Saturday, March 04, 2017

Trans fats vs linoleic acid

TLDR: Trans fats may not be as bad as they are made out to be.

This paper is comparing a high linoleate diet (using lard) to a soya oil derived (Primex) diet where much of the linoleate has been industrially hydrogenated in to trans fats and fully saturated fats.

A comparison of effects of lard and hydrogenated vegetable shortening on the development of high-fat diet-induced obesity in rats

The thing I like best about it is that, wait for it, they measured the fatty acid composition for their diets! HPLC and all that. Then, they put the results in the paper! On the down side their concepts about energy balance are pure CICO and rats have lingual receptors for fat which link to "hedonistic" centres in the brain. So they understand nothing, but we can forgive them for that.

Here are their results:

I think this might suggest that linoleic acid is obesogenic. Not that I've ever mentioned that before. Now obesity is obesity. What about health? Here are the numbers which matter, obviously insulin is the one we want to look at:

So, clearly, eating 11% of your calories as linoleic acid makes you fat and ill. The fat in the HVF (and the NF control) diet is made of:

"Diet compositions are presented in Supplementary Table 1. Primex pure vegetable shortening, a mixture of partially hydrogenated soybean and palm oil, was used by Dyets Inc. (Bethlehem, PA, USA) to formulate NF and HVF experimental diets".

Compositions panned out as :

Now these are measured, nothing is accepted from USDA food tables etc. So..... If we look at health outcomes (Table 3) in conjunction with diet composition (Table 1) it become pretty evident that, in these rats, trans fatty acid (TFA) feeding at 15% of total calories is positively health generating compared to linoleic acid feeding at 11% of calories. I did not expect this.

Aside: If any diet trial does not have its fatty acid composition measured you have no idea how much linoleic acid it contains. Usually more than you think. Also, joke of the century so far:

Question: If you are in a position of power over innocent folks who are trying to eat healthy food, which fat would you ban?

Answer: The wrong one!

End aside.

I was on a PubMed search looking for trans fat toxicity. In this current study trans fats not only fail to cause insulin resistance, they render insulin-glucose parameters identical to the NF fed rats, despite the TFA fed rats carrying an extra 100g of adipose tissue.

That is very interesting. You can't answer the whys and wherefores from this paper. The missing piece of information is probably FFAs.

Trans fats are very odd. On acute exposure to TFAs isolated adipocytes release stored FFAs. This is what Cromer et al have to say in their study:

Replacing Cis Octadecenoic [Oleic] Acid with Trans Isomers in Media Containing Rat Adipocytes Stimulates Lipolysis and Inhibits Glucose utilization

"Overall, results of this study clearly show that conversion of octadecenoic acid from the cis isomer [oleic acid] to the trans isomer [elaidic acid] in adipocyte media will substantially increase lipolysis and inhibit glucose oxidation and conversion to cell lipid".

Just to clarify: Acute exposure of freshly isolated adipocytes to trans-oleic acid causes fat release, decreased glucose uptake and decreased glucose incorporation in to lipids. Sounds like a weight loss drug to me.

I was expecting this lipolysis to show up as elevated fasting FFAs and elevated fasting insulin. But there isn't any insulin resistance under fasting conditions visible in Table 3 so I think it is reasonable to assume there is no elevation of FFAs at this time either......

The only explanation I can come up with is that the adipocytes of the TFA fed rats are not as "full" as they should be, due to trans fatty acid induced lipolysis. Giving some "space" within an adipocyte allows the insulin sensitising effect of PUFA oxidation to show, certainly while fasting and lipolysis is the correct state to be in. Hence the fasting insulin levels are normal despite the 100g of extra bodyweight.

Or you could theorise that excess lipolysis from the trans fats is being almost exactly matched by decreased lipolysis from the insulin sensitising effects of linoleic acid. The combination just happens to pan out close to normal, provided the rats carry 100g of excess adipose tissue.

Some degree of post prandial hyperinsulinaemia/insulin signalling seems essential just to maintain those extra 100g of accumulated fat, but it's clearly not visible in the post absorptive phase.

If anyone can come up with a better explanation, I'm all ears.

BTW, this obviously relates to Axen and Axen's work. Their hyper-obesogenic diet was based on something Crisco-ish from the 1990's:

"The hydrogenated vegetable fat contained ∼25% long-chain saturated, ∼44% monounsaturated and ∼28% PUFA, with ∼17% of total fat as trans fatty acids (manufacturer’s communication)".

With fat making up 60% of the calories in the diet, and that fat being 28% PUFA, this is somewhere around 17% of total calories as linoleic acid. That is a LOT of linoleic acid. At the time I thought that the trans fatty acids would be to blame. Nowadays I'm not so sure.

How much of the bad rap that trans fats have received is from the PUFA which travelled with them at the time? Without mentioning the amount of fructose in the biscuits. Modern Primex is much more hydrogenated, so lower in PUFA, than whatever Axen and Axen used. It's far less obesogenic too.


Other odd final thought: People who are obese and insulin sensitive: Are they the folks who eat most trans fats along with their hearthealthypolyunsaturated linoleic acid???????


Tucker Goodrich said...

Saw this study a little while ago. Had the same reaction to it as you.

Most of the research I've seen against TF is epidemiological, with a few that indicate metabolic dysfunction.

I wonder how much of the case against TF is just association w/ n-6?

The relation to insulin is particularly interesting, as I think LA is likely causal in IR, but haven't found a clear mechanism yet.

Peter said...

Yes, epidemiology confounds trans with w-6 and fructose I suspect. And of course trans fats are a whole ragbag of chemicals, some seem worse than others. I don't have any great love for them but I doubt they are as bad as linoleic acid!


Unknown said...

CLA has a lot of positive results when looked into. The assumption has always been the artificial nature of partial hydrogenation making the TFA so bad, but it seems more likely that it is the nasty chemicals that come along with them, plus the fact it is always LA heavy oils that are the base.

Jason Jodway said...
This comment has been removed by the author.
Chris Adams said...

Pretty sure Nina Teicholz concluded in her book that the evidence around trans fats being bad wasn't that convincing.

Samuel William said...

hmm I eat a LOT of Tofu (hermaphroditic times..), no trans fat but high linoleic, maybe I should find a trans fat pufa source and let you know the phenomenology, more experiments what fun! Hyperlipid gags. Shocking how each fuel has such different effects, relate to environment differently. The most locally overwhelming for me is alpha linolenic... but fun! As if it all just happens in the alchemist stomach...

DLS said...

it seems this LA is the plutonium! doesn't apply to peanuts ... im 175cm @ 63 kg still eating 100+ peanuts + a few almonds a day + industrial eggs n meat for extra (ω-6)... (3 years now) btw i know is not a "health metric" but... how much do YOU weight peter? 64? and do you have ANY abs (i do...) or are you rocking... healthy REGULAR flab? i trust.. .results. ... btw LAST Q, is LA -from peanuts- worse than 40gr sucrose a day? i don't think SO...

Robert Andrew Brown said...

Do you live mainly on heavily processed refined nutrient depleted/damaged including regular fried food, fried chicken plus skin etc. and rarely exercise.

I suspect not?

As previously responded to your post the issue is oxidized LA not LA per se.

See material on Kung San! and maybe read ch27-32 where I grind on about such issues.

The LA metabolic pathways are complex and lipid metabolism differs significantly between those that exercise, and or inter-meal fast, and those that do not.

Robert Andrew Brown said...

^ Please see fig 5 here for thought provoking data on obesity and chicken, shortening and soyoil intake intake.

What we are seeing is about the effects of these foods in the context of nutrient depleted overly processed food often produced to considerations of cost rather than primarily nutrient value to the best value that can reasonably be achieved.

IF food does not nourish humans if fails to meet its purpose . . .

DLS said...

thanks for your reply Robert! I missed your previous post (traveling + shitty cellphone) i tend to agree, but Peter seems to suggest that ANY kind of LA -refined or not - eventually leads to obesity n Disease. Im sceptic but also respect his input, tend to feel better "following" most of his cryptic advice. so much that i removed sucrose and milk from my diet since 2014 (insulin ) this with a 2% body fat and pure muscle... i was tired after eating and now im not.

Peter said...

DSL, absolutely not. F3666 ketogenic diet does not cause obesity or shorten lifespan, despite being very high in linoleic acid causing and low grade hepatic inflammation (w-6PUFA = hepatopathy). But if insulin is low as per ketosis, who cares about differential insulin sensitivity? You have your own ideas, they are working for you. They may or may not work for the next 20 years. That's fine. Record what you think and why. I just record what influences me. I object to being asked to perform like some gym rat with disturbing* abs.


*"Disturbing" was the descriptor applied by a lady in response to a selfie posted on Dr Bernstein's forum by an antipodean fitness guru, many years ago. Pretty good descriptor. Not one I'll forget.

Tucker Goodrich said...

@Unknown "CLA has a lot of positive results when looked into."

Yes, because it appears to block uptake of LA, this has been seen in breast cancer experiments.

Interestingly, the SkQ1 synthetic antibody blocks most bad effects of LA on mitochondria (by protecting it from oxidation) but most of the mice put on that regimen died of breast cancer. Diet not disclosed, but as this was done in Russia, one imagines they were getting a fair bit of LA.

Tucker Goodrich said...

@Robert Andrew Brown:

"As previously responded to your post the issue is oxidized LA not LA per se."

I don't think this is correct. LA accumulates in cardiolipin in mitochondria, where it is subject to oxidation from ROS generated from complexes I-III, but it also auto-oxidizes with iron-containing cytochrome-c. This generates ROS independently of respiration, along with toxic oxidized LA metabolites, inside the cell, leading to many bad outcomes.

High intake of high-LA-containing grains like wheat or corn leads to epidemic esophageal cancer. This is seen in China and Africa.

Tucker Goodrich said...

@Robert Andrew Brown:

So while I agree that it's oxidized LA, not LA, that causes the problems; since LA (and AA, apparently) can auto-oxidize where it can do the most damage, it's a moot distinction.

Tucker Goodrich said...

@Robert Andrew Brown:

As for the !Kung, dietary LA rapidly alters cardiolipin composition of LA. They eat their high-LA diet seasonally, not all the time, and this may be an important reason they don't seem to see ill effects.

They have a resting period to undo damage. That, combined with a constant exercise routine, is likely the difference.

DLS said...

so where back to LA is shit - at cellar level- mmm. btw peter im mot trolling u. like i said is not a heath metric. "gym rat" etc > common misconception @ bias against exercise? you don't even need to touch a gym to have great -or disturbing- ABS- they are made in the kitchen- so if you dont have them... maybe something is not so great... btw > "my gym" : push up stand, 2 big dumbells, chin up bar, the floor. and just one move per day, 5 to 6 sets. nothing at all!

Robert Andrew Brown said...

^ Hi Tucker

Thanks for the comments. We are not so far apart. Yes cardiolipin takes up LA, and oxidised LA in cardiolipin can lead to mitochondrial loss of efficiency and ultimately mitochondrial 'death'. There are multiple mechanisms for oxidation of LA in cardiolipin, including the LOX enzyme pathways.

Lipid membrane antioxidant options and pathways are somewhat limited in number.

Arguably during intermittent fasting or exercise related energy deficit the peroxisomes and related PPAR alpha pathways play a much greater metabolic role than realised creating short fats and ACoA that can be fed to the mitochondria.

At the same time PPAR alpha related antioxidant pathways are uprated, and it is possible oxygen consumption is reduced.

So I hold to the position that excess oxidised LA in the context of a antioxidant capacity depleted westernised highly process diet is the key issue. So as most people are on a westernised nutrient depleted diets, to varying extents for most people, excess LA is a health issue, but for those with better lifestyles and diets such as the Swedish in times gone past effects of excess LA will manifest less clearly, which explains some of the study outcomes.

San LA intake was seasonal - agreed - and agreed they were also lean, worked for their food ate nose to tail etc, so they had antioxidant rich diets, did not carry much LA in fat tissue etc.

So excess LA in the context of a western diet is an issue, as it will as you point out, get oxidized both auto and enzymaticaly . . .

The AA and LA oxidised products product pathways have commonalities but are also in many respects very distinct. The amount of oxidised LA products in plasma far exceeds AA products. . .

Robert Andrew Brown said...

HI Peter

Many thanks for an excellent post; some thought provoking material; I will come back on the insulin issues and give the trans fat observations on adipose metabolism some thought; they will take a bit of brain space and I have a 'paper' to finish.

Again many thanks


Tucker Goodrich said...

@Robert Andrew Brown:

Thanks for your thoughtful response! I agree, we're on the same page, as there aren't enough studies in this area, there's still plenty of room for quibbling over details!

"Lipid membrane antioxidant options and pathways are somewhat limited in number."

They're apparently non-existent, as this demonstrates:

"Impact of Antioxidants on Cardiolipin Oxidation in Liposomes: Why Mitochondrial Cardiolipin Serves as an Apoptotic Signal?"


"So I hold to the position that excess oxidised LA in the context of a antioxidant capacity depleted westernised highly process diet is the key issue."

Well, eating oxidized LA is clearly a major problem, as the toxic products can clearly enter the circulation. I don't argue with that. What I don't think that explains is the widespread existence of mitochondrial dysfunction through oxidized cardiolipin in the metabolic syndrome and related diseases, such as cancer:

"Cardiolipin and electron transport chain abnormalities in mouse brain tumor mitochondria: lipidomic evidence supporting the Warburg theory of cancer"

I don't think ingesting OxLA explains why it would appear to target inner mitochondrial membranes specifically, as toxic products such as 4-HNE appear to be short lived and diffuse randomly. What I'd like to draw your attention to is internal generation of these same products

Additionally, endogenous antioxidants such as glutathione appear to be specifically depleted in tissues experiencing oxidative stress. As glutathione is the detox pathway for 4-HNE, it appears that mitochondrially-oxidized LA can affect specific tissues, depleting antioxidants from the inside out, as it were. This would explain the repeated failure of exogenous antioxidants to be beneficial.

"Degree of Glutathione Deficiency and Redox Imbalance Depend on Subtype of Mitochondrial Disease and Clinical Status"

"The AA and LA oxidised products product pathways have commonalities but are also in many respects very distinct. The amount of oxidised LA products in plasma far exceeds AA products. . ."

AA is more susceptible than LA to degradation to 4-HNE, from what I've read, but as AA's rarer, it's less of an issue in plasma, as you note. This appears to be tissue-specific, however, as in prostate cancer, where depleted AA states such as diabetes reduce AA output and prostate cancer incidence.

AA status in plasma also appears to be fairly stable, but it accumulates in tissues differentially.

AA -> 4-HNE seems to be a key path for Alzheimer's, as 4-HNE is always present, and AA is depleted in affected tissues, indicating that a AA->4-HNE self-sustaining reaction is occurring.

"Altered neuroinflammatory, arachidonic acid cascade and synaptic markers in postmortem Alzheimer's disease brain"

"Four-Hydroxynonenal, a Product of Lipid Peroxidation, is Increased in the Brain in Alzheimer’s Disease"

"Oxidative stress and lipid peroxidation are upstream of amyloid pathology"

The brain-blood barrier in humans appears to block much LA (shown by the reduced incidence of LA in human brain cardiolipin), but AA does cross the BBB.

The later incidence of Alzheimer's supports your observation that AA -> 4-HNE is a lesser problem, unless it's occurring in tissues that are particularly susceptible to that pathway.

Thanks for your response again!

JohnN said...

IMO, it doesn't have to be trans fat vs linoleic acid. This issue began with the avoidance of saturated fat several decades ago from faulty research.

The (cheap) old Crisco formula fully hydrogenates linoleic acid (in cotton seed oil) into stearic acid (beef tallow).
If one has to consume synthetic fatty acid (when the natural version becomes unaffordable, in the near future) it's better to choose one that duplicates nature. Fatty acids are part of lipid membrane and also signaling molecules. The impact of trans may be far more reaching than insulin sensitivity.

Passthecream said...

Peter, what do you make of the significantly higher Homa-ir insulin resistance vs the highest LA intake?

Peter said...

Pass, these are adipocytes which have accepted most lipid under the influence of linoleate. They are distended so trying to push more fat in forces distension induced FFA release, hence increase IR. The TFA adipocytes have more room due to some degree of lipolysis, especially under peak insulin, so less fasting FFA spillage, less IR.

The only issue I have personally is that the TFA rats gained as much weight as they did, the linoleic acid in the diet was nowhere near as high as those fed lard but they made +100g cf the +200g of the lard fed.....


George Henderson said...

I have a possible explanation for the TFA lipolysis riddle;
Niacin doesn't inhibit lipolysis from TGs neatly - glycerol is released from adipocytes but the FFAs are retained.
"Although adipose tissue lipolysis, measured as glycerol release, increased, the lipolyzed fatty acids were retained in adipose tissue, suggesting an enhanced synthesis of glycerides both from exogenous and endogenous sources. The increase in fatty acid incorporation into adipose tissue indicates that the decrease in serum triglyceride levels produced by nicotinic acid treatment may partly be due to the fact that this drug promotes incorporation of fatty acids, derived from lipoprotein-carried triglycerides in the blood, into adipose tissue glycerides."

TFAs are not at all associated with type 2 diabetes in epidemiology, but they are strongly associated with CHD.
The mechanism proposed by Kummerow and Enig is that elaidic acid is structurally similar enough to block elongation of ALA to EPA and DHA.
A lot of the epi studies on TFA come from the US.
An American joke - "In some foreign countries they eat animals that Americans keep as pets, like fish".
Also, even ALA intakes were very low in the US during the earlier part of the trans fat era.
So, if trans fats are associated with a diet high in LA, and low in ALA, and hopeless for EPA and DHA, there's a context for harm, but it's not metabolic harm so much as dodgy blood clotting and regulation of inflammation that will kill you, so you won't see it in a short-term study.

Robert Andrew Brown said...

@ Tucker Goodrich

From a quick skim we are very much on the same page.

Thank you for the links which I will look at as soon as I have a moment.

I too think excess LA / related oxidative stress etc has a role in Alzheimers; LA does cross the BBB maybe in greater amounts than realized and likely metabolized by the peroxisomes . . .

Again see the link to the Springer Chapters, where I look at the same issues you are obviously thinking about.

Do you have a web site with an email contact?

DLS said...

evil peanut LA "bs" DEMOLISHED... ? Meet China’s ‘supercentenarians

Li Aizhu, though, has a simpler explanation. “We asked her once,” said the 113-year-old's great-granddaughter, Yi Mei. “She said its because she eats a lot of... peanut oil, that's her secret.” -

PEANUT ... OIL. 33% LA = 113 years. ok

Peter said...


Excellent, enjoy your peanuts!


DLS said...

D L S.. not only peanuts, almonds and 70% chocolate !.. vit E to "saturate" pufas (lol)

Peter said...

Oops, sorry!


DLS said...

nothing to apologize peter! LSD also works! btw i don't want to hijack the discussion but i have to admit... -once again- you were spot on regarding LA... im seeing some pretty spectacular changes just by reducing my "pig feed" to half. doh. im back to my original self! at least cosmetically... ( not going to post pics of disgusting ripped abs dont worry!) i will try 2 enjoy my peanuts, in moderation. (and maybe ditch them if i manage to find decent chocolate) THANKS!

karl said...

I have wondered if the data on trans-fats was confounded with LA. If you bubble hydrogen through PUFAs to make fake lard - some of the LA becomes trans-fats but not all.

There is another narrative that I've become quite doubtful of - that the O-6 (LA etc) pathway to AA and on to inflammatory prostiglandins is blocked by the consumption of the long chain O-3's in fish oil. I don't think this narrative is at all well grounded.

This matters, as there are a large body of 'diet experts' that claim(parrot and preach) that balancing O-6 with O-3 is some magic workaround. I don't think so - I think it might be a much better idea to just limit LA consumption to something like what was around before 1960.

It also appears that LA may do harm in several ways - HNE, protons and more - and these pathways have nothing to do with O-3s.

In 1960 when they started pushing concentrated veg oils - very few people were overweight. I think there are other possible changes - the young 20 year-olds today don't look healthy to me (total subjective speculation here out). Their skin looks older than their years - the muscles look flaccid (even the athletically active). I've been calling them generation F for Flaccid.

I remember spending some time in the Philippines back in the early 90's. They sold buckets of lard for cooking at the market - no one cooked in veg-oil. Few appeared overweight. Returning to the USA everyone seemed so fat.

Today they are selling cheap oil and obesity and T2D is booming -


If the new narrative for CAD is correct - that the initial damage is due to elevated insulin - then eating LA has yet another health effect...

Tucker Goodrich said...

@Robert Andrew Brown

"Do you have a web site with an email contact?"

I have a website:

I'm at tuckerg gmail if you want to email me.

I'm also on twitter @TuckerGoodrich

George Henderson said...

This post inspired me to leave a PubPeer comment on another paper

Peter said...

Well said George. Some of the "high fat diets" are utterly opaque. No one can replicate such studies, ever. It's bad enough coming out of the far east but Dom D'Adgostino is a co-offender. You can tell most scrutes are a) unpaid and b) don't much bother about the devil in the details. And boy, there are some devils in those details!


Tucker Goodrich said...

@Peter & George:

"Some of the "high fat diets" are utterly opaque. No one can replicate such studies, ever.... And boy, there are some devils in those details!"

It's even worse than that! The researchers buy standard diets, and the companies selling those diets don't even know what's in them!

"...I got an email today from Dr. Matthew Ricci, the Vice-President and Research Director of Research Diets, the company that produces the infamous 60% fat, lard-based rodent diet D12492. I've written about this diet before. The company had previously been using the USDA database to determine the diet's fatty acid profile, but recently had it directly analyzed, knowing that the fatty acid profile of lard can vary according to what the pigs are fed.

"It turns out that the diet obtains 32% of its fat from PUFA instead of the previously reported 17%. The ratio of omega-6 linoleic acid to omega-3 linolenic acid had been previously reported as 7.8 but is actually 14...."

karl said...

"Some of the "high fat diets" are utterly opaque. No one can replicate such studies, ever.... And boy, there are some devils in those details!"

I think calling them opaque is being overly kind - at the very least they are muddying the water rather than advancing science - at the worst they are being purposely dishonest.