A few thoughts on the paleo diet


As of September 2017, new Sweat Science columns are being published at www.outsideonline.com/sweatscience. Check out my bestselling new book on the science of endurance, ENDURE: Mind, Body, and the Curiously Elastic Limits of Human Performance, published in February 2018 with a foreword by Malcolm Gladwell.

- Alex Hutchinson (@sweatscience)


I was asked in an interview a few days ago for my take on the “paleo diet,” and I figured I may as well share those thoughts here. I’ll start by saying that I’m not an expert in this area — these are just my impressions from the outside! For anyone who’s interested in the scientific rationale behind it, there’s a very comprehensive review paper that was published earlier this year and is freely available online. Anyway, a few scattered thoughts:

It’s not a diet, it’s a lifestyle. I’ve seen this sentiment expressed on a number of paleo-oriented blogs, and I think it’s a very important point. If you want to argue that humans are uniquely adapted to the paleolithic environment because that’s where we spent the most time, it’s meaningless to just consider one part of that environment. If you spend the day sitting on your couch watching TV, then picking up the phone and ordering an authentic ancestral meal from McPaleo’s isn’t going to make you healthy. The review paper focuses on the following key elements of the paleolithic environment:

  • regular sun exposure for vitamin D
  • plenty of sleep, in synch with light/dark cycles
  • lots of physical activity!
  • no exposure to pollution
  • fresh, unprocessed food
  • short bouts of acute stress (tiger!) rather than chronic stress

All of this stuff sounds great — I’m absolutely in favour of every element of this lifestyle.

Plants vs. animals. In the fantasies of some people, going paleo means you get to eat enormous Fred Flintstone-style chunks of meat — for every meal. Not quite: here’s a passage from a paper published in the British Journal of Nutrition last year:

[I]n contrast to common belief, hunting probably played a less dominant role from a nutritional point of view compared with gathering, and on average, it makes up 35% of the subsistence base for present-day worldwide hunter–gatherers, independent of latitude or environment.

This is a point picked up by David Katz in an article earlier this summer: you’re still going to eat, as Michael Pollan would say, “mostly plants.”

The evils of wheat and dairy, and the pace of evolution. Okay, this is where I believe we start to drift away from well-supported science and into the realm of unsupported hypotheses. The basic idea is that, since humans only started farming about 11,000 years ago, our genome hasn’t had time to adapt to these foods. Moreoever, grains like wheat actually contain “antinutrients” that hinder proper digestion and cause chronic inflammation — in everyone, not just those with celiac disease or gluten sensitivity. These results are not widely accepted — or at least, I personally don’t find the evidence convincing.

Moreover, 11,000 years — or 366 human generations — is actually quite a long time. As a result, for example, it’s well understood that the gene allowing for humans to digest milk was selected through evolutionary pressure in populations that domesticated cows. The review paper I mentioned above notes this as a “key exception” to what they otherwise claim is the rule that humans haven’t had time to adapt to agriculture. I, on the other hand, would view it as “key evidence” that humans have had time to adapt to agriculture. Obviously, not all modern humans can digest milk — and those who can’t shouldn’t drink it! But I see no evidence that those who can drink milk should avoid it. Same goes for wheat: it’s certainly true that some people can’t process it adequately, but I’m not convinced that it’s full of antinutrients that are secretly poisoning the rest of us.

Overall, as Stephan Guyenet pointed out in his discussion of the review paper, the evidence seems to support the idea that “the main detrimental change was not the adoption of agriculture, but the more recent industrialization of the food system.” In other words, the diet we should be seeking to emulate is pre-1850, not pre-10,000 BC — which, not coincidentally, once again sounds a lot like Michael Pollan’s advice: don’t eat anything your grandparents wouldn’t recognize as food.

28 Replies to “A few thoughts on the paleo diet”

  1. Well summarized Alex. I think the arguments that go back and forth between paleo and vegetarian folks (essentially, very pro-meat and anti-meat) miss the point that it’s not the inclusion of meat in the diet that accounts for the majority of good human health – it’s the absence of highly processed foods. For the most part, you can get 90% of the way to great health just by avoiding processed food. Pollan’s book “In Defense of Food” really hits home in this regard.

  2. I have largely adapted a paleo diet over the past month, although had been leaning that way for about 6 months. Since cutting wheat and grain products, and increasing my vegetable intake I have a lot more energy. I’ve been able to cut my caffeine intake 25% and I don’t notice blood sugar swings in the afternoon (no pronounced 2 PM drop off.) I also stay full longer.

    Food quality is important. I used to eat cereal for breakfast everyday and have switched to eating eggs and greens, or meat and greens. It makes a huge difference.

  3. A couple of thoughts:

    Mass dental problems seem to have emerged after the end of the paleolithic, not the advent of industrial food (which did indeed cause dental problems to increase, as well as an associated increase in infectious diseases like TB.

    Extrapolating diets from modern hunter-gatherers in problematic, because modern HGs live in fringe environments. If you look at what a paleolithic style human ate when living in a prime environment, it tended to be more meat-centric. This is confirmed by the disappearance of animals whereever humans migrated to around the globe. We weren’t looking for plants.

    Plants are good, but a plant-centric diet is plan B for a human. Comparing our digestive tract to our primate relatives who do eat plant-centric diets makes that clear. As does the experience of the subcontinental Indians, who have a plant-centric diet in large part, and have been suffering from “industrial” diseases like diabetes for 2,000 years…

    Wheat as a problem in the human diet goes back 2,000 years to the ancient Greeks. And as Dr. Davis details in his book “Wheat Belly”, the wheat you eat today did not exist 50 years ago, let alone 500 or 5,000, it’s a novel plant, bred and genetically modified by humans, and contains proteins that traditional wheat did not. The rapid increase in celiac disease is indicative of something changing in wheat.

    I agree with you about dairy. The paleo anti-dairy argument doesn’t hold up when we’re discussing traditional dairy consumption, and modern dairy intolerance seems to be a by-product of consuming modern wheat.

    So going back 150 years or so is definitely a huge improvement in human diet, but that still leaves you with wheat flour and sugar, which were causing disease even then. Ditch those two, and any foods “invented” after then, and you’re on to something.

  4. Egads, I hate the “don’t eat anything your grandparents wouldn’t recognize” meme. Half of their siblings were dead. What did they eat?

  5. Thanks Alex…it is always nice to hear a moderate approach advocated for those of us who strive to eat healthy, balanced meals but don’t feel the need to omit major food groups for no reason! I think the “1850” type approach is great – get ready of packaged stuff and eat real food. Only problem is, where do sports drinks / bars / gels etc come in for people doing long endurance events? May need to make an exception for events since most food is unrealistic during races.

  6. The paleo diet assumes that human development was linear, and began very (relatively) recently. For an interesting opposing view, see The Yugas (http://www.crystalclarity.com/product.php?code=BTY). If the world does, as the authors claim (with substantial evidence) pass through 24,000-year cycles of ascending and descending civilization, then 11,000 years ago we may not have needed agriculture. This is, in fact, the claim of the sages of ancient India. But, you stopped reading long ago…

  7. Glad to hear you weigh in on this. While I think that “eating paleo” beats the hell out of the standard American diet, a diet of bug spray and Skittles probably does too (credit to Lyle McDonald, if I recall, for the words and sentiment there).

  8. Thanks for all the comments, folks!


    “If you look at what a paleolithic style human ate when living in a prime environment, it tended to be more meat-centric. This is confirmed by the disappearance of animals whereever humans migrated to around the globe.”

    “Plants are good, but a plant-centric diet is plan B for a human. Comparing our digestive tract to our primate relatives who do eat plant-centric diets makes that clear.”

    To be perfectly honest, I don’t find this convincing. Why does the disappearance of animals imply that humans weren’t still eating lots of plants? What do the digestive tracts of herbivores tell us about the dietary composition of omnivores?

    That fact is, we don’t know exactly what paleolithic humans ate. The BJN paper I cited above, which was written by paleo advocates including Loren Cordain, notes “the absence of accurate data on human nutritional (plant/animal en%/en%) subsistence ratios in the Paleolithic diets.” It explicitly argues that hunting likely played a less dominant role than gathering. If anything, the wide range of observed hunter-gatherer plant-animal ratios (from 30-70 to 70-30, roughly) tells us that there isn’t a single “natural” diet that defines us as human.

    Another prominent paleo researcher, Boyd Eaton, argues that our “ancestral diet” took about 50% of its calories from fruits and vegetables (compared to 16% among modern Americans). Considering how much more energy-dense meat is compared to plants, that means they would indeed have been eating “mostly plants” — including 100 grams a day of fibre! But this isn’t the message we hear in some paleo circles: it’s more like “We have to eat exactly like our ancestors ate, but only when those dietary decisions correspond to choices I wanted to make anyway.”

  9. @Malindi! Great point. Evolution did lots of great stuff for us, but in the modern world we give meaning to tasks that are totally unrelated to whether a caveman would starve to death or live to get laid. If you want to win a marathon (or, more generally, run a marathon as fast as you’re capable of), you have to ingest easily digestible carbohydrates at a rate that no caveman would have dreamed of. And similarly, if you want to be a concert pianist, you have to spend more hours sitting down than any caveman would have — and so on…

    Fortunately, I don’t think drinking sports drinks and taking gels and so on in the context of endurance activity is a problem — you’re just providing the fuel your body needs to perform that task. The problems arise when you give it that super-fuel without burning it, like drinking sports drinks just because you’re thirsty on a hot day.

    P.S. Great to see that you’re headed to Pan Ams!

  10. I’ve been playing with the paleo diet for 7-8 mo (after trying the Thrive diet for 6 months) and the shift for me has been quite dramatic. I definitely had some issues with wheat and as a runner and a vegetarian it had been a major component of my diet. On the flip side my wife just tried it for 6 weeks and only noticed minor improvements. We are definitely all individuals and have varying degrees of food tolerance. I’ve recently been reading some paleo advocates referring to paleo 2.0 (e.g. Dr. Kurt Harris), which reduces the emphasis on macro-nutrient ratios(I’ve always ate a higher carb version than most advocates recommend). The 2.0 version recognizes that the macro-nutrient ratios vary significantly between different present-day hunter gathers, and depending on activity level and goals (losing vs. maintaining weight), the carb-protein-fat ratio should shift for us too. Some grains, like rice and buckwheat, are OK as well. For me the biggest health improvements (I think) have been realized by eliminating wheat and increasing vegetable and good fat intake.

  11. If you take as a model the diet of the Hazda people in East Africa, who are assumed to have retained our ancestral hunter gatherer lifestyle to a very large degree, you’d have about 180 grammes of meat a day, but subject to large seasonal variation. Elephant steaks when available, but sometimes no meat for months.

    You would have to include scavenged meat in your diet, about 25 grammes a day on average. It should be noted that the Hazda have hunting weapons. If you take hunting to be a ‘modern’ development, the share of scavenged meat in a truly paleo diet should probably be much higher.

    (numbers from http://www.clas.ufl.edu/users/sdef/ANT3428/readings_spring_2008/Hadza_scavenging_O%27Connell_et_al._1988.pdf)

  12. @alex

    “What do the digestive tracts of herbivores tell us about the dietary composition of omnivores?”

    In primates, the dietary consumption of plants corresponds to the size of the colon. Bigger colon, more fermentation of fiber into short-chain fatty acids, which are a primary fuel source of herbivores.

    Apes, which do eat mostly plants, have a much bigger colon than chimpanzees, which eat less, and humans, which eat the least. This is pretty basic science…

    “The Human Colon in Evolution: Part 1, comparative anatomy”

    This explains why a high-plant diet results in malnutrition pretty reliably in humans. The less meat included, the more likely malnutrition.

    I’m not saying, btw, that we didn’t eat plants, but based on our physiology, they cannot be a primary source of nutrition. We cannot digest them well enough.

  13. The big objection to the Paleo diet that is never articulated well is that evolution through natural selection does not optimize a species’ genes for longevity and lifetime health, it optimizes them for reproductive fitness – and these two variables are not necessarily correlated.

    Even if it is the case that humans are most adapted to a paleolithic environment, it does not follow that this type of environment maximizes lifespan or health – it simply means that humans that adapted best to that type of environment were able to reproduce more successfully than their counterparts.

    Further, because reproductive success is the one single variable that evolution maximizes in a species, it would surprising if another variable (like longevity or health) that is not directly correlated with it, would be maximized as well. In other words, there is most likely “room for improvement” in every other quantitative variable in a species’ genome besides reproductive success.

    Whether reproductive success correlates with improvements in lifespan and health is another debate, but evolutionary theory says nothing about it.

  14. @Tuck “This is pretty basic science…”

    While I appreciate the tip, we clearly have different views of what constitutes science. You linked to someone’s homework assignment that summarizes the well-known hypothesis that eating meat allowed our ancestors to shift metabolic resources from the digestive system to the brain. This leads us to the staggeringly obvious conclusion that our ancestors ate meat — which no one remotely disagrees with. It doesn’t tell us anything about how much plant material they were also eating at the same time.

    “Apes, which do eat mostly plants, have a much bigger colon than chimpanzees…” I’m a little confused here, because chimpanzees ARE apes. But if you meant to say gorilla instead of ape, it still disagrees with the data in the link you provided, which says the colon takes up 53% of gut volume in plant-eating gorillas and 52% in omnivorous chimpanzees — hardly the evidence you claim that colon size correlates with proportion of plants in diet.

    The question we’re debating isn’t whether human ancestors ate meat, which is obvious. But even the source you linked to, in its next section, acknowledges that our ancestors likely ate up to 100 grams a day of fibre. So humans could certainly have gotten an evolutionary boost by incorporating meat in their diet, while still eating a roughly 50-50 caloric balance of meat and plants, as argued in the Eaton paper I linked to in my previous post. Again, 50-50 in calories means “mostly plants.” But you’re somehow arguing that this data proves that “based on our physiology, [plants] cannot be a primary source of nutrition.”

    As I said above, all of these hypotheses are based on very limited data. If I were to give you links to scientists — including people like Richard Leakey — looking at this same evidence and arguing that comparative anatomy and paleoanthropology “prove” that humans evolved to eat plants and nothing but plants, you’d probably say that those arguments reflected a simplistic, agenda-driven, and selective reading of the evidence. I agree, and I find that approach equally unconvincing on both extremes.

  15. @Aaron: Hope all’s well! So in going paleo (2.0), have you started eating meat? If so, isn’t that the biggest change? And if not, is it really paleo?

    @RH: Interesting stuff — but as you point out, no matter how “perfectly preserved” the Hadza diet is, it’s still just a single snapshot of a group in a particular ecological and geographical niche. In the BJN paper I linked to, they talk about “selective and non-selective savannah, savannah/aquatic and aquatic hunter-gatherer/scavenger foraging strategies” even within the context of the East African Paleolithic diet. That’s why I think it’s a mistake to get too dogmatic about the specific ratios or foods a particular sub-group of people might have been eating 500,000 years ago. Why is 3 million years ago too early to matter? Why is 10,000 years ago too late to matter?

  16. @Tom: This is a very interesting point. The big review paper I linked to in the post makes basically the same point in its introduction:

    “The impetus for these genetic changes was not to increase longevity and resistance to chronic degenerative diseases but rather to increase the probability of survival and reproductive success. Occasionally, mutations that had positive survival and reproductive value sometimes also caused adverse health effects in the postreproductive years.”

    So it makes this point — but then, as far as I can tell, never acknowledges it again or offers any counterargument.

  17. @alex
    Totally agree. If anything characterizes the human diet, it is variability and adaptability.

    I brought up the scavenging stuff to show that a paleo diet, if it resembles our ancestral diet at all (as you point out, if there is one such a thing), it is a tidied up, 21 century adaptation of it. Fresh hygienic steaks instead of random body parts that the hyenas left over a couple of days ago. A steady fixed portion every day instead of alternating between abundance and scarcity.

  18. @alex

    You seem to have done a fabulous point of missing the point of Melissa’s “homework assignment”, which I linked to because she does a pretty good job of summarizing the argument.

    In short, this sentence: “The major difference in this matter between humans and the other great apes is that apes such as the gorilla are able to use their larger colons to obtain as much as 60% of their caloric intake from SCFA alone (Popovich et al., 1997). Upper estimates for human caloric use of SCFA range from seven to nine percent. (McNeil, 1984).”

    So we we’re simply not capable of getting a large portion of our calories from fiber. And as any vegan can explain, attempting to do so requires spending your entire day chewing. We had some evolutionary cousins who did that, and they looked more like gorillas than humans. They lacked the hunting adaptations that humans had, and had the large chewing muscles that gorillas have.

    It doesn’t really matter if we know exactly what our ancestors ate, if they would not have been able to utilize it.

    BTW, since when is 50% “mostly”? since we’re arguing definitions…

    And I’m not advocating a Fred Flintstone diet here, as you’re attempting to paint it. Even primarily meat-eating societies like the Inuit ate vegetables as an important part of the diet, so an all-meat diet is likely unhealthy at some point.

  19. Any diet that rejects all the riches of agriculture–which, of course, is the basis for nearly ALL human culture (like religion, architecture, art, etc.) seems wrongheaded to me from the start. And the tendency of Paleo advocates to compare Paleo to processed is just plain silly.

  20. OMG! If you are going to take the paleo-lifestyle, please be serious, leave society, stop drinking the water, using the electricity, using cars or transportation of any sort that you did not build yourself from a tree and a few stones, etc. I see this as the typical insanity of yearning for a better time that never used to exist. No timeframe in history was magical. If you are 30 or over you’d be dead by now most likely in the paleo world. There exists basically no food today that existed then anyway – corn is different, apples are different, anything farmed has been genetically changed, and most “wild” fruits have been affected and changed by agriculture as well.

    Nutrient-enriched food is one of the leading reasons we live so long these days, along w/ modern medicine and dentistry (@Tuck: where did you get your dental information? Most remains from thousands of years ago show massive dental problems). For that matter, if you want to live a paleo lifestyle don’t use the medical or dental system.

    Yes, pollution seems to cause some problems as does repetitive tasks brought on by many modern jobs (I maintain that long-term stress has always been part of our lifestyle – I’m guessing night time, every night, was a stressful thing in the paleo world). You can find plusses and minuses in any era. But for the most part it is irrefutable that modern life extends lifespan greatly.

    And it definitely makes you a better athlete. Besides the fact that MODERN training methods continue to make better athletes through the use of MODERN technology, even the basics of food are much better: I eat cereal everyday. It is processed, but stock full of vitamins and minerals. I have it with milk that is processed using pasteurization among other processing methods. It is also advocated by USA Swimming to its athletes (among many other national sports organizations from many countries) as possibly the best single food you can eat (no, I’m not saying they advocate only eating cereal). Processed food is not evil – spiking your blood sugar beyond what your body is using repetitively and avoiding vitamins, nutrients, and fiber is probably not the best plan to feel healthy and to be strong and agile later in life, but that is a different thing than avoiding modern food (which, as I said, is actually impossible).

    The argument that we didn’t evolve for this or that is preposterous. So what if we didn’t evolve for it. We didn’t evolve to sleep in a bed, or in a house, to be warm and safe at night, to use medicine or modern transportation, and certainly not to go to school and learn to read and write and use our brain to do advanced mathematics and physics and write poetry.

    This reminds me of raw-fooders. After all, controlling fire for cooking food is a technology we once mastered (and of course there are some carcinogens from cooking food too). So my question for paleo-ists is: How far do you want to go back? To when life expectancy was 40, 30, or 20?

  21. @alex All is well. Love the blog btw… I started with the Thrive Diet, which is basically the vegan version of the Paleo diet (from an exclusion perspective) and I probably noticed 70-80% of the benefits from eliminating foods through following this diet. After being a vegetarian/vegan for 17 years adding meat has made eating such a restrictive diet significantly more satisfying and simpler, but the benefits (for me) are relatively few compared to eliminating grains. If I could go back in time and choose to either eat meat or exclude grains (specifically wheat) during my heavy training years, I would eliminate wheat hands down.

  22. @Rachel Stone
    religion? in an argument pro-agriculture and anti-paleo? really?

    When arguing (be it pro or con), people should really try to avoid fallacies. Taking part of an argument to extrapolate and/or prove it wrong while ignoring the other half isn’t quite the best way to go at it, for instance.

  23. “it makes up 35% of the subsistence base for present-day worldwide hunter–gatherers”

    Bzzzt. Wait … that doens’t mean H. sapiens only ate 35% meat 200K years ago. Today, the best land for humans and game are used by humans for cities and suburbs and farms. Modern hunter gatherers are paleo on marginal land and have over hunted their range.

    Face it vegans: Humans were designed to run down and eat antelope.

  24. Let us no forget that without the agricultural development of wheat, rice, corn and most importantly the introduction of the potato into Europe, we would not likely be having this conversation. We would be concerned where our next meal was coming from. I for one, feel ashamed that knowing the impact on agriculture due to meat production, we are even discussing a meat intensive diet while such a large population in our own country let alone the rest of the world go to bed hungry. What is the real impact on longevity and health? Not just the same old game of chasing numbers for more needless blood tests. When we have conquered hunger, and discussed birth control first, we can begin the discussion of diet without the guilt we should all have knowing there are those dying for lack of calories in the world. Perhaps we live too long for our species own good. Send the children outside to play and maybe join them; the key is how we live, not how we eat, for most of us.

Comments are closed.