What Makes Ultraprocessed Foods Addictive?

Articles in this report were published with editorial independence. The collection was made possible by the support of WVU Rockefeller Neuroscience Institute.

The recent surge in the use of GLP-1 weight-loss drugs has propelled addiction-adjacent terms such as “food noise” and “food cravings” into common vernacular. But can food actually be addictive? Now some neuroscientists and food behavior researchers are trying to understand if food—particularly ultraprocessed foods—can be addictive in the same way as other known substances, such as cigarettes, alcohol and cocaine.

For foods to be potentially addictive, “they’re created in a way that is most palatable and most delicious,” says Alex DiFeliceantonio, an appetitive neuroscientist at Virginia Tech. “When you look at the food environment, those tend to be ultraprocessed.”

On supporting science journalism

If you’re enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.

Scientific American spoke with DiFeliceantonio about research unpacking whether food addiction is real, whether certain types of foods might have more addictive qualities and how related eating disorders can be addressed.

[An edited transcript of the interview follows.]

What does it mean to have a “food addiction”?

When we’re thinking about food addiction and looking qualitatively at what people are eating when they are saying that they can’t stop eating, we have to put it in the framework of a substance use disorder. These disorders affect life in an untenable way. Food addiction isn’t in the Diagnostic and Statistical Manual of Mental Disorders (DSM) like substance use disorder is, but there is a proposal to have it put in the DSM.

We typically look to the Yale Food Addiction Scale for clinical evaluation. The scale was designed to assess the same criteria as the substance use disorder criteria in the DSM. The scale also contains what we call clinical indicators that a person is experiencing symptoms of an addiction and those symptoms are poorly affecting life—such as the ability to engage in social situations or engage in aspects of work or life. If we accept that food addiction exists—if you give the Yale Food Addiction Scale to large population-level studies and do it across multiple countries internationally—we generally find that around 12 percent of people [experience] it.

A combination of factors can lead to an addictive behavior. And the most common is the addictive potential of the substance combined with the vulnerability of the person. We think about both of those things with food, too: ingredients that could have addictive potential and the people who could be most vulnerable. We also look at food attributes, such as high refined carbohydrate content, which is known to trigger reward pathways in the brain.

Other aspects of substance-use-disorder criteria include loss of control over intake and highly patterned intake. That’s what we see in binge-eating disorder. Binge-eating disorder and food addiction are not the same thing, but they share similarities. If we look at the foods people report consuming when they binge eat, they tend to be things that would be classified as ultraprocessed—things like pizza, ice cream, candy, chips. They’re very rarely things like fruit, nuts, beans.

What do you consider an ultraprocessed food?

There are multiple definitions. I would say the one most studied and what we use in my lab is the NOVA [“new” in Portuguese] definition; it has four levels, and the fourth is ultraprocessed foods.

The NOVA level-four foods contain ingredients or processing methods that are not available to the home cook. You can think about additives like stabilizers, cosmetic additives that enhance color or flavor, or emulsifiers to maintain texture. If you add vitamin D or calcium—types of nutritional fortification—that doesn’t make a food a NOVA ultraprocessed food by itself. Ultraprocessed might also refer to foods produced with an industrial method, like making starch slurries that are then extruded, puffed, subjected to high heat or molded in ways that you really wouldn’t be able to make in your kitchen.

Why might ultraprocessed foods in particular fire up reward pathways in the brain?

The current scientific thinking is we have one reward system and lots of different things that can be rewarding. All addictive drugs increase dopamine in the striatum [a brain region beneath the cerebral cortex that is involved in motor and reward processing]. This has been the dogma since 1988 with [a paper by pharmacologists Gaetano Di Chiara and Assunta Imperato]. It’s the same thing [with certain foods]. If you infuse sugar and fat into the oral cavity of an animal, you see an increase in dopamine. If you infuse these things directly into the gut [of animals], you also see increases in dopamine. There is no agreed-upon threshold in which we say a substance that is addictive must increase dopamine in the striatum by x amount.

Modern ultraprocessed foods started to become widespread in the U.S. around the 1950s. Those foods are acting on a reward system that evolved to deal with natural rewards from the environment.

When we’re thinking about food addiction, we know that there are certain levers or ways to highly activate the reward system, and ultraprocessed foods seem to access the most levers. They elevate levels of sodium, fat and refined carbohydrates in the body. And this is aided in various ways—with emulsifiers, with texture changes, with flavor changes—ultraprocessed foods are made to be the most palatable, the most delicious. We don’t think about broccoli as an addictive substance; we think about foods that contain enough of these potentially addictive nutrients in combination to be addictive substances.

How might GLP-1 medications such as Wegovy and Ozempic factor in?

We still need more data on this, but looking at what people report on the GLP-1 drugs is that they feel less food noise. It is not surprising that the medication also works through other seemingly more food-specific mechanisms, through satiety mechanisms [or making people feel full]. Some of the reduction is because of that increase in nausea and satiety, but the other aspect of it that we’re still figuring out as a field is what proportion of it is from this change in the reward system.

You recently conducted a small study on young adults aged 18 to 25 who consumed a high ultraprocessed food diet. What changes in eating habits did you see?

What we found, I think, is really interesting. We did two test outcomes: participants either consumed meals that were 81 percent calories from ultraprocessed foods or zero percent calories from ultraprocessed foods for two weeks. All subjects ate both diets, which were nutritionally matched, and there was a four-week washout in between.

In one test, participants went to a buffet meal after each of the diet periods and could eat as much as they wanted. We found that the older group [aged 21 to 25] didn’t eat any more after the ultraprocessed diet than the minimally processed one. But the younger group consumed more at the buffet after the ultraprocessed diet than after the other diet.

In another test, called the absence-of-hunger test, the participants were presented with snacks or time to play on their phone. After the ultraprocessed diet, younger participants also ate more. So they were consuming more even when they weren’t hungry. This is mindless snacking.

The study didn’t directly measure addictive behavior, but it demonstrates how ultraprocessed food can lead to behavior change. We plan to publish a similar study looking into what foods have an addictive profile and what that looks like for different people.

Why might some people experience more addictive behaviors around ultraprocessed foods than others?

We know genetic variants that put people at greater risk for nicotine abuse or for finding cocaine addictive; we know genetic variants for alcohol use disorder. But we don’t know this for food addiction. This is, I think, where the field is right now, and where my lab really wants to make progress is understanding.

Why does “food addiction” ignite debate?

It’s a really thorny one. In the context of substance use disorders, we are moving from something that you don’t have to do to survive to something you do have to do to survive. You have to eat. So it’s really difficult to think and talk about food as an addiction.

One pushback I hear is we don’t want to overpathologize everything. But I think that if about 12 percent of a population is telling you that they have a problem, maybe we should look at it, or we should at least give it some concerted study and determine what it is. People also say it’s a behavioral addiction—you are not addicted to food as a substance; you are addicted to the act of eating. But that argument falls down pretty quickly when you look at what people are eating. If you were addicted to the act of consuming, you would be eating things that were hard or crunchy or that required a lot of work to consume. And that’s not really what we see. We see people losing control over intake for items that are high in fat and sugar—refined carbohydrates.

I also think a lot of the pushback is a moral tie—if you’re addicted to food, you’re a bad person. For the most part, a lot of people have let this go. We understand that alcohol use disorder, for instance, isn’t caused by a failing of willpower. [People with the disorder] cannot overcome it, and we have to help them. I’m always bringing that level of compassion to food addiction, too.

If ultraprocessed foods are truly “addictive,” what are some treatments? How should those treatments be tackled on a societal and individual level?

When someone has a substance use disorder, part of the treatment is for them to avoid the cues and context in which they use that substance. In our environment, how do we tell someone not to engage with food? It’s impossible. People are bombarded by food cues everywhere, especially for ultraprocessed foods.

Addiction is societal and behavioral. We decide what drugs are illegal. We decide at what age people have access to potentially addictive substances. Artificial refinements of foods—so pure sugar, fat, combinations of fat and sugar that don’t occur in nature—also activate our reward system. At what level is a rewarding substance one that we are willing to regulate as a society?

You need to eat to survive, but you don’t need the majority of ultraprocessed foods for human survival, food security and national security. We do have to process foods for national security reasons. We don’t have goiters or rickets, because we fortify foods with essential micronutrients. Canned vegetables and frozen fruit are good for food security, but you don’t really need to have extremely sugary cereal advertised to children. Those two separate worlds can exist.

I feel like a lot of times this gets framed as “you’re taking away people’s food when you regulate it.” To quote Nora Volkow, head of the National Institute on Drug Abuse, “Drugs rob the brain of the capacity to exercise free will.” We want to make decisions for our own health and for the health of our families. When we’re dealing with an addictive substance, you are no longer in control of that decision. That’s where it’s important to have policy in place. You’re not taking something away; you’re putting up guardrails and helping people make informed decisions—because you can’t make an informed decision about an addictive substance you’re taking.