“Diet advice that ignores hunger” – Gary Taubes answers his critics in NYT


If you want to lose weight, should you avoid fat or carbs? Recently a six-day long and artificial study was heralded by some as the final word on the topic (“avoid fat”). Even though much longer, bigger and more lifelike trials usually give the opposite result (“avoid carbs”).

Science journalist Gary Taubes has long championed the low-carb approach and obviously he does not agree that this six-day trial is the final word. Here’s his new opinion piece in The New York Times:

NYT: Diet Advice That Ignores Hunger

I wrote about the study two weeks ago, and basically bring up the same two main things – this study is very short and it ignores hunger. The participants were locked up and had no choice but to starve. In that situation any starvation diet will do.

If you’re not locked up you’d want a diet that reduces hunger so much that you do not even need to count calories (like low carb).

The thing that Taubes does not bring up is the real trick of the study. The low-carb group burned through their limited glycogen stores during the short study (that’s why they lost more weight). Once the glycogen is gone – after a few days – there’s no choice but to start burning even more fat. Quite a convenient time to stop the study, right?

It’s amusing that to finally “prove” that a low-fat diet works best for weight loss you have to…

  1. Lock people up so they can be starved,
  2. stop the trial after just six days, and…
  3. not care that they actually lost less weight.

If that’s your proof then consider me not so impressed.


Did a Low-Fat Diet Result in More Fat Loss?


  1. greensleeves21
    What Gary ignores is that diet advice isn't about health or even cost-effective positive outcomes. It's about the deep hatred, bias & cultural prejudice against overweight people. THEY MUST BE PUNISHED BY STARVATION FOR THEIR SINS. The core cultural attitude still hasn't shifted from Gluttony, A Deadly Sin.
  2. Scooze
    Greensleeves21, Gary Taubes doesn't mention the "sins" argument in his article, but he writes about it extensively in his books.
  3. Don
    From what I have read of the study the authors were not trying to prove whether cutting carbs or fat resulted in more weight loss in the real world. They were trying to evaluate whether calories are the determining factor (fat reduction or carb reduction make essentially no difference) or insulin is the key factor (carb reduction wins). I am completely on-board with the proposition that a high fat, high protein diet reduces hunger and can enable weight loss with no attention to calories. That has worked for me for going on four years and many studies support the proposition. But the underlying mechanism is still of interest and I wonder whether GT has modified his views on the roles played by insulin, gherelin, etc. It could be that all of GT's practical advice was valid but the underlying biochemistry was wrong. Worth revisiting as the science develops. I hope NuSI gets at this.
    Reply: #7
  4. Henrik
    I have refuted the claim that one must eat food and drink water to survive!
    How did I do? I did not eat and did not drink, and it worked.

    How long I held up? Not long, but I just wanted to disprove the claim, and I did ......

    Reply: #5
  5. bill

    Now just extrapolate out to perfect adherence
    over 6 months and ya got something!

  6. Nate
    I read the comments picked by the NYT readers as good to read. These comments included a comment by Dean Ornish. He states that "What makes the NIH metabolic ward study you (Gary Taubes) cited so important is they knew exactly what people were consuming." Unfortunately, getting one important variable in a study properly controlled does not make a trail worthwhile or even valid. Nature does not work that way.

    One must provide proper control, as much as possible, of all variables. The NIH study did not provide enough time for the metabolism of the participants to change in response to the change in macronutrients. That one lack of good control of an important variable means that the study is weak and maybe even worthless. Second, a diet with 30% carbs is not a low carb diet for many if not most. Not to acknowledge that everyone is different and thus their metabolisms react differently to carbs is a lack of control of that variable.

    Also, Ornish's comment about satiety ignores that the volume of food eaten only helps reduce hunger for a couple of hours after eating. Thus, ignoring the more important and longer lasting impact of reduced calories on hunger. And what about the lack of certain nutrients on hunger?

    I'm sorry that nature is so complicated and thus makes doing metabolic studies so difficult and expensive. But Ornish and the NIH must stop ignoring the beautiful complexity of nature and embrace it.

    Sadly, Ornish's comment shows that a person's pet idea and/or paycheck can cloud one's mind and prevent them from doing good science.

  7. murray
    Don. I agree that there is something to be learned from the study about carbs and insulin. The study seems inconsistent in reporting whether insulin fell on the low-fat or stayed the same. The discussion says one thing and the tables say another. My hypothesis is that the very low fat with meals meant either one or both of two effects of very low fat. First, with inadequate fat in the gut, the uptake of nutrients may have been slowed, which reduced the insulin response that would otherwise occur with even a sustainable low-fat diet. I have read elsewhere that fat with a meal can speed absorption and the insulin response. Presumably the designers of the experiment did too and saw an opportunity to find a quirky effect of a reduced fat diet at extreme low-fat conditions. Note that the macronutrient proportions in the "control" diet were rigged to result in very low fat but not genuine low carb. (Ivor Cummins pointed out the arbitrariness in the design this way.) Second, to the extent there was a post-meal insulin response, there would have been very little fat in chylomicrons to be delivered and stored in response to the insulin signal. Thus again, at a quirky boundary condition of extremely low dietary fat, the fat-storing effect of post-meal insulin would be mitigated. The carb load would have to be modest, so liponeogenesis from the carbs would be delayed and serum fat kept low while insulin was still elevated.

    So it appears the experiment was rigged to create peculiar boundary conditions with unsustainable levels of dietary fat to mitigate the insulin effect so that equal weight could be lost despite eating carbs. (There was no measured difference in change in body fat using body scans. The assertion of a difference was an inference from calculation of other variables and not an empirical measurement.)

    So what does the experiment show? Peculiar boundary conditions with calorically inadequate carbohydrates and unsustainably low levels of dietary fat appear to mitigate the insulin effect so that equal weight could be lost despite eating carbs.

  8. Henrik
    I have refuted the claim that one must eat food and drink water to survive!
    How did I do? I easily get to eat and drink, and it worked.

    How long I held up? Not long, but I just wanted to disprove the claim, and I did ......

  9. Henrik
    One thing that is wrong with the study. They do not test an existing diet!
    Or was there a diet, and where are the follower
  10. Snorri Godhi
    This website is, of course, focused on practical advice, not on physiology, and based on my limited experience the practical advice is sound, and wrt the specific study under discussion, it's good to see pushback against the extravagant claims at the BBC.
    Having said that, i'd also like our host to discuss the much more reasonable claims about physiology made by Guyenet on his blog; specifically, that the 6-day study does _not_ prove that a low-fat diet is better than a low-carb diet, but it _does_ disprove the insulin hypothesis of obesity.
    NB: also based on my experience, i believe that a low-fat diet does not necessarily lead to obesity (so i tend to doubt the insulin hypothesis) but i still would not recommend it, for other reasons. I'll give more details if anybody is interested.
  11. Niki

    Yes but but Gluttony is one of the deadly sins. Which is why eating meat was approved in the Bible, as it was the only thing that took care of the appetite. It has been twisted as being because people have appetites for meat so that something has to be Biblically OK, but I believe that it has been misinterpreted in and through the translation. Gluttony can only be controlled when there is a limitation on the excesses of fruits, grains and vegetables. Why even Cain sacrificed ( meaning for eating) vegetables, while Able ate the lamb. It was never the potato, the rice, the grains or lettuce that higher power was compared to. Gluttony is, therefore, a result of the misinterpretation of historical foods that humans have been eating for thousands of years--aka our modern food pyramid. Interesting that it is compared to the pyramid.

  12. cydills
    No matter what kind of diet you are on if you are hungry eventually you are going to eat. Especially in todays society where food is available 24/7. Every one should know at this point that fat satiates hungry more then grains or carbs. I admire Gary Taubes so much I hope he will be on the Low Carb Cruise next year so I can meet him.

Leave a reply

Reply to comment #0 by

Older posts