A spark of reason

I’ve been reading through many (not all) of the comments on Guyenet’s rebuttal of the carbohydrate hypothesis. Although it can be rather tedious, especially because of the apparent mental hospital escapee ItsTheWoo2 (who does have a few good points but a complete lack of putting those point out there in a sane way), but Stephan finally got to his senses and blocked the comments from her (him?).

After reading both the blog post and the comments I find that I am rather unsure of what we are discussing. It seems we are dealing with to poorly defined theories and that much of the seemingly contradictory nature of those theories stem from a lack of clear definition.

The only comment I felt was really worth noting was that of Dave “Spark of reason” Dixon who had this to say:

Just to clarify (and apologies if already discussed, I haven’t gone through all of the comments): what precisely is the «carbohydrate hypothesis» under discussion? Is it that carbohydrate is *necessary* for obesity, or *sufficient*? My reading of Taubes is the former, that you need carbohydrate to drive fat storage, but not that just eating carbohydrates will make you fat.
I don’t find anything in what you presented which is inconsistent with that view, e.g. examples of lean cultures with traditionally high-carbohydrate diets may simply indicate that other factors are needed to drive obesity, and particularly metabolic syndrome. Are there any examples of obesity occurring in the absence of dietary carbohydrate?

Isn’t this actually an important point? Does the carbohydrate hypothesis being discussed say that carbohydrates are necessary or sufficient for causing obesity? If, as Dave interprets GCBC, the theory says carbohydrates are necessary, then what exactly are people arguing about?

I don’ get it!

Macronutrients and food reward

If you see one bright red swan, you are not likely to give up a theory that says that all swans are white; you will instead go looking for the person who painted it. 

Imre Lakatos 

Much is being said on this subject. Bet many are getting pretty fed up by now. But I still think this is an interesting discussion and so I will take this opportunity to add some thoughts. After all, the goal here is to find the truth; to find out how the world works. In that respect, I would also like to say that I do not agree with any one side in this discussion. Scientifically speaking, agreeing is not very scientific. That would mean confusing matters of opinion with matters of fact. Things are just what they are.

Chris Kresser is of course right in that there is no single cause of obesity. In animal studies obesity can be induced in a number of ways, just as in humans. The fat tissue is a large part of our body and it has a wide range of receptors and interesting signaling, so it should not come as a surprise that there are many ways to become fat.

If we are to look for a general cause, we could say that western, post agriculture lifestyle is to blame for obesity and our lifestyle diseases. But that does not mean you cannot get fat eating paleolithic foods, although if you did, you should blame your parents for the lousy chromosomes.

Neither carbohydrates nor food reward is able to explain all the observations. They both explain a part of the observations and so are both likely influential factors. Just how big a role each plays is an extremely difficult question to answer. Thus the current discussion.

The key question is: Why is not hunger down-regulated in humans becoming fat?

The Guru Walla
From what I can see, the Cameroonian Guru Walla is a bland food, overeating, fat gaining rite.

In the Guru Walla ritual, young Cameroonian men consume a diet made of red sorghum and cow milk (makes up over 95% of calories). The young men isolate themselves in different houses with a female attendant devoted exclusively to the preparation of Guru Walla meals. The diet and exclusion is supposed to lead to a certain level of purity. The men eat every 3 hour for 60 days, during which time body-weight can increase by an average of 17kg. Only 64-75% of the weight gain is fat [1].

Traditional food amongst these Cameroonians is about 75% CHO, 10% fat and 15% protein. During the Guru Walla it is 70% CHO, 15% fat and 15% protein.

The Guru Walla food is obviously fattening; at least if force fed and combined with minimal physical activity. The question we need to ask is: Would the Cameroonians be overweight if all they consumed was the Guru Walla diet?

It seems that the Cameroonians do not get fat because of the food itself. Rather they become fat because they force feed themselves. The newly gained weight is also lost after the ritual.

The reason I called the Guru Walla food bland, is that it most likely is very bland after the first few days. Try eating any one food exclusively for 2 months, and eat it even though you are not hungry (vomiting is also a part of the Guru Walla). The dopamine reward response should be minimal. Remember the Twinkie Professor who ate nothing but Twinkies, Oreos, donuts and similar crap but who lost 27 pounds in a 10 week experiment. He did consciously under eat, but my point is that we need to ask ourselves how lack of variation affects reward.

Food reward
While food reward might help explain why we overeat at a biochemical level, there is little evidence to indicate that a fat loss diet needs to be unrewarding, if by unrewarding we mean less palatable. We also need to know if it is possible to unconsciously overeat (become fat) on rewarding foods if we have a working metabolism and the rewarding foods do not mess with our metabolism. If not, whatever caused the metabolism to go out of whack is the real problem.

Stephan’s bland food through a straw experiment does not necessarily support a theory claiming that the study participants lost weight because of an inherently unrewarding quality of that particular diet. The finding could easily also support the theory that eating only one food, no matter how rewarding it may be when consumed intermittently, will make people lose weight because the rewarding quality of that food declines with increasing intake.

So we need to know if people could lose similar amounts of weight eating other foods exclusively.

I am still having trouble seeing what’s the big fuzz about leptin. It is a signaling molecule. It signals energy surplus and the lack of leptin signals lack of energy. Leptin also increase fat oxidation. The leptin deficient animal models, that are obese, act and behave as they were starving and administering leptin normalize their behavior and induce weight loss. Either the body just needed to be told that it had stored energy to use, or we just needed to increase fat oxidation. If you increase fat oxidation by other means such as GH, ob/ob mice lose weight just as with leptin.

If for example high insulin levels cause leptin resistance, focusing on leptin does not add anything to obesity treatment. High insulin levels can also be caused (or at least be exacerbated) by factors other than carbohydrates. For example factors that messes with liver function.

“In particular, protein-rich foods such as beef can increase insulin secretion as much as certain starch foods such as pasta, or more.” 

The quote takes the results from trials out of context. It is an unfair statement, just like “proteins are inherently satiating” statement. A few days of beef eating will likely lead to lower insulin than a few days of pasta. I’ve written some about satiating proteins here.

In overweight people, as in overweight animal models, the key issue seems to be a reduced fat oxidation. Reduced fat oxidation with a high energy intake cause fat deposition in most all tissues and also insulin resistance.

Anything that increase fat oxidation in overweight animal or humans, cause weight loss and reduced food intake.

Lessons from insulin resistance
Stephan claims that overweight people have high serum free fatty acids. This is not completely true, at least if we are to listen to Keith Frayn at [2]. The claim may be true in general, but there are lots of overweight people with normal FFA levels. This however does not change Stephan’s argument. Generally the fat tissues of the overweight give out more FFA indicating adipose tissue insulin resistance.

Here is how we imagine insulin resistance to occur (roughly):

The pancreas has a direct route to the liver. The reason for this direct route is that the liver controls blood sugar level through its production of glucose. When blood sugar rises, the pancreas notice and secretes insulin. When the liver receives this insulin, glucose production is reduced. As the cells in the body are utilizing glucose for fuel, blood glucose level drop.

Somehow the liver becomes insulin resistant and keeps sending out glucose despite the insulin being sent from the pancreas. The reason seems to be inflammation and/or buildup of fat (NAFL). In this insulin resistant state, the muscles also fill up with fat. Once glycogen stores are full they become insulin resistant to avoid sugar poisoning, but keep taking up fatty acids. Because of the high carb diet and/or lack of physical activity the muscles do not burn fat and so it builds up. Also, there is some loss of muscle and liver mitochondria function and probably fatty acid transport into mitochondria.

The fat tissue takes up both glucose and fatty acids and expands if it takes up more than it gives out. The expansion of fat tissue eventually cause fat cells to send out stress signals (probably caused by endoplasmatic reticulum stress) and macrophages invade the tissue, gathering around dying fat cells. In this state, the fat tissue secretes a lot of fatty acids that wreak havoc around the body. But if free fatty acids are not burned they need to be re-esterefied. A high FFA level does not mean that we are not gaining weight or that we are losing weight (that more fat is leaving than entering the fat tissue). FFA are measured fasting and although the level might be higher in overweight and insulin resistant in that fasted state, this does not mean that over time more fat is leaving the fat tissue than are entering.

Stephan Guyenet takes the high FFA-level often observed in the overweight to mean that the fat tissue is insulin resistant and that they could not be gaining weight. This might be a wrong assumption. They have definitely been gaining weight and most overweight people are either weight stable or gaining weight. Is it impossible to gain weight while still having high FFA level?

Lean people also get insulin resistant. As do animals and humans with lipodystrophies. Many massively overweight do not become insulin resistant, and it seems that what causes the overflow of free fatty acids from adipose tissue is that it reaches its limit – a limit of course determined by both genetics and lifestyle.

In the insulin resistant state (metabolic syndrome), free fatty acids are usually high and fat builds up everywhere. Anything that increases fat oxidation helps. Pharmacological inhibition of the oxidation of fatty acids in the liver stimulates food intake in both humans and rats and stimulation of hepatic fatty acid oxidation reduces food intake, weight gain and adiposity in rats with diet-induced obesity [3].

FFA’s come from food, the liver or fat tissue. Carbohydrates are largely responsible for the amount secreted by the liver. At a cellular level, insulin resistance/metabolic syndrome seem to come from a high total energy intake. There is a surplus of both glucose (glycogen) and fat and the body can’t handle it all. Reducing the dietary fat load helps (at least if hypocaloric), but reducing dietary carbohydrate is the most efficient treatment to date. The question, though, is still why these people overeat.


“…for insulin to cause fat gain, it must either increase energy intake, decrease energy expenditure, or both.” 

“If calories and protein are kept the same, high-carbohydrate meals cause equal or greater satiety than high-fat meals, and equal or less subsequent food intake, despite a much larger insulin response)” 

Stephan Guyenet

Insulin will reduce hunger as long as there is energy coming from ingested food. Once that flow of energy stops or is reduced, a high insulin level cause hunger. In order for insulin to cause overweight, the level only needs to be high enough for allowing fat oxidation to be less than fat storage in that particular individual over time.

Injecting both glucose and insulin reduce hunger. Injecting insulin alone increase hunger. Long term satiety is better with low carbohydrate diets than high. We need to remember that we adapt to burning different fuels. If we normally eat high carb and suddenly eat high fat we are likely to be poor fat burners and thus more likely to get hungry. This might also explain higher leptin levels after high fat meals in acute feeding studies.

“If blood glucose decreases enough, it activates a system called the «counter-regulatory response», designed to maintain blood glucose at all costs to protect the brain from the effects of hypoglycemia. Part of this response is hunger and increased food intake. However, this system is not activated except in severe hypoglycemia, which is rare except in diabetics, thus it is not relevant to common obesity.” 

This quote seriously needs references. It seems very unlikely.

These are just some thoughts. Nothing more.


1. Pasquet P, Brigant L, Froment A, Koppert GA, Bard D, de G, I, Apfelbaum M: Massive overfeeding and energy balance in men: the Guru Walla model. Am J Clin Nutr 1992, 56: 483-490.

2. Taubes G: Insulin resistance. Prosperity’s plague. Science 2009, 325: 256-260.

3. Ji H, Friedman MI: Reduced capacity for fatty acid oxidation in rats with inherited susceptibility to diet-induced obesity. Metabolism 2007, 56: 1124-1130.

Food reward, a factor in obesity

 “I think food reward offers the most compelling explanation for the US/global obesity epidemic.” 

Stephan Guynet 
In studies where the food intake and energy expenditure of subjects are carefully monitored over a period of weeks to months (which tends to average out day-to-day fluctuations) a remarkable balance between calories consumed and calories burned was observed. When various mammals, from mice to monkeys, are either overfed or starved for a few weeks, their weight soon returns to normal levels when free access to food is resumed. Crucially, our mammalian bodies seem to be able to regulate feeding based on the amount of energy available in the food we consume, not just on the volume of that food.

The above quote David Linden, suggests the body controls body weight by registering the amount of energy in food. This theorizing usually leads to the white adipose tissue derived hormone leptin and particularly its effect on the hypothalamus. Leptin, in general, correlate with adipose tissue mass. But the energy the body needs comes from two places: food and stored energy. Increasing use of stored energy will make animals and human eat less. Any energy-sensing control system must register the total amount of energy available, both from foods and from fat and glycogen stores. From this point of view, obesity could to be caused by the body not having access to its own stored energy and so continues to signal for food intake. Alternatively, the stored energy is readily available, but somehow an obese person experience feelings of hunger and craving that overpower any signal telling the brain that there is enough energy available. This scenario makes it likely that obesity is caused by a malfunctioning control system.

Stephan Guyenet has created quite a stir lately with his posts about food reward. Several of the posts have over 100 comments, some more rational than others, but people clearly have strong feelings about this. I think one of the reasons some people feel personally offended by his writing, is that they feel food reward lends support to overweight being caused by lack of willpower. This is definitely not Stephan’s intent, nor does his reasoning indicate willpower as a major factor. Nevertheless, willpower is a major part of food reward, as one of the opposing forces to a physiological drive to consume rewarding foods.

The theory of food reward is a theory of how foods affect our feelings, stimulate our behavior and how some foods appear addictive and promote addictive behavior. This seems lost on many. Food reward does not and cannot explain how we fatten. To find the answer to this we need to look at the physiology of the fat cells. Basic biochemistry still applies and some foods are more fattening than others, although as it seems, Stephan does not think so and he even uses his belief that macronutrients doesn’t matter as an argument in favor of the food reward theory. But the theory itself is a theory of why we (over)eat, not a theory of how we gain fat.

Even though the theory of food reward is not about willpower, willpower invariably enters into the equation. Many physiological drives can be affected by conscious thought. Stick your hand in ice water and your head tells you to pull it out (or your spine), but you can keep it submerged by willpower, some longer than others. Highly rewarding foods do, according to Guyenet, cause obesity in susceptible people, just like drugs may make addicts of some (often the same people). Still, I cannot see that food reward argues more strongly for willpower as a part of obesity, than other rational theories.

The theory of food reward is actually mostly about carbohydrates as most of the data relating to it is from the studying of sweet foods. As Hans-Rudolf Berthoud put it 

For nonsweet palatable foods (typically high-fat foods), there is less convincing evidence for development of dependence…” [1]
So it seems the key questions are:

– do sweet foods cause addiction and increased energy intake with subsequent obesity?

– do sweet foods cause obesity (fat storage) with following addiction or addiction like behavior possibly caused by metabolic clues?

As we undoubtedly fatten differently and not everyone becomes obese despite similar obesogenic environments, we can conclude that addiction to high palatability foods is 1) genetic and that preexisting differences in reward functions cause obesity; 2) intake of palatable foods is in itself addictive and leads to obesity; or 3) obesity (the excess storage of energy in fat tissue) cause changes in reward functions thus further accelerating obesity.

As many lean people also eat large amounts of highly rewarding foods, it seems unlikely that the food itself can be to blame. So, either food reward is secondary to the harmful effects of sugars/grains (sweet food not found in hunter gatherers): these foods create excessive fat storage in obesity prone people and this cause addictive behavior towards the very same foods; or it is the primary cause of obesity: people prone to weight gain have physiological measurable differences in parts of the brain that cause an addictive intake of fattening foods.

Although I enjoyed the posts about food reward I was left with very many unanswered questions after reading:

The reasoning 
Stephan uses the fact that hunter-gatherers are lean in support of rewarding foods causing obesity in non HG societies, arguing that one of the reasons hunter gatherers are lean is because their diets are bland (although I think many HG’s would disapprove of their diets being called bland). This argument could go both ways. Because if the foods that drive fat gain also promote addictive intake of the same foods, then traditional diets can be as tasty as any, just as long as they do not contain these particular foods. As long as they don’t, there is no reason to think blandness is the cause of leanness.

Although lean traditional people’s diets are more unrewarding then say a SAD diet, this does not mean that we in the west become obese because our foods are not.

Also we have to ask: if obese people remove fattening foods, which are the same as those considered highly rewarding, will the addictive behavior/strong cravings for the fattening foods subside? I know from experience that many who struggle with strong cravings, lose their cravings when switching to a LCHF diet. The fact that some feel cravings even after some time on low carb diets, does not favor a set-point hypothesis. It could just indicate a dietary insufficiency, like the lack of salts or some fatty acids. As the cravings usually disappear before a considerable weight is lost, it is unlikely that the cravings were caused by the obesity itself. Often, it seems that cravings disappear when people regain the ability to burn fat.

Burning fat, or having a functional metabolism will make us eat less. The oxidation of fat in the liver offers a strong satiety signal [2]. So, even if lipolysis is high in obese, hunger will not go down if somehow the burning of fat in the liver is restricted. This is sort of a “metabolism argument”: One of the things that separate those prone to obesity and insulin resistance from the rest, is a poor and broken metabolism. They rely on glucose (glycogen) for fuel and have poor fatty acid oxidation in combination with blood sugar fluctuations and cravings, so fat is stored rather than burned as it should. Resolving the metabolism issues will in many reduce the cravings and rewarding foods are no longer an issue.

Another important question to ask is: how often during the day and how much hyperpalatable, highly rewarding foods do people who become obese actually eat?

If people become obese without consuming highly rewarding foods (something I consider very possible) then the theory of food reward argues strongly that this type of obesity is mostly due to lack of willpower, as there is no addiction to blame.

The “bland food” study from 1965 Stephan writes of can be used to support a “food reward” theory, but there are many other ways of explaining why the overweight people lost weight while the lean did, not. If the obese ate high sugar/grain and franken-fat diets, that also happen to be palatable once you get used to it, then of course they lost weight on the liquid diet.

The first volunteer continued eating bland food from the machine for a total of 70 days, losing approximately 70 pounds. After that, he was sent home with the formula and instructed to drink 400 calories of it per day, which he did for an additional 185 days, after which his total weight loss was 200 lbs. The investigators remarked that «during all this time weight was steadily lost and the patient never complained of hunger or gastrointestinal discomfort.» This is truly a starvation-level calorie intake, and to eat it continually for 255 days without hunger suggests that something rather interesting was happening in this man’s body.

This isn’t really that interesting. With all likelihood the man could have lost an equal amount of weight eating real foods that are far more rewarding but not fattening. It has been known to happen.

I think decreased fasting insulin occurs as a result of weight loss…

Stephan Guyenet 

Another important point is that the body fat “setpoint” is still a theoretical point, and any theory based on the setpoint hypothesis is equally hypothetical. 

As one would expect if food reward influences the body fat setpoint, lean volunteers maintained starting weight and a normal calorie intake, while their obese counterparts rapidly lost a massive amount of fat and reduced calorie intake dramatically without hunger. This suggests that obesity is not entirely due to a «broken» metabolism (although that may still contribute), but also at least in part to a heightened sensitivity to food reward in susceptible people. This also implies that obesity may not be a disorder, but rather a normal response to the prevailing dietary environment in affluent nations.

Lean people have good access to their own body fat and high fat oxidation rates. They have a better working liver than obese, and they definitely had a better pre experiment diet than the obese. The above results can be explained exclusively by a broken metabolism theory. There is no need to involve food reward.

Some people may be inclined to think «well, if food tastes bad, you eat less of it; so what!» Although that may be true to some extent, I don’t think it can explain the fact that bland diets affect the calorie intake of lean and obese people differently.

Most diets affect lean and obese differently. These people are per definition quite different metabolism wise, and foods affect metabolism. Once again, the fact that one of the many diets that affect lean and obese differently are bland, does not lend much evidence for palatability playing a major part in obesity.

Although the rewarding abilities of different foods might explain some of the reason we overeat on fattening foods there are very many other ways you are likely to gain weight. As David Pier points out in the comments section:

Excess fructose? Too high an omega-6/omega-3 ratio? Too much omega-6? Too little omega-3? Too much polyunsaturated fat in general? Too little saturated fat? Micronutrient (choline, minerals, etc.) deficiencies? Excess total carbohydrate? Superstimulating hyperpalatibility? Over-availability? Excess insulin (cause and/or effect)? Gut flora (cause and/or effect)? Lack of fiber (insoluble and/or soluble)? Multi-generational epigenetic changes? Artificial sweeteners? Endocrine disruptors? Sleep disturbances? Psychological causes essentially independent of all hormonal homeostatic mechanisms?

In his third post, Guyenet writes about the review of low fat non energy restricted diets where overweight lost more weight than lean:

In other words, low-fat groups reduced their calorie intake by an average of 271 calories per day, and lost 7.5 pounds (3.2 kg). When they considered only people who started off overweight, they lost 12.8 pounds (5.8 kg). The investigators noted that the results were similar no matter what the duration of the trial, because weight loss plateaued fairly quickly.

Then he writes

This is all without any instruction to reduce calorie intake, therefore we can assume these dieters were eating to fullness.

No you can’t assume that. These are participants included in non blinded weight loss trials. I would say it’s a safer bet that they were in fact restricting their food intake.

The best low-carbohydrate diet study I’ve seen was published in 2008 in the New England Journal of Medicine (3). 322 «moderately obese» participants were placed on a low-carbohydrate diet, a calorie-restricted low-fat diet, or a Mediterranean diet, for two years. The low-carbohydrate group’s carbohydrate intake decreased by 130 grams per day, which is about half of a typical person’s total intake, and neatly corresponds to the reduction in calories of 561 per day, despite not being instructed to reduce calorie intake.

At two years, the low-carbohydrate group had lost 10.4 lbs (4.7 kg), which is very similar to the average weight loss seen in low-fat diet trials.

There are two major issues here. 1) The study by Shai et al is a horrible study: The Atkins based diet came with recommendations of getting fat from vegetable sources; By 24 months, carbohydrates constituted 40% in the low carb group and 50% in the low fat group. The low fat diet went from baseline fat intake of 31,4% to 30% (no reduction at all); the aurhors left out baseline energy intakes and only reported reductions; The study also used intention to treat analyses. The weight loss in the low carb group for the 272 who completed the study was 5.5kg in the “low carb” and 3.3kg in the “low fat” group. After 6 months the study diets were not very dissimilar.

If this is the best low-carbohydrate study Guyenet has read, he needs to read the other studies. Low-carbohydrate diets usually outperform low fat diets, as long as carbohydrate intake is kept restricted. This outperformance is despite low fat groups having caloric restrictions while low carb groups can stuff themselves as much as they want. His reasoning that low carb and low fat perform equally is flawed in so many ways, and he uses this reasoning to support a food reward hypothesis.

I think the reason very low-carb ketogenic diets cause fat loss is the same reason extreme low-fat diets cause it: they have a greatly reduced reward value.” 

Stephan Guyenet 

The fact that participants in the Lindberg study lost weight without caloric restrictions does not mean food reward had anything to do with it. Once again, if certain foods themselves cause fattening, and we restrict these foods, weight loss is likely to occur. There is no need to blame blandness.

Messing about with dopamine signaling can cause obesity in animal models, and there are differences in dopamine receptors between «normal» people and those prone to addictive behavior. It is not strange that messing about with the brain will cause all sorts of things, but it does not mean obesity is caused by food reward.

There is more reasoning to discuss, but this post is getting way to long. Is there really enough available evidence to justify calling food reward a dominant factor in obesity? If there is, I can’t say I found evidence of this in Stephan’s posts.

And as Paul Jaminet pointed out:

Likewise, we’re all familiar with young people who eat massive quantities of junk food and remain slender. The high food reward diets, even toxic and malnourishing diets, seem not to cause weight gain until some kind of metabolic damage occurs.

It seems that metabolic damage – the disease of obesity – is a prerequisite for food reward to matter.

Obese people should eat boring diets
Guyenet even offers tips on how to make food less palatable and more bland. But does this mean that there are people out there who have tried all the obvious ways to lose weight, like reducing inflammation and cutting back on carbs, who have not succeeded and are left with trying to make food not taste good?

The most palatable foods are those packed with fat and sugar. These foods are the first to go on any dietary strategy. Do we need to make the rest of the diet bland?

Guyenet offers a range of advice for losing weight based on food reward theory. For example:

Don’t snack. In France and many other countries with strong food traditions, snacks are for children. Adults eat at mealtime, in a deliberate manner.

And yet, if snack in itself do not seem to cause obesity, why not snack?

Don’t add fat to your food. That doesn’t mean don’t eat fat, it just means keep it separate from your cooking. If you want to eat butter, eat it separately rather than mixing it in with your dish.

…I don’t know what to say about this…but I know I don’t like it.

Some of his advices are meant mostly for those who struggle to lose weight, but I fear if anyone would follow them, they would die of boredom instead:

Eat only single ingredients with no flavorings added. No spices, herbs, salt, added sweeteners, added fats, etc. If you eat a potato, eat it plain. If you eat a piece of chicken, eat it plain. It can be in the same meal as other foods, but don’t mix anything together. If you would like to keep salt in your diet, dissolve it in water and drink it separately.

There are more of course. Most make sense, but they also make sense without considering food reward.

Importantly, all the studies used to support the award theory can also be used to support different theories. While they do no not falsify a reward theory they do not provide strong supportive evidence. But this is how science works. Stephan is right in offering the theory and he might turn out to be spot on. It will be interesting to see what future studies will reveal. We need some RCT’s to enlighten the causation between food and dopamine response and function, well any kind of RCT in this field would be important. I would like to make foods that are highly rewarding (measured by dopamine response or something fancy, that make people crave them, and that does not contain anything inherently fattening. Then I would give people free access to it to see if they got fat. Wonder what it could be?

«Some people have lost fat simply by avoiding carbohydrate or fat. I’ve heard people say that a low-carbohydrate diet in particular curbs their cravings and allow them to have a healthy relationship with food again (although others have developed strong cravings on low-carbohydrate diets). I believe this is mostly, if not exclusively, driven by the fact that carbohydrate and fat are major reward factors.»

Stephan Guyenet


1. Berthoud HR, Lenard NR, Shin AC: Food reward, hyperphagia, and obesity. Am J Physiol Regul Integr Comp Physiol 2011, 300: R1266-R1277.

2. Friedman MI, Harris RB, Ji H, Ramirez I, Tordoff MG: Fatty acid oxidation affects food intake by altering hepatic energy status. Am J Physiol 1999, 276: R1046-R1053.

The setpoint hypothesis revisited

The homeostasis 

(hō’mē-ō-stā’sĭs): The ability or tendency of an organism or cell to maintain internal equilibrium by adjusting its physiological processes.

Humans are warm blooded and the enzymes that make us, require a certain temperature to operate optimally. The water and ion concentration of cells must remain at a certain level to allow for normal cellular processes to occur. Homeostatic processes make us able to adapt to environmental changes. Feedback mechanisms are responsible for maintaining particular levels of processes. Most processes allow for some variation, but if pushed too far in one direction the process may break down.

Body temperature is maintained by thermoreceptors communicating with our hypothalamus, by sweating, vasoconstriction and dilation, hairs standing on end, shivering, shifts in metabolism and other. All of these factors work in unison and affect each other and by doing so maintain our body temperature at around a rough mean.

During energy restriction we get hungry, our thyroid hormone levels decrease, non essential processes like that of reproduction is down regulated, and energy expenditure goes down. It is difficult to make people gain weight by overfeeding and it is difficult to make people lose weight by underfeeding. The feeding always affects energy expenditure. Thus body weight or some level of cellular mass is required for survival and the body is self regulating to make sure sufficient tissue mass is present for survival and eventually reproduction.

None of this justifies the use of the term “setpoint” in body weight regulation.

The set point 

“Again, this supports the idea that the body has a body fat mass ‘set point’ that it attempts to defend against changes in either direction. It’s one of many systems in the body that attempt to maintain homeostasis.” 

Stephan Guyenet 

Although I have much respect for Stephan Guyenet and appreciate his very fine blog, I don’t understand his focus on the body fat setpoint. I don’t understand anyones focus on body fat setpoint for that matter. I’ve never felt comfortable using the “setpoint” word when it comes to body weight regulation. It makes me think of a glowing red number etched into my hypothalamus. “15 kg fat” NO MORE, NO LESS. It does not feel very “organic” to have a set point and I’m pretty sure I am organic. But of course, no one is claiming there is a number etched in my brain.

When it comes to the body fat setpoint, I rather like the lake comparison. A lake can for those less informed seem to have a set point of water level. Despite rather large fluctuations in temperature, evaporation and water going into and out of the lake, the lake maintains it water level because the factors mostly responsible for the level influence each other. This does not mean that it is difficult to change the level, nor does it mean the lake «attempts to defend against change.» Build a damn dam and the water level will go up. Drain it, and the level goes down. It’s not very hard, you just have to push the right buttons.

If our fat mass changes through life and in various situations and in fact is not that difficult to change, then why the hell say it’s a set point? It’s obviously not very set, is it? Most of the time we are gaining and losing fat at pretty much equal speeds, although with constant variations around the mean. The fact that the fat storing and fat using process does not spiral out of control either way, that most of us don’t get very lean or very obese, would mean that there is a control mechanism, a feedback loop or a set point. But all the body’s processes are regulated. We are self regulated and all processes work best around a given mean determined by the nature of the very process and the processes it affects or is affected by.

A homeostatic process must have a level to operate on. It’s the way the world works. The level can usually be changed by modifying the factors that make the process.

Stephan Guyenet claims that one criterion for the cause of modern fat gain is that it “…has to cause leptin resistance or otherwise disturb the setpoint.”

What I don’t understand is, why can’t we say disturb the “regulation of body weight”, “the metabolism” or “the process.” Why “disturb the setpoint”?

The language

“..the body doesn’t want to lose weight. It’s extremely difficult to fight the fat mass setpoint, and the body will use every tool it has to maintain its preferred level of fat: hunger, reduced body temperature, higher muscle efficiency (i.e., less energy is expended for the same movement), lethargy, lowered immune function, et cetera.” 

Stephan Guyenet 

Whenever body fat setpoint is discussed it seems the body is at war with one of its processes. Why turn it into fight? Obviously it is not that difficult to “fight the fat mass set point”. I’ve asked people who’ve lost weight who said it was easy. What’s more is that they have stayed lean with ease. The setpoint hypothesis was invented mostly because people didn’t lose weight by eating low fat and that they often regained lost weight. Now, I could try to change the water level of a lake by drinking the water through a straw, and when failing I could conclude that the water level is “extremely difficult to fight”, but it wouldn’t be a very good conclusion.

I believe firmly that our language is a great obstacle for scientific progress. The main reason is that it is very hard to imagine things that cannot be put into preexisting words. Also our minds operate within the boundaries of our language and language will limit the things we can imagine. I think quantum physics is a good example of an area in which the language simply is not sufficient. The electrons can be in several places in the same time. Some things are both particles and waves. Matter consists mostly of empty space and so on. On top of it all, once words are set, once something is described in a particular way, it usually stays that way.

In Guyenets posts about the body fat setpoint, the word setpoint is often given with quotation marks, like in the above quote. I can only assume that this means that the word setpoint is not an accurate description of what he is talking about, that it is not actually a set point, but that it is used for the lack of a better word.

Guyenet also writes extensively about leptin. Leptin is an interesting protein, but it is just one of those signaling molecules involved in the homeostatic regulation. Of course weight will change if we or nature messes with leptin, its receptors in the hypothalamus or any other involved part. But the existence of leptin does not justify the use of the setpoint term.

Body fat mass can be quite easy to change so why the hell do we need the setpoint? I feel the word contributes nothing to our understanding and if anything just complicates the matter more than necessary. How about adding a muscle mass setpoint and a bone mass setpoint? How about a hair growing speed setpoint or a saliva setpoint? Does it help our understanding?

I do not think the use of the word setpoint is neutral. It affects out thinking and I am afraid its existence is a net negative contributor to the science of body weight.

On the website of one of the leading obesity clinics in Norway we are told that the clinic bases its work on the setpoint theory. They call it theory, but talk about it as an undeniable fact. This is a problem because the hypothesis rests on the assumption that weight loss is inherently difficult and that the body (in the words of Guynet) “will fight” that weight loss. It is not a very positive attitude. Remember that the setpoint hypothesis saw light of day partly because losing weight through traditional (eating less) methods did not give the expected (based on thermodynamics) results and that people constantly regained lost weight. The clinic thus operates on the assumption that it may take years of hard work to change the setpoint downwards and that it takes a very short time to change the setpoint upward (that the body «gets used» to a new higher level of fat mass).

What if the obesity clinic, instead of basing its treatment on setpoint theory (forcing overweight people to starve and doing insane amounts of exercise), said that weight loss does not have to be difficult and lasting weight loss is possible? What if they read some real science, pushed the right buttons and stopped torturing people? The studies are out there. It really does not have to be difficult. The metabolism must be altered, mostly through changing the hormonal milieu and reducing inflammation thus achieving a homeostatic regulation at a lower amount of body fat mass. And there is little evidence it has to take years to do so.

The conclusion

«When something seems ‘the most obvious thing in the world’ it means that any attempt to understand the world has been given up.» 

 Bertolt Brecht 

If the point of this rambling is lost on you, let me try to sum it up.

The body fat “setpoint” is just a word used to describe a level of a self regulating biological process.

To call a certain mean level of a homeostatic regulation a set point is unproblematic. As far as I can see, this is how Stephan Guyenet uses the term. But when medical professionals use the setpoint hypothesis to argue that body weight loss by nature is difficult and near impossible, we have a problem. The continued use of the expression seems to affect both medical professionals and researchers in a negative way, because it closes their mind for other possibilities, among them that weight loss from a physiological point of view is easy if you just push the right buttons. That we do not always know what these buttons are does not mean weight loss must be difficult, only that we are ignorant. Many aspects of metabolic regulation are under our control. Most important of all is that the level of homeostatic regulation of body fat called the setpoint, seems to be determined by what we eat.