Why We Get Fat ... Lessons from a Cafeteria Rat

BUMP!

In addition to blogging on several topics raised during Stephan Guyenet and Gary Taubes recent appearance on the Joe Rogan Experience podcast, I've decided to bump some of the past content on this blog relevant to some topics discussed and/or research studies mentioned.  I'll include a little commentary with each.

At around the 12 minute mark, Stephan makes a great opening play, and then gets distracted or flustered into fumbling the football.  Stephan asks what it is about the food environment, etc. that causes obesity, and distills it down to the simplest of questions (paraphrase):
Q:  What is the most fattening diet on the planet?
A:  Human junk food.
Nail.  Head.  Boom.  He then calls out some numbers from the research summary he posted on his blog in advance of the podcast.   The study I discuss in this blog post is the one included in "#52" that he's discussing in the podcast.    Joe seemed perplexed by this idea that we couldn't replicate the junk food fattening with sugar or even fat alone.   Joe seems fixated on junk food = sugary "bullshit diet".  He gets very confused, despite Stephan laying it out fairly well.  Give just sugar/refined carb or give just fat, and the rats don't get AS obese as they do when you give them a mixture of sugar/fat and FLAVOR in human junk foods.    Yes, Joe, soda is just sugar and is a junk food, but junk foods are  also things like donuts, cookies and ice cream or pizza, fries and Doritos.  Carb + fat + flavor. 

While the CAF rats are indeed rats, this study does a fairly good job in demonstrating that it's not just the macros.     Four diets:  Standard Chow (mostly grains, 12% fat) , Low Fat Refined (35% sugar, 10% fat) diet, High Fat diet (45% fat, as lard and soybean oil) and CAF -- a wide assortment of junk foods from sugary breakfast cereals, to Doritos, to pork rinds, to wedding cake.   If it's carbs and insulin, why don't the rats on the SC get fat?  If it's just refined carbs, why don't the sugar rats get fat?  If it's just the fats, why don't the lardos get fat?   These two groups of rats did gain more weight, but -- as Stephan mentions -- they don't gain a lot more, and more importantly, they don't overeat energy.  The CAF rats, meanwhile, consumed over 30% more food/calories and got a whole lot fatter. 

Gary's Response:  We're looking at this wrong.  Calories are the wrong way to look at this.  











{Original publish date:  2/25/11}


(Hat tip to Beth for bringing this to my attention)

This study used male Wistar rats

{eek ... I'm having flashbacks to a former career!}

This rat is not a genetic mutant designed to be predisposed towards obesity.  It is, however, often used in diet induced obesity (DIO) studies.

All of the rats were 7-8 weeks old and weighed approx. 200g when they began the study.  They were given a standard chow (SC) diet ad libitum for 2 weeks and achieved an average weight of 300g prior to being assigned to one of four ad libitum diets for 10 weeks:

  1. SC control (estimated <5% simple sugars, 12% fat)
  2. HF chow - 45% fat, lard
  3. LF - low fat (10% lard & soybean oil) matched to HF chow with sucrose substituted for fat content, thus this is a 35% sucrose diet
  4. CAF = SC + 3 human snack foods varied daily from a list

Here is the table of the foods made available to the CAF rats: {as always click to enlarge in full browser window}


Now many low carbers will probably take offense to the inclusion of pork rinds, cheese, pepperoni and hot dogs as being "unhealthy snacks", but these were nonetheless included in the rotation of foods offered to the rats.  Actual consumption was determined by weighing leftovers.  

Here's how the diets consumed panned out:  

Note that the total caloric consumption was:

HF (559) < LF (660) ≈ SC (663) << CAF (908)  

The authors make the following additional observations:
  • All rats consumed ~100 kcal/day/rat except for CAF-fed rats who consumed significantly more of about 30% additional calories.
  • Rats on the CAF diet consumed the standard chow at ~15% of that consumed by rats fed SC only.
  • HFD-fed rats, unlike CAF-fed rats, significantly decreased total gram intake of the calorically dense diet to normalize caloric consumption by week 2, an effect commonly seen in rodents placed on HFD
Not surprisingly the CAF rats gained the most weight.  Below is a condensed version of one of the figures showing just body weight change:
I find it interesting that both the sucrose rich LF diet (essentially isocaloric to SC) and the fat rich HF diet (consuming notably fewer calories) both led to significant weight gain vs. SC.  There wasn't a significant difference in the weight gains for LF v. HF, but gains were a bit higher for HF whereas calorie intake would predict them to be lower.   For rats, at least, this sounds like a bit of a metabolic advantage goes to sugar over fat, eh Dr. Eades?     In any case, I'm not surprised that either the HF nor the high sucrose diets caused weight gain vs. the standard chow.  Simple sugars and fats would be more efficiently digested/absorbed in either case -- food efficiency -- and converting carbs to fats requires energy (~25%) while dietary fat easily becomes body fat (no conversion!!).

Now the old cafeteria rat is off the charts!  Every rat in this group gained more than the fattest standard chow rat.  

I would make one additional observation:  Look at the spread of the body weights at the end of this study.  The rats eating the bland standard chow were pretty tightly clustered and gained about the same.  Given a suitable diet (even crap-in-a-bag as Peter/Hyperlipid would call it) there wasn't much individual variability.  

The rats in the other groups had much more of a spread in their weight gain indicating much more individual susceptibility to high fructose, high fat or low protein high fat/fructose junk.  The LF and HF chow fed rats were still somewhat clustered at the low end of weight gain with a few that gained a lot.  The CAF fed rats were pretty well distributed and here is where the most individual variability seemed to come into play.  

The rats don't have other external factors such as the attractiveness of their figures or healthfulness of their diets at play so they choose the CAF crap over the bland chow because those are their only choices.  Us humans have more choices.

Can't help but think of that new book Taubes is peddling as I look at the results here either.  The HF rats ate the fewest carbs by a long shot of all groups, and yet they are considerably fatter than the SC rats (and nominally fatter than the LF rats especially when considering they consumed fewer calories).  And since it's now the refined carbs supposedly wreaking all the havoc, one can't help but notice that the rats eating the most simple carbs fared no worse and perhaps a bit better than those eating less than half as many.

Of particular interest is that the CAF rats demonstrably preferred tasty snacks to the standard chow (surprise surprise??)  Could it be there's something to this palatability stuff leading to overeating after all?  I used to know a food taster for a major supermarket chain.  It's really no secret, nor something particularly nefarious, that food manufacturers and distributors try to maximize palatability and thus consumption/sales of their products!  

Even more interesting, however, is that the array of snacks varied from some super-high carb/sugary stuff to super high fat.  And yet ... of the total increased caloric intake (245 cal), while simple carb consumption almost tripled vs. SC (95 cal v. 36), total carb consumption actually decreased slightly for CAF (104.3 g) v. SC (110.5 g).  The increase is attributable to ... drumroll ...increased fat intake:  392.5 cal (CAF) vs. 69.4 cal (SC) for a difference of 323.1 cal ... exceeding the caloric difference (IOW % fat and total fat went up).  So much for overeating carbs ... huh?

Can Gary Taubes please explain how this is possible?  

To me it is simple why we overeat when we choose to eat crap.  Look at the protein column of the snack table compared to the standard (0.25g protein/g chow containing 3.87 cal,  about 26% protein) chow.  Only a few items on the list come anywhere close (we're talking 10-20% of the absolute protein content in most cases, and those where protein is high or higher come with considerable fat calories attached).  Passive overeating .....

Disclaimer:  We are not rats.

Comments

Stephan Guyenet said…
Ha, you beat me to it. I think this is a very important study, not because it contains any new information (caf diets are decades old and it's always been the best dietary way to make a rodent fat and sick), but because it contains so much information in one place.

High-fat refined diets generally increase adiposity in rodent models relative to low-fat refined diets. The HF vs. LF comparison in this study is atypical, although results like that have been reported before. Rats generally are a lot healthier on standard whole food chow than low-fat refined diets. No one studies the health effects of unrefined high-fat diets in rodents to my knowledge.

I agree the rodent lit as a whole is a major challenge to the idea that postprandial insulin release due to dietary carbs leads to obesity. This study and many others lead to a different conclusion that you mentioned: palatability. This is going to be a big topic of mine in the coming months.
Matt Stone said…
I'm tryin' out one of them good old-fashioned high-sugar diets right now myself, instead of my usual starchy fare. Let you know how it goes. I have already noticed things that are opposite to what rat studies show on a high-sucrose diet experience, such as a radical increase in vasodilation (rats show decreased vasodilation). So we'll see.

Here's some interesting reading for you. It shows more lean mass and less body fat on a LF vs. HF isocaloric overfeeding regimen, and explains the differences. There is a huge "metabolic advantage" to carbohydrate. Carbohydrate overfeeding raises the metabolism much more than fat overfeeding does...

http://www.ajcn.org/content/62/1/19.full.pdf+html
Matt Stone said…
Yes Stephan,

Palatability is a huge factor, and I think indicates strongly toward the possibility that pleasure centers in the brain can be one of the overriding factors in raising weight set point. However I noticed that you thought the potato diet guy had lowered his weight set point, but in reality when one returns to more palatable food the weight set point climbs back up again, often eclipsing the old set point with a new and higher water mark. You can read my thoughts about some of this from last year...

http://180degreehealth.blogspot.com/2010/08/dopamine-and-weight-loss.html
MM said…
CarbSane,

"...the fat rich HF diet (consuming notably fewer calories) both led to significant weight gain vs. SC."

How is this possible? According to the table the SC group was getting ~130 kcal/wk more than the HF group. Did the HF diet somehow slow their metabolism? I don't suppose that kind of data is in the paper. I imagine all the food was meticulously weighed so I doubt they made a mistake here.

Thanks for the great post.
Christian said…
Very interesting. Thanks for posting it.

Palatability is definitely a major contributor to the question of how much we eat of a particular foodstuff. If you take a milkshake full of fat and sugar and make it taste bitter or otherwise crappy, no one will drink it. Unfortunately the palatability of the SC, HF and LF chows are more or less impossible to assess - as is "absolute" palatability in general.

"Can Gary Taubes please explain how this is possible? "

Unfortunately it is still a shame that you are not able to present such studies without judgemental references to Taubes.

Controlling variables is key - as always. If you imagine the same experiment with humans, what conclusion can you seriously draw if you give one group cookies, burgers and french fries ad libitum and the other group high-fat meals or low-fat meals ad libitum that possibly taste like cardboard? Hmm ...
Nigel Kinbrum said…
"Apparently the main problem with Heroin* is it's rather moreish." Harry Hill (UK "Comedian")

*or Cocaine, or really yummy food.
John said…
Matt,

Based on the paper you referenced, I would say that there was a huge initial "metabolic advantage" to carbohydrates, which lasted less than 14 days, and only applied to lean subjects. The paper clearly shows that the "metabolic advantage" was highest at the beginning of the study. At the end the carb group was storing similar amounts of fat as the fat group due to a steady drop in fat oxidation (page 23).
The obese carb group gained more fat mass than the obese fat group (table 2). For the lean subjects it was the other way around.

The increase in lean mass for the carb group doesn't seem too surprising, as that group would've had much more glycogen than the fat group. Given the hydrophilic nature of glycogen, more lean mass would be expected in the carb group.


John
Anonymous said…
i dont think 'overfeeding' studies are of much help to anyone. for one, why is someone overfeeding themselves in the first place as opposed to listening to their body and eating? overfeeding studies shouldnt be of such an interest.

palatability was a huge factor in my ability to gain weight. if everything tastes the same, i dont care and a few bites fills me up. a big plate with a variety or tastes and textures and temperatures made it a lot easier to eat to gain
Bill said…
In mice, at least, a refined low-fat diet is not really "a lot healthier" than standard chow. Lower body weight? yes, but also more insulin resistant.
CarbSane said…
Welcome William! I don't think LF was a good designation for that formulation in this case as it wasn't really that much lower in fat than the regular chow (10% vs. 12%). It was a diet comparable in composition to the 45% fat HF diet with 35% of the fat in that diet substituted for sugar. With all that fructose it is surprising these rats didn't become insulin resistant or prediabetic as did the CAF rats.
CarbSane said…
@Stephan: One of my peeves with this whole insulin theory is that it concocts a complicated explanation to that which can be explained by simply looking at what obese people are eating. Having been one, I can attest, to get obese I simply ate too much. When one eats low protein high calorie foods, it is easy to overeat. There's a lot of science that goes into processed food development to enhance palatability but fat+salty and fat+sweet are time tested combos that work!

@Christian: Judgmental? No, I would like Taubes to actually address studies like this one and put his theory to the test he keeps saying nobody does. The HF diet in this study turned out to be pretty low carb (35%) vs. the standard chow diet (64% carb) and yet they got fat on the diet. We're told that it was when we started replacing fat in our diets with carb that that's when we got fat. Well, the LF group replaced 35% of the fat with sugar and they ate a little more but didn't get any fatter - on average they got nominally less fat than the HF rats. Of course now we'll get the "we are not rats" dismissal. But it seems many of us behave just like these rats choosing to eat tasty, protein deficient, high FAT & calorie foods to excess.
CarbSane said…
I'm going to edit this into my post but it's not enough of an edit to bump it so I'm also posting it as a comment:

I would make one additional observation: Look at the spread of the body weights at the end of this study. The rats eating the bland standard chow were pretty tightly clustered and gained about the same. Given a suitable diet (even crap-in-a-bag as Peter/Hyperlipid would call it) there wasn't much individual variability.

The rats in the other groups had much more of a spread in their weight gain indicating much more individual susceptibility to high fructose, high fat or low protein high fat/fructose junk. The LF and HF chow fed rats were still somewhat clustered at the low end of weight gain with a few that gained a lot. The CAF fed rats were pretty well distributed and here is where the most individual variability seemed to come into play.

The rats don't have other external factors such as the attractiveness of their figures or healthfulness of their diets at play so they choose the CAF crap over the bland chow. Us humans can choose otherwise.
Christian said…
Carbsane, I am not sure if you would get the "we are not rats" dismissal on this study, but the "control your variables" dismissal like I said.

"Having been one, I can attest, to get obese I simply ate too much."

So why did you eat too much then? ;) Well, probably because a) you were hungry and/or b) because you craved tasty food and/or c) something else. These would be actual explanations for your observation that you ate too much. And to dig deeper we could hypothesize that a) i.e. too much hunger could have been induced by sick hormonal regulation or b) food cravings could have been induced by palatabilty of foods and the effect on your Hypothalamus.

Eating too much is not an explanation for your observation that you gained weight: cum hoc ergo propter hoc.
CarbSane said…
@MM: The metabolic advantage gang always cites Jequier's thermogenic factors (don't quote exacts right now 2-3/fat, 6-8/carb, 20-25/protein). Thing is, protein about the same, carbs burn more. Rodents do convert way more carb to fat (stored) than we do - an energy utilizing/thermogenic process. So fats are easily socked away, carbs not so much.
Sanjeev said…
> So why did you eat too much then?

So what do all your proposed theories, standing alone without conservation laws tell you?

Nothing

"I was hungry so I ate" tells you zero.

Your isolated factoids are less than useless. Even less useless than Taubes' cherry picked BS.

"I was hungry so I ate" PLUS conservation laws explain everything. It's the only way individual factoids make sense.

> Eating too much is not an explanation for your
> observation that you gained weight:
> cum hoc ergo propter hoc.

Constant repetition won't make it true. Knock yourself out trying though.
Sanjeev said…
> Unfortunately it is still a shame that you
> are not able to present such studies without
> judgemental references to Taubes.

Taubes is the elephant in the room presenting cherry picking as if it were science.

The issue must be spoken to, even if the only thing it achieves is for someone to come to this article from Googling Taubes.
Sanjeev said…
malpaz wrote:
> i dont think 'overfeeding' studies are of much help to anyone. for
> one, why is someone overfeeding themselves in the first place as
> opposed to listening to their body and eating? overfeeding studies
> shouldnt be of such an interest.

wow. Something I completely disagree with.

I think eating without thinking is the standard behaviour, and I suspect eventually it will be found to be far more important than the fevered guesses about insulin. I won't go cherry picking studies to prove my point though.

Some natural human overfeeding:

Watching TV while polishing off a bag of potato chips (2,000 calories) or a package of oreos (even more calories). Is anyone going to claim one's insulin forces people to do this?
Eating to relieve boredom
Eating while driving
buying the large of popcorn in a movie theatre, thinking you'll only eat until you're full, but your hand keeps reaching in and the popcorn's gone by the movie's end
Sanjeev said…
Matt Stone said...
I'm tryin' out one of them good old-fashioned high-sugar diets right now myself, instead of my usual starchy fare. Let you know how it goes.
__________________

Lyle Mcdonalds' "UD2" is a GREAT diet to experiment with.

but your post reminds me of some of my experiences

Oversimplifying, There is a week long low carbohydrate and exercise phase which is meant to deplete one's muscle glycogen. This is followed by a 1 day massive carbohydrate over-feeding

I used to keep my living space at the time really cold and the first few times I did this carb load the place felt incredibly hot. I had to get rid of all my blankets when sleeping.

And similar to your experience, my veins grew incredibly prominent. I really notice this because for most of my life my veins have been invisible. The phlebotomists where I donate blood used to have to stab me repeatedly to find a vein.
Christian said…
Sanjeev: "I was hungry so I ate" tells you zero.

That's why I didn't propose or said that.
CarbSane said…
Christian, I would add that I do not plan to ignore Taubes, especially because he has this new book out that perpetuates many of the GCBC myths. I did not find this study or go looking for it to discredit Taubes, but it just so happens to do so pretty soundly and irrefutably. You see, the rats were not free living humans like in Shai. What they ate and the results on body mass could be carefully controlled, monitored and analyzed. While there are differences between humans and rats, I think we can all agree that the actions of insulin in response to carbs are the same. In his latest radio interview with Dr. Oz, Taubes states carbs make us fat and insulin makes us fat. Any study that shows otherwise that is discussed here will likely reference this nonsense. Because he now repeatedly calls CICO nonsense when it is his cherry-picked theory that in actuality is.
MM said…
CarbSane,

Thanks. I did some looking and I did find this study about that effect seen in Wistar rats. I can't get the full text, but here's the abstract.

http://www.ncbi.nlm.nih.gov/pubmed/20558837

It seems the high fat diet caused, or maybe just resulted in leptin resistance.
MM said…
CarbSane,

I seem to be slightly obsessed with these stupid Wistar rats. For some reason I found the leptin resistance explanation unsatisfying. So, I did a little more searching, and found a couple more studies. First, I found this one confirming the leptin resistance and looking at the actual gene expression in the hypothalamus. http://www.ncbi.nlm.nih.gov/pubmed/18721800

Second, I found this much more interesting and satisfying explanation in my opinion. http://www.ncbi.nlm.nih.gov/pubmed/18655006

In this study even though the rats lost total body weight they gained significant fat mass. It was 2.4 times the control after only 1 week. I wish I had the background to totally understand this study, but this sentence caught my eye.

"The visceral lipogenesis was supported by an increment in adipocyte PEPCK, aiming to provide glycerol 3-phosphate to triacylglycerol synthesis and this fat accumulation was accompanied by glucose intolerance."


(I can get you the full text of either of these studies if you want them.)
CarbSane said…
Do please send them through my profile MM! In the 2nd study they gained less, didn't lose but gained more fat. This is the opposite of what "insulin guided" fuel partitioning is supposed to do ... right?
MM said…
Emailed you the files. As to your question, uh... this is where my background fails me. :) But yes I think this whole study flies in the face of the insulin hypothesis.
MM said…
Looking at it again I realized I said the rats lost more weight, when I should have said gained less. This wasn't a weightloss study. Even so the rats were getting zero carbohydrates and put on more body fat.
Christian said…
CarbSane, this rat study definetely provides evidence against a purely "insulin-based" theory of obesity. But - as I would argue - it simply adds another very potent factor, i.e. the palatabilty of foods. Unfortunately, this "factor" is everything but new. My point is that if you think this study contradicts the notion that hormonal dysregulation of insulin -can- contribute to obesity - you simply have to pay more attentention to "controlling variables" when interpreting and comparing studies.

And again it still seems to me that you do not make an effort to distingish between CICO and insulin-based theories of obesity, while it is clear, that those two concept are not related to each other. CICO is called nonsense for a particular reason (independent of any theories about insulin causing fat accumulation and leading to obesity or studies contradicting this notion).
CarbSane said…
it simply adds another very potent factor, i.e. the palatabilty of foods. Unfortunately, this "factor" is everything but new.

No, it's not new, but the "engineering of foods" for this is, especially the degree to which folks include these REGULARLY in their diets.
Lisa said…
Hi CarbSane,

I love your blog even though I don't understand most of the science. I've been thinking about this study a lot. I've been wondering what do you think would constitute human standard chow? We're really picky about the quality of our ingredients, but I think the food we make for ourselves from scratch is just too tasty.

Lisa
CarbSane said…
Thanks Lisa and Welcome! I've actually got a "become a foodie" post in the works over at my other blog (should publish in a day or so). It may sound counter-productive, but I think it's more the "engineered palatability" of common SAD junk foods (in conjunction with protein and micronutrient deficiencies) that are the problem, and not cooking tasty foods.
Anonymous said…
Just wanted to throw this out - have you looked at any of the studies about how food variety impacts quantity?

The whole 'imaginary diet' study was pretty interesting in that it indicates there's a definite mental aspect to satiation - http://www.sciencemag.org/content/330/6010/1530.full

Also some of the buffet studies... palatability may be one factor, but the caf diet is also bringing in variety.

Also the SC diet had the most protein and caf the least. ;)

And hi! Been reading a while, haven't posted before ;) ... And I do think there's a big difference between the idea that insulin can _contribute to_ obesity and "insulin makes you fat". Certainly in this study, if insulin was a contributor to obesity (disregarding the insulemic effect of protein), it wasn't the primary factor.

I may also be biased by the fact that it seems like a lot of the people on the internets who say that CICO is meaningless also claim to eat 5000+ calories a day and to never count calories... at the same time.
CarbSane said…
Hi JJ! Thanks for reading and coming out of the internet-woodwork to contribute! Haven't looked much into the eating habit studies re: variety, but instinctively I think it plays a role in most.

I think if folks ate at McD's every day and could only eat burgers and fries they may well not overeat calories after a few days. I think there's something to the thought that food boredom plays a role in LC weight loss for some.
Sanjeev said…
>> have you looked at any of the studies about how food variety impacts quantity?

Brian Wansink's bailiwick.
http://www.brianwansink.com/

(not "Life of Bwian" Wansink, just Brian Wansink)
the one I most remember from his book was the study of multicoloured jelly bean (could have been M&Ms) bowls versus monochrome.

People ate more from the varied bowl than the mono-bean bowl.

There was also stuff about food colouring towards the book's end, I've forgotten specifics. something about fooling people into thinking they were getting lemon jello even though it was cherry jello with yellow food dye.
Lisa said…
CarbSane, I'm looking forward to that "become a foodie" post. I hope you're close.

I would like to see a study like this in rats and/or humans testing four ad libitum diets with 1) variety (not palatable), 2) palatability (no variety,) 3) variety and palatability, 4) neither. In the not palatable groups I wouldn't want yucky foods, just not particularly tasty ones. Clearly the 3rd diet would cause the most weight gain but I'd like to see the effect size compared with diets 1 and 2.
RRX said…
Hi Sane,

I have a question about this that kind of relates to Matt Stone's comment. The "metabolic advantage" that he mentions and cites is in relation to overfeeding. How does that relate to eating at maintenance calories? The study discussed in this post and in Stone's comment seems to me to matter when understanding weight gain. How about when maintaining weight? Have you come across studies that have looked at how variations in macronutrients affect fat deposits in that situation? Do they just stay as they are by the time weight stabilizes? I'm past the weight loss part (300 lbs down to 165 lbs) of my life and have been maintaining for years. My interest is departing from losing or gaining and is becoming more about studies on maintaining. I've pretty much lost interest in the Taubes v Everyone Else situation due to having maintained my weight loss for years now. For me, his or anyone's explanation of what I have already done is useless to me. I wonder if you can direct me toward any studies you know of that address this aspect (maintenance) of diet after a loss. Thanks!
CarbSane said…
Hi RRX, Unfortunately there aren't a whole lot of studies on maintenance. Most of the studies on the formerly obese tend to be a few months post weight loss. I do believe that in terms of regain (or better preventing it), carbs do have an advantage over fat b/c they do stimulate burnoff of occasional excesses.

I watch my fat intake a bit more these days as I eat a few more carbs. It's been a bit of an odd adjustment to recognize how much (volume) food I can eat.
CarbSane said…
BTW, CONGRATS!!! That is quite an accomplishment to have shed that much weight and KEPT IT OFF.
Anonymous said…
Prepping and Using Your Fleshlight
Adult ToyThe proper prep can make a difference to
the quality of the teaching or instruction has consistently
been lauded, if at times inconsistent.