea of dividing foods into simple categories of ‘good’ and ‘bad’ is understandably appealing, especially when people are frustrated by hearing that nutrition is complex,” counters Magkos. But “focusing on processing category risks discouraging the consumption of many nutritious foods, while drawing attention away from the dietary characteristics that are more likely to matter for health.”

The policy problem

Regulation around UPFs is, more than almost any other topic in the US today, a highly bipartisan issue. But with this mishmash of evidence, along with the lack of a workable definition of UPFs, the creation of such policies is destined to be fraught.

“You know, it's interesting,” says Mande. “We have this Dietary Guidelines process that we use every five years to set the government policy on nutrition. And […] this cycle, it happens to cross two administrations – the Biden administration did the first half of the work, and put together the advisory committee, had them meet, put out a report, and then the Trump administration has now taken that advisory committee report and written the dietary guidelines.”

“So, on this particular issue of ultra-processed food, the Biden advisory committee concluded that there wasn't enough evidence to make any statements about ultra-processed food,” he explains. “So had Biden or Kamala Harris been elected president, it's a sure thing that the dietary guidelines would not have really mentioned ultra-processed food at all.”

In contrast, the Trump administration has made UPFs a flagship issue, at least nominally. “[They’ve] put it front and center,” says Mande. “The main recommendation in their Dietary Guidelines is that we should eat less ultra-processed food.”

The time cost of preparing the food, and the knowledge to do it […] that is something that a lot of people don't have. And that's real.

Jerry Mande

Now, far be it from us to suggest handing it to RFK Jr – his focus on reforming nutritional standards in the US is, after all, unlikely to be the result of serious scientific research, and stands notably at odds with his department’s recent relaxation of regulations around peptides, hormones, and food safety in general. 

Still, the Trump administration’s purported push for better nutrition and higher-quality food in the US is one that’s been overall welcomed by establishment science. 

“But the question is, what are they going to actually do?” says Mande. “And that's been more unclear.”

After all, a full crackdown on UPFs could mean, for example, taking the vitamin and iron fortifications out of your flour and milk. It could make your kids’ wholegrain cereal much less healthy, or more expensive. Americans in general may support limits on UPFs, but in practice, when living costs are as tight as they are, the results may not be popular. 

“People often say it's expensive to eat,” says Mande. “The time cost of preparing the food, and the knowledge to do it […] that is something that a lot of people don't have. And that's real.” 

“That's why education labels aren't the solution, by the way,” he adds. “Those people who really don't have a choice, you're just reminding them that they're making bad decisions, making themselves sick.”

For the average consumer, food purchases are motivated by four things, Mande explains: taste, cost, convenience, and health. The problem is, one of those considerations is not like the others: “as a consumer, you can come home after a trip to the supermarket, and you can tell within an hour whether the food met your expectation for taste, cost, and convenience,” he points out. “But you have no idea on health.” 

“You won't know for weeks, months, years whether this food's causing diabetes, heart disease, cancer, obesity. You just won't know that.”

For corporations looking to increase profit, then, it’s the obvious thing to compromise on. After all, you make your product less tasty, or more expensive, and people simply won’t buy it. Make it less healthy – and spend decades investing millions of dollars per year in lobbying and lawfare efforts – and you’ll likely suffer little to no backlash at all.

“And that's what's happening in the US and [other countries],” Mande says. “Companies are swapping our health for their profits.”

“And they know they're doing this,” he adds. “That, to me, is the most enraging thing: that these CEOs know they're doing this. They're producing food that they know will make us sick if […] and they're doing it because they know they can get away with it.”

Just desserts

For all the talk of regulations and legislation, and how dangerous UPFs are or are not, perhaps the argument is moot. 

After all, after much hype around their promised crackdown on UPFs, the US Food and Drug Administration ultimately declined to follow through – instead settling for a watered-down regulatory measure that skipped defining “ultra-processed foods” entirely.

To many readers, that won't come as a surprise. Upon entering his second term, Trump pretty much crammed his various federal agencies with industrial lobbyists or stakeholders, including many with ties to Big Food – that is, people with a vested interest in avoiding regulations in the food industry. 

And, frankly, such moves would be wildly out of sync with established Republican tradition in any case: “you know, they sort of stand for, ‘we don't want to use the government levers of regulation or funding. We want to not use those’,” says Mande. “And if you're not going to use those, you're not going to change anything.”

In the meantime, few experts would argue with the longstanding advice to aim for nutritionally dense, high-fiber, whole foods – even if the odd UPF is, at this point, pretty much inevitable.

“We basically know that doughnuts and pies and lots of regular soft drinks […] is not a good thing,” says Mellor. “[But] the risk of this debate is we're focusing on the wrong end of the argument.”

“We’re looking at the ‘bad guys’ rather than focusing on the other end,” he tells IFLScience. “We know we should eat more of the good stuff […] We need to invest in the good message, rather than just tinker around about how we can make this thing that's technically ultra-processed no longer tick that box.”