Taubes-Guyenet debate. My analysis (IV)
(Versión en español: hacer click aquí)
My notes on the fourth segment of the debate (next-to-last) and my comments below them:
- [GT 1h33m] The sugar hypothesis: you add something about a western diet/lifestyle to populations, you get explotions of obesity and diabetes. My hypothesis is that sugar is the something that has to be added. Maybe it is sugary beverages.
- [SG 1h34m] It’s always possible to tell stories to salvage your hypothesis, but that does not necessarily make the story correct. We don’t have any evidence to support what Gary just said. Quote from Christopher Hitchens: «That which is asserted without evidence, can be dismissed without evidence». And that’s the way I feel about this particular story.
- [SG 1h35m] I am going to talk about three different cultures that eat a lot of sugar and don’t have obesity. The Hadza in Tanzania eat a lot of honey. 15% of their calorie intake comes from honey. This is as much as americans eat and they eat fruit on top of this. The pygmies in the Congo also eat a lot of honey. Up to 80% from calories can come from honey during the rainy season. And they are lean.
- [SG 1h38m] If you eat a lot of sugar but everything else is in place, the sugar is not enough to make them fat. I am not saying it doesn’t contribute, I think sugar does contribute, but it is not singlehandedly responsible as Gary has argued. The Kuna of Panama is the 3d culture. They eat white sugar.
- [Joe Rogan 1h39m] In non westernized-civilizations you burn off so many calories…
- [SG 1h40m] I 100% agree with you.
- [Joe Rogan 1h40m] The thing about burning off the calories is that in westernized lifestyles you are just existing, they are doing nothing to burn the shit off. I think what Gary is getting at is that if you have this lifestyle and then with that lifestyle consume sugar you are going to get fat. I don’t think you can compare that to athletes and I don’t think you can compare that to hunther-gatherers. I remember being working out 4h/day and eating everything I was in the mood (soda, cookies, etc.), and not gaining any weight other than muscle.
- [SG 1h42m] I agree that it is not as simple as sugar, but Gary argues that sugar is the primary cause of obesity and that physical activity does not matter, calorie intake does not matter.
- [GT 1h42m] When we talk about the cause of obesity, I don’t believe obese people become obese because they are sedentary.
- [Joe Rogan 1h42m] Don’t you think that an athlete has a high sugar demand and will burn that off?
- [GT 1h42m] Absolutely.
- [GT 1h43m] Can I go back to the Kuna? Stephan calls my stories, stories, which they are, but he calls his «evidence». About the Kuna [he reads a text from an e-mail exchange with Guyenet] they eat a low amount of sugar (32 pounds of sugar per year) and we don’t know how that amount has changed over the years. [Guyenet laughs out loudly while Taubes speaks] I also don’t believe the data from the urbanized Kuna, who apparently drink only 3 8-oz soda per week, and they consumed more sugar back where they used to live.
- [GT 1h45m] Stephan is constantly citing studies which only speak about one hypothesis, like the overfeeding studies, or that are poorly constructed and poorly done. About the Hadza, I already told him that I don’t think it is a refutation of my hypothesis, since my hypothesis is that when you add sugar to the diet of any population you get obesity. May be the Hadza have been eating honey for 1000s of years. May be when they inmigrated to this area they had obesity then. I don’t think finding a hunter-gatherer population that eats honey and doesn’t have obesity is a refutation, Stephan does.
- [GT 1h47m] My non-for-profit funded a study with 40 kids that had NFLD.They could eat fruit, as much as they wanted, but no added sugar and no sugary beverages. The fatty-liver disease resolved. We can’t talk about the mechanism. May be it was the weight loss.
- [GT 1h48m] About the metaanálisis from Hall, it is putting together all the junk-science that has been done in the last 50 years, while ignoring any quality of the studies. And they did a poor job. In one study I checked, they even confused a 400 kJ decrease in energy expenditure with a 400 kcal decrease [N.T. 400kJ=95kcal].
- [GT 1h49m] You have to do the right studies if you want to get the right answer.
- [GT 1h50m] Stephan openly rejected this study that NuSi funded with 12 million dollars. There are basically 3 preople out there that are convinced that everything people like me are saying is wrong. They keep coming over and over again. Kevin Hall claims that he has refuted the carbohydrate-insulin model. When his study comes out supporting it, he works to find out why the study is wrong. We all do the same thing. The goal is ultimately to do the right study, because we have to remember what is on the line here.
- [GT 1h51m] For decades people have thought as Stephan does, and there is an alternative hypothesis and we have to find if it is true because people are dying out there.
- [SG 1h51m] This alternative explanation has been investigated extensively, included studies that were funded by his own organization, NuSi. 2 out of the 3 studies were a clear refutation of his hypothesis. There is a remarkable correlation between studies undermining your beliefs and you thinking those studies are garbage. [Guyenet smiles]
- [SG 1h52m] About the Hill and Peters studies, if you can’t actually find specific problems with the studies, you can’t just dismissed them.
- [GT 1h53m] I gave you a problem with the study: it was designed based on the assumption it was supposed to test.
- [SG 1h52m] Ok, it wasn’t designed how Gary wanted it to be designed.
- [SG 1h53m] The Kuna. The primary basis for Gary’s book is that when sugar comes into this cultures they become fat. I am pointing out 3 cultures that eat high levels of sugar and are not fat. In reference  we can see that in the Cuban economic crisis the intake of sugar went up, 28% of caloric intake, double what Americans eat. Their calorie intake went down and they had to walk a lot because of the lack of gasoline. If Gary is right and calories don’t matter, only refined carbohydrates and sugar, obesity should have gone through the roof. The prevalence of obesity declined by half: from 14% to 7%. The rate of underweight increased only slightly. As soon as they went back to their original diet, obesity went up.
- [GT 1h57m] I would have to read the study to figure out what the problem is.
- [Joe Rogan 1h57m] What about the NuSi studies?
- [GT 1h57m] Well, this is a game people play. My issue with Stephan is that he speaks as if though he knows, with this authority.
- [SG 1h58m] The 2 first studies refuted Gary’s hypothesis. Gary is about the only person who thinks they did not refute his beliefs. The scientific community is pretty unanimous.
- [GT 1h59m] The first study was a pilot study that depends on what you look at, you can think it refutes a belief. Kevin Hall tends to believe that obesity is an energy balance problem, David Ludwig and his colleagues tend to believe what I believe. If you believe Kevin interpreted correctly his study (the one funded by NuSi), then it is not supportive of carbohydrates being responsible of driving insulin and insulin being driving fat accumulation. The David Ludwig’s study reported the opposite. David has critisized Kevin’s study and Kevin’s has critisized David’s study. The middle study was a free-living study done by Christopher Gardner. The low-carb diet was not a meaningful low-carb diet and they told both groups not to eat sugars and refined grains, which are by my hypothesis the most fattening carbohydrates. They tested two diets and neither of them had sugar nor white bread. It was a poorly done study, it didn’t answer the question.
- [SG 2h3m] Caloric deficit was self reported, so it is not actually accurate.
- [SG 2h4m] There was a 2-fold difference in carbohydrate intake.
- [GT 2h4m] If you believe the food frequency questionares that you just said we don’t know whether can be believed.
- [SG 2h4m] I agree with you that refined carbohydrates and sugar are the most fattening types of carbohydrates.
- [Joe Rogan 2h5m] OMG, we can do an agreement?
- [GT 2h5m] The question is why. Is it because people eat too much of them?
- [SG 2h5m] If you believe Gary’s hypothesis, since any type of carbohydrate increases insulin levels more than fat does, if that matters for fat loss, these groups had a 2-fold difference in carbohydrate intake, even though it was predominantly healthy carbs. You should have seen something, not the same amount of weight loss in both groups.
- [Joe Rogan 2h5m] But they are not high in sugar, right? Your argument has been always about sugar, right?
- [GT 2h5m] Stephan is right: I defend my hypothesis. He does the same. We all do it. And every study ever done has plenty of reasons not to believe it. This is why independent replications is something you always want.
The ad populum fallacy
[SG 1h58m] The 2 first studies refuted Gary’s hypothesis. Gary is about the only person who thinks they did not refute his beliefs. The scientific community is pretty unanimous.
The ad populum fallacy is the appeal to the popularity of an opinion as a reason to accept it. The number of people who believe a claim is true is irrelevant in relation to whether it is correct or not. Taubes is aware that his ideas go against the prevailing opinions.
Guna, Cuba and Hadza
I think it would be a mistake to ignore the examples of the Guna, Hadza or the economic crisis in Cuba. Let’s see how they live and how healthy they are.
The Guna are presented by Guyenet as consumers of a large amount of white sugar. Actually, according to his calculations, the «high» consumption of white sugar (sucrose) is 36g/d. That amount includes 1/3 cup of sweetened drink per day. That is a consumption of a 12-oz soda per 4-5 days. According to this study, 23% of their calories come from dietary fat.
Is this an adequate example that you can consume a large amount of sugar and still not have obesity? In my opinion, this example tells us nothing about what happens in a Westernized culture. If, for example, they live on islands and due to their high fish intake they have a good omega-3 to omega-6 ratio, it is possible that this protects them from their —apparently very low— white sugar consumption.
The case of the Cuban economic crisis from 1989 to 2000 is another of the examples presented by Guyenet. As we can see in the graph below, which is included in the article linked by Guyenet, the energy intake was reduced around 750 kcal in the center years of the aforementioned period.
More data: the diet was very low in fat (13% of calories in 1993) and in protein, with only 46 g/d, only 10% of caloric intake (see):
The population had to walk or to use bicycles everywhere due to the lack of motorized transport (see):
Starting in 1990, this new situation, which produced changes in the nutritional status of some groups of the population, combined with a forced increase in the level of physical activity. As a result of the general scarcity of motorized transport, the population began to walk and use bicycles for their daily movements.
There was an increase in the cases of: anemia in pregnant women, chronic protein deficiency, underweight pregnant women, underweight births, etc. In other words, with a nutritionally-very poor diet, based on sugar and rice, when forced to exercise, and consuming much fewer calories than recommended, this population lost weight in the short term (see). And this happened despite consuming around 150g of sugar per day.
The prevalences of obesity in Havana were 11.9 percent, 5.4 percent, and 9.3 percent in 1982, 1994, and 1998, respectively.
As a summary, they lost weight transiently by being hungry with an unsustainable low-energy diet, that was nutritionally deficient and very low in fat, while they were forced to have more physical activity, just as if they were in a concentration camp. Under these conditions, sugar was not enough to produce obesity in the short term. Okay.
As for the Hadza, they apparently consume a large amount of honey during part of the year (source) and generally they do not have excess weight. Assuming that the food that is brought into their camps represents the foods that are available in each month of the year, the following graph would show that a high consumption of honey only happens during, approximatedly, half of the year.
It should be noted that what they consume is neither white sugar nor refined honey: it is honey with bee larvae, which contains more protein and fat than commercially processed honey:
Although only small amounts of protein (mainly free amino acids) are found in liquid honey (Bogdanov et al. 2008), wild honeys contain higher levels of protein and fat, most likely because they contain trace amounts of bee larvae, whereas cleaned and commercially processed honey does not (source)
On the other hand, the diet of the Hadza seems to be very low in dietary fat, with only 11% of the calories coming from fat (see). Could this be the key factor that makes honey not fattening for the Hadza? In any case, there may be other relevant factors, such as the speed of ingestion and honey’s texture, or low levels of omega-6 in their diet.
The Hadza’s story is an interesting one.
Do these stories refute the sugar hypothesis? If they do, we need a better hypothesis that is also based on physiology, not to go back to worship the energy balance pseudoscience.
And if the sugar hypothesis is unambiguously refuted, whatever hypothesis steps up as the next prime suspect has to be very carefully considered. (i.e., not the simplistic notion that people eat too much and move too little). We need a hypothesis that holds the promise of explaining the epidemics everywhere. Gary Taubes
The three studies that NuSI funded:
- The pilot experiment (Kevin Hall)
- DIETFITS experiment (Cristopher Gardner)
- The weight maintenance experiment (David Ludwig)
As we have already discussed in the 3rd part of this analysis David Ludwig’s experiment, I focus on the other two.
» Energy expenditure and body composition changes after an isocaloric ketogenic diet in overweight and obese men «
In this 2016 video we have Kevin Hall bragging about refuting the carbohydrate-insulin model with a not-randomized pilot experiment!! (diets are applied in one sequence only and there is no control group) while on the background there is a poster whose title says the opposite of what he explains in the video (see): «Energy expenditure increases following an isocaloric ketogenic diet in overweight and obese men.»
The increase he found wasn’t enough for him, because he set a requirement as high as he liked so the difference between diets was reported as irrelevant (see): he demanded a difference of 300-600 kcal/d.
These data, although somewhat confounded by ongoing weight loss, suggest that large isocaloric changes in the proportion of dietary carbohydrate to fat transiently increase EE by only ∼100 kcal/d after adjusting for body weight and composition. […] the carbohydrate–insulin model predicts that the KD would lead to increased EE, thereby resulting in a metabolic advantage amounting to ∼300–600 kcal/d (21, 22). Our data do not support EE increases of that magnitude.
So, we see that Hall thinks of a 100 kcal/d difference between diets as small. It is awkward, though, because Hall believes that 10 kcal/d of caloric surplus explain the current obesity epidemic in the USA (see). In other words, he says a 100 kcal/d difference between diets is too little because he demands 300-600 kcal/day to the ketogenic diet, just because he wants to demand that. This is god-level zealotry. And he gets it published in a scientific journal!
We did not include a control group that did not receive the KD or a group that had the diets delivered in the reverse order.
The order of the diets can make, for example, the energy expenditure greater during the first diet, since no weight has yet been lost and there is more fat and non-fat mass. That’s why it was a pilot study, because it was non-randomized and not designed to detect differences between groups. It was not meant to draw conclusions, but rather to learn how to properly design a subsequent bigger study that would have enough statistical power (see). Did you say that the study has limitations? It doesn’t matter: when you think it can be useful to support your pseudoscience, you use the outcome as if it were a randomized experiment.
Moreover, in this experiment the authors (Kevin Hall among them) failed to keep the weight of the participants stable (see).
A major limitation of our study is the unintentional weight loss. Despite slight positive energy balance during the chamber days, the overall negative energy balance amounted to ~300 kcal/d and was likely due to greater spontaneous physical activity on nonchamber days.
It is an invaluable practical demonstration of the actual value of Hall’s ideas about how our body works. Bias? Cognitive dissonance? Zealotry? It’s Kevin Hall, PhD.
Guyenet summarizes this experiment saying that there are no differences. The article says the opposite of what Guyenet says, despite the limitations of the study, which probably played against the ketogenic diet:
The isocaloric KD was not accompanied by increased body fat loss but was associated with relatively small increases in EE that were near the limits of detection with the use of state-of-the-art technology.
«Was asociated.» Another inconvenient truth for Guyenet.
» Effect of Low-Fat vs. Low-Carbohydrate Diet on 12-Month Weight Loss in Overweight Adults and the Association With Genotype Pattern or Insulin Secretion «
This experiment, led by Christopher Gardner, is commented in this blog post. Since in that link we can read a detailed analysis, I go directly to the relevant facts. The participants were supposedly doing a caloric restriction of 500 kcal/d. We see in the following figure the evolution of their body weight:
After 12 months, one group lost a mean of 5.3 kg while the other one lost a mean of 6.0 kg:
Weight change at 12 months was −5.3 kg for the HLF diet vs −6.0 kg for the HLC diet (mean between-group difference, 0.7 kg [95% CI, −0.2 to 1.6 kg]).
There was a difference between diets, but it didn’t reach statistical significance. But the important thing is that in the second half of the study both groups gained weight back. Why did they regain part of the lost weight? Is this what the CICO hypothesis predicts? The problem for Guyenet is as follows:
- If he claims that the outcome shows no differences between diets, he has to assume that the intake data is reliable. Otherwise he doesn’t know if the caloric restriction was the same in both groups.
- If he thinks that eating less is a valid weight loss strategy, he can’t accept the intake data as correct.
If he says that the intake data of this study are reliable, this experiment shows that the hypocaloric diet is a failure as a weight loss diet. That would be another refutation of Guyenet’s hypothesis. Are reliable the intake data from this study, Guyenet?
Taubes makes this point: «it’s the same data that you said that they were not reliable because they were self-reported».
NOTE: Kevin Hall’s low level as a scientist and his lack of rigor are evident in his articles (see, see).
NOTE: it’s not obvious that nonindustrialized populations such as the Hadza have a high energy expenditure that allows them to eat as much as they want and not get fat (source).
Go to the conclusions
Go to the fifth part
Go to the fourth part
Go to the third part
Go to the second part
Go to the first part
30 kJ = 7.16538 kcal
7 kcal/d are enough if they are an «energy imbalance», but for the ketogenic diet to have a metabolic advantage he requires a 300-600 kcal/d difference. Who said bias?
Wiildcard critcism with no real arguments.