This is a transcript of the Gastropod episode, Nutrition Advice Decoded: What Foods Are Actually Good For Us, What Should We Avoid, and Why Is It All SO Confusing?, first released on March 25, 2025. It is provided as a courtesy and may contain errors.

DAN BUETTNER: We found our first blue zone about 125 miles off the coast of Italy, on the island of Sardinia. And not the entire island. The island’s about 1.4 million people. But only up in the highlands, an area called the Nuoro Province. And this is a place where people not only reach age 100, they do so with extraordinary vigor. Places where 102-year-olds still ride their bike to work, chop wood, and can beat a guy 60 years younger than them.

NICOLA TWILLEY: Oh, wow, is this the episode where we reveal these hardy old people’s secrets and finally become millionaires?

BUETTNER: Should you eat organic meat or should you be eating tofu? How about these hormones, or resveratrol?

CYNTHIA GRABER: I’d love one of those answers to be the key to eternal life, but sadly no. It turns out that the secret of these so-called blue zones has in fact recently been revealed! You know what it takes to ride your bike and chop wood like the younger guy? You are one of the younger guys! You’re not actually 100!

TWILLEY: Cynthia, how are we going to monetize that? I hate to say it, but you’re not doing this right.

GRABER: Sadly no, but hopefully we can get the science and history right. We, of course, are Gastropod, the podcast that looks at food through the lens of science and history, I’m Cynthia Graber—

TWILLEY: And I am a very disappointed Nicola Twilley. But this episode is a tale of high hopes and deep disappointment, plus a whole lot of confusion. For one, we’re telling the story of why the blue zones became trusted nutrition wisdom and why it seems as though it’s actually kind of a made-up story.

GRABER: This is part of a bigger story. It’s the tale of how modern nutrition research got started, some of its transformational and life-saving early successes, and then how things in nutrition science got…more complicated.

TWILLEY: With the result that many people have no idea whether eggs, or coffee, or red wine, or protein, or carbs, or saturated fats—whether they’re healthy or terrible for you. There is a lot of advice out there about what to eat and what not to eat, and a lot of it is contradictory—and this episode we want to figure out how we got in this terrible mess and what it will take to get out of it.

GRABER: This episode is supported in part by the Alfred P Sloan Foundation for the public understanding of science, technology, and economics, and by the Burroughs Wellcome Fund for our coverage of biomedical research. Gastropod is part of the Vox Media Podcast Network, in partnership with Eater.

[MUSIC]

SAUL NEWMAN: Blue zones are regions that are supposed to reach remarkable ages at remarkable rates. So the idea is that there are these special regions where people reach 100 at remarkable rates compared to everywhere else in the world.

TWILLEY: Saul Newman is a research fellow at the Oxford University Institute of Population Ageing. And recently he got interested in these blue zones, which were originally made popular by an author called Dan Buettner in 2005.

BUETTNER: So the premise of blue zone is if we can find the optimal lifestyle of longevity, we can come up with a de facto formula for longevity.

GRABER: Dan did a bunch of traveling and exploring, and he ended up coming up with five regions in the world filled with super old people, and then he came up with a story about what these regions had in common. And he turned it into big business.

TWILLEY: There were articles and then books and then more books packed with recipes, challenges, and lots of lessons for living longer and then of course there was a Netflix special.

DAN BUETTNER: What’s the best tea to drink on a daily basis?

OLD WOMAN: [SPEAKING ANOTHER LANGUAGE]

TRANSLATOR: Wine.

DAN BUETTNER: Ha! Ha! Ha! I like the way you roll.

[MUSIC]

DAN BUETTNER: These people live to one hundred at the highest rates in the world. And these secrets can help every one of us to get every good year we can get out of this body of ours. That is the promise… of blue zones.

GRABER: As we said, Dan claimed that there were a lot of things these places with an unusually high number of centenarians had in common, like for instance people who live in blue zones move around during the day a lot and have strong community bonds. But he also went hard on dietary recommendations and how that could explain how people in these regions live practically forever. Like that glass of wine.

BUETTNER: What—people who make it up to a hundred on average, they’re eating about 90 percent whole food plant based. The five pillars of every longevity diet in the world are whole grains, wheat, corn, and rice, greens, and of course, garden vegetables, they all have garden vegetables, tubers. Nuts, and the cornerstone of every longevity diet in the world is beans.

TWILLEY: Unlike Dan, Saul is actually a scientist and also a longevity expert. Well, he’s an expert on longevity in plants.

NEWMAN: You know, there’s a reason you can buy tulips anytime of the year, and it’s because plant scientists really understand how to modify lifespan in plants. We can do this to an extraordinary degree. So this was my day job. And while I’m carrying out this day job, someone sent me—I have a PhD in medical science, so someone sent me a paper on the limits of life in humans.

GRABER: Saul went down a rabbit hole that led him to the blue zones, and he found the research fascinating—but he thought it seemed a little off. So he did some mathematical work to figure out if this was likely to be true, that these regions had the highest number of centenarians, and what he found was that the data was mostly likely to be junk. That’s his word.

TWILLEY: His reasoning is a little abstract, especially if like me you’re not good with numbers. But basically, if someone makes a very normal paperwork mistake and gives a person a date of birth that makes them, say, ten years younger than they actually are—it’s just one mistake, but by the time that person reaches 100, at least on paper, they’re the last one’s standing.

GRABER: They’re actually 90, not a 100. And you’d never know from looking at them. And meanwhile, all the people who had correct birth certificates died. Who knows, maybe in their 80s or 90s, like most people.

NEWMAN: And so even if you start out with, let’s say, a 1 in 10,000 error rate, by the time you get to 100, the entire population is made up of errors.

GRABER: I’m going to admit that this takes a little thinking about to understand, but here’s an example of these paperwork errors played out in the real world. Saul did some digging in those blue zones to find the centenarians and figure out if they were really over 100. He found a great example of this, where paper errors created an entire fake blue zone: it’s in Okinawa in Japan.

NEWMAN: Okinawa was invaded, during World War II, obviously, by the Americans. And if you look at the pattern of centenarians within Okinawa, you could predict around 80 percent of the variation in where they were based on where had been firebombed. So effectively, if you’ve blown up the birth certificates, you get a lot more hundred year olds.

TWILLEY: Add to that the fact that the American government of occupation was in charge of issuing replacement birth certificates, and those folks mostly didn’t speak Japanese. So, the scope for errors was large.

NEWMAN: And this is part of the reason that in 2010 when Japan went looking for their hundred-year-olds, they found out that 82 percent of them were dead. Most of them had died in World War II, and simply hadn’t had their deaths registered.

GRABER: Funnily enough, even in places that weren’t taken over at a certain period by foreign soldiers, Saul found that most of the supposed centenarians were, like the Okinawans, just totally dead.

NEWMAN: So one of the early cases I discovered was that 72 percent of the people in Greece who were aged 100 were actually dead. At least 72 percent.

TWILLEY: In this case, firebombing and language barriers were not the problem. The issue was much more straightforward.

NEWMAN: There’s a motivation to commit this kind of fudging of people’s ages in Greece, which is that pension fraud is really, really common. You know, you just don’t register your neighbor’s death and his check will turn up every month. So, you know, we had nearly 8,000 people that were collecting their pension, but when you went and knocked on their door, they were dead. Very much alive on pension day.

TWILLEY: Dan Buettner had declared this very delightful Greek island in the Aegean called Ikaria—he had declared it was a blue zone in 2006. And then the Greek government decided to crack down on pension fraud in 2012.

GRABER: So in research that Dan Buettner cited, published in 2001 before this crackdown, there were supposedly 90 centenarians in the Ikarian blue zone. Saul told us that after the crackdown, there were only three living on the island.

NEWMAN: So, you know, you have this billion dollar industry. And it’s based on a sample size of three.

GRABER: Dan has admitted that at least one of the blue zones, he just made up. His editor said he had to find one in America, and so he decided that the vegetarian Seventh Day Adventists in Loma Linda California were a blue zone.

TWILLEY: What’s also interesting about the Loma Linda example is that really, it seems like Dan picked it just because it fit with his overall blue zone-derived dietary advice: eat lots of plants.

GRABER: We’ll get back to this particular diet advice and what science does and doesn’t say, but the problem with the blue zones is that it did NOT have the science to support that recommendation. Because while Loma Linda was largely vegetarian, the people who lived in the rest of the blue zones didn’t necessarily eat that way Dan was suggesting.

NEWMAN: And one of the best cases for this was Okinawa. Because the Japanese government really, really measures their citizens. And they’ve been doing this since they were handed Okinawa back. So we have these continuous, massive nutritional surveys in Okinawa since 1975. Okinawa has the worst body mass index in Japan. It has since 1975. Right? There’s never been the point where this has been the healthy, a healthy province. It’s always been the single worst. They eat the least vegetables. The least leafy vegetables, root vegetables, pickled vegetables. They have the third last rate of consuming beans. And they have extraordinarily high rates of eating meat. They have very low rates of eating oily fish. And this is all completely the opposite of what has been claimed.

TWILLEY: And they also don’t live to a 100 at high rates.

GRABER: Saul told us that the blue zones people come at him and say he doesn’t know what he’s talking about because he’s just a plant scientist, but he’s been working in demographic science for years. They say they have scientists who accurately verified the age of every centenarian they included, but Saul’s research has been available for years now, and nobody has been able to find major flaws with it.

TWILLEY: Bigger picture, this story—the idea that a particular way of eating is the secret to eternal health, until it turns out it’s actually not—this story is sadly extremely common. It happens with specific dietary patterns, like the blue zone or Atkins. It happens with different nutrients, like carbs or protein. And it happens with individual foods too.

BROADCASTER: On your health watch this morning, people who eat three or four eggs a day have a higher risk of both heart disease and early death. That’s according to a new study out of Northwestern Medicine.

REPORTER: Research now suggests that eating a dozen or more eggs per week may not have a negative impact on cholesterol for adults over 50.

GRABER: Eggs were demonized starting in 1968, but in 2015 the Dietary Guidelines Advisory Council said there was no evidence that you’d die early or get heart disease from eating eggs. And the confusion isn’t just limited to eggs.

BROADCASTER 1: Some really good news this morning if you’re reaching for that morning cup of coffee.

BROADCASTER 2: The health benefits it can offer are being revealed in a brand new study out just this morning.

NEWSCAST: This could add a bitter taste to your morning cup of joe. An agency says there’s a possible risk of cancer from coffee, and wants businesses in California to warn consumers.

TWILLEY: Yes, coffee is another thing that has been declared the key to health as well as basically poison. So has red wine. We mentioned this on our recent episode about non alcoholic beverages—in the 1990s, doctors literally recommended that perfectly healthy individuals add a glass of red wine to their diets to help their hearts.

GRABER: But it turns out that all of the studies on the chemicals in red wine that are supposed to be helping your heart—all the studies show that it’s bunk. They don’t help at all. And what seems like the constant back and forth has left Americans a little confused.

VOICEOVER: Simple question. What’s actually healthy for you to eat? If you’re confused over what is and isn’t healthy to eat, you’re not alone.

BROADCASTER: Apparently Americans have a problem with nutrition, at least according to a new survey from the International Food Information Council. It finds that fewer than half of us know what foods and nutrients are important for good health.

NEWMAN: I feel, I do feel bad for the public because they’re constantly asked to evaluate very complex claims. If you’re the average Joe, do you know if blueberries are good for you or not? I mean, even nutritionists are still arguing about whether eggs are good for us after 50 years. And how are you supposed to evaluate that as a member of the public? And most people, you know, I think quite rightly get… get a bit lost in this, get a bit frustrated.

TWILLEY: So how did nutrition advice get so messed up? Why does it seem so hard to know what to eat to be healthy? Why can’t nutrition scientists figure this out? Is it all a scam like the blue zones, or are there actually foods we should eat for a healthy life? We are getting to the bottom of this after a quick break.

[BREAK]

GRABER: People for millennia have been trying to figure out what foods will make us feel good and keep us healthy. But modern nutrition science didn’t begin until the 1800s, and it started to have its biggest wins once vitamins were understood.

DARRIUSH MOZAFFARIAN: Many people don’t realize how common vitamin deficiency diseases were, in the world, in this country. Even a hundred years ago. So, millions and millions of Americans were affected by these diseases. These were harsh, awful diseases. They caused muscle deformities, bone deformities, deaths—much suffering.

TWILLEY: Dariush Mozaffarian is director of the Food as Medicine Institute at Tufts University. And as you may recall, we’ve actually made an episode all about vitamins and the story of how they were discovered, it’s called V is for Vitamin and it’s great.

GRABER: In the 17 and 1800s chemists started to break food down into its parts, they were understanding protein and fat and starches. But vitamins weren’t understood until the early 1900s, really just a century ago.

MOZAFFARIAN: The explosion of vitamin science that happened in the 1930s and 1940s really led to, to almost the eradication of these diseases in most countries.

MARION NESTLE: It was thrilling research, and it didn’t get finished until 1948. So, the whole first half of the 20th century was one thrilling discovery after another. I mean, that was the absolute apotheosis of nutrition research. It was the height, the most thrilling time, it’s when everybody was just so excited about it.

TWILLEY: Marion Nestle is a professor of food studies, nutrition, and public health at NYU, and she told us that, after that roaring success, in the fifties, nutrition science really was on a roll. It seemed as though it was on track to discover the ingredients to guarantee public health, like the secrets of what to eat for perfect health were finally within our grasp.

GRABER: This new knowledge helped battle some misconceptions about diet—for instance, one bit of diet advice held that children shouldn’t eat any fruit. Another popular line of thought in the 1800s was that raw vegetables were bad for you. Understanding vitamins changed how we saw fruits and vegetables.

TWILLEY: But it also just saved lots of lives. Darriush gave us some examples of how astonishing this transformation in nutrition science was. So, like, as late as 1940, something like three million Americans had such a severe vitamin B deficiency, that they suffered skin disease, diarrhea, dementia and even death.

GRABER: But then, once scientists understood what this disease came from, that people didn’t have enough vitamin B3, food manufacturers could add it to foods like bread and cereal.

TWILLEY: This happened across a whole range of different vitamins and their deficiency diseases. Fortification was truly a revolution in public health.

GRABER: And this whole science, of getting at what was in our food, the parts and chemicals that seemed to make food healthy for us, this totally transformed our understanding of and relationship to food.

MOZAFFARIAN: From what it had always been before—where, you know, we need to have some food, we need to enjoy it, we need to taste good. Yeah, there might be some things in there that you need to have to be healthy, but nobody had really knew what they were. Changed it to this new science of if we have the right vitamins and we have enough calories, we can technologically manufacture the right healthy food. And so the scientific community, the government, the food industry said, well, what’s the next battle? And heart disease was the first target.

GRABER: Heart disease was the leading cause of death in the 60s and 70s. And so based on their early successes, nutritionists thought they could easily win the battle against heart disease

MOZAFFARIAN: And with that focus, everybody—government scientists, industry—all, you know, intuitively, but mistakenly said, let’s use the same approach. There must be some nutrient, some single factor we can figure out that is a problem for heart disease. And if we figure it out, we can deal with it. And so, the overly simplified approach was- was taken to look at heart disease, and the villain that came out was saturated fat. That saturated fat was the cause of heart disease.

TWILLEY: There was some early data that seemed to show that saturated fat and cholesterol was a problem, and, flush with their early success, the nutrition scientists and the government officials charged with translating that science into dietary guidelines went all in. Among other things, they told Americans to steer clear of eggs, because egg yolks are high in cholesterol. And that’s how eggs first became bad.

MOZAFFARIAN: Well, it was a mistake. It was a big mistake, right? We, we oversimplified cardiovascular disease to one nutrient and in fact they said, well, it really is saturated fat, but let’s just say it’s all fat because saturated fat will be too complicated for people. So it actually just turned to a low fat diet approach.

GRABER: As Darriush says it was a mistake, later research has shown pretty clearly that fats aren’t all the same and that some fats in particular are really good for you. But at the time, the thinking became, just take all the fat out of food.

MOZAFFARIAN: But if you start wanting to take fat out of dairy foods, want to take fat out of meats, want to take fat out processed products. You start introducing all kinds of additives, emulsifiers, stabilizers, flavorings, surfactants— you know, other things.

TWILLEY: One thing that companies trying to avoid using saturated fats did was start using things called trans fats instead, which didn’t turn out so well. Trans fats are like hardened vegetable fats, and they in fact turn out to be associated with heart disease, stroke and diabetes.

GRABER: But then also, as food companies tried to make tasty foods that didn’t have any fat at all, they just added lots and lots of sugar. The most famous example of this are those fat-free Snackwell cookies that were super popular in the 1990s.

WOMAN 1: Excuse me. Do you make these delicious fat free Snackwell devil’s food cookies?

COOKIE MAN: Why, yes I do.

WOMAN 2: Wanna tell us why we can’t find ’em in the stores any more?

COOKIE MAN: Wha- what is this?

WOMAN 3: You got some explaining to do, cookie man.

WOMAN 2: Yeah!

COOKIE MAN: Well, they were a little more popular than we expected!

MOZAFFARIAN: They tasted awful. But provided added value, you could put really cheap ingredients in it and make a profit.

GRABER: Snackwells were made from sugar, white flour, more sugar, and a bunch of additives. They cost almost nothing to make, but they sold for a premium because they were supposedly healthy

TWILLEY: Long story short, whereas nutrition science in the first half of the twentieth century had notched up some real wins and saved a lot of lives, things in the second half were not going so well. And these mistakes didn’t just lead to disgusting cookies, which is already a tragedy. They actually damaged people’s health.

GRABER: The belief that you could make a new fat and it would be healthier, those trans fats, they were harmful. So was all that sugar people were downing. And then there was also a mistaken belief that people in developing countries were sick not because they didn’t have enough food and they were hungry, but because they specifically had diets that were low in protein. They weren’t malnourished, they were protein deficient.

TWILLEY: And that led to development agencies and NGOs pushing protein-fortified baby formula and toddler formula over breastfeeding. Which is something we talked about in our baby milk episode—it resulted in millions of deaths. All of this came out of this approach that was focused on finding the one villain: the one nutrient that we were missing, or the one nutrient that was making us sick.

MOZAFFARIAN: But we’ve learned now, 40 years later, that that approach—of a single nutrient causing a single disease—falls apart for chronic, complicated conditions like diabetes, cancer, obesity, heart disease, and so on.

GRABER: But also it wasn’t just the approach, this focus on single nutrients, that caused nutrition science to go off the rails. Another issue is that there have been some problems in how nutrition research was done in the past.

MOZAFFARIAN: The original observational studies that led to the recommendation for low fat diets were what we call cross-national studies or ecologic studies, where you compare Greece to France to the US to, you know, to Italy. Like there was a study called the seven country study that did that. That’s a very, very flawed way to do science, to compare a whole country to another whole country and try to understand—you know, make cause and effect.

TWILLEY: Imagine all the possible things that could be causing a difference between one country’s health outcomes and another’s that are *not* to do with fat intake at all. I mean, income levels, air pollution, healthcare provision, stress, physical activity—all of these things could and do vary wildly between different countries, and we know they affect health outcomes.

GRABER: But then even at the level of just getting accurate data about what people were eating in order to connect diet and health outcomes? Nutrition scientists tried to do that in a number of ways. A major one is called recall, asking people what they ate, or how frequently they ate particular foods like last week or over the past month.

NESTLE: The hardest thing in the world in nutrition research is to find out what people are eating. Because, people eat very varied diets from day to day, from week to week and from year to year. They don’t remember what they eat. It’s very difficult. I can’t remember what I ate yesterday. So what you want to do is an easier way, and those are food frequency questionnaires. In which you give people a big questionnaire and ask them, did you eat this, this, this, or this? Within the last day, week, or month. And so people fill out these forms. And say whether they ate this, that, or the other thing. And there’s a great big long list. You know, for some things, you say: Well, okay, I have coffee every day. I know I have coffee every day. That’s one thing. But how many times did I eat fish in the last—? I mean, I just can’t.

TWILLEY: Recent research, using a super accurate way of measuring energy expenditure—its called doubly labeled water, and it works by measuring chemically labeled molecules in the body—this research has found that people’s estimate of their calorie consumption in these kind of food frequency questionnaires is off by about 60 percent. That’s a big difference between recall and reality.

GRABER: That’s just overall calorie consumption, like they may be eating 60 percent more sheer food than they think they are. But then you can imagine that the same thing would be true for individual foods that people eat, we just can’t get that right in surveys, we can’t accurately remember our past meals. And if you don’t know what people are eating, it’s hard to figure out how their diet is affecting their health

TWILLEY: You could have people write things down or even photograph it at the moment they’re eating it. But Marion told us this method, it gets around the memory problem, but it’s flawed too.

NESTLE: You have to measure every single thing that you’re eating, and the act of measurement changes what you eat.

GRABER: Nicky and I had to do this for our episode on personalized nutrition, and it definitely changed what we ate. Sometimes we couldn’t be bothered with all the rigmarole of measuring and weighing, and so we just didn’t eat something. It wasn’t an accurate representation of our usual diets.

TWILLEY: This Schrodinger’s cat problem, of when you study something you inadvertently change it—it happens on a more meta scale too. It’s why studying food is harder than studying, say, a new pill, in pharmaceutical trials.

NESTLE: It’s much more difficult. Because you can’t blind them. You can’t double blind them.

GRABER: Like for a new drug, a study participant is blinded because they don’t know what drug they’re getting, or maybe they’re getting a placebo. In a double-blinded study, even the scientists don’t know which one a particular patient is getting until the study is over and they’re ready to analyze the results.

TWILLEY: With food, you can’t do that. It’s recognizable.

NESTLE: And the minute somebody knows what’s being studied and what they’re eating, they have some idea of what they’re supposed to say. So that influences it.

TWILLEY: And in addition to all of these problems, of accurately measuring food intake and then not altering that intake by measuring, you have the more fundamental challenge that humans are humans and they’re not especially good at following instructions.

GRABER: This would be an issue for a type of study where people are told what to eat, and then scientists look at the impact of that.

NESTLE: So clinical trials for dietary intake are usually done by people who are free living. That means they’re doing their thing. And you tell people to eat one kind of a diet and you hope that they do, and maybe you’ll have a biochemical test that will show how much compliance they had with that diet. But mostly they’re really difficult to do and they don’t usually show much. Because people are people. And they don’t necessarily follow the protocol. It’s very hard to stay on a diet for a long period of time.

GRABER: And that question of time, doing something over a long period of time—that’s important, because foods can take a while to have an impact on our overall health, and so it’s important to study these things for a long time. People tend not to have heart attacks until they’re older and they’ve been eating their diets for decades.

TWILLEY: Time is an issue, eaters who won’t follow rules and don’t remember what they had for breakfast are an issue. But the other problem is people who have a story they’re trying to make the evidence fit—kind of like Dan Buettner and the blue zones.

GRABER: The whole red wine back and forth is a perfect example of this. By the end of the 1990s, there were literally hundreds of studies showing that if you drank red wine every day, you had way less chance of dying from heart disease.

TWILLEY: A lot of these were done because researchers were fascinated by this phenomenon that was called the French paradox. Basically, in the 1980s, a handful of French researchers were like wow, look at us, we eat triple creme brie and duck fat and pig’s heads and yet we have very low rates of heart disease.

GRABER: One of the leading theories was that somehow the French got away with eating a diet that would give everyone else a heart attack on the spot by washing it all down with red wine.

MORLEY SAFER: There has been for years the belief by doctors in many countries that alcohol, in particular red wine, reduces the risk of heart disease. So the answer to the riddle, the explanation of the paradox, may lie… in this inviting glass.

[CLOCK TICKING]

GRABER: When Morley Safer did an episode of 60 Minutes on this in 1991, red wine sales in America went through the roof. And researchers did a bunch of studies looking at what it was about red wine that was reducing heart disease risk, they concluded that it was probably because of a particular chemical in red wine called resveratrol.

TWILLEY: But hundreds of studies have since failed to show any positive impact of resveratrol. And when researchers went back to look at all this French paradox data again, they realized the original studies had failed to take into account a bunch of potential complications.

GRABER: For one, the French were under-counting coronary heart disease. They only recorded it as the cause of death if the heart attack was what killed you. But we record coronary heart disease as being present at death even if something else killed you in the moment. Like maybe someone had a couple of heart attacks, but died of something else. So it looked like Americans had more coronary heart disease than the French, but that wasn’t necessarily the case.

TWILLEY: For another, moderate red wine drinkers are often more educated and wealthy. They are also often likely to eat more vegetables, do more exercise, and have health insurance. Those things can definitely help reduce your risk of dying of heart disease.

GRABER: So sometimes people are trying to make the science fit a story they’d like to tell. But then, like in the case of eggs, the reason for the back and forth has to do with advances in science. Like we said, eggs had been demonized because they have cholesterol in them, dietary cholesterol, and scientists kind of assumed that eating cholesterol led directly to higher levels of cholesterol in your blood, and it turns out that’s not really the case.

TWILLEY: That’s one area where the science has just progressed. Which Darriush says is not surprising—that’s what science is supposed to do.

MOZAFFARIAN: Yes, absolutely, the science has changed. From 1980 when we thought low fat, low cholesterol, low saturated fat, eggs are bad—you know, let’s all have, fat free cookies, fat free frozen yogurt, fat free salad dressing. The science has changed. That was wrong.

GRABER: Science changes, that happens in all types of science, there are new studies and they may overturn some particular previous understanding. That’s normal. But Darriush says that we pay a lot closer attention to these changes when it comes to the science of what we eat, which makes them seem a lot more shocking.

MOZAFFARIAN: I’m a cardiologist, and in my career, we’ve had dramatic changes in the understanding of heart attacks and the treatments for heart attacks, from the kinds of blood thinners we use, the kinds of cholesterol lowering medicines we use, to the kinds of procedures we use to open the heart arteries, to how we do surgery. Many, many, many changes. Or, you could take another example of physics, right? In physics, in just the last 50 years, you know, we’ve gone—or the last 100 years, we’ve gone from Newtonian physics to quantum mechanics to the discovery of dark matter and dark energy. Totally upending previous, you know, lines of science. But you know, not everybody’s a physicist, not everybody’s a cardiologist. So when those, when those scientific fields change, most people don’t notice. But when nutrition changes, it’s personal, right? When, when somebody goes from telling you, hey, that dark chocolate was bad and now it’s good or vice versa. Or eat the steak or don’t eat the steak. It’s personal.

TWILLEY: At this point you might be thinking wow nutrition science seems like a hopeless case. But that’s really not true. Yes, doing good nutrition science is hard, and there’s no doubt that it can be really difficult to tease out the impact of diet from all the other things that contribute to the kinds of chronic diseases like heart disease, diabetes, cancer that we face in the west today.

GRABER: But also, there is a lot that can be figured out using the basic tools of nutrition research, even with their flaws. First of all, even the maybe weaker studies asking people about what they ate, if enough people are involved and if scientists do enough of them, they can find out things that are important—like, that’s how the finger was first pointed at trans fats.

TWILLEY: And of course like in many fields of science, nutrition science has made progress over time, and today we have better techniques that produce more robust results. So for example, rather than just looking backwards at existing data to try to see patterns over time—that’s called a retrospective study—scientists these days are more likely to measure people going forward, setting out their endpoints at the start, to try to avoid that kind of making the evidence fit the narrative problem.

MOZAFFARIAN: Modern nutrition science really uses prospective studies that get people at baseline and follow them forward in time for years, and assess their diets in lots of different ways.

GRABER: And then another thing scientists can do to make their data even more accurate is look at biomarkers in the blood. For some foods they can find things that will let them know how much of a certain food has been eaten or not.

TWILLEY: And increasingly, they can link certain biomarkers to specific longer-term health outcomes, to get around the issue of how long a study can be versus how long chronic diseases take to progress.

GRABER: All of these different types of studies are important, they all offer a different method of looking at nutrition: maybe trying to understand specific foods, or more general dietary patterns.

KEVIN HALL: And so it’s by putting together these pieces, from multiple different study designs, that allow you to kind of triangulate towards something closer to the truth.

TWILLEY: That’s Kevin Hall, he’s senior investigator at the National Institute of Diabetes, Digestive, and Kidney Disease, which is one of the National Institutes of Health in Bethesda, Maryland. And he is currently running a quite different kind of nutrition science study, designed to get to the bottom of why ultra-processed foods seem to be so bad for us. That story, after this word from our sponsors.

[BREAK]

GRABER: We told you that you can use large-scale studies that look at what people are eating in their normal life and try to understand something about their diet, these studies can help point the finger at potential problems. They’re called epidemiological studies, and lately scientists say that a bunch of them have shown that diets high in ultra-processed foods might not be great for us.

HALL: That was a kind of common recurring theme was that diets high in ultra-processed foods seemed to be associated with increased risk for obesity.

TWILLEY: Obesity is one of the most complex and common chronic health issues in the developed world today. Each year, the health impacts of obesity at a population scale are estimated to be responsible for millions of deaths from cardiovascular diseases, diabetes, certain cancers, et cetera, and the number of people who have obesity is going up every year. So if there’s something about ultra-processed foods that is causing obesity, we need to know what it is.

HALL: One of the challenges with epidemiological studies is that people often say, well, they can demonstrate an association, but they don’t really say anything about causation.

TWILLEY: Before we get into what kind of nutrition science experiment *could* demonstrate causation, let’s back up for a minute and get into what exactly ultra-processed foods are.

GRABER: In 2009 a Brazilian scientist named Carlos Montero was one of the ones who was pointing at ultra-processed foods, or UPFs, as a problem, and he defined them. He said the first category of food is not processed—like apples or eggs. Then there’s category two, that’s a processed ingredient like olive oil or butter. Category three is a lightly processed food, like cheese, pasta, or home-made lasagna.

HALL: And then there’s the category four foods, which are the so-called ultra-processed foods. And essentially it’s these combinations of ingredients typically that you wouldn’t be able to find in In your home kitchen or, or even a typical restaurant.

TWILLEY: What’s interesting is that different versions of foods can be in different categories depending on how they’re produced. So, like your artisanal sourdough that’s just made with wheat and water, that’s category three. But a supermarket sliced loaf that has lots of preservatives and dough conditioners and other stuff, that’s category four. Unflavored plain yogurt can be category three, but if you add flavors and emulsifiers and what not, then it’s a four. The difference is really whether it’s an industrial product or not.

GRABER: But what makes this approach different is that it’s not connected to particular nutrients.

HALL: Right, so, it doesn’t talk about saturated fat, doesn’t talk about sugar, doesn’t talk about high fructose corn syrup, directly. It doesn’t talk about, you know, the, the salt content or sodium content of, of the foods. But of course, a lot of the ultra-processed foods that are in category four tend to be high in saturated fat and sodium and sugar and things like that.

TWILLEY: These are things like industrially produced sausages and cookies and snacks.

GRABER: So a lot of these foods *do* have ingredients or nutrients in common, but not all of them do—like ultra-processed bread doesn’t necessarily have a lot of fat and sugar. That’s what makes this approach quite different from nutrition studies in the past, it’s trying to understand a category of food, not a single nutrient.

TWILLEY: Another big difference is how Kevin is studying it. Like Cynthia said, there have been a bunch of the standard kind of observational epidemiological studies that link eating lots of UPFs with negative health outcomes. But those kinds of studies can’t necessarily get at why—why a diet full of UPFs somehow seems to lead to obesity and obesity-related health issues.

GRABER: Kevin’s idea was that maybe there’s something about UPFs that just makes us eat more of them. That there’s something about them that just makes people literally consume more calories. And so he wanted to try to design the best possible study to really drill down and figure out if that was true. He wanted to get around the issues of food recall and of people not following the diets they were told to eat.

HALL: You know, the challenge is we’d like to be able to have, you know, ultimate power over the entire universe, of course, as scientists. And control everything. [LAUGH] Of course, we don’t have that kind of power.But our idea was, maybe if we brought people in to live in the NIH clinical center, which is a hospital setting. And if they lived with us 24 hours a day, seven days a week, we could take control over their entire food environment.

TWILLEY: So starting in 2018, that’s what Kevin and his collaborators did. Participants in his study had to come stay at the NIH clinical center in Bethesda for four full weeks, no weekends or evenings off. And during that time, they got two weeks of a 100 percent ultra-processed meal plan, followed by two weeks of a minimally processed diet, or vice versa.

GRABER: Overall, the meals as they were presented to the participants were exactly matched for nutrients—fiber, sugar, protein, fat.

HALL: And they’re given really simple instructions, just eat as much or as little as you want. We’re going to measure lots of stuff. We’re drawing blood all the time. We’re measuring their body composition changes, their body fat, their lean tissue mass and things like that.

GRABER: There’s no way for them to cheat and get ahold of some other food. Their exercise was carefully controlled, their leftovers were measured.

TWILLEY: And the diets are designed so participants couldn’t necessarily tell which is supposedly healthier and which is not.

KEVIN HALL: Yeah. So just an example of a, of an ultra-processed meal would be, kind of a typical box breakfast cereal.

TWILLEY: Honey Nut Cheerios, one of the preferred breakfast cereals of my youth, were on the menu.

KEVIN HALL: You know, with some milk and maybe a blueberry muffin that was pre-made. And, that would be a typical ultra-processed breakfast. A minimally processed breakfast might be some, kind of plain Greek yogurt with some bananas and berries and some, some mixed nuts or something like that.

GRABER: As Kevin said, the participants were allowed to eat as much as they wanted. They were asked how pleasant they found the food. And they didn’t rate the ultra-processed meals as better, they were both fine and dandy. And the participants were also periodically asked how full they were, and no matter their diet, they didn’t report being hungry.

HALL: But, despite the no differences in appetite, they ended up consuming about 500 calories per day more on average during the ultra-processed food environment. And they were gaining body weight, and gaining body fat. Importantly, they were blinded to their weight measurements. They kind of have their backs to the scales when they’re being measured every day and they’re wearing loose fitting scrubs. So they can’t tell if their clothes are getting tighter or looser. But they were spontaneously gaining weight on the ultra-processed food environment. And the same people, when they were in the minimally processed food environment, just spontaneously lost weight.

TWILLEY: So this is a significant result. And a real insight into why all those other studies connected eating UPFs with obesity. But it still doesn’t get at the why beneath the why—in other words, why people ate more foods that were ultra-processed.

GRABER: There are lots of ideas about why this might be the case. One of them is that the super industrial food is so soft that you don’t have to chew very much, and it goes down really easily.

HALL: And maybe by the time your gut is signaling to your brain that you’ve had enough to eat, it’s already too late, cause you’ve eaten the food so quickly,

TWILLEY: Another theory was that some of the ultra-processed foods are also something that researchers have defined as an hyperpalatable food.

HALL: Which are foods that have pairs of nutrients that cross certain thresholds. So in other words, instead of just looking at sugar per se, what about a food that’s high in both fat and sugar. Or a food that’s high in both fat and salt. Or both high in carbs and salt.

GRABER: The idea is that because they’re high in two of these things rather than just one, they really trigger our pleasure centers and make us overeat. And then maybe it could be something else—like maybe some of the preservatives are messing with our gut microbiome, and that’s making us absorb more calories. or it’s affecting our gut hormones that tell us we’re full…

HALL: And the list kind of goes on and on.

TWILLEY: So Kevin decided to do a second sleepover study designed to figure out what exactly in these ultra-processed foods is the trigger for eating more calories.

HALL: And we said on a meal by meal basis, what are the properties of the meals themselves that seem to kind of be best correlated with how many calories people choose to eat?

GRABER: This time there were four different types of meals. One week the participants ate only minimally processed food. The other three weeks they ate ultra-processed foods, but those weren’t all the same thing.

TWILLEY: They were all slightly differently formulated to try to tease out whether the issue was really hyperpalatability, or instead maybe all the stabilizers and emulsifiers used in a lot of industrially processed food, or what.

GRABER: This trial is still underway, Kevin’s working his way through 36 participants and only two can come to NIH at a time, this takes a long time to do. But he spoke publicly about the results at the midway point of the study earlier this year, because they’re looking pretty clear so far.

HALL: We got a pretty consistent picture, that, it seemed to be that meals that had higher energy density. Meals that were eaten more quickly. And meals that had more of these hyperpalatable foods tended to cause people to consume more calories on a given meal. And when we ended up looking at it, it seemed like the bigger effects were the hyperpalatable foods and the energy density.

TWILLEY: A great example of an energy-dense ultra-processed food is a Twinkie snack cake. Or Combos, those cheese or pepperoni-filled pretzel bites. These kinds of foods are low in fiber, low in water, and packed with fats and starch and calories. And according to Kevin’s study, those seem to be the kinds of ultra-processed foods that trigger people to consume more calories.

GRABER: Whereas maybe supermarket wheat bread or a frozen microwaveable meal, which also are ultra-processed, these don’t seem to be making people overeat, at least based on Kevin’s early findings.

TWILLEY: So if these results hold up, at the end of the study, Kevin will be able to say that within this whole huge category of ultra-processed foods, it seems as though there are particular ones that are causing people to overeat.

GRABER: And that matters, especially because more than half of what adults are eating in the US, on average, falls in this ultra-processed food category.

HALL: And some people have, you know, go so far as to suggest, oh, well, we just have to eliminate all ultra-processed foods. Well, they’re about, you know, depending on who you believe somewhere between, you know, 50 and 70 percent of the food supply in the US. And you know, most of us who—including me-—who still consume ultra-processed foods would have a very difficult time finding both the money and the prep time and the skills to prepare all of our meals from scratch.

GRABER: In other words, if something about ultra-processed foods is making people overeat, we need to know what that something is—because most people don’t think it’s realistic to just cut out the whole category.

TWILLEY: This is why Kevin thinks it’s so important to really try to dig into the mechanism behind the health impacts. There are so many possible causes for why ultra-processed foods might be bad for us, but doing the research to tease out what it is about them that is leading to obesity is a key first step.

HALL: Because if we could figure that out, then we could, you know, inform people who make policies about what it is that we need to do in order to shift the food supply to one that doesn’t drive excess calorie intake. We could inform consumers about what kinds of products are likely to cause overeating and excess calorie intake and weight gain. And we could inform manufacturers about what it is about some of their products, and maybe not others, that are driving these potential problems.

GRABER: Like if it’s not the preservatives, then you don’t want to set up a situation where what happens is policy makers will decide that particular preservatives are bad, the ones that are currently used are taken out of circulation, and then the food industry just finds a different one, or food goes bad—but also our health hasn’t improved. Like when we got rid of fat and everyone ate Snackwells, nobody got healthier.

TWILLEY: Kevin and his collaborators went to great lengths to make these studies as accurate as possible: weighing everything and matching the nutrients and so on. But there have still criticisms of his studies. For one, to bring the fiber levels in the ultra-processed meals up to the same level as the minimally processed meals, they had to put a lot of soluble fiber into the drinks.

HALL: And some folks came along and said, oh, well, wait a second. That’s not right. Because beverages have a very different effect on appetite than solid foods.

GRABER: This is something they tried to control for in the second study, but it’s an issue. Another one is that in the first study where people clearly ate a lot more calories on the ultra-processed diets, they seem to have front loaded that consumption—they ate a lot more at the beginning, but it kind of evened out by the end of the two weeks. So it’s possible that people wouldn’t have continued eating so much more of the ultra-processed meals.

TWILLEY: There’s also some things a study like Kevin’s can’t do. It can’t definitively say that *the* only way ultra-processed foods are bad for our health is because we eat more of them when they’re hyperpalatable and energy dense. It can’t rule out that other things might be going on that are also having a negative impact on our health.

GRABER: It could still be that there are also ways that these foods are changing our gut microbiome and that’s having an impact on our health in some way. It could be affecting our hormones. As we mentioned there are still all sorts of theories.

TWILLEY: The maybes and what-ifs can seem endless, but the way science works is you just have to keep accumulating evidence—ideally from a bunch of different well designed studies. You can’t just figure this stuff out from one experiment, no matter how compelling that experiment sounds and how clear the results seem.

HALL: There’s no one single study that should dictate our behaviors as a society or as an individual. That we have to look at multiple studies mounting on top of each other that paint a consistent picture about what’s going on. So, you know, as much as I’m flattered by how much our studies get attention and—you know, would I make recommendations based on only one study? No, absolutely not.

GRABER: Part of the reason Kevin’s research is getting so much attention is that it seems like the gold standard. He and his colleagues have everything carefully controlled, they change only one variable of what people eat, they measure everything they can.

TWILLEY: But like we said, even this kind of rigorous, randomized controlled trial has weaknesses and can’t tell us everything. That’s why to arrive at accurate nutrition advice we need all different kinds of studies: ones that measure biomarkers in blood, ones that look at populations and follow them over time, ones that test specific compounds from foods in the lab to see what happens.

GRABER: The problem is that doing all of this research is expensive, and Kevin’s in-house controlled study is particularly pricey.

NESTLE: Those are phenomenally expensive to do, can only be done on a few people at a time, and NIH seems extremely unwilling to pay for more of that.

TWILLEY: On top of that, right now, If the current administration has its way, the NIH won’t be able to pay for any trials like this, and, if you can’t do trials like this, then you are missing a vital tool to really understand how food affects our health.

GRABER: Industry could certainly fund it, and they do fund a lot of nutrition research.

NESTLE: And food companies are very happy to fund research in their interests. They are not so happy to fund research that doesn’t produce results that they like. But a lot of—you know, increasing amounts of nutritional research are being food- funded by the food industry. And there, it turns out that there is a rather large body of research literature that studies the effect of industry funding on the outcome of research. And it shows, quite conclusively, that industry funding influences the outcome of research. Surprise! People are nice to their funders. They like funders. But that, the influence is received unconsciously by the people who receive the funding. They didn’t intend to be influenced. They don’t believe they were influenced. They don’t recognize the influence. And they deny the influence.

TWILLEY: Darriush told us that even with this unconscious influence problem, he thinks there is room for industry funding, if there are sufficient guardrails in place. Marion and Kevin suggested one effective guardrail would be if the industry funds are pooled, and then parcelled out by an external institution, for example.

GRABER: Which, of course, industry does not want to do. The orange industry, for instance, doesn’t want to put money into a pool to fund who knows what, they want to fund research into oranges—in particular research that’s going to show how great oranges are for us.

TWILLEY: But all our experts are agreed—and we are, too—that government actually does have to step up to fund nutrition science. It’s so essential to our health, it needs investment.

MOZAFFARIAN: In the last 50 years, as a percentage of its research, research at the National Institutes of Health or NIH, on nutrition research has been flat. At, at a maximum, it’s about four percent of funding, but I think that’s a very optimistic estimate. It’s probably much, much, much less. And yet this is at a time when diet-related diseases have skyrocketed.

GRABER: To be honest, on the one hand, nutrition science has already come to some pretty clear conclusions about what the best basis of a healthy diet is, at least for most people. Amusingly, it’s similar to what the blue zone promoter was saying, except without the daily glass of wine and without fudging the data.

NESTLE: I think we know what people can eat to keep themselves optimally healthy, and that is, you know, eat food, not too much, mostly plants. Doesn’t mean no meat, it doesn’t mean no junk food, it doesn’t mean no ultra-processed foods. It means just keeping them… modest.

TWILLEY: On that level, nutrition science has already figured it out. Which is not nothing. Like we said at the start of the show, it’s impressive in just a century to have come from a world where deficiency diseases were common to one where we actually know how to eat for health, broadly speaking.

GRABER: And part of that came from that early nutrition science that helped us understand for example what nutrients are in whole grains versus white flour. That said, there is still a lot of misinformation out there—in part, sure, because science has evolved and changed, but also because people want an easy fix.

NEWMAN: Yeah, I mean, there’s a nice reason, you know, people have been selling aging cures forever. It’s kind of—we don’t like the actual cures. Jogging is hard, right? [LAUGH] It’s not—you want it to be tea. [LAUGH]

TWILLEY: I mean, Saul is not wrong—it would be nice if there was a painless silver bullet. Knowing something is good for you and doing it are two very different things.

GRABER: But Darriush said that even better communication and convincing people that eating whole grains and beans and vegetables and oily fish are great for you—that’s still only one part of the solution.

MOZAFFARIAN: Knowledge and education alone is not enough to solve the problem.

TWILLEY: There’s only so much science can do, even really great science, when it comes to actually fixing our food environment.

MOZAFFARIAN: When most adults are sick, when the vast majority of people are sick, it’s not a problem of individual willpower anymore. It’s a problem of a broken system. And a broken system needs system solutions.

GRABER: There are all sorts of issues with our food environment that make it unhealthy, and we’ve talked about them on Gastropod before. They’re beyond the scope of what most nutrition scientists are able to work on. Poverty, farm subsidies that support the production of commodity crops that go into ultra-processed foods, rather than fruits, veggies, whole grains and beans. Whether people can access fresh or frozen fruit and vegetables. Whether they have time to cook for themselves. How stress and discrimination are affecting their diets and their health. There’s a lot of systemic issues that affect what people can and do eat, and how their bodies respond.

TWILLEY: These are issues for policy makers. But there is still work for nutrition scientists to do, too. For one, they can try to help make the convenient industrial foods that form the basis of the American diet healthier, by figuring out what about them is so bad for us. That’s what Kevin is working on.

GRABER: And then, even though the broad principles of a healthy diet are well known and not up for debate, there are plenty of important and interesting questions for specific populations, in terms of what to eat for better health.

MOZAFFARIAN: If I have inflammatory bowel disease, does having omega-3s make a difference? If I’m gluten sensitive, you know, what does that mean? Why am I gluten sensitive? And is it actually the gluten? What if, especially for disease treatments, right? What if I have cancer—are, you know, certain ketogenic diets good for brain cancer, like glioblastoma? You know, so there’s so many questions that we need to answer.

[MUSIC]

GRABER: We do have one more piece of nutrition advice, and this is coming from us and from all of our guests this episode: do not trust one single study that tells you that one particular food is either awesome or heinous. You can and should ignore all those headlines.

TWILLEY: If that’s the only thing you take away from this episode, our work is done! Even though we won’t become millionaires with that kind of advice, dammit. But truly, that is where a lot of the confusion comes from these days—media reports about a new study that shows X or Y. When you hear the words “a new study,” just channel Kevin’s voice telling you that you cannot make dietary recommendations based on just a single study!

GRABER: Thanks this episode to Saul Newman, Marion Nestle, Darriush Mozafarrian, and Kevin Hall, we have links to their research and books on our website, gastropod.com.

TWILLEY: And as always huge thanks to our producer, Claudia Geib. We’ll be back in two weeks with another brand new episode for your listening delight. Till then!