Saturday, May 28, 2011

Who Said Paleo Diet Was High In Fat? Part 2.2

In her editorial comments and letter to the editor remarking on the Cordain et al paper on Plant-animal subsitence ratios of world-wide hunter-gatherers, Katherine Milton also made several other comments casting doubt on the idea that evolutionary human diets contained low proportions of plant foods and carbohydrate, and high proportions of animal protein and fat.  Rather than quote, I will paraphrase:
1.  Humans come from an ancestral lineage—i.e. primates–– in which plant foods have traditionally served as the primary source of energy.

2. The human gut displays a protracted transit time, averaging 62 hours with low-fiber diets and 40 hours with high-fiber diets.  “In striking contrast to humans and all great apes, all extant Carnivora show a rapid turnover of ingesta. For example, a 370-kg polar bear takes ≈24 h to digest a seal carcass.”

3. “To date, few genetic adaptations to diet have been identified in humans, suggesting that, in their evolution, humans tended to resolve dietary problems primarily by using technology rather than biology.”

4.  “The technologic abilities of humans derive from their unusually large, complex brain, a brain that, under normal conditions, is fueled by a steady supply of glucose. Consumption of digestible carbohydrate is the most efficient way for humans to obtain glucose for brain function. Potential alternatives—gluconeogenesis or the use of ketones to fuel the brain—represent alternative, more costly metabolic solutions.”

Our ancestral primate lineage consumed, so far as we can tell, a plant-dominated omnivorous diet.  This does not mean that humans should eat like other extant, closely related primates (and Milton does not suggest that we should).  We don’t have exactly the same digestive physiology as chimps or gorillas; we don’t have the equipment required for a raw plant-dominated omnivorous diet. 

However, as Milton notes, we also lack the digestive physiology of the top carnivores.  Even on a low fiber diet, due to our proportionately long small intestine, the passage of food through the human gut takes more than 2.5 times longer than transit of food through the guts of extant carnivores.   In this respect, the human gut functions more like that of other primates than that of top carnivores.  This strongly suggests that the human gut remains adapted to a plant-dominated diet.  

As well, the fact that humans produce more salivary amylase and have more copies of the AMY gene that codes for amylase production than other primates also suggests that evolutionary diets contained substantial amounts of starch.

Although we have clear evidence that ancestral humans succeeded in expanding the animal-source component of their diets and good reasons to believe that this may have played a role in expansion of the human brain by providing neural fatty acids, this evidence does not tell us that ancestral humans had abandoned the primate tradition of plant-dominated nutrition.  
The expansion of the human brain through evolution does not necessarily indicate a move to a diet composed largely of grassland animal meat and fat.   The idea that “meat-based diets support brain development better than plant-dominated diets” doesn’t gain much support from comparative anatomy.  Primates build larger brains (relative to body mass) than other species on largely vegetarian diets.  The chimps living on 95 percent plant food diets have a body:brain ratio (by mass) of 100:1.  Elephants eating plant-dominated diets have a ratio of 851:1.  Tigers living on virtually 100 percent grassland animal food have a body:brain ratio of 1000:1.  

 Thus, the 95 percent vegetarian primate eating a has a brain 10 times larger than the carnivorous tiger on a relative basis.  Moving the diet in the direction of the tiger’s (predominantly land animal flesh and fat) would seem to favor a smaller, not larger, brain. Notably, children recovering from malnourishment appear to build normal brains on diets containing as little as 8 grams of animal protein per day (one-third of total protein requirement) if provided adequate essential fatty acids.[1, 2, 3 ]

Some have suggested that consumption of brains and marrow supported development of hominid brains in evolution.  Of interest, Ta'i chimpanzees regularly eat the brain, eyes, and marrow of long bones of colobus monkeys [4 pdf] and presumably have been doing this for millennia, but the chimp’s brain is only one-third the size of a human’s.  Perhaps some other nutritional factor supported hominid brain expansion?

Dolphins have a brain slightly larger than human (1.8 kg vs 1.4) and a body:brain ratio similar to humans (dolphins, 42:1, humans, 56:1).   For development and function, the human brain requires specific vitamins (B12, folate, B-complex) and minerals (iodide, iron, copper, zinc, and selenium) in addition to essential fatty acids.  On a weight basis, shellfish, eggs, finfish, pulses, and cereals all provide greater concentrations of these minerals than meat.[5, pdf]

From this data, using single foods, a human can satisfy all requirements for brain-selective minerals by eating 2 pounds of shellfish, 5 pounds of eggs, 8 pounds of fish, or 11 pounds of meat.  With this basis, one can imagine that only small amounts of seafoods to a plant-dominated diet would provide a hominid with adequate nutrition for building a large brain.

The evolutionary expansion of the brain actually increased the demand for glucose provided by plant foods.   Nonhuman primate brains only use 8-9% of  resting energy expenditure, but the modern human brain uses about 20-25% of resting energy expenditure and two-thirds of all glucose used by the body, while the mass of the human gut is only 60% of expected for a similar sized primate.  Although the brain can adapt to a large extent to use of ketones instead of glucose, this is an inefficient way to fuel the brain and appears to stimulate the stress i.e. fight or flight response.  For example, it appears that the ketogenic diet’s antiseizure effect depends on stimulation of the sympathetic nervous system since disabling norepinephrine destroys the anti-seizure effect of the diet. []

The nervous system and red blood cells use an estimated 150 g of glucose daily.  Non-neural tissues will also use glucose for fuel when provided.  The conversion of amino acids to glucose unnecessarily burdens the liver with ammonia to detoxify, and the kidneys with sulfuric acid to eliminate.  Therefore, for smoothest operation, humans need food that supplies large amounts of glucose in a compact and readily usable form (i.e. not amino acids).  This is supported by the beneficial effect that high carbohydrate diets appear to have on mood [7], and reflected in the fact that the body preferentially burns glucose as fuel and stores glucose as glycogen in the liver and muscles rather than as fat.[8 , 9 ] Animal tissues are universally low in glucose, therefore not serving as the best source of substrate for brain metabolism. 

The body also needs glucose as a substrate for the manufacture of a number of other important functional and structural compounds including:

  • Glucuronic acid, which binds to substances to facilitate their transport around the body. In this way glucuronic acid is largely responsible for the elimination of toxins such as drugs, excess hormones, and foreign chemicals.
  • Hyaluronic acid, a constituent of extracellular fluids, needed for synovial fluid, cartilage, and skin.
  • Chondroitin sulfates (a type of glycosaminoglycan), constituents of cartilage
  • Immunopolysaccharides
  • DNA and RNA
  • Heparin (another glycosaminoglycan), an endogenous anticoagulant

None of this means we should “go vegan.”  Animal products of the best quality—particularly seafoods-- provide important nutrients not adequately provided by plants, not the least of which is vitamin B-12.  However, as a student of evolutionary nutrition, I find it noteworthy that we have enterohepatic recirculation of vitamin B-12.[10]   As noted by Herbert [8 ], this can prevent B-12 deficiency from occurring in a previously omnivorous adult vegan for 20-30 years:

“The enterohepatic circulation of vitamin B-12 is very important in vitamin B-12 economy and homeostasis (27). Nonvegetarians normally eat 2-6 mcg of vitamin B-12/d and excrete from their liver into the intestine via their bile 5-10 mcg of vitamin B-12/d. If they have no gastric, pancreatic, or small bowel dysfunction interfering with reabsorption, their bodies reabsorb ~3-5 mcg of bile vitamin B-12/d. Because of this, an efficient enterohepatic circulation keeps the adult vegan, who eats very little vitamin B-12, from developing vitamin B-12 deficiency disease for 20-30 y (27) because even as body stores fall and daily bile vitamin B-12 output falls with body stores to as low as 1 mcg, the percentage of bile vitamin B-12 reabsorbed rises to close to 100%, so that the whole microgram is reabsorbed.”

Looking at this from an evolutionary perpective, why would the body have such efficient recycling of vitamin B-12 but not of other B-complex vitamins?  Logically, nutrient recycling probably represents a response to scarcity and infrequent consumption, so I would guess that this system would most likely have developed in response to a diet that did not have a continuous rich daily supply of vitamin B-12, with high B-12 intake occurring sporadically. 

Of further interest, which animal products supply the most vitamin B-12 per serving?  From the Linus Pauling Institute page on vitamin B-12 [11] :

Shellfish, not land animals, appear the richest sources.  An adult human requires about 2.4 mcg of vitamin B-12 daily.  Given the fact that we recycle vitamin B-12 with a net loss of only 2-5 mcg per day from body stores, three ounces of beef or salmon or 4 large eggs daily, 3 ounces of crab once every other day, 3 ounces of mussels once every 3-4 days, or 3 ounces of clams once every ~20 days, would satisfy the requirement. 

So, reverse engineering from current human vitamin B-12 metabolism seems to suggest that for a very long period of time, long enough to establish our baseline B-12 metabolism, our ancestors infrequently consumed foods rich in vitamin B-12. 

On the other hand, applying the same thought process to modern human carbohydrate (glycogen) and fat storage suggests frequent consumption of carbohydrates and infrequent consumption of fats in ancestral diets.  But I’ll save that topic for a future entry.


Theo said...

Your comment about storing fat and being unable to store carbohydrate is interesting in support of the idea that we consumed carbohydrate frequently.

I wonder if there are animal examples to support this. For example, are animals that do not get carbohydrate often very good at storing it? Or are they just very good at making it from protein? Also, do carnivores, which would be getting a lot of fat in their diet, store significantly less body fat than herbivores? I know less about carnivores than herbivores so I'm not sure if this applies, but there are certainly very fatty and also very lean herbivores.

Eric said...

The whole "enterohepatic recirculation of vitamin B12" question is quite interesting. A similar system would appear to also be in place for vitamin D.

I also have always considered it interesting that breastmilk is so "relatively" low in protein (for a period where maximum growth is occurring) and high in carbs yet, how do we reconcile the fact that, by percentage of calories, it is also very high in fats?

Stephan Guyenet said...

Hmm, I'm not sure about Milton's numbers for GI transit time. I think traditional cultures eating relatively high-fiber diets in Africa have transit times <24 hrs, and the average Westerner is closer to 40. I know mine is on the order of 24 hours, sometimes less.

Don said...


We can store carbohydrate as glycogen, but much less than fat and in an interesting volume.


I think breast milk is high in fat to provide compact source of calories to facilitate weight gain in a growing mammal that has a metabolic rate 2-3 times that of an adult. Think about that...what if you aren't growing that fast and don't have half the metabolic rate?


Another source I consulted quotes Burkitt and Trowel as saying that transit time is ~30 h in humans on high fiber (100+ g/d) and >48 h in humans on low fiber (≤20 g/d). I would guess that it varies with fiber intake, so different researchers may report different times. I know mine was up to 96 hours on zero carb, but is closer to 24 hours on high fiber.

But the important thing is that the bear has a transit of ~24 h on zero fiber meal (seal), whereas rapid transit in humans occurs, in most people, to the best of my knowledge, only with high fiber intake.

Don said...


Re VT-D...yes. The fact that we have an endogenous system of production shows that in evolution, dietary vitamin D was scarce to non-existent, or at least insufficient to meet requirements, so we needed our own means of producing it from sunlight.

Don said...


Animals that don't get carbohydrate frequently, i.e. carnivores, have more efficient gluconeogenesis than humans, make glucose from protein more efficiently, but so far as I know they don't efficiently store it, since they don't make enough or eat enough to store. I don't know of any really fat would clearly impair performance in hunting to carry around much body fat, so evolution would select against fat storage in carnivores.

Bill Strahan said...

3 ounces of mussels...funny timing. I was having an INTENSE craving for mussels. Weird since I've only had them twice in my life. I ran out to Whole Foods and picked up 3 pounds of them.

After steaming in a garlic/butter/wine concoction, I removed the mussels from the shells while the juices reduced to a thick sauce. Out of curiousity, I weighed the mussel meat. 6.8 ounces!

I was hoping it would be closer to 10 or 12, but the edible portion actually ended up being 6.8 out of 48 ounces. About 14%.

Regardless, I combined that with 8oz blackened tilapia and about half a head of cauliflower. There was definitely something my body needed from the mussels. Within 20 minutes of dinner I was calm, relaxed, and totally satisfied.

Hmmm, come to think of it I've been craving shrimp and scallops a lot lately, so perhaps my diet (rich in grass fed beef, liver, eggs, and vegetables) wasn't providing something I require.

More shellfish. I'm on it.

Kor said...

So Don, when can we look forward to seeing some new additions to your My Meals section along the lines of these recent posts?

john said...
This comment has been removed by the author.
john said...

Comparing a "vegetarian primate" brain & diet to a tiger [carnivore] brain & diet is unfair, since considering dolphins, which are mentioned just after, counters that point--unless dolphin vitamin/mineral intake overrides the lack of plants/glucose. Across primates, there's a strong correlation between "diet quality," which is strongly dependent on animal food content, and brain:body ratio...chicken or egg?

The amount of required glucose can be met with a somewhat low plant intake (in terms of % calories). I don't see how the glucose-brain argument supports a high or low carb diet.

Regarding transit time, anyone ever eaten tom yum soup on an empty stomach?

Teech said...

Which herbivores are on high carbohydrate diets? From my understanding, herbivores like cows ferment fiber in their gut into fats and get most of their energy from fat, not carbohydrate. We do the same thing with greens, we ferment it into fat.

The above article suggests early hominds ate grasses which are low in carbohydrate.

Technically, if I ate only low carbohydrate vegetables I would be on a high fat diet.

Don said...


Primates violate the idea that diet high diet quality is required for large brain to body ratio. Tigers have smaller brain (relative to body weight) than elephants. That was my point, showing data that contradicts the claim that across species, high "diet quality" predicts larger brain relative to body weight.

Most likely, dolphins have better gluconeogenesis than humans.

The human brain uses more glucose than a chimp brain. Therefore, on a weight basis, humans require more glucose than chimps.

Don said...


Herbivores do ferment fibers to fats. They also have small brains with low glucose requirements. Grasses and leaves provide some sugars, as well as fermentable fibers, but not enough to support a human brain. If you eat only greens, you will have to eat a lot of protein to meet your glucose requirements, or force the body to run the brain on ketones. It sort of works, but its not efficient. You end up having to kill animals just for fats, wasting the muscle tissue. Anthropologists call it selective hunting. Its probably one of the reasons for the anthropogenic Pleistocene extinctions. By the extinctions, the course of Nature selected against humans continuing that way of life. It didn't work (because it couldn't efficiently meet human nutritional requirements) so people had to find a different way of life.

Michal said...


Aren't you contratidcting yourself by first gloryfing large brains of elepthants, and then saying
"Herbivores do ferment fibers to fats. They also have small brains"?

What you forget is what amount of time is actually available to utlize a large brain. Elephants spend ~80% of their non-sleep time for food-related stuff. Even gorillas have to waste >50 percent of their time for this. So the only possible usage of their brains is just for remembering a lot (superior elephant memory!), and recalling it when necessary. Rest goes for social activities and reproduction.

Eating fat liberates time (lowest chewing time/calories ratio), that can be spend on inventing stuff, doing art, learing, teaching, playing, training.

Teech said...

Thanks for the response Don. I like that your blog makes us think about our current ideas on diet.

Posted by Stephen Guyenet...Its a long read but what he says that Kitavans could be getting 30% of calories just from fat from the fiber they eat. Industrialized nations could be getting around 5-20% depending on how much fiber your eating. He also mentions how this fact tends to be ignored and diets may seem more carb heavy than previously thought.So we still could be eating high fat on a plant biased diet with a little animal fat thrown in. What do you think.

I also didn't fully appreciate the caloric contribution of fiber to the human diet. In industrialized countries, fiber may contribute 5 to 10 percent of total calorie intake, due to its conversion to short-chain fatty acids like butyrate in the large intestine (free full text). This figure is probably at least twice as high in cultures consuming high-fiber diets. It's interesting to think that "high-carbohydrate" cultures may be getting easily 15 percent of their calories from short-chain fats. Since that isn't recorded in dietary surveys, they may appear more dependent on carbohydrate than they actually are. The Kitavans may be getting more than 30 percent of their total calories from fat, despite the fact that their food is only 21 percent fat when it passes their lips. Their calorie intake may be underestimated as well.
Interesting. :thup: I still think it's misguided for The Authorities to recommend to the public that they eat upwards of 40 grams of fiber per day, especially since most of them will do it with All Bran bars and cereals, but it lines up pretty nicely with paleo philosophy.

If you going to go carnivore, be generous with the butter.


CarbSane said...

Teech - it seems you forgot the hyperlinks?

I've never really bought into the notion that certain species get XX% fat in their diets from fermented fiber.

The short chain fatty acids do not behave either physicochemically nor metabolically like long or even medium chain ones.

It is possible that VHC cultures like the Kitivan do eventually "burn" more fat as fuel due to de novo lipogenesis (an insignificant path otherwise), but it's a stretch to equate that with dietary fat.

Actually the absence of significant DNL coupled with the amylase for digesting starch would support a higher "bioavailable" carb intake in humans. We don't have a developed ceocum (sp?) while hind gut fermenters do.

Don said...


Are you claiming that people eating high carbohydrate (starch-based) diets don't have time to use their brains? That should be news to the Japanese. I've eaten every ratio from 70+% carb to 0% carb (by energy) and never did eating take up all of my time. I think you'll have a very hard time establishing that among humans a high fat diet leads to more creativity, playing, training, etc. than a high starch diet.

I am not contradicting myself and I didn't "glorify" the brains of elephants. I simply pointed out that an elephant has a larger brain:body ratio than a tiger, which contradicts the claim that meat-eating produces a higher brain:body ratio across species. Similarly, the chimp also demonstrates that the so-called "low quality" plant based diet doesn't prevent a chimp from having a higher brain:body ratio than the tiger on the "high quality" diet. On the other hand, RELATIVE TO HUMANS AND CHIMPS, elephants have a small brain: body ratio, so their demand for glucose would be smaller on a weight basis than for either the chimp or the human; hence, they many not need a high glucose diet for optimal brain function.

Don said...


So far as I know, in humans, fermentation of fiber is an insignificant contributor of fat to the body's fat economy.

If it was, it certainly would add even more mystery to the fact that overfeeding high carbohydrate and high fiber intake results in lower body fat gain than overfeeding fat or protein.

And subjects need more kcalories to maintain weight on low fat diets than on high fat diets:

Don said...


Grass-eating hominins...imagine that. Its only a short step from eating grass to eating grass seeds. Maybe grains have been in the diet longer than paleo theorists would like to think?

Tuck said...

Burkitt and Trowell, in their book "Western Diseases, Their Emergence and Prevention", cite transit times of healthy Africans on traditional diets of between 24 and 36 hours.

They didn't test Africans on a low-fiber diet, as they didn't test the Masai.

I don't recall any of the transit times being much longer than that, but I'm unable to locate my copy at the moment.

Alan said...

eating protein is an inefficient way to get glucose calories for my brain and red blood cells to run on?

Sounds like just what I want... less total calories to get stored as bodyfat!

Alan said...

The problems of our era are: (a) people using food as a hedonistic pleasure source, notwithstanding that they're no longer particularlly hungry; and, (b) eating excess amounts of calories before reaching a feeling of satiation.

Copying primitive man's dietary intakes does not address either of those problems.