Wednesday, June 13, 2012

New Study Demonstrates that Sugar has to be Palatable to be Fattening in Mice

Dr. Anthony Sclafani's research group just published a study definitively demonstrating that high palatability, or pleasantness of taste, is required for sugar to be fattening in mice (1).  Dr. John Glendinning was lead author. Dr. Sclafani's group has done a lot of excellent research over the years.  Among other things, he's the person who invented the most fattening rodent diet in the world-- the 'cafeteria diet'-- composed of human junk food. 

Mice and rats love sweet food and drinks, just like humans.  If you give them a choice between plain water and sugar water, they'll overconsume the sugar water and become obese.  I have argued, based on a large body of evidence, that the reward value and palatability* of these solutions are important to this process (2, 3, 4).  This is really just common sense honestly, because by definition if the solution weren't rewarding the mice wouldn't go out of their way to drink it instead of water, the same way people wouldn't go out of their way to get soda if it weren't rewarding.  But it's always best to confirm common sense with research.

This study didn't address reward in any case-- it was focused specifically on palatability.  In mice as in humans, sweet tastes are detected by special proteins on the tastebuds, and the signal is transmitted within the cell by another set of proteins, and the genes encoding these proteins are known.  Dr. Sclafani's group asked a simple question: "is sugar still able to make mice obese if they can't taste its sweetness?"  They did this by genetically "knocking out" two different proteins, one in each line of mice, that are required for the perception of sweetness on the tongue (5).

The investigators showed that mice lacking these proteins have a normal food intake and body fatness when fed standard lab chow, but unlike normal mice both mutant strains are almost completely resistant to fat gain when given a sugar solution.  This is despite the fact that they drank a similar amount of sugar as the normal mice, which became obese**.  Basically, they drank the sugar water but it was no longer fattening once it didn't taste sweet.

But here's the coup de grace.  To make sure the mice weren't just resistant to obesity in general for some reason that has nothing to do with palatability, they altered the sugar solutions by adding a very small amount of fat emulsion, so that the solutions were once again palatable to all groups of mice.  Now, suddenly, the mice that couldn't taste sweetness got just as fat on the sugar solution as normal mice! The paper's main conclusion:
Our results suggest that nutritive solutions must be highly palatable to cause carbohydrate-induced obesity in mice...
I couldn't imagine a better way to test this hypothesis, and even I wouldn't have expected such a striking outcome.  This study completely shuts down the argument-- never convincing to begin with-- that food palatability is not relevant to body fatness.  Consistent with a large body of evidence in animals and humans, it supports the prevailing view among obesity researchers that excessive food palatability is an important factor in the development of obesity.  It also supports the argument I made in my post "Is Sugar Fattening?", that added sugar can be fattening because it increases the energy density, palatability, and reward value of food, rather than due to a metabolic effect that occurs after ingestion. 

* Reward.  The brain contains a "reward" system, whose job it is to gauge the desirability of food (among other things) and reinforce and motivate behaviors that favor the acquisition of desirable food.  For example, if you eat a strong cheese for the first time, maybe it won't taste very good to you.  As it's digested, your reward system gets wind that it's full of calories however, and the next few times you eat it, it tastes better and better until you like the flavor.  This is called an acquired taste, and the reward system is what does the acquiring, motivating you to obtain a food it has deemed safe and desirable.  Eventually, you may go out of your way to purchase the cheese or beer at the grocery store because you like it so much, and maybe you'll consume cheese or beer even if you aren't hungry or thirsty.  This is an example of the reward system reinforcing and motivating behaviors related to foods that it considers desirable.  Processed "junk foods" such as ice cream, fast food, sweetened soda, cookies, cake, candy and deep fried foods are all archetypal hyper-rewarding foods. 

Palatability is a related concept-- it is simply the pleasantness of a food; how much a person enjoys eating it.  Palatability is determined in part by inborn preferences (e.g., a taste for sugar and energy dense foods), and in part by the reward system (acquired tastes).  Palatability is governed by the hedonic system in the brain, which is closely integrated with the reward system.

The reward system is what motivates you to get food and put it to your lips, every time you eat.  When scientists shut it down in mice, they completely cease eating (6).  The hedonic system influences how much you eat once you begin a meal-- highly palatable food generally increases food intake by activating this system (7).  Together, reward and hedonic circuitry in the brain determine in large part how often you eat, what you eat, and how much you eat, and this is influenced by the attributes of the food that's available.

** Basically, they still find the sugar solution rewarding, just not palatable.  It has been known for a while that sugar receptor knockout mice still find sugar rewarding.  It just takes them longer to develop that reward association, as Dr. Sclafani's paper demonstrated.  The post-ingestive effects of the sugar allow them to form a conditioned preference for the sugar bottle, and therefore seek it out and drink a bunch of it, despite the fact that it doesn't taste sweet to them.  


Christopher in Seattle said...

So is the implication here that in order for the obesity to take place, the brain has to be stimulated in a reward sense? If I'm understanding this correctly, it emphasizes the fact that the brain is regulating body composition, probably via hormonal pathways(?).

A related question I have about all this sugar debate has not to do with obesity, but rather other biological effects sugar has on our protein structures. It's my understanding that sugars cross-link with proteins and cause dysfunction physiologically over time. If this is correct is it safe to say that obesity may be cause only if there is a palatability to the sugar, but does the sugar still cause other problems even if not causing obesity?

psychic24 said...

Just to clarify, "This is despite the fact that they drank a similar amount of sugar as the normal mice, which became obese**", the 'similar' means close to but less than right? Because if there was no caloric difference then they'd have to get obese as well...

Also, you mentioned that these mice took longer to associate reward due to lack of palatability. Did they end up getting obese further down the road (longterm)? Or was this just a short term experiment?

Stephan Guyenet said...

Hi Christopher,

You said "in order for the obesity to take place, the brain has to be stimulated in a reward sense".

I wouldn't go quite that far. I would say in order for obesity to occur in this context, the brain has to be stimulated by high food palatability. The brain is regulating body composition, but it may or may not involve hormones. The brain controls energy expenditure and the release of fatty acids from fat cells.

Hi psychic24,

I was hoping I wouldn't have to explain this, haha, because the study was more complicated than what I related in my post. There were two strains of taste mutants, T1R3 and TRPM5 knockouts (nulls). They also tried two different palatable solutions, sucrose and polycose (a glucose polymer solution that rats love). The T1R3 null mice could taste polycose but not sucrose, while the TRPM5 null mice could taste neither. Polycose and sucrose were fattening to the mice that could taste them, but not to those that couldn't. As I said in the post the study is really quite striking.

Anyway, to answer your question, the T1R3 null mice ate a little bit less sucrose than normal mice but still obtained ~2/3 of calories from the sugar water. Just like normal mice, they increased their total calorie intake when offered the sucrose solution, but did not gain fat. This suggests that in the absence of high palatability, they're able to mount a compensatory response and burn off the excess calories.

The TRPM5 null mice also got about 2/3 of calories from sugar water, but total calorie intake was not increased and neither was fat mass.

When 1% fat emulsion was added to the sugar solutions, making it highly palatable to all groups, they all ate the same number of calories and gained weight at the same rate.

The experiment was 38 days long, which is not a short-term experiment for a rodent study. It was long enough for the animals to more than double their fat stores.

I may discuss some of these details in a further post because they're important. They indirectly support my hypothesis that reward/hedonic circuits directly impact hypothalamic circuits that regulate energy balance and body fatness.

Jared M Johnson said...

How can you conclude:

"that added sugar can be fattening because it increases the energy density, palatability, and reward value of food, rather than due to a metabolic effect that occurs after ingestion."

How do you know that the difference in taste does not have metabolic effects? Don't artificial sweeteners have clear metabolic effects?

Stephan Guyenet said...

Hi Jared,

Yes it is at least partially due to metabolic effects, but the point is that the metabolic effects are due to the taste, not due to some effect the fructose has on the liver or whatever else after it's absorbed.

Josh said...

Wow, very interesting. I look forward to further discussion of palatability/reward driving increased calorie consumption via hedonic pathways vs palatability/reward increasing fat mass via homeostatic pathways.

Gabe A. said...

Are you proposing that some of the metabolic derangements are a result of the hedonic/reward system? This would mean that genetic obesity researchers are looking in the wrong place, would it not? This is all very nebulous to me.

Colldén said...

Lest people get the wrong idea from the title, one thing of note is that the glucose solution was more fattening than the sucrose solution in WT mice, even though they did not consume more calories (so likely has nothing to do with palatability). Though it is in line with much earlier research showing that sucrose reduces feed efficiency.

Renaud said...

So, calories are not fatening, pleasure is ;-)

The role of palatability make obvious sense to me. Reward is a bit more nebulous, but could be interesting also.

I can perfectly see how those mecanisms can lead to overconsumption of "dangerous" foods, or bad eating patterns, that could result in endocrine disruption... wich could also happend for other reasons (genetics, toxics...).

Maybe some can not resist palatability and grow fat, some take this way to induce metabolic dysfunction, and some more have fucked up metabolisms to begin with ?

The one size fit all explanation of obesity does not exist.

Insulin is not fatening ? No, it is not. But fucked up insulin (and other hormones) signaling and/or sensitivity can be. As well as palatability or reward can have some impact.

majkinetor said...

This reminds me on apetite reducing sympthoms of staffed nose. The effect is similar to that of mentioned knock out freak animals apart from the fact that it is not sugar selective. So you don't eat much or at all, you eat because it is programmed, not because it is pleasurable.

So how do we know what amount of reward is normal and expected and what is overly rewarding to the point that it induces consuming in excess of metabolic needs ? You wont eat cardboard or clay, right ? Its common sense.

It is also hard to conclude side effects of gene removal as in this case. Taste receptors, as all other receptors in the body, are not 100% specific so removing them probalby drastically shifts taste profile. There are other side effects, for instance saliva is constantly bathing taste recptors and that triggers specific neuronal pathways. With no receptors this is totally absent.

I am not against reward hypothesis, but I am not sure this study could be interpreted as proof that reward is dominant factor in obesity or something...

This study completely shuts down the argument-- never convincing to begin with-- that food palatability is not relevant to body fatness
I think nobody disputes that food rewad is irelevant, just that its not dominant.

RobR said...

Are any researchers working on a way to temporarily disable the sense of taste / smell in humans? I would think this would be a good way to bypass all of the reward drug problems (Rimonabant suicides); by instead going right at palatability.

I have heard that this can have conflicting results on weight. Some people lose a lot, some people gain weight in spite of losing their sense of taste/smell.

Seth Roberts has people nose-clip while eating for this effect. But I find it tedious as it interferes with normal breathing and swallowing.

As a no-nothing lay person I have this opinion that taking away the pleasure of eating might be the best cure for most obese people. Instead of limiting their reward by avoiding junk, all food becomes essentially tasteless.

Gretchen said...

This experiment doesn't make sense to me from an evolutionary point of view.

Mice find food. Mice eat food. Only mice who enjoyed it gain weight. Food runs out. Mice who didn't gain weight starve to death.

Does this suggest that evolution selects for hedonistic behaviors? If so, why?

EF said...

Dr. Guyenet,

To me this sounds like strong support for the Shangri La diet and the concept of adding tasteless calories to your diet by either noseclipping some food or ingesting the oil. thoughts?

John said...

This study has some interesting information, but there are also many implications about related ideas. First, this, like many other interventions, shows calorie number per se should not be the focus!

It is sort of inconsistent with some of the other posts like "Calories Still Matter." The "idea" that "calories matter" is inherent in our understanding of physics; one cannot argue against or for that statement.

It also somewhat counters the reward hypothesis, unless palatability and reward cannot be completely uncoupled. We should investigate the physiology of these different mice more [on these different solutions]; maybe there already is further info, but we have to buy the full text.

Anonymous said...

All that junk food is so addictive! Interesting that the sugar has to be palatable first.

Jimmy Gee said...


You replied to Jared:

"Yes it is at least partially due to metabolic effects, but the point is that the metabolic effects are due to the taste, not due to some effect the fructose has on the liver or whatever else after it's absorbed. "

Also to Christopher you responded:

"I wouldn't go quite that far. I would say in order for obesity to occur in this context, the brain has to be stimulated by high food palatability. The brain is regulating body composition, but it may or may not involve hormones. The brain controls energy expenditure and the release of fatty acids from fat cells."

So are you saying that the brain is the main control for fat storage leading to obesity and any peripheral effect is more or less secondary?

Gadfly said...

Saying anything is "definitive" using one study is laughable.

Anonymous said...

This is fascinating!

Three things come to mind: do you think this is exclusive to sugar/sweet taste receptors? Or would a similar experiment targeting fat taste receptors (do those even exist?) have the same output?

Secondly, could this also prove some data that shows that replacing sugar by artificial sweeteners isn't necessarily effective for weight loss?

Lastly, do you believe there's an evolutionary reason for this? For instance, hedonistic behaviors would be selected because would allow animals to "stock" more calories (assuming sweet food could be sparse in some contexts and for some animals).

Ed said...

Dinis, given that fat in the solution caused fat gain, I would think that a mouse with its fat taste buds nullified would not gain weight with the fat in the solution. Seems potentially like an obvious follow-up study for Sclafani's group.

John said...

Great study!

"It is sort of inconsistent with some of the other posts like "Calories Still Matter.""

Calories still matter, but you need to look at the complete picture: calories in and calories out. The latter likely went up due to the sugar, allowing a similar intake without the weight gain.

"It also somewhat counters the reward hypothesis, ..."

Sugar water supplies calories and thus it provides a reward. But without taste/texture/smell the brain has nothing to associate this reward to. Only if an association is made between the experience of eating a food and its delayed reward (i.e., the calorie load after digestion and absorption), the brain can anticipate the reward the next time that particular food is eaten.

The way I look at the reward hypothesis is a bit like a special type of classic conditioning: stimulus (= food eating experience) and delayed response (= how the brain reacts to the caloric load of the food).
In this mouse study, the stimulus was simply removed. If Pavlov's dog would have been deaf, ringing a bell wouldn't have worked either. However, that wouldn't have countered the conditioning hypothesis.

John (capital 'J')

George Adventures In Health said...

Not only a great post, but also a great set of comments.

I will have to come back again to this post to ensure I get the updates in the comments.

I feel that we are starting to get on the right track with the brain being the central controller in all this.

i don't, however, agree that our diets must be composed from tasteless pap for them to help us lose our excess fat successfully. When me and my clients do paleo fully, the results are always great, with a little fine tuning for each person in terms of macros I can't see why this wouldn't work for 98% of people.

I could be off the mark there though, but just not with the people I've worked with so far...

Keep up the good work,

Pranay said...

Hi Stephan,
Was there any significant gain in fat mass of the control rodents before the fat emulsion was added than the gain in fat mass after it? Basically, I'm asking if this study at least partially supports the hypothesis that fat and sugar, when combined, are more fattening that either one alone. This could be due to the palatability/reward system. However, since you said only a bit of the emulsion was added, and since the study was probably not trying to test this hypothesis, it might be hard to conclude anything...

Robert Andrew Brown said...

Thanks Stephan fro another fascinating and highly thought provoking post.

Did they measure total calorie intake including the intake of chow?

Did the taste insensitive mice automatically adjust their calories in any way by reducing their intake of chow.

It would be interesting to try the same study with fructose and glucose as well as sucrose.


Puddleg said...

What if eating tasteless food is distressing to mice? After all, an inability to taste is a pretty strong signal that something has gone wrong somewhere.

Did the taste-free mice eat as much?
You haven't really analysed how much calories mattered.

Do some humans overeat because they have better taste receptors?

If this was the case, wouldn't our best chefs be the fattest of us all?

Does insulin or any other hormone enhance the functioning of taste receptors?

Do sweet taste receptors cause obesity by inducing insulin secretion?
This paper seems to reconcile food reward with the insulin hypothesis:

Diabetes Metab J. 2011 Oct;35(5):451-7. Epub 2011 Oct 31.
The Role of the Sweet Taste Receptor in Enteroendocrine Cells and Pancreatic β-Cells.
Kojima I, Nakagawa Y.
SourceLaboratory of Cell Physiology, Institute for Molecular and Cellular Regulation, Gunma University, Maebashi, Japan.

The sweet taste receptor is expressed in taste cells located in taste buds of the tongue. This receptor senses sweet substances in the oral cavity, activates taste cells, and transmits the taste signals to adjacent neurons. The sweet taste receptor is a heterodimer of two G protein-coupled receptors, T1R2 and T1R3. Recent studies have shown that this receptor is also expressed in the extragustatory system, including the gastrointestinal tract, pancreatic β-cells, and glucose-responsive neurons in the brain. In the intestine, the sweet taste receptor regulates secretion of incretin hormones and glucose uptake from the lumen. In β-cells, activation of the sweet taste receptor leads to stimulation of insulin secretion. Collectively, the sweet taste receptor plays an important role in recognition and metabolism of energy sources in the body.

Puddleg said...
This comment has been removed by the author.
Puddleg said...

" in the absence of high palatability, they're able to mount a compensatory response and burn off the excess calories."

That would be, in the absence of the full-blown insulin response.
Because knockout mice will have the T1R protein lacking in all cells, including pancreatic beta-cells.
So they won't make insulin in response to the sweet taste, only once the glucose actually enters the beta-cells in some quantity.

The possibility is that the insulin response is attenuated or modified in the knock-out mice, and this accounts for the decreased caloric intake and the decreased obesity.

John said...

John (with capital J),

As I said, it is pointless to say that "calories matter."

What is the conclusion based on the last several posts?

...There are too many inconsistencies/holes: "Calories matter," and sugar itself isn't a problem; except, a sugar solution is a problem, except when the mouse can't taste sweet, so calories "don't matter;" but the slightest palatability makes "calories matter" again. There *needs* to be an explanation beyond, "The lack of sweet lowered their set-points."

Insulin naysayers are quick to note that the FIRKO mouse has no use when looking at human physiology; how could a mouse without sweet taste possibly be better? What practical insight emerges, when even *minor* complexity/taste makes a significant difference?

cityofelephants said...

Thank you for inspiring us to write about health in our fashion blog come by and check out the health tips on our blog and feel free to follow and leave some of your amazing health advice we would love to hear it from you please make our day like you do with all of your posts and follow

Stephan Guyenet said...

Hi john,

I'm fortunate to count myself among the "insulin naysayers", which includes almost all researchers who actually study insulin biology, and the large majority of those who study obesity as well.

Comparing the taste receptor knockout to the FIRKO mouse is like comparing apples to oranges. Properly functioning fat tissue is required for the development of obesity; no one disputes that. When you completely eliminate insulin signaling on fat tissue, it doesn't work anymore and mice can't store fat regardless of what you throw at them. We could infect their fat cells with ebola virus and it would do the same thing. This has little or no relevance to what happens in common obesity, although it is interesting from a basic science standpoint.

The taste receptor KO mice are impervious to the fattening effect of sugar, and this is due specifically to the loss of palatability, as the study clearly showed. These mice are still capable of getting fat-- their fat cells and everything else work fine, as shown by the fact that they get fat on the sugar solution once it's rendered palatable to them by the addition of a small amount of fat emulsion.

Here are the conclusions from the two studies:

FIRKO: you have to have some amount of insulin signaling on fat cells for them to be able to store fat. We already knew this because T1 diabetics can't keep on fat, but again the relevance to common obesity is murky at best because you're generating non-functional fat cells. Knocking IR out of fat cells is a major physiological perturbance.

Taste receptor KO: sugar is only fattening in mice because it's palatable, not because it causes insulin resistance due to some effect on the liver or whatever. I would say this is actually relevant to common obesity because it's consistent with a variety of other studies that suggest palatability is a key player. Knocking out specific taste receptors is not a major physiological perturbance.

I would be careful about dismissing mutant animals offhand. These animals are some of the best tools we have for understanding mechanism because they allow us to precisely tease apart cause-effect relationships in a way that is typically impossible in humans. But you have to think about what the muation is and what relevance it has to normal physiology.

Puddleg said...

Is fattening an animal the same thing as diabesity in a human?

Fat storage is not pathological, and not every fat person is sick.

An animal doesn't need insulin resistance to get fat (or "sleek" in the traditional use of the word); a human who has IR, hyperglycaemia and hypertriglyceridaemia has a problem, regardless of their weight.

Bill said...

@ Stephen
“they drank the sugar water but it was no longer fattening once it didn't taste sweet.”



p.s. kudos for [finally] so clearly spelling out food reward philosophy.

Grinch said...

I'm still confused how palatabilty per se can lead to fat gain. What about the two dozen metabolic ward studies that show the only real predictor of fat gain is calorie intake?

Gabriella Kadar said...

many years ago I read that a person can gain weight just reading recipes and salivating at the prospect of consuming the food. If true, we need to stay clear of food porn. That would take all this business about palatability, neurotransmitters, hormones to a higher level.

Stephan Guyenet said...

Hi Dave,

A few responses:

1) The metabolc ward studies have shown that the only real predictor of fat gain is energy balance, not just energy intake. Some people ramp up energy expenditure in response to overfeeding, which is what these mice did. See this human overfeeding study for example:
2) Only one of the strains overate in terms of total energy intake. The other strain obtained most of its calories from sugar but total energy intake didn't increase.
3) none of these metabolic ward studies have examined the effects of palatability on energy expenditure in humans. This would be interesting to see.
4) The effects of palatability on energy expenditure were relatively small in the current study and may be below the detection threshold of more noisy human studies.

Regardless of whether palatability influences energy expenditure in humans or not, high palatability certainly increases energy intake, so to some extent the question of whether or not low palatability increases energy expenditure has limited practical value, although it's interesting to think about.

Grinch said...

You're right, about calorie balance being the factor. I guess since varying macro-nutrients didn't seem to have any significant effect on EE, I have doubts that palatability would have any effect.

I could see if palatability simply increased caloric intake, that's certainly good enough to lead to obesity even without any change to EE.

John and Melissa said...

Based on the abstract that George Henderson posted, I would ask if the sugar was even absorbed in the mice that had no sweet taste receptors? Is it possible that the reason they didn't gain weight was because they simply did not absorb the sugar?
Here's the part of the abstract that makes me wonder: In the intestine, the sweet taste receptor regulates secretion of incretin hormones and glucose uptake from the lumen

Melchior Meijer said...

Interesting line of thought, John en Melissa.

"In addition to the regulation of incretins, the sweet taste receptor also controls glucose uptake from the intestine. Glucose uptake from the intestinal lumen by absorptive enterocytes is mediated by two different types of glucose transporters, SGLT-1 and the facilitative glucose transporter GLUT2 [31]. SGLT-1 has an apparent Km of 8 to 20 mM and is important at relatively low luminal concentrations of glucose. The expression of SGLT-1 in brush-border of enterocytes is regulated by a glucose sensor facing the luminal membrane.

Margolskee et al. [20] reported that glucose absorption and the expression of SGLT-1 in enterocytes are regulated by dietary sugars and artificial sweeteners in normal mice. The effects of sugars and artificial sweeteners are attenuated in knockout mice lacking T1R3 or gustducin [20]. They also found that T1R2, T1R3, and gustducin are expressed in enteroendocrine cells [20]. Given that the sweet taste receptor regulates secretion of GLP-1 in enteroendocrine L-cells, they suggest that dietary sugars and artificial sweeteners act on the sweet taste receptor in L-cells, which stimulates GLP-1 secretion and, in turn, upregulates SGLT1 expression in enterocytes, thereby increasing glucose absorption [20]."

Like George Henserson pointed out, the sweet taste receptor is also present in pancreatic beta cells. When the pancreas is unable to 'taste' glucose, there might be a smaller insulin response in these mice after glucose ingestion:

"The physiologic significance of the sweet taste receptor in pancreatic β-cells is unclear at present. Unlike the sweet taste receptor in taste buds and gastrointestinal epithelium, the sweet taste receptor in β-cells is exposed to the bloodstream. Hence, sweet substance(s) in plasma may be a regulator of the receptor. An obvious candidate is glucose, a major energy source in the body. Since β-cells sensitively detect plasma glucose concentration, it would be interesting to clarify the role of the sweet taste receptor in the context of the glucose-sensing machinery of β-cells. As β-cells express the functional sweet receptor signaling system, and since this receptor is activated by glucose, the most straight-forward interpretation is that this receptor is part of the glucose-sensing machinery of β-cells. A critical question is to what extent the sweet taste receptor is involved in glucose-sensing in β-cells. In this regard, the sensitivity of this receptor to ambient glucose is not high, at least in a heterologous expression system [20]. Thus, it is uncertain whether or not the sweet receptor signaling system, as activated by ambient glucose, is reasonably detectable. This is an important issue and should be determined experimentally. Given that the sweet taste receptor is activated by various compounds with completely different chemical structures, it is also possible that some endogenous agonists apart from glucose exist, either in the extracellular fluid or cytosol. Such an endogenous compound, if any, would therefore modulate insulin secretion. This intriguing possibility should be examined experimentally. The sweet taste receptor expressed in β-cells has a unique signaling system in that it activates both the calcium and cAMP signaling systems. Since agents which increase cAMP production in β-cells would protect these cells from various stresses and apoptosis, the sweet taste receptor may be a potential molecular target for developing novel therapeutic agents to treat diabetes."


Melchior Meijer said...

Sorry, this was the link:

Jean said...

Wow! Great post! Nice information. :)

Scott Russell said...

I'm also wondering how much of the sugar they ate was actually absorbed. Since one of the principle regulators of glucose uptake is the sweet taste, how much glucose uptake was actually going on?

I also wonder how this would affect the digestion of the sugars. Does lack of sweet taste impact secretion of sucrase and polycase?

Would these results have been different if they had been fed simple sugars, and therefore the digestive enzymes were not necessary?

An easy way to test this is through a feeding-tube diet. Overfeed people a diet of sucrose that bypasses taste receptors, and see how they handle the bolus of calories absent any taste response (aka zero palatability.)

niceneotech said...

As a child of God, I am greater than anything that can happen to me.Pediatric Intensive Care Unit

Jane said...

I suspect Stephan knows how this works, and doesn't say because he'd be accused of speculation, a terrible crime among scientists.

Here's what I think: palatable food releases beta-endorphin in the hypothalamus, which inhibits the sympathetic nervous system, reducing lipolysis and thermogenesis.

'Activation of hypothalamic beta-endorphin pools by reward induced by highly palatable food'

'Effect of beta-endorphin on sympathetic nerve activity to interscapular brown adipose tissue'

John said...


That's a perfectly reasonably response, but it's not like the palatability is the ultimate cause of their fat storage. There has to be some more investigation on brain and adipose processes in the different mice.

Why aren't more people clear about their views on the insulin hypothesis? It's a big difference to say "...variations seen in humans don't play a/the significant role in human obesity," vs "Insulin doesn't affect fat storage; it's 'calorie balance.'" I also don't see why most people have an all-or-nothing approach. There are numerous interventions that prevent/decrease body fat independent of calories, insulin secretion being one. On Hyperlipid you said blocking hyperinsulinemia doesn't prevent fat, but oxidized oils and diazoxide both do.

Oh, I'm not dismissing mutants. I agree they can be useful.

Jane said...

Yes the palatability IS the ultimate cause of the fat storage. Palatability--> beta-endorphin--> inhibition of the sympathetic nervous system--> less lipolysis, ie fat storage.

The sympathetic nervous system is regulated by events in the hypothalamus, and it innervates adipose tissue where it activates lipolysis.

Isabel said...

My first question would be: what esle may happen when you knock down two different proteins related to the mice ability to perceive sweetness on their tongues? One possible answer:

George A. Kyriazis, Mangala M. Soundarapandian, and Björn Tyrberg
PNAS Plus: Sweet taste receptor signaling in beta cells mediates fructose-induced potentiation of glucose-stimulated insulin secretion
published ahead of print February 6, 2012

Stephan Guyenet said...

Hi John and Melissa and Melchior,

The experiment was controlled for that possibility. When the sugar solution was made palatable by a small amount of fat emulsion, the taste receptor deleted mice fattened on it just fine. This demonstrates that they are not malabsorbing the sugar or having some altered physiological response that prevents fat gain as a result of the knockout.

Hi john,

I know Peter likes to cite the diazoxide stuff, but that's actually pretty poor evidence for the insulin hypothesis. For one, diazoxide is not specific to insulin secretion-- it has many effects on the body. Peter likes to cite the diazoxide trial where it lowered body weight, but he never seems to cite the trial where it greatly reduced insulin levels but had no effect on body weight using a similar study design. I wonder why??

"These findings do not suggest that hyperinsulinaemia per se contributes to maintenance of the obese state, and insulin secretion inhibition seems not a promising drug target."

Isabel said...

My first question would be: what esle may happen when you knock down two protein related to the mice ability to perceive sweetness on their tongue. One possible answer:

George A. Kyriazis, Mangala M. Soundarapandian, and Björn Tyrberg
PNAS Plus: Sweet taste receptor signaling in beta cells mediates fructose-induced potentiation of glucose-stimulated insulin secretion
; published ahead of print February 6, 2012

Isabel said...

The first question that comes to my mind is: what esle may happen when you knock down two proteins related to the mice ability to perceive sweetness on their tongue. One possible answer:

George A. Kyriazis, Mangala M. Soundarapandian, and Björn Tyrberg
PNAS Plus: Sweet taste receptor signaling in beta cells mediates fructose-induced potentiation of glucose-stimulated insulin secretion
PNAS 2012 109 (8) 2713–2714; published ahead of print February 6, 2012

John said...


Umm, are you sure you know what "ultimate" means?

...You supported my statement with your flowchart thingy.


Yes, there is some inconsistency diazoxide; but, there are other related interventions, and you just said yourself insulin/insulin signaling was necessary [but not sufficient].

Why is it common to designate it [insulin] as irrelevant when the evidence is so clear? I don't think (and I doubt any reasonable person does) insulin is the only factor regulating fat storage, but manipulating it does work in several studies. To me, it's plain silly to think it's not part of the puzzle.

Stephan Guyenet said...

Oral diazoxide gets into the hypothalamus and alters neuronal activity there:

"A drug called diazoxide (which is not used for treating diabetes) was administered to ten healthy subjects, drug that leads to the opening of K channels in the hypothalamus. The resulted hormonal secretion of the pancreas was closely monitored to detect any change in glucose production that occured due to diazoxide administration. Once the drug was administered the scientists observed that the liver slowed its glucose production.

The researcher then administered diazoxide orally to rats and noted similar findings. A sufficient concentration of diazoxide passed through the blood-brain barrier and sucesfully stimulated the hypothalamus K channels. Afterwards researchers blocked the K channels using another drug administered directly into the brain."

Paper ref:

The hypothalamus plays a key role in energy balance, so I don't see any reason to speculate that diazoxide's effects on the pancreas are what causes fat loss in some studies. To accurately interpret the results of drug studies, you have to have a solid understanding of what the drug actually does. Diazoxide opens a specific type of potassium channel that is used by a variety of cell types, including insulin-secreting beta cells and hypothalamic neurons, to regulate their activity.

Since diazoxide crosses the blood-brain barrier, and is known to influence the activity of the hypothalamus when administered orally, this is the most obvious candidate for its weight loss effects in some studies.

Stephan Guyenet said...

Another point to add. The insulin-lowering somatostatin anologue octreotide that Dr. Lustig used in his studies also does not offer much support the insulin hypothesis, for similar reasons to diazoxide.

Octreotide is an extremely nonspecific drug that "inhibits secretion of many hormones, such as gastrin, cholecystokinin, glucagon, growth hormone, insulin, secretin, pancreatic polypeptide, TSH, and vasoactive intestinal peptide"

It's also a potent mu opioid receptor antagonist ( What do mu opioid receptors do? They're one of the key elements of hedonic (palatability) signaling in the brain. Other mu antagonists such as naloxone reduce food palatability and cause weight loss in animals and humans. The mu knockout mouse is resistant to diet-induced obesity. In Lustig's trial, participants had improvements in their "carbohydrate addiction score", which was basically how drawn they were to junk food. This is exactly what you see with other mu antagonists like naloxone that have no effect on the pancreas.

So again, as with diazoxide, the effects of octreotide are consistent with actions in the brain and there is no evidence to suggest that their effects on insulin secretion are the reason why they case fat loss.

Scott Russell said...

If it is as you suggest, and there is not a failure to absorb the excess sugar calories in the sweet receptor knock out mice, what happens to those extra calories?
(I don't have access to the full article so I don't know the intake and expenditure, but you make it sound like there was not a compensatory decrease in food consumption to justify them not becoming obese.)

I would like to know how sweet taste receptors affect sucrase excretion in these mice. You can't get fat off of sucrose without the sucrase.

Sanjeev said...

you folks can do some simple "reality checks"[0] on your suppositions about knocking out taste receptors

in non-KO rats and humans, does fructose release more insulin than glucose (which is supposedly not sweet at all, according to some ("tasted like tiger piss"))

does Aspartame (200 times sweeter than sucrose) release much more insulin than sucrose? Splenda (600 times sweeter), saccharine, cyclamates, ?

Do all of the above (among others) have massive hormonal and digestive effects as they trigger all those receptors in non intestines and pancreases?

[0] not to be confused with experiments, just "consider opposite angle to my proposal"

Sanjeev said...

sorry for the typo: remove "non" in this

> trigger all those receptors in non intestines and pancreases

Sanjeev said...

> Does this suggest that evolution
> selects for hedonistic behaviors? If
> so, why?

If there is no plausible answer what is your conclusion?

Sanjeev said...

> The "idea" that "calories matter" is inherent in our understanding of physics;

inherent ... dropped fully formed into peoples' heads at the same time that they were creating physics.

inherent ... no one had to formulate the idea

inherent ... no one had to fight to get it accepted

inherent ... no one had to work hard to figure out exceptions

inherent ... riiiiiight

it's an hypothesis. People had to work to formulate it, then they and others worked to prove or disprove it.

Just because all working scientists and engineers use it doesn't mean it dropped, fully developed, magically into peoples' heads as physics developed. and no one had to do any work to formulate it or prove it.

Your assertion is factually incorrect

There's nothing "inherent" about it.

(this isn't the earlier post I mentioned)

Sanjeev said...

> It is sort of inconsistent with some of the other posts like "Calories Still Matter." The "idea" that "calories matter" is inherent in our understanding of physics; one cannot argue against or for that statement.
trying to verbally minimize or sideline an idea or distract from it to insert your own agenda IS arguing against it.

Do you enjoy doing things at the same time you claim they can't be done?

Gretchen said...

@sanjeev I think the idea that glucose isn't sweet comes from antifructose people. It's very sweet. Buy some glucose tabs sometime and try them. Or try some clear Karo syrup, which is mostly glucose.

Sanjeev said...

Gretchen, i commented either here or at carbsanity that I find dextrose almost as sweet as fructose and asked if anyone else has had this experience and how individual that is.

I've been eating and using dextrose for a long time. Around 3 years ago local grocers (small and big chain stores) stopped carrying it so I started buying it at wine making shops.

Amazingly, some local grocers are carrying pure powdered fructose in kilo bags (no pure dextrose).

Steven J said...

Hi Stephan,

I was just wondering if you could clarify the energy expenditure part of the food reward/palatability theory. Is it that increased food reward increases food intake and expenditure is unable to compensate to the same degree, unable to compensate at all, or actually decreases expenditure? And what is the reasoning for whichever mechanism it is? Thanks a lot for any clarification.

Bill said...

Do the Trpm5 KO data directly contradict the Food Reward hypothesis? Trpm5 KO mice aren’t attracted to the taste of polycose or sucrose, but get fat on either of them (independent of overeating). I.e., “food reward” is not necessary for obesity. But more importantly, sugar is fattening whether or not you like it.

John said...
This comment has been removed by the author.
RobR said...

Stephan: I found this article a little confusing.

So the conclusion it reaches is that food reward per se isn't as fattening (driver of overconsumption) as food palatability?

And that food palatability somehow influences energy out (exercise + NEAT) ?

Are there any other studies examining palatability/reward affecting energy out?

I look forward to your next article.

Sanjeev said...

> It is sort of inconsistent with some of the other posts like "Calories Still Matter." The "idea" that "calories matter" is inherent in our understanding of physics; one cannot argue against or for that statement.
That's what you wrote before.

You admit it's terrible, terrible, terrible by not even trying to defend it directly.

It was so bad you are trying (in vain) to completely changing what I replied to.

Try (in vain) to change the plain meaning as much as you want.

Please try again. More name calling will help your case immensely.

my prediction: you'll never answer my direct response with an on-topic, relevant response.

OR you'll invent some far-fetched, illogical craptrap to justify what you wrote.

Sanjeev said...

Where did the word inherent go john?

That was a big part of the comment I replied to

What happened to "cannot argue for or against it"

That was another big part

What happened to you claiming it can't be argued, then making arguments[0] against it?

are you capable of staying on topic john? Capable of understanding

English and logic?

[0] invalid, illogical, cr*p but still arguments

John said...
This comment has been removed by the author.
Melchior Meijer said...

Hi Stephan,

I suspected something like that. Thanks. It was a nice try ;-).

Jane said...

The link between hyperinsulinemia and fat storage might be due to failure of autophagy, which people are beginning to blame for practically everything, with good reason.

Cells do autophagy all the time. If they don't, they don't work properly. Pancreatic beta cells for instance might become less sensitive both to glucose and to feedback inhibition by insulin. This would mean hyperinsulinemia.

We now know that autophagy breaks down fat as well as everything else. They call it lipophagy. This means, failure of autophagy would be the ultimate cause of both hyperinsulinemia and fat storage.

People used to blame insulin resistance for everything. Well, we now know insulin resistance is linked to excess mitochondrial superoxide, which is the same thing as deficiency of MnSOD. MnSOD is responsible (in fruit flies at least) for inducing many autophagy genes. This is supposed to be why it's associated with longevity: autophagy clears out the junk so everything works better so you live longer.

Stephan Guyenet said...

Hi Bill,

No offense intended but you must have misread the study. The TRPM5 KO mice were almost totally resistant to fat gain from either sucrose or polycose (fig 2). They gained a slight but significant amount of fat on both, but the normal mice gained 4-7X more fat.

Basically, mice of every genotype got fat on the solutions they found palatable, but they didn't fatten (or scarcely) on those they didn't find palatable. The result could not be more clearly supportive of the food reward hypothesis.

bentleyj74 said...

This is going to sound really stupid but could you humor me please?

The knock out mice consumed less than the intact palatability mice until another mode of reward [fat] was introduced...correct or incorrect?

I was sort of unclear whether you were saying they didn't fatten despite comparable intakes or that they weren't interested in consuming to the same degree.

Colldén said...

I think "got fat" is somewhat exaggerated. The sucrose-fed mice had about 0.6g more fat by the end, but they ate 114 kcal more over the course of 38 days, this means that less than 4% percent of the extra energy they consumed was stored as fat.

The control mice had about 3% BF and the sucrose-fed WT mice about 5% BF at the end of the experiment. Compared to high-fat diets that often send them up into 30-40% BF this is a miniscule increase, especially considering that high-fat fed mice usually aren't significantly hyperphagic beyond the first days.

Puddleg said...

it's not necessary for sweetness receptors to increase insulin secretion (or sugar absorption) by very much at all, given the minor weight gain involved.
Nor are we, by mentioning this possibility, saying "Insulin Hypothesis Q.E.D.".
We are merely saying that, given that these receptors are involved in insulin mechanisms, a study that does not mention, quantify, or eliminate this effect cannot falsify the insulin hypothesis.

As for the calories in = calories out shibboleth, this is supposed to apply to closed systems.
Few if any human diet studies even approach being closed systems.
The diet studies that take place in a sealed box, with duplicate meals to burn, and all excreta carefully tested for calories, are the only studies for which calories in = calories out.
Of necessity they last days rather than months and have small n= numbers. They can't tell us much about the sort of thngs we want to know.
Energy is like time; it can't be created or destroyed, but it can be wasted.

Sanjeev said...

My responses to john were specific to his comment and the invalidity I read there.

I intended no wider meaning than those specific crappy argument tricks.

> Nor are we, by mentioning this possibility, saying "Insulin Hypothesis Q.E.D.".

I neither wrote nor believed that anybody did claim that, I was suggesting a procedure, a thought experiment, that could be used to put the proposals in a wider context.

> As for the calories in = calories out shibboleth

it's NOT a shibboleth

in the context it's clear what it means

> a principle or belief that is considered to be old-fashioned and no longer important

I can GUARANTEE conservation laws are being used right this minute for design and analysis ... computers are running simulations on airplanes and cars and HVAC systems right now solving equations derived using conservation laws
> A shibboleth is like a motto or catchphrase that members of a group tend to say, like the conservative shibboleth that the only good government is a small government.


a belief, principle, or practice which is commonly adhered to but which is thought by some people to be inappropriate or out of date
a custom, phrase, or use of language that acts as a test of belonging to, or as a stumbling block to becoming a member of, a particular social class, profession, etc


A word or pronunciation that distinguishes people of one group or class from those of another.
A word or phrase identified with a particular group or cause; a catchword.
A commonplace saying or idea.
A custom or practice that betrays one as an outsider.

You could go with the ones right at the end of course

Sanjeev said...

> The diet studies that take place in a
> sealed box, with duplicate meals to
> burn, and all excreta carefully tested
> for calories, are the only studies for
> which calories in = calories out.

I think you have an argument with both john and js290 then, because according to them it's not only those studies for which calories in = calories out

here's their opinion

> john said...

> Sanjeev,

You seem to be some kind of insane, so I'll
> comment one more time. The following is pretty
> much how I see it, as commenter js290 says,
> "Conservation of Energy is a boundary condition
> in the control system you're interested in, in
> fact it's a boundary condition in ALL systems in
> this physical universe. Hence, it's a necessary
> effect.

see that?

calories in must = calories out for all systems in this physical universe.

They don't qualify it with
" ... are the only studies for which calories in = calories out."

This kind of thing happens when people discuss concepts they've never applied to a real HVAC system or never had to use in modeling fluid dynamics or heat transfer.

Sanjeev said...

or use in designing electronic circuits or use in modelling explosion dynamics of volcanoes or industrial safety accidents (all applications of conservation laws) ...

Sanjeev said...

john and js290 also seem to give the "only applicable to closed systems" thing a pass.

IMHO you may be misunderstanding the closed systems thing: it doesn't tell people "don't use conservation laws" or "conservation laws are only valid in some extremely proscribed circumstances"; it's prescriptive:

When you set up your system for analysis draw your lines/boundaries (define your system) so it is closed (or as close to "closed" as you want to get for the sophistication of the analysis)

Sanjeev said...

sigh ... I edited 2 sentences together & left one word unedited ...

> only valid in some extremely proscribed circumstances

replace proscribed with "rare and restricted".

Puddleg said...

In fact, I doubt whether even sealed-box diet-exercise experiments ever match calories in to calories out exactly to the nearest whole calorie, especially if they run for long periods.

But the average diet study makes large assumptions about calories. You will find very few in which excreta are weighed, let alone analysed, for example. Exhaled ketones are not counted. Temperature is not constantly recorded at as many locations as possible (and temperature can vary by a degree or more between sides of the body). Exercise may be assigned a value determined in an earlier experiment.
Some assumptions about what is and is not important are being made, not tested, more often than not.

If energy is lost through heat and this is not being measured, then this aspect of the system is open.
If the subject is free to fidget, it is open. An open system is just one where all energy is not accounted for by the observer or returned to the system.

The insulin hypothesis is surely the parsimonous explanation of weight gain (if Occam's razor can be usefully applied to something as hairy as biology).
Food reward studies and the facial expressions of mice may help to modify the theory by filling in some details.
And perhaps by contributing to the psychological factors (after all, hypnosis has some success too).
The problem is that the insulin hypothesis is the rationale for diets that help people lose weight.
The fattening carbohydrate has been a useful concept for nearly 200 years.
The food reward hypothesis hasn't launched any diet books yet, has it?
I've seen a couple of tubefed crash dieters on TV recently, that's all.

rahul said...

I am happy to see the post here thanks for it.
Wholesale Vaporizers

Reijo said...

Thanks for an interesting post again. It's fun to read these. A recent study suggested that a sight of food is enough to increase levels of grehlin (hunger hormone).

Have you ever written about TV commercials and their effect on food intake?

John said...


I apologize about the personal insults. Let me clarify: to me it's clear that attempting to argue for or against the idea that "calories matter" is to basically argue for or against an analytical truth, and perhaps a tautology depending on context, as there is no counter argument that stays within our understanding of science/physics. So when I say, “It's true," I'm not trying to persuade someone with evidence or anything—it’s as if someone made an addition error and I pointed out the correct sum, or confirmed a correct addition.

And yes, I am giving the "closed system" condition a “pass,” but actually a pass shouldn't be necessary--what is their point when they say humans aren't a closed system? ...The first law applies to all systems; you just need to know the definitions of systems for it to be consistent (and make sense)!

Alex said...

Infants are able to put on fat easily with a feeding tube diet of formula. I don't know if the same thing occurs in adults but this is a clear example of a mammal putting on fat despite not having any taste of what they're eating.

John said...

George Henderson,

Are you mixing up "closed" with "isolated"?

An exchange of heat and/or work with the outside environment is within the conditions for a closed system (but matter is constant). A human/cell is an open system though.

Amy said...

If we're voting on glucose being sweet; such as in the dextrose form, I'd say no. If I use real sugar in my tea I use much less sugar (1 teaspoon), than if I add dextrose (4 teaspoons). I'd say that glucose is less rewarding of a food, but I tend to gain more weight on it because it is easier for my body to absorb and it doesn't make me as full as fructose. One can of soda can keep me full for an entire day. Fructose has a more severe effect on me because I have a mild fructose intolerance, but other people have commented to me that they fell less full eating glucose based foods; like white rice. I suspect that is because glucose is so easy for the body to absorb.

Puddleg said...

and because the human is open to exchange work and heat, noise and fumes, etc with the outside - except under very rare experimental conditions where there is an attempt to measure ALL of these scrupulously - any citing of the first law of thermodynamics, conservation of energy, mass etc. in the context of diet-exercise is irrelevant, because there is a second law (entropy or increasing disorder) which is being ignored.

An analogy: sometimes you don't know how energy-efficient your heat pump is until you get the power bill. And if you keep leaving the bloody door open, you'll never know.

Insulin resistance? Food reward?
Who needs them if obesity is caused by a virus that promotes insulin sensitivity:

Always Sick Chick said...

Seems pretty obvious to me. They actually did a study on this?

Bill said...

@ George Henderson,
I totally agree with your summary of the necessity for a comprehensive assessment of energy balance, but I don’t think I have EVER seen one published… have you?


Jane said...

'..The insulin hypothesis is surely the parsimonous explanation of weight gain..'

There is actually an even more parsimonious explanation than insulin: stress. We know stress at work is 'an important risk factor for the metabolic syndrome'
..and stress means changes in autonomic nervous system activity.

The sympathetic branch of this system controls adipocyte function via oscillations in lipolysis.

'Rapid oscillations of visceral lipolysis have been reported. ...FFAs and glycerol oscillated in phase at about eight pulses/hour. ...These results show that lipolysis in the fasting state consists of an oscillatory component, which appears to be entirely dependent upon sympathetic innervation of the adipose tissue, and a non-oscillatory, constitutive component...'

We don't know what happens to these oscillations in stress, and there are reports of both under- and over-activity of the SNS in obesity. But clearly, derangement of the oscillations is likely to disrupt the balance between lipolysis and lipogenesis.

I don't think stress can explain obesity in the absence of micronutrient deficiencies. But then, the same can be said for the insulin hypothesis.

JT said...


Great post!

Just to make sure I understand: It was the palatability that drove the overconsumption of the sugar solution and it was the sugar in the solution that drove fat gain. Without palatability, there was no fat gain from the sugar.

So what would happen if there was palatability and no sugar? Would there be overconsumption of the solution but no fat gain since there are no (or few) excess calories to consume? Or would the palatable solution drive the rats to seek more rewarding food since there is no reward from the water?

I guess the essence of what I'm trying to ask is this. How dependent is overconsumption on the presence of both palatability and reward? Must you have both? Or is an excess of one or the other enough to drive overconsumption and thus fat gain?

Stephan Guyenet said...

Hi JT,

You're asking some important questions. I didn't get into the details of the study because it was too complex for a single blog post.

When all of the details of the study are considered, it offers remarkable support for the idea that palatability influences fat gain (and completely deep sixes the idea that post-absorptive effects of fructose are relevant to fat gain, at least in this context).

The study actually examined two different palatable carbohydrate foods, sucrose and a sort of short-chain starch called Polycose. Polycose is made of starch oligomers (therefore it's 100% glucose), tastes a little bit sweet to us but rats/mice have taste receptors for it and they LOVE it (humans may or may not have receptors for these compounds-- currently unclear).

Anyway, normal mice can taste sucrose and Polycose, the TAS1R3 mice can taste Polycose but not sucrose, while the TRPM5 mice can taste neither. Normal mice got fat on both, TAS1R3 mice only got fat on Polycose, and TRPM5 mice got fat on neither. So mice only fattened on the ones they could taste. It's worth mentioning that normal mice got the fattest on Polycose-- the one that's 100% glucose oligomers and contains no fructose.

You asked about the different roles of energy intake vs. energy efficiency (i.e. fat mass accumulated per calorie eaten) in this study. What the study showed is that both are important. The TAS1R3 mice didn't gain fat on sucrose despite the fact that they drank a lot of it and their calorie intake increased just as much as normal mice. This suggests that in the absence of the palatable taste, they were able to burn off the excess calories through a normal homeostatic response. Another way of putting this is that their 'setpoint' was not raised by the sugar so they 'defended' a lower level of body fat mass.

However, in the TRPM5 mice, they still ate tons of sugar and Polycose, but these displaced an equivalent amount of chow so total energy intake remained constant. So they 'defended' leanness in a different way-- by not increasing calorie intake in the first place.

As soon as the solutions were made palatable once again by adding a small amount (1%) of fat emulsion, all strains fattened similarly on sugar, proving that both strains are physiologically capable of absorbing and metabolizing sugar normally, and putting on body fat mass normally, as long as the palatability is there.

So to answer the rest of your question, here is a succinct summary:

-Food reward is one of the major determinants of what and when you will eat. It is the motivating ability of a food; how 'tempting' or 'seductive' a food is. e.g. if soda were not rewarding, no one would go out of their way to drink it instead of water. No one would eat donuts, despite knowing they're unhealthy, if donuts weren't rewarding. This is a common sense but key point that Taubes and the other denialists have failed to grasp due to a lack of understanding of the concept. If you think sugar-sweetened beverages are relevant to obesity, then you think food reward is important-- there's no way around it.

-Palatability is one of the major determinants of how much you'll eat once you begin a meal.

-In addition to determining when and what you eat, there is accumulating evidence that reward and/or palatability determine what you do with the calories you eat, because of their effects on brain centers that regulate fat mass. In other words, X number of calories that are highly palatable will be more likely to be stored as fat than X calories that are less palatable, at least in the context of excess energy intake (it may not matter if you don't overeat in the first place, it's unclear). That's one of the key findings of Sclafani's paper. I may discuss it in more detail at some point on the blog, but it would be somewhat technical.

Stephan Guyenet said...

To add a little bit more to that comment, reward is what gets you to seek a food out and begin to consume it, and palatability is one of the things that determines how much of it you eat. They work together to drive the overconsumption of unhealthy foods in the modern food environment.

Puddleg said...

"To add a little bit more to that comment, reward is what gets you to seek a food out and begin to consume it, and palatability is one of the things that determines how much of it you eat."

But surely good French quisine is the most rewarding and palatable food in existence; even junk food addicts would probably choose from the Larrouse Gastronomic/ Michelin Guide first if they had the choice;
and yet we have a French paradox.
I don't see Jamie Oliver as the enemy of food reward or palatability...

To a heroin addict, heroin is very rewarding, but to a naive user, there is considerable aversion to its effects. The compulsion to seek the reward develops as a result of exposure.
If people eat high-carb food on weaning because they are hungry and it is all that is offered, they can develop a compulsion to seek it in future; but this does not mean that it is intrinsically more rewarding than other foods.
What we eat when very young is more strongly associated with the cessation of hunger than novel tastes. But this might be a conditioned response, not an intrinsic property of the food.

A person who is starved as an infant may develop cravings for food you or I would consider highly unpalatable. Some people overeat in adulthood because they were conditioned not to waste food in infancy - completely regardless of what is actually on their plate.

Stephan Guyenet said...

Collectively, I've spent several years in France, over the course of my lifetime, I grew up with a French father, and I have French friends both here and in France. I know a thing or two about how people eat in France in their actual homes on a daily basis.

Point #1: people don't eat at Michelin-ranked restaurants every night, in fact generally they eat at restaurants much less then we do here. Nice restaurants are expensive in France and traditionally they were considered a rare treat, i.e. once a month if you could afford it.

Point #2: on a daily basis, French people tend to eat relatively simple home-cooked food made from scratch. It's tasty and satisfying because it's made from fresh ingredients and composed with skill, but it is by no means hyperpalatable. French people don't have the time to make tiramisu and duck a l'orange every night.

People get the wrong idea when they go to France as tourists or get a French cookbook. This is the same mistake people make with Asian food. The average person in the countryside in most of Asia is eating food that would be relatively low palatability to us (very low in fat and meat, repetitive and often heavily based on grains and legumes).

Puddleg said...

I hear you, I don't disagree.

But surely the point of home cooking, unless you are a health nut, is to make the meals your family finds rewarding and palatable?
If the main is sufficiently rewarding they don't demand dessert or snack afterwards.

As for people who do fatten in Michelin restaurants, didn't Brillat-Savarin identify the offending foods long ago?

Two differences between the French way of life and those of the US or UK would be, lower sugar consumption, and greater retention of scheduled eating (less snacking) which go hand-in-hand.

But I didn't really mean to cite France as proving anything, the "french paradox" was a throwaway line, one can eat fine quisine anywhere in the world.
It is surely as rewarding and palatable as junk food, but less fattening.
The gourmets would be the bariatric avante garde if food reward and palatability decided obesity.

Stephan Guyenet said...

I have to disagree about chefs, they are quite often overweight or obese. But even if they weren't, it wouldn't tell us much because 1) chefs are on their feet all day being active. It's not a sedentary occupation. 2) being a top chef is not just about maximizing food palatability and reward. Presentation, creativity and atmosphere are just as important. 3) chefs are not exactly at the bottom of the socioeconomic barrel-- that also influences obesity/health risk.

I don't know why you say that gourmet food is not as fattening as junk food-- do you have any evidence to support that statement? Of course it depends on the type of gourmet food we're talking about. If you go to a restaurant where the plate only has two artfully arranged slivers of radish on it, it's not going to be fattening.

But I can speak for myself that I will often emerge from a fine dining experience far more side-splittingly stuffed than I would from a fast food joint if money is no object. There's no question that I would gain fat if I ate like that every night.

Puddleg said...

Why does being at the bottom of the socioecomonic bucket affect weight?
Because surely it does.
Obesity as a form of malnutrition.
I think I read about that in "Good Calories/ Bad Calories".

The "let them eat carbs" answer to poverty from government; the "because you don't have time to cook" theme that weakens the role of scheduled mealtimes; the cheap sugar and flour that makes this possible.

Mice in labs are homogenous (very closely related); humans are heterogenous; taste responses are likely to vary far more than other, more essential, aspects of metabolism (like glucose/ insulin) between unrelated individuals.
Even then, conditioning can over-rule such responses.
As can immune response;
If you are gluten intolerant, you can reach a point where something that was once addictively tasty comes to burn your tongue.

Stephan Guyenet said...

The American poor are not exactly eating a low-fat diet. White flour consumption has dropped by ~50 percent in the US in the last century. Added fats have skyrocketed, and total fat consumption has increased in the last 100 years. Refined seed oil is one of the cheapest calorie sources there is today.

In terms of nutrient composition, the US has departed from its historical diet patterns mostly via increased consumption of added fats and sugars, and this is true at all socioeconomic levels because everyone can afford both.

Puddleg said...

New Zealand has the highest per-capita honey consumption in the world, and the highest female youth suicide rate in the OECD.
But are the people eating the honey the ones committing suicide?

Some people have gotten thinner while the US diet has changed.

Correct me if I'm wrong, but the most obese Americans are eating more seed oils, more sugar, more corn products. It's possible that any increased sweetener intake has been at the expense of starch, but perhaps not so much grain starches.

It is really what the obese eat that we need to question.
They don't eat Atkins or low GI or diets designed to minimize insulin resistance or keep blood glucose stable.
Granted they probably don't eat low-fat overall, although they probably do eat a lot of low-fat and fat-free foods and drinks.
People that need to be cut out of their houses do seem to eat in ways that support the insulin hypothesis, such as it is.

Consider the Rutherford-Bohr model of the atom; it was soon rendered obsolete by quantum physics and closer observation.
Yet every metallurgist, every meth cook, every nutritionist still acts as if the Rutherford-Bohr model were all there is to the atom. Only quacks try to drag quantum physics into these practical uses.
This is how I see the insulin hypothesis; it can perhaps be easily falsified in various ways,
it fudges detail and oversimplifies, yet it will continue to be the preferred model, because of its practicality.

Puddleg said...

And maybe this can be clarified by a simple thought-experiment.

Suppose a version of the food-reward/palatability hypothesis is accepted as the new paradigm.
Scientists advise governments, and food manufacturers are regulated to comply.

What then? What are the practical applications?

At present, anyone with a passing knowledge of the insulin hypothesis can find information on food packaging, and online for non-packaged foods, that allows them to construct the dietary balance that they suppose to be least fattening and/or least inflammatory.
One can read the % (g per 100g) of carbohydrate, sugar, fat, sat fat, protein, and even the calories or kilojules if one is so inclined.
It is even reasonably easy to estimate PUFA where this information is not supplied.

Under a food reward doctrine, will we have labelling information about palatability or reward?
Out of five stars, perhaps?
Values determined by a panel of scientists, a panel of fatties, or a panel of mice with expressive faces, interpreted via the lastest algorhythms perhaps?

a) will this tell us what we don't already know?
b) will it tell us as much as the advertisers are already paid to tell us?
c) what are we supposed to do with this information?
d) given there is limited space on packaging, will this new information come at the expense of other data?

Sanjeev said...

Every question you've raised has been hashed out here and several other places many times. carbohydrate ->insulin -> obesity guess/proposal is parsimonious, but that doesn't help any because the weight of the evidence says it's simply incorrect.

If you want more information there is a TON of it. Like I wrote above, just about every question you asked here has been hashed out many many times and as a former Atkins/Taubes high-fatter/low-carber I'm now convinced, CIGuess is WRONG.

> What then? What are the practical applications?

here's a simpler thought experiment:

a person goes low carb for a year, and since he believe calories don't count he keeps no track

At the end of the year he's stuck to the diet and gained 20 pounds.

He's STUCK. How can he troubleshoot? CAN't DO IT ... He kept no records;

If he had behaved as if CICO is correct he would have records to use, a trail of bread crumbs to figure out what went wrong, and maybe why.

CICO: tools, methods, procedures, techniques, avenues of inquiry

CIGuess: nothing. Zip. Zilch. Nada. Ne pas de zero.

Here's another practicality: the Atkins diet WAS TRIED - it was given the same chance, played on the same field, as any other diet that makes it into the public eye and it has fared no better.

It has isolated "success stories" but ALL diets do.

"NO body did it properly" is not an excuse: they did it as properly as they did any previous diet

For all of Taubes' demonization of the effectiveness of previous diets, where's the teeny admission that Atkins was tested the same way?

The silence on the fact that Atkins/Taubes failed miserably on the same playing field as most of what came before is proverbially deafening.

Sanjeev said...

none of this has any bearing on the correctness of the idea

I've heard no one propose that any governments or countries do anything with this

> Under a food reward doctrine, will we
> have labelling information about
> palatability or reward?

> Out of five stars, perhaps?

> Values determined by a panel of
> scientists, a panel of fatties, or a
> panel of mice with expressive faces,
> interpreted via the lastest
> algorhythms perhaps?

> a) will this tell us what we don't
> already know?

> b) will it tell us as much as the
> advertisers are already paid to tell
> us?

> c) what are we supposed to do with this
> information?

> d) given there is limited space on
> packaging, will this new information
> come at the expense of other data?

Sanjeev said...

Thanks for the reconsideration john. I had not expected it, thank you for being the rare person who surprises me these days.

I could have chosen a less hyper-sarcacastic tone, so apologies from me there.

We'll have to agree to disagree. From BBS to USEnet days to today I don't know why this issue still riles me the way it does.

I don't think anyone MUST count calories and I don't think people have as much conscious control over eating as the strict (mostly straw man) CICO folks

It MUST, MUST MUST be kept there at least in the background because it's the only universal troubleshooting framework that can be applied[0] to any specific technique

[0] in the worst cases, as post-mortem after a technique gone wrong.

John said...


People who "overeat" on low carb do not "believe" that CICO is incorrect. They are told that calories from fat do not get stored as fat; there's a difference.

The problem with the popular low carb community is that they think nothing beyond carbohydrate-->insulin-->fat storage, and the problem with the "CICO" community is that they have an implied belief that physiological processes play no role in fat storage--ie, once you eat enough, fat "just grows."

You can certainly make animals that have physiology that makes them almost unable to store fat, despite high calories, palatability, reward, etc. This doesn't violate any laws of thermodynamics, as idiot nutritionists seem to love to use as some kind of argument.

The reason there is no wideranging super successful diet is because any "diet" requires a certain amount of willpower; donuts and cupcakes taste good, are unhealthy, and are available. There is no magic bullet diet that removes all unhealthy food cravings and adjusts your appetitie perfectly. In general, not enough people are motivated, and the education is terrible--bad combination.

John said...


Regarding your most recent comment, I'm not knowledgeable about what is practical for fat people. I am lucky in that I have never had to consciously count calories and am just trying to learn and make continual improvements. Sometimes I think I see flaws in people's arguments, and with that in mind, I think itsthewooo and Petro of Hyperlipid have the most coherent explanation of fat gain. It is probably not complete however.

Stephan Guyenet said...

Hi George,

If I understand correctly, you're arguing that CIH is useful because of its practical value for creating dietary guidelines. Personally, I think it's possible to acknowledge the usefulness of LC for the treatment of obesity without saddling it with a bunch of pseudoscience. Although I've devoted several blog posts to it, Taubes's version of the CIH is actually remarkably easy to falsify. One of the easiest ways is to look at the relationship between fat mass and the release of fatty acids from fat tissue. CIH proposes that obesity is due to a reduction in the ability to release fatty acids from fat stores. What the evidence shows is the opposite: the more fat mass a person carries, the more fatty acids their fat tissue releases.

What this paper clearly shows is that fat is not 'locked' in fat tissue, by insulin or anything else, at any level of fat mass-- to the contrary. Lustig's version of the CIH is more difficult to falsify because Lustig has a much more sophisticated understanding of human physiology than Taubes and so he avoids the obvious pitfalls. However, his idea also doesn't line up well with the evidence as a whole.

The main point I want to make is that although CIH may be useful in the sense that it helps obese/overweight people come to the LC diet who may be really helped by it, it also has unfortunate consequences. It implies that lean, healthy people will eventually get sick from eating carbohydrate per se, which is utter nonsense. I personally ate LC for about 6 months after reading GCBC. Fortunately I snapped out of it pretty quickly once I began investigating his claims on my own and realized that they didn't make sense. I derived no benefit from eating that way, although to be fair it didn't hurt me in any obvious ways besides inconvenience and expense. But I get e-mails frequently from people who were lean and healthy, decided to go LC to protect their health, and ended up gaining fat and feeling like crap. Swapping fat for carbs, they typically lose the excess fat and feel better.

CIH is not the CAUSE of obesity, although it does lead to some practical conclusions that are useful for the TREATMENT of established obesity in some people. But we don't need to accept CIH to accept the practical value of LC diets.

Jane said...

Wow. Incredible discussion.

samreen said...

Amazing how simple it can be to communicate with people and have them understand a certain topic,
mumbai hospital

Puddleg said...

Good response Stephan; I agree that it is syllogistic to argue that, because LC cures diabesity in some, therefore the opposite of LC must be the cause.
(Though of course such effects must supply potentially useful hints).

My understanding of Taube's fat cell picture was a little more nuanced; that insulin forces the cells to retain fatty acids up to the point at which they become resistant, at which point the FAs start to spill over.

Of course the liver is also releasing more glucose from glycogen and this extra glucose maybe does not get enough credit as a contributor to IR.
(if fructose replaces liver glycogen without the intervention of insulin, then it's perhaps a good candidate for the initial cause of this glucose leak).

You are, I think, right to say that a lean, healthy person has nothing to gain by changing diet (beyond removing obvious toxins which increase the odds of future problems; replacing bread and HFCS with carbs from tubers, and seed oils with healthy fats, for example).

Doctors have a saying "I don't know what you've been doing right, but I'll advise you to keep doing it". Robert Atkins compared the efficacy of ketogenic dieting to that of hormones; you probably wouldn't take hormone supplements if you didn't need to.

There is a practical application of the food reward science; it is likely to produce insights into how appetite suppressant drugs and supplements can be improved.

Sanjeev, I was having trouble seeing the point of replacing a hypothesis which allows people to be advised about health and become knowledgable about metabolism, with one that doesn't lend itself to helpful advice and useful knowledge. People read these blogs, in part, because Stephan, in his cautious, take-it-or-leave-it way, is actually a brilliant health coach. Does food reward theory have implications that can lead to better practical advice, or is it an interesting but ultimately unhelpful academic sideline?
Are there readers who have been helped by applying food reward theory in their daily shop-cook-eat cycles, and how has this manifested itself?

Stephan Guyenet said...

Hi George,

Yes I've noticed that the latest incarnation of the CIH is that fat cells hang on to fat (presumably due to elevated insulin) until you've reached some level of fatness, and then they release it again. Since the proponents of this idea have yet to produce any evidence to support it, it strikes me as a rather ad hoc way of salvaging the CIH. If the idea is correct, one should be able to observe during fat gain that circulating free fatty acids decline, or that FFA release from fat tissue declines, and then rises again once a person's weight has stabilized, but this has never been observed to my knowledge, despite the many studies on FFA kinetics in lean and obese people.

Regarding the practical value of the food reward idea, that's an area where I'd like to see more studies. The literature as a whole is consistent with the idea that it's quite effective for fat loss in humans, but no weight loss studies have deliberately targeted reward/palatability to my knowledge.

But realistically, people apply the food reward idea all the time intuitively. Most people understand that industrial food promotes overeating and simple homemade food is a better way to preserve leanness and health. I've offered people practical guidelines for using the idea on my blog, and a number of people have used it successfully, including to break weight plateaus reached on LC and/or Paleo diets.

Sanjeev said...

click here

or copy and paste into your URL bar:

I still would track my eating behaviour (I did so when I was doing level 5) for real-time feedback on whether my appetite actually IS being controlled.

Stephan Guyenet said...

Hi Sanjeev,

I'm curious, what was your experience with level 5? Most people don't track calorie intake so this is interesting to me.

Sanjeev said...

I didn't expect several posts to show up before I clicked "publish"

"click here" in my comment above: takes you to recommendations and actions you could take.

Puddleg said...

I think of a poor person in a restaurant; they eat all the bread or rice because it is free, and will otherwise go to waste, not because it is flavoursome.
I simply can't agree that industrial food is more rewarding or palatable than the best home cooked meals. Maybe to some, but that's subjective. There are other reasons - such as ease of availability - why people might eat more of it. There are toxins that might be addictive or damage delicate controls (MSG, acrylamide, fructose).

The fat-storing cells are also fat-producing cells that are capable of converting glucose to fat. But if your liver is releasing extra glucose and lipids and your fat cells are releasing extra fat, you clearly have a hypercaloric situation. The body is juggling energy because there is too much to burn. Or because it is not set to burn the dominant form available (say, the fat) because the high sugar/insulin prevents this.

Puddleg said...

Or because the mitochondria are too few, or too obstructed, by any one of the hundreds of toxic agents, including pathogens, or micronutrient deficiences known to inhibit oxphos, damage mitochondrial DNA, or decrease mitochondrial membrane potential...

It is only possible to falsify the CIH because units exist for measuring insulin, glucose, free fatty acids, calories, and body weight.
What units currently exist for measuring palability or rewardiveness, that allow these to be quantified?

Stephan Guyenet said...

Hi George,

Palatability is often measured in humans using a visual analog scale or other related measures, and reward is measured by using progressive ratios and other related methods. Palatability and reward can also be quantified in animals, and this is done routinely (although palatability is harder to quantify in animals than humans). These are some of the methods that have been used countless times to quantify palatability and reward in the literature.

I have previously reviewed a number of these studies on this blog. There have been many, many opportunities for the FRH to be falsified, but it has received remarkably consistent support from the literature. It's not every day that you come across a hypothesis in the health/nutrition world that is so well supported by a variety of different fields of research.

Puddleg said...

Insulin, lipids, glucose are measured by detection and mass analysis of actual physical particles present in the blood stream. By counting or weighing something that is really there. That is actually physically present; an object.
"Human insulin is a peptide hormone composed of 51 amino acids and has a molecular weight of 5808 Da."

Whereas moderation is an abstract concept, and palatability is a kind of emotional recognition, from what I can gather.

As far as I know, no-one has falsified Bishop Berkley either.

FRH may be true; it certainly sounds plausible (and so simple that the analysis perhaps exceeds the phenomenon in complexity), and advice to avoid industrial food is always going to give good results.
But I feel as if the FRH and the CIH currently exist in different realms, and it is futile to use the evidence for one to try to contradict the other.
Gary Taubes, much as I love him, is a journalist who wrote a good book some years ago, and no doubt would correct some pages today. Rather than posing as an expert, he was trying to report the sayings of experts who impressed him. Shouldn't critics of CIH be engaging with those scientists now, rather than the journalist who tried to summarize their thinking back then?

Stephan Guyenet said...

Reward can be measured objectively and quantitatively by directly measuring how hard a person or animal is willing to work for something, and it is not an abstract concept.

Palatability is measured objectively in animals using stereotyped facial expressions. In humans it's measured subjectively by asking how good it tastes. Some people might think this is wishy washy. Why bother studying something subjective like that? By this logic, we would also stop studying pain, depression, PTSD, and most psychological disorders. Subjective does not mean that the tools of scientific inquiry can not be applied.

You said "it is futile to use the evidence for one to try to contradict the other". Absolutely. Gary Taubes tried to do this in his screed on food reward, but I have never used the argument that one being true weakens the other. Each hypothesis must be evaluated on its own terms independently of the other.

You also said "Rather than posing as an expert, he was trying to report the sayings of experts who impressed him." This is the opposite of what happened. Taubes has been trying as hard as possible to pose as an expert himself, even going so far as to publicly disagree with and insult a number of researchers with far more knowledge and experience than himself. He cherry picks a few experts whose views he can sort of cram into his own, but the rest he basically thinks are idiots and he has said as much.

Puddleg said...

What I liked about this study is that it appears to open a portal between the subjective palatability world and the concrete insulin world, with the mention of sweetness receptors (and receptors and their binding can be quantified - there's a future objective measure of palatability, perhaps). Receptors which also happen to be part of the insulin system.

It seems a not-unreasonable view that acquired metabolic diseases
are multi-factorial, and that what we need is a unified theory that shows how these factors might interact, e.g. at some critical junction like the Fox01 transcription factor (impacted by insulin, fructose, and mitochondrial ROS, which may come from pathogen disruption of OXPHOS as well as toxins and micronutrient deficiencies).

It might take a computer to make such a model; it may already be beyond human ability to do the data justice. It will certainly take computers to adapt the analysis to individual genomes in the future.
On the issues of saturated fats, statins, or whether red meat is bad for you many venerated experts are in fact idiots, and it does no harm to point this out.
We should perhaps cherry pick those experts whose methods we trust; this is how the Cochrane Collaboration works.

John said...

There is a problem with palatability and reward not being inherent food properties: the practical advice is extremely limited. Since they are based on behavior, people would simply have to eat less of their most frequently eaten foods, as it is implied that those foods are most rewarding and/or palatable [for most people]. But that offers nothing beyond the "character defect" hypothesis where the advice is, "Eat less."

Travis Culp said...

Might it be that this species of mouse seasonally fattens in the wild with sugary foods being a signal for it to commence?

If this is the case, it seems unlikely that it would carry over to humans, given that the bulk of our evolution occurred so close to the equator.

On the other hand, if we are influenced by such things, it may explain the fattening effect of diet soda.

Puddleg said...

That is in fact a theory; that fructose and PUFAs, being seasonal, are fertility regulators and especially potent growth signals. In the wild, fructose and vegetable PUFA are abundant in the time of plenty, and scarce when food in general is becoming scarce.
They also appear in the foods that don't keep so well (compared to roots and tubers in the ground and animals on the hoof or wing), so there's a practical point to eating more of them and storing the energy away.

Comfort for carb-lovers; from The Great Hunger by Cecil Woodham-Smith, on the Irish diet before the potato blight. Before the famine, Irish population growth was double that of England and Wales despite appalling poverty and a lack of medical facilities:

"There was an abundant supply of incredibly cheap food, easily obtained, in the potato, and the standard of living of the time was such that a diet of potatos was no great hardship. With the addition of milk or buttermilk potatos form a scientifically satisfactory diet, as the physique of the pre-famine Irish proved.
The potato was moreover, the most universally useful of crops. Pigs, cattle and fowls could be reared on it, using the tubers which were too small for everyday use; it was simple to cook; it produced fine children; as a diet it did not pall."
As well as milk, occasional ham, lard, and bacon (from the pigs that were surely not kept as pets), eggs, and cheese would have supplemented this diet - but to a very small extent, if the diets of better-off Enlish labourers of the period are any guide.

We don't need to look for Kitavans or coconuts and other superfoods to find people doing well on high-carb diets; as long as grains and sugar are largely absent from the diet, and some whole animal food is present, people seem to do well.
Incidentally, growing potatoes was not hard work, peat fuel was free and cabins warm, unemployment ran to 75%, so we cannot put everything down to exercise.

Jane said...

George, I am greatly impressed by the extent of your knowledge. At least, I was until you said high-carb diets are OK as long as grains are largely absent. Haven't you read about McCarrison's work? There is no doubt at all, despite what some people will try to tell you (not Stephan), that grain-based diets can produce excellent health.

Jane said...

George, I imagine you are thinking of what happened in Iran, where wholemeal bread produced some teenagers with short stature and late puberty.

As I understand it, and I can't find the article I read about it now, this was a new phenomenon which appeared during the green revolution when Iranian farmers tried to grow high-yielding crops on mineral-poor soil. In villages where the bread was leavened, it didn't happen. It apparently needed poor soil, unsuitable crops AND unleavened bread.

Unleavened wholemeal bread was eaten by the people of northern India studied by McCarrison 100 years ago, and it caused no problems. Nor did it cause any problems in the rats he fed on a northern Indian diet. They were remarkably healthy, like the people. Other rats fed the white-rice-based diet of southern India were extremely unhealthy.

LeonRover said...

"Other rats fed the white-rice-based diet of southern India were extremely unhealthy. "

And what about the S Indian population eating like "other rats"?

Jane said...

LeonRover, here's a couple of quotes from a 1924 McCarrison paper on beri-beri.

'But while it is not true of endemic areas of the Madras Presidency that beri-beri never appears when under-milled rice only is used, it is true of non-endemic areas. In the former I have found the disease to be nine times more common where the use of polished rice is fairly general than in country districts where the use of home-pounded rice is general. In the latter the disease is very rare or unknown.'

'Among the 800 men there occurred thirty cases of beri-beri, or
a case incidence of 3'7 per cent. The outbreak, which was of a cardiac type, was promptly cut short when a proportion of the rice was replaced by atta (whole wheat flour).'

McCarrison says the pathology of beri-beri includes 'enlarged and degenerated heart, enlarged liver, enlarged adrenals, polyneuritis, and hydropericardium with or without other evidence of oedema.'

The degenerated heart is interesting in view of what Cleave says in The Saccharine Disease about the incidence of heart disease in India being 7 times higher in the south than in the north, despite a far higher consumption of saturated fat in the north.

Puddleg said...

Well rice may be a special case. I am interested by the fact that potatoes produced a "healthy physique" and "fine children" compared to the grains of europe. Potatoes are about 4% protein and 40% carb (today), whereas wheat is about 11% protein. But perhaps rice and potato are similar in that the protein, though low, is complete protein. And the potato needs no milling to be edible, so micronutrients aren't lost. And the Irish diet would have been high in ascorbate by European standards.
By 1840 the sugar/tea/jam combination was beginning to made inroads into English labourer diets, at the cost of some protein and fat; it does not thus appear in Ireland, which did not have the kind of cash-and-carry economy to support a widespread trade in imported foods.
When Indian corn was imported into Ireland during the famine, the country lacked the mills to grind it into flour due to the prior lack of demand for bread.

One of the very earliest diet-health documents is a treatise on the health effects of grain avoidance, found in a Han tomb in China and dating from the 3rd century BC.
There may still be healthy grains out there somewhere, but that is no longer a quest I want to waste any more of my health pursuing.

Jane said...

George, I looked up 'China health effects of grain avoidance' and found this in Wikipedia under 'Bigu (avoiding grains)'.

One of the striking things about the texts we have reviewed is that most of them offer very little by way of an internalist critique of grains or other everyday foods. That is, they all recommend avoiding grains and offer what they tout as superior alternatives, but on the question of precisely why grains are such inferior nourishment they have little or nothing to say. What little internalist critique we do find comes quite late — apparently Eastern Han at the earliest — and does not seem well developed: ordinary foods, described as rotten and smelly, impurify a body that must be brought into qi-based resonance with heaven. This impurity is located specifically in the intestines. […] In most discussions, then, it is not that prescribers and practitioners of transcendence arts portrayed ordinary food as harmful; it is rather that they had what they considered superior alternatives to offer. [… But,] why these diets of qi or of rare herbs and minerals should be regarded as superior to one of ordinary food is a question that very often remains unanswered; we are merely, but repeatedly and in diverse ways, told that they are superior.

BigWhiskey said...


Puddleg said...

Thanks Jane, I was wondering how that reference (in the introduction to the Tao Te Ching) could be taken further.
Bigu is also seen as fasting
and probably has more in common with calorie restriction than with paleo.

"There is no history of Bigu being used for weight loss in ancient China. The reason might be that there were not as large a percentage obese people as in modern times. People did more physical work and ate less fattening food. Further, being overweight was considered a sign of wealth. Recently, Bigu has become popular for weight loss (4), (5); (6). You don’t have to worry about counting calories, choosing and preparing food or do strenuous exercise.

Bigu should be investigated as a possible treatment for diabetes. One participant, described in (6), was in Bigu for forty days. She was able to decrease the amount of insulin as well as the number of injections per day. Some days she was down to one injection while teaching, performing ballet, coaching gymnastics and swimming."

They go on to compare Bigu to Dean Ornish's diet. Which I kind of assumed was wholegrain based.

The Wikipedia article is fascinating. It can be seen as a history of the Neolithic revolution preserved in traditions.

"By dispensing with starches a man can only stop spending money on grains, but by that alone he cannot attain Fullness of Life. When I inquired of people who had been doing without starches for a long time, they replied that they were in better health than when they were eating starches. When they took thistle and nibbled mercury and when they also took pills of brown hematite twice a day, this triple medication produced an increase in breaths, so that they gained the strength to carry loads on long trips, for their bodies became extremely light in weight. One such full treatment protected the patients' inner organs for five to ten years, but when they swallowed their breaths, took amulets, or drank brine, only loss of appetite resulted, and they did not have the strength for hard work. The Taoist writings may say that if one wishes Fullness of Life the intestines must be clean, and if immortality is desired the intestines must be without feces; but they also say that those eating greens will be good walkers, but at the same time stupid; that those eating meat will be very strong, and also brave. Those eating starches will be wise, but they will not live to an old age, while those eating breath will have gods and spirits within them that never die. This last, however, is only a biased claim advanced by the school that teaches the circulation of breaths. (15, tr. Ware 1966:243-4)

Jane said...

Yes, it's more like calorie restriction. The goal is to increase the energy flow, the qi or chi. Qi is not a figment of ancient Chinese imagination, it's real. It was studied by Robert Becker, twice nominated for the Nobel Prize, and appears to be composed of electrons and protons which flow through the extracellular matrix (see 'Charge transfer in the living matrix', Oschman 2009) and along membranes, DNA, and the cytoskeleton.

The other goal of fasting was of course to increase autophagy. The physical structure of the body must be as close to perfect as possible to encourage the energy flow. So the junk must go.

There are three ways to increase autophagy: fasting, exercise and proper mineral balance, which means a high ratio of Mn to Fe.

Mn2+ is extremely magnetic. It is my belief that the magnetic fields which direct the energy flows are dependent on Mn2+, which is known to bind to extracellular matrix, cytoskeleton, membranes and DNA.

Puddleg said...

Magnetism and electricity are examples of variable energy outputs that aren't usually measured in diet-exercise studies. Maybe not significant ones, but who knows in every case?

Look at this:

Sci. Signal., 28 February 2012
Vol. 5, Issue 213, p. ec64
[DOI: 10.1126/scisignal.2002999]


How Sweet It Is

Wei Wong

Science Signaling, AAAS, Washington, DC 20005, USA

Fructose is the sweetest natural sugar. It is detected by a heterodimer of the sweet taste G protein–coupled receptors T1R2 and T1R3, which are present not only on the tongue but also in other organs, including the intestine and the pancreas. Kyriazis et al. found that fructose potentiated insulin release from isolated mouse islets when added in combination with a concentration of glucose expected to induce insulin release (8.3 mM) but not with a concentration that would not be expected to induce insulin release (3 mM). In β cells from wild-type mice, the increase in intracellular calcium concentration (which is required for insulin release) elicited by 8.3 mM glucose was potentiated by fructose, an effect that was not seen in β cells from T1R2–/– mice. In islets isolated from wild-type or T1R2–/– mice, increasing the glucose concentration from 8.3 to 16.7 mM induced a substantial increase in insulin release (a phenomenon called glucose-stimulated insulin secretion). This increase was further enhanced by adding fructose in combination with the higher glucose concentration to islets from wild-type mice but not those from T1R2–/– mice. Similarly, when administered in vivo, fructose potentiated glucose-induced insulin secretion in wild-type mice but not in T1R2–/– mice. The ability of fructose to potentiate glucose-induced increases in intracellular calcium required phospholipase C (PLC) β2 and the cation channel TRPM5. Fructose potentiated glucose-stimulated insulin secretion in human islets, an effect that was blocked by pharmacological inhibition of T1R3. The authors note that sweet taste receptors in the intestine stimulate the absorption of ingested glucose and the release of glucagon-like peptide, a hormone that promotes glucose-induced insulin secretion. They speculate that sweet taste receptors in the intestine and pancreas could be activated by dietary fructose, including that found in high-fructose corn syrup, a sweetener widely used by food manufacturers that has been controversially linked to metabolic diseases.

G. A. Kyriazis, M. M. Soundarapandian, B. Tyrberg, Sweet taste receptor signaling in beta cells mediates fructose-induced potentiation of glucose-stimulated insulin secretion. Proc. Natl. Acad. Sci. U.S.A. 109, E524–E532 (2012)

Search Engine Optimization said...

The knowledge you provide is a real asset we have. Keep on posting such valuable articles. I like your blog design as well.
Drug rehab centers oklahoma

Unknown said...

Nice post Christopher

William Eden said...


You posted this article:

in support of the idea that the fatter someone gets, the more FFA is in circulation, so therefore insulin couldn't be responsible for locking fat inside of fat cells in obese people.

This is true - however, if you read the paper it is pointing out that FFA per kilogram falls logarithmically with increasing fat mass! Thus each individual fat cell is releasing much less fat in obese people than lean people, though this is not enough to make up for the enormous amount of fat they are carrying. The difference between 10 kg and 20 kg of fat is enormous, though it levels out quite quickly.

This particular paper actually seems to support the insulin hypothesis of obesity, though much of your other work and citations don't line up well.

Thanks for this fascinating discussion!


Stephan Guyenet said...

Hi Will,

Yes, you are correct about that paper, and kudos for taking a close look at it. FFA Ra per unit fat mass decreases with increasing obesity. However, think about it this way. What really determines whether your fat mass is increasing or decreasing is the total amount of fat going into vs out of fat cells, not the amount per unit fat mass. So in the obese, the total amount of fat exiting fat tissue is greater than the total amount exiting a lean person's fat tissue.

Therefore, suppression of fatty acid release due to elevated insulin cannot explain the development or maintenance of obesity, because total fatty acid release increases in parallel with body fat accumulation.

jo said...

Hi Stephan,

I'm loving your blog. It's so fascinating to finally be getting some clear answers about the whole weight loss conundrum.
I was wondering though - about artificial sweeteners - as I do use them alot.
They have a sweet taste so would they promote over eating and reduced fat burning?
Also - was wondering about Seth Robert's sugar water idea - as he said that the sweet taste caused him to lose weight - as long as it wasn't associated with a flavour.

your thoughts would be greatly appreciated!

jo said...

Hi again Stephan!

This whole new paradigm has got me wondering how observed facts fit in with the theory and how a better understanding could help us make better food choices.

I was wondering about protein - as there are alot of studies which seem to show protein to be more filling.
I was wondering if that was due to protein having inherent filling properties or whether that may be due to a slower calorie signal (to use Seth Robert's terminology)i.e. a lower GI.
Also - I was wondering if low carb eating doesn't work that well long term (I think)because the body eventually cottons on that higher fat food does result in higher calories (despite the calorie signal being spread over a longer period of time due to it's lower GI).
So I was wondering if those studies about eggs being more satisfying than carbs , for example, may only be true in the short term, as the body eventually cottons on that eggs are associated with more calories ( due to a higher energy density) and in the long term, eating a lot of eggs might not actually be that helpful for weight loss.
Do you think that in the long term, energy density may be the most useful indicator of how fattening a food is, as the greater volume and weight of low energy dense foods counters the fattening-ness of calories ( whether low or high GI).
And lastly - what do you think of the validity of varying the flavour mix each time you eat something (so that no strong flavour/calorie association forms).
This could be more fun than eating bland.
Sorry to bug with you questions - would love to know your thoughts to get a bit more clarity on it.

Puddleg said...

of course, high-fat food only means more calories if you actually eat more calories from a high-fat diet than you did from whatever other diet you eat.
Imagine a food that gave a hundred calories per gram, astronauts could eat it in pill form, they would not necessarily be obliged to overeat.
Studies that compare diets show lower calorie intake by high-fat dieters over time.
Calorie-dense foods (such as cheese) can be some of the most nourishing and least fattening, while some fattening foods such as sodas have quite low caloric density.

Meegs said...

" In other words, X number of calories that are highly palatable will be more likely to be stored as fat than X calories that are less palatable, at least in the context of excess energy intake (it may not matter if you don't overeat in the first place, it's unclear). That's one of the key findings of Sclafani's paper. I may discuss it in more detail at some point on the blog, but it would be somewhat technical."

I think this is the part that is confusing people.

So if you eat less palatable food you're more likely to exercise?