Why I like Gary Taubes but don’t believe he’s the Messiah


As of September 2017, new Sweat Science columns are being published at www.outsideonline.com/sweatscience. Check out my bestselling new book on the science of endurance, ENDURE: Mind, Body, and the Curiously Elastic Limits of Human Performance, published in February 2018 with a foreword by Malcolm Gladwell.

- Alex Hutchinson (@sweatscience)


Weight loss is a pretty controversial topic, so I always expect to take some flack when I write about it, from people who are convinced they know “the truth.” (The fact that I hear from so many different people with so many different versions of “the truth” is a good reminder that it’s impossible to make everyone happy.)

I certainly don’t know the full truth about weight loss, but I did my best in Monday’s Globe article to summarize my understanding of the current literature. The most common complaint I’ve heard so far relates to my treatment of Gary Taubes’s ideas. Early in the piece, I referred to his latest book as an “anti-carb polemic.” Later in the piece, I wrote “Mr. Taubes’ core idea, that refined carbs cause a damaging spike of glucose that can affect insulin function (and thus fat storage) is backed by quite a bit of science.” To me, that seems like a fairly even-handed treatment — if anything, I’m coming down on his side. Nonetheless, I’ve heard from several people who say I’ve careless misinterpreted or misunderstood Taubes’s point.

For the record, I think Taubes’s critiques of existing nutritional orthodoxy are enormously important, and I made significant changes to my diet after reading Good Calories, Bad Calories. However, I also think he does exactly what he accuses his opponents of: drawing conclusions from epidemiological and mechanistic studies without confirmation from intervention trials. Or to put it another way, just because everyone else is wrong doesn’t mean Taubes is right and his orthodoxy can’t be questioned.

I had a very long and very interesting interview with Taubes a few years ago. In case anyone’s interested, here’s a transcript of one of the last questions I asked him. This question — and his response — is one of the reasons I don’t believe he has the “final solution” to understanding weight loss, diet and health.

AH: The last thing that I was trying to reconcile is the role of exercise. If you were to take a sample of 100 serious marathon runners, you’d have a fairly emaciated group who, traditionally, have been instructed to eat as many carbs as they can.

GT: Here’s the way I see it, and I’ll give you my bias. Growing up, I had an older brother who was always lean. He used to be famous, at least in our family, for saying he never got full, he just got bored of eating after a couple of hours. But he would run 10 miles a day, and he was a rower. I played football, and even when I was running a lot, the most I could do was 8 miles before my body broke down. I’ve always been thicker than my brother. In high school, he was 6’5”, maybe 180, and I was 6’2”, 210.

The question is, was my brother leaner because he ran more, or did he run more because he was leaner? I think the answer is the latter, because of a concept called fuel partitioning. You imagine two people eating 1,000 calories per day. One of them is predisposed to store those calories as fat, so he stores maybe 700 calories as fat tissue, and has 300 calories left for energy to run his body on. The other one—call him Lance Armstrong—his fat tissue that doesn’t want to store energy, so he only stores say 200, and has 800 calories left for fuel. Which one’s going to be more energetic? Now Lance has 800 calories in his blood stream impinging on his muscle cells, like somebody’s putting their foot down on the accelerator and he’s a car. So he’s gonna go.

The Lance Armstrong type eats a 1,000-calorie meal and then goes for a three-hour bike ride, because he’s now got 800 calories floating around that he’d like to get rid of. The Gary Taubes character, meanwhile, eats a 1,000 calorie meal, but there’s only 300 calories left for fuel and he’s storing 700, so he wants to conserve his fuel, and he goes and takes a nap.

The pre-World War II clinicians used to talk about the “impulse to physical activity.” So the question is, does someone who’s a runner have that impulse because their body doesn’t want to store calories as fat? The more carbs they eat, the more energy they get from it. But if your body does want to store calories as fat, the more carbs you eat, the more calories you store. And this is what fuel partitioning determines, and that’s driven by the body’s response to insulin.

What you can’t tell, by looking at somebody or at a group of people, is where the causality lies. Are runners lean because they run marathons, or do lean people run marathons because they’re constitutionally lean. The only way to find is to find out is to do these clinical trials where you, say, take 100 people and randomize 50 of them to exercising daily and 50 of them to being sedentary, and try to maintain some kind of stable diet, and see whether the people who exercise lose weight or not.

I just don’t buy that. I know plenty of people who’ve discovered running or other forms of exercise partway through life, after being sedentary and overweight. Did they undergo some genetic change that suddenly created the “impulse to exercise”? As Matt Fitzgerald said to me an interview on a completely different topic, “It’s like Freud saying, ‘It’s all sex.’ Sure, sex explains a lot — but not everything.” Until I encounter further evidence, that’s my attitude to carbs: they explain a lot, but not everything.

15 Replies to “Why I like Gary Taubes but don’t believe he’s the Messiah”

  1. I don’t think Taubes is the Messiah. I was just completely surprised by the tone that was evident in the article. The paragraphs that really made me sit up were these two:

    “For starters, it’s clear that losing weight isn’t just about eating certain foods. Last fall, Mark Haub, a nutrition professor at Kansas State University, subsisted for 10 weeks on 1,800 calories a day – about 800 fewer than usual – of Twinkies, Doritos, brownies and other convenience-store staples. In the process, he lost 27 pounds and improved health markers such as cholesterol levels.

    While Dr. Haub’s experience runs counter to the advice in anti-carb polemics like Gary Taubes’ new book, Why We Get Fat, his slim result was no surprise to obesity researchers.”

    I don’t know if Dr. Haub has written up his “experiment” in a paper or not. I couldn’t find any link to it on his faculty web page. In a CNN article, though I read: Before his Twinkie diet, he tried to eat a healthy diet that included whole grains, dietary fiber, berries and bananas, vegetables and occasional treats like pizza. (http://www.cnn.com/2010/HEALTH/11/08/twinkie.diet.professor/index.html)

    In Why We Get Fat Taubes points out that most calorie-restricted diets involve reductions in total carbohydrate, which reasonably explains any weight loss. Given Dr. Haub’s pre-diet BMI of 28.8 and his description of his pre-experiment diet I’m sure that his 1800 calorie junk food diet included fewer total grams of carbohydrates, even if he was eating mostly junk food. Taubes discusses this phenomenon and logical error in Chapter 15 of Why We Get Fat. (Did you actually read all of Why We Get Fat, or did you just assume it was slimmed down, layman’s version of Good Calories, Bad Calories?) Taubes also points out that a NEJM editorial stated that researchers now think of HDL as a “biomarker for dietary carbohydrate” (footnote, p. 188) so it wouldn’t be too surprising that Dr. Haub’s HDL went up.

    I don’t think it’s an exaggeration to say that you were quite dismissive of Taubes based on the juxtaposition of the Haub anecdote. But a bigger problem is even mentioning the Haub anecdote. Why not talk about the science? Countless studies have shown weight loss on all kinds of diets. Does the weight stay off? Is the twinkie diet something you’d advocate? I think a reader of the globe article could be forgiven for thinking that you seriously consider it as good as anything else.

    Furthermore, you quote this in support of your even-handedness: “Mr. Taubes’ core idea, that refined carbs cause a damaging spike of glucose that can affect insulin function (and thus fat storage) is backed by quite a bit of science.” Isn’t this a bit like saying “evolution is supported by quite a bit of science.” It implies there is still some doubt or controversy. Is there? I don’t think so. Please correct me if I’m wrong.

    You may think I’m a zealot, but I’m not, and I do approach all of these ideas with an open mind. I don’t buy all of Taubes suppositions either, including the one you mention in this post. But he’s pretty clear when he’s talking about his hunches (which can’t be supported by any studies because the right studies haven’t been done) and when he’s discussing the science. To be given such a stage and to write about Dr. Haub’s twinkie diet instead of so many other things you could have written about (one example of dozen possible: the “A TO Z” study by Christopher Gardner) just seems a shame. But it’s a bigger shame for those who will read your article and come away thinking that they can eat ice cream as long as they spend 30 minutes doing it instead of 5.

  2. There is another component in the modern dietary affecting fat metabolism that doesn’t get much press – omega-6. Taubes has made it clear he’s not interested in it. However, over the years I’ve seen evidence that the body processes omega-6 fats differently from the short and medium chain saturated fats.

    A recent experiment by Susan Allport (Google “Susan Allport omega-6”) indicates that omega-6 lenoleic acid promotes fat storage and lean tissue loss.

  3. I think Gary Taubes tends to overstate some of his conclusions because he wants to be understood by a lay audience. In relation to the example about Gary and his brother, you are no doubt aware that he makes the same argument about Lance Armstrong – Armstrong isn’t lean because he rides his bike; it’s the other way around.

    I don’t think Gary is saying that ALL people are more active because they are lean, or that when fat people lose weight they don’t become more active. Or even that no fat people are acitve. He’s just pointing out that our ability to perform at a high level in sports is in part driven by our genes. I have to agree with him.

    On a personal note, I love learning langauges. And guess what, I have a really easy time learning them. So, do I love languages, and a result, find them easy to learn? Or do I find them easy to learn, and therefore I can see the beauty of grammar and the ways the words can keep their meanings across related languages? I believes it’s the latter.

    Of course this does not mean that you can’t become really good at understanding several languages even if you find it hard. It’s just unlikely that you would go through the struggle.

    As far as Lance Amstrong goes, I looked up some of the physiological traits that he is said to have:

    -Armstrong’s oversized heart can beat over 200 times a minute and thus pump an extraordinarily large volume of blood oxygen to his legs.

    -His VO2 max, the maximum amount of oxygen his lungs can take in, an important measurement for an endurance athlete,is extremely high.

    -He produces one-third less lactic acid than do other top cyclists and delivers oxygen to his legs at a rate higher than all but maybe 100 of his fellow earthlings.

    -Stacy Ingraham, Ph.D. lecturer in Kinesiology, says Armstrong’s entire body is “built for the bike.”

    So, maybe if Armstrong had been born in a poor country and didn’t have the same opportunities, he would have become the fastest runner in his village, or something similar.

    I was disappointed to see your reference to Taubes as a polemicist. Especially if you think is critiques are very important and your own life was changed by his work. Most people don’t know anything about him, or the arguments that he makes. I feel that the article could have been a good opportunity to introduce your audience to an important science critic.

  4. @Jennifer D.
    “I was disappointed to see your reference to Taubes as a polemicist.”

    I’m guessing my use of the word “polemic” is the reason so many Taubes supporters are mad at me. My dictionary says a polemic is “A controversial argument, especially one refuting or attacking a specific opinion or doctrine,” which seems to me to be a pretty accurate description of Taubes’s books. In hindsight, though, I wish I’d used a different word, since it seems to have distracted many people from my actual point.

    “I feel that the article could have been a good opportunity to introduce your audience to an important science critic.”

    I think this is also the crux of the criticism — people think my article would have been a good opportunity to write a completely different article. This wasn’t an article about Gary Taubes, the cellular mechanisms of fat storage, or historical trends in macronutrient consumption.

    My editors asked me to evaluate the evidence behind currently popular diet approaches, and that’s what I’ve tried to do. I included data on one of the randomized controlled trials that compared the Atkins diet to other approaches like Zone, Ornish and Weight Watchers (there were some graphs in the paper version). They were all roughly the same. So why should I have turned this article into a lecture about Gary Taubes’s theories?

  5. @Brian
    “I think a reader of the globe article could be forgiven for thinking that you seriously consider [the Twinkie diet] as good as anything else.”

    Come on, Brian. Are you seriously saying with a straight face that you finished that article thinking I advocate the Twinkie diet? Really? Or are you just trying to find something to criticize?

    “Given Dr. Haub’s pre-diet BMI of 28.8 and his description of his pre-experiment diet I’m sure that his 1800 calorie junk food diet included fewer total grams of carbohydrates, even if he was eating mostly junk food.”

    Again, really?! The guy goes on a diet that consists essentially of nothing but highly processed carbs along with a 30% calorie reduction, and your critique is he was probably eating MORE carbs before? That sounds to me like someone who’s trying to bend the facts to fit a theory, not someone who’s approaching the facts with an open mind.

    To reiterate: for this article, I went and looked for peer-reviewed clinical data on weight loss using various dietary approaches. In the comparison studies I found, there were rarely any major differences between approaches. Taubes’s response to that is that compliance was very low among subjects on the Atkins diet, which is why they didn’t lose (and keep off) more pounds.

    Maybe so. But that’s the point. I wasn’t writing about the theory of fat metabolism, I was writing about what has been shown to work in the real world. Now, personally, I’m a big fan of running. I’m pretty confident that if everyone ran 130 km per week, obesity would be a thing of the past. But even if that’s true, that’s unlikely to be the solution we’re looking for, because many people are unable or unwilling to do that. No doubt that Taubes’s approach works for some people, in the same way that running works for me. Being convinced that something WILL work and having evidence that it DOES work are two different things — and I think that distinction is important in science journalism.

  6. Let’s clear up one thing quickly before this turns into an ugly flame war so typical of ‘net discussion forums. I admire your writing and your insight greatly. I’ve followed your blog daily for about eight months (I’m always disappointed when there is no new post). I’ve gone back into the archives to read the majority of posts older than eight months. I follow a number of science and exercise blogs and read the science and health sections of a number of periodicals daily including the New York Times, the Wall Street Journal, the Guardian, Scientific American, Science Daily, New Scientist, and Discover. New journal articles with intriguing findings are obviously reported on in numerous places, and so I get the chance to see how different writers respond to the same research. It’s in this context that I respect your take on things. I’ll sometimes get a copy of the original study that you or other writer’s mention. I have a brother with a PhD in molecular biology who works for a big pharmaceutical company and he can get me virtually any journal article I ask him for. He usually reads them and gives me his take on things, too, giving me yet another perspective on the same information.

    The day the Globe article came out I was asked by a number of colleagues at work what I thought of it. I don’t read the Globe and hadn’t seen it. My colleagues know I follow this sort of thing. I’ve told them all about Taubes and they dutifully make fun of my low-carb diet. So, naturally they were interested if I had any thoughts on the article. The article gave them the impression that scientists don’t really have a clue what makes people fat (I’m not making that up). I went online during a break and took a look at it. I honestly did not notice that you had written it. I read the whole thing and came away with a distinct impression that the author was trying to dismiss or marginalize the low-carb theory of weight loss and I felt the manner in which this was done was not particularly scientific or rigorous – that’s why I was surprised and disappointed when I went to check your blog and realized you’d written the article! So much of the article seemed to consist of sound-bites from researchers as opposed to comments on the actual journal articles. Was this dictated by the Globe editors?

    Let me offer a reasoned explanation for this feeling. As I mentioned before, you brought up Dr. Haub’s junk food diet experiment and then immediately followed up with your comment about Taubes’ book Why We Get Fat. Now, OK, I’m not seriously suggesting that you are advocating a junk food diet. Let’s say, though, that you were using this anecdote in support of the hypothesis that the nature of the calories is not important, just the amount. Fine. Use science to back up this idea. Why would you use an anecdote? This isn’t scientifically honest.

    I still contend you haven’t read all of Why We Get Fat (I think you’re probably too honest to dissemble – like a good politician would, you’ve dodged the question instead). I really don’t care if you have. I’m not on a witch-hunt. It is, after all, a briefer version of Good Calories, Bad Calories and you’ve read that and claim that it changed your own eating habits. The only problem is that Taubes specifically adds a couple of arguments to Why We Get Fat to counter critics. So, in fact, Dr. Haub’s experience doesn’t run counter to the advice in Why We Get Fat. The idea that Dr. Haub’s diet resulted in a lowering of his carbohydrate total is based on Taubes’ idea, and I don’t think it’s a stretch at all. It’s just arithmetic. Let’s take a look at the carbohydrate content (by percentage of calories) of some of the things that Dr. Haub claimed he ate on his diet:

    Little Debbie Nutty Bar 40%
    Little Debbie Swiss Cake Roll 55%
    Doritos 49%
    Oreos 60%
    Powdered donut 44%
    Green beans 71%

    Some of the articles called the diet the Little Debbie diet, but are those snacks really that high in carbs? Let’s say for argument sake that Dr. Haub’s experimental diet was 60% carbs by calorie (it was probably closer to 50%). Given that he claimed he was eating a maximum of 1800 calories per day, that means he was taking in about 1080 calories in carbs. According to him, his pre-experiment diet consisted of “whole grains, dietary fiber, berries and bananas, vegetables and occasional treats like pizza”. His BMI was 28.8. My guess would be his pre-experiment diet consisted of more than 2600 calories, but even at 2600 calories, a carb total of 1080 calories would represent 41.5%. It’s more likely he was eating 55% (an often quoted number for a typical American diet, and low given his stated preferences), which would mean 1430 calories of carbs. So, even with what I think are conservative estimates he reduced his carbohydrate intake by 24.5%. It’s quite likely he dropped his carbohydrate intake by more than 30%.

    Regardless, nobody should be too surprised that Dr. Haub lost weight on a restricted calorie diet. We’re not shocked when a hunger striker loses weight, or when we see pictures of POWs. It’s routine to hear about a friend or celebrity (Oprah) losing weight on a semi-starvation diet. The big question is what happens when they go off the semi-starvation diet, because they invariably can’t maintain it for an extended period of time unless they’re forced to.

    In reply to Jennifer you wrote “My editors asked me to evaluate the evidence behind currently popular diet approaches, and that’s what I’ve tried to do.” [So why the Haub anecdote? It’s not evidence any more than my own personal experience on a low-carb diet dieting is evidence.] “I included data on one of the randomized controlled trials that compared the Atkins diet to other approaches like Zone, Ornish and Weight Watchers (there were some graphs in the paper version).” [The online version of the article didn’t include any reference to the “A TO Z” study, unfortunately.] “They were all roughly the same.” But, they’re not all the same. In the “A TO Z” study even Gardner conceded that the Atkins diet did the best. And (without meaning to sound like an acolyte), as Taubes points out the big difference in most of the experimental comparisons between Atkins-style diets and other diets is that the calories are never restricted in Atkins-style diets. The studies are not really controlled studies.

    There’s also a bit of a straw man fallacy in your opposition to Taubes. You don’t buy his “Lance Armstrong” argument. I don’t buy it either. It seems reasonable that a person who is actively fattening will be either hungry or sedentary because most of his energy is being partitioned into his fat cells. But it doesn’t explain why a fat person with a stable weight is sedentary. And, it also doesn’t explain why there are lean people (yes, even those on low-carb diets) who don’t seem to have much of an urge to do anything active. When I’m dragging my butt out the door to go for a run in the rain I don’t feel like I have any particular urge aside from my desire to follow a training schedule. Sometimes I do feel like I have lots of energy. Nevertheless, Taubes doesn’t claim this is anything more than a hunch or untested hypothesis, at best. But you don’t seem to spend much time attacking any of the science that he uses to support his main ideas.

    Here’s an element of post hoc ergo propter hoc fallacy: “Now, personally, I’m a big fan of running. I’m pretty confident that if everyone ran 130 km per week, obesity would be a thing of the past.” It’s likely that most of those who are running 130 km per week (I’m assuming you’re in that group) are pretty lean. But, is it always the case that someone who exercises this much becomes lean? No. I had the opportunity to do body marking at Ironman Canada last year, and I can certainly tell you that there are obese competitors (obese at least in the sense of BMI>30, with visible rolls of fat). These competitors are most likely logging more hours than anyone running 130 km per week. Yet, they retain their fat, and often get fatter according to some I’ve talked with. This isn’t evidence one way or another, but neither is your observation. If your idea is correct, why isn’t there more evidence to support it? Indeed, why did the American Heart Association and the American College of Sports Medicine write the following statement in August 2007 in their joint guidelines: “It’s reasonable to assume that persons with relatively high daily energy expenditures would be less likely to gain weight over time, compared with those who have low energy expenditures. So far, data to support this hypothesis are not particularly compelling.” That doesn’t exactly sound like a ringing endorsement of the idea that all you need to do is exercise to lose weight.

    It’s odd how some debates move into an entirely different sphere – low-carb vs. low-fat, barefoot vs. shod, evolution vs. creation, Democrat vs. Republican, Yankees vs. everyone else! I tell anyone who’s interested about Taubes, but I read lots of authors. A number of them cover the same ground and often go further afield. I think Taubes does it the best, though. I read but don’t tell people about Barry Groves, Uffe Ravnskov, Loren Cordain, Joe Friel, Mark Sisson and any number of others, because they don’t seem as rigorous or even-handed. Taubes seems the least polemic of this group. And I don’t limit myself to low-carb proponents. I read Pollan, Ornish and others. I just don’t find them particularly compelling.

    I’m sure you’re going to do your best never to write about this topic again, as it’s probably already consumed far more of your time than you’d like. I look forward to reading your next post on an exercise related topic. Please write something on marathon training, or the optimal running speed for interval training, or anything not about diet, so I can go back to looking forward to your posts and also not spending inordinate amounts of time writing comments.

    And, my sincere thanks, for all the insightful posts in your blog, a blog free of annoying ads, I might add. Here’s to a healthy 2011!

  7. Brian — thanks, once again, for taking the time to share your thoughts, and for your kind words. You’re obviously very well-versed in this topic area, and have plenty of insight to offer. That’s one of the main purposes of this blog: to provide a venue for extended discussions — and disagreements! — about the ideas I blog and write about, most of which are far from settled scientifically.

    It’s clear that, while we may agree on many things, we disagree about a few key points in this area. Perhaps the disagreements are just semantics, I’m not quite clear. In my article, the core message I chose to focus on is that weight loss is fundamentally about calories, rather than “good” and “bad” foods. Taubes points out, quite rightly, that “calories in = calories out” is meaningless if you don’t understand what drives both sides of the equation. Where I differ is in the belief that it’s carbohydrates that are the sole drivers — I’m certainly convinced that carbs are one factor, but I see obesity (both on an individual and societal basis) as multifactorial. The remainder of my article explored (superficially, given the space) how diverse those factors are — sleep, psychology, eating habits, etc. as well as diet.

    You started the discussion with the assumption that my conclusions could only differ from yours because I’d either failed to grasp or hadn’t read Taubes. I’m asking you to leave that assumption behind, and consider the possibility that I’ve read the same things you have but reached a different conclusion. That being the case, let’s forget about exhaustive disquisitions on my “tone” and talk about the substance of the ideas!

    For example, what do you think would happen if Haub repeated his stunt from the same starting point, but added 3,000 calories of pure butter to his daily regimen? Would he achieve exactly the same level of weight loss?

    (Don’t worry, I’ll be blogging about some new topics very soon, so you’re free to ignore this if you’ve had enough of calories versus carbs!) Best, Alex

  8. We’re on the same page Alex. Excellent analysis.

    I think Taubes is good because he questions norms and makes people think.

    However, he overstates some of his claims which aren’t quite definative answers like he or others make them out to be.

    Just my opinion, but his running/exercise idea is laughable at best to me.

  9. I also read and liked Good Calories, Bad Calories but tend to agree with Michael Pollan’s critique of the book that demonizing a single nutrient makes Taubes somewhat guilty of what he criticizes others of doing to fat. Nonetheless, the book made me think, and my diet has indeed changed as well since reading it. I tend not to worry so much about fat intake anymore, and avoid processed carbs to the extent that I can.

    As for the exercise bit – I agree with you and Steve’s comment above that it’s ridiculous. I was overweight and sedentary when I became a runner, and since becoming a runner have lost about 20 pounds. I am a N of one, but my story is most definitely not unique. I started running because i was getting fat, and if I stop running, I can pretty much guarantee that I would put that weight back on.

  10. @Pete I’m interested to know more about you losing 20 pounds when you started running. Were there any other factors that changed when you started running?

    Taubes argues in “Why We Get Fat” that those who make a conscious decision to change their sedentary lifestyle will also change their eating habits. Exercise and a changed diet will go hand in hand. Honest question — could what you think you lost due to running actually be attributed to a change of what you ate (and a lowering of the overall carb intake)?

  11. I agree. Taubes’ conclusions on how to lose weight I think are correct but ecological fallacies and intra-population analyses are invalid. He should know the latter as he quotes GEOFFREY ROSE

Comments are closed.