Monday, February 05, 2018

Follow on to Tucker's post on PUFA in rats

Tucker posted an excellent discussion of this paper on his blog. Go read it:

Fat Quality Influences the Obesogenic Effect of High Fat Diets

The basic conclusion is that feeding rats a high fat diet makes them fat. If it is PUFA based, including a generous amount of omega 3 alpha linolenic acid, it will cook their liver (figuratively speaking... in actuallity it converts their liver to being full of peroxidised PUFA, en-route to cirrhosis). I have an anecdote-type post on the problems of being married to a cardiologist if you happen to be alcohol addicted somewhere. I really ought to dig it out and hit post.

So. The problems with the paper:

The rats on the PUFA diet, with the gross fatty livers, were less obese than the lard fed rats, had better lean body mass percentage and much better brown adipose tissue hypertrophy and fat oxidation.

The bottom line: If you want look slim and well muscled in your coffin then a safflower oil diet with a heavy dash of varnish might be a good choice...

How come?

The paper was not looking at insulin levels or insulin signalling so it doesn't provide the data we need to come to any conclusions but it has resonances to the comment Zoran made on the previous post.

The Protons Credo (believe if you so wish!) for the situation:

PUFA, of a carbon chain length which targets them for mitochondrial oxidation, input less FADH2 at mitochondrial electron transporting flavoprotein dehydrogenase (mtETFdh) than do saturated fats or MUFA. This lack of FADH2 input limits the ability to reduce the CoQ couple and facilitates electron flow down the electron transport chain (ETC) and so limits the generation of reverse electron transport through complex I. This damped RET limits the ROS generation (superoxide and H2O2) necessary to initiate insulin signalling under fasting and to limit excessive insulin signalling in the fed state.

So on a whole body basis PUFA maintain insulin sensitivity. Insulin acts, rather well, under PUFA compared to under saturated fat, in the fed state. It works less well in the fasted state.

A fed, insulin sensitive animal will do two things of interest on a medium carbohydrate, generous fat diet. It will utilise glucose easily in muscles to burn calories and it will continue to use glucose in adipocytes to esterify FFAs with glycolysis-derived glycerol, to store fat.

So the Protons thread expects insulin sensitivity to cause fat accumulation because of maintained insulin sensitivity in adipocytes at high levels of insulin signalling. The cost of this insulin sensitivity is obesity.

PUFA = obesity, soybean oil is the best, they used safflower here.

Slight aside: The insulin resistance associated with obesity is nothing to do with insulin per se. It is triggered by the fact that very large adipocytes leak free fatty acids irrespective of insulin levels. At elevated FFA levels more insulin is needed to translocate GLUT4s than at low FFA levels.

Back to the rats.

The lard fed rats are the most obese. The PUFA fed rats the least obese.

The lard fed rats are on about 10% of their calories as PUFA in their diet. They are probably almost as fat as a 10% PUFA diet would like them to be, ie their adipocytes are almost as distended as a 10% PUFA diet dictates. The rats are almost as fat as they need to be. They are doing this on 380kJ per day. Because the rats are only allowed a total of 380kJ of energy per day. Did you pick that up in the methods?

The PUFA fed rats want to be truely, grossly obese, much more so than the lard fed rats do, because they are on somewhere between 50% and 60% of their energy intake as PUFA. But there, in the hopper, is that same old 380kJ per rat per day. It doesn't matter how much your adipocytes are crying out for more fat, how empty they feel, how hungry they tell your brain to feel. There, in the hopper, is 380kJ.

These rats are intensely insulin sensitive because their adipocytes are "empty" compared to how thery would like to be. They are "starving" compared to how they would like to be. Their muscles respond to insulin's anabolic effect and I'd be willing to bet their growth hormone levels are through the roof and IGF-1 through the floor (another post there, GH, IGF-1 and starvation). Insulin is going to be low because any glucose released from the liver is easily utilised in the fed state. In the fasting state insulin fails to act effectively so that, while FFAs may be the same as in the lard fed rats, we know (from Figure 2) that lipids are being oxidised much more rapidly on a 24h basis.

PUFA sensitise adipocytes to insulin. Given the choice the animal will eat until obese and become insulin resistant due to adipocyte distension. Combine PUFA with starvation and insulin sensitivity will be maintained. Or enhanced.

Just the Protons view. Any other explanations welcome.


Of course people should ask how the action of PUFA compares to the action of metformin. They are superficially similar. That might need more doodles I'm afraid!


raphi said...

This papers makes seed oils seem all that more prominent in terms of fatty livers, obesity and diabetes

Peter, what confidence level do you have in measures of RET in human cells? Any hopes of direct in vivo measures on the horizon?

Peter said...

It's really difficult raphi. The problem is that some of the inputs to the CoQ couple are tonic and some are intermittent. Mitochondrial preps are ephemeral and you can't supply direct inputs using reduced electron transporting flavoprotein, one of the main background inputs. Adding palmitoyl carnitine turns the whole Krebs Cycle, obviously including complex II. That's why the Protons ideas are a bit of a credo. I find it supportive that there are serious researchers who view NADH as the cytoplasmic energy indicator and CoQH2 as the mitochondrial marker. I should go and follow on from the further publications by these groups.....


Peter said...

raphi, first hit!!!! Talk about confirmation bias..... Haven't read the paper yet, just the abstract......


raphi said...

@peter fancy that!

just looked at their methods and they don't say much beyond MitoSOX and dichlorofluorescein with regards to how RET is measured. This paper gives a summary of the problems with techniques seeking to identify where ROS originate in mitcohondria:

"Essentially there are two major hindrances to the accurate identification of sites and rates of ROS production in cells. The first is the intracellular ROS-detection methods that currently are available. The accessible and widely used fluorescent ROS-detecting probes, such as dichlorodihydrofluorescein and dihydrorhodamine, have pitfalls that lead to interpretation issues (Wardman, 2007). However, it is possible that probes such as MitoSOX and dihydroethidine will provide more reliable results for the accurate measurement of ROS in cells (Mukhopadhyay et al., 2007; Song et al., 2007). Second, the use of inhibitors raises the same issue that we encounter in experiments with isolated mitochondria. It is difficult to tease apart the origin of the ROS in a complex cellular system without the use of inhibitors, and yet inhibitors enforce non-physiological constraints on the system. Also an issue in cells is that ROS may arise from sites other than mitochondria, such as the NADPH oxidases or cytochrome P450 enzymes (Caro & Cederbaum, 2004; Babior et al., 2002), and this complicates the interpretation of ROS observed in vivo even further"

Peter said...

Huh, I've blogged about that paper before.....


DLS said...

hi peter, i know you posted this for me! tanks! lol
yep still eating 100g peanuts +15g almonds
+ 5- 6 eggs a day 24x7x365- my total cholesterol is... INSANE! (even higher than my "milky" days)
no health issues . basically im the leanest 39 old guy over here.
- ifin effortlessly and taking 1000 mg metformin. lol

macros according cronometer...

113 prot
38 carbs
142 fat... which 61 mono / 48.2 sat / 16.6 poly !!!

11% pufa?.. so what kind of rat im i?

looking for some kind of "low GI snack" 2 replace the nuts for my pre meal... any ideas? fiber helps my ifing so no greek yogurt, coffe and cream plz.

Peter said...




DLS said...

i wish u had those!...
thanks Peter!

DLS said...


Tucker Goodrich said...

Glad you liked the paper, Peter. I figured it was right up your alley,but forgot to send it.

Glad you saw it anyway!

DLS said...

i -suspect- the real conclusion here is..." a high safflower,linessed, sunflower - insert refined shit- diet is super unhealthy in sedentary caged rats" !!

on the other hand i have hard time believing i may have or develop fatty -anything- eating somewhat "natural" -lolz- occurring pufa.... just looking at the mirror, at my waist , n sustained low bodyweigth... i mean no offense Truker i know you avoid pufas like the plague but im -literally- half the man you are, i do not run ( xpet chased!) barely train, and most of my 64 kgr are lean mass... ( btw not only that...recent MRI shows non existent intra-abdominal fat...) im my exp rice, potatoes, dulce de leche, starch, fructose + any fat is WAY more destructive than pufa -from pnuts- maybe is the choline, vitamin E, metformin? or then again my liver could be extra fried by now! the science looks solid...

DLS said...

human study

Both groups gained similar weight. SFA however markedly increased liver fat compared with PUFA and caused 2-fold larger increase in VAT than PUFA. Conversely, PUFA caused a nearly 3-fold larger increase in lean tissue than SFA. Increase in liver fat directly correlated with changes in plasma SFA and inversely with PUFA.

see... but what if you eat SFA in the form of cream an buter and not muffins high sfA from... palm oil!

Peter said...

Well, that's a fascinating study! Got some thinking to do there, first read and the study sounds quite genuine...


Unknown said...

Interesting post. What I'm even more curious about is your slight aside:

Slight aside: The insulin resistance associated with obesity is nothing to do with insulin per se. It is triggered by the fact that very large adipocytes leak free fatty acids irrespective of insulin levels. At elevated FFA levels more insulin is needed to translocate GLUT4s than at low FFA levels.

Rather than me guessing, can please you elaborate more about what you're saying here.


raphi said...


I did a podcast episode with Gabor Erdosi (of the Lower Insulin Facebook group) where the gist is that enlarged adipocytes overloaded with lipids lead to insulin resistance

Monica said...

Speculation on the muffin study:
1. Sounds like the 'pizza effect' with sugar on top. People with diabetes have high blood sugar for days after eating pizza (or any sat. fat & high carb combination I guess). The experimental subjects were young and healthy, and the blood sugar didn't go up, but the liver is the gatekeeper for the incoming carbs, so it was affected. Plus, almost half the carbs were actually sugar.
2. I noticed that the subjects in the SFA group tended to have slightly more liver fat to begin with, despite having slightly less visceral fat... predisposition to fatty liver?

Peter said...

Monica, yes. I've been speculating in my head long those lines. Ultimately you can't say but it's studies like this one and some of the references cited which I find really interesting. There are so many aspects which must carry an explanation other than PUFA are great for you... The Protons idea suggest PUFA offset the effects of sugar but these should be at the cost of obesity and fatty liver. What is it about the participants here, or the set up, which just shows the beneficial aspects? Interesting.


Peter said...

John, I've not had time to listen to raphia's discussion yet so don't know what it covers. The uncontrolled leakage of FFAs from distended adipocytes is pretty non controversial and the insulin resistance caused by FFAs is also pretty well accepted. Whether this is from the Randal Effect or from cytoplasmic FFA derivatives interacting directly with the electron transport chain is much less clear cut... Or from RET as per Protons.


Typo and slight addition

Eric said...

Ok, I admit I am lazy and have not looked up the papers today. Two things buzzing around my brain:

1) rats/PUFA: did they measure the composition of the lard or assume FDA defaults? there was this much publicized study that today's lard is essentially linoleic acid, Peter, you even quoted the study.

2) people/palm oil muffins: if the carbs were mainly sugar, then half the sugar was presumably fructose. From previous blog entries by Peter, PUFA + fructose = fatty liver, SFA + fructose = healthy liver. what is different in this study?

Peter said...

"what is different in this study?" Exactly! There are others too, in the refs section of is paper.


Dophamn said...

I've thought quite a bit about this in the past (i.e. fatty liver and fat distribution as effected by fatty acid and carb sub-fractions). I don't currently have any of the studies at my disposal, and the following are just my vague ideas from memory:

- although it's popularly assumed/implied (by those who write studies in journals) that things like fibrosis, cirrhosis, "metabolic syndrome" et al are a "natural" follow-on from basic fatty liver (aka steatosis, hepatosteatosis and goodness knows what else), this doesn't really seem to be the case (i.e. the more serious things don't seem to result automatically when some simple threshold level of liver fat is reached, and could potentially take root at relatively modest levels of liver fat, or be more or less absent at relatively high levels of liver fat)

- I've seen some conflicting effects on liver fat from studies using various species, feeding regimens, control methods etc. etc., but things like fibrosis and cirrhosis (wherever found) always seem to be linked to sugars and/or PUFAs and/or ethanol

- SFAs and MUFAs with 16-18 carbon chains are the body's "preferred" form in which to store fats, and in which to export them into other tissues, so that the liver and other tissues are probably less liable to "panic" (and thus less liable to upregulate metabolism/excretion or other things) in the face of a modest shortish-term over-abundance of "preferred" fats

- with just a modest over-abundance of "non-preferred" fats, it's quite conceivable that (especially over the shortish term, but possibly long-term as well) the liver/body's reaction will be quite "anabolic" (inefficient fat storage/retention, insulin sensitive muscle cells that stock up on glycogen and water)

- with a more wanton and/or sustained over-abundance of "non-preferred" fats (as more typically employed in animal studies), or with more modest and short-term exposures in persons who are already insulin resistant with fatty and/or fibrotic livers, it is quite conceivable that irreversible catastrophe shall ensue (especially if the "non-preferred" fats are administered in a readily oxidizable form)

- I probably won't die on account of my fondness for consuming hundreds of grams of lightly roasted pistachio nuts on a semi-regular basis

- most "natural" high-fat foods (regardless of fatty acid composition) also contain assorted other factors that make it incredibly challenging to develop "metabolic syndrome"; isolated oils are not the same (especially when added to fructose-fortified muffins)

Both groups were deliberately overfed. Both gained the same amount of weight. The palm-fructose muffin group gained additional fat in a not-evidently-pathologic way. The PUFA-fructose muffin group gained (probably) additional "lean water".

Dophamn said...

A couple of relevant recent studies I found:

- an isoenergetic high-carb, high-fibre, low-GI diet (52% carbs, 30% fat) had no effect on liver fat over 8 weeks, whereas an isoenergetic high-MUFA and higher-GI diet (40% carbs, 42% fat) reduced liver fat (and seemed to increase insulin sensitivity as well): (full text available), (insulin, abstract)

- a week of wanton overfeeding with glucose (3g/kg), fructose (1.5g, 3g or 4g/kg) or "30% excess energy as saturated fat" resulted in notable degrees of liver fat accumulation in all groups (except 1.5g/kg fructose added to drinks!), but liver insulin sensitivity was maintained only on the "30% excess energy as saturated fat" diet (the full text doesn't condescend to specify the source from which this was taken nor the form in which it was given):

Zoran Markovich said...

Hello Peter,

Thank you very much for your direct answers. "PUFA in RATS" further elaborates this topic.

Yes, I support lower liver insulin resistance but not more insulin signals. Even better, I support less insulin signals but more insulin signalling. Or do they have to be hand in hand?

And, I want minimal liver insulin resistance, but not more NADH vs. NAD+. Does it always go together?

Re. rats. They gave them linolenic acid. Suppose that rats were fed with EPA and DHA , is it possible that oxidation via peroxisomes and through mitochondria shows a neutral position regarding insulin resistance? Why then take omega 3 fats at all? For me personally it is important for their anti-arrhythmic properties. May plasmalogens prevent oxidation of omega 3 in lipid membranes?

I hope that some of my questions will be interesting enough for you to answer.

Best regards,

Zoran Markovich

Dophamn said...

It behooves me to add that I have also happened upon some recent-ish studies (in rats and their close relatives) that don't fit well with the working model I'd formed in my memory from before:

- 3-month-old male wistar rats were given FREE ACCESS, for 8 weeks, to diets containing 35% carbs (mostly corn starch) and EITHER 45% safflower oil OR 45% cocoa butter (my personal favourite form of fat); the outcomes were that the ones given cocoa butter ate more food/calories but didn't gain more total weight/fat, but that the ones given safflower oil gained MORE liver fat ... and yet the ones given cocoa butter (guaranteed very-low-PUFA) exhibited MORE evidence of liver fibrosis and inflammation (and other forms of stress/injury): (one of the annoying studies where the methods, design info etc. are at the end, after results+discussion)

- 6-week-old male Sprague-Dawley rats were initially restricted to half rations of chow pellets (60% carbs, 30% protein, 10% fat) for 14 days, and then (for a further 14 days) given (with 21% carbs and 21% protein) 58% calories in the form of fat from EITHER lard (ostensibly low-PUFA) OR mixed safflower/linseed oils, in CONTROLLED hyper-caloric amounts that matched the observed intakes of spontaneously growing weight-matched control rats; the outcomes were that the high-PUFA re-fed rats gained less fat and less VAT and WAT but more lean tissue (similar to the human study from 2014), along with more BAT (brown adipose tissue) and more UCP-1, and exhibited less liver inflammation and lipoperoxidation:

- rodents additionally given non-fat cocoa components seem to become highly resistant to liver fat and WAT accumulation etc., and have more UCP-1, more PPARa signalling, more liver insulin sensitivity etc.:,,

Peter said...

Argh, double edited for typos!

Hi Dophamn,

Another interesting study in

You have to be very careful with your terminology here. I would agree that the stearic acid rats stayed comparable in weight to others despite eating more calories than either CIAB or PUFA fed rats. There is NO evidence that they developed inflammatory changes in their livers! They had a statistically (but biologically significant?) significant increase in messenger RNA expression for seven genes associated with inflammatory liver disease. The question is whether these mRNA changes actually result in inflammatory changes in the liver, or are they markers of the normal response to reverse electron transport though complex I derived superoxide which might also trigger life extending increases in SOD and/or catalase?

So what we need to know is whether there is histological evidence of NASH development. After all, we know from the methods that they took terminal samples of liver and snap froze them in liquid nitrogen. Either sticking some in formalin at the same time or getting histology done on the frozen samples (not ideal from the histologist point of view but quite possible) would allow them to correlate their mRNAs with actual damage in the liver. They didn’t do this.

So why did they freeze liver? As they say:

“Liver alanine aminotransferase

Liver tissue was homogenized in buffer (100mM Tris, pH 7.8) and alanine aminotransferase (ALT) concentration was determined from SUPERNATANT via manufacture instructions (Cayman Chemical, Ann Arbor, MI)”.

In the results section Figure 4 this is converted to:

“PLASMA alanine aminotransferase concentration was higher in SAT compared with CON and PUFA”.

[My shouting capitals on supernatant and plasma]

Plasma???? No. Liver homogenate! Plasma ALT is an absolutely routine, standard, everyday marker of liver damage. It is a surrogate for hepatocellular damage, i.e. a normal component of liver cytoplasm which has leaked in to plasma in response to liver injury. It’s measured every day in any patient undergoing any sort of health/illness monitoring blood work. It is a COMPLETELY normal cytoplasm component while it is contained within the liver hepatocytes. It is LEAKAGE we are interested in as a surrogate for hepatic damage. The rats all had terminal blood samples taken. The group could have measured ALT for a few pence in real plasma, or for a few pounds at a commercial lab, on this blood. They didn’t. They homogenised liver and measured ALT in the supernatant. They described this as “plasma”. All we can say from Fig 4 is that the liver of stearic acid fed rats has more of ALT within its hepatocytes. ALT is a normal peripheral component for interconverting certain components of the TCA. Who knows why it is increased under stearate feeding, but it's not a marker of hepatocellular damage unless it is released...

Why, why, why did they do this??????

The related studies they cite are equally interesting. As always, my fascination is about the mindset of the research group.

Who decided to homogenise liver to get “plasma” ALT? Who decided to bin the histology friendly liver samples?



cfmorais said...


And about Erythritol action?

Another mode of action similar to metformin?

Peter said...

That would need a lot of reading! You have to wonder how much of the effect might be through the gut microbiota... Ditto metformin, in addition to systemic effects.


Dophamn said...

Alas, although I must have read through thousands of study papers over the past decade-plus, and think I have an excellent working knowledge of nutrition, I still understand very little of the methods sections (just an endless succession of exotic-sounding substances, processes and apparatus about which I know nothing in practice). So it's easy to miss some methodological sleight of hand, and to attribute anything that seems a bit off to my ignorance (and/or to the shit writing abilities of the research team). I just glide through the methods to get a handle on the design and whether/how any potential variables were controlled, but then have to take it on faith that the technical stuff was sound.

In regard to mRNA, I have noticed in the past (in other realms unrelated to nutrition) that higher RNA levels of a thing don't always go hand in hand with higher protein levels of that same entity, so that it isn't even possible to be sure what it represents.

It's unsettling to encounter findings that contradict my working models, but I never want to sweep them under the carpet. I want to be able to explain everything.

The trouble, though, is that, to the average person who never reads studies (who doesn't even realize how many millions of them there are) and who believes that nutritionists etc. are in receipt of divine inspiration (and that any remaining ailments, after years of rigid adherence to their wisdom, must be punishment for original sin and/or for festive indulgences in "high-fat" bags of Smarties), seeing studies like that (e.g. "PUFAs make you jacked and shredded, and SFAs initiate NAFLD and metabolic syndrome") would seem like the most amazing, ultimate, extra-uber final proof that saturated fats are alien toxins that were disseminated across earth by anthropomorphic reptiles.

I like the idea of SOD extending life-span. Personally, I know that I was in a very mediocre state of health into my mid 20s, before I'd even grasped the concept of "nutrients". I know that my skin didn't look very healthy when (after my initial, naïve, over-enthusiastic forays) I adhered to a very-low-fat, high-starch, high-fibre, low-GI diet with a high PUFA:SFA ratio and "perfect" n3:n6 ratios. I know that now, nearing 40, after more than a decade of consistently consuming ~100g of SFAs every day, I'm in better health than back then and that people still assume I'm in my mid 20s.

But many of the same people who are surprised by my health and youthfulness also react to me chastisingly when they see me snacking on chocolate (dark or not-so-dark) or when they learn that cheese and egg yolks were the primary contributents to my breakfast. They react as though I'm behaving in an immoral and degenerate way, but getting away with it (for now) thanks to some special, lucky genes that were of no use to me up to my mid/late 20s (back when I ate in the ways that most other people eat, either blindly or in zealous adherence to nutritionists or others who are assumed to be "in the know").

As for the fluctuating low, not-so-low or casual/intermediate carbohydrate content of my diet, my experiences have always seemed very similar to those of the people in this study: (i.e. always feeling more healthy in general when eating lots of fat with low-carbs/PUFAs, but finding certain more "intense" aspects of some forms of exercise harder and showing diminished performances on some tests, and yet simultaneously finding it easier to recover from exercise and feeling more enthusiastic about indulging in those aspects of it that feel innately enjoyable). The "Athletes' Experiences" section of the Discussion contains some interesting snippets.

Peter said...

Well Dophamn, as you realise, the world is mostly populated by sheeple. At least if there are now LC extremists out here then there is an option for sheeple to follow towards better health rather than worse health........


Peter said...

BTW, you might have notice that I tend to read the methods. Discussions which miss quote their own results are a gauge of the cognitive dissonance of the authors. Alternatively, the methods is where they cook the books to avoid that cognitive dissonance, assuming they see it coming! I was taught how to read a study as part of my BSc degree back in the mid 1970s. Has stood me in good stead.


Peter said...

Cathy Wilson! I've been trying to remember the tutor who explained how "most people" read a study (abstract and discussion at the max, usually just the abstract). Results and methods become essential if you are challenged by the abstract and discussion. Many thanks to Cathy, all those years ago. She did a good job. Threw good parties too.