Thursday, September 03, 2009

When is a high fat diet really a high fat diet?

The paper I mentioned some time ago about the preservation of both fat oxidation and cardiac function on a high fat (60% of calories from fat) diet compared to Western diet (45% from fat) is now available in full text for free. The paper is long (10 pages) and detailed and makes anyone touting an apoE-/- mouse, or a fat stupid one, look pretty dumb. It will take some analysis and it's the World Congress of Veterinary Anaesthesia at the moment so I'm not doing detailed reading this week.

A minor snippet in abstract form which further backs up the idea that you have to manipulate your experimental set up to show the adverse effects of high fat diets is this one. Having lots of mitochondria appears to be a good thing. This fits with Power, Sex and Suicide. One approach to minimising free radical damage is to have lots of mitochondria running at tickover (with uncoupling proteins in abundance). It also suggest that if you are going to have a coronary occlusion, better be in ketosis if you wish to recover! Of course the occlusion is a bit unlikely if you are in ketosis anyway....

Are either of the papers dubious? Well, they fit with my personal experience of what it feels like to eat a high fat diet. But then I'm not eating 16% of calories as refined sucrose!



Stephan Guyenet said...


The high-fat diet also contained the least sucrose.

Low fat = 350 g per kg
Western = 173 g per kg
High fat = 69 g per kg

I think the problem with the "Western diet" may be the combination of PUFA imbalance (industrial lard plus soybean oil; industrial lard has a fair amount of n-6 and very little n-3) and excess sugar. The low-fat mice escape problems because they eat so little PUFA that they can tolerate the sugar.

That's my take. But feeding rodents high fat diets typically causes fat accumulation regardless of the fat source. They become hyperphagic. In that sense, it doesn't line up with how humans react to high-fat diets.

Aaron said...

I think it's valid to worry about mitochondrial ROS alot -- it is the foundation of the aging process.

The study in question makes me question heavy protein intake for sure.

Fats or carbs dont seem to matter much in terms of mitoROS

Seems as if one should raise carbs to just the point where you can satisfy your bodies' requirement for protein with the minimal amount (so you don't need to convert much protein by gluconeogenesis)-- at least in terms of methionine.

Peter, what are the best ways to raise uncoupling proteins-- I always thought polyunsaturates do that the best-- and we usually try to minimize them.

We have to remember-- if carbs indirectly cause less mitoROS by creating individuals with less muscle mass, less mitochondria- and therefore less damage from them-- carbs may be good for longevity as long as you keep yourself thin somehow.

Look at the Japanese, low muscle mass, high uncoupling proteins from omega 3s and slight CR- and high carb intake-- and good longevity.

While I know that on a hyperlipid type of diet you will have low insulin-- do we know for certain if someone on a high fat- low protein diet will have a lower mitoROS then someone on a higher carb, lower protein and fat diet?

The low carb camp has to show that you can have lower mitoROS on its diet irrespective of insulin to make me think there are longevity benefits to the high fat/low carb way of eating.

Otherwise- I'd use low carb for building muscle and staying thin-- and eat more carbs if i wanted to live longer (if i could figure ways to stay thin)

Stan Bleszynski said...


That paper says:

Previous studies have shown that the decrease in mitochondrial
reactive oxygen species (mitROS) generation and oxidative damage to
mitochondrial DNA (mtDNA) that occurs during life extending dietary restriction also occurs during protein or methionine restriction, whereas it does not take place during carbohydrate or lipid restriction.

That would confirm Dr. Jan Kwasniewski's advise that protein have to be limited (to 1g/kg/day for humans, +/-15%) just as carbohydrates.

He was also quoted as saying that if you really have to overeat something on a HF diet, it is better to exceed carbohydrates than protein. Extra carbohydrates will just get converted to extra fat by the liver and stored.

Stan Bleszynski said...


Very interesting: - this is the first study (I see) that seems to confirm Kwasniewski's 45% fat "forbidden zone"! That is that the intermediate fat diet, between the low fat and the real high fat is more harmful than either one!

Aaron said...

Stan, not sure if my wording was right-- but my comment already agreed with what you said-- especially for low protein!

Remember, he found no difference for fat or carb mitoROS-- but what happens over time when weight changes- muscle mass changes and so on? I'm arguing that a lower muscle mass may turn out longevity benefits by lowering the amount of mitochondria present in the body. Growth hormones increase on higher fat diets-- a boon for muscle-- maybe no so for longevity.

I don't worry about insulin levels if your bodyweight is kept low-

optimal carb/fat intake for longevity in not clearcut yet.

ItsTheWooo said...

Sephan, Peter, others... any possibility that it's not the diet at all and perhaps the underlying cause of metabolic disorders in the west is actually related to industrial pollution? I am rather fond of this idea, Jenny at diabetes 101 is too. It's well known many industrial pollutants damage mitochondria and induce insulin resistance as a compensatory adaptation (if you don't have enough mitochondria to process normal influxes of glucose after eating, cells develop secondary insulin resistance as a way to trigger hyperinsulinemia thus the repartitioning of glucose into fat, which spares mitochondria from more toxicity secondary to high levels of fatty acids and glucose which are bound to occur with disabled mitochondrial function... a very low carb is merely a symptomatic work-around for this metabolic disorder which avoids the "glucose influx" part and it's not necessary for people without debilitated metabolism, such as non-industrialized people).

I don't see how PUFA is the problem, if PUFA is only converted into inflammatory cytokines when hyperinsulinemia already exists (and, the hypothesis goes that this inflammation then causes insulin resistance). The enzyme is under control of insulin, so high insulin levels need to be present for PUFA to induce more insulin resistance. PUFA may contribute, but never cause because of this fact. High insulin had to be present before.

Actually humans do respond to high fat diets with metabolic disorders and hyperphagia. The most fattening diet is going to be a carb/fat mixed diet. Donuts, fried food, it's all high in fat and carb.

If you reduce the fat content of a mixed diet, without increasing carb, people will see reductions in metabolic symptoms (glucose, insulin, weight). Low fat diets do technically "work" (when compared to how typical americans eat), they only don't work when compared to the much more effective very low carb/high fat diet (which generally works the best for metabolic disorders).

ItsTheWooo said...

BTW, carbs tend to augment muscle mass (assuming you are not insulin resistant) because insulin / leptin (the "fed state") increase IGF and free testosterone. Carb restriction will lower body weight as well as lean mass for this reason... reduced growth/repro hormones. Fat people lose both fat tissue and lean mass (bone/muscle) when they go on slimming diets, carb restricted or otherwise. A decline in insulin means a decline in IGF-1 and often a decline in reproductive hormones all of which promote muscle mass.

Now, I always associated uncoupling proteins with low insulin/high fat. UP is a fatty acid thing.

I always knew protein sucked and was marginally less deletarious than carbohydrate (sometimes carbohydrate is actually advantageous... I've noticed a little bit of sugar with fat tends to give me huuuuge energy). I try to avoid over eating protein since it invariably makes me very hungry and depressive too but I also find it difficult to eat "low protein". I couldn't do 1 g/kg, that would give me about 55 grams per day and I would be too hungry on that. I need at least 80 grams of protein to be satisfied and avoid protein-craving hunger. But then when I eat more than that (like, 120+) I start getting hungrier and hungrier.

Aaron - increases in growth hormones on high fat diets are secondary to increases in insulin, because most "high fat" diets are actually carb/fat diets which are insulinogenic. So, any changes observed on a growth-hormone raising diet must be observed int he context of an insulin raising diet, because growth hormones generally can only ever increase when insulin does. And we are all familiar with the problems associated with hyperinsulinemic conditions, regardless of whether or not it was fat or carb induced.

I would definitely worry about insulin levels even if body weight is low. First, high insulin with low body weight is absolutely pathological. High insulin with high body fat may also be pathological but not as much as the previous condition (which suggests not only metabolic disorder but one that is atypical and less understood). Insulin is an inflammatory hormone (think cardiovascular diseases) and also regulates lifespan in lower animals. Since insulin also increases expression of growth hormones (body fat/muscle/bone) this has some relationship to cancers, particularly sex steroid cancers (prostate/breast) and these are all more common in hyperinsulinemics and those with probable hyperinsulinemia (obese).

Thin people should have low insulin, healthy people should have low insulin. High insulin translates into difficulty using nutrients on a cellular level, inflammation, and general poor health.

Faith said...

Hi Peter,

Awesome site. As someone with HLA-B27, I am trying to understand the implications of eating starch. Are you saying that everyone with the HLA-B27 antigen will undergo this process in which their immune system attacks pullulanase resulting in the collateral damage to the collagen, irregardless of whether they have AS, presuming they have any level of klebsiella in their gut? Therefore, anyone with HLA-B27 should avoid/reduce starches, irregardless of whether they have AS? And that they are also at risk of developing cervical spinal arthritis even if they don’t develop AS? What about other forms of spinal arthritis?

I was tested for AS in 2005 and clear on CT scan. Gave up gluten, casein due to intolerances (double DQ1 genes) and went on low starch diet for several years anyway due to gut and systemic problems. There was indeed great resolution of the stiffness. Nevertheless, I’ve developed severe food intolerance probably from an already bad gut and continuing to eat high oxalate food (had crystals in urine) and have been unable to eat meat for the last year. So I ate non-gluten grains for a year, now just white rice, but getting the soreness in the hips and low back. Sounds like this should be a red flag for me from what you are saying and that I may be endangering my collagen.

Any input you have would be greatly appreciated. It’s hard to dig out of this hole.

Thanks again for all the information.

Anonymous said...

soo it would make sense, as i gather this in, to have a 75-80% fat diet...with 15-20% protein? or should protein be lower, or should there be some carbohydrate sources? eating only enough protein for the body should keep insulin levels steady right?

blogblog said...

the Japanese live only 1% longer than Westerners. However when you adjust for homicides and car accidents (no guns and extremely strict road rules in Japan) the Japanese have a lower life expectancy than Americans.

blogblog said...


only guinea pigs, bats and higher primates ever suffer ischaemia regardless of what they are fed. They are also the only mammals that don't synthesise vitamin C. So the most logical explanation for CHD in humans is a lack of vitamin C. This is the basis of the Pauling-Rath Hypothesis.

Wild primates, bats and guinea pigs have extremely high intake of vitamin C and antioxidants (equivalent to >2500mg/day for a human >60x the RDI). Some wild fruits have vitamin C levels more than 50x as high as oranges.

Contrary to widespread mythology the Masai eat a considerable amount of phytonutrient-rich plant materials including bark and roots. The Inuit have very high selenium intake which is an extremely potent antioxidant.

Cardiorespiratory exercise also stimulates the production of extremely powerful endogenous antioxidants.

It is important to note that people with Gilbert's Syndrome (high bilirubin production) have reduced atherosclerosis. Bilirubin is a very powerful antioxidant.

blogblog said...


All mammals evolved from insectivore/carnivore ancestors.

Wild mammals typically obtain 3-5g/kg/day protein either from direct consumption or gut fermentation. Even cattle on a grass diet obtain nearly all of their nutrition from volatile fatty acids and protein not carbohydrates.

Stefansson and Andersen ate only ~1.5kg/day of meat and offal during the Bellevue Hospital Study.

Peter said...

Hi Stephan,

There have been one or two other papers where a diet of almost complete sucrose has actually come out quite well, provided the fat intake is low enough. It does get a bit tedious trying to work out how these diets have been developed to produce the desired end point without any information on the tweaks which were needed to get there. Just occasionally, as for the apoE-/- fiasco you just have to try and work out what's going on. Never did get a reply re Ca/PO4 amounts or ratio from Rosenzweig on this, or from Murray on anything to do with his fat stupid rat diet. Wasting funding showing that high fat diets were beneficial in the longer term was not on the agenda!


Peter said...


I would guess all high fat diets raise UCP3. I'd see the logic to PUFA working better than saturated fats (if they do, I've not chased this) is that PUFA need burning asap. They seem to beta oxidise perfectly well and no one would want excess linoleic acid hanging around to go rancid in their cellular systems. Bit like glucose really. I don't see glucose as a technique for minimising ROS, there seems to be more scope for ROS than with fats (complex 1). And you need fewer mitochondria as glycolysis augments oxidation...

Interesting thread on methionine. I see Stan commented further down the list that the OD is an adequate protein diet. If we need 40g/d of top grade protein then 60g/d seems a reasonable amount to aim at. On the longevity front I think the answer is "who knows?" Personally I'm satisfied that what I'm doing sorts out multiple medical problems. Living to 70 on beta blockers can be converted to living to 70 without. It seems inconceivable to me that a move which gets rid of the medical problems should actually shorten life. If there's no life extension, at least I can feel good in the interim!


Peter said...

Stan, I still feel that Stephan is on a reasonable track with the PUFA aspect as well as the fructose. We would need information on a low PUFA low fructose forbidden zone diet to decide. I think it might be reasonable to assume that many of the benefits of LC are from fructose restriction. Getting PUFA truly low is quite hard even if you are dairy based, as even a small percentage of PUFA in 80% of your calories adds up. As we both know JK has no specific advice about PUFA except to avoid them but is happy with the amount in pork lard... If you dropped the PUFA would the fructose be less of a problem???????


Peter said...


I see absolutely no problem with bisphenyl A etc as a fat damaging chemical but I can't see it being relevant to the observations of Stefansson, Cleave and Weston Price. We keep compounding problems but I think there is a problem there to compound. The ancient Egyptians seems to manage type 2 diabetes perfectly well without industrial pollutants...


Peter said...

Hi Faith,

Sorry for the delay: In general starches are bad news if you are HLA B27 positive and personally, I would avoid them even if I was asymptomatic for low back pain. Prof Ebringer doesn't feel cervical spinal arthritis is particularly B27 related but I would expect something similar to be working. I did get myself blood tested and an B27 negative, but still my spinal problems have resolved completely on LC eating, but with starch reduction rather elimination. If you are symptonatic it is just a matter of assessing how low you need to go.

I put my approach to RA here. As would be similar but with the added avoidance of starches.

I find meat intolerance very difficult to understand in the absence of other gut pathology, perhaps triggered by casein if you are unlucky. I certainly feel that a fast followed by fat, followed by animal based protein, followed by enough LC veggies to keep you out of ketosis and oxalate free in your urine would be the approach. Ultimately you could replace the veggies with dextrose monohydrate in Optimal ice cream if casein is OK. I would be very curious to see whether oxalate crystals were a problem if you got sorted on other aspects of your diet.

Hope that helps

BTW there are good and bad HLA B27 alleles too, they're not all the same!


Peter said...

Malpaz2003, I think my diet works out at 10-15% protein but of course this varies with total calorie intake, it's 50-70g/g. I would just emphasise that what any of this will do to longevity is speculation. My interest is in feeling good and being healthy now, one day at a time, for the rest of my life... It's definitely a one day at a time thing!


Peter said...

Hi Blogblog,

I have a great deal of time for the ascorbate hypothesis of IHD. Its big stumbling block is that humans have been without vitamin C for around 40 million years and have significant CVD for perhaps the last 10,000 and the acute coronary became commonly recognised maybe 100 years or so ago. I'm less interested in putting the clock back 40 million years, I'd settle for 10,000. Stopping the nutritional clock 10,000 tears ago might stop evolution towards a grain based diet, but there are plenty of people happy to suffer the disease pressure of grain adaption. Got to keep the doctors in jobs somehow, just I'd rather it wasn't me with systemic lupus.

On the antioxidant front I would also suggest uric acid. This is a vitally important human antioxidant and adds significantly to bilirubin in the antioxidant capacity of human plasma. Plant antioxidants we excrete as rapidly as possible, with very occasional exceptions (lutein and xeazanthin seem to have some use in the retina, I doubt their plasma concentrations do much). No one should trust a plant to save their life.


Faith said...

Thanks for your generous input, Peter. I have definitely been symptomatic from eating starch, so guess I'd better stay away given the HLA-B27+ status.

About the meat intolerance, I think it could be the amines since it's difficult for me to get fresh meat. I have an elevation of mast cells (elevated serum tryptase) and many people with the same problem (e.g. mastocytosis, but in my case not necessarily primary) cannot tolerate most meat, poultry, fish.

I'm also casein intolerant, so no go on the ice cream!

I'll look at your approach to RA and consider your fast and reintroduction suggestion.

Thanks for the great blog. Faith

Faith said...

Oh, what does this mean:

"BTW there are good and bad HLA B27 alleles too, they're not all the same!"

How so you know if yours is good or bad?

Peter said...

Back when I was doing a lot of reading about HLA B27 (I tested negative despite a history of low back problems and IBS) I found that there are shades of grey of the B27 gene. It's not just a +ve or -ve and some +ve variants cause AS and some don't. The ultimate arbiter is the low starch diet... Plus I guess normalising D3, balancing out omega three to six, eliminating trans fats, going to chronic normoglycaemia and any other anti inflammatory dodge you can put in place, short of trusting plants...

Yes, the amine issue is a bummer. One of my cats vomits spectacularly on raw beef (usually aged in the UK) but is fine on chicken or pork (no one ages these unless culturing campylobacter or salmonella). I suspect amine intolerance rather than allergy for him.


donny said...

Lowering methionine changes the ratio of methionine to other amino acids in the diet. If you give rats extra methionine, you can increase the delta 6 desaturase activity, increasing production of arachidonic acid. Glycine supplementation reduced this effect of methionine here;

I don't know that methionine would lower lifespan in rats by decreasing the saturation of cell membranes, but it seems like something to consider.

Most carnivore species eat a fair amount of bone and cartilage, which would increase the glycine/methionine ratio.

To make it easier to manipulate methionine levels in the diet, you also have to either go to a diet of purified amino acids, or maybe add some purified methionine to a natural protein and call that the control diet. And with the purified aminos, what do you replace the methionine with, so that you know that it isn't just protein restriction that's working?
I think they replaced it with Glutamic acid in the methionine-restriction study that started this whole mess, assuming that that was a particularly inert amino acid to use. But is Glutamic acid so mild-mannered after Methionine leaves the room? Maybe replacing Methionine with a different amino acid wouldn't be so effective.

"Glycine intake decreases plasma free fatty acids, adipose cell size, and blood pressure in sucrose-fed rats"

Glycine connects to Methionine restriction (maybe), through it to omega-6's (or at least what happens to them in the liver, once theý're in the body), and protects rats from some of the dangers of sucrose. (According to the study I posted, the list of dangers includes, in addition to those in the title, hyperinsulinism, high triglycerides, and visceral fat.)
Almost like sugar causes a glycine-deficiency. Maybe on a very low carb diet, more glycine would be available to protect against the life-shortening effects of methionine. If there really are any.
I hope I used enough maybes' to avoid coming across as a total lunatic. %)

Aaron said...

Peter, I've noticed that a few researchers out there actually implicate high uric acid with a whole slew of diseases-- not to mention higher blood pressure and inability to lose weight-- (theres even a whole book about reducing uric acid, "the sugar fix" by richard johnson, MD

many plant anti-oxidants work by raising uric acid levels-- not sure this is a good thing

its also not surprising that fructose increases uric acid levels the most-- if the above book is correct- you can partially quench the weight gaining effects of fructose if you can quench the corresponding increase in uric acid-- (not that i want to have much fructose-- but interesting none the less)

as a side note-- I've looked through your site and am very interested to know your view on probiotic bacteria-- because you eat so much fat (which isn't food for them) and don't care to eat too much fiber or other vegetable matter (which they would use) do we really need beneficial lactic acid bacteria down there? are they important for the Butyric acid they produce which stabilizes the gut-- or what about b vitamins and such?

Anonymous said...

Fascinating paper! Whaty's most fascinating is that they seem to have set out with the intention of proving fat causes heart disease, found it wasn't so but published anyway.

It lends some credence to the concept that fat is only a danger when added to toxic levels of carbs while also permitting a low fat high carb diet to (marginally) improve on the SAD. Another U curve, or J curve more likely - and shows that adapting to fat metabolism is something that once done you may not want to reverse.

"its also not surprising that fructose increases uric acid levels the most"

Hmmm, also interesting, mother has had attacks of gout while avoiding purines and is pretty convinced the culprit is rhubarb. So the fructose bangs up the uric acid and the oxalates may be the rest of the equation and all the Usual Suspects are innocent. May also explain why I had gout as a vegetarian, eating too much healthy fruit . . .

Peter said...

I've got a fascinating paper on dogs and insulin resistance which gives some suggestion that total caloric intake is involved. Of course, total caloric intake can only be high if there is enough insulin to store the excess energy, which generally requires more than a certain threshold of carbs. Dogs and humans have a huge amount in common. I think it was Stan who suggested that it wasn't for vegetable leavings that dogs teamed up with humans... Though Hills would tell you different!