Guyenet v. Taubes: Thoughts on Stephan's Demolition of the Carbohydrate Hypothesis of Obesity
I've been laying kinda low on Stephan Guyenet's disassembling of Taubes' "Carbohydrate Hypothesis" and just sort of soaking in some of the responses. However as the mood hits, I'll be sharing some thoughts on the reaction in the coming days.
This post, however, relates to his blog post. As I said in the comments on his blog: Well done! Most of the arguments he makes are not new to this blogger, but I think these things need to be stated by as many bloggers as many times as is needed -- and judging from the reactions the work in this regard may never be done!! -- until the following acronym becomes as common in LC circles as acronyms like ACE, OWL and CCL are to Atkins dieters. When Taubes' theories become known as Taubes' Wrong Insulin-Carbohydrate Hypothesis On Obesity - TWICHOO. It kinda rolls off the tongue!
I would first like to highlight what I consider the most important part of Stephan's post: Identifying YET another one of Taubes' GCBC references where he misrepresented what the reference actually stated. Indeed I was just compelled to post GCBC Reference Check ~ Part VI of ? ~ The Massa. I add that to the list of Taubes' own references where he either offers up a misleading interpretation or outright misstates what the reference actually said. I don't care if there were a thousand references, these six -- so far -- and the way in which they were used to support his hypothesis are more than sufficient to dismiss this book as credible. This is not to say he's wrong about everything in the book, but getting it wrong on key points in this manner speaks volumes to the "scholarly incompetence" Gary prefers be used to excuse his professional lapses. We can add this to the previous five:
- GCBC Reference Check ~ Part I of ? ~ Metabolic Adaptability & Energy Balance
- GCBC Reference Check ~ Part II of ? ~ Insulin Resistance (again)
- GCBC Reference Check ~ Part III of ? ~ Is glycerol phosphate rate-limiting?
- GCBC Reference Check ~ Part IV of ? ~ Kipnis
- GCBC Reference Check ~ Part V of ? ~ Insulin Resistance: Taubes v. McGarry
I personally find these transgressions to be far more damning of the man than that his hypothesis has more holes than Swiss cheese. If it weren't for this, I could abide those justifications that GCBC helped so many to lose weight or improve their health. But when one realizes that he was paid almost three-quarters of a million dollars in advance to write a book (that ended up taking five years to deliver) these sorts of "errors" are inexcusable. I'm beginning more and more to think that he didn't even bother to read many of his references which makes all of those mentions of the number of them all the more meaningless.
Back to Stephan's post.
Part I: I would have liked to see Stephan elaborate a bit more on how Taubes does not offer us much in the way of what initiates this defect, but overall solid.
Part II: I have a couple of quibbles with this section. First, Stephan addresses elevated insulin levels and energy expenditure in the obese:
At least two studies have shown that higher fasting insulin is associated with a higher resting energy expenditure, independent of body fatness, not a lower expenditure (14, 15). This is consistent with the idea that elevated fasting insulin opposes body fat accumulation, not that it contributes to it.
Perhaps Stephan got a little caught up in a point/counter-point mindset, so that if elevated fasting insulin doesn't cause fat accumulation per Taubes he's stating it must be consistent with the opposing it. Rather a third option is available: that elevated fasting insulin is not associated with fat accumulation, as I demonstrated here with the scatter plot of change in fasting insulin and weight loss from Kipnis. A little later Stephan states:
Obese people are obese despite having higher fasting insulin, not because of it.
Here I would restate this as I don't really agree that obese people are obese despite the elevated insulin. I believe that there is a clear association between obesity and fasting insulin (on average) and that the obesity is the cause of the increased basal insulin production and not the result. More on that when I address Andreas Eenfeldt's response in a future post.
It appears that Stephan has finally tired of ItsTheWoo2's commentary on his blog and deleted some (all?) of her comments. But I'm presuming she meant the above discussion by Stephan and somewhat misquoted it. She apparently attributed the following quote to Stephan, but I don't find this in my browser search of his post: "This is consistent with the idea that elevated fasting insulin protects against body fat accumulation, not that it contributes to it."
I have commented sparingly over there and in response to this I did add the following:
Here I have a minor quibble. High fasting insulin is the result of obesity and adipose IR not the cause of it. Elevated fasting insulin is not protective against body fat accumulation, it is the body's attempt to protect it from its already accumulated fat! This is what a huge body of post WWII research clearly demonstrates. It's somewhat of a fat cell/pancreatic feedback loop: Overstuffed adipocyte becomes IR -> releases too much NEFA -> stimulates basal insulin production so as to compensate for adipocyte IR and prevent excessive NEFA release.
Never one to miss an opportunity to portray me as being incapable of understanding plain English (let alone all that science stuff), and fresh off of slagging me over at Paleo Hacks, Kurt Harris donned his white knight outfit and posted up this:
CS (me): "Elevated fasting insulin is not protective against body fat accumulation, it is the body's attempt to protect it from its already accumulated fat!"
S.G.: "This is consistent with the idea that elevated fasting insulin protects against body fat accumulation, not that it contributes to it."
KGH: I actually took Stephan to mean "protects (the body) against the already accumulated fat" as in, " a defense against that fat and the NEFA it contains". And the fat storage itself can be conceptualized as a defense in the first place. So I think all three of us are actually in agreement. It was just awkward wording.
Not wanting to be accused of spamming SG's blog, I responded "fair enough". Now perhaps Stephan has changed the wording from "protects" in the intervening days, but I contend that the new wording is equally incorrect. In any case, however, it is clear that SG did not mean what KGH tries to make it seem he meant so that for a change of pace I was in agreement with the big boys with higher degree lettering after their names.
On the topic of "starving cells" I only wish Stephan had done a little name dropping. And no, I don't mean yours truly as it appears Taubesians far and wide are already a little miffed at him for posting his comment here and making them feel all dirty landing at my blog from the various obscured links to it around the web. No, I'm talking about Keith Frayn whom Gary is now quoting in his lectures in addition to his frequent references to having discussions with him, some seemingly rather recent. Frayn's body of material is extensive, but he's also of the habit of writing (or coauthoring) comprehensive and comprehensible review/summary articles. That by this late date Taubes has not availed himself of these is inexcusable when they so definitively refute this notion of rogue fat cells hoarding energy and starving the rest of the body. Why We Got Fat was released in Dec. 2010 and he has to keep the schtick up long enough until at least to the point where those book sales taper off. If we ever get an acknowledgement from Taubes of his errors, it won't be coming much sooner than that.
Onward ...
Stephan puts quite a few nails in the coffin of TWICHOO, but his next section on genetics (it should perhaps have been Part Ib instead of IIb?) I think is his strongest argument yet. In fact, while all that came before is damning, this perhaps cuts more to the core. If obesity is a result in a defect in fat tissue regulation, then genetic obesity should result in a defect in something in the fat tissue regulatory process. As Stephan points out, every genetic cause of obesity identified to date involves leptin signaling. I do believe it's a bit unfair for him to state "body fatness is normally regulated by the brain, not by fat tissue". After all, this leptin acting on the brain is produced by fat tissue, but clearly leptin signalling between fat and brain is key. Nora Gedgaudas, to name just one other hopelessly mired in flawed dogma speaker at AHS describes leptin as controlling the other hormones if I understand correctly.** In any case, for TWICHOO to be TRICHOO (R = right), the genetics would have to point to something involving insulin and triglyceride/fatty acid cycle regulation and this does not appear to be the case. Game, set, match Stephan Guyenet.
On to the remainder ...
I don't really have much to say about the rest other than that Stephan dismantles Taubes' new tack which is to stress the degree of refinement of the carbs as a factor. He did this in one interview (no link handy) by first claiming that the Japanese eat brown rice, but he's been ratcheting up the insulinogenic argument quite a bit lately. He told Fat Head that he wanted each and every one of us to forget about calories, thus there must be something particularly insulinogenic about nuts and cheeses. He has dug his trench deeper and deeper that "carbs make you fat" ... and yet he also says that it's quite possible that w/o sugar it may be impossible to get obese. Ummmm.... In any case, this section does a good job of undermining any validity to that argument. His graphic on the macronutrient distribution through the years is also compelling.
More to come with commentary on the commentary ....
** 1. Please correct me if I've misstated Nora's position on leptin
2. Sorry, but it's impossible for me to take someone seriously who can write that all body fat comes from glucose and defend that statement to me in an email. Indeed whenever I read her name it goes something like Nora Good-Gawd-help-us if folks are learning human metabolism from her. It even seems from this that she said something similar in her book. I'll not waste money or a trip to the library for that book, but if anyone has an ebook to donate to the cause, I'd be grateful.
Comments
Huh?
I'm sure there are some that do. Just as some Italians eat brown rice. And some French. And some Irish.
So, yeah, they do.
Here is my rationale for the statement about fasting insulin and energy expenditure. Insulin acts in the brain in a manner similar to leptin (i.e. fasting insulin correlates roughly with body fatness, and it causes negative energy balance by acting on hypothalamic circuits), just not as potently. So if you control for fat mass, which the studies did, it would not be surprising if higher fasting insulin was associated with higher energy expenditure, just as increasing leptin increases energy expenditure. That would be consistent with the hypothesis that insulin signaling constrains fat mass in a manner similar to leptin.
Then again, it also depends on the degree of insulin sensitivity in the hypothalamus. If it is reduced, then the actual amount of insulin signal transduction in hypothalamic neurons wouldn't necessarily be higher than in someone with lower fasting insulin. There's no way to tell those things apart from an observational study, but I just found the observation interesting. That's why I said it's "consistent with" that idea rather than it "demonstrates" that idea.
One thing i recently learned (not from Seth Roberts) is that human mothers' milk tastes different based on many factors, including stress and diet (I'm skeptical so this is provisional). Modern formula is completely uniform. This fits in so well with Roberts' ideas.
This casts even more doubt over Lustig's fevered fructose theorizing.
I can still see a lot of itsthewoo2 comments.
"This is consistent with the idea that elevated fasting insulin protects against body fat accumulation, not that it contributes to it."
Taubes in a nutshell: everywhere you have fire trucks, you have fire.
to get rid of fire your city should HAVE NO FIRE TRUCKS !!!
Next: Gary Taubes rids your city of crime ... (something to do with cops ...) and cancer (something to do with chemo drugs and radiation)
I'm going to pretend for a minute that I'm buying that without being able to verify it.
Is that strictly desireable? Does it mean you starve 1000 cals easier than someone who is running on carbs? What are the consequences of that for the metabolism or the body as a whole?
As an aside, after that post I returned to my beloved banana smoothie for breakfast. OK, first I spiked a football in the front yard and did an excerpt from the thriller dance but after that...smoothie.
Liked the addition of IIB to the argument.
It would not surprise me at all if insulin (in a normal physiological range) had no significant effect on energy expenditure. I didn't intend to imply that I believe it does. The studies that showed it causes weight loss when you infuse it into the brain I think showed that the mechanism was mostly via appetite suppression.
Hi Sanjeev,
I incorporate Seth Roberts' idea of "ditto foods" into my reward theory if that's what you're referring to. I don't recall if I expressed that in my posts, but I mentioned it in my AHS talk. Consistency of flavor is a reward factor, monotony lowers reward. The potato diet is a good example.
How can it be used as a model to propose functions of insulin or extrapolate what insulin does since a bolus dose of insulin suddenly appearing in the brain is not exactly representative of practical occurances?
We've missed out all the other hormonal signalling that occurs in the rest of the body in response to ingestion of insulinogenic food.
@bentley: This is a very good point. It's like saying I'd rather run my car on crap gas because I can fill my tank with it while the good gas to drive my car the same miles only fills it to 3/4. IF there's anything to these futile cycles and uncoupling and such these are all ways to deal with excess and metabolism in excess seems to produce a lot more nasties like ROS.
I was thinking along similar lines. Would I buy/design a car with the promo shot including someone saying "Look how much more gas it takes to do the same thing!"
That seems like an inefficient ideal. From a survival standpoint [presumably what people do when they haven't got three grocers every mile or so] getting adequately nourished on the least amount of calories would be the game changer and not the reverse. That seems like it would be the design function...use contrary to that is likely to come with some consequences.
The whole notion of "getting" to eat more carries some implications that [imo] support the "food reward" theory.
I've seen two or three comments of late, some in response to Mike Howard's recent ranking of sites and some at places like Eenfelt's blog, and they take the form of "I gained weight eating (1300-1800) calories a day, and I am losing weight eating (2000-3500) calories a day." It's tough to pin down if they are referring to short term dieting, but the numbers being thrown around indicate that there is a third option to the ones you are considering: their stories are false.
It may be unintentional, it may be due to exaggeration, or they may simply be lying, perhaps to themselves as much as to us. Even Dr Eades, who champions the metabolic advantage concept, states that it accounts for 200-300 calories. But these people are indicating that they are experiencing a swing of 500-1300 calories in some cases. That's a hell of a lot of water retention!
The problem is that if this sort of anecdotal story becomes common, it will poison the discussion and allow LC defenders to dig their heels in even further. And we wind up in exactly the situation that led most of them to become LC acolytes in the first place.
"One wrong idea is that floating stools are caused by an increase in the fat content of the stool. In fact, it is increased gas in the stool that makes it less dense and allows it to float."
I also remember immediately after eating that food I would instantly regret it because I knew it was making me gain weight.
@Kindke: Emotional eating or whatever you want to call it is clearly at play here. Whenever I read the "I had a bad day yesterday and ate a whole bag/box/bowl of ....." Binges were triggered more by my thoughts and less by any particular food.
Everything you just said points directly to using food as a stimulant. You could have replaced "ate" with "did a line of coke" without missing a beat.
I agree with that usage but it's not what I'm seeing in comments where emotional eating is being referenced.
If someone goes on a 1000 cal/day diet and sticks to it they simply have to lose weight. I've seen a 450+ pound woman claim she's gained weight or not lost eating like 500 cal/day for several days. Either she's a liar or she has serious water imbalance issues. And if that same person can eat 2000 cal/day of protein and fat and lose weight, a serious fat malabsorption problem.
I don't know who these folks think they're helping or what the purpose of such posts is. Especially since many making the claims remain significantly overweight despite their outlier metabolisms.
A stimulant for some, a relaxer for others. For me, the latter. I was put in a very sedated and pleasant mood by binging on sugary items. Of course, it's the whole deal: mouth feel, general palatability, etc. I wouldn't have derived the same satisfaction from eating plain sugar, or sugar mixed with butter. In fact I was never a cookie dough type. Things had to be very particular....then, Katie, bar the door.
Not saying I agree with all that btw but it is what's used.
By stimulating the pleasure centers of the brain you could easily induce a "chill out" effect.
That's why I almost exclusively see people describing it in regard to depression type scenarios. Common depressive symptoms include loss of interest and lack of emotion. Actual heightened emotion is usually already seeing a lot of stimulation and often results in things like shaking/nausea.
Saying that a person eats to feel pleasure is a LOT different than saying that emotions are making them HUNGRY.
The ritalin per se is almost always excitatory.
The person claiming can gain at 1000 cals high carb etc made comments at Stephan's and is at a normal weight. She seems genuine but still think she miscalculated cals, didn't give the 1000 cals much time to work or something else - fat malabsorption another possibility.
That said, CarbSane touched in this post on something I was fumbling at in Guyenet's comments-- the variation in energy measurements of different types of macronutrients. Apparently the 4/4/9 is derived by setting food on fire and measuring the output. Plus there are differences in how you cook stuff as well as bodily malabsorption issues.
CICO isn't bad, it's just that fixation on exactness with it leads to a false sense of precision that isn't really possible and also leads to people freaking out over tiny fluctuations in intake.
Hardly unique to dieting. I can't remember how many times I've told people to clean up their significant figure handling ("your measurement methods don't even support theoretical 3 digit significance at most, and 2 digits realistically, yet you're reporting results to 8 digits) and been insulted for it.
For diet, even when one reaches the best available general answer one is left with the personal fight to be flexible.
And the fixation on digits is a SMALL part of it. The much bigger issue is the inflexibility has people abandoning the diet because of one binge.
I've never understood how anyone would think the nutrition data for meat was exact. There is no way the database can know how marbled your meat is, how much fat cooks out or how "waterlogged" your meat is. I've also noticed a huge difference in leanness between ground beef sold in different stores, even though they all say 85% lean, for example.
more emotional than others.
I could even suppose that freaking out helps some folks deal with some issues, it's just their way of working stuff out as best they can.
When the freaking leads to abandonment though ...
How can we seriously have an healthy debate with people who use n=1 as evidence and believe that their personnal experience disprove a study? Once again it made me realize how futile it is to speak about nutrition with most individual. It is saddening me to see that nutrition is not a science to most people. It's getting a cliché, but it's still damn true, nutrition is actually a religion, and BGGC is currently the ''bible'', a i've read many time before already.
Indeed, we never see them in studies, and top coach never see them in real life, but they exist, when they are alone in their kitchen doing confirmation bias on how they can eat so much on a LC diet.
I recently read this blog post by Doug McGuff: http://www.bodybyscience.net/home.html/?page_id=57 where he says about CT scans of obese people: "The most readily obvious finding is the bizarre appearance of a normal size person entrapped in a fatty prison." and "Even more astounding than the abundance of fat, was the lack of muscle. The atrophy was truly profound."
Now this is not a rigorous study, but it seems to indicate that fat people who consider going to this particular gym do not seem to have more lean mass than normal people, maybe even less, or maybe only a little more, or maybe only in their arms and/or legs (outside the CT scans).
I couldn't readily find anything on pubmed that compared lean body mass between normal and obese people. So I'm still wondering if obese people indeed have more lean mass than same-height-and-skeletal-build normal people.
If you are just comparing the general populations, without taking activity levels into account, I don't know how the comparison would go. But if you normalized things by weight, then the answer would be pretty obvious. If you took Michael Jordan and the fat guy who is also 6 foot 6 and weighs the same as Jordan, my money is on Jordan having more lean mass.
So much of this sort of conclusion will depend on how someone decides to slice the data.
Wout: It might depend on how you compare them. If you normalize the groups for level of activity, then it would make sense that fat people have more lean tissue. Take two six feet men who are basically sedentary, but move around a little during the day and maybe go up and down stairs a few times a day
________________
In the lecture I heard McGuff was speaking directly to this point: this thought experiment predicts the obese will have more muscle. It's not there in the people McGuff looked at.
But since it's a completely uncontrolled "study" presented with zero detail (conclusions only) we have no idea how DM thinks the 2 groups he looked at could be comparable ...
I don't have a problem with anecdotal cases per se. The problem is with people claiming results that are so exaggerated that even many typical possibilities (disease, other health conditions, improper counting of calories, etc) cannot be reasonably applied.
If I'm remembering correctly, Rebecca claims that she gained weight on 1100 calories a day and is losing weight on 1800 calories a day, a carefully controlled self-experiment that she has tracked for at least two years. In effect, she is not claiming that metabolic advantage trumps CI/CO; she is claiming that calories do not matter AT ALL.
She is not describing a swing of 700 calories. It has to be more based on her claims, since we have to factor in weight gain/loss. So she's probably claiming a metabolic advantage of at least 1000 calories a day, and possibly as many as 1200-1300 depending on her rate of weight gain or loss. That is not n=1, that is n=BULLSHIT.
She's not the only one making such outrageous claims that cannot be verified. And it will simply muddy the waters because people will point to these apparent outliers as proof that CI/CO is irrelevant and only food choices matter. I'd be interested to see any of them post their diets and see if others have similar success on them. I would not be surprised if either they decide not to publish their diets, or if people who try it don't see such startling results.
So these people add carbs to their diet then freak out when the scale goes up a couple pounds. Anyone who has dieted on low carb diet knows the massive water retention that can occur after re-introducing carbs and it can take a while for your body to adjust and release the water weight.
But if you're actually in a calorie deficit you are losing fat but the water retention can mask it.
This would be true only if a defect in fat tissue regulation was the sole cause of obesity. If it isn't, genetic obesity could result from any other salient factor.
I agree with everything you said. I'd just add that I also think n=1 are good when it comes to ''fine tuning'' a diet to fit the individual, but obviously they are useless when it comes to make general recommandation for the population as in the case with science and saying that n=1 disprove a study is beyond ridiculous, or it certainly could be explain by the fact that, as she said herself, she has no scientific background.
Post a Comment
Comment Moderation is ON ... I will NOT be routinely reviewing or publishing comments at this time..