This morning, while tidying up my office, I found an open box of packaged coconut and chocolate cookies that I'd bought sometime last year. The "use by" date had come and gone more than eight months ago. Curious, I took a small bite. They still tasted pretty darn good.
A closer look at the ingredient list revealed some things I've certainly never baked with, including carrageenan and sorbitan tristearate, additives used to do things like thicken, emulsify and preserve the flavor and enhance the texture of food.
Welcome to the world of ultra-processed foods – edible products made from manufactured ingredients that have been extracted from foods, processed, then reassembled to create shelf-stable, tasty and convenient meals.
"These are foods that are industrial creations," says Allison Sylvetsky, an associate professor in the department of exercise and nutrition at the George Washington Milken Institute School of Public Health.
And we're eating a lot of them. Ultra-processed foods currently make up nearly 60% of what the typical adult eats, and nearly 70% of what kids eat.
The category includes everything from cookies and sodas to jarred sauces, cereals, packaged breads and frozen meals, even ice creams. You might not realize you're eating one, but look close and you'll see many ingredients you wouldn't find in your kitchen – think bulking agents, hydrolyzed protein isolates, color stabilizers, humectants.
They dominate the food supply. And a large and growing body of evidence has consistently linked overconsumption of ultra-processed foods to poor health outcomes.
"Four of the top six killers are related to an inadequate diet, which in the U.S. is probably largely due to convenient, safe, inexpensive food that we eat too much of," says Christopher Gardner, the director of nutrition studies at Stanford University, who has spent decades studying the links between diet and chronic disease. "Too much of it leads to obesity and type two diabetes and heart disease and cancer."
Gardner says the emergence of ultra-processed foods led to products that were inexpensive and safe to eat for longer periods of time. "But it just went too far."
High consumption of ultra-processed foods has been linked to health concerns ranging from increased risk of obesity, hypertension, breast and colorectal cancer to dying prematurely from all causes.
While there's clearly a link to health, researchers are still not completely sure what it is about this category of food that appears to make us sick. But one researcher, Kevin Hall, has a hunch.
One reason ultra-processed foods likely contribute to health issues seems obvious: They tend to be low in fiber and high in calories, salt, added sugar and fat, which are all linked with poor health outcomes when eaten in excess.
But back in 2009, researchers in Brazil proposed that it wasn't just the nutrients that made these foods unhealthy, but rather, the extent of processing these foods undergo.
Kevin Hall, a senior investigator at the National Institutes of Health, where he studies obesity and diabetes, says when he first heard this theory, he was skeptical.
"It struck me as a really odd way to think about nutrition science, because, after all, nutrients seem to be important for nutrition," says Hall.
Hall designed a study to find out whether there was anything to that theory. Now, most studies that have linked ultra-processed foods to harmful health effects are observational, which can only show correlations – they can't prove cause and effect. What Hall did was different: In 2019, he put together the first randomized controlled trial to compare the ultra-processed diet to one based on less processed foods.
Hall used the NOVA classification system — developed by the Brazilian researchers — which categorizes foods based on how much processing they undergo. It distinguishes between unprocessed or minimally processed foods, like an ear of corn or frozen peas; processed foods – like tuna canned in oil or smoked meats, which generally have two or three ingredients; and ultra-processed foods, which are created with formulations of ingredients made using industrial techniques.
Hall recruited 20 healthy adult volunteers to stay at an NIH facility for a four-week period. Participants were randomly assigned to either an ultra-processed or minimally processed diet for two weeks at a time, then switched to the other diet for another two weeks. People on the ultra-processed diet were fed meals like canned beef ravioli, chicken salad made with canned chicken, tater tots and hot dogs. The unprocessed diet mainly featured fruits, vegetables and unprocessed meats – think baked cod served with a baked potato and steamed broccoli.
Both groups were served twice as many calories as they would need to maintain their body weight, and they were told to eat as much or as little as they wanted. Both diets were nutritionally matched, so each meal contained essentially the same total amount of fat, sugar, salt, fiber, carbohydrates and protein.
The results took Hall by surprise.
"I had sort of expected that ... there wouldn't be any difference," says Hall. "But in fact, what we saw was that when they were on the ultra-processed diet, they were eating about 500 calories per day more than when they were on the unprocessed diet and they were gaining weight and gaining body fat" – they put on about 2 pounds on average.
On the unprocessed diet, people ended up eating less and lost weight.
The findings strongly suggested that it wasn't just salt, sugar and fat, but something about the highly processed nature of these foods itself that was propelling people to overeat and gain weight. Gardner, who was not involved in that study, agrees that there's a signal in the data that needs to be explored. "There's something there," he says.
But what, exactly, is that something?
Hall is currently running another randomized control experiment designed to help suss that out. This time, he is offering participants variations on ultra-processed diets to hone in on why we overconsume them.
He says one reason might be that these foods tend to pack more calories per bite, in part because they often have water removed from them to make them shelf stable. Or it may be that they tend to feature irresistible combinations of fat, salt, and sugar more frequently – and in higher levels – than normally occur in unprocessed foods.
"If we can figure out what it is about ultra-processed foods that drives people to overeat and gain excess weight, then we can at least then target which ones to avoid," says Hall – and perhaps eventually, figure out how to re-engineer these foods into healthier products that still have the benefit of being cheap and convenient.
"I think that's really a path forward" for packaged foods, says Hall.
Even among the range of processed foods currently on the market, "not all ultra-processed foods are bad," says Dr. Fang Fang Zhang, a nutritional epidemiologist at the Friedman School of Nutrition Science and Policy at Tufts University who has studied the rise of ultra-processed consumption and correlated health risks.
For example, she points to packaged whole grain breads. Under the NOVA classification system, most mass-produced breads are considered ultra-processed. But if they have little or no added sugars and high levels of fiber, they could be a healthy option, Zhang says. Her research has found that yogurt consumption was linked to a lower risk of colorectal cancer in women, even though the yogurts in the study qualified as ultra-processed because they contained added sugar.
Those kinds of examples are one reason why some critics want to see the NOVA classification system tweaked to allow for more nuanced distinctions, so that not all ultra-processed foods are vilified.
Even so, Zhang says there is enough research linking ultra-processed foods to health issues that it makes sense to try to cut back on our consumption in general – especially when these foods are high in salt, sugar and fat.
And that goes for kids, too, she says. While the evidence is limited linking ultra-processed food consumption in children to health concerns, Zhang notes the evidence is pretty strong in adults – and "dietary habits in children often carry over into adulthood." So cutting back now could help set kids up for better health down the road.
Christopher Gardner of Stanford agrees, though he'd prefer to focus the message on what people should be eating instead. "If you're going to tell them what to avoid, Americans are often clever enough to choose something else that's just as bad or worse," Gardner says.
He worries, for example, that if a family on a tight budget hears that jarred spaghetti sauce is ultra-processed, instead of opting for a fairly healthful homemade meal of pasta served with jarred sauce with vegetables and lean meat tossed in, they might opt for fast food.
His advice? Focus on eating a diet that's primarily plant-based – though that doesn't have to mean no meat whatsoever. And if you're considering eating a packaged food, read the ingredient list. "If you really have no idea what some of those ingredients are, it probably went too far," Gardner says.
Gardner sits on the independent advisory committee that reviews scientific evidence and makes recommendations that will inform the development of the next iteration of the U.S. dietary guidelines to be issued in 2025. He says ultra-processed foods are one of the issues they're considering. Some countries, including Brazil, Peru and Uruguay, have dietary guidelines that specifically recommend freshly prepared meals and avoiding ultra-processed foods.
Ultimately, Zhang says the burden of making healthier food choices shouldn't fall solely on consumers – especially when it comes to setting kids up for better health long term.
"It does need the whole society to pay attention to this, to work together, including the food industry, including the government, to be able to reduce the amount of ultra-processed food our kids are consuming in their day," she says.
