Saturday, December 11, 2021

Protons (67) a formula revised for butter oil

A couple of months ago Tucker emailed me this study

Docosahexaenoic acid lowers cardiac mitochondrial enzyme activity by replacing linoleic acid in the phospholipidome

The study itself, while interesting, is not the point. The point is that the diet used contained the infamous 42% of calories from fat and was obesogenic. In common with a number of other studies I have been mulling over, the fat was butter oil. This is the butter oil composition

which has 2.3% linoleic acid in 42% of calories giving just over 1% of calories as LA. All I am interested in here is the comparison between 14 weeks on control low fat diet (CD) vs 14 weeks on a low linoleic acid Western Diet (WD). Here are the fat mass data (bodyweight mirrored this)

and here are the IPGTT results. We're just comparing the black control (CON) with the red western diet (WD) lines

So I think we can say that 14 weeks on butter oil is obesogenic and has the appropriate insulin resistance as we expect from any obesity with its increased basal lipolysis of large adipocytes coupled with Protons effects on the non adipose tissues.


The diet looks like this:

If anyone thinks this looks like fudge, here is the recipe for a kilo of the stuff

Take 340g of table sugar, add 200g of anhydrous butter fat, bulk it up with 150g of corn starch to help it set and throw in some casein if you feel like it. Mix and extrude (you could cut in in to cubes for human consumption).


Looks yummy. Rewarding. Don't snigger!

End aside.

Butter oil is one of those features of study designs which produce obesity without linoleic acid.

Again, prompted by Tucker, I re-drafted the F:N ratios of MUFA and PUFA in the last post to account for the consumption of one NADH to provide an NADPH for rearranging each double bond during beta oxidation and we can add these revised ratios and some for selected saturate values to the composition table of butter oil like this:

In red are linoleic acid and the short chain fatty acids of an equal or lower F:N ratio cf LA. I threw in oleic acid and C8 caprylic (blue numbers) too to point out that caprylic acid, though higher than LA, is now lower than oleic (I view oleic acid as the mammalian default for "normal" insulin sensitivity) and so might be obesogenic.

So, from the F:N ratio Protons perspective, we have a modest supply of short chain fatty acids of potentially greater insulin sensitising ability (hence obesity promoting) than linoleic acid itself.

The effect of butyrate as a dietary supplement on obesity is controversial and reviewed here

Butyrate: A Double-Edged Sword for Health?

Conclusions totally depend on how you set your study up, what you consider good and what you consider bad. Bear in mind that butyrate is the darling of fibre-philes so consider publication bias too. Conversely it is to a large extent consumed by the colonic epithelium, so not a lot gets through to the systemic circulation. But some clearly does. The snippet of Figure 2 which caught my eye was this section:

I would just point out that anything which decreases lipolysis and increases adipocyte glucose uptake is NOT going to make you skinny. It might make you insulin sensitive (bravo), at least until you get fat enough to leak excess FFAs via augmented basal lipolysis.

You could of course just say butter oil fudge is highly Rewarding, so makes rats and mice fat. How can you tell it's Rewarding? Because it makes rats and mice fat. Except corn oil is also very Rewarding but ad-lib preferential consumption fails to induce obesity. For obesity there is the absolute necessity of calories to enter adipocytes and then stay there. Long term. Dopamine release in the brain might make you choose to eat something over something else but without pathological energy storage... Shrug.

My own concept of how butter oil/sucrose causes obesity is limited by the clear fact that there is no way of simply saying a given F:N ratio will always produce obesity. Too many variables for this to be set in concrete, which allows the hypothesis to side step conflicting evidence. You have been warned.

Random thought: Sucrose (when it doesn't produce a slim insulin sensitive phenotype) usually produces hyperinsulinaemia and insulin resistance (often "skinny fat"). The higher the insulin levels the more effective insulin sensitising dietary components (linoleic acid and now possibly SCFAs) are at allowing that high level of insulin to generate obesity. Probably why the cornstarch is added to the fudge, to augment the insulin response.

As always, alternative explanations welcome.

My thanks to Basti in the comments after the last post which crystalised a lot of this current post. And to Tucker twice over.



Tucker Goodrich said...

Wow. Cool... I will need to ponder on this one!

cavenewt said...

That recipe took me back to the days when I baked. Substitute wheat flour for the cornstarch and you've got cookie dough. Rewarding indeed!

Peter said...

I believe the "cookie dough" was originally TM-ed by some obesity researcher who's written some book about Reward and the like. Dunno really.

Tear it apart please Tucker. I'm too biased to see if it is reasonable. Obviously I think it's okay, but then, I would...


Lucho said...

Hello Peter, great post.
Based on this new "hypothesis" that SCFA seems to be obesogenic... what would you say is the best type of meat to stay lean? I know that beef, lamb and goat are all high in oleic, palmitic and stearic...
What do you think?

WilliamS said...

So butter is fattening?? I’m selling all my stock in GT (Gary Taubes) immediately. Also dumping all shares of FB (Fire in a Bottle).

Peter said...

Lucho, the spilt seems to between dairy and ruminant meat. There are lots of reports of stalls in weight loss helped by dropping dairy. And the more ketogenic you are the less it matters.

WilliamS, I think it's a matter of degree, if I'm correct. And the model discussed has to be explicable, I can't let it go as "just a rodent study"...


BigWhiskey said...

Well, dang!

Passthecream said...

That is very close to what I once suggested as a nougat diet. Just change the protein to egg white and add a few nuts.

I also still think that such a high portion of energy as fructose via the sucrose, combined with fat plus glucose from both sugar and starch, is innately fattening. I think any significant carb intake on a high fat diet is going to be dangerous s.t. if you can't cut the carbs then drop the fats. What concerns me more is where does all the circulating glucose end up that isn't being forced into adipocytes?

A: Eyes, ears, kidneys, joints, endothelium, RBC, AGEs, etc It oozes out however and wherever it can.

Tucker Goodrich said...

We need to remember, these C57BL/6 mice, despite being advertised as "wild-type" are not. They are selected to be susceptible to obesity.

"But because it uses the broken C57Bl/6 as an example of defective respiration it does bring home how irrelevant this particular mouse might be to more humans with a more normal metabolism."<\ahref="">

Peter said...

Tucker, I have many thoughts and some I write down and some I don't. I have wondered for quite some time whether the failure to assemble ETC super complexes by Bl/6 mice results in less, rather than more, ROS for a given input to the ETC of an ATP replete cell. This would be the equivalent of feeding a low F:N PUFA when you are actually oxidising a high F:N saturate. However butter (usually with sucrose) seems pretty fattening in Wistar rats too which, as far as I know, have normal functional SAPs...


Tucker Goodrich said...

"Rats are used in DIO studies (Table 2). Sprague–Dawley rats are considered a good model for inducing obesity through diet, since they have a behavior similar to humans with regard to excessive food consumption, which can cause weight gain and changes in lipid metabolism [22]; however, Wistar rats are more susceptible to the development of obesity through diet, since they usually consume a higher amount of high-fat diet than the Sprague–Dawley. Also, differences in lipid metabolism, as fatty acid uptake and lipogenesis, as well as the interaction between genes and diet, make Wistar rats more susceptible to DIO [19]."

“Diet-induced obesity in animal models: points to consider and influence on metabolic markers”

I don't know the answer, but lab economics takes precedence over model validity.

Peter said...

Yes, until whatever genetic defect triggers excess fat storage actually messes with your own (possibly not obesity related) research, no one is going to work out what is wrong with a given strain to make it obesity prone/resistant...

Though I wonder if this has been looked at as a drug target opportunity? Dunno.


WilliamS said...

What is the F:N ratio for butter as a food, i.e. considering all of its constituent fatty acids, both short and long chain?

WilliamS said...

Since the short chain fatty acids are such a small percentage of the overall fatty acid content I wonder if, as a whole, butterfat is still pretty good from the protons perspective.

Anecdote. I was quite lean before Covid hit in my area (longtime low carb/paleo eater). I always ate out very frequently. (A man needs his vices.) When society got shut down restaurant food consumption dropped dramatically. As compensation I added half a pint of full fat Rebel Creamery ice cream to my daily routine. At home we cook with butter and ghee and I eat quite a bit of full fat cheese as well. So a lot of dairy fat, along with plenty of ruminant fat.

With no effort whatsoever I lost about six pounds and nearly another pants size during the siege. When society gradually reopened here, and I ate out more and more, the weight crept back on. Then I renovated my kitchen and went to 100% restaurant food for a month. Even more weight came back, and more quickly.

More recently, as an experiment I voluntarily cut back quite a lot on the restaurant food again, retaining the Rebel Creamery compensation. Like clockwork the weight began to decline again and continues to do so. I’m almost back to where I was at the height of the siege.

My guess is that the main factor here is some combination of zero seed oil and generous saturated fat with home cooking, but the unavoidable seed oils and fewer saturates in restaurants. Not sure what else could explain this pattern. Carbs remained about the same. In fact the most recent period of weight loss coincided with adding back in sweet potato, with copious butter, which had no discernible negative effect.

So, for me at least, lots and lots of dairy fat seems to have been slimming even with little body fat to lose. Hence my question about the overall effect of butterfat from a protons perspective.

Still those butter oil mouse experiments do need to be explained. I know it’s out of fashion these days, but evidence contradicting hypotheses should get all of our attention.

Tucker Goodrich said...

@WilliamS "...but evidence contradicting hypotheses should get all of our attention."


raphi said...

@Peter and @Tucker,

2 years ago I made Wistar rodent model of schizophrenia by injecting the pregnant dams with methylazoxymethanol acetate. I fed those offspring and WT Wistar rats a ketogenic diet (Dyet 104425; AIN-93G purified rodent diet with 30% ketone ester by calorie) or a low-fat control diet ad lib for 5 weeks.

This is what the weights of the males looked like:
This is what the weights of the females looked like:
This is composition of the ketogenic diet:
This is composition of the low-fat chow:

The ketogenic rats tended to weigh less. And no, I don't know what their body composition was like unfortunately

Malcolm said...

Did you test the schizo symptoms? (how?!)

Tucker Goodrich said...

That's a fairly dreadful keto diet! Starch + ketones?

Kajus said...

High deuterium hide cvirus in the bodyfat!

Peter said...

Tucker and WilliamS,

Thinking about this lots.


Peter said...

raphi ditto!


raphi said...


I did the novel object recognition (NOR) test over 1 hour and 24 hours which is looking at memory/cognition over those time periods. I also did the social (SN) test.

- only my WT saline rats were paid normal/greater amounts of attention to the novel object in the 1 hour NOR test (the rest were uninterested in the novel object, which shouldn't be the case)
- both WT saline and WT aripiprazole rats did better than WT olanzapine rats on the 24 hours NOR test
- 11.24% of variation in my results are attributable to the 'drug x model' interaction
- 7.19% of variation in my results are attributable to the drug alone

overall my results are inconclusive. either I because
- I did something wrong
- the tests aren't precise enough to show meaningful differences
- the drugs suck
- the rodent model sucks
- a combination of all/some of the above

what's certain is that my PhD supervisor who's a prescribing psychiatrist in a hospital was not happy with the drugs looking not that useful

raphi said...


There was 630 kcals of soybean oil out of 3,765 kcals total (for 1 kg) so that 16.7% soybean by kcals - dreadful indeed! I'm less worried about the carbs kicking them out of ketosis as 30% monoester by kcals is pretty powerful. But when I run their blood samples I'll be able to tell you how ketogenic the diet actually was - I'm betting > 1 mmol/L, but we shall see!

Tucker Goodrich said...

I don't think it's possible forthe carbs to kick them out of ketosis with a ketone ester.

That's then problem, it's a completely non-physiological diet.

But curious to see the results...

raphi said...


non-physiological is the right way to put it. I got the ketone ester diet for free from Donna Herber (worth $5k!)

Tucker Goodrich said...


I commented the other day on how lab economics drive animal models to non-physiological states that may not correspond well to humans. LOL. Thanks for the demonstration!

Passthecream said...

Tucker, that's a bit weird. It raises the definitional question of what do you mean by ketosis exactly, since you still have all the carbs swishing around, going places, maybe mostly by concentration grafients. Insulin is ... where, what, how?

IRC that ketones inhibit fat burning too.

Perhaps it would be more interesting to be exploring the behavioural consequences of uninhibited fat burning, which is incidentally what I would usually think of as ketosis? But I can see the relevance of testing ketones as nutritional, possibly prophylactic supplements.

Also I remember the effect of cosuming distilled low weight sfa from coconut oil, quite a bad hypo episode plus a case of the munchies. I wouldn't have been paying much attention to novel stimuli either!

Passthecream said...

Peter, did you say a while back that it's difficult to get rats into a ketogenic state?

Passthecream said...

Maybe before trying diet experiments on lab animals it would be useful to test the ingredient mix on humans (eg lab techs?, Scientists?) who can communicate their subjective experiences more accurately?

Subject to ethics approval of course!

Adrian said...

Hi Peter,
Thats a very great post that also somehow induced the following thought in me...

its often said, that insulinsensitivity is reduced late in the evening/night (usually argued via higher post prandial blood sugar levels).

If this is correct,
1. do you think its rather the adipose which is insulin resistant or the lean tissue?
2. do you think it might be more slimmming to eat later in the evening compared to during the day?

Im really looking forward to your opinion.