A look at recent DONNUT Lab research exploring why ultra-processed foods drive persistent cravings, plus practical habit changes for better appetite control.
On September 17, 2026, USA Today published a report detailing new insights from the DONNUT Lab at Virginia Tech. The facility is directed by Alex DiFeliceantonio, who serves as an appetitive neuroscientist at the Fralin Biomedical Research Institute. The lab gets its name from the "DiFeliceantonio Neuropsychology of Nutrition." Her scientific team is actively investigating why certain products are remarkably difficult for people to resist.
The primary focus centers on highly palatable products like doughnuts, cookies, chicken nuggets and chips. These ultra-processed foods are specifically engineered for maximum convenience, long shelf stability and strong flavor. The news report contrasts these packaged items with minimally processed whole foods. Examples of minimally processed options include fruit, vegetables, beans and whole grains.
Instead of relying on outdated ideas about personal failure, this team asks a fundamentally different question. They want to know exactly which properties of ultra-processed foods encourage overeating. They are also investigating why different human brains and bodies may respond entirely differently to the exact same snack.
For years, I watched smart, capable people blame themselves when standard diet advice failed them. They would cut calories drastically, run themselves into the ground and inevitably regain the weight. It was heartbreaking to see. I realized we were treating a complex biological and psychological system like a simple math problem.
That was the turning point when I knew we had to focus on metabolic health and habits rather than just restriction. Cravings are not a character flaw. They are often a direct biological response to how modern food is formulated. The DONNUT Lab uses multiple advanced tools to map this exact response mechanism.
Researchers employ MRI scans, blood sampling and taste tests to gather concrete data. In one specific experiment, participants view photographs of ultra-processed foods like potato chips alongside minimally processed foods like baked sweet potatoes. Researchers simultaneously use MRI technology to measure live brain activity. Another study monitors physiological reactions to various combinations of sugar and artificial sweeteners.
The research team measures blood glucose and hormones, including insulin, through blood samples collected while participants consume test beverages. A further study sends participants home with specific drinks featuring different sugars and flavors. They consume these beverages outside the laboratory for two weeks. Afterward, they return for taste testing, preference discussions and an MRI scan.
DiFeliceantonio hypothesizes that consuming some of these foods may produce mind and body effects resembling addictive drugs. Understanding the neuroscience of food noise and managing persistent cravings helps individuals see this as a biological challenge. Willpower is a finite resource that frequently fails when people are tired or stressed. This research treats human appetite regulation as a complex physiological system instead of a moral test.
Mainstream media headlines often present early research as absolute proof that all modern food is toxic. The current coverage occasionally borders on this exact sensationalism. However, DiFeliceantonio’s comparison to addictive drugs is a research hypothesis, not a completed clinical conclusion. The current evidence does not prove that every single ultra-processed food meets the strict diagnostic criteria for substance-use disorders.
USA Today notes a recent study published on Nature’s platform regarding grocery availability. That study found that more than 70 percent of foods sold in U.S. grocery stores are ultra-processed. With such massive market dominance, expecting people to achieve absolute dietary purity is completely unrealistic. Food access, time, cooking facilities and income dictate what many families can actually eat.
DiFeliceantonio rightly points out the severe difficulty of asking a single parent to rely exclusively on minimally processed foods. A caregiver living in a food desert and working multiple jobs cannot easily abandon convenient packaged items. Practicality matters just as much as nutritional theory. Instead of extreme restriction, prioritizing healthy convenience foods that fit busy schedules offers a realistic middle ground.
Ultra-processing is also an incredibly broad category. Packaged foods can differ substantially in calories, protein, fiber and sweetness. They also vary significantly in portion size, texture, eating speed and palatability. Treating every product in this massive category as equally harmful simply overstates the current scientific evidence.
A much more useful behavior change approach involves targeted self-observation. People can identify the specific foods that trigger their repeated, unplanned eating episodes. Once identified, they can change the availability or portion format of those specific items. Keeping satisfying, minimally processed alternatives highly visible makes healthier choices far more convenient.
The recent USA Today report mentions a completed DONNUT Lab diet study. This trial compared an ultra-processed food diet directly against a low-processed food diet. Participants experienced both diets in a varied order. The reported meals differed significantly, with the processed lunch featuring deli meat, packaged sliced bread and ultra-processed American cheese.
The comparison meal included whole chicken breast, homemade bread and natural cheddar cheese. However, the news report does not share the sample size, intervention duration or statistical results of this specific study. Without these metrics, we must look to established clinical trials for concrete data. The best-established directly relevant feeding-trial context is a 2019 National Institutes of Health inpatient crossover study.
This important trial involved 20 weight-stable adults who spent two weeks eating an ultra-processed diet. They then spent two weeks eating an unprocessed diet. Both diet phases were carefully designed to be broadly matched for several nutritional characteristics. Despite this matching, participants consumed approximately 508 more calories per day during the ultra-processed food phase.
This occurred completely spontaneously, meaning participants were not instructed to overeat. The trial demonstrates a sizable difference in spontaneous energy intake under experimental conditions. However, the small sample size and tightly controlled inpatient setting are important study limitations. This specific environment limits how directly the result can be generalized to all midlife adults or all grocery store environments.
The weight changes followed this spontaneous calorie increase very closely. Participants gained about 0.9 kilograms over the two weeks they consumed the highly processed diet. Conversely, they lost approximately the same amount of weight during the unprocessed food phase. These controlled findings confirm that modern food processing influences how much energy our bodies naturally seek.
The emerging research agenda is finally moving beyond broad dietary associations. Scientists are now investigating specific mechanisms like gut-brain signaling, hormonal responses and eating behavior. Future studies must determine exactly how different flavor, sweetness and processing levels interact to influence our daily appetite. We also need to understand whether adolescence is a particularly sensitive period for these food-related brain changes.
The DONNUT Lab included many teenagers in its recent diet study to examine this exact question. Understanding these developmental windows could reshape how we feed future generations. Adults managing significant weight changes, diabetes or medication-related appetite shifts should always consult a qualified clinician. The current DONNUT Lab findings do not constitute individualized medical advice.
Until more definitive clinical protocols emerge, everyday adults should focus on manageable daily habits. Meals that combine minimally processed carbohydrate sources, adequate protein and high-volume foods offer a sustainable path forward. Track observable outcomes like between-meal hunger, evening snacking, daily energy levels and sleep quality. You can steadily improve your environment by breaking the restriction overeating cycle with flexible eating.
Gradually replacing one routinely consumed packaged item with a whole-food alternative builds lasting metabolic health without unnecessary stress.
WeightRestart shares research-led guidance on weight loss, metabolism, nutrition, strength, appetite, sleep and recovery. Our goal is to make complex health information clear, practical and useful for people building progress they can maintain.
Learn how to build a weight-management approach around better information, realistic expectations and habits you can keep using.
read the blog
