Monday, January 2, 2012

High-Fat Diets, Obesity and Brain Damage

Many of you have probably heard the news this week:

High-fat diet may damage the brain
Eating a high-fat diet may rapidly injure brain cells
High fat diet injures the brain
Brain injury from high-fat foods

Your brain cells are exploding with every bite of butter!  Just kidding.  The study in question is titled "Obesity is Associated with Hypothalamic Injury in Rodents and Humans", by Dr. Josh Thaler and colleagues, with my mentor Dr. Mike Schwartz as senior author (1).  We collaborated with the labs of Drs. Tamas Horvath and Matthias Tschop.  I'm fourth author on the paper, so let me explain what we found and why it's important.  

The Questions

Among the many questions that interest obesity researchers, two stand out:
  1. What causes obesity?
  2. Once obesity is established, why is it so difficult to treat?
Our study expands on the efforts of many other labs to answer the first question, and takes a stab at the second one as well.  Dr. Licio Velloso and collaborators were the first to show in 2005 that inflammation in a part of the brain called the hypothalamus contributes to the development of obesity in rodents (2), and this has been independently confirmed several times since then.  The hypothalamus is an important brain region for the regulation of body fatness, and inflammation keeps it from doing its job correctly.

The Findings

Inflammation occurs in many tissues when rodents are placed on a fattening diet, but this usually takes weeks or months to develop, and therefore it is often considered secondary to the development of obesity.  One of the things we were able to show in this study is that in rodents, we can detect signs of inflammation in the hypothalamus within one day of exposure to a fattening diet, making it the earliest known inflammatory event during the development of obesity.  This is consistent with the idea that dysfunction of other tissues is the result, at least partially, of dysfunction that occurs first in the brain.

Our study also addresses the second question: why is obesity so difficult to treat?  We know that brain inflammation contributes to obesity in rodents, so one possibility is that the hypothalamus sustains damage during this process, which durably elevates the "setpoint" around which body fat mass is defended by the body.  In other words, damage to the part of the brain that regulates body fatness makes the body "want" to carry more fat and resist fat loss attempts.

This is the hypothesis we began to test in this study, and we found evidence that supports it.  Rodents placed on a fattening diet show evidence of neuron injury in as little as three days.  Our study was not able to say that neuron injury causes obesity or resistance to fat loss, only that obesity and neuron injury are associated with one another.  This is the first step in this line of investigation, and we are currently following up with related studies that will expand the findings. 

The most striking finding in the paper was contributed by our collaborator Dr. Ellen Schur, who showed by MRI that obesity is associated with a marker of neuron injury in the hypothalamus of live humans. As you might imagine, this increased the impact of the study considerably!  It is important to point out that we found the suggestion of neuron injury specifically in the hypothalamus, not in other brain regions.  We did not find changes in brain regions responsible for thought, language, movement, etc., that might be expected to impact a person's ability to think and function.

The Diet

The "high-fat diet" that was used in this study is Research Diets D12492 (3), and the comparison diet was normal unpurified rodent chow.  Normal rodent chow is a whole food based pellet that is mostly composed of unrefined corn, soybeans, a small amount of meat and animal fat, and added micronutrients.  It is very low in fat, typically ~14 percent of calories.  They do just fine on this food, particularly if they are given a running wheel and other forms of environmental enrichment. 

D12492 is a "purified" diet that is fundamentally different from unrefined rodent chow.  It is one of the many rodent diets that were created because investigators needed a highly consistent food to minimize experimental variability.  Whereas unrefined chow can have a different composition based on the variety of corn/soybeans used, the terroir, the particular growing conditions of that year, etc., purified diets are composed entirely of highly refined ingredients so they are much less variable.  However, it rapidly became clear that rodents don't do as well on these diets, even if macronutrients and micronutrients are approximately matched to the unrefined chow diets.  D12492 is composed of non-hydrogenated lard, soybean oil, maltodextrin, sucrose, casein for protein, cellulose for fiber, and added vitamins and minerals.  It does contain sugar, but the amount is modest (6.8 percent of calories).

D12492 is 60 percent fat by calories.  Although rodents love the taste of it, it is not good for their health-- susceptible strains of rats and mice will begin gaining fat in as little as a week on the diet, become morbidly obese by three months, sometimes develop diabetes, and live less than half a typical rodent lifespan if they are allowed to age (generally they are not). Personally, I refer to this diet as a "purified high-fat diet", because that acknowledges that not only is it high in fat, it is also composed of refined ingredients.  In our study, there is no way for us to know if what we observed in rodents was due to the dietary fat per se, some other aspect of the diet, or both.  Based on other findings, I strongly suspect that it is both.

We deliberately use rodent strains that are susceptible to obesity on this diet.  Some strains are more resistant to obesity than others, but a comprehensive look at the literature reveals that high-fat diets are generally not good for rodents, and most strains tend to gain some amount of fat and develop long-term health problems on high-fat feed.  There are a few exceptions in the literature if you look hard enough for them, but they are drowned out by the much greater number of studies showing harm.

So if we're deliberately selecting rodent strains that are particularly sensitive to fat gain on a purified high-fat diet, how can we generalize from this and say that dietary fat causes damage in the brain and obesity?  The answer is that we can't, and we haven't.   Nowhere in the paper does it say that dietary fat per se causes damage to the brain, or even causes obesity, and Drs. Thaler and Schwartz were careful not to say that in interviews either.  We choose rodent strains that are susceptible to obesity on purified high-fat diets simply because we're studying obesity, and we know that feeding this diet to the right strains of rats and mice produces it readily.  

The fact is, we don't know what aspects of D12492 cause injury to neurons and obesity in rodents, and we don't know if those are the same factors that cause obesity in humans.  That was not the point of the study!  The point was to try to understand what's happening in the brain during the development of obesity.  Based on previous studies, the dietary fat itself is probably an important component that makes D12492 fattening in rodents, but whether high dietary fat would lead to obesity in the context of a well-composed whole food-based high fat diet, and a running wheel, is not known. 

The Implications

This study once again highlights the importance of the brain in the development of obesity, and shows directly for the first time that similar changes occur in humans.  Our findings also raise the possibility that injury to the hypothalamus may contribute to the fact that obesity is so difficult to durably reverse.  That does not mean that fat loss efforts are hopeless!  I believe that with the right techniques, many obese people can lose a substantial amount of fat and keep it off.  But realistically, we already know that it is rare for a long-term obese person to attain a totally lean state.  Persistent changes in the function of the hypothalamus are a logical way to explain this, although more research will be required before we can say it conclusively. 

I do not think this study suggests that dietary fat is inherently fattening or causes damage to neurons in humans.  The question of whether or not dietary fat is inherently fattening is controversial, but our study did not address it.  Based on my reading, studies show overall that dietary fat is not fattening in humans as long as total calorie intake remains appropriate.  However, adding fat to food that is otherwise low in fat does facilitate overconsumption of calories in some people, by increasing energy density and food palatability/reward, and this has been demonstrated many times.  I think the fact that low-carbohydrate diets cause fat loss in many obese people offers support to the idea that dietary fat is not inherently fattening in humans. 

This was not a diet study, but if there is a dietary message in it, it is this: eat a whole food based diet that allows you to reach or maintain a healthy weight naturally. 


Stephen Boulet said...

"Nowhere in the paper does it say that dietary fat per se causes damage to the brain, or even causes obesity, and Drs. Thaler and Schwartz were careful not to say that in interviews either."

This really highlights how easy it is for people to jump to faulty conclusions. I can only imagine how it must feel for researchers to see their work taken out of context. Maybe the conclusions section of the paper should always delve into the the study's limitations.

Chris Kresser said...


I wonder if you could comment on the choice of the title for this paper. If you call the paper:

"Eating a high-fat diet may rapidly injure brain cells that control body weight"

it's not at all difficult to see why people would interpret that as "a high-fat diet causes brain damage and obesity."

This is something that really irritates me about a lot of the published research in this area.

Why not be more specific and say "purified high PUFA diet"? Soybean oil is very high in n-6 PUFA, and lard is relatively high depending on the source.

Serdna said...

Sorry, but you have a typo just in the abstract of your paper: where it says "high-fat diet" it should say "high omega-6 diet". After that, don't expect any reporter to grab a clear idea of your work and its limitations. Whenever you do the same experiment with butter instead of modern lard and soybean oil, then we will have at least a piece of useful information.

Best regards.

Stephan Guyenet said...

Hi Chris,

I'm not sure what you mean. The paper was titled "Obesity is Associated with Hypothalamic Injury in Rodents and Humans".

Soybean oil is only about 10% of the fat content of the diet; the other 90% is lard.

Hi Serdna,

Sorry to disappoint you, but it has been shown many times that butter causes fat gain in rodents as well.

SamAbroad said...
This comment has been removed by the author.
Beth@WeightMaven said...

Chris, the title of the actual paper is not "Eating a high-fat diet may rapidly injure brain cells that control body weight" ... it's "Obesity is associated with hypothalamic injury in rodents and humans."

Marwan Daar said...

Chris Masterjohn has an interesting piece on the D12492 diet:

Chris Kresser said...

Ah, my mistake! I saw it at the bottom of the Science Daily article and assumed that was the title.

James W. Hill, MD said...


Supportive of your research, further evidence of obesity-associated hypothalamic damage -- reduced neurogenesis in this case --was reported by Harvard scientists last week:

Perfect Parallel said...

Very interesting - another example of how a diet that a species is poorly adapted to drastically affects health markers.

LeonRover said...
This comment has been removed by the author.
yolio said...

Cool results! Congratulations.

Tony Mach said...

If they had replaced the soybean oil with coconut oil, or olive oil, would they had the same results?

Capcha: palma :-)

LeonRover said...

Nice paper and GREAT presentation of data.

Merci beaucoup for free online viewing.

It also shows how difficult it is to get low residual variance data from human measurements.

I note the difference between mouse data with an unexplained variance of 30% vs human data with unexplained variance of 62%.

Oh and BTW - while many obesity researchers allow their human subjects to overfeed on Real Fast Food, they conduct caloric restriction with the human equivalent of Refined Rat Chow.

I have noted Liebel feeding his subjects:
Fat (corn oil) 40%
CHO (PolyCose) 45%
Pro(Casein hydrolysate) 15%

Deirdre said...

Hi Stephan,

As a researcher in the field of childhood obesity, my perspective is from the “new” food frontier that future generations of children are now being raised in (i.e., GMOs, hyper-engineering, etc.). One of these massive grey areas is the adoption of novel proteins engineered into our food supply – in particular, the Bt/HT “proteins”/toxins produced in genetically modified crops such as soybean, corn and cotton. There’s already plenty of evidence that the Bt-toxin produced in GM crops is toxic to humans and mammals and triggers immune system responses.

With regards to your study ...

Currently, 94% of commercially grown soybean crops and 65% of commercially grown corn crops are genetically modified using these engineered toxins. My question to you is this … If you used commercially grown corn and soybean products in the rat chow, could this possibly have influenced the inflammatory response that you and your colleagues observed in both your animal and human subjects?

If so, this would be an interesting segue into a study on GM organisms, and whether or not they can pass the blood brain barrier. And if proven … then that would be a pretty huge discovery, IMHO.

Emily Deans said...

Hi Stephan,

Very nice post! I'm sure you saw this paper from the beginning of the year - done in mini-pigs and using SPECT scans: Changes in Brain Activity after a Diet-induced Obesity

Not humans, of course, but likely relevant.

All best,


RobR said...


Do you have any thoughts on the role of BDNF in protecting and possibly regenerating damaged areas of the hypothalamus?

There is some evidence that BDNF produces weight loss in rats.

It would be interesting to know how the damage differs on this high-fat chow with rats put on an IF schedule, HIIT, sleep deprivation, or dosing of curcumin+piperine.

Do you suspect low-reward feeding will repair damage to the hypothalamus? Would the BDNF produced by calorie restriction be a plausible repair mechanism from low-reward feeding?

Steven Hamley said...


Why do researchers refer to the highly refined and toxic diet as high fat? Why not just call it the the 'obesogenic refined and toxic diet', or the 'obese RAT diet' for short.

Derek H said...

Stephan (or others),

This may have been discussed in the past, but when it comes to research, what parameters define "obese". How much bodyfat accumulation are we talking about? Thanks.

john said...


Check out BDNF and intermittent fasting.

ivat said...

Congratulations: this is valuable work.

At what point does inflammation to the hypothalamus result from underfeeding, do you think? Would this also begin in the short term?

Serdna said...

Okay, so butter is fattening in murine species. Is high sugar also fattening for them? Is any fattening murine diet doing the same neurological damage? Since in humans butter doesn't seem to be fattening, is a murine model of obesity validated enough for drawing conclusions? I suppose as long as the model centers on "diet change inducing obesity" as input variable and not on a specific fattening diet it will be more or less validated for that specific variable, but you certainly are not getting the right message through to the mass media. Nevertheless, studies on omnivorous species with the same fattening diets, such as the one pointed out by Emily Deans, seems more convincing.

Best regards.

gunther gatherer said...

Hi Stephan,

Do you have any evidence that inflammation starts in the hypothalamus before obesity? Because it could be explained that obesity just IS inflammation and that lots of body tissues become inflamed when you get fat, including the hypothalamus.

So does a damaged hypothalamus cause obesity or does obesity come with a damaged hypothalamus, amongst lots of other things?

Very revealing TEDx talk, BTW. If vegetable oil consumption has risen so much, could we conclude that "n-6 ruins hypothalamus causes obesity"?

JBG said...

" certainly are not getting the right message through to the mass media."

Golly, yes! The titles and the content of the four links given at the start of the post all shout a profound mis-message. Not sure what could be done to prevent that, but all the authors must find it perplexing and embarrassing.

A particular question that crossed my mind-- Why would damage to the hypothalamus simply cause the body to defend a higher setpoint? Why wouldn't it abolish the setpoint mechanism ouitright? (In fact, if rats and people persist in getting fatter, isn't that, at least in effect, what happens?)

Marwan Daar said...

JBG, read Stephan's four part series on body fat and set-point. It discusses some of the mechanisms involved in leptin signalling and inflammation.

Sara E Anderson said...

The hypothalamic involvement in cementing a set-point for body weight makes me wonder what this might have to do with the way that many women seem to see their set-point revised upwards after pregnancy. Can you comment on a possible relationship?

mscott99 said...

After opening the Spokesman-Review and reading about this study, I was really wondering how they concluded that "after humans and rodents eat a high-fat diet, their brains begin to show signs of injuries in just 24 hours."

I'm frequently bothered by the way the media tries to simplify complicated studies into headlines. Basically:

Eric De las Eras said...

Very Interesting topics!

Eric De las Eras said...

Very Interesting topics

Todd said...

@ gunter gatherer

To answer your question, there is plenty of evidence that obesity starts in the hypothalamus. I've summarized that evidence in this post:

One of the most striking studies, by Ono et al. (, showed that hypothalamic insulin resistance precedes — and probably causes — obesogenic insulin resistance in other organs and tissues. Ono found that feeding rats a high fat diet induced insulin resistance in the hypothalamus after only one day, with no concurrent hepatic insulin resistance! It took a full 3 days on this diet for insulin resistance to show up in the liver, and 7 days for the muscles and peripheral tissues to become insulin resistant. The mechanism of inflammation was the activation of the mTOR/S6K pathway by exposure to fatty acids. The S6K protein apparently inhibits insulin signaling in the arcuate nucleus of the hypothalamus, activating the orexigenic NPY/ArGP neurons and inhibiting the POMC neurons.

bharatbook said...

Really obesity became big issue for our health..
market research

FocusAppsStore said...

Excellent information.I have not seen this before.Obesity is a condition of acquiring as much as 20 percent of more fat than a normal human body can embrace.
For more information related to this please visit : understanding obesity and medical animation

Jane said...

So is the fat doing this by causing oxidative stress? I think it must be. Fatty acids activate NADPH oxidase, which makes superoxide, and fat metabolism in mitochondria produces superoxide too.

Superoxide releases iron from ferritin, and free iron produces toxic hydroxyl radicals.

Saturated fatty acids are linked to insulin resistance, which itself is linked to excess mitochondrial superoxide according to a paper Stephan gave us:

Mitochondrial superoxide is normally detoxified by manganese, which appears to counteract the oxidative stress caused by excess iron.

Everybody thinks manganese deficiency is rare. Maybe it isn't. After all, white flour, white rice and white sugar have very little or none, despite the need for manganese in carbohydrate metabolism. White flour and white rice even have added iron.

So what does this have to do with a high-fat diet? It turns out that saturated fat promotes iron absorption and inhibits manganese absorption.

Himanshu said...

yes this is true we are really facin these problems in our daily life due to improper consumption of food and unhealthy diet

Himanshu Girdhar
Health BPO

berto said...

Hi Todd,

"The mechanism of inflammation was the activation of the mTOR/S6K pathway by exposure to fatty acids."

Could you highlight where they actually state this in the study?

I only noted that they claimed exposure to excess nutrients and hyperinsulinemia.



JBG said...

Marwan Daar said...

Chris Masterjohn has an interesting piece on the D12492 diet:

= = = = =

Masterjohn's very interesting article says that D12492 is not only a very specific "HFD" (as a number of earlier commenters here have said) but it is not even the specific HFD that the company selling it thought (and said) it was.

Reminds me of the breast cancer debacle where a decade of research was carried out with a cell line that was not what researchers thought it was.

Anyway, I would say that the use of the phrase "high fat diet" in the article Stephan's post is about, with apparently no discussion of just what Research Diet D12492 is (all I could find was a bare mention of it -- not even a reference), fairly compelled the media to report as they did. If the phrase "high-fat diet (HFD)", which occurs in the first line of the abstract had read instead, "engineered obesity-inducing diet (OID)" or "...inflammation-inducing...(IID)" or something similar, and that phrasing maintained throughout the article, perhaps the entire misdirected media frenzy could have been avoided.

Todd said...


My statement regarding the inflammatory activation of the mTOR/S6K pathway by fatty acids is not a direct quote, but a synthesis of what Ono et al say in their final discussion section:

"Here we demonstrate...that a single day of exposure to a HFD [high fat diet] blunts both insulin signaling in the hypothalamus and the ability of hypothalamic insulin to suppress HGP [hepatic glucose production]. This blunting is accompanied by activation of S6K, suggesting a role for this molecule in the mediation of central insulin sensitivity and peripheral glucose homeostasis...Our observations support the idea that the mTOR/S6K pathway is critical for nutrient sensing in the arcuate nucleus."

Hope that helps.


JBG said...

Reading back over the comments, I see that Steven Hamley earlier made my point about the phrase "high fat diet", and several others said things pointing in the same direction.

Stephan, in the original post you are very explicit at making clear that the paper is about the consequences of a certain kind of hypothalamic injury, and NOT about how humans should eat. Presumably the other authors were also clear about that. Was there any discussion among the authors about use of the phrase "high fat diet" and its likely media result? Do you think those involved are likely to look for more restricted phrases in future?

Stephan Guyenet said...

I do wish we could use a different name for the diet, but "high-fat diet" is standard research terminology for this type of diet. When you're publishing a paper designed to convey information to the scientific community, you use terminology that other people in your field will recognize. Making up new abbreviations will leave people scratching their heads.

But let's be honest here: D12492 IS a high fat diet, and the fat itself is almost certainly one of the main reasons why it's fattening in rats (purified low-fat comparison diets are generally less fattening, even if they're high in sugar). Our study did not state or imply at any point that dietary fat itself causes neuron injury and/or obesity in humans. Period. Any such statement or implication was entirely created by the media. As frustrating as it is, we can't control how the media choose to report our study. Scientific papers like this are written with the academic community in mind, and it is easy to misunderstand them if you don't have a background in the subject.

Personally, I favor the term "purified high-fat diet", because it is an accurate descriptor and it is standard terminology used by the American Institute of Nutrition (which originally designed these diets). However, the abbreviation will still be HFD so that researchers will recognize it.

Travis Culp said...

Strictly speaking, this was a high fat diet and probably matches fairly well the composition of fats that the average American consumes (and becomes obese eating). Those eating more ancestral types of high fat diets always become so knee-jerk defensive of their precious dietary fat, but if we want to understand the obesity epidemic in the United States, for example, we wouldn't use butter, whole cream, tallow etc. as it's not what the vast majority of vast people actually consume. If anything, this 60% fat diet actually contains healthier fats than the 45% (or whatever it is now) fat diet that the average person in America is consuming. It's all toxic, inflammatory seed oils with a lot of trans fats.

Thanks for the head's up on this one, Stephan. It's becoming quite clear that the major issue with obesity is when conditions are imposed that lead the hypothalamus to incorrectly estimate the body's fat stores and induce the austerity measures/starvation response that would be appropriate at a far lower body fat%. With any luck, some reliable method will be found for re-teaching the hypothalamus how to accurately gauge fat stores and respond appropriately.

bentleyj74 said...

"Scientific papers like this are written with the academic community in mind, and it is easy to misunderstand them if you don't have a background in the subject."

Thanks for saying that :)

Armi Legge said...

Hey Stephan, thanks for writing up the details of the study in such a clear fashion.

If excess calories cause inflammation, does this mean that people who are trying to gain weight are creating an inflammatory state by overeating? Excess calories are needed to gain weight, so shouldn't this cause inflammation in any form? If it doesn't, could the small amounts of refined/low quality ingredients be the key differentiator between fat formation or hypertrophy?

Thanks again, great work as always :)


Armi Legge said...

Sorry, let me clarify. If people who are trying to gain muscle overeat, shouldn't this also cause inflammation?


bopes said...

I nominate "Obesity Inducing, Non-Ketogenic" (OINK) diet.

; )

Btw, has anyone seen this:

This article in Science Daily reports on it:


Because RGS9-2 is normally expressed in the brain's striatum, a section of the brain involved in both motor control and reward responses, Kovoor and his fellow researchers thought that the weight gain could be a result of an increased reward response triggered during eating.
"You would expect more eating from the mice without RGS9-2 (because they were the ones that gained weight), but that was not the case," Kovoor said. "Studies with humans, rats and mice implicate RGS9-2 as a factor in regulating body weight. But we had to look at another factor other than feeding behavior."
"Our research shows that the striatum, through RGS9-2, has a role in regulating body weight that is independent of the motivation, movement and reward responses," Kovoor said. "We have identified a new gene that likely regulates weight gain through metabolism."

Pete said...

According to this article (, lab mice have been experiencing an obesity epidemic way before us humans. How can feeding an animal that's already sick a high-this-or-that diet prove anything?

JBG said...

Stephan, I'm afraid I can't agree.

The paper in question is a technical one concerned with brain physiology. There is no legitimate reason for it to come to the attention of the general public at all.

The way the diet used in the paper is phrased (it is not anywhere in the paper actually described) would lead almost any lay reader to the wrong conclusion. To expect a science reporter, or even a medical reporter, to figure the situation out (ie, to find the single mention of D12492 and look it up on his own) is unreasonable. The media is not to blame for the problem in this case.

Surely it would be possible to craft a sentence, a single sentence, to include in the abstract that would tip off reporters to what the case actually is. I see it as a responsibility of academic authors to have the media problem in mind and to take appropriate action to fend off likely misinterpretations. The human diet situation is complicated enough without such unnecessary additions to the confusion.

Stephan Guyenet said...

Hi Armi,

I'm not sure. If you're overeating but also working out such that the energy is going into muscle building, I wouldn't necessarily think it would cause inflammation.

Hi bopes,

Interesting. I don't think they demonstrated that the effect was not via reward circuits though. They assumed this was the case because food intake wasn't different, but they never actually tested reward function. Actually, their measurement of food intake itself seems to have been pretty cursory so they could have easily missed a difference that occurred at an earlier time point.

But the point is that reward circuits connect with hypothalamic circuits that regulate body fatness (as they mentioned in the paper). These hypothalamic circuits influence both food intake and energy expenditure. So it's entirely possible that the mice gained fat due to reduced energy expenditure because of abnormal activation of reward circuits connecting to the hypothalamus.

Hi Pete,

Standard lab rodents are not sick. We measure their body fat content, insulin/leptin levels, food intake etc. all the time in my lab. On regular unpurified chow, young adult rats and mice are ~5-12% body fat as measured by MRI. If you let them age for a while, they do get fatter, but that's mostly because they're sedentary. If you put a running wheel in the cage, they stay much leaner.


Our primary objective is to do good science and communicate it accurately to the scientific community, and that is what we have done. We were careful in the wording not to imply things that we didn't find. The fact is, many media outlets reported the finding accurately, so it's not as if it was written in a way that made it difficult to understand. Those outlets that didn't report it accurately, did so because they wanted the story to be 1) more sensational/impactful, and 2) more actionable. This is how the media works, and we can't change that.

As far as I can understand, the objection that some people have to the paper's wording revolves entirely around the use of the abbreviation "HFD" to describe the rodent diet we used. As I said, this is standard research terminology, and it isn't exactly an outrageous way to describe a diet that is 60% fat, 20% carb and 20% protein. I'm sorry, but that does not constitute being misleading, given the careful wording we used throughout the paper. It doesn't take a genius to figure out that the paper does not state or imply anywhere that dietary fat causes neuron injury in humans.

Asim said...


Not all inflammation is bad. It's a biological response against pathogens and tissue damage. When your "ripping" your muscles through intense weight training, your creating a state of inflammation to repait the muscle tissue. In the case of exercise, your going to have to increase your caloric intake.

When people of science, not pseuod-science, speak about inflammation being a problem, it's in reality, in the total context of the "Western" diet. If your constantly triggering inflammation, in the long wrong you'll see a rise in metabolic disorders.

Asim said...

Here is a great link on inflammation and excercise:

Armi Legge said...


I agree, and I understand not all inflammation is bad. My question was not in regards to the effects of inflammation from exercise. It was whether or not exercise augments the inflammation caused by excess calories. According to Stephan's response, it would appear that since exercise (depending on the exercise) enhances nutrient partitioning, it is not the same kind, or level of inflammation.

Dave Liddy said...

Another egregious misreporting of the article

The quote from the bot at the Baker-IDI Institute just shows how bad we have it here

JBG said...

Stephan, I do not mean to be argumentative, but observe how easy it would be to virtually exclude the troublesome misinterpretation (initial sentence of abstract):

Rodent models of obesity induced by consuming the artificial high-fat diet (HFD) D12492 are characterized...

You say, “...many media outlets reported the finding accurately...” but the four you name all got it wrong, in a couple cases blatantly wrong. In order:

• Researchers...fed an American-style high-fat diet to lab rats...

• "....brain injury may be a consequence of the overconsumption of a typical American diet ..." Joshua Thaler...

• Pass it on: Eating a high-fat diet may change the brain in ways that makes it harder to lose weight.

• [Headline:] Brain injury from high-fat foods may be why diets fail

The fact that the paper used “careful wording” that “does not state or imply” the problematic misinterpretation is logic chopping, given that almost anyone NOT in the researcher in-group almost certainly would get the wrong idea. I would say that taking a big step to prevent that would be worth three extra words.

Maxxy Pane said...

Obesity does not just happen instantaneously, it produces progressively from poor living alternatives and, somewhat, from your body's genes. Way of life alternatives are an essential aspect in having an influence on your weight. Eating more calories than you need may be down to unhealthy diet.

Insane Detox

hithaeglir said...

Hi Stephan,

Excellent post!

This was another classic example of lazy media misreporting of science research. It was refreshing to see one of the authors of this study willing and able to point this out in a candid way.

As a "specialist" science reporting source, Science Direct in particular should be thoroughly ashamed of their negligent "headline".

It was also good to see the paper published as open access. This enabled me to read the original paper and conclude that the "sensational" Sydney Morning Herald headline sent to me by a colleague was simply not correct. More papers should be open access, instead of being locked behind expensive pay-walls affordable only be wealthy individuals and large institutions.

JBG said...

Although the proposed added three words (or something like them) at the beginning of the abstract to provide a red flag that an attentive person could be expected to notice would be the crucial change for the present case, it would also be desirable down in the Methods section to provide a sentence or two describing D12492. All this thinking, of course, is intended to influence future practice, the present case being water over the dam.

A thought has occurred to me, however, that leads me to offer to desist altogether from pursuing this. I realize, Stephan, that even if you were persuaded, it would be awkward for you to say so, since it would imply criticism of decisions embodied in the paper that were made (consciously or not) by the senior author/s.

So, having said my piece, I’m willing to shut up.

Stephan Guyenet said...


I deliberately selected media headlines for my post that presented the finding in a sensationalistic way. I was trying to make the point, subtly, that the finding had been misrepresented by some sources.

I agree that we should have described the diet in more detail in the methods section. I didn't write the manuscript (although to be clear I did review it) and I wasn't first or last author, so I didn't have much control over the wording. I did have some control, since I reviewed it, and that is something I should have emphasized but it slipped under my radar. The methods section often doesn't get scrutinized as much as the rest of the manuscript.

Note that Josh Thaler, in the interview, did not state that eating fat causes neuron injury, he stated that eating the "typical American diet" MAY lead to neuron injury. That is a very accurate representation of what we found.

Robert Andrew Brown said...

Fascinating research Stephan

Well done for all you hard work and dedication to your regular posts.

The sad truth is that this comment may represent the reality. Your paper fairly raises the issue.

Science has as yet to sort out and agree why, but it is certain that oxidative stress will be there in the mix.

• "....brain injury may be a consequence of the overconsumption of a typical American diet ..." Joshua Thaler...

inmybackpages said...

D12942 is not 6% sugar, it's 20% sugar. Your calculation is based only on sucrose and you left out maltodextrin.

montmorency said...


Did you actually measure the respective proportions of saturated fat, unsaturated fat, and polyunsaturated fat in the lard?

Theoretically, lard is supposed to be "only" about 11% PUFA, but I believe it can be much higher, depending on what the pigs were fed. I don't live in the USA, but I've heard that some commercial American lards can be not far off being liquid at room temperature, which would suggest a high level of unsaturation.

You quoted:

Note that Josh Thaler, in the interview, did not state that eating fat causes neuron injury, he stated that eating the "typical American diet" MAY lead to neuron injury. That is a very accurate representation of what we found.

Does the typical American eat lard as 90% of the fat content of her diet? I would be surprised, and if not, then I don't see how your last sentence follows.

Sarah Barracuda said...

@inmybackpages - Whether maltodextrin is classified more as a sugar or a starch depends on the degree of hydrolysis of the glycosidic bonds. It's a POLYsaccharide, not even necessarily an OLIGOsaccharide. I refer you to any of the biochem experts swimming in the blogosphere if you would like to pursue this further.

Maybe you're not trying to nitpick, but that's how it comes across to me. So if you're going to nitpick, at least do your homework first.

Also, I'm pretty sure the breakdown of D12942 is recorded pretty clearly somewhere easily accessible to obesity researchers, and Stephan did not have to (mis-)calculate anything himself.

Mary Parlange said...

This is a perfect example of the disconnect between the way the scientific method works and the way that science is communicated to the media and the public. THe study wanted to look at if and how obesity was associated with changes in the brain. They can't experiment on humans (because it involves slicing up and examining brains), so they use the lab mouse. To do the study, they have to make the mice obese. There's an easy, reliable, fast way to do that (the HFD), so that's what they do. The result is very interesting, if you are a mouse.

THe problem is that the media and their audience are interested in humans, not mice, and in the whole diet/obesity/brain spectrum, not just the little bit that's analyzed here. The disconnect comes when the results are applied far beyond the field where they are, in fact, applicable. (i.e. to humans and to human nutrition). This happens all the time.

I worked for many years in the media/press office of a scientific institution and I know from personal experience that you have to be VERY, very careful how you word things in order to avoid misinterpretation, and often, even if you're careful, the research results are misinterpreted anyway. ANd I also know from first-hand experience that many people working in institutional press offices aren't very careful because their job is to get the name of the institution in the news in as many outlets as possible - so they deliberately make things sensationalist in order to attract attention. Sometimes the scientists can override them, sometimes they can't. (it's not always easy to get a good explanation out of a scientist, either, BTW. I'd have been happy writing a press release on the research based on Stephen's explanation!)

Bottom line - go to the source if you find a news item about science interesting, and good luck with the jargon.

bharatbook said...


Advertising Market

rick said...

What about arachidonic acid and neuro inflammation

Eades, Kresser and McClearly never replied!/rmorranis/status/97354379929198592!/rmorranis/status/97642734344613888

chief gabril said...

Drugs are good, because they cure diseases, let us good. They can be toxic when they are granted. There are tips that will help you know what you need to do, especially when you are in any drug or treatment.

Stair lift

rrustad said...

Maltodextrin is ~derived~ from various starches (depending on the country of origin) not classified as a starch. Maltodextrin is classified as a sugar based on dextrin equivalent (DE). DE is a dry measure of reducing sugars in a sugar product.

Starch has 0% DE. Glucose/dextrose has 100% DE. Maltodextrin ranges from 3 to 20% DE. Maltodextrin is a common additive in soda, candy bars, and beer.

viji said...

This work must be really appreciated.

Hoodia Weight Loss Pills

josh said...

hi stephan - i understand your intent to communicate with the academic community and the resulting choice of words, but it would be irresponsible and potentially unethical to pretend things happen in a vacuum. obviously the media will pick this up, and i think that the authors have a bit of a responsibility to be as clear as possible in order to do everything they can to avoid misrepresentations. it isn't a perfect world and some media will still twist words and facts to sell their product, but that doesn't absolve people in your position.



Taylor said...


I'm curious whether you believe this hypothalamic change to be pathological or just adaptive? If it is adaptive it raises the question of what advantage it confers.