Paleolithic Nutrition: Twenty-Five Years Later

Melvin Konner and S. Boyd Eaton wrote a review article for Nutrition in Clinical Practice that update’s their seminal New England Journal of Medicine  paleo nutrition article of 1985.  They took a fresh look at recent  data on modern hunter-gatherer societies as well as advances in anthropology. 

NEJM likely has much wider circulation, so it’s too bad the update wasn’t published there.

I’ve written previously about the history of the modern paleo movement.  I consider S. Boyd Eaton and Melvin Konner two of the founders. 

I wanted to share a few tidbits from the new article:

  • The transition from hunting/gathering to farming (about 10,000 year ago) saw a decrease in body size and robustness, plus evidence of nutritional stress
  • Modern humans migrated away from Africa between 50,000 and 100,000 years ago
  • Levels of muscular and aerobic fitness in ancestral groups are much higher than modern societies, with a concomittant higher level of calorie consumption
  • Average life expectancies in pre-industrial hunter-gatherer (H-G) groups was only 30-35 years, but much of this low number simply reflects high infant and child death rates
  • H-G deaths overwhelmingly reflect infectious diseases
  • H-G groups had a high degree of dependence on plant foods
  • Cooking has been important to humans for at least 230,000 years, if not longer
  • Fish and shellfish are more important food sources than these authors thought 25 years ago
  • Ancestral H-G groups derived 35 to 65% of diet (calories, I guess) from animal flesh
  • Game animals have more mono- and polyunsaturated fatty acids than supermarket meat
  • H-G diets are higher in fat than they once thought: the new range is 20 to 35% of calories
  • H-G diet omega-6: omega-3 ratio was 2:1, in contrast to the modern Western ratio of 10:1
  • H-G groups eat more protein than they once thought
  • Carbohydrate consumption of H-G groups varied from about 35 to 65% of calories.  (I’ve written elsewhere about the carb content of paleo diets.)
  • Nearly all H-G carbs are from vegetables and fruits, which have more favorable glycemic responses than grains and concentrated sugars
  • Uncultivated fruits and veggies have much more fiber than commercial ones (13 versus 4 g fiber per 100 g of food)
  • H-G diets have at least 70 g of fiber daily
  • Sodium in H-G diets is very low: 800 mg/day

The Diet-Heart Hypothesis is the idea that dietary total and saturated fat, and cholesterol, cause or contribute to atherosclerosis (hardening of the arteries), leading to heart attacks and strokes.  Konner and Eaton still believe the theory is valid for fats, but not cholesterol.  The latest evidence is that even total and saturated fat are minimally or unrelated to atherosclerosis

They also believe total fat, due to its caloric load, is an important contributor to obesity and type 2 diabetes.  (I agree that my be true if you eat a lot of carbs with fat.)

Konner and Eaton review the very few clinical studies—a grand total of four—that apply a paleo diet to modern humans.  Results support their theory that paleo-style eating has healthful metabolic effects.

Their conception of a paleo food pyramid is a base of high-fiber vegetables and fruits, the next tier up being meat/fish/low-fat dairy (!?) (all lean), then a possible tier for whole grain (admittedly very unusual), with a small peak of oils, fats, and refined carbohydrates.

They recommend high activity levels, including resistance exercise, flexibility, and aerobics, burning over 1,000 cals/day exclusive of resting metabolism.

They also seem to favor small amounts of alcohol—not generally considered paleo—to reduce heart disease risk, admitting that “…the HG model cannot answer all questions.”

Eaton and Konner suggest the following as the “estimated ancestral diet“:

  • Carbohydrates, % daily energy                 35-40
  • Protein, % daily energy                                 25-30
  • Fat, % daily energy                                          20-35
  • Added sugar, % daily energy                            2
  • Fiber, g/day                                                        >70
  • EPA and DHA*, g/day                                    0.7-6
  • Cholesterol, mg/day                                       500+
  • Vitamin C, mg/day                                           500
  • Vitamin D, IU/day                                4,000 (sunlight)
  • Calcium, mg/day                                       1,000-1,500
  • Sodium, mg/day                                         under 1,000
  • Potassium, mg/day                                         7,000

*Eicosapentaenoic acid and docosahexaenoic acid

Konner and Eaton call for more randomized controlled studies of the paleo diet.  These studies will need to define the paleo diet carefully.  Their definition is probably as good as any, if not the best.  Of course, the listed nutrients should come from minimally processed, natural foods.  We just need Loren Cordain‘s and Staffan Lindeberg‘s  input on a concensus definition of the paleo diet, then we’re ready to rock’n’roll.

Steve Parker, M.D.


Konner, Melvin and Eaton, S. Boyd.  Paleolithic Nutrition: Twenty-Five Years LaterNutrition in Clinical Practice, 25 (2010): 594-602.  doi: 10.1177/0884533610385702

Eaton, S.B. and Konner, M.  Paleolithic nutrition: a consideration of its nature and current implicationsNew England Journal of Medicine, 312 (1985): 283-289.


16 responses to “Paleolithic Nutrition: Twenty-Five Years Later

  1. What about seasonal variation? Winter, Spring, Summer, Fall diets we different even in my rural youth.
    What about location variation? Inland adjacent to lakes; fish, else no fish.
    There is no one Paleo Diet. Whole real food, some fresh, some less so, what ever they could find, and the survivors reproduced. The remainder were part of natures cull. But what do I know.

  2. The extremely high fiber intake is problematic. The thing is, lets say they get 30-40% energy as carbs from fibrous tubers like yams, fruits, root veggies and greens… This would probably mean their getting anywhere from 100-150 grams of carbs, including fiber. I am willing to be that including fiber the number is at the higher end, hell it could be 200 including fiber. But… it takes about 3lbs of sweet potato to get to simply 40 grams of fiber… and that right there takes you up to 330 grams of carb (including fiber). You can add in some root veggies, fruits, greens, this will probably get you up to 60 grams total in fiber but now your looking at close to 400 grams of carb including fiber.

    The potassium recommendations are similar. The highest potassium foods that we know of are yams, potatoes, and fruits like papaya and coconut water… this are all incredibly carby. 3lbs of sweet potatoes again will get you to about 5700mg of potassium… to get to 7000mg we still need more fruits and veggies… which puts us even higher in the carb ranges.

    Either these HG’s ate waaay better plants than we did, or this data just isn;t working out. Honestly, it looks incredibly realistic to me (the macro breakdown), so I support where there going with this. A higher carb intake to reflect the high plant intake seems more accurate that the VLC ZC extremely high meat and fat diets that many Paleo’s endorse. But the fiber and potassium recommendations are just not possible (with our neolithic foods I guess) with a carb range even as moderate as 30-40% of calories… it would have to be much higher, say 50-70%. Further, if they did have better crops…. then how can we work around this? The other idea is that maybe they ate upwards in the 3000 of calories a day, and thus these numbers might work out with the calculated macros… but most studies seem to show general caloric restriction in many HG societies.

    I am paleo myself but I am eager to see what is done about some of these puzzle pieces that don’t quite fit.

    • Thanks for your input, Bill.
      Eaton and Konner have estimated that our Paleolithic ancestors indeed ate in the 3000 calories a day range.
      The high carb consumption you note might be problematic for diabetics, depending on residual beta cell function, diabetic drugs, other macronutrients in the meal, etc. Extremely high fiber consumption, on the other hand, would tend to counteract the blood-sugar-raising effect of the digestible carbs.
      I’m thinking about the clinical paleo diet studies of Staffan Lindeberg et al: I doubt fiber intake was even above 50 g/day.


  3. Steve,

    Makes sense. I really dig the objectivity of your blog.

  4. Steve
    Thank you for bringing this Update/Revision of Paleo H-G nutrition to my attention.

    It feels much less “ideological” than the conclusions drawn by Cordain from the Eaton papers of the 80’s.

    Arithmetically the macro proportions centre on 1/3 1/3 1/3 which ought to be regarded as “moderate” as no macro is either “high” or “low”.

    Whenever I make a “stew”, I seem to arrive (roughly) at these proportions and let it blend via hours of slow cooking.

    Slainte, Desmondo
    P S I do not have hyperglycemia.

  5. Pingback: Random Thoughts On Paleo Eating For People With Diabetes | Paleo Diabetic

  6. Pingback: Random Thoughts On Paleo Eating For People With Diabetes | Paleo Diabetic

  7. Pingback: Random Thoughts On Paleo Eating for People With Diabetes | Advanced Mediterranean Life

  8. Pingback: Random Thoughts On Paleo Eating for People With Diabetes | Diabetic Mediterranean Diet

  9. Pingback: Denmark’s Paleolithic Diet | Paleo Diabetic

  10. Pingback: More Potassium and Less Sodium May Reduce Cardiovascular Disease and Prolong Life | Advanced Mediterranean Life

  11. Pingback: New Cochrane Review Supports Population-Wide Sodium Restriction | Paleo Diabetic

  12. Pingback: Scientific Reviews Support the Paleo Prescription for Potassium and Sodium | Paleo Diabetic