Effects of consuming a high carbohydrate diet after eight weeks of exposure to a ketogenic diet

Thanks to Leon for finding this :)

Effects of consuming a high carbohydrate diet after eight weeks of exposure to a ketogenic diet

Basically they took two groups of rats and fed them (ad libitum) either a ketogenic diet (KD) or regular chow (CH) for 8 weeks.  Then they switched the diet of the KD group to CH (KD:CH) and continued to feed the rats for an additional 8 weeks.  As this was a rat study, the rats gained weight/grew for the entire 16 week duration of the study.  

The diets:  KD = 5% carb/15% protein/80% fat             CH = 60% carb/23% protein/17% fat

Now I'm not a rat, but I do find these results both interesting and a bit surprising.  Here are the graphics for caloric intake and weight:

caloric intake: 

Surprise #1:  The KD rats consumed pretty consistently fewer calories (although the difference is not noted with an * for statistical significance) for the first 8 weeks and yet gained consistently more weight.  I'm not sure if this difference was stat.sig. either, but it was mentioned in the results:
After 8 weeks of consuming a ketogenic diet, KD rats had increased adiposity and plasma leptin levels, and reduced insulin, as compared to CH controls. 
So much for the insulin-induced fat accumulation theory once again.  And so much for the metabolic advantage -- if anything, this demonstrates an advantage for the high carb diet.  But CS, doesn't this violate calorie balance?  Nope.   I'm sure there's a TEF or RMR change in the KD rats vs. the CH ones that would explain this seemingly nominal discrepancy.  Furthermore, the KD rats had significantly more epididymal fat (one type of visceral fat) than the CH rats.  Is this a good thing for a rat? .....

Surprise #2:  Initially after the switch, the KD rats didn't go on a soaring blood glucose fueled carb binge.  They actually reduced caloric intake significantly in the first week after the switch.  Presumably these animals had a bit of insulin resistance and blood coursing with postprandial glucose during this first week of the abrupt switch.  And yet this didn't send the rats into a carb frenzy.

After that "honeymoon", however, the KD:CH rats increased their caloric intake over the rats who just ate the regular chow diet the whole time.  This increase is consistent but does not become statistically significant until the last three weeks of the study.  Is this an overcompensation for metabolically "starving" for extended periods?

The hormone level results are totally antithetical to the whole insulin/fat accumulation theory.  KD rats had significantly higher leptin and lower insulin levels than the CH rats.  However this would predict lesser adipose accumulation in KD vs. CH right?  Not the case.  The leptin remained elevated for several weeks (4) following the switch for the KD's.  REPEAT:  lower insulin in KD correlated with increased adiposity (visceral at that).

From the discussion:
Whereas KD rats had significantly increased fat pad weights and plasma leptin levels as compared to CH rats, resuming the chow diet prevented a further increase in adiposity and leptin over time. Rats that consumed chow for the entirety of the study increased fat pad weight and leptin to resemble those of KD:CH rats by the end of the study. In addition, plasma insulin levels in KD:CH rats were not different from CH rats one week after returning to the chow diet, although it was significantly increased after 8 weeks of consuming chow after the ketogenic diet.
If I had my druthers:
1.  A third group kept on KD would have been interesting.
2.  Since they started to see this significance towards the end of the study, why not extend it a bit longer?

Personal aside:  Since I adopted my "cheating" method for LC weight loss and maintenance, this result is consistent with my experiences with going off LC.  In the past, I never "gave up" on LC, just strayed from the strict version of it and never got back to it.  I think there was a small honeymoon period whereby everything still fit and I didn't seem to gain so there was no big motivation to jump back gung ho on the wagon.  (I have no idea if the scale said so or not, but I'll go by size and ignore if there was any water fluctuation in there).  However after that brief period I didn't binge but the pounds sure piled back on fast.  The first 100 lb swing occurred in no more than 2 years but could have been mostly accomplished in one!  The second, in roughly the same time frame.  So fast that before I knew it I was back past where I started. I experienced weight losses in the wakes of some of my cheats during my 10 months or so of rapid losses.  I think because for long weekends or a few weeks (vacation) I probably ate less eating higher carb.  But now I'm a smarter cookie.  I know that I cannot continue like that because whatever the cause, I lose my sensitivity to how much I'm eating after a while of consistent carbs.  


Sanddog said…
Rats do not eat high fat diets in the wild. When was the last time anyone heard of a rat bringing down a cow and chowing down on spare ribs?

My point is, it's kind of stupid for humans to feed a rat an unnatural diet and then try to apply the results to humans. Perhaps this silliness is the result of human low fat diets.
Nita said…
Not only that, but the last time I saw a study similar to this, they were feeding the rats 60% of calories from what was essentially Crisco. Hydrogenated vegetable oils that contained 17% trans fatty acids. I don't care what diet you think is healthy; that crap ain't food.

I'd be very interested to see what that 80% fat was, because not all fats are created equal. And certainly not all fats are even remotely healthy, for rats or humans.
CarbSane said…
Welcome Ms & Nita!

While I believe all rat (mice and all other animal) studies must be taken in stride with "humans are not rats", a great deal has been learned about human metabolism (and other functions) through such research. Good animal models allow in vivo manipulations that are either impractical or unethical in humans. In nutritional studies, there are many advantages to using a rat model such as being able to control and monitor the diet and do "long term" studies (in the life of a rat) in the fairly short term.

An 80% fat diet may not be natural for a rat, but it is not natural for a human either. Our paleolithic ancestors didn't tackle fatty livestock and eat just the fat, no butter existed, and they didn't squeeze the fat from olives and coconut meat. The Inuit are an isolated, highly adapted culture, and the fat they ate differs widely from the fats most eat on a high fat diet.

One thing that is certain is that to create an obese rat via diet, a high fat diet is used -- although it usually mimics the Western/SAD diet in the 40's percent range with carb. It is worthwhile noting that in this study, the KD rats didn't gain a lot more than the CH (low fat) diet, e.g. they did not become obese, but it was noted that there was increased adiposity as epididimyl fat. The CH diet was considerably higher in protein, and it may even be that factor at play here.

Keto diets have been known to stunt growth, if anything (epileptic children) in humans. Since most can go off that diet in adulthood, it would be interesting to do a follow-up to see how that effected long-term metabolism.

My own (n=1) experience has been one of marked changes in my fat distribution since my first LC weight loss, and only becoming truly obese after going off the first LC diet. I'm curious why, and the longterm health implications.

I'll write the authors and see if I can't get the diet compositions.
CarbSane said…
Posting for Reader Harry:

Hi there Carb Sane,

I tried to post the following comment on your article “Effects of Consuming high carbohydrate diet after eight weeks of exposure to a ketogenic diet”...but no luck.

Here’s the comment (perhaps you could post it up for me?)

Thank you for the article (and your analysis) CarbSane...very interesting as always (although I do concur with the other comments regarding the applicability of this study to humans; especially seeing as many other rat studies show that they have a very special ability to gain fat in all sorts of hormonal and food-supply situations, an ability not typically shared by humans).

*** I suspect trouble posting comments is due to the length. This is why sometimes I'll make a separate blog post out of lengthy comments I may make. The remainder of the email/comment will be posted in my personal blog HERE***
CarbSane said…
I agree with some of the concerns regarding the applicability of rat metabolic experiments with humans. However not to the extent to which some take it by totally discounting any rat experiments.

It is RARE for a human to get obese on a low fat diet. Just as rats gain some fat mass, but do not become obese on regular chow (usually 15-20% fat). The obesity epidemic was not caused by society going low fat. All of the oft quoted stats demonstrate a slight reduction in the percent of fat in the average diet, but this is because carbs make up a higher percentage of the 300-500 cal/day we have ADDED (on average) to our intake.

The SAD is a high fat, high carb, high calorie diet.

Anecdotally, I put weight on much faster after going off LC and my weight skyrocketed to highs several (60-80+ pounds more) than they ever had previously through countless yo-yo cycles. Back when I did Atkins the first time, I shared that and loaned my books out to several people who also tried it. While they didn't experience the extra weight rebound, they did gain back their weight seemingly more rapidly (and shared that with me).

As I've blogged before, a ketogenic diet is the equivalent to "starvation" -- even though your body is getting protein and fat, the shift in metabolism is the same as that of extended fasting. Perhaps this causes overeating if you go back to eating carbs in any considerable amount.

If this is the case, then Atkins should come with a disclaimer: Don't try it if you don't plan to do it for life! To their credit, the authors of The New Atkins hint that "some" people need to keep carbs low for life. Perhaps that should read "most".
Unknown said…
"An 80% fat diet may not be natural for a rat, but it is not natural for a human either. Our paleolithic ancestors didn't tackle fatty livestock and eat just the fat, no butter existed, and they didn't squeeze the fat from olives and coconut meat. The Inuit are an isolated, highly adapted culture, and the fat they ate differs widely from the fats most eat on a high fat diet."

Nice way to skirt the issue CS. Where is the evidence to support your statement? Fact is the type of fat matters and you cannot conclude anything relevant from this type of animal study. Were it the other way around...

A ketogenic diet is NOT a starvation diet because the calories you are not eating due to satiety come from the fat in your fat cells.

ALL mammals eat a high fat diet as what they eat converts to fatty acids in the bowel. But the animal needs to eat what is natural to it.
CarbSane said…
Welcome to my blog Georgia!

I don't see where I've skirted any issue. I've blogged on "the evidence" previously here and here. I also compared seal oil to beef fat here. As you say, all fats are not alike. Most of the fats I see high fat low carb adherents consuming differ considerably from the omega 3 rich fats consumed by the Inuit -- the poster culture for high fat diets.

FWIW, I do not consider any macronutrient "fattening". Least of which protein, but carbs or fat consumed in the context of whole foods either. The fact remains that when we want to make a rat obese through diet, it is almost impossible to do so on a low fat diet. A high fat diet may not be normal for a rat, but neither is either the SAD (a diet high enough in fat to make rats obese) or OD, etc.

As to my comment re: ketogenic = "starvation", note the quotations. It is a fact that the fat-adapted metabolism (gluconeogenesis, glyceroneogenesis, upregulated ketones and lipid oxidation) mimics that of a fasted/starving person. See for example here and here and some of the references from the subject of that latter blog. There may well be long-term changes in the metabolism of someone following a ketogenic diet whose body "thought" it was starving (hormonally) for an extended period of time.

There are implications here for recommending a keto diet for various reasons. Seems to me that if it is not a diet for life, there are potential implications for weight management, and obesity-related diseases.

As to: ALL mammals eat a high fat diet as what they eat converts to fatty acids in the bowel. But the animal needs to eat what is natural to it.

I'm not sure what you mean by this. It is only soluble fiber that gets converted to short chain fatty acids by gut bacteria, however these fatty acids are not digested, stored and/or metabolized in any way like the long chain fatty acids we commonly consider fats in the diet. I would be glad to discuss this point further if I'm misunderstanding something.
LynMarie Daye said…
Sounds to me like Georgia is confusing ruminants with all mammals.
CarbSane said…
Perhaps! I've seen many references on LC forums that soluble fiber need not be counted as carbs because what is fermented is converted to short chain fatty acids ... aka fat.

This is erroneous because, if we're talking metabolism, the molecular designation of fat by structure is not the same as the metabolic pathways said molecules feed into. I contend the latter is more relevant.

Even MCT's are different from the long chain fatty acids we commonly refer to as "fats". Lauric acid is sometimes packaged into chylos, but some is not. Shorter FA's do not get packed into chylos, indeed a goodly portion if not the majority of MCT's are absorbed intact and transported to the liver to be metabolized immediately. SCT's and MCT's are far more "carb-like" to our bodies than LCT fat.

Soluble fibers contain on average 2 cal/gram energy from the byproducts of bacterial fermentation. This is for all intent and purposes carb, and although will not impact the primary carb concentration (blood glucose), will still suppress lipid (long chain) oxidation.
CarbSane said…
I want to add: I find MCT supplementation an odd phenomenon in the LC community vis a vis weight loss.

If one looks in the literature regarding these fats, there is a plethora of information on their use to maintain weight/prevent wasting in diseases such as AIDS and cancer or conditions where normal fat absorption is impaired. IOW, they are a means to pack a caloric punch while sidestepping the need to emulsify, break down, absorb and re-esterify and package into chylos necessary for LCFA's.
David Isaak said…
Sorry, since when did butyric acid and its relatives become "for all intent and purposes carb"?

So the short-chain fatty acids in butter and milk are, you're claiming, "effectively" carbs?

That's one of the oddest propositions I've ever heard.
CarbSane said…
Welcome David!

I actually have a rather long draft in the hopper that relates to this, but I've got other things higher on the list at the moment. Nonetheless I'll try to reply to the core of the issue.

To me, when our bodies *see* fats, we're talking long chain fatty acids. These must be broken down to FA's + glycerol to be absorbed and re-esterified to TAG's and packaged into chylomicrons for transport. They must then again be broken down into FA's + glycerol so the FA's can be transported into cells (and may again be re-esterified for storage). The LCFA's go through the "fatty acid spiral" or FAS/beta-oxidation breakdown to produce acetyl CoA as the ultimate byproduct. This pathway is the last in the heirarchy of our "fuel mix", so other pathways to produce acetyl CoA suppress beta-oxidation -- most notably glycolysis. If you check out this post, focusing on the fermentation byproducts, you'll see that, although *technically* fatty acids, short chain fatty acids are intermediaries/substrates in metabolic pathways involving carbs. Therefore they will be metabolized as a higher priority just like glucose. Hence my characterization of them being "effectively" carbs.

FWIW, even MCT's are somewhat carb-like in their effects on fuel partitioning in human metabolism. I believe only some lauric acid is transported in chylos, the rest and all shorter MCT's are not. They are directly absorbed and immediately processed by the liver into AcCoA. This would suppress beta-oxidation of LCFA's (and MCT's induce an insulin response BTW).

This is what I mean by characterizing the behavior of these molecules that are, by chemical classification, fatty acids, but, by virtue of how the human body processes them far more akin to carbohydrates.