I recently featured Dr. Walter Willet's take on the obesity epidemic in an extensive interview for PBS's Diet Wars special. Here is what Gary Taubes has to say. It is a bit of a long read, but well worth it:

What made you go after this topic in the first place?

Two things. I'd been reporting on salt and blood pressure, which is a huge controversy, and some of the people involved in that were involved in the advice to tell Americans to eat low-fat diets, and they were terrible scientists. These were some of the worst scientists I'd ever come across in my 20-odd year career of writing about controversial science.

I literally called up my editor and said, "I just got off the phone with so-and-so, and he's [taken] credit for getting Americans to eat less eggs and less fat. This guy's one of the worst scientists I've ever talked to, and if he was involved in this, then there's a story there." And that was it. I didn't know what the story was. I just knew there was a story.

[Was there a personal motivation?]

Before I did it, I was up at MIT, interviewing an economist about another story, a guy who runs a laboratory of financial engineering. He told me about being on the Atkins diet, and how effective it was. He was an Asian-American who had lost 40-50 pounds by giving up white rice, in effect.

I thought I would try it as an experiment, since I was going to write about fat and whether it really did cause heart disease and weight loss. I tried it, and it was amazing. You know, it's everything -- the 20 pounds that I'd never been able to lose, in six weeks, and I stopped exercising. It was kind of a surreal experience, and probably, in a sense, informed my opinions from there on in. I mean, after that happens, you say, "I want to know what's happening, and I want to know why."

Why is it so easy for us to believe that fat is a bad dietary ingredient?

The idea is that fat has nine calories per gram, and carbohydrates and protein have four calories per gram, and somehow the theory is that the denser the calories, the easier it is for us to eat more of them. What happened is in the '50s and '60s, when researchers started fingering fat as a cause of heart disease, the obesity researchers, the obesity community started advocating low-fat diets, which they had never done before. A low-fat diet is by definition a high-carbohydrate diet.

But you had this sort of synchronicity where you had the heart disease people saying, "Give up fat, saturated fat, for heart disease," and the obesity people started saying, "Give up fat because it must be the best diet because fat is the densest calories." They moved from there without ever testing actually either of those hypotheses, so the obesity people start recommending low-fat diets; the heart disease people are recommending low-fat diets. They have actually no idea whether it's going to cure heart disease, and the obesity people have no idea whether these diets even work. But because they believe that it's only the calories that [are] important, obviously if you give up the major source of calories in the diet, you must lose weight.

You get this hypothesis that animal fats are the worst kind of fats. That seems reasonable.

That came out of studies where you compare the fat consumption in various countries versus the heart disease rates. Basically that's what we still believe, that the Japanese have a very low fat consumption. Greeks have very low animal fat consumption. They have low heart disease rates. The U.S., Sweden, Finland have high fat consumption, they have high heart disease rates, and that's the genesis of that whole belief. "It's a worthless exercise," is what one researcher in the '50s called it. You cannot say that because fat consumption associates with heart disease, that that means it causes heart disease, because a lot of other things, for instance, associate with fat consumption. Wealthy nations have a lot of fat. They eat a lot of fat; they eat a lot of sugar; they get less exercise; they smoke more cigarettes; they drive more cars; they have more televisions.

There's a world of difference between the countries that eat low-fat diets and the countries that eat high-fat diets. And to finger fat because that's what you have in your mind to go in [to the study], is just bad science. But that's what they did, and that's how animal fat came out of it. We knew that animal fat, saturated fat, raised cholesterol, LDL cholesterol, the bad cholesterol, and it was just this sort of series of suppositions.

And we knew that cholesterol was associated with heart disease. The higher the cholesterol, the higher the heart-disease risk. Although when you actually look at the studies, [it's] kind of amazing. If you or I were to reduce our cholesterol levels by 30 milligrams per deciliter, we would probably increase our chances of living an extra 2 years by one-thousandth of a percent or something. For the actual individual who doesn't have extremely high cholesterol, lowering cholesterol makes very little difference in how long you're going to live.

There were several studies done in the late '80s, where they actually calculated how much longer you would live if you cut back on saturated fat. If everyone in the country cut back on saturated fat to that recommended by the government, and cut back their total fat consumption, you could then calculate from these studies how much longer you would live. And the answer was a days to a few months. And as the authors of this study pointed out, that was published in the Journal of the American Medical Association, those are at the end of your life. It's not like you get an extra month between 46 and 47. It's, you're in the nursing home and you die at, say, 77 and three weeks instead of 77 and one week. One commentary that was published along with one of these results said this is the equivalent of rearranging the [deck] chairs on the Titanic.

So you think right off the bat something went wrong. Clearly the differences between these countries has to be environmental factors?

Yes. The idea was, when you follow immigrants from one country to another, they tend to adopt the heart disease rates and cancer rates of the country they've moved to, so that suggests that it's not genetic, it's environmental. And then the question is: What is the environmental factor?

In the '50s, '60s, and '70s, there were a school of British researchers who said it's sugar, flour, white rice, what we now call "easily digestible carbohydrates" or "high glycemic-index carbohydrates." The diet doctors pushing low-carbohydrate diets, like Atkins and Taller and people like this, were sort of disciples of these British researchers. They read some of their writing, ad the idea was, primitive peoples, when they adopt Western diets, [they] adopt Western diseases as well: diabetes, heart disease, obesity, the foremost ones; some cancers -- colon cancer and breast cancer.

These British researchers pushed this theory and it kind of got run over by the dietary fad, cholesterol, heart disease dogma. [Other] researchers said: Well, if sugar and refined carbohydrates don't raise cholesterol, then they can't cause heart disease. Or if we can't prove beyond a shadow of a doubt that every country that has a high sugar consumption has high heart disease, then that means the theory's not true.

There were two different standards at work. In the heart disease dogma, every piece of positive evidence supported the hypotheses and moved it forward, and every piece of negative evidence, contradictory evidence, was ignored. In the refined carbohydrates theory, every piece of negative evidence was proof that the theory was wrong, and every piece of positive evidence was ignored. So you had two entirely different standards. One of them moves forward to become the theory we're living by today, this idea that if we cut back on fat, we'll be healthier. And the other sort of gets squelched.

You start to see the quality of this science, the foundation, is not what you thought?

Well, the first thing I did, I wrote an article for the journal Science, where I just looked at the question of fat and heart disease. And it was fairly clear from that, that if saturated fat has any effect on our heart disease risk, it's small. Having done that, and not being enraptured by this idea that saturated fat is evil, I then actually just pitched a story to The New York Times Magazine, saying, "I want to find out what started this obesity epidemic."

The obesity epidemic starts between 1976, say, and 1986. We're fairly confident about that because there're these series of National Health Examination surveys, and we know that in the third NHANES survey, obesity rates are still 14 percent.

Up until about 1980.

Yeah. Basically, up until about 1980, the obesity rates in this country are 12 to 14 percent. And then somewhere in that period between the late '70s and late '80s, they shoot up to 22-25 percent. That's known as the obesity epidemic, and the idea is: What explains it?

From my fat research, I already knew that there were two major changes in the country during that period. One was, high-fructose corn syrup came in as sort of the primary caloric sweetener in America, which was my personal bias. I thought that it was high-fructose corn syrup because I'm allergic to high fructose corn syrup. …

The other theory was that we started pushing the low-fat diets during this period. Starting in 1977, the government started telling all Americans to eat less fat, and starting in the mid-'80s, we started producing these low-fat products that in effect replaced the fat in the yogurt or the cookies or the whatever with carbohydrates. We went from being a country that ate about 40 percent of their calories in fat and 45 percent carbohydrates, to 34 percent fat and that much more carbohydrates. Conceivably, this belief that set in, that carbohydrates could be eaten to excess and wouldn't cause weight gain, that they were both heart healthy and the ideal diet, might have had some effect on weight.

I just went off to try and find out what the answer was. I didn't know when I went in, and the more research I did, the more it became clear that this argument, this hypothesis, the alternative hypotheses that carbohydrates cause weight gain, had validity. Didn't mean it was true. Just meant that it had validity. It could be true, and if it could be true, then the way you would check is to put people on low-carbohydrate diets. You go and look at the low-carbohydrate diet idea, and lo and behold, there's Robert Atkins, who's been pushing low-carbohydrate diets for 30 years, and people swear by them. I had my former experience where I knew that at least for me, I knew it was very easy to lose weight on this diet.

I found out while I was doing my reporting that there had been five studies recently done, clinically controlled trials comparing high-fat, high-protein diets like Atkins to low-fat, low-calorie diets of the kind the American Heart Association was recommending. And in each case, the people on the Atkins diet had twice the weight loss, and their cholesterol profiles, if anything, got better than the people on the American Heart Association diet. So I felt confident saying: Here's an alternative hypotheses that has validity. Here is one set of tests from the hypotheses, that seem to confirm it. Doesn't mean it's true, but it seems to confirm it, and what now needs to be done is more studies.

So your article comes out. … What was the reaction?

Some people said it was the best article they ever read, went on the diet. I got crucified in a variety of publications. A Washington Post reporter went after me, who had been advocating low-fat diets since the early 1980's. The Center for Science in the Public Interest went after me for having the nerve to suggest that low-fat diets might not be healthy. They'd been probably the primary force in the 1970s pushing low-fat diets, and through the 1980s. It was fascinating. They go after the messenger as much as the message. ...

[Talk about the difference between diet and other topics you have covered.] Is there more ideology?

Certainly diet became a religion. The whole low-fat idea, as much as anything, came out of the counterculture and Berkeley and San Francisco in the '60s, this idea that eating fatty meat, in effect, is the dietary equivalent of conspicuous consumption. There were famines going on around the world, people were starving, and here in America we were eating eggs and bacon for breakfast and huge steaks for dinner. This was just unacceptable politically, sociologically, ideologically. It merged with this idea that fat might cause heart disease, and then blossomed in the '70s.

Many of the health reporters who cover this, many of the researchers who report on it, are vegetarians or close to vegetarians. So it becomes much more than just a subject -- I mean, people are more polarized in this than they are in politics. I've had friends who have accused me of having a brain transplant, because suddenly I turned around and said maybe low-fat diets don't work, and may low-carbohydrate diets are the answer. It's as though the data becomes irrelevant. The evidence becomes irrelevant. Everyone knows what the answer is, and it's a little frustrating even from my point of view, because no matter how much research you do, you're going after a monolithic dogma, in effect. And dogmas protect themselves.

That's what I found out when I wrote my Times piece. Everybody involved says, "We're right. We've invested our whole lives in this. We believe it." Everyone believes it, and it's hard, once it's established, to even criticize it or be skeptical without being pegged as somebody who's self-interested or somebody who doesn't care that Americans are out there dying of heart disease.

Nutritionists do admit anecdotally that Atkins works. How do they explain it away?

The first Atkins-like diets were explained away because the people who advocated them suggested they had a metabolic advantage over low-fat diets. Say you have two diets that are 1,200 calories, one's low fat, one's low carbohydrate. These people said the low-carbohydrate people will lose more weight on the same amount of calories. Researchers came along and they did studies and they said, "No, if we give both groups 1,200 calories, they're going to lose the same amount of weight." So that allowed them to say the [low-carbohydrate] diets are no different than any other diets.

Then the diets kept coming up. They kept being popular. And the next question was: Well, the people on the low-carbohydrate diets don't eat as much calories, so even though you tell one group of people, "Eat less fat and east less calories," [on] a calorie-restricted diet, and [for] the Atkins-like diet, you say, "Eat as much fat as you want, eat no carbohydrates, eat as much food as you want. So you can go out and have 10 porterhouse steaks a day, but as long as you don't eat carbohydrates, you're going to lose weight."

First the critics said this is thermodynamically impossible. Who knows? I have no idea. If you eat 10 porterhouse steaks a day, maybe you'll lose weight, maybe you won't. That's another long story.

So [low-carb advocates] said, "Okay, if people are losing weight on this diet, we're telling them to eat as much as they want but they're eating less. So why are they eating less? Why is it so easy for them to eat less?" And the answer was: Well, they're either getting nauseous, because this state of ketosis which is caused by these diets can sometimes create nausea in the beginning, or you just can't replace the calories. If you give up the carbs in the diet, you can't eat enough fat and protein to make up for it -- which is kind of ironic because since the early 1980s, we've been told that what makes us fat is not the bread but the pat of butter we put on top of it, not the baked potato but the sour cream that [we] put on top of it. It just seems that they want it both ways. On one hand they say, if we just add one pat of butter and sour cream to a baked potato, on our white bread, we're going to get fat, and it's the butter and the sour cream that do [it]; but if we only eat the sour cream and we only eat the butter, we can't replace the bread and the baked potato with enough meat or fish or cheese or whatever, to replace those calories. ...

The other response to the Atkins diet has been to say it was unhealthy.

Yeah, there're two factors. The Atkins diet, because it's very low in carbohydrates, it can be ketogenic. That's why your body, in effect, stops running on glucose, on blood sugar, and starts running on fat, and the fat's broken down into these ketone bodies. Ketosis is a mild version of diabetic ketoacidosis, which is the state that occurs in uncontrolled diabetes, and it's fatal.

This whole medical society grew up basically thinking ketosis is bad, ketoacidosis is bad, and therefore ketosis in these diets is bad and you should do anything you can to avoid these ketogenic diets. In fact, many of the compromise diets from the 1930s through the 1990s were diets that cut back carbohydrates to just the level above which you won't have ketosis.

So on one level, ketosis is bad. That was always the message, and the other level was, these diets are high in fat, and if they're high in fat, they're going to cause heart disease. One study was done by this fellow John La Rosa, who went on to be a big administrator in the American Heart Association. He did a study in 1981 where he said he put these people on a Atkins diet and their cholesterol levels skyrocketed. Their weight plummeted, and if you actually look at the study, it's almost impossible [that] what he says happened. It's really fascinating that you take somebody from the American Heart Association, you have him do a study on the Atkins diet in 1981, and the cholesterol levels skyrocket, even though the people lose like 20-30 pounds. What he said happened should not happen, even if the diet was atherogenic, as they say.

Then 20 years later, you have unbiased people do the same experiment. Lo and behold, cholesterol doesn't go up at all. [However], once [the American Heart Association] showed that cholesterol went up, that was the end of the Atkins diet. Cholesterol caused heart disease. You go on this diet, it may reduce your weight, but you're going to die of a heart attack.

One of the things I had always tried to understand, the orthodox wisdom is [to] cut back. Just eat less, and yet people eat less and they don't lose weight. I've eaten less and I've not lost weight. I mean, it doesn't seem to work, and I kept saying to myself: Why can you tell people over and over again [to] eat less, and yet they don't lose any weight? One possibility is that every time you tell them to eat less, you tell them to eat less fat as well. That's what we did in the '80s and '90s. We even stayed away from avocados and peanut butter, which we now know have good fats in them. ...

And now the consensus is that the Atkins diet is not necessarily dangerous but it's a trick.

Yeah, that's the compromise position: "Okay, it doesn't raise cholesterol. We were wrong about that." Other people said, "well, there're probably other reasons saturated fat causes heart disease." It's fascinating. There was this train of logic that said saturated fat raises LDL cholesterol and that's a risk of heart disease. Then if they come up with a diet that says saturated fat doesn't raise LDL cholesterol, then they turn around and they say: "Well, saturated fat must cause heart disease in other ways." So saturated fat, by virtue of being saturated fat, is now unhealthy. There's nothing you can do to change that. Even if you have a diet high in saturated fat that doesn't raise cholesterol, it's still a risk factor for heart disease. The logic is circular and sophistic. …

The thing they never pay attention to -- and I've interviewed probably 300 or 400 researchers in the obesity/heart disease field in the past four to five years, and in obesity in particular -- they do not consider hunger a physiological phenomenon. They will talk about hunger hormones, hunger genes ... But hunger to them has always been something that's purely psychological on some level, so they could put people on a diet.

Now, here's the thing. If you just reduce the amount of calories, you'll lose weight, even though there's not a single study that's ever shown that. But this [is] what they know for sure, so we're going to put people on a diet, we're going to tell them to eat less calories, and if they don't lose weight, that's because they didn't eat less calories. If they didn't eat less calories, that's because they don't have any willpower, and willpower's some psychological concept that you don't have but I do. I'm thin because I have willpower. And that's what it comes down to. …

When you look at the history of these Atkins-like diets, like the protein-sparing modified fast -- which is a sort of fancy name for an Atkins-like diet with a little more protein and a little less fat -- when you look at the history of these studies, even the ones that said they didn't have a metabolic advantage over low-fat diet, they all said the patients, the subjects didn't seem to be hungry. Study after study from the '40s onward. We put them on the diet, if they eat less calories, they do it because they just don't seem to be hungry. That's a physiological phenomena, and you have to be able to explain it. ...

So high-carb diets, even with less energy density, make you hungrier and therefore make you fatter.

That would be the theory. Exactly. You're hungrier. Insulin actually used to be known as a hunger hormone up until about the '60s; the idea being, diabetics are hungry to begin with because they don't have the insulin to push the blood sugar and the fat into the cells, so their cells don't see it. There used to be a treatment for anorexics. You would just inject them with insulin and they get hungry and eat. The question was: Does this cause some kind of hypoglycemia, very low blood sugar? Is that how it works? Or does it work in some other way? It was always assumed that it works by just causing very low blood sugar, which in effect causes a state called hypoglycemia, and you get nauseous and dizzy and tired, and you have to eat.

But [now] everyone agrees that insulin is the hormone that controls the deposition of sugar and carbohydrates and fat in your body. They agree that if insulin levels are high, you'll preferentially store calories as fat; and that as long as insulin levels stay high, you won't be able to get to that fat to use it for fuel. They agree that carbohydrates will raise insulin levels more than -- fat doesn't have an effect on insulin, although if you force-feed enough calories, you can [raise] it. All of that is given.

What they don't agree is that somehow the carbohydrates, the actual macronutrient content of the diet, will do this. [Scientists] will say a calorie is a calorie is a calorie. They'll admit that a calorie of carbohydrates has an entirely different effect on your hormonal system than a calorie of fats. They'll admit that your hormones can control your weight; that insulin and estrogen have effects on weight, hunger, and body weight regulation. But they will never go from the step where they say: Hey, maybe the amount of carbohydrates and the kind of carbohydrates in the diet will have an effect -- through their effect on insulin, through insulin's effect on the deposition of calories, through that effect on hunger -- [on] being a functional diet.

To me, it's almost mystifying, because I've interviewed people who have done research on each step of the way. And then I say, "This should be the null hypotheses. You should assume carbohydrates cause weight gain until proven otherwise, not the other way around. Why don't you believe that?" And they'll say, "Well, because my rats get fat on fat." And it's true. Rats will get fat on dietary fat, fatter than they will on carbohydrates. So then the question becomes: Are rats a good animal model for human obesity?

So this theory would explain a big part of the rise of obesity on the high-carb diets?

This theory would explain a big part of the rise of obesity on the high-carbohydrate diet. Yes. Basically it says: The things that are making us fat -- sugars, high-fructose corn syrup, these kinds of easily digestible sweets, flour -- some people may get fat even if they eat vegetables. I really don't know because they've never done the research. Because they've been so convinced that somehow the type of calorie is irrelevant, I'm not sure this is still an open question. But the theory would explain why we started putting on weight just when we started thinking "if we would just reduce the amount of fat in the diet, that'll make the difference."

It's not just the calories. That's the point. It's the effect of the calories on the hormones, and the effect of the hormones on how your body decides to use the calories you're eating -- is it going to burn them as fuel or store them as fat -- and that effect on hunger.

But clearly the quantity of calories has to be going up. You're saying the mechanism for that is that the carbs make us hungrier?

The quantity of calories appears to be going up. Whether we would get fat anyway -- we've all had this experience where our weight fluctuates, apparently independent of how much we seem to be eating. If you chronically change your insulin levels, you will change your weight by doing it. So the question is: Can you do that by increasing calories, or do you do that simply by increasing carbohydrates and keeping the calories constant? Good research in the '80s by established researchers showed you can do that just by changing the amount of carbohydrates. ...

Does your story have big holes in it too? Every time you try and reduce things to single macronutrients, you run into--

No, actually, you look back to one of the things that prompted the whole dietary fat and heart disease theory was the reduction in heart disease rates in occupied Europe during World War II -- and diabetes rates and cavities, I might add -- they plummeted. People said they're eating less cholesterol. Then that was knocked down. Then they said they were eating less fat, and that was the theory.

Out of these studies that came out of occupied Europe, that was what generated one of the major legs of the theory that fat -- it's a tricky question because you can also argue that doctors in occupied cities in Europe and Scandinavia might not be making the effort to diagnose heart disease and diabetes as they were during the war, or before the war or after. The evidence that something dramatically changed the incidence of disease during the war years is pretty compelling. It could have just been less calories. We're now fairly confident that if you cut back on calories, you'll cut back on disease rates, and not just heart disease but cancer. But the question is: Then what makes you cut back on calories? And now you're back to, you know, Ancel Keys tried to get his conscientious objectors to cut back on calories. So now you're back to what causes hunger. And it takes back into this same hypotheses.

Most researchers now would say that calories are probably the most important factor, calories and weight, so you're still left with: What is it that makes you eat more calories?

What would be your prescription for curbing the obesity epidemic?

There's compelling evidence that low-carbohydrate diets work. Clinical trials suggest they work, [as does] anecdotal evidence for 150 years. People, when they wanted to lose weight, gave up sugar, gave up starch and we all grew up believing this, even into the '70s. Jane Brody, the New York Times health reporter, writes a bestselling book in 1985 where she tells everyone to go out and eat starches to their hearts' content, but first she says we all grew up believing that what causes obesity is bread, rice, pasta, potatoes, sugar.

It's fascinating. For 150 years since Jean Anthelme Brillat-Savarin in 1825 writes The Physiology of Taste and says, "I know what cause obesity. Just talk to fat people. They eat too much starches and sugars." He says this. "I have 500 conversations over the year with stout people, and each one, they're telling me, 'I love the potatoes. I love the rice. I love the bread.'" Since then, it's sort of been institutionalized that carbohydrates -- starches and sugars -- make you fat.

The '70s come around. We decide that fat gives you heart disease. We have to eat low-fat diets, low-fat diets heavy in starches and sugars, and we start getting fatter. ...

You're not talking about a diet with no carbs, just a reduced amount?

The underlying philosophy is this kind of Paleolithic diet theory. It's what we ate during the 2 million years that we were hunter-gatherers on this planet. The fact that we were hunter-gatherers for 2 million years suggests it was an extraordinarily successful evolutionary adaptation. The question is: What did we eat during these 2 million [years] when we left the jungle, the trees, went down into savanna and started surviving on whatever we could hunt or gather? That's the philosophy. The answer is, probably considerable meat, very low glycemic index, hard-to-digest roots and starches, and fruits and berries that look nothing at all like the beautiful Fuji applies you can buy at your local market now. Some carbohydrates, but whatever it was, it wasn't refined. It wasn't sugar. It wasn't flour. It wasn't easy to digest. That's my going theory. If this theory's right, the diet we evolved to eat is probably the correct diet.

Just as [with the] low-fat fad, when manufacturers took fat out and put carbs in, and people ate a lot of low-fat ice cream because they thought it was not fattening, and they got fat, might not the same thing happen if you produce low-carb products?

It's conceivable. The idea that it's what we evolved to eat, we never evolved to eat refined protein either, which is basically what they're giving us. I have no idea what those foods will do in the long run. I mean, I've actually tried quite a few myself, and they don't seem to affect weight in any way. They do seem to keep insulin levels down, if that's what's actually happening. ...

Unlike Walter Willett, who says some fats are good but others are bad, you'd say a quarter house steak would be a perfectly decent meal?

Let's put it this way. When you look at the historical record and the literature, you find plenty of reasons to blame the refined carbohydrates on the chronic diseases that affect us, and it's hard to come to a point where you can finger saturated fat as actually being meaningful. There's this interesting theory, originally pushed by a fellow named Crawford, who was a researcher in Kenya in the '60s. It's this idea that the grazing animals eat a much wider variety of fats and carbohydrates. They have much great sort of tapestry of fats in their body, and they have much more unsaturated fats than saturated fats, and therefore that constitutes a healthy meal. That's a shift in the spectrum of fats we're eating, not the introduction of an entirely new type of food, like refined carbohydrates.

The question is: Can that shift in the spectrum from more unsaturated fats to more saturated fats really be that important, after you've removed this dramatic shift of no refined carbohydrates to refined carbohydrates? I don't know, and Walter doesn't know either. He's just telling you that he believes it's the case. ...

The other theory has to do with activity. We're an increasingly sedentary society.

One of the problems with that theory is that the highest obesity rates in this country are in the poorest members of the population. [Those scientists] have to explain, for instance, why it is that those individuals who are most likely to do manual labor are the ones who are most likely to be obese. Hispanics, for instance, come to America, they get high obesity levels here, and they're not coming here and taking white-collar jobs. You have to be able to explain that before you can say it's obesity.

You have to be able to explain why it is they've never been able to show that if you exercise more, you'll actually lose weight. When you read the research articles from the '70s and '80s and '90s, they're fascinating. They'll always start with a introduction that says exercise and physical activity is crucial to weight loss and weight maintenance. Then they'll move into the main part of the paper, where they'll go through study after study after study, where they acknowledge that these studies fail to show that you could actually lose significant weight by exercising or being more active. Then they'll get to the conclusion, and they'll give you a half a dozen techniques by which you can then make exercise part of an important weight loss or weight maintenance program. It's surreal reading these things. The book ended [with] the advice that you have to exercise, but the actual studies show that doesn't do any good.

They say the people who control their weight the best tend to be more wealthy, better educated. What are they doing right?

First of all, they say that the same people also experience an obesity epidemic. Their absolute levels of obesity are lower than the lower classes, but the amount that obesity increased was the same. Among poor, rich, it didn't matter. They all went up like 7 percent during this 10-year period of time.

Second thing is, carbohydrates are the cheapest calories. So the poorer you are, the greater the amount of carbohydrates in the diet. You go back and you read the obesity textbooks and the obesity monographs in the '60s and '70s, and people have to explain why there's so much obesity among poor people, and they invariably say, "Well, that's because they eat so much fattening starches." You had this concept once again of fattening starches. ...

What do you hope to achieve, writing your book? You're making a case that there's a lot of bad science, food ideology. Are the vested interests too powerful to change?

It's interesting. In my wildest fantasies, people are sort of liberated from this belief that they have to reduce the fat in their diet, and they try it. If they're overweight and they want to lose weight, they try going back to the old-fashioned way of giving up sugars and starches and seeing what happens. In writing the book, I want to explain what to me is a fascinating episode in science sort of gone awry, and in the same time write about some fascinating science that's simply been ignored, some beautiful, elegant studies that people have paid no attention to, and maybe show some people not just how body weight regulation works in the human body, why we gain weight, why we lose it, and what to do, but also explain how science works and how it doesn't work, and when it is successful and when it fails, and what's good science and what's bad science. That's what I've been writing about in my whole career.



---
Tried a Paleo or carb-restricted diet? Join the Registry today!

In 2004, PBS did a special entitled "Diet Wars," interviewing experts from all over the dietary advice spectrum, such as Dean Ornish and Gary Taubes. Below are selected exerts from one of the interviews. At first glance it may sound like Gary Taubes, but in fact it is Dr. Walter Willet, the chair of the department of nutrition in the School of Public Health at Harvard Medical School. 

Worth noting: early in the interview, Dr. Willet explains that  "farmers have known for thousands of years that you can make animals fat by feeding them grains, as long as you don't let them run around too much, and it turns out that applies to humans." Yet at the end of the interview, Dr. Willet states that at the base of the Harvard Food Pyramid (PDF) are exercise and whole grains. They have since updated their recommendations to the plate. It seems difficult to reconcile this advice. 

This low-fat dogma was incorporated into the USDA food guide pyramid in 1992?

Well, there's a little bit of complexity. It's interesting. In the '70s and early '80s, while heart disease rates were going bad, we were not talking about low-fat diets. We were talking about replacing saturated fat with a healthy fat, polyunsaturated fat. But somewhere in the mid-1980s, we lost that message. It's perhaps partly because some nutritionists felt it was too complicated to talk about different types of fat, and developed the notion we should just reduce all types of fat across the board. That was really the beginning of the low-fat, high-carbohydrate crusade.

How is the simplification -- eat less fat -- manifest in this food guide pyramid?

Well, the food guide pyramid that was developed in 1991 really is based on the idea that all fat is bad. Therefore [if] fat is bad, and you have to eat something, carbohydrate must be wonderful. So the base of the pyramid is really emphasizing large amounts of starch in the diet. We're told we can eat up to 11 servings a day, and if that wasn't enough starch, the pyramid puts potatoes along with the vegetables, so you can have up to 13 servings a day. That's a huge amount of starch.

Where's fat?

Fat's up at the top of the pyramid, and where it says explicitly "fats and oils, use sparingly." It doesn't make any distinction about the type of fat, and it tells us to eat basically as little as possible.

From a public health standpoint, how would you characterize this pyramid?

Well, this pyramid is really not compatible with good scientific evidence, and it was really out of date from the day it was printed in 1991, because we knew, and we've known for 30 or 40 years that the type of fat is very important. That was totally neglected.

[What were the] unintended consequences? The food industry started using vegetable oils, but baking was difficult so they made a technical modification.

In some ways, we do have to credit the food industry with being responsive to what nutritionists were saying. They did believe or accepted the evidence that vegetable fats, vegetable oils, would be better than animal fats, and that really led to the development and promotion of the margarine industry and Crisco, baking fats that were made from vegetable oils. But they were made by a process called partial hydrogenation, which converts a liquid oil, say like soybean oil or corn oil, to something like margarine or vegetable shortening. As it turns out that was a very disastrous mistake, because in the process of partial hydrogenation, a totally new type of fat is formed called trans fat. The evidence has now become very clear that trans fat is far worse than saturated fat.

So when people were told to switch from butter to margarine?

Unfortunately, as a physician back in the 1980s, I was telling people that they should replace butter with margarine because it was cholesterol free, and professional organizations like the American Heart Association were telling us as physicians that we should be promoting this. In reality, there was never any evidence that these margarines, that were high in trans fat, were any better than butter, and as it turned out, they were actually far worse than butter.

People read on the label "cholesterol free."

Right. This is a good example where just focusing too much on one particular piece of the diet, one particular substance or nutrient, can really mislead us. It is true that these vegetable shortenings and margarines were cholesterol-free, and that was pushed. ... Even though these products were cholesterol-free, the trans fats in them raised our blood cholesterol much more than actual cholesterol in the margarines would have done.

"All fat is bad" led to many low-fat products, some of which had beneficial value. Talk about one example.

This campaign to reduce fat in the diet has had some pretty disastrous consequences. ... One of the most unfortunate unintended consequences of the fat-free crusade was the idea that if it wasn't fat, it wouldn't make you fat. I even had colleagues who were telling the public that you can't get fat eating carbohydrates. Actually, farmers have known for thousands of years that you can make animals fat by feeding them grains, as long as you don't let them run around too much, and it turns out that applies to humans [my emphasis]. We can very easily get fat from eating too many carbohydrates, and the public was really directed to only focus on fat calories, when we really have to keep an eye on calories no matter where they're coming from.

With more fat-free products than ever, Americans got fatter.

Right. The reality is that during this campaign for fat-free and reduced-fat products, actual fat consumption did go down, but Americans got much fatter during this period of time. Now of course lots of things were going on at the same period in time, but I think it's highly likely this focus only on fat calories to the neglect of carbohydrate calories has contributed to this epidemic of obesity.

Is it fair to blame the food pyramid when people don't follow it anyway?

The food guide pyramid has actually had a substantial impact on the diets of Americans. If we look back compared to 20 years ago, the percentage of calories from fat in the American diet is quite a bit lower compared to earlier days. Second, there're some important indirect impacts of the food guide pyramid, in that many tens of billions of dollars of federal food policies have to be compliant with the food guide pyramid. So many programs -- for example, what's fed to young children, to pregnant mothers, to low-income families -- have to be consistent with low-fat diets. So the impact really has been, overall, substantial.

Behind the food pyramid were a collection of constituencies. Were people receptive when you raised these criticisms in the '90s?

There was not much receptivity in the 1990s, when we raised these criticisms of the food guide pyramid. It was almost an accepted religious belief that fat was bad and carbohydrates were good. Then there were lots of economic interests behind the food pyramid as well. Clearly the dairy industry is extremely well represented in the food pyramid. The beef industry is there, and it's very convenient that beef is combined along with fish and poultry and nuts and legumes. So each one of those industries can say: It's healthy to have three servings a day of our product.

What makes it so difficult to study the relation between diet and health?

The relation between diet and health is unusually complicated to study, and that's probably why we don't have all the final answers even yet. In a trial of a drug, for example, you can randomly assign people, one group to a placebo and another [group] to the new drug, and see what happens to them. But obviously you can't do that with nutrients or a whole diet very easily.

Second, we have to follow people for many years to learn all the consequences of diet, and perhaps even almost a lifetime, because some of the effects of diet may be operating during childhood and not be manifested until adult life. So there's no one simple kind of study that will give us all the truth about how diet influences our health, and we need to combine the results of many different studies to have the best and most complete picture.

One [kind] of study, for example, involves feeding small groups of people very controlled diets for short periods of time, and seeing what happens, for example, to their blood cholesterol levels. But blood cholesterol levels are only part of the picture, so we are also conducting very large studies where we enroll tens of thousands of people and follow them for many years, all the way along, finding out what they're eating, and then what happens to them in terms of their risk of heart attacks, cancers, and other conditions. And of course in those studies we need to control for many factors, like whether they smoke, how much exercise they have, their family history of various diseases. It's really only when we put those kinds of studies together with the more detailed metabolic, biochemical studies that we can have the best information about the effects of diet.

So you find out what they eat, rule out the bad things they do, and look for associations?

Right. What we do is find out what people eat through very detailed, structured questionnaires, observe what happens to them in terms of disease rates -- heart attacks and cancers, for example -- and then put these together, controlling for other aspects of lifestyle like diet and activity.

What did you notice that seemed to clash with the low-fat dogma?

When we began our studies back in the late 1970s, we expected that we would find a relationship between, say, fat intake and breast cancer, because that was almost an accepted relationship. But as the data started coming in over the years, we just did not find any higher risk of breast cancer among women who consume more fat in the diet. And the same was true for colon cancer and for heart attacks and risk of type 2 diabetes. In fact, the percentage of calories from fat in a diet has not been related to any important health outcome.

Amount of fat has no relationship to coronary heart disease?

The amount of fat had no relationship to risk of coronary heart disease, but the type of fat was extremely important.

So the advice we were getting was not just misleading but dangerous?

Right. The evidence that we accrued really suggested not only that the type of advice that people were getting was not useful, but it actually could be dangerous, because some people were eliminating the very healthy types of fat that actually reduce heart disease rates.

Certain fatty acids can dramatically reduce the incidence of cardiac arrhythmia?

One of the important findings, not just from our studies but several trials conducted by other people in Europe and also some careful animal studies, have very clearly indicated that some types of fatty acids in the diet, in particular the omega-3 fatty acids, can actually reduce the heart arrhythmias that really cause people to drop dead in the street. We call that sudden death. And that's very important because some people were eliminating those critically important fatty acids from their diet because they were told that all fat is bad. ...

Traditionally people think of carbohydrates as made up of simple sugars and complex starches. What's wrong about that?

Right. The thinking in nutrition about carbohydrates really had broken them down into two classes: sugars and so-called complex carbohydrates, which are mostly starches. ... The idea has been pushed that all forms of so-called complex carbohydrates are really the poster child of nutrients, and we should be eating them in large amounts. That's what the pyramid tells us to do. But in fact, these kinds of starches -- white bread, white rice, potatoes -- are starches that are very rapidly converted to glucose, really pure sugar, and almost instantly absorbed into the bloodstream. And these are the kinds of carbohydrates that we really should be minimizing in our diets.

A sugar rush from a potato?

Actually, careful studies have shown, demonstrated that you get a bigger rise in blood sugar after eating potatoes, a baked potato, say, than you do from eating pure table sugar.

Really!

Really.

That seems pretty extreme. Why is that?

... There are several problems with these rapid rises in blood glucose after you ingest large amounts of a rapidly digested form of carbohydrate. First of all, when the blood sugar goes skyrocketing up, the body wants to bring it back down. So our pancreas pumps out a big blast of insulin, and as a result, the blood sugar comes crashing down rapidly. In fact, in many people, after three and four hours, it overshoots and actually become a little hypoglycemic, and that rapid crashing down of blood glucose and insulin stimulates hunger. That would be no problem, except that it's often all too easy to go in the refrigerator or find a snack, and if we do that frequently throughout the day, that can add up to too many calories over weeks and months and years, and contribute to obesity.

Second, these high rises in blood glucose and insulin have a bad metabolic effect on the blood cholesterol fractions. Specifically the HDL, the good cholesterol, is driven down, and triglycerides, another type of fat in the blood that leads to heart attacks, goes up.

Third, after many years of demand for high amounts of insulin, the pancreas tends to give out. And at that point in time, we've got type 2 diabetes.

[What is the role of insulin?]

The role of insulin is to transport glucose from the blood into the cells, like into muscle or into fat cells.

What can go wrong with this system?

Well, the problem really comes about when we develop insulin resistance. And that means that the cells, like the muscle cells, become more resistant to the action of insulin transporting the glucose inside the cell. And we can become insulin-resistant in several different ways. There's some genetic component, and as it turns our, for example, many Asians tend to have more insulin resistance. Also if we have overweight or low levels of physical activity, we'll be more insulin resistant.

Many people argue the virtues of Asian diets, with a lot of white rice.

Right. The Asian diet as it's traditionally been used raises some very interesting issues in nutrition, in fact, some of the most important findings during the last several years. What we've come to realize is that if we have a higher degree of insulin resistance, then we much less well tolerate a high-carbohydrate diet.

Interestingly, in traditional Asian societies, people were very lean, very active, and therefore had low insulin resistance. They could eat large amounts of rice, even white rice, in the diet and have low heart attack rates and have low rates of type 2 diabetes. But if you take that same person, and they [now] may be living in Beijing and driving a car and watching a television, and they put on a few pounds, they're going to have much more insulin resistance. So if you take that same diet, high in carbohydrate and white rice, they will have a much worse metabolic response and much higher rates of type 2 diabetes.

If you're American and overweight, won't you be put on a low-fat diet?

If you're overweight and living in the United States, and you go to a hospital and see a dietician, almost for sure, you're going to be put on a low-fat, high-carbohydrate diet.

What's your view of that?

The problem is that that's really the wrong diet for an overweight person. Because the person is overweight, in general they're going to have quite a bit more insulin resistance and much less well able to tolerate low-fat, high-carbohydrate diet.

But fats have twice the amount of calories per gram as carbohydrates. Doesn't it make sense to push the low-calorie diet and therefore a diet low in fat?

There's been a very simplistic idea: Just because fats have more calories per ounce than carbohydrates, we should be eliminating fats or reducing fats to control our total caloric intake, in other words, to help control our weight. What's really important though is how satisfying a diet is, because we have very complex mechanisms that control our total intake of calories, and it's become pretty apparent that if we have a high-carbohydrate diet, particularly high refined carbohydrate, it makes it much more difficult to control our total caloric intake. That's probably because when we eat refined carbohydrates, we get these swings in blood glucose and insulin that lead to hunger between meals; whereas if we have a diet that's somewhat higher in fat, we tend to be more satisfied over the long run.

Is this what Dr. Atkins was saying 30 years ago?

Dr. Atkins was saying as much as 30 years ago, that if we reduce our carbohydrate intake to quite low levels, that will make it easier to control our caloric intake and thus promote weight loss. As it turns out, there is a strong element of truth in that. A number of studies in the last year have looked in a very careful way, comparing low-fat, high-carbohydrate diets with reduced-carbohydrate diets, and in general people have done better on the reduced-carbohydrate diets in terms of their weight.

Even though each ounce of fat has twice as many calories, you eat fewer calories because they're more satisfying?

Yeah. Well, first of all, this idea of how many calories per gram of fat versus how many calories per gram of carbohydrate is a little simplistic, because we almost never eat foods that are pure fat and pure carbohydrate. They come in foods as mixes, and often with a lot of fiber and a lot of water, and all of those things make a difference as well. But what is really important in the long run is how satisfying a food will be. And as it turns out, that many high-fat foods -- sometimes like meat, but even think of a handful of nuts -- are often very satisfying even though the physical amount is not very great.

One of the interesting observations in the recent studies that have compared high-fat diets with high-carbohydrate diets is that there are a lot of differences between people; that it's not that everybody loses a certain amount of weight on a reduced carbohydrate diet. Some people lose a lot of weight. Other people hardly lose any, so one of the things we have to come to understand better is the reasons why there's such a difference in response to these diets.

Talk about your food pyramid.

We've tried to put together an alternative food guide pyramid that would be based on the best available science. And of course it is a little bit challenging to boil down a tremendous amount of complex science to a simple graphic, and you really have to focus on the things that are well established and important.

Right at the bottom, we put weight control and regular physical activity, because overweight is the number one nutrition problem in the United States. Almost everyone is going to need to exercise on a regular basis if they want to control their weight over the long run. And this also is a message that all forms of calories are important.

On the next level, we put healthy forms of carbohydrate, meaning whole-grain carbohydrates and healthy forms of fat, meaning from vegetable oils, in the recognition that in most people's diet, most calories are going to be coming from some source of carbohydrate and fat. And what's really important is that those both be healthy sources.

Some of those fats have come right from the top of the other pyramid.

Right. What we've done is, we've brought some of those fats from the top of the pyramid down to the base, because healthy fats are an important part of a healthy diet.

The rest of the pyramid?

Our alternative pyramid, like the USDA pyramid, does emphasize plenty of fruits and vegetables, but we've taken potatoes out of the vegetable group. We've put legumes and nuts as a layer. If you want to be a vegetarian, those are good protein sources. But moderate amounts of poultry, fish, and nuts can also make a diet be a non-vegetarian diet and still very healthy. And up at the top we've put red meat and dairy products, dairy fat, because those are high in saturated fat. ... At the top of the pyramid, we've put foods like white bread, white rice, white pasta, and sweets as those that should be used sparingly. And that was really the base of the USDA pyramid.

Some nutritionists have criticized your pyramid as "floating on a lake of olive oil."

The formal studies that had compared a more moderate fat intake as we've suggested, with low-fat diets, have actually consistently shown that people did as well or better controlling their weight on a moderate-fat diet compared to a high-carbohydrate, low-fat diet.

Even good fats are more fattening than good carbs. So they think you're contributing to the obesity epidemic, or there's a risk of that. A tablespoon of olive oil is 14 grams of fat.

There are all kinds of beliefs about the amount of fat in a diet, tremendously strong opinions. What we really need is sound data, and the studies that have been done show that people actually end up controlling their weight at least as well, and usually better, on moderate-fat diets compared to low-fat, high-carbohydrate diets.

Is it okay to get more than 30 percent of your calories from fat?

The evidence is quite clear that it's perfectly fine to get more than 30 percent of your calories from fat, and probably, in fact, it's even better to be getting more than 30 percent of calories from fat, if it's the healthy form of fat. ...

When you look at the causes of obesity, what do you find?

The causes of obesity and the obesity epidemic in the United States are extremely complex. In fact, obesity is sort of a tip of the iceberg of tremendous social change that's been going on during the last few decades. First of all, our activity patterns have changed greatly. We have children and adults now watching on the average, about four hours of television per day, and in study after study, we've seen just the number of hours of television watching being the strongest predictor of obesity. When I was young and came home after school, we'd all go out and play, and our mothers would have to drag us in for dinner. That doesn't happen very often anymore. Often the mothers are not there. Kids are inside. The television is used as baby sitters. We've also made it dangerous and uninviting to walk to places, to walk to stores, to ride bikes in urban areas, and that's removed an important amount of physical activity from our lives as well.

So the physical activities pattern's changed, but the food environment has also changed. We have food available fast and very low-cost and very convenient, almost everywhere. The food industry has invested many tens of billions of dollars in making their products more attractive, more sweet, more salty, more sexy, more seductive in every way that they can, and we're vulnerable to that promotion, and we are eating more. So you put these two factors together -- reduction in activity, heavy promotion of food -- and you've got, not surprisingly, an epidemic of obesity.

Which we're exporting to the rest of the world.

Unfortunately, what we've created is now being exported to the rest of the world. And in almost every country, in Europe as well as Africa, Latin America and Asia, rates of obesity are climbing rapidly.

When you think about solutions, how can we proceed?

The causes of obesity are very complex, and therefore any response is going to have to be complicated and multi-faceted as well. There is no single solution to this, but we have to do many things to really solve this problem.

We've actually looked at this in a very detailed manner over the past two years in the New England states, and identified eight different areas that we have to really modify. For example, schools have to change what's served, what's promoted in schools. The healthcare providers -- physicians, nurses -- have to be providing much better and effective advice to their patients about controlling their weight. Work sites have to change. The whole food environment and what we promote to children has to be somehow modified, and we do have to protect children from these aggressive advertisements. We have to change the physical environment to make it more conducive, more attractive to bicycle to work, to destinations, not just drive places. Many other things need to be done if we're going to solve this problem, and it's time we really began.

The national obesity crisis, I mean, it's really severe. And to hear you talk about it, it sounds like it's hopeless to overcome it.

The crisis we're facing is severe, and it's getting worse. Interestingly, the full consequences won't be seen for another 30 or 40 years, because it takes that long a time for the risk of diabetes and the complication of diabetes to play out. But it's not hopeless, and I think that's the good news. There are places, if we look around the world, where people have been controlling their weight. For example, in Japan and in Sweden, women have really not been gaining weight as they've gone through midlife and gotten older, so we do need to learn more about how they're doing it.

Also in the U.S. there are many people who are controlling their weight successfully. Not everybody is going to be as successful, even if they do the same diet and do the same amount of physical activity, but many people are doing pretty well. And one piece of evidence we have is that upper-income groups actually only have about half the rate of obesity compared to lower-education groups. So it's not hopeless, but we are going to have to devote many more resources to doing this if, as a whole country, we're going to be successful. ...

The advice for so long was: Buy something that says "fat-free," or at least "low-fat" on it. All these [salad] dressings, fat free. But not necessarily a good idea?

In fact, a pretty bad idea, and that was really unfortunate that the crusade against fat really led the food industry to produce all these fat-free salad dressing products. The problem is that the fats that were in the salad dressings were the healthy fats, the ones you should be including in the diet. They were the unsaturated fats that reduce our blood cholesterol levels, and we've also seen that they reduce heart attack risk and type 2 diabetes, as well. So people were giving up these healthy fats, and what we've seen is actually an increase in risk of heart disease among people avoiding full-fat salad dressings.

Of course one of the important roles of a full-fat salad dressing is that it makes salads taste good, and we want people to be eating those vegetables. So it's not that we're telling people to gulp down bottles of salad dressing. We want them to use it on a salad and enjoy it. ...

[At] the base of your pyramid, after exercise: whole grains. Right?

Right. On the base of our pyramid, side-by-side, we put healthy fats and whole grains. And that's because we do see positive benefits in study after study, that there's lower risks of type 2 diabetes and heart disease, and probably better weight control as well, with some whole grains in the product. And that's because these carbohydrates are generally absorbed more slowly, they give less of a spike in blood glucose, and they come with a whole package of minerals, vitamins, and fiber that all have positive health benefits. ...

---
Tried a paleo or carbohydrate restricted diet? Join the registry today and contribute to a better understanding of this way of eating!