Weight gain, heart disease, type 2 diabetes, cancer, memory loss, depression, Crohn’s disease. The list of maladies linked to ultra-processed foods seems endless. Are those foods actually to blame? Here’s what we do—and don’t yet—know.

1. What’s ultra-processed? You may be surprised.

What makes a food ultra-processed? Nearly all studies use the NOVA system to decide (see below).

NOVA divides foods not according to their nutrients or impact on health, “but according to the extent and purpose of industrial processing,” explained NOVA creator Carlos Monteiro, professor of nutrition at Brazil’s University of São Paulo, at a June webinar held by “A Healthy Diet for a Healthy Life,” an initiative partly funded by the European Union.

And the purpose of ultra-processed foods, noted Monteiro, is “to replace all the other NOVA food groups, traditional dietary patterns, and then at the same time maximizing industry profits.”

That means using “cheap food substances and additives,” he added.

“They have affordable prices because of low-cost ingredients, they are convenient...they are designed to be ready to consume and to have long durations. They are engineered to have craving-like palatability. And they are aggressively marketed.”

Indeed, ultra-processed foods now comprise more than half of the average American’s diet.

But they’re not all what you’d call “junk foods.” When researchers looked at the ultra-processed foods eaten by roughly 200,000 health professionals, the category that contributed the most was “breads and breakfast foods.” That included nearly all packaged cereals and breads, whether white or whole-grain, raisin bran or Froot Loops.

And the second largest category was “fats, condiments, and sauces,” like ketchup, salad dressings, mayonnaise, soy sauce, and spreads.

Who would have guessed?

Granted, the third and fourth largest categories—“sweet snacks and desserts” and “beverages”—could easily qualify as junk food.

Although the definition of ultra-processed foods isn’t based on harm—and it’s not clear which of their many features might be harmful—studies have linked them with disease after disease.

The question is: Why?

Where we get ultra-processed foods

---

2. Are ultra-processed-food eaters different?

Dozens of studies that tracked thousands of people for years have linked ultra-processed foods with a higher risk of obesity, type 2 diabetes, cardiovascular disease, breast cancer, colorectal cancer, memory loss, non-alcoholic fatty liver disease, kidney disease, Crohn’s disease, depression, frailty, the risk of dying of any cause, and more.

“How can ultra-processed foods underlie so many different diseases?” asked Edith Feskens, professor of global nutrition at Wageningen University in the Netherlands, at the webinar in June.

“Is it something else? Is it maybe the cost?” she suggested. “Is it socioeconomic position?”

Researchers do try to take those and other possible “confounders” into account.

For example, “people who consume a high level of ultra-processed food may also be less physically active or may consume more alcohol or be more likely to smoke,” says Fang Fang Zhang, associate professor of epidemiology at Tufts University.

“We adjust for those and many other factors, but something we don’t know about could also play a role.”

The only way to eliminate confounders is to randomly assign people to eat either unprocessed or ultra-processed foods.

“But diseases like cancer take decades to develop,” notes Zhang. “Unless you do a study for 20, 30, or 40 years, it’s hard to measure cancer in randomized controlled trials.”

Instead, she looked at the diets of about 200,000 health professionals who were tracked for 24 to 28 years.

“A higher consumption of ultra-processed foods overall was associated with an increased risk of colorectal cancer in men, but not in women,” says Zhang.

But the results varied by food group.

“We found that a higher consumption of ready-to-eat meats—for example, hot dogs, salami, sausages—and sweetened beverages was linked to a higher risk of colorectal cancer in men.”

“In women, ready-to-eat mixed dishes—frozen pizza, for example—were linked to a higher risk, and ultra-processed yogurt and dairy desserts were linked to a lower risk.”

What could explain that lower risk?

“Calcium or milk may be protective for colorectal cancer,” suggests Zhang.

In another study, health professionals who ate the most ultra-processed foods overall had a 28 percent higher risk of type 2 diabetes than those who ate the least.

But, as Zhang found, some food groups (in this study, ones like refined breads, sweetened beverages, and ready-to-eat mixed dishes) were linked to a higher risk, while others (like whole-grain breads, cereals, and yogurt and dairy desserts) were linked to a lower risk.

But even if only some ultra-processed foods cause harm, we can’t ignore them, argues Zhang.

“Adults get 57 percent of their daily calories from ultra-processed foods,” she notes. “The evidence on cancer is limited, but with so many studies showing a link with obesity, type 2 diabetes, and heart disease, we need to pay attention to ultra-processed foods now.”

3. Why ultra-processed foods may pile on the pounds. 

Only one well-controlled trial has tested what happens when people are randomly assigned to eat either ultra-processed or unprocessed foods.

“We wanted to test whether ultra-processed foods—independent of their salt, sugar, fat, and fiber content—can drive overeating,” says Kevin Hall, a researcher at the National Institute of Diabetes and Digestive and Kidney Diseases.

For two weeks, 20 people were offered meals and snacks consisting largely of either ultra-processed foods or minimally processed foods.

“To my surprise, people ate about 500 calories more per day on the ultra-processed diet than they ate on the unprocessed diet,” says Hall.

After two weeks on the ultra-processed diet, the participants had gained about two pounds. And after two weeks on the unprocessed diet, they had lost two pounds. 

“The question is: Why?” says Hall.

When his team started to dig into the data, some explanations came up empty. For example, “there weren’t large blood sugar spikes on the ultra-processed diet,” he notes.

But two possibilities stood out: calorie density and hyper-palatability.

Calorie density

People eat more when they’re offered foods that have more calories per bite (like potato chips and chocolate) than when offered foods that have fewer calories per bite (like fruits and vegetables).

So the diets were matched for calorie density...but only if you include both the food and drinks.

“If you take out the beverages, the ultra-processed foods had a much higher calorie density,” says Hall.

“And if beverages don’t contribute to satiety as much as solid foods do, you might expect ultra-processed foods to lead to weight gain.”

Hyper-palatability

Research by University of Kansas assistant professor Tera Fazzino suggests that foods are hyper-palatable if they are high in:

• fat + sugar (like cake or ice cream), 
• fat + sodium (like bacon or pizza), or
• carbs + sodium (like pretzels, popcorn, or crackers).

Sure enough, more foods on the ultra-processed diet fit the bill.

“Calorie density and hyper-palatable foods each appeared to explain about 40 percent of the effect of ultra-processed foods on calorie intake,” says Hall. “But to prove it, you need to do a new study. We have one underway and hope to have an answer in the next year or two.”

4. Does the food’s structure matter?

In a 2022 study, people ate more calories from all-you-can-eat lunches made with ultra-processed rather than minimally processed foods. But they also ate more calories when foods were soft rather than hard.

A new, not-yet-published study tested both processing and whether foods were “fast” or “slow.” “Slow” meant harder, chunkier, more viscous, or more solid.

For “breakfast, for example, unprocessed slow was a fresh mixed fruit bowl,” Wageningen University’s Edith Feskens told the June webinar. “In the fast version of the unprocessed food, it was a homemade smoothie.”

The fast ultra-processed breakfast “was a store-bought smoothie with some, let’s say, additives for increasing the shelf life,” she added. “And in the slow ultra-processed, it was a can of mixed fruit.”

The results: People ate 22 percent more calories per day when offered the “fast” rather than the “slow” foods, regardless of whether they were unprocessed or ultra-processed.

“This indicates that [processing] doesn’t matter so much,” suggested Feskens. “It is the texture.”

Of course, most ultra-processed foods are “fast.” And even simple processing can make foods less filling.

For example, people ate less at lunch when given peeled apple wedges instead of applesauce before lunch was served. Apple juice curbed their appetite the least, even when scientists added an apple’s worth of fiber to the juice.

In a 2020 study, MRI scans of the stomach helped explain why. It took longer for whole apples to exit the stomach than either purée or juice.

So ignore claims like “5 g fiber” on ultra-processed foods. It’s not just the fiber, but the intact food structure—in whole fruits, vegetables, beans, whole grains, and nuts—that matters.

5. Could additives or contaminants be to blame? 

Scientists have a long list of potential culprits that may explain the links between ultra-processed foods and disease. They include contaminants from packaging (like phthalates) or processing (like acrylamide) and food additives (like food colorings).

Among the top additive contenders: carboxymethylcellulose (often called cellulose gum) and polysorbate 80. Like other emulsifiers, they’re found in heavy whipping cream, ice cream sandwiches, pickles, relish, and more.

But so far, the evidence comes largely from mice given high doses (in their water).

“Feeding polysorbate 80 and carboxy­methylcellulose to mice changes their gut microbiota and promotes a state of low-grade inflammation,” explains Andrew Gewirtz, distinguished university professor at the Institute for Biomedical Sciences at Georgia State University.

The mice also ate more food and gained weight, and their blood sugar levels rose above normal.

How might the two emulsifiers lead to weight gain?

“We think they impede satiety signaling,” explains Gewirtz.

“Normally, when somebody eats a meal, signals tell the brain to stop eating. But in states of inflammation, the receptors of those signals don’t function properly.”

That wasn’t the only problem. Gut bacteria also encroached on the layers of mucus that ordinarily protect the intestinal lining from those bacteria.

“And in mice that were genetically predisposed to develop colitis, the emulsifiers hastened the onset of colitis,” says Gewirtz. “Colitis is a severe inflammation of the colon similar to Crohn’s disease.”

But evidence in humans is sparse.

Researchers have reported a higher risk of Crohn’s disease in people who eat the most ultra-processed foods. But emulsifiers may not be to blame.

And in a small study, Gewirtz’s team randomly assigned 16 people to get foods with or without a high daily dose (15 milligrams) of carboxymethylcellulose.

After 11 days, the gut bacteria had encroached on the layer of mucus protecting the intestinal lining in two of the seven people who got the carboxy­methylcellulose.

“We think some people can eat large amounts with no problems but others are sensitive,” says Gewirtz.

He’s not worried about all emulsifiers.

“For example, in mice, soy lecithin is less harmful, maybe because it’s a natural component of many foods, so we’ve evolved eating it.”

The upshot: “More testing is needed,” says Gewirtz. “But our research in mice indicates two things that are detrimental about ultra-processed foods. One is emulsifiers, and the other is their lack of fiber.”

He’s talking about the fiber in fiber-rich vegetables, fruits, beans, whole grains, nuts, and seeds, not ultra-processed foods with added fibers like inulin.

Gewirtz’s bottom line: “Limit highly processed foods both because of the additives they contain and the natural components of food they lack.”

6. Do ultra-processed foods alter your gut microbes?

Ultra-processed foods may also cause harm by failing to feed our gut microbes. Why? Their nutrients are absorbed so quickly that they never reach the microbes that are in the colon.

“A typical U.S. diet doesn’t have enough fiber and nutrients for our gut microbes to do what they’re meant to do,” says Rosa Krajmalnik-Brown, director of the Biodesign Center for Health Through Microbiomes at Arizona State University.

“Gut microbes are meant to ferment the undigested food we eat. That produces short-chain fatty acids like butyrate, which feeds the cells that line the colon.”

Butyrate may also boost satiety hormones like GLP-1 and PYY and curb high blood sugar levels.

So her team created a “microbiome enhancer diet” with four features:

• few ultra-processed foods,
• large food particles (think whole nuts rather than nut butters),
• high-fiber foods, and
• high levels of resistant starch. (Resistant starch isn’t broken down by our digestive enzymes, so it ends up in the colon. Beans, pasta, and potatoes are rich sources, though the study also used a specialized corn starch to boost levels.)

“Our key goal was to see if the micro­biome enhancer diet would send more food to the gut microbes than a typical Western diet,” says Krajmalnik-Brown.

It did. “On the microbiome enhancer diet, the gut microbes got more calories, they grew more, and they produced more short-chain fatty acids, and we saw more calories in the fecal samples.”

On average, people on the microbiome enhancer diet excreted roughly 100 more calories per day.

“The participants were getting fewer calories because some were going to the microbes,” explains Krajmalnik-Brown.

More studies are needed. Until then, “eat whole, unprocessed foods with a lot of fiber to feed your microbes.” 

The bottom line

Fill half your plate with whole fruits and vegetables, and the rest with beans, whole grains, nuts, fish, poultry, or dairy.

Love your flavored yogurt, meatless burgers, or ice cream? Don’t worry. In some studies, people who ate the fewest ultra-processed foods—and had the lowest risks of disease—still averaged three servings a day of foods like those.