Does eating fat stimulate fat burning?

Boost metabolism?

Well ... apparently not in rats.

full text in Google Docs HERE

I came across this recently.  The first and last results in the title list were nothing new, but the noting of decreased energy expenditure jumped out at me.  This would be in contrast to claims.  So they fed rats at about pre-study ad libitum levels one of two diets:  HiCHO (69C/10F/21P) or HiFAT (20C/59F/21P) for 4 weeks (24 days prior to cannulation, 3-4 days post-op where applicable).  The rats were cannulated to allow for glucose and insulin infusions to measure insulin response in Expts 1&2.  
Experiment 1 ~ In vivo insulin action:  Insulin was infused at mid-physiological level and glucose was infused at so as to maintain basal glucose levels.  Glucose infusion rate was a measure of whole-body response to the insulin infusion.  At 75 minutes, a bolus of non-metabolizable radiolabeled (H3 and C14) glucose analogs was injected.  Tracer levels were measured periodically for 45 minutes.  At the end of the study muscle and fat tissues were collected and analyzed for incorporation of the radiolabels.  
Experiment 2 ~ Endogenous glucose production:  Radiolabels were used to determine glucose production in rats treated similarly to Expt. 1.
Experiment 3 ~ Heat Production:  The rats were housed in metabolic cages and O2 usage and CO2 production were measured and heat calculated.  This was done for 3 hours to assess basal levels.  The rats were then gavaged (feeding tube) with either a dextrose or a lipid load with care not to stress the animals and the heat production determined for another 3 hours.  The volume and calorie content of the CHO vs. FAT "meals" were the same.
The results and discussion are extensive which is why I'm sharing this through Google Docs for anyone interested.  I'll highlight the major findings that were summarized in the abstract and discussion.
  • In the "pre-experiment" feeding period, all rats gained approximately 70g total body weight regardless of diet, weighing in at around 375g.  However, the distribution of weight was significantly different.  The HiFAT rats were significantly fatter -- roughly 65g fat vs. ~50g fat for the HiCHO rats.  This was at the expense of lean tissue (~61g vs. ~69g for HiCHO) and minor differences in water weight.
  • In the first experiment, they infused a constant amount of insulin (4.1 mU/kg·min) and the glucose infusion rate to maintain blood glucose levels was a measure of insulin sensitivity.  The more sensitive the tissues, the more glucose the given insulin concentration would push into cells, the more would be required to be added by infusion to keep blood glucose levels from dropping.   The effect was profound.  The glucose infusion rate was nearly double in the HiCHO vs. HiFAT rats (10.4 vs. 5.6) repeated in the second experiment (now 10.6 vs. 4.0).
  • The HiFAT diet suppressed glucose uptake from almost 30 to approximately 60% in the major oxidative skeletal muscles vs. HiCHO.
  • Glucose uptake into brown adipose tissue (BAT) was suppressed by 90% which accounted for approximately20% of the whole-body effect. 
  • Basal metabolic rate did not differ, but post meal thermogenesis for the glucose meal (Expt. 3) was considerably greater for the HiCHO vs. HiFAT
  • The HiFAT rats accumulated less BAT and more WAT.
  • Interestingly, HiFAT  feeding resulted in  major decreases in basal and insulin-stimulated  conversion of  glucose to  lipid  in  liver (~30-60%), BAT (about 90%) and WAT (no effect to about 40%).
The authors write:
This  [insulin] resistance is seen at basal insulin  levels,  is pronounced at  insulin  levels in  the midphysiological range, and is accompanied by  reduced energy expenditure leading to obesity. 
Now, it bears mentioning that fat tissue physiology is rather different in rodents and humans.  First, adult humans don't rely nearly as much on BAT for thermogenesis/body temp regulation as small animals like rodents.  Second, the conversion of glucose to fat by de novo lipogenesis is a more significant metabolic path in rodents than in humans.

Still, in this study what happened to the HiFAT rats is everything the TWICHOO predicts should happen for the HiCHO rats.  That is, it is supposed to be carbs and the insulin spikes they elicit that leads to hyperinsulinemia and insulin resistance ... where the fat remains insulin sensitive to the bitter end resulting in fats trapped in the fat cells while lean tissues starve.  By contrast, when carbs are low, the fat elicits no insulin response.  Although the rats weren't super low carb, they still would have lower insulin levels during the almost 4 week pre-feed.  Thus we would expect lower insulin to mean more energy available to the cells, ratties going wild with energy, reduced fat mass, etc.  Perhaps the most surprising result of all for the Taubesians whom I call TWICHOOB's, is that the glucose uptake into fat tissue was considerably more in the HiCHO rats, but they accumulated considerably less fat than the HiFAT rats.

The more carbs, the more insulin sensitive the fat tissue, but the more postprandial thermogenesis carbs generated so less fat was accumulated while more lean was created.  True, we're not rats, but human metabolic ward studies seem to agree with the results here.


Stephan Guyenet said…
Hi Evelyn,

This is typical of rodent high-fat diet induced obesity. They go through a hyperphagic period that lasts 7-10 days, after which their food intake declines back to the control level. Yet they continue gaining fat, due to a reduction in energy expenditure. Food intake may eventually rise again.

You mentioned that in rodents, DNL is more significant than in humans. I'm not convinced it's that different. Typical rodent chow is only ~14% fat. Would they have as much DNL on a 40% fat diet like what is typically eaten by humans? Would humans eating a 14% fat diet long term see an increase in DNL? I don't know the answer to those questions, but I'm not sure there's a clear-cut species difference.
Unknown said…
Hi Stephan,

About DNL and lipogenic capacity in rodents vs. humans:

About DNL in humans:
Justin Cascio said…
It's significant to me that the fat in the high fat diet is derived from safflower oil. In fact, the whole diet in Table 1 doesn't look anything like what I would eat: highly processed foods, mostly of vegetable origin. I'm not going to worry about what this study comes up with; I already know it isn't a good diet they have those poor rats on.
CarbSane said…
Welcome Justin! While you make a valid point, this is both a drawback and convenience of rodent studies. Feeding them real food would lead to variability. So chow it is. I would note that the quality of the fat in high fat rodent diets seems to make little difference. The high fat, low insulin does not seem to produce the fuel partitioning and metabolic effects claimed ... at least in rats. It's possible it does in humans, but more than a few metabolic ward studies say not.
benn686 said…

I've heard it said that if you eat carbs after a no-carb diet period of time, you may gain unexpected weight because the body has to re-adjust to burning it's preferred carb fuel!

So is perhaps fair to conclude that eating a low-carb, hi-fat diet will make your body more efficient at burning fat as energy (not that it converts to glucose)?

Also, I assume weight gain is *strictly* Calories In greater than Calories out, but can the type of foods you eat today, impact how future calories are burned?

If so what diet approach would you recommend? Should specific types of foods be minimized... rice, carbs, fats, fructose, wheat gluten / fiber, etc, etc..?
benn686 said…
Oops forgot the link..
Mirrorball said…
I would like to suggest this study too: