Loading ...

Stroll down almost any supermarket aisle and the modern foodscape reveals itself in bright boxes, colorful liquids, and long ingredient lists: shelf-stable snacks, sugary breakfast cereals, and frozen meals engineered for convenience.

Partly because of their ubiquity, these products — ultraprocessed foods (UPFs) — are at the center of growing scientific scrutiny. A raft of recent studies has tied heavy consumption of UPFs to various adverse health outcomes, including increased risks for certain cancers and higher mortality among cancer survivors.

However, critics counter that the evidence is largely observational and subject to biases. Perhaps more importantly, focusing on the degree to which foods are processed may not be all that useful.

What’s In a Name?

What makes processed food ultra? Generally, researchers go by the NOVA food classification, which was initially developed in 2009 by Brazilian epidemiologist Carlos Augusto Monteiro. It separates food into four categories, based on the extent and purpose of their industrial processing:

• Unprocessed or minimally processed, which includes fruits and vegetables, nuts, legumes, dairy milk, and meat, for instance; these products may undergo some processing intended to extend their shelf life or make preparation easier: Frozen vegetables, rice, and dried pasta are a few that make the cut
• Processed culinary ingredients, such as butter, vegetable oils, salt, and sugar
• Processed, which are foods from the first category that have salt, sugar, fat or other culinary ingredients from category 2 added to them; examples include salted nuts, canned fish, vegetables packaged in brine, and freshly made bread and cheese
• UPFs

Unlike foods that are merely processed, UPFs typically offer little to no intact whole food, and their ingredient lists boast substances foreign to home cooks such as high-fructose corn syrup, protein isolates, preservatives, cosmetic additives such as artificial and “natural” flavors and colors, or emulsifiers.

The UPF umbrella covers many of the convenience foods long considered “junk” or at least less than healthy: packaged snacks and sweets, mass-produced breads and cheeses, breakfast cereals, hot dogs, and canned soups, to name a few.

However, UPFs also include foods commonly viewed as “healthy,” such as plant-based milks, flavored yogurt, meat substitutes, and instant oatmeal. Meanwhile, full-fat dairy milk gets the health halo of being minimally processed, and nuts coated in a sugary glaze can dodge the UPF label.

So while the NOVA classification is “useful,” it’s far from perfect, said Urvi A. Shah, MD, a hematologist-oncologist who conducts nutrition research at Memorial Sloan Kettering Cancer Center in New York City.

“Ideally, we would want clearer definitions for unhealthy foods beyond their level of processing,” Shah told Medscape Medical News.

Another expert went further, saying ultraprocessed is “not a useful distinction” at all.

“The ‘worst’ [NOVA] category isn’t meant to tell us about whether food is necessarily good or bad for our health,” said Gideon Meyerowitz-Katz, PhD, an epidemiologist at the University of Wollongong in Wollongong, Australia. “It’s meant to tell us about processing, specifically.”

And there’s little evidence that processing methods, per se, are what make certain foods bad for us, Meyerowitz-Katz told Medscape Medical News.

“If I bake cookies myself, are they more or less healthy than buying cookies at the grocery store?” he said. “The answer is maybe.”

The Epidemiologic Evidence

Despite the NOVA system’s imperfect definitions, research using it is gaining momentum. That includes several recent studies focused on cancer.

A study published last year in Thorax used data from the PLCO Cancer Screening Trial to link high UPF intake to increased odds of developing lung cancer over 12 years. Among over 100,000 participants, the quartile with the highest UPF intake at baseline developed lung cancer at a rate of 1.6 per 1000 person-years vs 1.1 per 1000 person-years in the quartile with the lowest UPF intake. After adjustment for factors such as smoking status and family history of lung cancer, high UPF intake was associated with a 44% relative increase in risk.

Another large study, published in JAMA Oncology, explored the theory that UPFs could be a contributor to the rising rate of colorectal cancer in younger adults. It found that among nearly 30,000 women from the Nurses’ Health Study II, those in the highest quintile for UPF intake had a 45% higher risk for colorectal adenomas before age 50 than those in the lowest quintile.

Some research has also begun digging into whether UPF intake could affect long-term survival after a cancer diagnosis. A study published earlier this year found that among roughly 800 cancer survivors, those who consumed the most UPFs had a 48% higher risk for death from any cause and a 57% higher risk for cancer-related death over 15 years, after adjustment for multiple health and lifestyle factors.

But all of those studies had key limitations such as use of self-report or issues in the analysis that could have created a false association between UPF intake and outcomes.

Andrew Chan, MD, MPH, a gastroenterologist at Mass General Brigham Cancer Institute in Boston, was senior author on the JAMA study that looked at colorectal adenomas.

He said that the rapid rise in early-onset colorectal cancer is so far largely unexplained, and it’s critical to understand what modifiable contributors, such as UPF consumption, might be involved.

At the same time, Chan said there are few longitudinal studies that have characterized people’s diets in enough detail, and over an adequate time period, to fully assess the association between UPF intake and long-term health outcomes.

Chan was highlighting some well-known limitations of nutritional epidemiology in general, particularly when it comes to an outcome as complex and long term as cancer.

One is its reliance on food-frequency questionnaires. In the case of UPF research, that presents particular barriers, Shah said, because standard questionnaires “are not built” to specifically ask about UPFs in depth.

“We don’t really have granular detail in those kinds of studies,” she said, noting that it’s unclear whether UPFs, in general, are associated with cancer risk or whether certain ingredients might be.

Again, Meyerowitz-Katz went further. “Food-frequency questionnaires are notoriously unreliable because people are terrible at remembering the things that they’ve eaten,” he said.

It’s hard to pin health outcomes on certain food categories or, even more specifically, processed-food ingredients when you can’t know whether study participants actually ate those things, Meyerowitz-Katz said.

Still, some researchers have tried to get that specific, with recent studies tying some food additives to cancer risk.

Earlier this year, a study published in The BMJ involving more than 100,000 participants found that consumption of certain food preservatives was associated with increased cancer incidence, particularly breast and prostate cancers.

At the time, critics pointed to the same methodological issues that Meyerowitz-Katz raised, including imprecise information on the amount of preservatives study participants consumed.

They also highlighted other red flags, such as the finding that moderate intake of vitamin C (but neither low nor high intake) was tied to increased cancer risks, as well.

What Clinical Trials Say

It’s unlikely there will be a long-term randomized trial that assigns people to eat a diet heavy in UPFs and follows their cancer risk. However, some short-term trials have tried to demonstrate the effects of UPF consumption on more immediate outcomes like calorie intake and weight gain.

One frequently cited clinical trial, published in 2019, randomly assigned people to spend 2 weeks on an “unprocessed” diet and 2 weeks on a UPF diet. Meals were matched for calories and macronutrients, and participants were free to eat as much or as little as they wanted. Overall, the trial found that people ate about 500 calories more daily and gained some weight on the ultraprocessed diet vs the unprocessed one.

“You could argue the RCTs [randomized controlled trials] are more robust” than the observational research, Meyerowitz-Katz said.

But the problem, he explained, is that existing trials have not studied foods that are different only in terms of processing; they’ve compared diets that are on opposite ends of the nutrition spectrum.

A randomized crossover trial published in Cell Metabolism last year examined the effects of a UPF diet vs unprocessed diet on metabolism and sperm quality in 43 healthy men. The main findings: Participants weighed about 3 lb more after 3 weeks on the UPF diet vs 3 weeks on the unprocessed comparator (due to weight loss on the latter) and showed an increase in cholesterol levels.

That’s unsurprising, Meyerowitz-Katz noted, given what participants ate. The unprocessed diet featured green salads, chicken, fish, and snacks of fresh fruit, while the UPF diet offered burgers, gummy bears, chocolate bars, potato chips, and soft drinks.

“There’s a big difference between giving people gummy bears and cake as their snacks and giving them fresh fruit,” Meyerowitz-Katz pointed out. And such trials cannot illuminate what effects food processing, per se, has on outcomes.

Advice for Patients

One fact is indisputable: Americans eat a lot of processed food. And studies looking at ultraprocessed products, specifically, suggest that they make up nearly 60% of US adults’ calorie intake.

Based on that, and the science thus far, it’s certainly reasonable to encourage patients to eat fewer UPFs, according to Chan.

As for addressing confusion about what the term means, he said: “One of the things that I tell my patients is that if the food you’re eating is something that you could not have prepared in your own kitchen using standard ingredients, it’s probably ultraprocessed.”

According to Chan, it’s not about singling out specific foods as “bad,” and he said diet should be viewed in a holistic way. For some people, cutting back on UPFs may be a straightforward way of improving their overall diet quality.

Similarly, Shah said it helps to read labels because processed foods are clearly not all the same. She recommended checking nutrition labels for fiber, saturated fat, added sugar, and salt content — and scanning the ingredients to see if there’s a long list of additives and other substances not found in a typical home kitchen.

In the end, Meyerowitz-Katz believes that the current focus on UPFs in research and public discussion simply comes back to the time-honored advice to eat less junk food.

“If I was in charge of American dietary guidelines, I might mention [UPF] as something that you could avoid,” he said. “But I think that the general public already knows which foods to avoid.”

Meyerowitz-Katz had no financial disclosures. Chan reported receiving personal fees from Pfizer Inc. and Boehringer Ingelheim. Shah reported being a principal investigator on the NUTRIVENTION trial and receiving research funding and/or personal fees from Celgene/BMS, Janssen, and Sanofi.