The world of Gary Taubes

Science writer Gary Taubes has a recent New York Times commentary criticizing Kevin Hall’s recent paper on the differential metabolic effects of low fat vs low carbohydrate diets. See here for my recent post on the experiment. Taubes is probably best known for his views on nutrition and as an advocate for low carb diets although he has two earlier books on the sociology of physics. The main premise running through his four books is that science is susceptible to capture by the vanity, ambition, arrogance, and plain stupidity of scientists. He is pro-science but anti-scientist.

His first book on nutrition – Good Calories, Bad Calories, was about how the medical establishment and in particular nutritionists have provided wrong and potentially dangerous advice on diets for decades. He takes direct aim at Ancel Keys as one of the main culprits for pushing the reduction of dietary fat to prevent heart disease. The book is a great read and clearly demonstrates Taubes’s sharp mind and gifts as a story teller. In the course of researching the book, Taubes also discovered the biological mechanisms of insulin and this is what has mostly shaped his thinking about carbohydrates and obesity. He spells it out in more detail in his subsequent book – Why We Get Fat. I think that these two books are a perfect demonstration of why having a little knowledge and a high IQ can be a dangerous thing.

Most people know of insulin as the hormone that goes awry in diabetes. When we fast, our insulin levels are low and our body, except for our brain, burns fat. If we then ingest carbohydrates, our insulin levels rise, which induces our body to utilize glucose (the main source of fuel in carbs) in favour of insulin. Exercise will also cause a switch in fuel choice from fat to glucose. What is less well known is that insulin also suppresses the release of fat from fat cells (adipocytes), which is something I have modeled (see here). This seems to have been a revelation to Taubes – Clearly, if you eat lots of carbs, you will have lots of insulin, which will sequester fat in fat cells. Ergo, eating carbs makes you fat! Nutritionists were so focused on their poorly designed studies that they missed the blatantly obvious. This is just another example of how arrogant scientists get things wrong.

Taubes then proposed a simple experiment – take two groups of people and put one group on a high carb diet and the other on a low carb diet with the same caloric content, and see who loses weight. Well, Kevin Hall anticipated this request with basically the same experiment although for a different purpose. What Kevin noticed in his model was that if you cut carbs and keep everything else the same, insulin goes down and the body responds by burning much more fat. However, if you cut fat, there is nothing in the model that told the body that the fat was missing. Insulin didn’t change and thus the body just burned the same amount of carbs as before. He found this puzzling. Surely there must be a fat detector that we don’t know about so he went about to test it. I remember he and his fellows labouring diligently for what seemed like years writing the protocol and getting the necessary approval and resources to do the experiment. The result was exactly as the model predicted. We really don’t have a fat sensor. However, the subjects lost more fat on the low fat diet then they did on the low carb diet.  This is not exactly the experiment Taubes wanted to do, which was to change the macronutrient composition but keep the calories the same. He then hypothesized that those on the low carb diet would lose weight and those on the low fat, high carb diet would gain weight. Kevin and a consortium of top obesity researchers has since done that experiment and the results will come out shortly.

Now is this surprising? Well not really, for while Taubes is absolutely correct in that insulin suppresses fat utilization the net outcome of insulin reduction is a quantitative and not a qualitative question. You cannot deduce the outcome with formal logic. The reason is that insulin cannot be elevated all the time. Even a continuous grazer must sleep at some point where upon insulin falls. You then must consider the net effect of high and low insulin over a day or longer to assess the outcome. This can only be determined empirically and this is what Taubes fails to see or accept. He also commits a logical fallacy –  Just because a scientist is stupid doesn’t mean he is wrong.

Taubes’s recent commentary criticizes Kevin’s experiment by saying that it 1) is a diet that is impossible to follow and 2) it ignores appetite. The response to the first point is that the experiment was meant to test a metabolic hypothesis and was not meant to test the effect of a diet. My response to his second point is to stare agape. When Taubes visited NIH a few years ago after his Good Calories, Bad Calories book came out I offered the hypothesis that low carb diets could suppress appetite and this could be why they may be effective in reducing weight. However, he had no interest in this idea and Kevin has told me that he has repeatedly shown no interest in it. (I don’t need to give details on how people have been interested in appetite for decades since it is well done in this post.) I came to the conclusion that appetite control was the primary driver of the obesity epidemic shortly after arriving at NIH. In fact my first BSC presentation was on this topic. The recommendation by the committee was that I should do something else and that NIH was a bad fit for me. However, I am still here and I still believe appetite control is the key.

15 thoughts on “The world of Gary Taubes

  1. The acylation of ghrelin by c8:0 via GOAT enzyme and other medium chain fatty acids is a de facto physiological sensor. Fat alters CCK in an appetite suppressing manner is a second example of FA sensing.

    I agree with you that appetite regulation is key to the low carb dieting. Increased fat intake and protein intake is clearly far more satiating than carbs.

    A second effect of glucosing sensing neurons also alters appetite.

    Ketones that increase in low carb diets restore insulin sensitivity. Therefore insulin in smaller quantities is much more effective.

    Talking about insulin effects for weight gain without thinking about insulin receptor states in endocrine/brain vs adipose vs liver vs muscle is truly forgetting the physiology and biochemistry of such systems! Indeed the monitoring of the single effect of insulin activity, blood sugar, is so myopic it’s astounding. Insulin does so much more than that including reducing appetite!

    For example at a steady state of food intake and an initial physiology of “normal” insulin receptor activity, any increase in insulin resistance will result in a positive feedback response. Say that just the insulin receptor in brain responsible for reducing hunger becomes resistant. The others are working just fine (so far). With this scenario, in order to get the same steady state of appetite, you would require more insulin to be secreted to get the desired effect. But how do you get the increased insulin?? You’d have to eat more carbs! Now you have an aberrant system no longer in balance. The increased carbs results in increased fat formation relative to the originap rats of fat formation vs fat utilization. Fat accumulates slowly at first. But well known is the effects of obesity to increased insulin resistance! Now other tissues become involved with increased resistance. More insulin is required to cause the same effects as before. This is the essence of the positive feedback result.

    Therefore Taubes is also correct in the final result.

    Now to be sure, the rate of insulin resistance is highly variable from one individual to the next. It’s quite likely that there are individuals that just never have issues with insulin resistance. Based upon the current trends of obesity, I’d guess that group is less than 20% but that’s purely speculative for the record.

    Because of this effect by insulin receptors (and similarly for leptin, PYY, PP, CCK,….x >100 peptide receptors), a low carb diet is likely to help many overweight people overcome this cycle.

    Liked by 1 person

  2. El experimento de Hall Et al. NO demostro disminucion de grasa por DXA. Lo hizo por un modelo matematico (fat balance). No teniendo presente cuanto grasa se usa en vivir..cuanta nobes absorbida..etc. fue un pobre estudio que como no les cerro usaron un modeo matematico que no se adecua a la realidad de un ser humano


  3. I wonder if the model also predicts that you dont have to be a scientist to be stupid. ‘even if i’m wrong i’m right’ (fixing a hole’—great beatles song about metalogics). (I wonder how the beatles fit in with Hsu’s view that classical music attracts high iq people—i consider the beatles to be brilliant, as well as some jazz and other musics (punk, hip hop…) . I guess one would have to give some tests, and then place all species in a hierarchy using the ‘well ordering principle’ (axiom of choice) assuming its unique and not like a mobious strip. Maybe this would be about ‘multiple intelligence’ ideas —pop musicians and writers like mark twain are smart in a different way, as are spiders.

    This reads like a murder mystery and i dont have energy or concentration to follow it—so i’ll just blab instead. (I mostly read 1 paragraph abstracts of articles—some of them i dislike, mostly in social sciences and some in what are called crackpot science because they say sometjhing like ‘it is well kniown and proven here that everything other people know, think or beluieve is wrong so in this paper i explain the correct theory, the one truth—which usually turns out to be either some slightly concealed plagiarism of others’ theoires or some sort of semi coherent rant like one would hear in a psychiatyric hospital)..

    it sounds like there are alot of feedback loops here—even a bit like a thermostat, or some set of coupled chemical reactions with different kinds of reservoirs..

    my experience with appetite is that sweets and potato chips are addictive so you can put on 5 or 10 pounds easily. (this i think happens at holidays, which i rarely celibrate as a sort of policy—and usually only do to be civil—my view is often there is little to celebrate and what there is is not really connected to human religious or national rituals. my last holiday celebration—new years with family—i took a turn on the guitar playing blue skies’ by willie nelson—-thery told me it was ‘unnaceptable’ and ‘got to bed’ andf that spoiled the party. i used to do this intensionally as a ‘juvenile deliquant’ at some of the wealthy hotels in dc—we were having our own party— but it was also sending a message. Even physically fit sports fans get to have a riot after the game—which is why taxes should be used to fund stadiums and billionaires rather than wasted on welfare, national parks, research and art etc.

    also, there is nonlinearity—the more weight you put on (which does sequester in fat) the less you will excercize and hence may keep or put on weight (this can also have a ‘psychological’ component since you may become a ‘recluse’ if your appearence seems or is bad—people may not want you around—and this may also lead to a ‘bad attitutude’ which compounds the problem).

    fat sequestrion seems to have a kind of intertial effect —its less ‘liquid’ or fluid and may drag you down since its less quickly available than say a ‘sugar buzz’ (which i have found at times to be so uncomfortable i have run it off–its almost like being on fire, which i have also been). reminds me of keynsian economics—friction—or TARP (bank bailouts, not jail)..

    i also think if you dont eat any or very minimal protein eventually you will lose weight —at least for me but that was also accompanied with a whole lot of hiking (i ran ouit of food a few times in mexico except for fruit and ladakh india (himalayas). Protein seems to be a cofactor neccesary for fat sequestration (sort of like wearing a coat in winter–its insulation so you dont lose the calories immediately).
    Protein seems to sequester in part in its own reservoirs—bones, etc. Perhaps even brain cells and nerves.

    i googled taubes and had seen the ‘cold fusion’ book (i use cold fusion in my apt since its cheaper than pepco); also he took out after carl rubbia in physics (barely heard of him). (Some of my favorite science types are the ones like brian josephson or the guy who developed some DNA technique who then later become aids denialists, proponents of telepathy, big bang deniers, nazis, religious fundamentalists (eg Price of group selection) etc. It sometimes seems like ideas/theories/techniques like math/ ideologies etc are basically platonic memes (as chomsky has suggested—words and ideas just exist ‘out there’ and they sort of just get randomly combined in people brains (which is the view of some econophysicists for the way income is aquired—its just maximum entropy. That may be a ‘principle of impotence’ or symmetry, or of maximum ignorance, which just seems to work heuristically as well as any other reason (eg god) and may not explain anything)..


  4. ps every calorie is the same every other one, and e=mc2 but the periodic table, like national currencies have sort of different kinetic constants, so you can use arbitrage to at least temporarily have a sort of perpetucal motion machine (i think Prigogine discussed this as the ‘spin echo experiment—temporarily reverses the arrow of time, though poincare and tiipler (physics of immortaily) might say it can be done permanently or that really is no arrow of time or time either—-carlos rovelli and coonnnes i think make this argument though its been around) I’m personally am agnostic so can argue on both sides—eg its due to merit, or chance, and one can never know (a la quantum mechanics—there are detection loopholes, slutsky (spurious) cycles, or maybe ‘statistical properties of chaotic systems’ (BAMS 1991).


  5. I don’t remember if I got this idea from you or somewhere else but:

    Models like yours show that a not-that-large calorie imbalance leads to fairly large changes in weight which, for the average, healthy, young person with a good diet, are not observed (in my experience). For such people, and for what seems like almost all animals, calorie balance just sort of works out, without calorie counting. Appetite just happens to work out over the timescale of weeks and months. So for those people who are not in calorie balance (or who weren’t while they were gaining weight), hyper-appetite seems like a reasonable place to look. And a poorly composed diet may be a source of hyper-appetite.

    Unrelated: doesn’t any theory of obesity that relies on how many calories are being burned throughout the day (and how this might be driven by diet) make predictions about the heat emitted by the person on that diet? Unless there are somehow changes in the efficiency of cellular metabolism itself? So can we measure ambient heat changes in a well-insulated room?


  6. @rick I just realized that I forgot to post about this paper Chow, C. C. & Hall, K. D. Short and long-term energy intake patterns and their implications for human body weight regulation. Physiology & Behavior 1–6 (2014). doi:10.1016/j.physbeh.2014.02.044 that deals with your question. Intake can fluctuate wildly and the body weight will be stable because of the long time constant.

    Regarding heat, yes that is what a direct calorimeter would measure. This is possible but much more expensive than an indirect calorimeter, which measures oxygen consumption.


  7. If you look at NHANES, the population has a lumpy distribution of micronutrient deficiencies. Or look at the Westphal 1995 study of women becoming malnourished on the military diet. Recently the AF started screening for iron anemia (see also IOM study). Roughly 20% of women and 6% of men arrived iron anemic. When they corrected the anemia and issued a multi to the women, their attrition and injury rates went down dramatically.

    The A-to-Z Diet study basically showed that most people were unable to eat a prescribed diet (whether they read the diet book or not) because they can’t translate the book into actual meals. The YouTube presentation has more insights than the research paper. Although the Atkins group did best re the outcomes, all groups tended towards the mean over time.

    Realizing there have been massive socio-economic changes in the last few decades. Dual-income or single parent families (usually working) are the norm. Most “meals” are eaten on-the-go (IOM Weight of the Nation).

    Frankly arguing over macronutrient ratios without further definition is pretty much like arguing about what sounds dinosaurs might have made.

    You should publish your comp blood chemistry. It might be interesting. BTW iodine levels have fallen 50% (CDC) so we may start seeing goiters again.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s