Tuesday, July 14, 2020

Protons (55) The miracle of fish oil (2)

I have a feed to my email account which has worked out that I am interested in longevity studies and particularly the role of PUFA in the inner mitochondrial membrane. This paper popped out today:

Dietary fatty acids and oxidative stress in the heart mitochondria

The diets were roughly 16% of calories as coconut oil, olive oil or fish oil. Fed to rats for 16 weeks, which is a fair length of time in the life of a rat. They were interested in the effect of unsaturation on the measurable oxidative damage done to mitochondrial proteins and the peroxidation of inner mitochondrial membrane lipids.

TLDR for the paper itself: If I was taking fish oil I would stop.

But of course I'm more interested in the body weights.

Here is the summary of the lipids in the diets, butchered out of Table 1:

And here, in its entirety, is Table 2 giving the weights at the end of 16 weeks. We can ignore the fish oil plus probucol group, except to note that they were even fatter than the fish oil group, don't you love those good old antioxidants:

Obviously none of the weights are significantly different from each other and 10-20 grams on a 500g rat is not a huge difference. Except food supply was limited to a fixed, slowly increasing amount as the rats grew. There is no mention of uneaten food so I think it is reasonable to assume all rats ate all of the food offered. So on a rigidly fixed calorie intake the fish oil fed rats were heaviest. I won't mention Arnie rats or C57Schwartz6 mice after my embarrassment in the comments to the last post.

Very roughly the modest excess weight goes up with the double bond index of the diet. On a rigidly fixed, mildly hypocaloric diet, even if p stays stubbornly above 0.05.

Also distinctly non-significant but appropriately trending are the fasting glucose readings. Those are in Table 3:

Highest in the Coconut oil group, trending down to lowest in the Fish oil group. Fish oil leaves you insulin sensitive.

Insulin signalling in adipocytes makes you fat.

If the rats were allowed to eat ad lib then the calories lost in to adipocytes would be replaced by eating more food. Eating the food would get the blame for the adipocytes being bigger than they ought to be.




Eric said...

Hmm, I don't quite see that the olive oil mice put on more weight per se, but their BMI increased more, though probably not significantly.

A question I have asked before in the context of the croissant diet (TCD): is it good to be insulin resistant and have all that glucose floating about?

Reasons why it might not be good:
- Glycation
- Pancreas cranking up insulin production even higher?

So what should we be eating? Fully hydrogenated coconut oil (we can buy it in Germany under the brand name Palmin) and meat? Just tastewise, I prefer virgin olive oil over Palmin or virgin coconut oil any day...

Eric said...

Oh, forgot, has anyone found a source of food grade stearic acid this side of the pond? Palmin is 100% saturated but not as long chain.

Stefan said...

Not stearic, but I often use Ossewit, or blanc de boeuf. That's for sale in the deep frying department in the Netherlands. It's rendered beef fat.

Eric said...

Oh good, I need to look the next time I'm there.

Eric said...
This comment has been removed by the author.
Peter said...

Eric, the DBI goes from 23 to 94 to 374. I'd guess that 23 was close enough to 94 to have its effect lost in the noise...


Passthecream said...

Probucol " may also cause QT interval prolongation."

Hey, maybe it too will be useful against CV19??? /sarc

Unknown said...

What does Probucol mean?

ctviggen said...

Let's assume that your theory about O3s is correct. Could there be any other benefit to taking fish oil or eating fish (well, I stopped taking fish oil a while ago, and just eat fish...sometimes)? I mean, there are very few people (maybe Cate Shanahan?) who are against O3s.

Dr. Shanahan was my primary care doctor for a short while:


Are all of the so-called good studies of fish/fish oil and in particular heart disease just group-think?

Eric, it's a difficult question. I tried a very high saturated fat diet after seeing Brad's croissant diet. I also added in a "targeted keto diet', where I ate higher carb after my workouts (only). I used mashed potato or sweet potato and Brad's high stearic acid ghee, even some pasta with super high butter fat/cream sauce. I do think the high saturated fat part of this did blunt my hunger, often significantly.

However, I find I can't handle sweet potatoes, as they cause issues (eg, heartburn, which I NEVER get if I eat just meat). I also some issues from potatoes and pasta. I also did not have a CGM (continuous glucose monitor), so I could not get a good gauge of what happened to blood sugar. A lot of plants and I don't get along, and wheat in particular is not great for me.

And I wonder about how much PUFA is too much? Last night, my wife made "pigs in a blanket" which are hot dogs normally covered in a flour/bread mixture. She made them from a Diet Doctor recipe, so I'm sure it used almond meal/flour. Is that enough PUFA to cause me to be hungrier? Did I eat more last night or will I eat more today because of that?

It's super difficult to tell.

Over time, I have lessened the amount of olive oil I use and try to use more butter and beef tallow (rendered beef fat, in the US). I still eat olive oil, though, and sometimes a good olive oil is quite tasty.

Anyway, for real people eating low carb/keto (not you carnivores), there are a lot of issues about PUFAs, and I haven't come up with good solutions -- other than try to eat more (evil!) red meat from ruminants and reduce fatty chicken and pork, and anything with nuts. Can't get away from chicken, pork, and nuts, though, at least in my house.

cavenewt said...

It's still puzzles me why so many people trying the croissant diet think they must eat carbs with it. In his initial post describing the diet, Brad said he included the croissants to prove that it wasn't ketosis causing his weight loss. Also it made the diet more palatable. You can soak up an awful lot of butter with pancakes. Also he used to be a chef and just wanted to make croissants.

I've embraced stearic acid, cooking with a lot of butter, beef tallow, cocoa butter, and some coconut oil, but have largely cut out olive oil, chicken, and pork. Not for weight loss really, but I had a cancer blip over the winter and it appears that monounsaturated and polyunsaturated fats are tumorigenic. Not to mention polyunsaturated fats are looking more evil all the time.

Tangentially related question: I've got a 2-pound chunk of lamb suet in my freezer. Do I need to render that into tallow in order to use it for cooking? Or does that just make it keep longer? Or what?

cavenewt said...

Peter, you included several tables from the rat diet study. But you left out the one where they interviewed the rats about how hungry they felt during the course of the experiment. That would be some useful information.

Puddleg said...

Peter, if I was taking fish oil at 16% of calories I would stop, even before reading the paper. I'm taking 3g a day, which I make out as around 1/10th as much, and around what I'd get if I liked fish at around an average amount.
I don't remember to take it every day, but it makes me less grumpy and my joints don't ache.
Both these things were hugely reduced from baseline by LCHF of course but it's good to have backup.
In the dietary fat and alcoholic liver disease literature, 35% of calories as fish oil is even worse than corn oil, but much smaller intakes seem to have benefits. Basically, you want the stuff to be either fed into PPAR-alpha or esterified etc, you don't want to eat so much of it that there's a lot washing around as TGs and FFAs.

Puddleg said...

P.S. I love figure 1 in this experiment, for what it says about appetite, dietary fat, and the set point. It couldn't be clearer what "normal" appetite looks like.
The results are interesting too - an anti-inflammatory effect of palmitic acid at sn-2, where you will find it in animal triglycerides vs plant.

"The increase of PEA in adipose tissue was associated to a reduction of plasma TNF alpha after treatment with LPS, while other pro-inflammatory cytokines such as IL-1 and IL-6 did not change significantly. Whether the effect of Hsn-2 PA is limited to TNF alpha, as it has been shown for PEA [13], or the anti-inflammatory activity is not sufficiently strong to also influence IL-1 and IL-6 levels, remains to be elucidated."


Passthecream said...

Cavenewt, to make pure fat you can render your suet by chopping it up and putting it with some water and heating the lot gently in a large pan. Simmer for some time and keep an eye on the temperature. At the point when most of the water has boiled off the temperature will suddenly rise and that's when you stop. The crispy bits of cellular debris left in the pan are very tasty snacks.

I like to add maybe 20% suet, as-is, to other meat that I'm mincing as a fat boost. This makes great sausage filling or patties/rissoles. The minced suet makes the minced meat light and juicy.

My spell check wants suet to be diet. That's not right!

Jason Cuadra said...

Anyone know if the fish oil in this experiment was fresh and high quality, and not rancid?

Peter said...

Jason, “The t-butylhydrozyquinone is used as an antioxidant in all the diets to preserve the fatty acid free of oxidation”. Stored frozen.

Unknown, “A fourth diet was composed of fish oil but with
the addition of 1.0% probucol as a liposoluble antioxidant.”

Haha cave!

ctviggen, interestingly Tucker has a paper which shows nasties like 4HNE are disposed of by beta oxidation and, obviously, this is up-regulated in lc/keto eating. On a level above insulin signalling, that’s another plus for keto…

George, I can see that. My overall feeling is that as a signalling molecule DHA has a lot going for it. If it is beta oxidised in peroxisomes it will be mitochondrialy benign. The obesogenic effect should be demonstrable when omega 3s are being metabolised as bulk energy sources in mitochondria and especially when insulin signalling is high, so there is lots of scope for tweaking their effects by adjust the dose and the local metabolic environment.

The paper is fascinating. Overall I would consider increased food conversion ratio to be advantageous, evolutionarily. Ditto slower weight loss under fasting, provided that, under both circumstances, the organism is fully functional. I also wonder whether the position in the triglyceride determines oxidation vs storage on presentation to a cell, especially an adipocyte…


Puddleg said...

Looking at the sn-2 summary of Bill Barendse, who ought to know if anyone does, it seems that 99% of fat is absorbed wherever it is on the TG, except for stearic acid.

So it is not that significant amounts of non-sn-2 FAs are being lost (although FAs will be consumed by gram +ve microbes, and do affect stool consistency).
It's possible that sn-2 C12-C16 SFAs are more likely to be oxidised intrahepatically (to energy and ketones), lowering TGs, and that this spares longer-chain FAs for storage, longer chain FAs containing more energy.
Something like that seems to be happening here

JT said...

Abstract on PUFA.. Haven't read the full article to judge it's merits, though I tend to the pessimistic side of 90% of papers :)


Peter said...

JT, interesting but mealy-mouthed: "increased" and oxidised FFAs act on Hsp70.1. But only if they are PUFA. What they are not saying in the abstract is that the lipid hypothesis causes Alzheimers. I guess they daren't say that chips cooked in beef fat don't increased lipid peroxides, however high FFAs are raised. Perhaps they are dreaming of an Hsp70.1 blocking drug? We'll have to wait for SciHub to release the paper to the world before we can tell.

George, I guess you could simply say that increasing saturated fat by just 4% of calories in the diet (of which 1.2% happened to be myristic acid) improved dyslipidaemia. I wonder if they could have done exactly the same with just palmitic acid but realised that that would be academic suicide?


Anand Srivastava said...

I had read long back from Stephan Guynet, that anything more than 4% PUFA in diet is harmful. And it should ideally be between 1:2 and 2:1 of Omega6:Omega3. Obviously 16% of Fish oil will be very harmful. It would mean at least 5% of Omega3. I would expect that there was some more PUFA in there. For people who are consuming meats and vegetables, we are bound to get much more omega6 than omega3. A spoon (5ml)of Fish oil, ie around 1.5gm of Omega3s, 12kcal or .5% of 2500kcal diet would help, and cause no problems. I do try to get a spoon of Cod Liver oil a day, and it has helped the immunity of my family. We have experienced the loss in immunity when we don't consume it over several months.

Passthecream said...

Peter: "What they are not saying in the abstract is that the lipid hypothesis causes Alzheimers. I guess they daren't say that chips cooked in beef fat don't increased lipid peroxides, however high FFAs are raised."

I long for the good old days when all the deep frying around here was done in tasty beef dripping. There is a catch in heating any triglycerides to the smoke point however whether they are saturated or not and that is the production of acrolein which is nasty stuff. It is generated by the breakdown of glycerine. The sensible thing is just to not get any fat too hot. Cook fatty foods gently or just boil them.

I have been experimenting with other methods of turning triglycerides into FFA. Simmering together with acidified water ( citric acid or vinegar) has been interesting --- TAG can be broken down by acid hydrolysis almost as easily as by basic saponification. Even hot salty water can slowly hydrolyse TAG to FFA plus glycerine. One industrial method uses plain water plus heat plus pressure which is a bit risky for a domestic kitchen unless maybe the KFC people would lend one of their lipid based pressure cookers? No matter which method, the liberated glycerine stays mostly in the aqueous phase so can be drained and washed away.

Separating the pufa from the sfa afterwards is a bit more tricky. Probably need a kitchen centrifuge for that.


Btw after separating the resultant gylceroyl citrate from a reacted solution of citric acid and TAG, this can heated to make a type of polyester.

Tomasz said...

what's your take on this article by Matsuo et al:

Body fat accumulation is greater in rats fed a beef tallow diet than in rats fed a safflower or soybean oil diet".

There are other articles from the team referenced below.

Btw, I'm currently limiting my fat sources to beef tallow and I'm not overly concerned ;-)

I tried this stearic acid,

and this beef tallow.

Frunobulax said...

@Tomasz I always wonder if there is something that we don't control for. I posted as a comment that I suspect that we'll eat more if we're deficient in micronutrients. Maybe this could be a confounder?

I'm not a big fan of rat experiments in general. From what I gather, rats get fat if you feed them ad lib HFLC diets. Humans don't. Until we understand why, we should be careful to use rat experiments as benchmark.

Obviously it's easier to use rats than humans for such experiments. But we should still be aware of the limitations.

Peter said...

Anand, as to George, I would accept that small amounts of omega 3s as signalling molecules might be useful but I would be cautious about hard numbers.


Peter said...

Repost to correct a typo.

Hi Tomasz,

It’s excellent. I think I have it lying around on my hard drive somewhere but never got round to blogging about it. Look in to Table 3 and ask which rats where lightest? That’s after five weeks at peak growth rate, when calories will go to growth rather than adipocytes. What would happen over the next 50 weeks as middle aged spread sets in because lean mass growth is then much slower? The initial changes are already there but can be ignored because p is greater than 0.05, for groups of five rats over 5 weeks. It’s also nice to see the (ns) weight gain was lowest on beef fat, intermediate on pure omega 6 and highest with some added omega 3.

Because experiment 1 didn’t produce the dirt on saturated fats they did experiment 2. This time they semi starved the rats. If anything is going to make PUFA look good it is caloric restriction. But even so there is no difference in visceral fat. All we can say is that under hypo caloric conditions saturated fat preserves body energy stores better than PUFA do. None of this happens voluntatirly in the real world outside of WeightWatchers because we all know how much fun semi starvation is…

Pass, a kitchen centrifuge??????? What sot of kitchen do you have?

Fruno, I love rats. They have better mitochondria than Bl/6 mice. The devil is in the methods most of the time.


Puddleg said...


depending on feeding conditions there is usually more C16 than C14 at sn-2 in dairy - however, C14 seems to have the strongest effect on TGs and VLDL in hepatocytes, even noted at 10 mmol/l glucose (removing TGs from VLDL and producing native IDL and LDL, hence the cholesterol-raising effect).
"The combination of elongation and beta-oxidation results in the rapid disappearance of C14:0 in hepatocytes whereas C16:0 is esterified to form glycerolipids. This study provides evidence that myristic acid is more rapidly metabolized in cultured hepatocytes than is palmitic acid."

PUFAs undergo 2-carbon cycling and are substrates for cholesterol and SFAs, which could explain reduced weight gain, as it would involve a loss of energy. Carbon from palmitate isn't recycled in the same way. (Cunnane's research)
But once endocannabinoids are skewed, omega-6 produces weight gain by increasing appetite and adipocyte growth.

Passthecream said...

Peter: "? What sort of kitchen do you have?"

Whatever it takes!


I do have a couple of centrifuges in my workshop! One bench one, one floor model. Last time I got my phone wet it was very handy.

Peter said...

Thanks George, that Norwegan paper is nicely confirming much of the impression I have gained from looking at omega3 research. Context is everything. And mice and humans are remarkably similar.

Haha Pass!