Shopping for yogurt, bread and granola bars might feel like a healthy decision. The dairy seems like a calcium-boosting choice for kids, the whole-grain bread looks better than the white bread, and granola bars appear so much better for you than chips or gummy bears – and in many ways, they are.
But a growing number of grocery-store foods – even ones that appear healthy – are what scientists today call “ultra-processed”: fruit-flavored yogurts packed full of sugars, flavorings and thickeners like guar and carob bean gum; or packaged bread, with ingredients like soy lecithin and monoglycerides slipped in alongside the flour and water.
These industrially formulated products, which are often high in fats, starches, sugars and additives, make up 73% of the US food supply today. Yet research is increasingly tying ultra-processed foods to a myriad of health concerns, like diabetes, obesity, cancer and depression. Despite those risks, the average American gets more than 60% of their daily calories from ultra-processed foods – more than in any other country in the world.
A new way of thinking about nutrition
The phrase “ultra-processed foods” made its first appearance in 2009 when the Brazilian nutritionist Carlos Augusto Monteiro published a paper that made a bold argument: “The issue is not food, nor nutrients, so much as processing.”
Most food is processed in some way, whether that’s an apple that’s been waxed to shine brighter at the grocery store or milk that’s been pasteurized for safety. Processing in and of itself isn’t bad – in fact, vitamin enrichment and preserving techniques (like canning and fermenting) made the food supply safer and rid much of the world of hunger by ensuring that shelf-stable, nutritious foods are available year-round.
But while some food is “minimally processed” (like shelled nuts or washed vegetables) and other food is simply “processed” (think tinned fish, frozen vegetables or cheese), Monteiro identified another type of processing that emerged in the 1980s and 90s – ultra processed foods (like many breakfast cereals, packaged snacks and sugary beverages) – which, he wrote, are industrially formulated to be “edible, palatable and habit-forming”.
For years, nutritionists had focused on the nutrients in food – potassium and fiber were good, while sugar, salt and saturated fat could be concerning in high amounts. But in the early 2000s, Monteiro and his colleagues at the University of São Paulo noticed that rates of diet-related diseases (like obesity and type 2 diabetes) were rising, even though Brazilians were buying less sugar.
They posited that while eating large amounts of sugar wasn’t necessarily good for consumers, there was more to it than that. While fruit, like mangoes and bananas, are high in sugar, no one eats a dozen of them in one sitting. But something about ultra-processed foods, like candy bars and packaged cookies, made it difficult to eat just one. Scientists would later wonder whether that had to do with the “food matrix”, or chemical and molecular structure of food: the sugars in whole foods like fruit are packaged alongside dietary fiber and vitamins that make them recognizable and more satisfying to our bodies.
A growing body of evidence
Since Monteiro and his colleagues first defined ultra-processed foods, or UPFs, dozens of researchers have asked: what effect do they have on the human body? And why?
For years, scientists had been conducting observational studies that looked at the associations between diet and health outcomes. But the most convincing evidence that UPFs actually were doing something different from other foods finally came in 2019 when Kevin Hall, a senior investigator with the National Institute of Diabetes and Digestive and Kidney Diseases, published the first randomized, controlled study on ultra-processed foods.
Over four weeks, 20 healthy adult volunteers agreed to eat either an ultra-processed or a minimally processed diet for two weeks, then switch to the other diet. Importantly, the diets were matched nutrient for nutrient – those eating the ultra-processed diet would consume just as much sugar, fiber, fat, salt and carbohydrates as those eating the minimally processed diet – and both groups were encouraged to eat as much or as little as they wanted.
By the end of the study, Hall found that participants had eaten 500 calories more each day during the ultra-processed diet weeks and gained more weight. Something about ultra-processed foods left people hungry and wanting to eat more.
That was a monumental finding when, at the same time, the food industry was funding research trying to shift the blame for obesity away from food and to exercise.
Since then, the evidence that ultra processing has negative health impacts has continued to grow.
A 2022 study published in the British Medical Journal found that men (but not women) who consumed a large amount of UPFs had a 29% greater risk of colorectal cancer (findings that have drawn particular attention as the rate of colorectal cancer in young adults rises). At the European Society of Cardiology Congress in Amsterdam in August 2023, a research team presented a study concluding that a 10% rise in daily UPF intake was linked to a 6% rise in heart disease risk. And in September, a Harvard study found that the women who ate the most UPFs were 50% more likely to develop depression than those who ate the least.
Earlier this year, researchers at Deakin University in Australia, Johns Hopkins University, the Sorbonne and others published an umbrella review of the existing research into the health impacts associated with eating ultra-processed foods. They found that UPFs were directly linked to 32 harmful health impacts, including hypertension and anxiety. The paper suggests that consumption of ultra-processed foods is most strongly linked with cardiometabolic diseases, common mental health conditions and higher mortality rates.
“While the associations between ultra-processed food consumption and other health issues are less certain, such as respiratory and gastrointestinal health, some cancers, and intermediate cardiometabolic risk factors, this doesn’t imply these links are unimportant,” Melissa Lane, a postdoctoral research fellow at Deakin University’s Food and Mood Centre and one of the authors of the study, said in an email. “Rather, they highlight areas where further studies are necessary to clarify these potential impacts.”
Because UPFs are made up of dozens of ingredients that are relatively new to the food supply, “the possibilities for physiological effects are endless”, said Josiemer Mattei, a nutrition professor at the Harvard TH Chan School of Public Health.
Why do UPFs do what they do?
The concept of ultra-processed foods has been a helpful lens for scientists to use in evaluating the impacts of industrially manufactured products, said Leigh Frame, executive director of the office of integrative medicine and health at George Washington University. But as researchers go deeper, they are beginning to distinguish among types of UPFs.
Last year, a Harvard study noted that refined breads, sauces, condiments, artificially and sugar-sweetened beverages, animal-based products and ready-to-eat meals were most closely associated with a higher risk of developing type 2 diabetes (as opposed to other UPFs such as cereals, whole-grain breads, fruit-based products and yogurt). Similarly, another Harvard study published last year found that eating UPFs – but particularly artificially sweetened food and beverages – was associated with a risk of developing depression.
“One of the challenges is that UPF comprises all sorts of ingredients and compounds, and these can operate in multiple ways in the body,” said Mattei.
As scientists try to drill down into which UPFs cause the worst health impacts – as a way of asking why UPFs do what they do – they are also trying to conduct more rigorous research, like Hall’s study (although Mattei notes that “as evidence mounts on the potential adverse effects of UPFs”, it may become unethical to conduct future trials). Stronger data could eventually shape nutrition policy in the US.
To change dietary policies in the US, “you basically have to have a randomized controlled trial that shows definitively that this happens, and we don’t quite have that yet,” said Frame, who says National Institutes of Health studies are getting close. Given the poor state of nutrition-science funding in the US, though, she says, that’s a high bar.
Although researchers are wary of UPFs, Frame says that she doesn’t think any food should be 100% off-limits. “Such forbidden food only causes problems, specifically an increase in wanting,” she said. “I am more concerned with the overall dietary pattern than any one individual food.” If people can increase the amounts of whole, minimally processed foods they eat compared with UPFs, that’s a win.
Read more from this series:
