Wednesday, May 18, 2011

Melting Temperatures of Dietary Fats

Click on image for larger version.  Source:  Bettelhiem, Brown, and March:  Introduction to General, Organic, & Biochemistry, Sixth Edition.  Page 472.

The human body operates at about 95-100 degrees F. Notice that the saturated fats don't melt--i.e. become liquid--until you reach more than 110 degrees.

 The unsaturated fats all remain liquid at temperatures well below human body temperature.

Fats travel in the body packed in triglycerides:  three fatty acids bound to a glycerol backbone.  The following depicts a triglyceride composed of three saturated fatty acids:

In this picture, the little grey dots represent hydrogen atoms.  Notice that in such a triglyceride, the little grey dots lie close to each other, stacking cleanly like Lego blocks.   When triglycerides containing a high proportion of long-chain saturated fatty acids are packed close to one another, the hydrogen atoms of fatty acids on one will bond to those on another by London dispersion forces, which makes the fat firm at low temperatures; to melt it, you have to add heat to break the bonds between the hydrogen molecules.  In contrast, the following shows a triglyceride containing one unsaturated fatty acid (the one with the bend):

 In this type of triglyceride, the top two fats form bonds between 9 hydrogens, but because the bottom, monounsaturated fat has a bend in it, the middle and lower fats bond only at 6 hydrogens.  Because of this, a mix of triglycerides containing unsaturated fats have weaker bonds to one another and this makes the fat less firm; it will melt at a lower temperature.

If you substitute a polyunsaturated fatty acid for one or more of those fats, the triglyceride will form weaker bonds to others.  Also, if you use shorter chain saturated fats (e.g. coconut oil), there are fewer hydrogens available for binding, lowering the melting point of the fat.

Most dietary and endogenously produced triglycerides are like the second example, a mix of fatty acids, some saturated and some unsaturated.  Nevertheless, the higher the number of saturated fatty acids in a mix of  triglycerides, the higher its melting point.  Thus, beef tallow, a mix primarily of long-chain saturated and monounsaturated fats, has a higher melting point than coconut fat, a mix of primarily medium and short chain fats, and both of these have a higher melting point than flax oil, in which the triglycerides consist primarily of polyunsaturated fats.

This is really obvious in our experience.  Beef triglycerides are solid at room temp, and start to melt around 100 degrees.  Coconut triglycerides melt at about 75 degrees, and fish and flax triglycerides stay fluid even at temperatures below freezing. This means beef triglycerides are more viscous/less fluid at body temperature than flax triglycerides.

Similarly, compare the texture of cold raw beef to cold raw salmon.  The former is stiffer, in part because the fats in beef are more solid at refrigerator temperature than are the fats in salmon.

Nature also distributes the fats differently. Warm blooded animals and plants living in warm enviroments tend to have higher proportions of the saturated fats with higher melting points, e.g. beef and palm oils.  Cold blooded animals and plants and animals originating in cold environments tend to have higher proportions of unsaturated fats with low melting points, e.g. reptiles, salmon, and flax.  If salmon had a high proportion of saturated fats in their tissues, their bodies (muscles) would not be fluid enough to swim in Arctic oceans.  If flax plants produced saturated fats, they would not be fluid enough to transport about the flax plant in the native environment. 

The body uses different types of fatty acids specifically implementing their abilities to form more or less solid structures.  For instance, the fat pads of the palms and feet have a high proportion of saturated fats, because this makes a more firm, noncompliant structure for padding.  On the other hand, nerve membranes have a high proportion of highly unsaturated fats (specifically, arachidonic acid and DHA), presumably because these membranes must have a less firm structure in order to serve their purposes, which involve rapid fluctuations in sodium and potassium flow across the membranes.

These examples show that the melting points of fats (triglycerides) are very important in biological functions, and that nature naturally favors the use of unsaturated fats to keep things fluid and flexible, and the saturated fats to make things more stiff and noncompliant.  

Taking a look at how nature uses these fats for differential effects, we might predict that these fats will have different effects on blood flow.  increasing the amount of saturated fats in the triglycerides in mammalian blood stream will increase blood viscosity, which would impede flow of blood through the tiniest capillaries, impair delivery of oxygen and nutrients to and removal of wastes from the tissues (thus promoting ischemia and toxicity), and increase blood pressure.  Since oils in general are viscous compared to water, it also predicts that reducing total fat, or replacing high-melting-point fats with low-melting-point fats, would have the opposite effects.

Tai et al fed rats diets supplying either 10% of energy as soyabean oil (control group) or 40 % energy from soyabean oil (USFA), palm oil (SFA) and vegetable shortening (TFA) for 8 weeks.  They found:  "rats fed high-fat diets exhibited significant increases in serum TAG levels (P < 0.01), plasma viscosity (P < 0.01), whole blood viscosity (P < 0.01) and internal viscosity (P < 0.01) compared to the controls."

Hall reviewed the available data and reported that although good-quality intervention data on dietary fatty acid composition and vascular function are scarce, so far what we have indicates the following:

1) A single high-fat meal can impair endothelial function compared to a low-fat meal, apparently due to increased circulating lipoproteins and nonesterified fatty acids which may induce pro-inflammatory pathways and increase oxidative stress.

2) Cross-sectional data suggest that saturated fat adversely affects vascular function whereas polyunsaturated fat (mainly linoleic acid (18 : 2n-6) and n-3 PUFA) are beneficial.  

3) The superunsaturated, very low-melting-point fats EPA (20 : 5n-3) and DHA (22 : 6n-3) can reduce blood pressure, improve arterial compliance in type 2 diabetics and dyslipidaemics, and augment endothelium-dependent vasodilation.

Steer et al performed another human study.  Saturated fatty acids impaired endothelial function, and alpha-linolenic acid (omega-3) improved endothelial function. 

These findings alone might help explain why the Masai have atherosclerosis.  The dairy-based Masai diet would increase blood viscosity, impair blood circulation, and induce vascular tissue ischemia, oxidative stress, and inflammation, compared to a lower fat diet.

As noted by Stephan, Mann et al tried to blame the atherosclerosis of the Masai on foods other than dairy fats, based on the fact that the disease appeared prevalent only in Masai after the age of 40 years, after the Muran period during which they lived almost exclusively on dairy products.   Mann et al believed that some noxious agent introduced after the Muran period had to account for the atherosclerosis:

"We believe... that the Muran escapes some noxious dietary agent for a time. Obviously, this is neither animal fat nor cholesterol. The old and the young Masai do have access to such processed staples as flour, sugar, confections and shortenings through the Indian dukas scattered about Masailand. These foods could carry the hypothetical agent."

This line of reasoning has a major flaw: Simply, if flour, sugar, etc are the cause of the atherosclerosis, and both young and old Masai eat these things, then it should appear in both the young and the old Masai...both before and after the dairy period. But in fact, their own data, in the chart below, shows it substantially increasing in frequency only after their long period of a high fat dairy based diet.

If those non-dairy foods were the cause of Masai aortic fibrosis, and those foods were eaten by both young and old, but not those in the 20-40 age group, then the frequency of fibrosis should be high both before the Muran period (ages 10-20) and after (ages 40+).  But in fact the frequency is substantially elevated only after the Muran period.

The data actually suggests four possibilities that occur to me: Either 1) aortic fibrosis is simply a fact of life for Masai after the age of 40, due to some aging factor, or 2)  it is the 20 years of eating a high dairy fat diet that eventually causes the atherosclerosis to develop, or 3) the late life combination of dairy fats and flour etc. promotes atherosclerosis, or 4) there is some dietary factor other than the flour, etc. that promotes atherosclerosis in the aging Masai.

Given what we know about the effect of saturated fats on the viscosity of blood and endothelial function, I feel inclined to consider their high dairy fat diet a contributor to their atherosclerosis.  The fact that the frequency of fibrosis in the young Masai (under 20 years of age) is practically the same as that of the Masai during the late Muran period (30-40 years of age) argues against non-dairy foods contributing to the process.  I feel, at this time, more inclined to view their atherosclerosis as the result of 20 years of their vascular systems having to respond to high fat meals.

Song et al found that "rats on a diet rich in either saturated or unsaturated fat had higher blood pressure compared with chow-fed rats (approximately 130 vs 100 mmHg, respectively), along with hyperlipidaemia and insulin resistance."  Repeat:  High fat diets induced high blood pressure in rats, regardless of whether the diet was high saturated or high unsaturated fat.

Only a rat study?

Let's compare two human primitive populations, one with a long history of a high unsaturated fat diet--Eskimos--and one with a low fat, high carbohydrate diet--Yanomamo.
Andersen et al found elevated blood pressure but without modernized risk of ischemic heart disease in non-Westernized Inuit past the age of 40:

"Among the 812 Inuit aged 18 years or above blood pressure was unaltered until the age of 39 years (systolic, p=76; diastolic, p=0.36) and increased subsequently (both, p=0.001). Systolic blood pressure ≥140mmHg was more frequent when aged >40years (p=0.001) and diastolic blood pressure ≥90mmHg was more common in men (p=0.001) and in men and women aged ≥40years (p=0.001)."

"Blood pressure rose only after the age of 40 years in pre-western Inuit. Left ventricular hypertrophy peaked among 30-year olds and was independent of elevated blood pressure. It may be speculated that the common left ventricular hypertrophy was due to marked physical activity that contributed to the low occurrence of ischemic heart disease among pre-western Inuit."
In contrast, Yanamamo natives of Brazil demonstrated no age-related increase in blood pressure.  Yanomamo live largely on plantains, sweet potatoes, manioc, and various fruits, supplemented by insects, grubs, and hunted meat.  The Inuit have a much higher intake of omega-3 fats, and a much higher total fat intake.  These findings support the idea that chronic high fat intake promotes age-related elevations in blood pressure, even if a large portion of the fat consists of omega-3 fatty acids, even absent Western foods, and even if your ancestors have been eating an Inuit diet for more than ten thousand years.  

As I have said before, we have no reason to believe that all primitive diets had the same health effects.  The differences in blood pressure between Yanamamo and Inuit clearly provide evidence that different primitive diets have different health effects.  By the way, the study of the Yanomamo was conducted in 1989 as part of the Intersalt project, while the study of the Inuit was conducted in 1962-1964.  Both of these populations had exposure to Western civilization, the Yanomamos for 25 more years than the Inuit.  The investigators of the Inuit blood pressure were satisfied that the Inuit they studied were not Westernized.  Yanomamo in fact have the high carbohydrate diet that introduction of sugar, flour, etc would induce.  I would find any attempt to blame the elevated blood pressure of Inuit on carbohydrates in the diet unconvincing. 

 Kjaergarrd et al also report that found electrocardiograms indicating coronary ischemia in 5.5% of non-westernized Inuit.  It is a low rate compared to modern populations, but indicates that isolated Inuit were not totally immune to ischemic coronary disease.

By the way, Masai appear to have a special adaptation to dietary cholesterol not present in Caucasians. According to Taylor and Ho, controlled studies on the Masai show that they "have a much larger capacity for intestinal cholesterol absorption than whites and a greater ability to suppress endogenous cholesterol synthesis, averaging 50.5%, for compensation of their intestinal absorption of dietary cholesterol. This efficient feedback control is the only homeostatic mechanism that protects the Masai from developing hypercholesteremia."  In other words, if you aren't of Masai descent, your high-dairy diet mileage may vary.


jacob said...

Great argument - but a false premise? What's the source of that chart?

My coconut oil is definitely liquid at temperatures a lot lower than 136- more like 80!

Steven Groh said...

Yeah, I think that if it is warm enough in your mouth to melt it, then it is warm enough in your blood stream to keep it molten. Unrefined coconut oil is liquid at 76 degrees, and definitely melts in one's mouth.

Justin said...

My answers: Simple physics predicts that increasing the amount of high-melting-point, high-viscosity fats in the mammalian blood stream will increase blood viscosity, impede flow of blood through the tiniest capillaries, impair delivery of oxygen and nutrients to and removal of wastes from the tissues (thus promoting ischemia and toxicity), and increase blood pressure.

I'm by no means an expert, but this is a ridiculous statement.

It's like saying that eating rock candy will send little grains of sugar flowing through our bloodstream, or that eating salad will increase the amount of chlorophyll in our blood, and then we'll all bleed like Vulcans, oh, god, the humanity!

My point being: the melting point of a fat (or anything for that matter) is the temperature at which molecules start bouncing around fast enough that they're no longer in a rigid, ordered structure with other molecules (which is what creates their solid form).

In other words, the melting point of a fat is irrelevant once it's broken down into individual molecules. It literally has no impact. Zero. None.

If little globules of fat were rushing through our veins, then melting point might be an issue. My understanding is that there aren't.

Mark Muller said...
This comment has been removed by the author.
Daniel said...


What about fat consumed from the body's fat reserves? Isn't this all saturated fat? I find it hard to believe that saturated fat is bad for us when it's also the body's main fasting fuel source. Are there any studies showing that fasting, a 100% saturated fat diet if you will, promotes atherosclerosis?

- Daniel

Stephan Guyenet said...

Hi Don,

The table you posted was for non-esterified fatty acids (NEFA). There's only a tiny amount of NEFA in the blood (micromolar range), not nearly enough to influence viscosity, and in any case they're complexed to albumin, not floating around freely. The vast majority of fat in the blood is in triglyceride form within lipoproteins, and in that context the saturation level will have no effect on blood viscosity.

Triglycerides within saturated dietary fats all have a lower melting temp than human body temp, all you have to do is put some in your mouth to figure that out...

If saturated fats impair vascular function and that has some effect on cardiovascular risk, why do people who eat more saturated fat not have more heart attacks than people who eat less?

Abe said...

Are there any differences in absorption based on this information? For example, do we absorb more of the unsaturated fats than the saturated fats, since the saturated fats apparently are little hard clumps of unmelted fat in our intestines? If so, and if what you say is true, wouldn't we then want to bias the fat intake in our diets towards saturated fats, if only in order to reduce our intake of fats?

Don said...


So you think that the physicochemical properties of fats disappear on the molecular scale? I don't know of any chemist who would agree. Degree of saturation and chain length are what produce the gross physical properties of fats. These affect their behavior at the both macro and microscopic levels.

If they didn't, then there would be no difference in fluidity between a cell membrane with or without omega-3 fatty acids in it. In fact, cells with high saturated fat and low unsaturated fat contents of cell membranes have very different properties. We see the difference in the flesh of fish (high unsaturated cell membrane content) and that of commercial beef (high saturated cell membrane contents. Which is stiffer when refrigerated? Which takes higher heat to soften?

If you saw a video of the viscosity of blood high in triglycerides, you wouldn't be making the statement you made. High levels of fat in the blood definitely make it more viscous, in part simply due to physicochemical properties.

"The results showed that rats fed high-fat diets exhibited significant increases in serum TAG levels (P < 0.01), plasma viscosity (P < 0.01), whole blood viscosity (P < 0.01) and internal viscosity (P < 0.01) compared to the controls."

In comparison, feeding omega-3: "Blood viscosity levels decreased. "

This is in part due simply to the physicochemical properties of omega-3s (very low melting point=very low viscosity) compared to other fats.

Moreover, the data I produced to follow up is consistent with the physicochemical prediction, even if my explanation was wrong. Trials show the effects that I predicted based on the physicochemical properties.


Coconut oil is a mix of fats, some with low melting points, others with higher (lauric acid).

Is coconut oil more viscous than water, or not?

Source of the chart: Bettelhiem, Brown, and March: Introduction to General, Organic, & Biochemistry. Page 472.

Justin said...

Don, I wasn't commenting on anything other than the statement I copy/pasted, because frankly, I don't know enough about blood viscosity and chemistry to put up any real debate.

That said, I do know just enough about physics to know that melting points are essentially irrelevant at a molecular level (that is, "melting point" comes into play when you start lumping molecules together). To try and use it as a measure of nutritional value is a stretch at best (unless you're just using melting point to indicate level of fat saturation, in which case, your logic in this post becomes a bit circular).

Don said...


If they don't affect viscosity, then how do you explain the findings of Tai et al?

Triglycerides certainly have lower melting points than body temperature. But are you saying that a saturated triglyceride, say, tallow (which, from my experience, does not melt very well in my mouth) has no higher melting point or viscosity than an unsaturated triglyceride, say, flax oil? That beef fat and flax oil have the same melting qualities and temperature? Not my experience. I don't see any reason to believe that saturated triglycerides should behave exactly like unsaturated triglycerides.

Don't French have more heart attacks than Japanese? Yes. IHD is 40% more common in France than in Japan. 2.5 times more common in Switzerland than Japan. In the U.S. population, variations in saturated fat consumption probably aren't great enough to detect the effect (e.g. the GE study), unless you compare vegetarians to omnivores. I think it is far from settled that saturated fat intake has no effect on risk.

So are you saying that those studies that showed increased viscosity, disturbed endothelial function, etc. were all made up? That those researchers didn't really find what they reported? They lied...for what motive?


The body transports dietary fats differently from endogenous fats. This would affect their impact on cardiovascular health. Also, when you burn body fat, you release it small amounts at a time, often to local tissues. You don't get a huge bolus in the blood all at once. This is evident from the fact that when fasting, serum triglycerides are low, but following a high fat meal, serum triglycerides are very high.

Theo said...


This argument sounds a lot like the argument that consumption of carbohydrate will lead to raised blood glucose (in the long term).
I do not think having elevated triglycerides in the blood is good, just as elevated glucose is not good. But I do not think that, in a healthy person, increased consumption of fats would increase triglycerides in the blood (except immediately after a meal) any more than increased carbohydrate consumption would increase levels of glucose.

Don said...


When fats are in the blood, they are stacked together, in triglycerides, which allow intermolecule bonding via London dispersion forces. A triglyceride containing saturates has very different properties from a triglyceride mostly unsaturated. I find it hard to believe that anyone would claim that primarily saturated triglyceride will have the same physicochemical properties in the blood as a primarily unsaturated triglyceride. This amounts to denying any differences between fatty acids.

el66k said...


Tuck said...

Don, this is not a compelling argument.

You cite a couple of studies, and yet the guy who's spent the last 30+ years studying saturated fat and heart disease has thrown in the towel and declared:

"...there no significant evidence for concluding that dietary saturated fat is associated with an increased risk of CHD or CVD."

That link is to Stephan's blog. How do you reconcile that statement and the evidence behind it with your belief?

SamAbroad said...


Don't Switzerland and Japan have very similar life expectancy? (82 vs 82.9) So if you don't die of heart disease in Japan you die of something else?

Also how do you explain the study that within the Japanese more saturated fat results in less CHD risk?

You need to be wary of extrapolating short term feeding studies to long term life. As you have said, initial beneficial adaptations are exhausted over time.

FloridaSusan said...

OMG, my apples don't melt at room temperature either.

Erik said...

Ever make simple syrup? It's pretty damn viscous, too.

Dave said...


This seems so ridiculous on the face of it, that I wonder if you don't have some other purpose here, such as playing devil's advocate. But on the off chance you're serious, I'll play along.

People who argue on the basis of "simple physics" often have no clue about the physics they use as support for their argument. Note I said "often", not "always". Those who do have a clue can back up their arguments with simple calculations. Physics is a largely quantitative science, and is used to make quantitative predictions. So I would ask that you provide some fluid dynamics calculations to support your position, otherwise it's all angels dancing on pinheads. One can make any argument based on "simple physics" if not constrained by those annoying equations.

Good luck with that. I taught graduate fluid dynamics, and doubt that even the author of our textbook would consider flow of a multi-component fluid through non-rigid tubes of varying sizes driven by a variable pressure to be anywhere near the realm of "simple". Computer simulations are acceptable.

gallier2 said...

The melting point of triglyceride is irrelevant in the blood. The quantity may play a role, but not the viscosity. Why, because they are not free, they are emulsified and captured in lipo-protein balls (LpA, LDL, HDL, VLDL, chylomicron). The size of these balls has an impact on viscosity but not their content, they are isolated.
It's like rubber balls in water, the quantity and the size of the balls plays a role, but not their content.

On the other hand, glucose has a direct, straight forward impact on viscosity.

Don, on your actual crusade against sat fats, this was really a lame argument.

There is also one point I would like to bring to the table in the argument for fats, you forgot fasting. In a fasted state, the overwhelming part of needed calories comes from fat. It would be highly (biologically) illogical that a diet having nearly the same metabolic impact than a fast, to be much more detrimental.

Chris Masterjohn said...


I agree with what Stephan said and would like to add a few points.

First, coconut oil contains a mix of different fatty acids. It is not a brick of lauric acid. Coconut oil's melting point is in the low 70s F.

Second, when you eat fat, fatty acids with chain lengths similar to lauric acid or shorter, as well as some highly unsaturated fatty acids, are preferentially delivered to the liver via the portal vein. Because of this and because medium-chain and short-chain fats do not require the carnitine shuttle to enter the mitochondria, and thus their oxidation is not suppressed by carbohydrate and insulin, these fats are preferentially oxidized at a rate higher than usual and the bulk of them may never see the bloodstream.

Third, you assume that viscosity and endothelial function are primarily determined by physics, when, in fact, they are primarily determined by physiology. Viscosity is going to be determined by many factors including basal platelet activation, endothelial nitric oxide production, nervous system activity, and so on.


Chris Masterjohn said...

[...continued (part II)]

As Stephan pointed out, most of the TG circulating in plasma don't have direct access to the plasma because they are shielded by the phospholipid membrane of the lipoprotein. But beyond this, the effects are determined by physiology and not "physics," except insofar as physical principles are harnessed by regulatory physiological systems.

Fourth, saturated fats have been tested against polyunsaturated fats in clinical trials, including one going over eight years and double blind, and these trials miserably fail to support the attacks against saturated fat.

Fifth, your analysis of native diets and native health involves many confounders, and you acknowledge some but not all of them.


Chris Masterjohn said...

[...continued (part III and final)]

Mann showed that serum cholesterol and atherosclerosis was *lower* in "warrior" aged Masai than in young Masai, but you only look at fibrosis. Mann observed that the lumen increased with age rather than decreasing because of age-related dilation, which you do not discuss. Mann discussed the fibrous coverings of the plaque as a positive development, which they are, and, as Mann pointed out, they were responsible for the complete lack of unstable plaques in the Masai. Diet is confounded by age, exercise, and so on. In the case of the Inuit you have not only diets very low in plant foods, but very high in highly oxidizable marine fats. You have climate, latitude, partial modernization. Again, too many confounders.

Sixth and finally, a chylomicron's half life is 10-20 minutes. The TGs are quickly removed via lipoprotein lipase and the remnants are quickly taken up by the liver. The loss of TG from plasma begings occurring instantaneously.

A VLDL particle's half life is closer to 4 hours. It is degraded downward to IDL and LDL and taken up again. Then the VLDLs are resecreted by the liver all over again. The plasma sees much more VLDL/IDL/LDL than chylomicrons regardless of whether the fats are from diet or carbohydrate.

I hope that makes some sense.

Take care,

Eric said...

Thank you Chris... And Stephan...

Don said...


I updated my post. I think it is very clear that nature has selected across species for use of unsaturated fats to increase fluidity and compliance, and saturated fats to increase solidity and non-compliance. I stand by my analysis.

I am fully aware that natural fats consist of complex triglycerides, etc. If you think it doesn't matter physiologically what type of fat occurs in triglycerides, fine. If you don't think viscosity or melting point plays a role in biology, fine. Nature disagrees, it is very clear from the natural distribution of these fats.

The funny thing to me is, I provided the data showing that fats affect blood viscosity and blood pressure, but you all seem to ignore the data because you don't like it. Good luck. I prefer to side with Nature and Michael Crawford, author of Nutrition and Evolution, who has done more bench biochemical research on the effects of fatty acids on cardiovascular and neural physiology than all of you put together.

Jason said...

Wow. This is an irresponsible post. I thought you'd like to know that you've lost one reader.

Don said...

Basically, those who disagree with my analysis are saying that melting temperatures of fats have no biological relevance, and not even facing the evidence (i.e. research results) I presented that the amount and type of fat in the diet does influence blood viscosity, blood pressure, etc.

I love how when I show evidence that Eskimos have high blood pressure, people start screaming about confounders, etc. It seems you just can't accept that different primitive diets have different health effects, that not all primitive diets are healthful. Why is that? It sure looks like you go out of your way to deny the validity of data that doesn't confirm your preconceptions. If it doesn't confirm your preconceptions, it means it is "confounded" by various factors. But if it does confirm your preconceptions....confirmation bias abounds.

Those saturated fats have to be safe in any quantity, or not safe at sense of the basic principle of toxicology there: The dose makes the poison.

Don said...


Good bye. I don't see anything irresponsible in my post at all. I presented good data. Good luck.


I am not on a "crusade" against saturated fats. I am looking for a reasonable perspective on saturated fats and total fat content of the diet. I am presenting data that appears ignored by most paleo bloggers.

Look, as a health care provider, I am responsible for the health of my clients. I have seen people applying the dictum that saturated fats are 'healthy,' as a result of my uncritically passing around the current 'primal' infatuation with saturated fats.

Results in several cases: weight gain (in spite of low carb intake), endocrine disturbances, increased blood pressure, increased blood cholesterol, lethargy.

You are free to believe what ever you want when you aren't responsible for guiding others to health.

And Jason's remark shows, I guess I need to get ready to be excommunicated from the high fat religion. So be it.

SamAbroad said...

Ehh, Don you are doing the exact same thing you are accusing us of.

You are deciding that the longevity of the Japanese is due in part to the low saturated fat and that the equal longevity of the Swiss is irrelavant because they die more at the same age of something different.

Basically the crux of this post comes down to one short term feeding study. I'm open to having my mind changed but you're gonna need better evidence than this.

Don said...

BTW, I stated that physics would predict an effect...I didn't say that physics/fluid dynamics alone would produce the effect.

The prediction is that fat quantity and quality will influence hemodynamics.

Whether or not physics/fluid dynamics is solely responsible for the effects observed in the multiple studies I cited, the results of those studies were consistent with the prediction...showing that the physical properties of the fats may predict their physiological effects.

That's all. Now if you disagree, you need to show that the studies I cited did NOT really show that the amount and type of fat in the diet affects viscosity, blood pressure, etc.


Re Eskimos, I showed that they get high blood pressure on a high fat diet. Your response is that the Inuit diet is not only low in plants, it also has a lot of oxidizable marine fats, and the Inuit live in a cold climate, latitude, partial modernization.

I am unaware of any evidence that climate or latitude affect blood pressure. I already addressed the partial modernization...namely, the researchers were satisfied that they were not modernized, but even if they were, it would mean eating less marine fat, less animal protein, and more plant foods.

But the Yanomamo eat even more plant foods, and have at least as much contact with Westernized habits, but have low blood pressure, so the idea that plant foods caused the Eskimos to have hypertension, suddenly, after the age of 39, just doesn't hold water.

If oxidized marine fats are the cause, then Westernization would reduce the intake of these, which according to you, could reduce blood pressure (because apparently you think oxidized marine fats could explain the sudden onset of hypertension in aging Eskimos)...but that didn't happen either (ie. blood pressure went up, not down).

I just don't get why people want to defend the Masai diet as infallible. Even Weston Price considered it suboptimal, rating the physiques of fish- and plant- eating tribes as better than the Masai. Why should a diet composed of milk of bovines be optimal for any human being? The protein, fat, and carbohydrate in bovine milk is all 'designed' by nature to produce rapid body growth in calves, a small-brained species. Humans grow slowly and have large brains; and in adulthood, aren't growing at all. I don't see any reason to expect that a food specifically produced to nurture the growth of calves into small-brained cattle should be ideal for a full grown, large brained human. The body might get large but the brain might not get all it needs.

Don said...


Now, you have a point: Swiss live almost as long, on average, as Japanese.

Switzerland also has a fairly low rate of obesity, 7.7%, far better than the U.S. and only about twice that of Japan.

Still, double that of Japan.

This is a data point I needed.

I would disagree that this boils down to a short term study. The Eskimo data is not a short term study, nor is the Yanomamo.

Kamal said...


Question. The strongest conclusion of the studies you posted (temporarily ignoring all the cohort data because of obvious interpretability problems) might be that high fat diets increase blood viscosity.

However, the "high fat" diet fed to these rats consisted of soy oil as well as SFA and Trans fat. How is that applicable to high-fat paleo diets?

Second, do you happen to have studies on blood viscosity and health outcomes, controlling for other factors? (I'm not saying there aren't any, I'm just curious as to what they sayd--e.g. if it's just a blood pressure thing or if there are other impacts)

Don said...


The one study used palm oil, which has a fatty acid profile very similar to lard. So, basically it showed that a diet with 40% fat could increase blood viscosity regardless of fat type.

Besides blood pressure, this could affect inflammation and aging. Simply, increased viscosity reduces flow rate, which reduces delivery of oxygen and nutrients to the tissues, and removal of wastes from the tissues. Tissues experiencing this stress would tend toward inflammation.

I haven't yet found studies linking viscosity to other outcomes. This blog records a discovery process. But, I will say that impaired circulation is a common feature of the aging process. Anything that retards circulation would promote, rather than retard, the aging process, by favoring impaired cellular nutrition and metabolism.

dynacore said...

In the studies you cited, were the rats and the humans being tested fasting or post prandial? Seems to me that in the fasting state, the amount and type of fat you ate yesterday should have very little effect on blood viscocity or pressure.

Stephan Guyenet said...

Hi Don,

Come on now... fat is transported through the bloodstream in liporoteins; it's not fatty acids and triglycerides floating around in serum. In other words, it's packaged up in particles with lipid/cholesterol cores and a hydrophilic phospholipid outer layer. The saturation level of the fats inside the particle have no effect on blood viscosity because they aren't even touching the blood... they're on the inside of the particle.

If SFA were influencing endothelial function, blood viscosity or something else related to heart attack risk, then you'd expect it to cause heart attacks... yet the evidence suggests it doesn't.

Mirrorball said...

I'm biased towards more carbohydrate rather than more fat, but just for the sake of being thorough, the Japanese become overweight at a lower BMI and waist circumference than Western peoples. So it's important to state how obesity was measured when comparing different groups, because BMI > 30 doesn't work well for the Japanese. I know that the World Health Organisation was considering lowering the overweight threshold from 25 to 23 for the Japanese. I personally know older normal-weight Japanese women who have all the symptoms of the metabolic syndrome.

dynacore said...

In the studies you cited, were the rats and the humans being tested fasting or post prandial? Seems to me that in the fasting state, the amount and type of fat you ate yesterday should have very little effect on blood viscocity or pressure.

dynacore said...

In the studies you cited, were the rats and the humans being tested fasting or post prandial? Seems to me that in the fasting state, the amount and type of fat you ate yesterday should have very little effect on blood viscocity or pressure.

Grow Up said...

If high saturated fat consumption increases blood viscosity, surely that would slow bleeding?

But Hugh Sinclair found the exact opposite effect in his experiment of living on seal blubber for 100 days, cutting himself every day, and measuring the length of time it took for the cut to stop bleeding. He began with a time of 2 minutes, and wound up with 45 minutes…


for a report of a report—the original paper is not available online that I have found. This BBC report from 2005 still suffers from the "lipid hypothesis", which we know is wonky.

Don said...

Grow up,

Seal blubber is a marine fat, rich in anticoagulant omega-3 fatty acids.

Don said...


I agree with you, but if the quantity and quality of dietary fats don't affect blood viscosity or other hematological qualities, then what are you saying about the studies I presented? Since they appear to show effects of quantity and quality of fats on viscosity and blood pressure?

Is it not possible that what we think we know about transport of fats in the blood is not the whole story? Or do you consider physiology a completed science, nothing more to learn?

As I said, I was not claiming that the physicochemical properties were solely or even partly responsible for the effects, I only said that one might predict from the physicochemical properties certain effects. Then I cited research that seems to support the prediction regardless of actual mechanism.

Heart attacks are typically proximally caused by blood clots and/or unstable plaque broken loose and getting caught in arteries, plugging the vessels. The question is, what causes the degenerative process that scars the arteries over time, and what promotes the coagulation or hypercoagulability. No one is claiming eating saturated fat once or having a big milk shake once will cause an ischemic attack. What people are trying to understand is what is the chronic process that produces the late life conditions that underlie ischemic attacks. We have a pretty good guess that the underlying process involves inflammation of arteries. The question is, are there effects of high fat diets, particularly diets high in saturated fats, that may contribute to this inflammatory and hypercoagulability process, either directly (e.g. by causing inflammation of artery walls) or indirectly (e.g. by increasing blood pressure, which, over time, increases mechanical wear and tear on arteries).

Don said...

We all know that ischemic heart disease associates with aging, so it is most likely a result of chronic cardiovascular insults. There are many neolithic suspects, all of which I acknowledge. But I also consider dietary saturated fats suspicious because I see many reasons to believe that consistently lifetime high saturated fat intake is a neolithic event also, certainly especially if consumed as dairy fat.

For example, Michael Crawford considers saturated fats hazardous for several reasons, not the least of which is that if saturated fats replace unsaturated in vascular cell membranes over time, the arteries and capillaries become less compliant. He suggests that saturates compete with and interefere with the utilization of essential unsaturates...He also cites their effects on blood pressure and stimulating blood coagulation.

He comes at this entirely from an evolutionary perspective as well i.e. he considers dietary saturates suspect simply because foods available to humans during evolution could not have provided high saturated fat intake.

Chris Masterjohn said...

Hi Don,
Thanks for your response.
I realize you presented some in vivo studies, but my comments were meant to demonstrate that your explanation does not make physiological sense. I agree that you merely said you could “predict” the effects on viscosity using physical data, but someone with a knowledge of physiology would not make this prediction precisely because it doesn’t make any physiological sense for the reasons I stated. Moreover, the fact that you lead with this, and then continually invoke the melting fat, indicates that this is the explanation you favor – again, an explanation that doesn’t make any physiological sense.

I agree that Price did comment on the physique of the Masai, but this was a passing comment that he gave no further weight to. He states clearly at the beginning of NAPD that his purpose was to control for various factors such as latitude, altitude, temperature, and race by making comparisons of isolated and modernized peoples *of the same genetic stock* in many different environmental niches and areas of the world. Virtually the entirety of the book deals with comparisons meant to control for hereditary factors such as genetics.

Your analysis of the Eskimo transition is internally inconsistent. You state that EPA and DHA decrease viscosity in humans because of their high melting point. Then you blame the age-related increase in blood pressure among the Eskimo on the high fat content of their diet, even though their fat is primarily marine oils rich in EPA and DHA. Then when someone points out that one of their prized fats, seal oil, increases bleeding time, you acknowledge that this is because it is rich in the anticoagulants EPA and DHA. Well, I agree – but not because of the melting point, but because of the well known physiological roles of these fatty acids. You then point out that with modernization their blood pressure went even higher, despite the fact that they ate less traditional fat and more refined carbohydrate, which you intend to refute my comments but which clearly refutes your position that their fat intake is responsible for their age-related increase in blood pressure.

I merely pointed out numerous confounding variables present in the comparison between the Eskimo and Yanamomo. I didn’t aim to isolate “THE” cause of the Eskimo’s high blood pressure and claim it was the oxidizability of their dietary lipids. Their diet is clearly not *just* high in fat. In any case, the two groups have different ancestry, and therefore the comparison should be given little if any weight at all.

I never defended the Masai diet as infallible.


Richard Nikoley said...


I'm wondering if you've read Stephan's series on the Tokelauans and how you might interpret that.

Lucas Tafur said...

So do you also recommend drinking warm water with your meals to "melt" the fat and prevent it from "solidifying" and clogging your arteries?

Don said...


I didn't state that EPA and DHA are anticoagulant BECAUSE they have a low melting point. You will not find anywhere any such statement. I know that it is because of their effects on prostaglandin synthesis. I merely pointed out that these fats that reduce coagulability ALSO have a low melting point. All I said, and meant to say, is that the physical properties of these fats appear to PREDICT their physiological effects.

Sometimes it pays to disregard received knowledge and simply view things afresh. So, when writing this, I didn't think the whole thing through considering 'known' mechanisms. I should have written, "IF the body was just a physicochemical mechanism, we would predict that SFAs would do X, etc. etc. Now, that isn't how the body works, but, do we have any evidence that in the body, the natural melting point of fats predicts whether or not they play a role in fluidity or solidity/coagulability or the blood?"

I repeatedly invoked the melting point only to draw attention to the connection, i.e. that fats with a high melting point appear involved in physiological processes that promote increased viscosity (i.e. solidity) of the blood, and fats with low melting points appear involved in physiological processes that reduce viscosity (i.e. increase fluidity) of the blood.

But the profit of discarding/forgetting physiological knowledge for a moment and making a 'mistake' with a beginner's mind was a new perspective.

Namely, the melting points of fats, or their viscosity at body temperature, may indeed predict their effects on blood viscosity, for example, through physiological processes.

Perhaps you do not, but I certainly find it interesting that fats with a very low melting point (omega-3s) also reduce COAGULATION which is a species of SOLIDIFICATION. In other words, through a physiological process (prostaglandin synthesis), they 'melt' the blood, keep it fluid, keep it from freezing solid.

Why didn't nature put saturated fats to that task? I only suggest that saturated fats can't do this task simply because they are designed by nature to create solidity, firmness, noncompliance; so it would not suprise me at all if they facilitate physiological processes that, for example, promote blood solidification, i.e. clotting...and indeed, some evidence does suggest this happens. On the other hand I would be extremely surprised if anyone found evidence that unsaturated fats promote processes that increase viscosity, and saturated fats the opposite. It would be completely contrary to the nature-wide functions of these fats.

So you may say that someone with knowledge of physiology would not make this prediction. On the contrary, you just saw someone with a knowledge of physiology just make it.

Don said...


You completely twisted what I said about Eskimos, so much so that I won't even try to untangle it. I didn't anywhere say that modernization of the Eskimo diet made their blood pressure go even higher. I didn't talk about modernized Eskimos at all. You were the one who implied that the high blood pressure was due to modernization. I was saying that it was NOT due to modernization, and why.

So we should not compare any diets of any human groups to any others and compare health effects? Because each group is a separate species?

Emily Deans, M.D. said...

Don, I am confused as to your description of increased SFA content in cell membranes being implicated in the pathogenesis of hypertension. Hypertension is a long term process of stiffening of the elastin (most likely through cross linking of cysteine via inflammatory processes) and hypertrophy leading to increased intima-media thickness that is measurable - a smooth muscle phenomena, not a cell membrane phenomena. Certainly fatty acid content of TGs can affect membrane fluidity and cell signaling - lack of omega3s being more important here than excess SFAs as far as I can tell.

Emily Deans, M.D. said...

(TGs in lipoproteins, I should say!)

Chris Masterjohn said...


I appreciate your attempt at objectivity on this issue, and that you do not accept every darling of the "paleosphere."

I also appreciate you explaining that you led off and repeatedly invoked something that was completely wrong not because you thought it was right but because you thought it would offer a fresh perspective. This is not my approach and I won't claim to empathize much with it, but I'm glad you explained it.

You list arachidonic acid, which is involve in initiating platelet aggregation, as having a melting point of -56. I don't like to cite Wikipedia, but I also don't have much time to devote to this and it is the first thing that pops up on Googling -- it lists the melting point of DHA as -47 and its data for arachidonic acid agree with yours. It doesn't list one for EPA, but Sigma lists it as -63 or -64 F.

These values are fairly similar, but the physiological roles are very different. As a general summary, AA can initiate platelet aggregation and other inflammatory processes, while products of both AA and DHA can resolve these processes, and EPA will inhibit them. By comparison, other fatty acids do not have direct roles in the process at all, but most of them have radically higher melting points.

I think you have to draw from the data very selectively to make a correlation between melting point and physiologic effect on blood viscosity.

I agree with you that these fatty acids can have effects on viscosity of the things they are dissolved in, but as Stephan repeatedly pointed out, and as I did also, these are not "dissolved" in the blood, nor are the triglycerides they are made of, and as I am trying to emphasize blood viscosity is controlled by physiological processes, of which prostaglandin synthesis is but one.


Chris Masterjohn said...

Don, this is what you said:

"If oxidized marine fats are the cause, then Westernization would reduce the intake of these, which according to you, could reduce blood pressure (because apparently you think oxidized marine fats could explain the sudden onset of hypertension in aging Eskimos)...but that didn't happen either (ie. blood pressure went up, not down)."

I interpret the last sentence as indicating that with Westernization blood pressure went up. If not, I'm not sure what it means.

You can, of course, compare different populations with different diets if you want, but such is an "ecological" comparison that is widely regarded as the most confounded of all epidemiological observations. I don't mean we should not make the observations. I said we should give them very little weight when considering cause-and-effect relationships.


Tuck said...

"Why didn't nature put saturated fats to that task? I only suggest that saturated fats can't do this task simply because they are designed by nature to create solidity, firmness, noncompliance..."

You appear to be confusing the role of fats in cold blooded creatures like fish with the effect in man. Fish in colder latititudes do indeed tend to have higher concentrations of PUFAs. They also lack the ability to thermoregulate, which allows humans to keep saturated fats as liquids.

I'll note that if your analysis is correct, then human mother's milk should be a solid, since you class it with butter and coconut.

Ryan J said...

Hi Don,

First, one key point that seems to be missed by so many is that the strict 10% fat diet has extremely significantly reversed atherosclerosis(not to mention all kinds of other related problems, diabetes,etc.) in very, very sick people. The Esselstyn 15 year trial proved this.

Then the fact that a huge percentage of the worldwide population eats starch-based seems quite clear what is closest to optimal for most, but hey.

Anyways, I'm curious on your take on linoleic acid. There are quite a number of studies looking at blood phospholipid concentrations of the fatty acid, and many show beneficial outcomes regarding CHD,diabetes, etc.

There have been a few walnut studies showing improved post meal FMD versus MUFA/SFA,but perhaps this was simply a result of the walnut ALA, yet, still, walnuts have such a huge percentage of N6.. Also, there was the sunflower versus coconut oil study a while back showing significant improvement in HDL capacity to improve inflammatory status of endothelial cells(in vitro, but still interesting), while HDL from the coconut was highly inflammatory.

There's also a few long term(~5 year) studies, by Rudel, where SFA vs MUFA vs PUFA enriched diets were fed to 2 different species of Monkey, that were fed cholesterol to greatly increase their serum cholesterol, and the PUFA fed groups at the end of 5 years had significantly less athero. that the other monkeys.

It seems to me that linoleic acid has, surprisingly, a fair amount of research to show that it's actually quite beneficial, but the only point that puts me off is that basically none of the high starch, seemingly very healthy populations, look towards nuts for a significant amount of energy.

Would love to hear your thoughts.

el66k said...
This comment has been removed by the author.
Adam Haritan said...


Thanks for the exciting post. Regardless of who is "right," I think you have sparked a fascinating discussion on the topic of fats and the human body.

Could you comment on this? ...

You say that nerve membranes have a high proportion of highly unsaturated fats. The myelin sheath is primarily made up of oleic acid. However, in this study ( researchers demonstrate that dietary oleic acid intake does not affect the oleic acid level in the brain. What's more, they suggest that "the brain, like the peripheral nervous system, synthesizes oleic acid from stearic acid."

Would that suggest then that although unsaturated fats may be necessary for brain health, dietary saturated fat intake, namely from stearic acid, would be just as crucial?

Theo said...


You wrote "This line of reasoning has a major flaw: Simply, if flour, sugar, etc are the cause of the atherosclerosis, and both young and old Masai eat these things, then it should appear in both the young and the old Masai...both before and after the dairy period. But in fact, their own data, in the chart below, shows it substantially increasing in frequency only after their long period of a high fat dairy based diet."

I am not aware of any culture where atherosclerosis is commonly seen in 10-20 year olds. My understanding is that it is a long process that takes many years to develop. Therefore, you assertation that if it was the refined foods that made them develop fibrosis then the children would have it too doesn't make any sense to me.

Don said...


That's facetious.


Sorry for the convolution in this:

"If oxidized marine fats are the cause, then Westernization would reduce the intake of these, which according to you, could reduce blood pressure (because apparently you think oxidized marine fats could explain the sudden onset of hypertension in aging Eskimos)...but that didn't happen either (ie. blood pressure went up, not down)."

To untangle it.

1) You suggested that the Eskimos with HTN were partially westernized.

2) I disagreed, but chose to grant you the possibility.

3) If I grant you that they were Westernized in part, then this would mean that they ate less oxidized marine fats (having Western carbohydrates as replacements).

4) you suggested that oxidized marine fats would be among the possible causes for HTN among Eskimos. I was attempting to point out that if they had Westernized, this would mean that they ate less marine fats, and if oxidized marine fats cause hypertension, then the replacement of marine fats with carbohydrates would help prevent hypertension. But the fact is, these Eskimos had HTN, which argues against reduction of marine fats as a protection against HTN.

I hope that clarifies that question.

Eskimos get hypertension, but Yanomamo don't. This I take as evidence that something the Eskimos do differently from Yanomamo promotes hypertension. We have numerous studies supporting the idea that dietary fat quantity and quality influences blood pressure, so I start there. The real test would be, take some of those hypertensive Eskimos and put them on a Yanomamo diet, and see if their blood pressure reaches Yanomamo levels. I would be inclined to do the test, because I consider Eskimos and Yanomamo both humans with similar physiology, but some would not even consider it because they think Eskimos are adapted to their diet and not to the Yanomamo diet. I personally consider the presence of hypertension among nonWesternized Eskimos a sign that they are not adapted to their diet.

Yes, the Eskimo diet contains a large proportion of omega-3 fats, which, in the right amount, appear to help reduce blood pressure by known and probably unknown mechanisms (we never know what we don't know).

But some nutrients have one effect in an appropriate dose and the opposite effect when overdosed. Vitamin A provides a handy example. In the right amount it is essential for building bones in concert with D and K. But overdosed relative to D and K, it stimulates bone mineral resorption and probably contributes to osteoporosis. So if high doses of omega-3s lose their positive effects and have opposite negative effects, this would not surprise me, in fact it would confirm a Chinese medical principle that extreme excess of anything tends to induce the opposite (like hypothermia inducing a fever, or fever inducing chills), but not always in a healthy way.

Sue said...

"I am unaware of any evidence that climate and latitude affect blood pressure"

Maybe this study:

Don said...


No, I have not read that series.


Everything you state is conventional wisdom. I don't think I said anything specifically about hypertension, but I did say that Crawford appears to me to imply that changes in fatty acid ratios in vascular smooth muscle membranes to lower P:S will increase arterial stiffness, leading to less compliance, which could play a role in vascular diseases in general.


No, my analysis does not imply that mother's milk should be a solid. Mother's milk contains a mix of fats including short chain saturates, polyunsaturates, and superunsaturates, all with low melting points, that keeps the included saturates from forming bonds that would produce solids, so keeping it fluid at body temperatures and lower. Same with cow's milk.


I appreciate the reminder about Arachidonic acid promoting platelet aggregation. AA clearly contributes to membrane fluidity. Here we might have a case of either a neutral effect of AA, or a situation where, in low concentrations, it promotes fluidity and in high concentrations, the opposite. I don't want to oversimplify, but this principle seems in operation throughout physiology, which leads to a lot of confusion among scientists and lay people both, so that for example AA is often called proinflammatory, although it also appears to have anti-inflammatory effects. I only suggest that the dose determines the effect, and that the more potent the item is, the less required to create the overdose that has the opposite effect.

Chris Masterjohn said...

Hi Don,

The paper you cited is a reanalysis of data taken from others decades ago and they say, with no reference, that "limited westernization" had occurred, with no data on dietary intake. Thus, I do not know how much or what kind of westernization took place, but their phrasing seems to allow for some.

In any case, I'll try not to let us get too side-tracked. I was merely listing a number of probable confounders when trying to look at the health of one population and attribute a particular health characteristic to a particular dietary characteristic, or when comparing two genetically dissimilar populations living in two radically different climates in different parts of the world eating different diets.

I fully agree with you that the test of your hypothesis would be to see whether the Yanamomo diet improves age-related blood pressure increases in Greenland Inuit. I fully agree with you that this should not be dismissed based on the dogmatic assumption that the Inuit are fully adapted to their traditional diet.

I also agree entirely that one can overdose on fish oil.

However, I seriously doubt this is because of a direct effect of these fatty acids on the viscosity of the blood. Likewise, if the Inuit diet promotes age-related increases in blood pressure, I seriously doubt it is mediated by a direct effect of the fat in their diet on the viscosity of their blood.

If you want to make a dietary inference from data like this, the best way would be to find several populations with similar diets but different genetics, and several popualtions with similar genetics but different diets, and then make the comparison, so that one can get a sense of how much variation is due to genetics and how much to diet. All the better if you can repeat this at different altitudes, latitudes, and climates, which is what Weston Price did.


Don said...


The body can manufacture stearic acid from surplus carbohydrate, so even though it is essential for nerve myelinization, it is not essential as a dietary item.


Autopsies of American soldiers killed in the Korean war showed significant atherosclerosis in 18-20 year olds. Don't have time to put my finger on the report right now, but I am sure that a little search can turn it up.

Theo said...


I found the study you mentioned ( According to this, 9% of soldiers had narrowed as much as 75%. Mean age was 26. If such an extreme degree of atherosclerosis is common in people that age, why do the not display symptoms until many, many years later? I imagine that these men displayed an unusual amount of atherosclerosis due to the wartime conditions, i.e. loads of stress and overwork of the cardiovascular system.

Don said...


I know, very low fat diets have shown therapeutic value, in Caucasian patients and other ethnic groups as well. I myself used very low fat diets in my own health care in the past, with some significant improvements in health.

I believe that we have to take this into account to understand human evolutionary diets. That doesn't mean that a 10% fat diet is optimum for the long haul, but its therapeutic value definitely tells us something about how the body responds to dietary fats, and their severe limitation, even if we don't understand all the mechanisms involved at this point.

Looking at the composition of the African wild game meat that probably gives us a good proxy for the meat humans ate during evolution, suggests that we evolved on fats containing a fairly high percentage of linoleic acid. I too have noticed that contrary to the LA bashing I see so often, many clinical studies seem to indicate positive effects of LA, which seems to jibe with the proposed higher LA content of evolutionary diets. Here again we probably have a case of an appropriate dose of LA (similar to evolutionary diets) being beneficial, and an excess, particularly if from refined sources, having opposite effects. !Kung eat a lot of mongongo at times getting 50% of calories from those nuts, which have high LA contents, and they have low incidence of vascular diseases. So I think LA in natural packages (whole foods, like nuts and wild game or grass-fed meat), probably has positive effects if the diet has no refined seed oils.

Don said...


I have no doubt that stress played a role, since we see very high rates of vascular disease in war-torn nations. However, I seem to recall that a similar study was done on Asian soldiers without finding this prevalence of atherosclerosis. Again I don't have the study on hand and hopefully I am remembering correctly. If I am, this suggested that Western diet also played an important role, unless you think that Americans of various races are just genetically prone to atherosclerosis.

Not sure why you conclude that the individuals did not have symptoms. Perhaps they had angina or other complaints. This study did not address that so I don't know, just saying that it seems possible that they did have symptoms.

Don said...


When I suggested that excess omega-3s might have opposite effects in overdose, I didn't mean to imply that the effect would necesarily be a result of blood viscosity changes. Perhaps their creation of some opposite effects (e.g. HTN) in overdose occurs through another mechanism, known or unknown. The principle I am suggesting does not specify the mechanism for overdose effects will be the same as for proper dose effects, only that overdose will likely produce opposite effects.

Theo said...


Allow me to clarify. I was not saying that the soldiers did not have symptoms. I am saying that, as far as I know, few 25 year old males in the United States show signs of severe atherosclerosis.

There is probably a double effect of the stress and diet that produced that severity which would not be found in nearly the same degree with only one of those factors.

Brandon Berg said...

It may be worth noting that in the Tai study, the palm oil diet was deficient in n3 fat, containing about 0.07% of calories from 18:3 and 4.8% of calories from 18:2. The control diet had about 0.6% of calories from 18:3 and 5.3% of calories from 18:2. The control diet had nearly ten times as much 18:3 and a much lower n6:n3 ratio (9:1 versus 70:1). Might this explain the results?

Also, hi, Chris.

Kamal said...

Okay, back to the viscosity bit. I took a brief look at the paper, and the results do not seem at all straight forward. Trans fat sucked the hardest in all parameters. But saturated fat did better soy oil in plasma (whole blood showed different results at different shear levels):

"...the SFA group showed a significantly lower plasma viscosity
compared with the TFA and USFA groups."

Furthermore, the discussion section mentions that the viscosity issue is complicated, and human trials show different correlations. A quick pubmed search yielded an old study showing no correlation of nutrients (as assessed by 7-day records) and viscosity:

What I'm trying to say is that the results from this study, feeding rats a very controlled diet, may actually not fully mirror results in humans. At the very least, inclusion of more papers on fat intake and viscosity would be wise, especially in humans with mixed diets. Otherwise, conclusions and extrapolations would be premature.

Emily Deans, M.D. said...

Don, I think perhaps you are confusing "conventional wisdom" with "basic knowledge of physiology.". You make a few interesting points, but frankly they are so mixed up with baffling and, incorrect, and implausible biologic explanations that I am simply floored.

Jay said...

Different game meats have different fatty acid profiles:
"Comparison of some physical and chemical properties of selected game meats" C. A. Onyango, M. Izumimoto and P. M. Kutima in Meat Science
Volume 49, Issue 1, May 1998, Pages 117-125

Quotes from the abstract:
"Game meats were similar in composition to beef, though beef had lower crude fat contents than expected....Kongoni, oryx and beef lipids consisted mainly of saturated fatty acids while those in zebra lipids were predominantly unsaturated."

and from
"Chemical characteristics of blesbok (Damaliscus dorcas phillipsi) meat"Louwrens C. Hoffman,Karen Smit and Nina Muller
in Journal of Food Composition and Analysis, Volume 21, Issue 4, June 2008, Pages 315-319
quote from the abstract:
"The saturated fatty acids palmitic acid (16.36%) and stearic acid (26.08%) were found to be the main fatty acids in blesbok meat. Blesbok meat has a ratio of polyunsaturated to saturated fatty acids of 0.92."

Chris Masterjohn said...

To Brandon, Don, and others,

First, Brandon, hi! Long time no talk. As you point out, the n-3/n-6 balance is out of whack in the high-fat diet. This is virtually always the case, and is one of the numerous reasons why it is preposterous to blame the effects of a "high-fat" rodent diet on "fat" per se.

Don, thank you for clarifying. It seems based on our comments in the discussion that we are largely in agreement because you are no longer attributing the effects of fatty acids on blood viscosity to their direct physical properties but rather to their physiological properties as metabolic substrates.

That said, I don't think your post in its current form really reflects this line of thought. Your revisions made the introduction more accurate by clarifying the behavior of fatty acids and triglycerides, but you've done this at the expense of devoting even greater attention to the direct physical properties. This encourages the reader to view your post as promoting a direct cause-and-effect relationship between melting point and blood viscosity, which you say you didn't mean.

If you were talking about the different roles of these fatty acids in membrane fluidity, I think you could make a good physics point if you were to take into account lipid bilayer dynamics, as well as enzymatic regulation of membrane composition.

Instead, you have poorly controlled intervention studies where you attempt to show an in vivo effect of dietary fat on blood viscosity. These include rodent diets with deranged n-3/n-6 ratios, or human diets where "fruits, vegetables, and grains were increased" and somehow we are to assume it is the low-melting point linoleic acid in these foods that is the active component.

Your conclusion still reads the same, but still doesn't make sense. As I said before, chylomicrons begin losing TGs immediately to generate chylomicron remnants and their half-life for total uptake is only 10-20 min. VLDL has a half life of hours and is perpetually recycled. Furthermore, as Stephan repeatedly tried to communicate, the phospholipids do not *dissolve* the fats in blood, but rather *compartmentalize* them so they are essentially not in contact with the blood. Thus it is not plausible that this would make a difference between dietary and endogenous fats both because of the fats are never directly dissolved in the blood and because virtually all fats will exist in the blood primarily in lipoproteins secreted from the liver and not in chylomicrons.

Again I appreciate your skepticism and scientific inquiry and I am glad you are not intimidated by readership loss or what you see as dogmatism within the blogosphere. But I would encourage you to be more clear when you are using metaphors so that you do not generate misunderstandings.


mario_encinias said...

Well, you won't be excommunicated by me Don. I think what you're doing is refreshing and may lead to a much needed paradigm shift in the paleo community.

I find the high-fat-low-carb-paleo position in this debate suspect for two reasons. The first is that it not only failed me as a weight loss regimine (and I am very committed on these matters. I almost never deviate from paleo foods), but I became incredibly ill after gaining a lot of weight, even in spite of maintaining ketosis throughout. It was an incredibly frustrating experience especially since the insulin theory of fat regulation seemed so uncontraversal. This experience suggested confounders not easily explained by the low-carb theory.

But secondly, the out-of-Africa theory of human origins seems fairly obvious these days, especially after works stemming from the human genome project. And if writers like Chris Wilson are correct, the Kung of south east Africa boast the oldest known genotype on earth. Their diet may not be a perfect analogue for the diet of early humans but it is a pretty good candidate since they are from the same region, share the same genes, and are subject to the same ecological and economic constraints as our early ancestors. And their diet is precisely the type of diet you are advocating: high in tubers and seasonal produce, and low in large game. They delight in big kills and organ fat, but the rate of success for killing these animals is not only low, but is further constrained by compulsory meat sharing. The ethnographic accounts of Borshay Lee, the Marshall family, and others are particularly helpful on this point. Needless to say, a high saturated fat diet is not possible for these folks.

I also find it suspicious that nature would select for very strong cravings for animal fat, had this item not been scarce in our ancestral environment. Our cravings drive human food habits that were intended for an environment that was characterized by foods with "low reward properties" as Stephen Guyenet put it, and large game fat posseses high reward properties.

So for now, I'm throwing my lot in with Steffan Lindeberg, Loren Cordain,and bloggers like yourself. Keep on trucking Don. Its not easy smashing idols they way you're doing. We need it though. Take care. Mario

Sue said...

Mario, why would you keep doing paleo and remain in ketosis if it made you put on a lot of weight and become ill?

CarbSane said...

@Chris & Stephan: I think you both are overexaggerating "physiology" trumping physics. The different physicochemical properties of various fatty acids don't magically disappear when they are bound to proteins.

It may not be that melting temps have any relation to viscosity per se, but the physicochemical differences still determine how FA's must be "packaged" for transport.

At least in the obese, there IS a significant contribution to NEFA from dietary fats b/c of failure of adipose tissue to effectively trap these FA's. So this may or may not effect viscosity.

However, I believe it's really chylomicrons that influence viscosity the most. Abstract only but would seem to support that.

Interestingly enough, though, one factor influencing viscosity should be chylomicron size. The greater intermolecular forces between the relatively straight chain saturated fats - responsible for their higher melting temps - also results in smaller chylos. Those chylos are also apoB enriched and/or have a higher apoB/trig ratio and seem to be cleared faster than the larger lower apoB chylos of PUFA's. (sorry, don't have these refs at the moment). This should favor SFA to PUFA unless a higher number of SFA-rich chylos makes for more viscosity vs. lower number of larger chylo. Fluid dynamics is way cob-webbed in this brain.

I'm not sure if the chylo size differences are because of tighter packing of SFA's or that the higher protein/lipid ratio is needed to "dissolve" (and I use that term loosely) the SF trigs for transport in blood, but ...

In any case, high fat meals do induce temporary vascular impairment. Any differences between LCFA's degree of saturation seems to be physiological, not physical.

Theo said...

I think that our analysis of modern day hunter gatherers in Africa, i.e. saying that the !Kung diet is the ancestral human diet, MIGHT be flawed for several reasons. First, their ecological niche has almost certainly been altered by the impact of modern life (relocation and fewer game to hunt). Depending on where they live, that may have a very large or a small impact.

Second, the !Kung are pretty puny, not exactly robust specimens. Maybe that is environment or genetics, it doesn't really matter. It is my hypothesis that many of the tribes we see today are still around because they are such bad hunters, meaning that they failed at strength and innovation so much that they were not seen as threatening and were never wiped out.

I'm not trying to disagree with the ideas about starch intake, etc. I'm just trying to say that it is entirely possible that our ancestors ate a lot more meat and fat than the modern !Kung can get their hands on. I mean, hominid ancestors are blamed for the mass extinction of several species in Europe. It that is true, it indicates a hunting prowess way above the !Kung's abilities.

mario_encinias said...

@ Sue..I was wedded to the idea that the insulin theory of fat regulation was infallibly correct and that perhaps I was either doing something wrong and/or my body just needed more time to adapt.. Many low-carb advocates mentioned this possibility. Eventually I had to admit that there were other factors I and many others had simply missed. I was especially disheartened when i encountered other low-carb devotees with similar stories. Considering some of my symptoms like inflammed bowel, cognitive difficulties (I left a semester of my PhD as a result), and chronic fatigue, I think it may have been a type of protein poisoning or other issue of toxicity. Belief systems are incredibly difficult to relenquish, and I was hard pressed to relenquish mine.

mario_encinias said...

@ Theo...The Hadza are not puny at all and their diet is practically identical to the Kung. Read Frank Marlowe's quantitative account of these folks.

Theo said...

Thanks mario_encinias, I'll look into that.

The thing I like about these discussions is that the all the ideas seem to come out -- the good ones and the bad ones. The good ones are selected and discussed. It really helps one refine one's understanding of the material.

Grandfather Smiles said...

I second what Theo said.

Having been eating Paleo for about half a year, I must say, blog debates like this one are what help me learn about nutrition and health at a far greater pace than any other learning method. They get me to question my current food ratios and help me find weaknesses in my own approach. I am then able to research more on presented material and, perhaps, refine what I am doing.

So thanks to all!

John Doe said...

I have read through some of the comments posted in response to the authors blog post and, although I find the discussion fascinating, most of it is above my head. From a simpletons perspective I have a simple question for Don. Did you change your diet based upon scientific theories alone or was there a more practical circumstance which led you to choose a different diet? Were you feeling sick a lot, did you have high blood pressure or high cholesterol, etc. Isn't a simple blood cholesterol test one of the easiest ways to determine whether the foods you are eating are making you healthier or slowly killing you, at least in relation to the topic of this blog? Science is a wonderful thing but when you can confirm or deny science with reality, that's when you can really stand strong in your belief system.

I also find it fascinating that coconut oil is a saturated fat that many studies have shown to have incredible health benefits. So when people like yourself, Don, argue against consuming saturated fats are they really arguing against the consumption of saturated fats universally or are they making distinctions between plant-based saturated fats and animal fats?

Don said...

John Doe,

Read my post, Farewell to Paleo. On an animal-based high fat diet, I had among other issues a high serum cholesterol level.

I do not agree that "many studies" have shown that coconut oil has "incredible health benefits." Dr. Greger has some nice videos on this topic, with photos from peer-reviewed reports of the negative effects of coconut on cholesterol levels.