Bad Science and 5 Nutrition Ideas

Understanding how research goes wrong can help you make smarter food choices.

Nov 17, 2022

One of the toughest parts of eating healthy is wading through the swamp of conflicting nutrition advice from very smart people. Whatever diet you want to follow (or are told to follow) you can find a peer-reviewed study or doctor to support it. Vegan, paleo, vegetarian, fasting, carnivore, low-fat, low-carb, and other diets all pull from research to argue their diet is the best.

A few years ago, I wrote an investigative feature story for Men’s Health on saturated fat and its role in health. But the story ended up being a much larger commentary on the state of nutrition science—how we know what we know and how that should guide our eating habits. This week’s 2-Percent newsletter will dive into the shortcomings of nutrition science. If you just want advice on how this information can guide your eating, scroll to the bottom. Sign up here if you were forwarded this email.

To understand the history of nutrition in the United States of America and what we’ve thought is “right” over the last four decades, you have to understand nutrition research at Harvard University. And to do that you have to understand Walter Willett.

Willet, 77, is 6’2” and slim but sturdy. Head of silver hair and thick, wispy white mustache. Picture a soon-to-retire park ranger who probably has a kickass woodworking setup in his garage.

Willett graduated from the University of Michigan Medical School Summa Cum Laude when Nixon was in office. “When I was practicing medicine, I became frustrated because my patients had conditions like diabetes, hypertension, and heart disease that I couldn’t cure,” said Willett. “I wanted to understand what was causing the conditions to prevent them in the first place.”

So he entered the Harvard School of Public Health for a doctorate in epidemiology. He joined the department as a professor after graduating in 1980 and began studying nutrition.

Until roughly the 1940s, most Americans didn’t worry whether food would blow up their waistlines or clog their arteries. People then were far more likely to die of malnutrition or infectious diseases. Food was medicine in the sense that eating more of it moved you away from death or helped you recover from our microbial killers. The government gave advice like “eat a lunch that packs a punch!”

Malnutrition, scientists were finding, was often caused by deficiencies in a single nutrient, like iodine or vitamin b3. So the government began to fortify foods. For example, iodizing salt and adding vitamins and minerals to cereals. This is a practice still continued today that saves lives and prevents disease around the world.

By the 1950s, the modern food and medical systems were rendering malnutrition and infectious diseases moot. We had food, plenty of it, and ate it like times were still lean. And that, to make a long and complex story short, is when too much food, rather than too little, started causing our health problems. But the success of fortification gave scientists the mindset that disease was determined by single ingredients within foods. For example, salt, sugar, or saturated fats, which are the type of fats mostly found in meat and dairy.

In the 1980s, Willett developed rigorous diet questionnaires and distributed them to 120,000 nurses. The nurses reported the foods they ate and how often. Willett would track what sort of diseases the nurses got, looking for links between eating habits and health. “We set the study up to look at the type of fat because that was the concern at that time,” Willett told me. The work moved the ball downfield significantly. He discovered the heart harms of artificial trans fats, which were blindly concocted as a “healthy replacement” for saturated fats. Thanks to Willett’s research, fake trans fats were banned by the FDA in 2013.

Willett’s work has also consistently shown that for heart disease (the number one killer of Americans) saturated fats aren’t bad. But they also aren’t good. “It’s really about comparison,” said Willett. “If you compare saturated fat to trans fats, then saturated fat looks good. But if you compare it to unsaturated fats, then saturated fat looks bad.” Unsaturated fats are the type found mostly in plants. Olive oil, nuts, avocados, etc. This is why Willett calls saturated fats “middle of the road” fats.

In a recent review of the research, Willett showed that heart disease rates drop slightly when people swap excess saturated fat for unsaturated fat. And so, “reducing red meat and dairy products in a food supply and increasing intakes of nuts, fish, soy products, and (vegetable oils) … will have a markedly beneficial effect on rates of coronary heart disease,” Willett wrote in the study.

“Saturated fat can never be zero, because it’s intrinsic to a lot of foods, a lot of healthy foods too,” said Willett. “But if you ask, ‘what does an optimal diet look like?’ then most people will agree that it will be relatively low in saturated fat.”

Willett’s research is the reason why the government nutritional guidelines tell us that saturated fat should be less than 10 percent of what we eat. Yet we suck at this. About 70 percent of us overdo saturated fat and 1 in 4 of us die of heart disease.

But there’s a catch to most nutrition research.

Willet’s big studies—and nearly all epidemiological nutrition research, which is research conducted on large populations of people—hinge on what researchers call “food frequencies questionnaires.” Harvard’s version, written by Willett, asks participants to recall exactly what they ate and how often over the last year.

“The single best way to understand the shortcomings of fruit frequency questionnaires is to take one,” my friend Tamar Haspel, a columnist for the Washington Post who recently wrote a wonderful book(!), told me. Seriously: Take one.

Food frequency questionnaires include hundreds of questions about how often you eat specific quantities of specific foods. The questionnaires have various answers ranging from “never,” to “once a week”, to “2 or more servings a day,” etc.

For example, one asks, how many times in the last year did you eat three slices of peppers? How many times did you eat a ¼ cup of salsa, picante, or taco sauce? How many times did you eat one muffin or biscuit? How many times did you eat 4 to 6 ounces of steak? How many times did you eat one tablespoon of peanut butter?

The researchers then comb through all the participants’ answers and health records to find links between food and good or bad health.

“So let’s say you find that people who eat lots of red meat have more heart attacks than people who eat less red meat,” the nutrition scientist Stephan Guyenet told me. “You could see that and say, oh, red meat causes heart disease. Or you could say, well, actually maybe the same people who eat red meat also smoke more, or exercise less, or both, or any other things that might impact heart disease.”

Guyenet said he’s “very uncertain about what exactly this epidemiological research means. I’m not even really sure how to interpret it. Is it pretty reliable? Is it not reliable at all? I really don’t know.”

In 2013, Stanford researcher John Ioannidis penned a now-famous BMJ piece titled “Implausible results in human nutrition research.” He pointed out “almost every single nutrient imaginable has peer-reviewed publications associating it with almost any outcome.” These mixed outcomes are why news outlets one week will report that a food like coffee or dark chocolate is a ticket to a quick and painful death and the next say that the food is the fountain of youth.

“The food frequency questionnaire approach is inherently unscientific. People consistently misjudge serving sizes and everyone underreports massively,” Trevor Kashey, Ph.D., a former cancer researcher (and friend of the 2% Newsletter) who now owns Trevor Kashey Nutrition explained.

Another hiccup: The less healthy a person is, the less likely their answers will be accurate. One analysis discovered that people who are at a healthy weight underestimate their daily calorie intake by 281 calories, while obese people underestimate by 717, the equivalent of a Big Mac and small fries. Now extrapolate that over 364 more days.

“So if the data is that convoluted, then it probably means the true impact on human health is just as convoluted,” said Kashey. “But that doesn’t mean big groups of humans are not worth studying.”

Willett said he always knew that his questionnaires could never be perfect. “The topic is infinitely complex and if you want to have perfection you will never learn anything,” Willett told me. “Just go through a list of pretty much everything important to public health, like environmental contaminants. We make decisions off epidemiology.” An example: His trans-fat work still to this day saves 50,000 American lives each year, according to the National Institutes of Health.

Which is all to say that the topic is infinitely complex. Diets make big claims based off of very uncertain data. Given how convoluted so much of our nutrition science is, how should we act? Here’s what I think (for now, anyways):

1. Eat foods that help you keep your weight in check

Some nutrition research might be shaky, but the medical research is conclusive: Being too big is associated with all kinds of health problems that make us more likely to die early.

And when more in-depth nutrition studies (read: not the big survey studies we’re talking about above) control for calories, scientists find no difference in health markers between people who eat a diet composed of heavily-processed foods versus “healthy” foods like fruits and vegetables, whole grains, and lean meats.

It's surprising. But what this tells us is that keeping your weight in check is likely far more important than any one food you eat. It’s probably not as bad as you think to eat “unhealthy foods” so long as you can do so without gaining weight. Because remember, the study findings above only apply when the study participants don't gain weight.

That said, some foods allow us to keep our weight in check easier than others. And so …

2. Eat mostly single-ingredient foods

Some people can live on hyper-processed junk and stay at a healthy weight (It’s not genetics. To know why, read The Comfort Crisis).

But most people most of the time have an easier time keeping their weight in check when their diet contains mainly those aforementioned “healthy” foods. It’s not that vegetables, fruit, meat, rice, oatmeal, etc are magic. Rather, they make us feel fuller on fewer calories. And that keeps us from overeating and breaking rule one.

3. Eat a variety of foods

When I reported that story for Men’s Health, some Ph.D.-holding nutritionists told me we should limit eating red meat to once a week. If we must—and it's probably best to go vegan. Other Ph.D.-holding nutritionists said eating two pounds of red meat a day was healthy. For fans of the former, we have the vegan or vegetarian diet. For fans of the latter approach, the carnivore diet.

But Tamar Haspel (mentioned above) once told me something smart. She said that if the research is all over the place, with every food being linked to both health and disease, we probably don’t want to go all-in on any one food or method of eating. Think of it like your financial portfolio. Diversifying is a hedge if scientists end up being blatantly wrong about any one food. If you eat red meat three times a day and the "meat is unhealthy" research was right, you'll find yourself in quite a pickle.

4. Question any research that suggests that one food or one type of food is magical for health

It’s not. Humans are amazing in our ability to eat and process a wide array of foods.

5. Don’t be a food fascist

Yes, some people have allergies and should worry about not eating, say, peanuts or gluten. But you only live once. Eat carbs. Eat fat. Eat dessert. Do double deserts for all I care. So long as you follow rule one and stay active, you’re doing what most of the research suggests is most reliably good for you.

2% Top Two

My two favorite things this week ...

1. A Thanksgiving Must-Have

My friend Casey Bard, who owns Tacticalories Seasoning Company, makes the greatest turkey brine ever. I used his Gobbler Hollow Bird Brine on a turkey last Thanksgiving. It changed my family's mind about just how good turkey can be. Support delicious food and buy the brine here.

2. Spotify for Deadheads

I’ve come to peace with the fact that I love hippy music. The Grateful Dead is my favorite band (fun fact: Joseph Campbell researched the Dead because of the mythic themes of their lyrics and their following … so you can understand why I’m into them). I recently subscribed to Nugs.net. It’s like Spotify for Deadheads. It’s a massive repository for live concert video and audio. It has a ton of live stuff from jam bands like the Dead, but also bands like Pearl Jam, Metallica, etc. (Note: The seven Deadheads who subscribe to the 2% Newsletter are very appreciative of this tip, I’m sure, while the rest of you are all rolling your eyes.)

Thanks for reading.

-Michael