Paleolithic Nutrition Twenty-Five Years Later

Abstract

A quarter century has passed since the first publication of the evolutionary discordance hypothesis, according to which departures from the nutrition and activity patterns of our hunter-gatherer ancestors have contributed greatly and in specifically definable ways to the endemic chronic diseases of modern civilization. Refinements of the model have changed it in some respects, but anthropological evidence continues to indicate that ancestral human diets prevalent during our evolution were characterized by much lower levels of refined carbohydrates and sodium, much higher levels of fiber and protein, and comparable levels of fat (primarily unsaturated fat) and cholesterol. Physical activity levels were also much higher than current levels, resulting in higher energy throughput. We said at the outset that such evidence could only suggest testable hypotheses and that recommendations must ultimately rest on more conventional epidemiological, clinical, and laboratory studies. Such studies have multiplied and have supported many aspects of our model, to the extent that in some respects, official recommendations today have targets closer to those prevalent among hunter-gatherers than did comparable recommendations 25 years ago. Furthermore, doubts have been raised about the necessity for very low levels of protein, fat, and cholesterol intake common in official recommendations. Most impressively, randomized controlled trials have begun to confirm the value of hunter-gatherer diets in some high-risk groups, even as compared with routinely recommended diets. Much more research needs to be done, but the past quarter century has proven the interest and heuristic value, if not yet the ultimate validity, of the model.

Full text

594
Nutrition in Clinical Practice
Volume 25 Number 6
December 2010 594-602
© 2010 American Society for
Parenteral and Enteral Nutrition
10.1177/0884533610385702
http://ncp.sagepub.com
hosted at
http://online.sagepub.com
Just over 25 years ago, an unusual article, “Paleolithic
Nutrition: A Consideration of Its Nature and Current
Implications,” was published in a respected journal.1
In it, we described a new paradigm for prevention based
on very old human experience: nutrition during the course
of human evolution. Drawing on modern studies of hunter-
gatherers (HGs) and also on archeological and paleonto-
logical evidence, we argued for the discordance hypothesis,
which in its simplest form states that our genome evolved
to adapt to conditions that no longer exist (the environment
of evolutionary adaptedness, or EEA), that the change has
occurred too rapidly for adequate genetic adaptation, and
that the resulting mismatch helps to cause some common
chronic diseases.
Among these “diseases of civilization” are atheroscle-
rotic cardiovascular disease (most coronary artery disease
and cerebrovascular accidents), type 2 diabetes mellitus
(T2DM), chronic obstructive pulmonary disease, lung and
From the 1Department of Anthropology and Program in Neuro-
science and Behavioral Biology; and 2Departments of Radiology
and Anthropology, Emory University, Atlanta, Georgia.
Address correspondence to: Melvin Konner, MD, PhD, 1 Dep-
artment of Anthropology and Program in Neuroscience and
Behavioral Biology, Emory University, 1557 Dickey Drive, Atlanta,
GA 30306; e-mail: antmk@mindspring.com.
colon cancers, essential hypertension, obesity, diverticu-
losis, and dental caries. In another study, we to showed
that serum cholesterol concentrations, aerobic fitness,
and diabetes mellitus prevalence in nonindustrial, espe-
cially HG populations, revealed low risk of the diseases
that most plague advanced societies.2 Indeed, by the time
of our first publication, it had been shown that former
HGs in Australia who were suffering from T2DM showed
marked improvement in their carbohydrate and lipid
metabolism when they were experimentally returned to
their former lifestyle.3 Also by that time, archeologists
working around the world had shown a decrease in body
size and robusticity and an increase in markers of nutrition
stress during the transition between hunting and gather-
ing and agriculture,4 suggesting that some aspects of the
discordance began as long as 10,000 years ago.
The general criticism that the mismatch has been res-
olved by evolutionary adaptation since the HG era has not
proved convincing. It is true that since modern humans
left Africa between 100,000 and 50,000 years ago, genetic
evolution during subsequent millennia has continued—
for example, pigmentation changes (hair, eyes, skin), intes-
tinal lactase retention beyond infancy, and adaptive
defenses against microorganisms (such as hemoglobin-
opathies and immune system modifications). New ana-
lytic methods are revealing subtler genetic adaptations to
A quarter century has passed since the first publication of the
evolutionary discordance hypothesis, according to which depar-
tures from the nutrition and activity patterns of our hunter-gath-
erer ancestors have contributed greatly and in specifically definable
ways to the endemic chronic diseases of modern civilization.
Refinements of the model have changed it in some respects, but
anthropological evidence continues to indicate that ancestral
human diets prevalent during our evolution were characterized by
much lower levels of refined carbohydrates and sodium, much
higher levels of fiber and protein, and comparable levels of fat
(primarily unsaturated fat) and cholesterol. Physical activity levels
were also much higher than current levels, resulting in higher
energy throughput. We said at the outset that such evidence could
only suggest testable hypotheses and that recommendations must
ultimately rest on more conventional epidemiological, clinical,
and laboratory studies. Such studies have multiplied and have
supported many aspects of our model, to the extent that in some
respects, official recommendations today have targets closer to
those prevalent among hunter-gatherers than did comparable
recommendations 25 years ago. Furthermore, doubts have been
raised about the necessity for very low levels of protein, fat, and
cholesterol intake common in official recommendations. Most
impressively, randomized controlled trials have begun to confirm
the value of hunter-gatherer diets in some high-risk groups, even
as compared with routinely recommended diets. Much more
research needs to be done, but the past quarter century has
proven the interest and heuristic value, if not yet the ultimate
validity, of the model. (Nutr Clin Pract. 2010;25:594-602)
Keywords: Paleolithic diet; hunter-gatherers; ancestral diet
Paleolithic Nutrition
Twenty-Five Years Later
Melvin Konner, MD, PhD1; and S. Boyd Eaton, MD2
Financial disclosure: none declared.
Invited Review
by guest on February 14, 2011ncp.sagepub.comDownloaded from
Posted by nutritional anthropologist Geoff Bond
www.naturaleater.com

Paleolithic Nutrition / Konner, Eaton 595
dietary and other ecological niches, including different
allele frequencies associated with dependence on cereal
grains as opposed to roots and tubers.5,6 However, the
importance of these differences is not clear, but we know
that core biochemical and physiological processes have
been preserved.7 Furthermore, it is now widely agreed
that hundreds of thousands of preventable deaths in the
United States alone are attributable to dietary and other
lifestyle risk factors similar to those in which advanced
countries differ from HGs.8 No one proposes that genetic
adaptations could have caught up with dietary and life-
style changes over the past 2 centuries.
On the basis of published data on the nutrient con-
tent of both meat and plant foods consumed by HGs
around the world, together with anthropological data on
the composition of HG diets, we put forward a model
consisting of estimates of macro- and some micronutri-
ents consumed by HGs and argued that this model was a
reasonable approximation of the nutrient composition of
the typical diet of our ancestors during the early history
and evolution of our species. We then compared these
estimates to published data on the average American diet
in the mid-1980s as well as to then-current recommenda-
tions from relevant government institutions and other
health authorities. It was also apparent that calorie output
and throughput were much higher in hunter-gatherers
than in the United States and similar societies and that
levels of both aerobic and muscular fitness in HGs were
maintained through most of life by their patterns of activ-
ity. Although beyond the scope of this article, our estimates
of HG activity levels strongly suggested a need for large
increases in all forms of exercise in modern populations.
Our papers and book9 were greeted with a certain
amount of media attention, including ridicule, some of it
based on the short life span of hunter-gatherers. This point
had of course not been lost on us; as we had shown in
extensive reviews of the primary literature,1,2 30 to 35 years
was roughly the average life expectancy at birth of prein-
dustrial populations generally, due mainly not to the abs-
ence of older people but to extremely high infant and child
mortality. Death in HGs was overwhelmingly due to inf-
ectious diseases we now control, and older HGs rarely got
or died of coronary artery disease, diabetes mellitus, or
chronic obstructive pulmonary disease, among other ail-
ments common in societies like ours. We had not proposed
that they were healthier in absolute terms, just that absent
infection and osteoarthritis, they rarely had the chronic
diseases we commonly have. Our review of various health
measures in HG and other nonindustrial populations showed
that average HG serum total cholesterol was always below
135 mg/dL, aerobic fitness of average men was in the ath-
letic range for Western populations, and diabetes mellitus
prevalence was very low.2
In the mid-1980s, the standard recommendations,
based on clinical and experimental research, were urging
Americans in most cases to change their diet in a direc-
tion consistent with the HG model. However, there were
important differences. Intakes of saturated fat, salt, and
refined carbohydrate levels were markedly lower in HGs
than in the standard recommendations, whereas protein
and fiber content were far higher. Cholesterol intake was
also higher; both cholesterol and carbohydrate intake
were roughly the same in HGs as in the average American
diet, although the spectrum of carbohydrates was very
different. We did not then and do not now propose that
Americans adopt a particular diet and lifestyle on the basis
of anthropological evidence alone; formal recommenda-
tions must rest on carefully executed laboratory, clinical,
and epidemiological studies. Rather, we suggested that
the standard recommendations of the time needed more
research in light of the HG model. Here, we assess how
that model has fared in relation to further analysis of HG
diets, both in the archeological/paleontological record and
in studies of recent living HG groups, as well as in com-
parison with the standard recommendations (then and
now) in light of a quarter century of further research.
Hunter-Gatherer Diets: How
Well Did We Characterize Them?
Some analyses in the past decade have suggested that we
underestimated the proportion of meat in HG diets.10-12
This is of substantial potential importance in estimating
the intake of total fat, protein, carbohydrate, and fiber in
those diets. However, this position has not gone unchal-
lenged.13 It is clear that ancestral human diets derive
from higher primate diets that were overwhelmingly plant
based,14 supplemented by insects and (in some species) a
small amount of animal flesh. Fossil evidence shows that
this pattern continued to be true of early bipedal homi-
nids (between 6 and 2 million years ago [mya]),15 with a
likely particular emphasis on underground storage organs
(USOs; tubers)16 and on large protected nuts and seeds.17
Reliance on animal flesh increased substantially after 2 mya
with the evolution of Homo habilis and especially Homo
erectus, a species clearly capable of hunting large game,
an ability shared by modern humans. However, much evi-
dence points to continued significant (if not predominant)
dependence on plant foods.
Consider the human gut. It is substantially smaller
than the value predicted from the primate regression of
gut on body weight (indeed, it is almost gram for gram
reduced in proportion to excess human brain weight).18
This is due in part to concentration of calories, both in
plant foods (fruit and nuts as opposed to leaves and grass)
and, later, meat, but it is clear that cooking played a key
role beginning at least 0.23 mya and perhaps much ear-
lier, reducing the need for human digestion of both plant
foods and animal flesh.19,20 However, the human gut retains
by guest on February 14, 2011ncp.sagepub.comDownloaded from

596 Nutrition in Clinical Practice / Vol. 25, No. 6, December 2010
many structural and metabolic features of the herbivore/
frugivore higher primate gut, departing in important ways
from the typical gut of top carnivores.13,21
Furthermore, the archeological/paleontological record
makes it clear that ancestral populations relevant to the
origin of our species (anatomically modern Homo sapiens)
relied on a variety of food sources in significantly varying
environments; indeed, flexibility in adaptation may have
been central to human evolution,22 and we now speak of
EEAs rather than a single EEA. This range undoubtedly
included substantial reliance on plant foods in many times
and places.13,23,24 To support a very long human childhood,
a unique human pattern of postweaning provisioning evo-
lved,25 including contributions of animal flesh from fathers26
and of plant foods from mothers and grandmothers.27,28
Fathers also contributed meat to their pregnant and nursing
wives.29 Children themselves foraged in many HG groups,
collecting substantial amounts of plant foods, shellfish,
and some small game such as lizards and birds,30-33 although
mastery of hunting was delayed well into adulthood.34
Finally, it has become clear since our initial publications
that marine, lacustrine, and riverine species were important
sources of animal flesh during the evolution of modern
Homo sapiens35 and may have played a role in the evolution
of brain ontogeny.36 In any case, shellfish and other aquatic
animal species must be considered part of the spectrum of
ancestral nutrition adaptations. Thus, there have been
changes in the way we estimate the likely diets of ancestral
HG populations, admitting more variability. However, the
best current estimates restrict most of that variability to a
range from 35% to 65% animal flesh, including substantial
marine animal resources for at least 0.2 million years. As we
will see, these new estimates do not affect the direction of
the great majority of our recommendations.
The Discordance Model of
Chronic Disease Prevention:
How Well Has It Fared?
We now consider macro- and selected micronutrients,
touching on our original estimates of levels in HG diets;
changes and controversies about those estimates in the
intervening years; current American intake levels; changes
in standard guidelines for nutrition and health parame-
ters in the United States; and accumulating evidence about
the value for disease prevention of a return to or toward
HG levels.
Fat and Saturated Fat
It was widely accepted by the late 1980s that saturated fat
(SF) intake in the typical modern diet is far too high and
that the C-14 and C-16 fatty acids are a major contributor
to endemic atherosclerosis underlying most coronary artery
disease and stroke, the first and third leading causes of
death. Through energy load, total fat (TF) intake is an
important contributor to endemic obesity and the grow-
ing epidemic of T2DM. Standard recommendations sug-
gested that TF be reduced to no more than 30% of
calories and that the ratio of SF to unsaturated fat be
reduced markedly. At the time, we estimated that in the
HG diet, TF contributed about 20% of calories, including
about 6% SF, a level of restriction deemed by most
authorities to be too difficult to achieve. On the basis of
new analyses of HG diets, we have raised the estimated
range of their likely TF intake to 20%-35%. Both low-fat
(20%) and high-fat (40%) diets have been shown to aid in
weight loss given appropriate caloric restriction and adh-
erence,37 but it has also been shown that very low TF may
not only prevent or retard atherosclerosis but, combined
with other lifestyle changes, partly reverse established athero-
sclerotic plaques.38,39
However, TF is only part of the story. Game animals
have more mono- and polyunsaturated fatty acids (MUFA
and PUFA) than supermarket meat.10 It used to be rec-
ommended that SF intake be less than 10% of total energy,
but according to the Institute of Medicine (IOM), any
increase raises cardiac risk.40 (However, recent evidence
suggests that the C-14 and C-16 saturated fatty acids,
and not C-18 stearic acid, are the chief serum cholesterol-
raising components of animal fat.41) HG trans-fatty-acid
intake was a small fraction of our 2% of total calories. Esp-
ecially given their high estimated intake of marine animal
flesh,35,36 PUFA intake would have been nearly twice the
present level of 15 g/d, due almost entirely to a greater
proportion of cardioprotective ω-3 forms. The resulting
ω-6:ω-3 ratio of about 2:1 contrasts with the current ratio
of about 10:1, with 8:1 recommended.40 We predict that
future recommendations for this ratio will decline further.
Dietary Cholesterol
We reported that HG cholesterol intake was similar to or
higher than that of modern Americans. Since muscle cell
membranes have as much cholesterol as fat cell mem-
branes, low-fat game meat and fish still had high choles-
terol content. HG diets suggested that discordance did not
apply to dietary cholesterol levels and that concern about
them would lead to unnecessary restriction of low-fat
meat and fish. It has since become clear that dietary cho-
lesterol is not a major independent driver of serum cho-
lesterol or its fractionation. The major dietary determinants
are the cholesterol-raising fatty acids, with an additional
contribution from dietary refined carbohydrates, suggest-
ing that for most people, restriction of these 2 compo-
nents of diet is sufficient to keep serum cholesterol and
the low-density lipoprotein (LDL)/high-density lipoprotein
(HDL) ratio very low.42-45 HG cholesterol intake is estimated
by guest on February 14, 2011ncp.sagepub.comDownloaded from

Paleolithic Nutrition / Konner, Eaton 597
at 480 mg/d, nearly 200 mg/d higher than recommenda-
tions. This level would be expected to elevate serum cho-
lesterol about 0.2 mmol/L (8 mg/dL), but the impact is far
outweighed by their lower intake of cholesterol-raising fatty
acids. In addition, high-protein intake can be expected to
have further mitigated the atherogenic effects of fat.
Protein
In the 1980s, most dietary advice called for a reduction in
protein intake, especially in the form of meat. This was not
consistent with the HG model, and we reasoned that the
ill effects of meat were mainly due to its almost inevitable
association in our culture with high cholesterol-raising fat
intake. Another concern raised by some authorities was
that nitrogen load might become excessive with high meat
consumption. Subsequent analyses have substantially
increased the estimate of HG protein intake,10-12 but there
is no evidence as yet that lean meat (similar to wild game)
intake corresponding to the levels in the average HG diet
has adverse health consequences. It has, in fact, been
shown that although diets rich in lean beef raise arachi-
donic acid concentrations (a negative), their long-chain
ω-3 PUFA content, plus the intrinsic hypocholesterol-
emic effect of protein, results in a serum lipid profile
thought to be protective against atherosclerosis.46 To the
extent that HG diets included aquatic species,35 this effect
would have been further enhanced.
Carbohydrates
Americans obtain about half their daily energy from carbo-
hydrate (CHO), including 15% from added sugars. HG
CHO consumption ranged widely, from about 35%-
65%,10,13,47 with perhaps 2%-3% from honey. Cereal grains
(85% refined) are our largest single CHO source, with dairy
products another significant contributor. HGs had little of
either, so nearly all CHO came from fruits and vegetables
(adding up to less than a fourth of current CHO), which
generally yield more desirable glycemic responses. The IOM
recommends a CHO range from 45%-65% of total energy,
with no more than 25% from added sugars.40 This recom-
mendation would approximate HG total CHO intake, but
qualitative equivalence would require that nearly all CHO
come from fruits and vegetables, with a minimum from
cereal grains, none refined. It is of particular interest that
a randomized controlled trial of the Mediterranean diet com-
pared with a simulated HG diet found the latter to be more
effective in improving insulin resistance and cardiovascular
risk factors in T2DM (see below for further discussion).48,49
Fiber
Uncultivated vegetables and fruits are markedly more fibrous
(13.3 g fiber/100 g) than commercial ones (4.2 g/100 g).50
Our 1985 estimate was limited to crude fiber, but soon
thereafter data on total fiber content became available
and suggested that total fiber intake (TFI) would have
averaged 150 g/d. With lower estimates of total HG CHO
intake, the estimate could be as low as 70 g/d but not
lower. The IOM suggested 25 g/d for women and 38 g/d
for men, a bit more than twice the current median intake,
but found insufficient evidence to set a tolerable maxi-
mum.40 High fiber intake may adversely affect mineral
bioavailability, especially in the presence of phytic acid, a
prominent constituent of many cereal grains but minimal
in uncultivated fruits and vegetables.50 Fruit and vegetable
fiber is also more completely fermented than cereal fiber,
enhancing its advantages. Finally, the HG ratio of insolu-
ble to soluble fiber was much higher than at present,
approximately 1:1.
Sodium and the sodium/potassium ratio
Both sodium (Na+) intake (768 mg/d) and the sodium/
potassium (Na+/K+) ratio were found to be extremely low
in HG diets. Although it is widely agreed that secondary
prevention of hypertension (HTN) should include lower-
ing a very high salt intake, debate has continued over the
importance of these measures in primary prevention. The
INTERSALT study demonstrated a relationship between
dietary Na+ and blood pressure (BP). With a 100 mmol/d
lower urinary Na+, population systolic pressure would rise
9 mm Hg less from age 25 to 55 years, corresponding at
age 55 to a risk reduction of 16% for coronary death and
23% for stroke death.51,52 Subsequent analysis showed that
the same difference in Na+ excretion would correspond to
a systolic/diastolic BP difference of 10-11/6 mm Hg.53,54
Critics noted that 4 of the 52 centers in the study accoun-
ted for most of the observed relationship.55
However, only in these 4 (Kenya, Papua New Guinea,
and 2 Native American groups in Brazil) was Na+ intake
near the range we found for HG populations, which show
little or no rise in BP with age. The Yanomamo, for example,
had a very low urinary sodium excretion (0.9 mmol/24 h),
mean systolic/diastolic BP of 95.4/61.4 mm Hg, and no
cases of HTN. Their BP did not rise with age, and within
the population, urinary Na+ was positively and urinary K+
negatively related to systolic BP, after controlling for age
and body mass index (BMI).56 Epidemiological research
rarely includes a group with low enough Na+ intake and
Na+/K+ ratio to be in the HG range; this limitation might
help to explain why clinical studies did not initially show a
strong relationship between these electrolytes and HTN.
However, recent work helps resolve the uncertainty
in favor of the discordance model, even without including
very low salt intake populations. A meta-analysis of pro-
spective observational studies conducted from 1966 to
2008 (19 independent samples from 13 studies including
177,000 participants) concluded that reducing salt intake
by guest on February 14, 2011ncp.sagepub.comDownloaded from

598 Nutrition in Clinical Practice / Vol. 25, No. 6, December 2010
from the estimated adult average of 10 g/d to the World
Health Organization (WHO) recommendation of 5 g/d
would be associated with a 23% difference in the rate of
stroke and a 17% difference in the overall rate of cardio-
vascular disease (CVD), preventing more than 4 million
deaths worldwide annually.57 Two randomized controlled
trials in which dietary interventions reduced Na+ intake
by 25%-35% achieved small but significant reductions in
BP over a 1.5- to 4-year period, but follow-up 10-15 years
later showed a reduced risk of cardiovascular (CV) events
of 25%.58 Most recently, a well-validated computer simu-
lation projected the effect on U.S. mortality from CHD
and stroke of linearly reducing salt intake by 0-3 g/d from
current estimates of 10.4 and 7.3 g/d for adult males and
females, respectively; the estimated reduction of the
number of deaths per year was 44,000 to 92,000.59 An
accompanying editorial called the evidence “compelling”
and the pot ential benefits “huge.”60 This is a long way from
the seemingly equivocal evidence relating salt intake to
illness that was available when we first pointed out an
order of magnitude difference between HGs and our-
selves in this dietary risk factor.
Electrolyte and Acid Base Balances
Due in part to the changes in the K+/Na+ ratio, acid base
balances have also changed markedly. In addition to the
roughly 10-fold difference between estimated HG Na+
intake and ours, their K+/Na+ ratio was probably at least
5:1; now Na+ exceeds K+, due to added Na+ and low con-
sumption of K+-rich fruits and vegetables.61 The latter
would also have driven systemic pH toward alkalinity,
whereas cereal grains and most dairy products are net
acid yielding.62 Over decades, the corrective metabolic
measures needed to offset acid-yielding diets cause uri-
nary calcium loss, accelerated skeletal calcium depletion,
calcific urolithiasis, age-related muscle wasting, and dete-
riorating renal function.62 A recent effort to model the net
endogenous acid-producing potential of the diets of 229
HG groups suggests that the majority had a net positive
acid load,63 but many of these had adaptations that could
not have been ancestral ones (eg, equestrian hunting and
circumpolar residence). Those HG diets that were pre-
dominantly plant based (such as those of ancestral East
African populations) would have had a more favorable net
negative acid load, so that an earlier estimate, which sug-
gested an overall alkaline net load for HG diets, remains
pertinent to the model.62
Experimental Clinical Studies
As noted above, an early study returning sedentary Australian
former HGs with T2DM to their traditional diet and life-
style for a period of 7 weeks lowered fasting and postprandial
glucose, increased insulin response, and markedly lowered
fasting plasma triglycerides.3 We have considered such
studies to be of the greatest importance and urged further
clinical experiments, especially with people living in modern
industrial states, to test the discordance model. Fortunately,
this work is now under way.
In 1 noncontrolled challenge study, 9 nonobese, sed-
entary, healthy volunteers consumed their usual diets for
3 days, then 3 “ramp-up” diets with increasing fiber and
K+ intake for 7 days, and finally an HG-type diet of lean
meat, fruits, vegetables, and nuts for 10 days, omitting
cereal grains, dairy products, and legumes.64 Participants
were monitored to ensure absence of weight loss. They
experienced modest but significant reductions in BP with
imp roved arterial distension; decreased insulin secretion
(area under curve, AUC) in a 2-hour oral glucose toler-
ance test (OGTT), with a marked reduction in insulin/
glucose ratio; and 16% and 22% reductions in total serum
and LDL cholesterol, respectively.64 These outcomes
seem remarkable for such a short-term intervention.
More interesting still are results from randomized con-
trolled trials (RCTs). In the most persuasive study to date,
29 patients with ischemic heart disease and either glucose
intolerance or T2DM were randomized to 12 weeks of a
“Paleolithic” diet (n = 14) based on lean meat, fish, fruit,
vegetables, root vegetables, eggs, and nuts or a Mediterranean-
like “Consensus” diet (n = 15) based on whole grains,
low-fat dairy products, vegetables, fruits, fish, oils, and mar-
garines.49 In OGTTs, the Paleolithic group showed a 26%
reduction in AUC glucose compared to a 7% reduction in
the Consensus group. There was a greater decrease in waist
circumference in the Paleolithic group (–5.6 cm) than in the
Consensus group (–2.9 cm), but the glucose reduction was
independent of that measure.
In a second randomized crossover pilot study, the start-
ing point was 13 patients (3 women) with T2DM who were
placed on a Paleolithic diet based on lean meat, fish, fruit,
vegetables, root vegetables, eggs, and nuts, and a Diabetes
diet according to the American Diabetes Association guide-
lines65 (evenly distributed meals with increased vegetables,
root vegetables, fiber, whole-grain bread and other cereal
products, fruits, and berries, but decreased TF, especially
cholesterol-raising SF).48 Participants were on each diet
for 3 months. Compared to the Diabetes diet, the Paleolithic
diet produced lower mean levels of hemoglobin A1c, tria-
cylglycerol, diastolic BP, weight, BMI, and waist circum-
ference, and higher mean HDL.
Although these are small studies, it is very gratifying
that the era of explicit experimental study of the discord-
ance model has begun and that initial results are consist-
ent with our original predictions. It is especially noteworthy
that 2 of the studies were randomized trials that compa-
red the HG diet to other recommended model diets rather
than to a baseline or typical Western diet. We hope and
trust that this work will continue.
by guest on February 14, 2011ncp.sagepub.comDownloaded from

Paleolithic Nutrition / Konner, Eaton 599
Discussion
Although not an across-the-board vindication of the HG
model, and despite some changes from our macronutrient
estimates as originally presented, research in the past
quarter century has vindicated the clinical and epidemio-
logical relevance of the model. Without supplying numbers,
some of which might be controversial, we can confidently
estimate the direction and magnitude of the modern diet’s
deviation from the HG diet in the range of EEAs (Table 1).
More notably, research has suggested that where the model
departed from standard 1985 recommendations, a shift
toward the model would contribute further to primary
prevention of several important diseases. Indeed, in some
instances, the standard recommendations have already
shifted in that direction (Table 2). This is the case for
total serum cholesterol; it is now considered highly desir-
able to be under 180 mg/dL, whereas in 1985, the thresh-
old was 200. We predict that the threshold will be lowered
further in future recommendations.
The HG model and the discordance hypothesis sug-
gest that meat and fish consumption can safely be higher
than in current recommendations. Recent dietary fads,
based on unproven theories of metabolism, claim that very
low carbohydrate intake combined with high protein and
fat consumption can safely produce weight loss. That this
kind of diet can produce at least temporary weight loss has
been demonstrated,66,67 and several studies now show that
levels of lean meat and fish intake higher than those in
many officially recommended diets are as safe or safer for
some groups of patients. We continue to believe that the
risk associated with the consumption of meat is almost
entirely explained by the high proportions of TF and espe-
cially SF in commercial meats. Neither the protein con-
tent of meat nor the cholesterol content of cell membranes
has been shown to adversely affect health at the (fairly
high) level characteristic of HG diets.
Reduction of carbohydrates to extremely low levels is
not consistent with the HG model, but neither is a very
high CHO, “meat as a condiment”–type diet; furthermore,
CHO sources are important. HG CHO came from fruit,
vegetables, and nuts, not from grains. Refined, concen-
trated CHOs such as sucrose played virtually no role, and
the consumption of plant CHO necessarily resulted in
high fiber intake. If we were to rebuild the food pyramid
along HG lines, the base would not be grains but fruits
and vegetables, which could be chosen to provide ade-
quate fiber content. The second tier would be meat, fish,
and low-fat dairy products, all very lean. Whole grains
might come next (although even these were very unusual
for HGs), whereas fats, oils, and refined carbohydrates
would occupy the same very small place at the top, essen-
tially functioning as condiments in a healthy diet. These
guidelines would not exactly replicate the HG diet in terms
Table 1. Widely Agreed-on Qualitative Differences Between Average Ancestral (Hunter-Gatherer) Diets and
Contemporary Western Diets
Ancestral (Hunter-Gatherer) Contemporary Western
Total energy intake More Less
Caloric density Very low High
Dietary bulk More Less
Total carbohydrate intake Less More
Added sugars/refined carbohydrates Very little Much more
Glycemic load Relatively low High
Fruits and vegetables Twice as much Half as much
Antioxidant capacity Higher Lower
Fiber More Less
Soluble:insoluble Roughly 1:1 <1 insoluble
Protein intake More Less
Total fat intake Roughly equal
Serum cholesterol-raising fat Less More
Total polyunsaturated fat More Less
ω-6:ω-3 Roughly equal Far more ω-6
Long-chain essential fatty acids More Less
Cholesterol intake Equal or more Equal or less
Micronutrient intake More Less
Sodium:potassium <1>1
Acid base impact Alkaline or acidic Acidic
Milk products Mother’s milk only High, lifelong
Cereal grains Minimal Substantial
Free water intake More Less
by guest on February 14, 2011ncp.sagepub.comDownloaded from

600 Nutrition in Clinical Practice / Vol. 25, No. 6, December 2010
of food categories, but it would do so roughly in terms of
macronutrients.
Na+ and the Na+/K+ ratio no longer provide a chal-
lenge for the HG model since large prospective epidemio-
logical studies and randomized clinical trials have recently
shown a clear correlation between dietary sodium and the
risk of CV disease, even for differences within a range
much higher than HG intake. However, since sodium
intake levels in those studies have rarely reached down
into the HG range, it remains possible that much greater
gains could be achieved than those suggested by current
studies.
As for other aspects of lifestyle, tobacco products, rare
in HG environments, have been the targets of inc reasingly
strong societal restriction, and we know that the frequency
and duration of exercise, including walking and other less
intense exercise, should be much higher than it is and
should include resistance and flexibility as well as cardio-
vascular components. Interestingly, we had been ske ptical
of the notion, common in the 1980s, that walking was not
an adequate exercise because half of our HG ancestors—the
women—exercised in this manner and did very little run-
ning. The subsequent finding that walking and other mod-
erate exercise also reduce the risk of cardiopulmonary
disease was consistent with the HG model.
Further research on HG populations themselves in
the past quarter century has confirmed most of our earlier
generalizations about them.68,69 Unfortunately, Wester-
nization worsens their diet and their health indicators, for
example among the Australian Aborigines.70,71
Not every prediction of the HG model will result in
the best recommendation. The case of ethanol consum-
ption, extremely low before the invention of agriculture,
dep arts from the model. A number of studies show that
mild to moderate ethanol intake reduces cardiovascular
risk, at least against the background of a modern diet.
The ease with which ethanol intake progresses to levels
that pose a wide range of other health risks suggests that
we were not set up by our evolution to handle this com-
pound without difficulty, but the positive value of small
amounts shows that the HG model cannot answer all
questions.
Still, a review of research since our original publica-
tion largely vindicates the model we presented 25 years
ago. Common arguments against the approach have been
effectively answered.72 Ridicule notwithstanding, the HG
paradigm offers a good provisional alternative to the some-
times confused, occasionally conflicting, and often inad-
equately prioritized stream of research findings.72 It is
almost certainly superior to the vast majority of diet fads,
the scientific basis of which is almost always dubious, and
which have failed to halt what has been called an obesity
pandemic73 and an ominous rise in T2DM,74 even in chil-
dren and adolescents.75
Table 2. Changing Recommendations for Diet and Biological Markers, as Compared With Current Estimates for
Hunter-Gatherers in the Range of Environments of Evolutionary Adaptedness
Recommendations
Pre-1990 Current Estimated Ancestral
Nutrients
Carbohydrate, % daily energy 55-60 45-65 35-40
Added sugar, % daily energy 15 <10 2
Fiber, g/d — 38 male; 25 female >70
Protein, % daily energy 10-15 10-35 25-30
Fat, % daily energy 30 20-35 20-35
Saturated fat, % daily energy <10 <10 7.5-12
Cholesterol, mg/d <300 <300 500+
Eicosapentaenoic acid and
docosahexaenoic acid, g/d — 0.65 0.7-6.0
Vitamin C, mg/d 60 90 male; 75 female 500
Vitamin D, IU/d 400 1000 4000 (sunlight)
Calcium, mg/d 800 1000 1000-1500
Sodium mg/d 2400 1500 <1000
Potassium mg/d 2500 4700 7000
Biomarkers
Blood pressure, mm Hg <140/90 115/75 110/70
Serum cholesterol, mg/dL 200-240 115-165 125
Body composition, %lean:%fat
Females — <31% fat 35-40:20-25
Males — <26% fat 45-50:10-15
Physical activity, kcal/d — 150-490 >1000
by guest on February 14, 2011ncp.sagepub.comDownloaded from

Paleolithic Nutrition / Konner, Eaton 601
Unfortunately, a great many Americans have yet to
accept the basic facts and theory of evolution, an obvious
obstacle to offering everyone the paradigm we advocate.
However, most people respond to the notion of a “natu-
ral” diet and lifestyle, and the HG model is the first and
only scientific approach to that notion. Certainly most
physicians and medical scientists can accept it as one
organizing principle for past and future research. Although
an anthropological model cannot be accepted at face
value—only the best clinical, experimental, and epidemi-
ological research can finally justify recommendations—
we can be increasingly guided in this research by such a
model. Meanwhile, we can keep an open mind about
what we may learn from our remote ancestors.
References
1. Eaton SB, Konner M. Paleolithic nutrition: a consideration of its
nature and current implications. N Engl J Med. 1985;312:283-289.
2. Eaton SB, Konner M, Shostak M. Stone agers in the fast lane:
chronic degenerative disease in evolutionary perspective. Am J Med.
1988;84:739-749.
3. O’Dea K. Marked improvement in carbohydrate and lipid metabo-
lism in diabetic Australian Aborigines after temporary reversion to
traditional lifestyle. Diabetes. 1984;33:596-603.
4. Cohen MN, Armelagos GJ, eds. Paleopathology at the Origins of
Agriculture. New York: Academic Press; 1984.
5. Hancock AM, Witonsky DB, Ehler E, et al. Colloquium paper:
human adaptations to diet, subsistence, and ecoregion are due to
subtle shifts in allele frequency. Proc Natl Acad Sci U S A. 2010;
107:8924-8930.
6. Hancock AM, Alkorta-Aranburu G, Witonsky DB, di Reinzo A.
Adaptations to new environments in humans: the role of subtle
allele frequency shifts. Philos Trans R Soc Lond B Biol Sci. 2010;
365:2459-2468.
7. Smith E, Morowitz H. Universality in intermediary metabolism.
Proc Natl Acad Sci U S A. 2004;101:13168-13173.
8. Danaei G, Ding EL, Mozaffarian D, et al. The preventable causes of
death in the United States: comparative risk assessment of dietary,
lifestyle, and metabolic risk factors. PLoS Med. 2009;6(4):e1000058.
9. Eaton SB, Shostak M, Konner M. The Paleolithic Prescription: A
Program of Diet and Exercise and a Design for Living. New York:
Harper & Row; 1988.
10. Cordain L, Eaton SB, Miller JB, Mann N, Hill K. The paradoxical
nature of hunter-gatherer diets: meat-based, yet non-atherogenic.
Eur J Clin Nutr. 2002;56(suppl 1):S42-S52.
11. Cordain L, Miller JB, Eaton SB, Mann N. Macronutrient estima-
tions in hunter-gatherer diets. Am J Clin Nutr. 2000;72:1589-1592.
12. Cordain L, Miller JB, Eaton SB, Mann N, Holt SH, Speth JD. Plant-
animal subsistence ratios and macronutrient energy estimations in
worldwide hunter-gatherer diets. Am J Clin Nutr. 2000;71:682-692.
13. Milton K. Hunter-gatherer diets: a different perspective. Am J Clin
Nutr. 2000;71:665-667.
14. Copeland SR. Potential hominin plant foods in northern Tanzania:
semi-arid savannas versus savanna chimpanzee sites. J Hum Evol.
2009;57:365-378.
15. White TD, Asfaw B, Beyene Y, et al. Ardipithecus ramidus and the
paleobiology of early hominids. Science. 2009;326:75-86.
16. Wrangham R, Cheney D, Seyfarth R, Sarmiento E. Shallow-water
habitats as sources of fallback foods for hominins. Am J Phys Anth-
ropol. 2009;140:630-642.
17. Strait DS, Weber GW, Neubauer S, et al. The feeding biomechan-
ics and dietary ecology of Australopithecus africanus. Proc Natl
Acad Sci U S A. 2009;106:2124-2129.
18. Aiello LC. Brains and guts in human evolution: the expensive tis-
sue hypothesis. Braz J Genet. 1997;20:141-148.
19. Wrangham R, Conklin-Brittain N. Cooking as a biological trait.
Comp Biochem Physiol A Mol Intergr Physiol. 2003;136:35-46.
20. Karkanas P, Shahack-Gross R, Ayalon A, et al. Evidence for habit-
ual use of fire at the end of the Lower Paleolithic: site-formation
processes at Qesem Cave, Israel. J Hum Evol. 2007;53:197-212.
21. Milton K. A hypothesis to explain the role of meat-eating in human
evolution. Evol Anthropol. 1999;8:11-21.
22. Potts R. Environmental hypotheses of hominin evolution. Am J
Phys Anthropol. 1998;41(suppl 27):93-136.
23. Hockett B, Haws J. Nutritional ecology and diachronic trends in
paleolithic diet and health. Evol Anthropol. 2003;12:211-216.
24. Kellner CM, Schoeninger MJ. A simple carbon isotope model for
reconstructing prehistoric human diet. Am J Phys Anthropol. 2007;
133:1112-1127.
25. Lancaster JB, Lancaster CS. Parental investment: the hominid
adaptation. In: Ortner D, ed. How Humans Adapt. Washington, DC:
Smithsonian Institution Press; 1983:35-56.
26. Kaplan H, Hill K, Lancaster J, Hurtado AM. A theory of human life
history evolution: diet, intelligence, and longevity. Evol Anthropol.
2000;9:156-185.
27. Hawkes K. Grandmothers and the evolution of human longevity.
Am J Hum Biol. 2003;15:380-400.
28. Hawkes K, O’Connell JF, Blurton Jones NG. Hadza women’s time
allocation, offspring provisioning, and the evolution of long post-
menopausal life spans. Curr Anthropol. 1997;38:551-577.
29. Marlowe FW. A critical period for provisioning by Hadza men:
Implications for pair bonding. Evol Hum Behav. 2003;24:217-229.
30. Bird DW, Bird RB. Martu children’s hunting strategies in the Western
Desert, Australia. In: Hewlett BS, Lamb ME, eds. Hunter-Gatherer
Childhoods: Evolutionary, Developmental & Cultural Perspectives.
New Brunswick, NJ: Aldine Transaction; 2005:129-146.
31. Bird DW, Bliege Bird R. The ethnoarchaeology of juvenile forag-
ers: shellfishing strategies among Meriam children. J Anthropol
Archaeol. 2000;19:461-476.
32. Blurton Jones NG, Hawkes K, O’Connell JF. Why do Hadza chil-
dren forage? In: Segal NL, Weisfeld GE, Weisfeld CC, eds.
Uniting Psychology and Biology: Integrative Perspectives on Human
Development. Washington, DC: American Psychological Association;
1997:279-313.
33. Tucker B, Young AG. Growing up Mikea: children’s time allocation
and tuber foraging in southwestern Madagascar. In: Hewlett BS,
Lamb ME, eds. Hunter-Gatherer Childhoods: Evolutionary, Develop-
mental & Cultural Perspectives. New Brunswick, NJ: Aldine Transaction;
2005:147-171.
34. Gurven M, Kaplan H, Gutierrez M. How long does it take to become
a proficient hunter? Implications for the evolution of extended
development and long life span. J Hum Evol. 2006;51:454-470.
35. Marean CW, Bar-Matthews M, Bernatchez J, et al. Early human
use of marine resources and pigment in South Africa during the
Middle Pleistocene. Nature. 2007;449:905-908.
36. Broadhurst CL, Wang Y, Crawford MA, Cunnane SC, Parkington JE,
Schmidt WF. Brain-specific lipids from marine, lacustrine, or ter-
restrial food resources: potential impact on early African Homo
sapiens. Comp Biochem Physiol B Biochem Mol Biol. 2002;131:
653-673.
37. Sacks FM, Bray GA, Carey VJ, et al. Comparison of weight-loss
diets with different compositions of fat, protein, and carbohy-
drates. N Engl J Med. 2009;360:859-873.
38. Ornish D, Brown SE, Scherwitz LW, et al. Lifestyle changes and
heart disease. Lancet. 1990;336:741-742.
by guest on February 14, 2011ncp.sagepub.comDownloaded from

602 Nutrition in Clinical Practice / Vol. 25, No. 6, December 2010
39. Gould KL, Ornish D, Scherwitz L, et al. Changes in myocardial
perfusion abnormalities by positron emission tomography after
long-term, intense risk factor modification. JAMA. 1995;274:
894-901.
40. Institute of Medicine. Dietary Reference Intakes: Energy, Carbohydrate,
Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids (Macro-
nutrients). Washington DC: National Academies Press; 2005.
41. Hunter JE, Zhang J, Kris-Etherton P. Cardiovascular disease risk
of dietary stearic acid compared with trans, other saturated, and
unsaturated fatty acids: a systematic review. Am J Clin Nutr. 2010;
91:46-63.
42. Schonfeld G, Patsch W, Rudel LL, Nelson C, Epstein M, Olson RE.
Effects of dietary cholesterol and fatty acids on plasma lipoproteins.
J Clin Invest. 1982;69:1072-1080.
43. Bronsgeest-Schoute DC, Hermus RJ, Dallinga-Thie GM,
Hautvast JG. Dependence of the effects of dietary cholesterol and
experimental conditions on serum lipids in man: III. The effect
on serum cholesterol of removal of eggs from the diet of free-
living habitually egg-eating people. Am J Clin Nutr. 1979;32:
2193-2197.
44. Edington J, Geekie M, Carter R, et al. Effect of dietary cholesterol
on plasma cholesterol concentration in subjects following reduced
fat, high fibre diet. Br Med J (Clin Res Ed). 1987;294:333-336.
45. Forsythe CE, Phinney S, Fernandez ML, et al. Comparison of low
fat and low carbohydrate diets on circulating fatty acid composi-
tion and markers of inflammation. Lipids. 2008;43:65-77.
46. Sinclair AJ, Johnson L, O’Dea K, Holman RT. Diets rich in lean beef
increase arachidonic acid and long-chain omega 3 polyunsaturated
fatty acid levels in plasma phospholipids. Lipids. 1994;29:337-343.
47. Cordain L, Miller JB, Eaton SB, Mann N, Holt SH, Speth JD. Plant-
animal subsistence ratios and macronutrient energy estimations in
worldwide hunter-gatherer diets. Am J Clin Nutr. 2000;71:682-692.
48. Jönsson T, Granfeldt Y, Ahrén B, et al. Beneficial effects of a
Paleolithic diet on cardiovascular risk factors in type 2 diabetes: a
randomized cross-over pilot study. Cardiovasc Diabetol. 2009;8:35.
49. Lindeberg S, Jönsson T, Granfeldt Y, et al. A Palaeolithic diet improves
glucose tolerance more than a Mediterranean-like diet in individuals
with ischaemic heart disease. Diabetologia. 2007;50:1795-1807.
50. Jenike MR. Nutritional ecology: diet, physical activity and body
size. In: Panter-Brick C, Layton RH, Rowley-Conwy P, eds. Hunter-
Gatherers: An Interdisciplinary Perspective. Cambridge, UK: Cambridge
University Press; 2001:205-238.
51. Stamler J, Rose G, Stamler R, Elliott P, Dyer A, Marmot M.
INTERSALT study findings. Public health and medical care impli-
cations. Hypertension. 1989;14:570-577.
52. Elliott P, Marmot M, Dyer A, et al. The INTERSALT study: main
results, conclusions and some implications. Clin Exp Hypertens A.
1989;11:1025-1034.
53. Stamler J. The INTERSALT study: background, methods, findings,
and implications [erratum appears in Am J Clin Nutr. 1997;66:1297].
Am J Clin Nutr. 1997;65(suppl):626S-642S.
54. Elliott P, Stamler J, Nichols R, et al. Intersalt revisited: further ana-
lyses of 24 hour sodium excretion and blood pressure within and
across populations. Intersalt Cooperative Research Group [erratum
appears in BMJ. 1997;315:458]. BMJ. 1996;312:1249-1253.
55. Freedman DA, Petitti DB. Salt and blood pressure: conventional
wisdom reconsidered. Eval Rev. 2001;25:267-287.
56. Mancilha-Carvalho Jde J, Souza e Silva NA. The Yanomami
Indians in the INTERSALT Study. Arq Bras Cardiol. 2003;80:
289-300.
57. Strazzullo P, D’Elia L, Kandala N-B, Cappuccio FP. Salt intake,
stroke, and cardiovascular disease: meta-analysis of prospective
studies. BMJ. 2009;339:b4567.
58. Cook NR, Cutler JA, Obarzanek E, et al. Long term effects of
dietary sodium reduction on cardiovascular disease outcomes: obser-
vational follow-up of the trials of hypertension prevention (TOHP).
BMJ. 2007;334:885-888.
59. Bibbins-Domingo K, Chertow GM, Coxson PG, et al. Projected
effect of dietary salt reductions on future cardiovascular disease. N
Engl J Med. 2010;362:590-599.
60. Appel LJ, Anderson CAM. Compelling evidence for public health
action to reduce salt intake. N Engl J Med. 2010;362:650-652.
61. Eaton SB, Eaton SB III, Konner MJ. Paleolithic nutrition revisited:
a twelve-year retrospective on its nature and implications. Eur J
Clin Nutr. 1997;51:207-216.
62. Sebastian A, Frassetto LA, Sellmeyer DE, Merriam RL, Morris RC
Jr. Estimation of the net acid load of the diet of ancestral preagri-
cultural Homo sapiens and their hominid ancestors. Am J Clin
Nutr. 2002;76:1308-1316.
63. Ströhle A, Hahn A, Sebastian A. Estimation of the diet-dependent
net acid load in 229 worldwide historically studied hunter-gatherer
societies. Am J Clin Nutr. 2010;91:406-412.
64. Frassetto LA, Schloetter M, Mietus-Synder M, Morris RC Jr,
Sebastian A. Metabolic and physiologic improvements from con-
suming a Paleolithic, hunter-gatherer type diet. Eur J Clin Nutr.
2009;63:947-955.
65. American Diabetes Association. Nutrition recommendations and
principles for people with diabetes mellitus (position statement).
Diabetes Care. 2001;24(suppl 1):S48-S50.
66. Foster GD, Wyatt HR, Hill JO, et al. A randomized trial of a low-
carbohydrate diet for obesity. N Engl J Med. 2003;348:2082-2090.
67. Johnston CS, Tjonn SL, Swan PD. High-protein, low-fat diets are
effective for weight loss and favorably alter biomarkers in healthy
adults. J Nutr. 2004;134:586-591.
68. Naughton JM, O’Dea K, Sinclair AJ. Animal foods in traditional
Australian aboriginal diets: polyunsaturated and low in fat. Lipids.
1986;21:684-690.
69. O’Dea K. Traditional diet and food preferences of Australian abo-
riginal hunter-gatherers. Philos Trans R Soc Lond B Biol Sci. 1991;
334:233-240; discussion 240-241.
70. O’Dea K. Westernisation, insulin resistance and diabetes in Australian
aborigines. Med J Aust. 1991;155:258-264.
71. Rowley KG, O’Dea K. Diabetes in Australian aboriginal and Torres
Strait Islander peoples. P N G Med J. 2001;44:164-170.
72. Eaton SB, Cordain L, Lindeberg S. Evolutionary health promotion:
a consideration of common counterarguments. Prev Med. 2002;
34:119-123.
73. Katz DL. Pandemic obesity and the contagion of nutritional non-
sense. Public Health Rev. 2003;31:33-44.
74. James WPT. The epidemiology of obesity: the size of the problem.
J Intern Med. 2008;263:336-352.
75. Kempf K, Rathmann W, Herder C. Impaired glucose regulation
and type 2 diabetes in children and adolescents. Diabetes Metab Res
Rev. 2008;24:427-437.
by guest on February 14, 2011ncp.sagepub.comDownloaded from