Probing the Gray Areas of UPF Research

When Kevin Hall picked up Stephen Hawking’s A Brief History of Time as a high school student, it sparked the kind of curiosity that would eventually pull him toward science. Although he formally studied physics, drawn to the discipline’s emphasis on first principles and empirical rigor, his interests evolved. Hall found himself increasingly fascinated by questions far less tidy than the motion of planets or particles, such as how the human body regulates weight, responds to diet, and adapts to an increasingly industrialized food environment.

That shift led him to the National Institutes of Health (NIH), where he spent more than two decades applying mathematical modeling and controlled metabolic studies to some of nutrition’s most stubborn puzzles. His analysis of television contestants on The Biggest Loser helped document how the body defends its weight long after dramatic dieting ends. And in 2019, his controlled feeding study on ultra-processed foods demonstrated something both simple and unsettling: even when nutrients are matched, people tend to consume hundreds more calories per day when eating highly processed diets.

Although Hall is the first to admit that the mechanism remains unclear, the implications have reshaped the scientific debate.

Hall resigned from his role at NIH in April 2025, frustrated by what he’s described as scientific censorship under the current administration. He’s since released Food Intelligence, a new book cowritten with journalist Julia Belluz that attempts to demystify that complexity for a broader audience. Rather than offering another set of dietary rules, he says it aims to explain how nutrition science works, why it often confuses the public, and how to think more clearly about food, uncertainty, and evidence.

In a wide-ranging discussion, edited for length and clarity, Hall spoke about his research, the challenge of communicating unsettled science, and that nagging debate among his peers about the definition and possible health impacts of ultra-processed foods.

When you set out to write this book, what was your intent and who was your primary audience?

I thought back to myself in high school. I got excited reading things like A Brief History of Time and other books that were written for the lay population but really tried to instill a fascination for the science and the fact that it hadn’t all been worked out yet, that it was still evolving.

There wasn’t really a book like that in nutrition and metabolism science. There were books about various diets, there were books about debunking myths, there were books about some new theory about obesity or something like that. But there really wasn’t the sort of foundational book with a little bit of the history of how we know what we know and what’s left to find out.

That is very different than my training in physics. Physics is so tied to the history. We learned about the Michelson-Morley experiment. We knew about the gravitational lensing of Eddington’s. We knew about these very historical studies that shook the foundations of the science and progressed it forward as the theory developed.

When I was at NIH, I would have students and postdocs with training in nutrition and metabolism. They knew their stuff and they were up to date, but then I would ask, “Do you know where that idea came from? Do you know what the key experiments were?” They had no clue. And at first, neither did I. So, I spent a lot of time exploring the history of nutrition and metabolism.

The thing that kept popping up in that history was a recurring theme: People would think that they’d figured out some major advance in nutrition, and then they would very quickly rush it to an audience of people to capitalize and commercialize those findings in ways that often were not tested.

And the more times that we’ve done that as a field, the more people have been confused. Why do we need all these supplements? Should we really be consuming this very low-carb diet or this very low-fat diet? And what’s the basis for the health claims behind those prescriptions?

Do you feel that rush to commercialize has had unintended consequences on the science itself? Are the right questions being asked in the right sequence of priority?

I don’t think it’s just about commercialization. The field rushes to overhype the science as well. And that occurs because people follow the funding stream. And if there’s a hot area of research that you can contribute to, scientists will often follow the money. I would d argue that ultra-processed foods are currently one of those hot areas.

It’s now in vogue to say, “Oh yeah, ultra-processed foods are horrible for you, and we need to devote all of our attention to going back to more natural foods.” And that’s very much along the rhetoric that we hear from some of the political leadership now, as well as some public health folks.

Have there been any reactions to the book that have surprised you?

One legitimate criticism has been that we focus very strongly on the role of the food environment in driving much of the health and nutrition problem and downplay the amount of conscious control and willpower that individuals might have in resisting the food environment. There are some people who obviously have had success and are making changes in their lifestyle.

We mentioned that in the book, but we don’t really discuss it that much. But almost every other book in this space is talking about specific diets or lifestyle hacks people can adopt to be successful. We’re trying to shift the conversation more towards the structural difficulties that people have in making those changes.

You talk about the ‘food environment’ as a framing device. That’s a complicated concept.

We try to emphasize that the food environment is not just the foods themselves. It’s not just the marketing of the foods. It is also the social environment of food and eating. Even those things have changed quite a lot. I don’t think that there were many people showing up to their office with doughnuts back in the 1970s. But nowadays, that’s not an uncommon phenomenon.

Your research came nearly a decade after Carlos Monteiro introduced the Nova framework to categorize foods based on degree of processing. Ever since then, the debate around ultra-processed foods has largely focused on definitions and how to properly frame the questions: Are diet-related health issues tied to the processing, the formulation, or something else entirely? Your research seems to imply that it really is a combination of factors.

People get very hung up on the definition and the label of ultra-processed foods. First, it’s a bit of a misnomer because it implicates only processing and not formulation. But practically, ultra-processed foods are recognized by their ingredients and not about how those ingredients were processed in a way to arrive at the final food product.

So, then the question is, do ultra-processed foods harm health? And a lot of the observational epidemiology says that diets higher in ultra-processed foods, even though it’s a very broad category, increase risk for a variety of poor health outcomes. There are many, many, many cohort studies and cross-sectional analyses of large groups of people that show that.

For me as a researcher, the question is what’s underlying those associations? Are they causal? And if they are causal, what are the mechanisms? Maybe there’s no mechanistic link or maybe there is. It’s complicated.

People have a very knee-jerk reaction to this concept of ultra-processed foods. And it seems that the originators of the concept have embedded in its origins a judgment that these foods are unhealthy. Fortunately, that conclusion is not part of the Nova definition of ultra-processed foods because that’s an empirical question. But in the papers where ultra-processed foods were originally described, they are described with a negative connotation in terms of disrupting traditional food systems as a result of increasing industrialization and capitalism.

I’m just really interested in, are there mechanisms of ultra-processed foods in relation to health going beyond the kinds of things that we’ve traditionally thought of in nutritional science? The Nova system ignores the nutrients and just looks at what they call the extent and purpose of processing and formulation. Alternatively, you view foods through a more traditional nutrient profiling lens and just look at how much saturated fat, sodium, added sugars, whole grains, fiber, and things like that are in the food. And I think these complementary approaches are kind of interesting.

What are the biggest questions still unanswered? And where are you with your own research?

We did a follow-up study to try to better understand the mechanisms by which people overconsume diets high in ultra-processed foods. In our first study published in 2019, people were presented with two diets that both had double the number of calories required for energy balance and matched for a bunch of nutrients but varying widely in ultra-processed foods. One of the things that’s a challenge in nutrition science is that when you’re using real foods, lots of other stuff varies. Even though you match two diets for carbs, fat, protein, sugar, sodium, fiber, glycemic load, it doesn’t mean that there aren’t other potentially important things that are different.

And just because you’ve labeled one ultra-processed and the other one minimally processed, there’s other stuff varying between the diets. So, we couldn’t conclude from our first study that ultra-processing per se caused people to overconsume calories. When we explored other factors that were different between the diets, one of the things that we noticed was that even though the overall energy density of the two diets was matched, in the ultra-processed diet that was because we gave them more beverages with dissolved fiber supplements so that overall fiber was matched to the minimally processed diet. There had been some discussion and debate in the past about whether beverages can contribute to feelings of fullness and satiation and delaying your next meal. Some folks said [that] you’ve got to look at the non-beverage energy density of the meals.

And indeed, the ultra-processed foods were more energy dense, but not because they had more fat, which is more calorific than carbs and protein, but because the ultra-processed foods had essentially been dried out. The food matrix has been disrupted, and water had been extracted because you want to increase shelf life and prevent bacterial growth, things like that. That process concentrates the calories.

Another factor that we found was different between the two diets in the original study was the number of so-called “hyperpalatable” foods. Again, I think hyperpalatable may be a bit of a misnomer because they are defined not by assessing subjective palatability but rather by whether combinations of nutrient pairs exceed prespecified objective thresholds. The three types of hyperpalatable foods are high in both fat and sugar, fat and salt, or carbs and salt. I say hyperpalatable might be a misnomer because despite the ultra-processed meals having more of these hyperpalatable foods, people didn’t report the meals being any more pleasant to eat.

So, to test the effects of both energy density and hyperpalatable foods in our new study, we designed three diets with about 80% of calories from ultra-processed foods. One diet is high in both non-beverage energy density and hyperpalatable foods, similar to our first study, another is high in energy density and lower in hyperpalatable foods, similar to the minimally processed diet, and the third ultra-processed diet is lower in both energy density and hyperpalatable foods.

That study is finished. We’ve analyzed the primary outcomes, and the short answer is that the ultra-processed diet high in energy density and hyperpalatable foods still resulted in overconsumption of calories compared to the minimally processed diet. But by reducing both the energy density and the number of hyperpalatable foods, we prevented overconsumption despite still being 80% ultra-processed foods. So, I think it’s encouraging that we’re starting to get at some of the likely main mechanisms by which ultra-processed foods typically drive excess intake.

Another factor that we found in the first study was that ultra-processed foods tend to be eaten more quickly, possibly preventing the normal oral sensory processing that can potentially drive satiation and termination of a meal. In our first study, people ate ultra-processed meals more quickly in terms of the grams per minute. But in the new study, they didn’t. I’m not sure why, but eating rate was consistent across all the different dietary patterns. This doesn’t mean that food texture and eating rate can’t play important roles influencing calorie intake. Other researchers have demonstrated that they can influence calorie intake. But these factors don’t seem to have played a role in our new study where energy intake varied by many hundreds of calories per day between test diets.

Overall, this research shows that we are beginning to reveal the mechanisms by which ultra-processed foods contribute to overeating and weight gain. I think understanding these mechanisms is important for informing recommendations to consumers and food manufacturers as well as policymakers.

You’ve just described multiple shades of gray during a time when the public and policymakers want simple answers. You left the NIH last spring over concerns about the science being muzzled. What’s your philosophy on how unsettled science should inform policy and regulation? It’s an ongoing tension.

Yeah, and the appetite for making food policy decisions obviously varies widely in different administrations. If you’re going to make a policy change, I would hope that you would want it backed by the best science. And I worry with the current administration that they’re not consulting the best scientists on those topics. I worry that they’re not particularly interested in science because science complicates their preconceived narrative. At least that was my experience when I tried to communicate one of our study results.

And despite all of the rhetoric about making America healthy again, none of the things that I’ve heard discussed so far are going to meaningfully improve public health when it comes to diet-related chronic disease, which is now quite dire.

The fact is we have to change our food environment in a way that is beneficial for public health, while at the same time, we have to plan for feeding the population that will plateau in the near future at about 10 billion people without destroying the planet in the process. This is what we’re getting at in the last chapters of our book.

We need all these different sectors working together. It blew my mind when I first realized that people in nutrition science don’t really talk to people in food science and industry. And they don’t really talk much with the folks in agriculture. And they don’t talk to people who are worried about the environmental consequences of food and agriculture.

It seems like a huge, missed opportunity that the MAHA (Make America Healthy Again) Commission, which spans the Environmental Protection Agency, U.S. Department of Agriculture, and Health and Human Services, could have convened experts and stakeholders from all those fields together to make real progress in planning for the future of food while protecting public health and the environment.

Public research funding and academic institutions are under massive stress. What’s your assessment where research goes from here?

I wish I knew the answer to that question, but I think you’re right. Nutrition and metabolism science has always been underfunded at the federal level and I fear that the situation is only going to get worse. I hope I’m wrong, but I suspect that we need to start thinking more creatively to fund this science.

Among the things that are likely going to turn the tide on diet-related chronic disease are the new obesity medications that are quite effective. Right now, they’re very expensive. But there’s more coming down the pipeline and some of the current medications will go generic so prices will likely come down. Some newer medications are in pill form rather than injectables that have to be kept cold, so that will facilitate wider distribution around the globe.

So, I think that there’s a potential role for pharma and food companies to fund mechanistic studies to better understand how these drugs work and how different dietary patterns might be used synergistically with obesity pharmacotherapies. How can we identify who is going to most benefit from various dietary patterns and food technologies alongside different therapeutics? I think these are really interesting questions.

One of the multiple constituencies for your book includes your peers across the food sciences. What is your takeaway message to them?

Society faces very big problems, and we all need to work together to fix them. For too long, the focus has been maximizing production of calories and corporate profitability while externalizing costs to the environment and public health. For too long, we’ve been siloed and not necessarily rewarded for reaching across fields to make real progress addressing the critical problems of the 21st century.

Food and nutrition scientists need to work with folks in agricultural research to ask, “what high-yield crops could we grow sustainably without harmful pesticides, herbicides, or fertilization methods to make products that have better nutritional properties to feed the population?” We need to say, “Hey, I’m a nutrition scientist, you’re a food scientist. Let’s work together to reengineer foods for better health and do the clinical trials to demonstrate their efficacy and safety.” These sorts of cross-sector projects are where we can identify the end point that’s going to have value across the different sectors and improve the health of both people and the planet.ft