Sunday, November 30, 2014

Did Hunter-Gatherers Eat at least 100g of Fibre per Day?

One figure that gets mentioned occasionally (for example, by David Gatz, Jeff Leach and Marlene Zuk) is that hunter-gatherers consumed an average of 100g of fibre per day.  It’s often mentioned to argue that hunter-gatherers ate a plant-based diet and/or that Paleo dieters aren’t eating a ‘true’ Paleo diet unless they’re eating 100g of fibre
Estimating Fibre Intake of Hunter-Gatherers
Where does this figure come from?  Here’s what I came up with following a brief look at the literature:
·         In Eaton and Konner’s origin paper, ‘Paleolithic Nutrition — A Consideration of Its Nature and Current Implications’, they estimated that hunter-gatherers had an average fibre intake of 45.7g based on nutrient analysis of wild plant foods and plants providing 65% of total calories [1]
·         However, in a more recent paper Eaton, et al estimate hunter-gatherers consumed an average fibre intake of 86g [2]
·         “Analysis of vegetable foods consumed by foragers in this century (Table 1) and evaluation of archaic native American coproliths suggest that ancestral human fiber intake exceeded 100 g/d (Eaton 1990). Rural Chinese consume up to 77 g/d (Campbell and Chen 1994), rural Africans up to 120 g/d (Burkitt 1983)” [2]
·         Australian Aboriginals have been estimated to eat 40-80g of fibre based on the nutrient analysis of native plant foods and plants providing 20-40% of total calories [3]
·         Jeff Leach (Human Food Project) and Kristin Sobolik estimated a hunter-forager group in the northern Chihuahuan Desert consumed an average of 150-225g of fibre per day, with 135g coming from “inulin-like fructans” from plants such as agave, sotol and onion.  However, this hunter-forager group may be an outlier as those plants are unusually high in fructans.  The paper estimates that 25% of the fructans would be degraded by cooking [4]
It seems like there is some evidence to support the idea, but there is huge inconsistency too.  Also, these estimates should be taken with a grain of salt (quite literally, because the estimated sodium intake of hunter-gatherers is really low):
·         These estimates may not adequately factor for cooking and other processing
·         They may not approximate what plant foods hunter-gatherers actually ate.  For example: simply averaging all wild plants may overestimate vegetable intake, thereby overestimating fibre intake.  “This category included the leaves, leaf buds, pith and stalks of various plants. As a group they played a fairly minor role in AA diets”* [3]
·         Another reason to be somewhat sceptical is that other estimates of hunter-gatherer nutrient intake are unrealistic.  For example, the estimates for protein are very high – 19-37% [5], 34% [1], ~50% [3] – considering that maximal urea synthesis equates to a protein intake of 29.7–40.9% of total calories in a 3000 calorie diet [5].
Putting 100g of Fibre into Perspective
The Australian adequate intake (AI) for fibre is 30g for men and 25g for women.  This is based on the median population intake + a few grams because resistant starch isn’t measured as fibre [6] (this is consistent with measurements from the ABS [7]).  The suggested fibre intake to reduce chronic disease risk is 38g for men and 28g for women, which is intake of the 90th centile [6].
The US AI** for fibre is 14g/1000kcal (28g in a 2000 calorie diet, 42g in a 3000 calorie diet) [8], which is similar to the Australian recommendation.  However, average fibre intake in the US is ~15g [9], quite a bit lower than Australia
The current fibre recommendations and the average intake in both Australia and the US is a long way off 100g, but is this a realistic target?
Using data from the USDA nutrient database and the NUTTAB nutrient database I’ve estimated what it will take to meet 100g of fibre.  The numbers are the medians for a category of food and the number is in red if a food category exceeds 2000g or 3000 calories.  These numbers may be a slight overestimation if fructans and galactans are classified as sugars or starches in the nutrient databases rather than fibre.  I also excluded categories of food with low/no fibre (animal foods, beverages, junk foods) to improve readability.
Data from ~200 foods (raw) in the USDA nutrient database.  Download here
Weight (g)
Non-Starchy Vegetables
Starchy Vegetables
Sugary Fruit
Whole Grains
Immature Legumes
Mature Legumes
Data from the entire NUTTAB nutrient database.  Download here
Weight (g)
Herbs, Seasonings and Spices
Bread and Bread Products
Breakfast Cereals
Flours, Grains and Starches
Nuts and Seeds
Data from the entire USDA nutrient database.  Download here.  The USDA nutrient database may overestimate weight and calories needed to meet 100g of fibre because food categories often include both whole foods and food products.
Weight (g)
Spices and Herbs
Breakfast Cereals
Fruits and Fruit Juices
Vegetables and Vegetable Products
Nut and Seed Products
Legumes and Legume Products
Cereal Grains and Pasta
American Indian/Alaska Native Foods
It’s quite evident how extremely difficult it would be to get 100g of fibre with commonly available foods.  Almost all plant foods would need to be eaten in huge quantities, with a few exceptions.  Even still, to be most efficient, you would need to eat ~1000g and almost ~2000 calories of mature legumes to meet 100g of fibre, leaving little room for animal foods.  One might argue that plant foods these days are higher in sugar/starch and lower in fibre than those in the Paleolithic, but that doesn’t seem to be case [10].  So in conclusion I think it’s unlikely the average hunter-gatherer group consumed 100g fibre, particularly with plant foods providing 35% of total calories on average [5]
* Unfortunately Cordain, et al didn’t give a fibre estimate in the paper ‘Plant-animal subsistence ratios and macronutrient energy estimations in worldwide hunter-gatherer diets’.  Though I suspect it would have been similar to the estimate of Australian Aborigines given that the macronutrient composition of wild plant foods came from that paper [5] and because the both papers had a similar estimated energy intake from plant foods 20-40% [3] vs. 35% [5]
** The US adequate intake (AI) is defined as: the recommended average daily intake level based on observed or experimentally determined approximations or estimates of nutrient intake by a group (or groups) of apparently healthy people that are assumed to be adequate—used when an RDA cannot be determined.” 
*** Denis Burkitt may have been another source of the idea that hunter-gatherers consumed 100g of fibre per day, but unfortunately I can’t track down his peer reviewed papers.  His hypothesis that diverticulosis and colorectal cancer are due to low fibre and can be improved by increasing fibre is not well supported [11] [12] [13]

Thursday, November 27, 2014


I’m about one and a half years late reading ‘Paleofantasy: What Evolution Really Tells Us about Sex, Diet, and How We Live’ by Marlene Zuk, but I thought reading and writing about it would be informative because it seems that so few people in the Paleo community have actually read it.
What is meant by ‘Paleofantasy’?
The term Paleofantasy was coined by Leslie Aiello and refers to:
“…stories about human evolution based on limited fossil evidence, but the term applies just as well to the idea that our modern lives are out of touch with the way human beings evolved and that we need to redress the imbalance”
Marlene Zuk also states that:
“The paleofantasy is a fantasy in part because it supposes that we humans, or at least our protohuman forebears, were at some point perfectly adapted to our environments”
She emphasises this point quite strongly, and uses this, along with evidence of rapid evolution, to argue against mismatch theory, which then forms a basis for her arguments throughout the rest of the book*.  While evolution won’t result in a species being ‘perfectly’ adapted to an environment, this is somewhat of an academic point as the premise of mismatch theory works just as well if we’re comparing environments a species is well adapted to vs. poorly adapted to.  Also, while evolution can be rapid in response to strong selection pressures**, these changes are often relatively superficial, such as height or beak size, and rarely substantially effects a species’ physiology/biochemistry, which is what we’re concerned with when discussing diet, exercise, circadian rhythms, etc.  Her general dismissal of mismatch theory has also been challenged in a paper titled: Throwing Out the Mismatch Baby with the Paleo-Bathwater.  As the paper discusses, mismatch theory has been very informative and helpful.
But it’s also important to remember that mismatch theory generates hypotheses using whatever information we have about our past.  Sometimes the information is faulty and at other times parts of the old environment are harmful (such as infectious disease) or part of the new environment are beneficial (such as medicine).  As with all hypotheses of this nature (such the lipid hypothesis) we can’t just assume it’s universally true for all elements of the old and new environments and shouldn’t stop testing each element following the first success or disappointment.  Matt Lalonde made a similar point in his AHS 2011 presentation (see here).
“As an anthropologist, I observe that Zuk’s use of the term ‘fantasy’ is just an emphatic way of describing the hypothesis-forming that is essential to evolutionary science. We play with hypotheses, explore their predictions and try very hard to falsify them. So it is, in a way, unremarkable that so many hypotheses proposed by anthropologists about ancient environments now seem to be wrong — and, in a few cases, even ridiculous.  It means that science is working.”John Hawks
Also, a point that shouldn’t need to be repeated so often is this: this is not about re-enactment.  The purpose of formulating hypotheses using mismatch theory together with evidence from the ancestral environment is to improve quality of life.  Emulating specific elements of the ancestral environment is a means to an end (quality of life), rather than an end in itself.
With that in mind let’s look at some of the chapters.
* Regarding this point, it’s informative to consider her perspective as an academic, namely that her research on the rapid evolution of crickets in Hawaii.
** Rapid evolution would have occurred following the adoption of agriculture, as agriculture would have exerted a strong selection pressure on early agriculturists.
Two of the ten chapters in the book are on diet.  The first is on dairy and the second is on grains and meat.
The chapter on dairy is about the genetics and advantages of lactase persistence and that lactase persistence is evidence that we have adapted to dairy.
While the hardcore Paleo view is essentially dairy = not Paleo, that hasn’t been the main view in the Paleo community for a while.  Dairy foods (milk, cheese, yoghurt) are reasonably nutrient dense and full fat dairy is associated with better health outcomes in observational studies [1] [2]*.  But some people don’t tolerance dairy well, whether it’s due to lactose intolerance, an allergy to dairy proteins or a milder adverse response (such as acne, clogged sinuses or mild GI symptoms).  The advantage with something like the Whole 30 followed by a reintroduction period is an awareness of how certain foods affect you that you may never have had before.  In my perspective this seems to be the main view in the Paleo community and is a far cry from ‘dairy = not Paleo, therefore don’t eat’.
* It’s surprising that full fat dairy was associated with better health outcomes given that it’s likely associated with an unhealthy lifestyle, and therefore is unlikely that this result is due to confounding variables.  Also see SuppVersity for articles on the science relating to dairy.
Grains and Meat
There actually wasn’t that much on grains, though in one section of the chapter she says Paleo is anti-starchy vegetables and grains, then proceeds to debunk this partial strawman (the strawman here bring that Paleo is anti-starchy vegetables) by saying we have adapted to increased carbohydrate consumption through more gene copies of salivary amylase.  However, the main Paleo view is that adaptation, or lack thereof, to grains as a staple in the diet never had anything to do starch digestion or carbohydrates, but rather to potentially harmful compounds in grains and the lack of bioavailable nutrients in minimally processed whole grains due to compounds like phytic acid.  That humans don’t produce phytase (an enzyme to break down phytic acid) is an argument that we are not well adapted to consume minimally processed whole grains as major dietary staple (which is what’s being recommended by dietary guidelines).  Of course there are ways of preparing foods to maximise nutrient absorption and minimise toxins, cooking being an obvious example.  Marlene Zuk mentions this elsewhere in the chapter (see below) but most national dietary guidelines ignore this while recommending minimally processed whole grains be our major dietary staple even though average micronutrient intakes are far from ideal [3].
“George Armelagos, an anthropologist at Emory University, suggests that distinctive cuisines developed as a way of restricting the potential for eating poisonous or otherwise unsafe foods in the environment.”
This chapter sets up the strawman that Paleo is essentially carnivorous, then debunks it, heavily featuring Katherine Milton in the process.  She makes several bizarre statements in this section such as quoting Katherine Milton saying: “…regardless of what Paleolithic hunter-gatherer societies were eating, there is little evidence to suggest that human nutritional requirements or human digestive physiology were significantly affected by such diets at any point in human evolution.”.  It’s unlikely, to say the least, that vitamin B12 and K2 being essential nutrients, our high requirement for choline/betaine, the extremely poor conversion of ALA to DHA and our smaller large intestine was not affected by diet in human evolution.  I also don’t think it’s a coincidence either that, in relation to our nutrient requirements, there are certain nutrients that animal foods are good sources of but plant foods aren’t and vice versa and therefore our nutrient requirements seem to be ideally met with combination of plant and animal foods.
“If the proponents of paleo diets were correct, however, those hunter-gatherers with higher levels of protein in their diets should be less likely to suffer from modern maladies like obesity and hypertension than do cultures consuming more plants and starches, but no evidence exists to support such a claim. Such diseases are indeed absent in most foraging societies, but they are more uniformly absent than would be expected if a high meat intake were the preventive.”
I don’t know where she got this idea from.  While protein and meat are considered beneficial in the Paleo community, just like plants and starches vegetables are, most dietary explanations of disease from the Paleo community are mainly centred on too much junk food which is pretty much universally absent from all hunter-gatherers.  Then she just uncritically accepts that “The biggest problem with many Western diets seems to be their energy density…”
“I want to make it clear that I am not discussing the relative health merits of the paleo versus Atkins versus Mediterranean diets, or any other particular way of eating. I am also not concerned with the details of exactly how much animal protein in the diet is necessary to constitute a real paleo diet, or which micronutrients would have been present in which quantities of food in the period before agriculture. I also realize that not all paleo dieters are doing exactly the same thing, and that not all are really attempting to replicate ancient meals.”
“What’s more, although I confess to a certain bemused fascination with the minutiae of, for example, how large a yam one can consume and still have the overall diet be considered paleo, I realize that virtually all diets— vegan, vegetarian, organic, and so on— have their fringe adherents, some of whom can become quite vehement.”
However, you don’t get that message anywhere else as the book:
·         Paleofantasy is dominated by quotes and views from fringe adherents on online forums, rather than more balanced views, either from scientists or from the people who have the most influence in a particular community (something else discussed in that paper).
·         She also does discuss the health merits of diets and does so in perhaps the worst way possible, firstly by mentioning the U.S. News & World Report Diet Rankings.  In these rankings Paleo is last, behind a variety of popular diets such as Slim-Fast, veganism and raw food.  I won’t waste my time and yours discussion how stupid the rankings are because it should be quite obvious.
·         She also misrepresents the studies on the Paleo diet, saying “A few studies have attempted to examine the effects of switching to a paleo diet on various health measures, including weight, of groups of subjects”.  But then only discusses one study that didn’t have a control group even though two Paleo RCTs were published years before Paleofantasy was published.
The chapter on exercise begins by dismissing Crossfit, Arthur DeVany’s Evolutionary Fitness and the exercise like a hunter-gatherer paper by O’Keefe, et al for pretty much no other reason than the evolution/HG argument behind them.
She then transitions to discussing how it’s essentially all about sitting and a lack of non-exercise activity thermogenesis (NEAT), which along with the thrifty gene hypothesis (TGH) she accepts uncritically.  This surprised me somewhat, because the TGH is perhaps the most commonly mentioned ‘just so story’, which also happens to be poorly supported by evidence [4] [5] [6].  It’s also interesting she put the TGH in the exercise chapter because the TGH relates more to diet than exercise, and exercise is generally not all that helpful for weight loss.
This is followed by a section on endurance running and walking/running barefoot, which understandably is generally mentioned quite favourably.  As she says, persistence hunting makes sense given our anatomy.
* FYI: “The success rate of such persistence hunting is difficult to gauge, since so few hunts were recorded, but Liebenberg suggests that it may be about 80 percent, which compares very favorably with other hunting methods, such as hunting with clubs and spears or using dogs. The latter technique is the one that seems to have yielded the highest amount of meat per unit of effort, though again the number of documented hunts was relatively small. Such a success rate is also considerably higher than that obtained by mammalian predators such as lions or raptorial birds”.
** She mentions many Paleo advocates are against running long distances, but I think this has to do more with long distance races (high intensity, long duration) rather than persistence hunting (low intensity, long duration).
The main message in the chapter on disease was arguing against hunter-gatherers being free of disease and that disease and death are inevitable parts of evolution.  But most people aren’t looking for immortality or perfection.  Reducing the risk of chronic disease is a means to an end, such as improving quality of life.
She discusses how hunter-gatherers get cancer, but just because some hunter-gatherers had cancer (or another non-communicable disease) doesn’t mean the risk of cancer (and other diseases) can’t be modified by diet and lifestyle.  There is an explanation for everything and diseases have underlying causes.  There would be an explanation for how some hunter-gatherers had cancer, such as infectious disease.  This argues against sweeping statements that hunter-gatherers had no disease but doesn’t argue against a similar idea, that humans are not broken, or that our disease risk is not solely determined by our genes.
She also says that cancer in an occasional accident of multi-cellular life just as autoimmune disease is an occasional accident of having an immune system*, which is true, but incredibly lazy and reinforces the prevailing message of hopelessness regarding disease.  “But that fine-tuned ability can malfunction, and when it does the result can lead to the unregulated growth that is cancer, as well as a host of other maladies”.  The same could be said of many diseases, but how this ability malfunctions and what we can do about it are the real questions, which understandably is beyond the scope of the book.
* I suppose it’s better than the people who argue that certain chronic diseases are evolutionarily adaptive or a sign that the body is working properly (such as the thrifty gene hypothesis, some parts of evolutionary psychology, the carnivore connection hypothesis, type 1 diabetes being cyroprotective and acne signalling immaturity.
Other Chapters
There were chapters on what evolution says about family and sex, which I found interesting.  I don’t know much and don’t have much opinion on either subject.  In isolation, I thought they were mostly well covered, but having read the previous chapters I wondered if she had given these topics a fair trial, which some argue is not this case [7].  These chapters were also quite dismissive, for example: “We have obviously already altered child-rearing practices in many different ways around the world, and most children grow up just fine.”  I found this to a surprising statement for a scientist to make considering the overall goal of science seems to be to advance knowledge and improve life.
The final chapter was on the question of whether we are still evolving.  Once again, it’s pretty clear from the selected quotes of others that some people have a poor understanding of evolution.  The answer is yes, which can be summed up by a quote in the book from Mary Pavelka “The question ‘Are humans still evolving’ should be rephrased as ‘Do all people have the same number of children?’”.
If Paleofantasy was just about correcting misconceptions of evolution and emphasising the uncertainty of drawing inferences about our past I wouldn’t mind at all and in fact it’s pretty clear that needs to be discussed as some people at least have a very poor grasp of evolution and ‘just so stories’ are quite ubiquitous.  However, neither of these is the case.  While Paleofantasy does discuss misconceptions of evolution it’s clear that either Marlene Zuk or her editor/publisher/etc (to create a bit of controversy) has an axe to grind with the Paleo diet and pretty much anything else that uses mismatch theory as a justification.  Rather than using constructive criticism (which I wouldn’t have minded either), the tone of the book is highly dismissive and ideas are ‘debunked’ often by using strawmen and poor arguments (many of which weren’t included in this blog post).  Although, she is far more reasonable in interviews, where the primary topic is misconceptions of evolution.
For these reasons, I would only recommend Paleofantasy to people using evolution or mismatch theory to promote something (diet, exercise, etc), as the book emphasises the uncertainty of the evidence and can help refine how concepts like evolution and mismatch are discussed.

Sunday, November 16, 2014

Carnitine and Cardiovascular Disease

Since we ‘know’ red meat increases cardiovascular disease (CVD)* researchers have been investigating possible mechanisms.  Attention has recently turned to carnitine (and choline but this post will focus on carnitine) as the components in red meat responsible for red meat supposedly increasing CVD [1]
Essentially the argument is that carnitine and choline increase TMAO in the blood, which is associated with CVD and in high amounts of TMAO can induce CVD in genetically susceptible mice [1].  (Which sounds an awful lot like the arguments against SFA, such as SFA >> LDL-C >> CVD)
* Even though a recent (2010) systematic review and meta-analysis found that no RCTs were identified that evaluated effects of red, processed, or total meat consumption on CVD or diabetes events” and in prospective cohort studies “red meat intake was not associated with CHD” [2], which is consistent with data from the NHANES III survey [3]
What is Carnitine?
Carnitine is a compound that is found almost exclusively in meat (hence ‘carn-’) and can be synthesised by humans from methionine and lysine using vitamin C dependent enzymes.  However, vegetarians tend to have slightly lower blood carnitine levels [4]
The main function of carnitine is to transport fatty acids into the mitochondrial matrix to be metabolised*.  Carnitine also functions as an antioxidant and supports mitochondrial function.
Examine has a page on the evidence regarding the effects of carnitine supplementation and it seems carnitine is beneficial compound in a variety of contexts [5].  For example, carnitine reduces blood pressure [5], improves insulin sensitivity [5] and reduces cognitive decline in the elderly [5]
* Due to this function, carnitine has been promoted as a ‘fat burner’.  However, it doesn’t appear to increase fat oxidation unless one has inadequate carnitine [5]
Does Carnitine cause CVD?
The proposed mechanism of carnitine >> TMAO >> CVD (therefore red meat >> TMAO >> CVD) has been thoroughly critiqued by Chris Masterjohn [6].  Other people have also critiqued the study (though not as well as Chris Masterjohn), but almost all these critiques have focussed on the study design and/or the mechanisms, with very little attention paid to the real question: ‘does carnitine increase the risk of CVD?’  Fortunately we have some fairly good evidence to help us answer that question
A systematic review and meta-analysis of clinical trials (13 trials, N = 3629) was conducted in 2012 looking at the effects of carnitine supplementation on CVD events and total mortality in secondary prevention (people who have had a heart attack) [7].  It found that carnitine associated with:
·         “A significant 27% reduction in all-cause mortality
(odds ratio, 0.73; 95% CI, 0.54-0.99; P=.05; risk ratio [RR], 0.78; 95% CI, 0.60-1.00; P=.05)”
·         “A highly significant 65% reduction in ventricular arrhythmias
(RR, 0.35; 95% CI, 0.21-0.58; P<.0001)”
·         “A significant 40% reduction in the development of angina
(RR, 0.60; 95% CI, 0.50-0.72; P<.00001)”
·         “No reduction in the development of heart failure
(RR, 0.85; 95% CI, 0.67-1.09; P=.21)
or myocardial reinfarction*
(RR, 0.78; 95% CI, 0.41-1.48; P=.45)”
These results are superior to statins (see table below), and unlike statins (which increase the risk of type 2 diabetes [8] and may impair cognitive function), carnitine seems to improve cognitive function and insulin resistance [5].  Carnitine supplements are fairly cheap, so I don’t see much reason why they aren’t prescribed to people who have had a heart attack
The Number Needed to Treat (NNT)
(lower is better)
Statins [8]
Carnitine [7]
Total Mortality
Non-Fatal Heart Attack
Ventricular Arrhythmias**
This is secondary prevention so it’s unknown whether carnitine will have a similar beneficial effect in primary prevention (people who have not had a heart attack).  However, based on these results and overall effects of carnitine (see the Examine article) it’s likely that carnitine is beneficial for primary prevention as well.  At the very least, it would be surprising if something which is beneficial for secondary prevention is harmful for primary prevention
* There were only 38 myocardial reinfarctions which makes the study pretty underpowered (by comparison there were 250 deaths)
** Ventricular arrhythmias and angina would be expected to have a lower NNT (due to higher incidence) and therefore shouldn’t be compared to the statin NNT for non-fatal heart attack.  I included those two figures just to emphasise how therapeutic carnitine seems to be
The Limitation of Mechanisms
The take away from this is that biological mechanisms are just hypotheses for what happens in real world.  If we reduce most mechanisms to A >> B >> C*, then logic tells us that we can only be certain that A causes C if at least one of these conditions is true:
·         A only causes B
·         B only causes C
This is rarely the case in biology, as biological systems are highly interconnected and biologically active compounds have many effects**.  In this case, although carnitine may have some adverse effect by increasing TMAO, carnitine also has many beneficial effects, resulting in a net positive effect for CVD events and total mortality in secondary prevention.  The only way to test this is to do clinical trials
* Carnitine >> TMAO >> CVD
Saturated fat >> LDL-C >> CVD
** Also, when we get carnitine from meat we are also ingesting many other nutrients and various compounds in meat, all of which have their own biological effects.  With this in mind, reducing a given food (or food group) to one nutrient (or several) ignores the net biological effect of all the nutrients and compounds in a given food, not to mention any nutrient interactions.  A good example of this is demonising fruit by reducing it to sugar
·         There have been no clinical trials to examine the effect of red meat on CVD
·         Red meat is not associated with CVD in observational studies
·         Carnitine is a beneficial nutrient in a variety of contexts
·         Carnitine reduces CVD events and total mortality in secondary prevention
·        Just because A causes B and B causes C doesn't mean A causes C
If you haven’t already, I recommend you read the Examine article on carnitine and Chris Masterjohn’s critique