To:     All Paleodieters
From:   Reply to Loren Cordain, Part I, from Mary and Sally

We apologize for taking so long to reply to Loren Cordainís submission on
October 9.  We will take his points in order.

1.  We stated that the necessary macronutrients and known vitamins and
minerals can all be obtained from animal foods.  We speculated that the
various phyto-chemicals MAY be necessary for optimum health, at least to
some individuals.  Certainly plant foods add variety to the diet and make
eating more fun.  But plant foods can have their drawbacks, particularly if
improperly prepared.  Grains, legumes and other seed foods contain anti-
nutrients such as phytic acid, enzyme inhibitors and goitrogenic substances
(including isoflavones, genistein and diadzen in soy foods.)(1)

2.  We did not say that hydrogenated oils are atherogenic BECAUSE they
raise cholesterol.  We have stated that cholesterol is a non issue--neither
serum nor dietary cholesterol is good predictor of CHD.  Trans fats
interfere with many processes on the cellular level that can lead to CHD.
Several recent studies have linked consumption of trans fats with increased
CHD.(2) Nor did we say that polyunsaturated fats contribute to heart
disease BECAUSE they raise cholesterol  It is generally accepted that
polyunsaturated oils LOWER cholesterol in the short term.  We stated that
excess n6 oils is a problem because, among other things, it disrupts
prostaglandin balance.  In this regard, we highly recommend a recent paper
by Okuyama, "Dietary Fatty Acids--the n6/n3 balance and chronic elderly
diseases.  Excess linoleic acid and relative n3 deficiency syndrome seen in
Japan." (3)

The fact that saturated fats raise cholesterol in short term feeding
studies, many in metabolic ward conditions, has no relevance to our
argument.  Often studies done in metabolic wards involve the use of
purified fats, cholesterol, proteins, etc., which are not the same as
eating real food.  A Finnish study found that over the long-term, the
reduction of saturated fats in the diet does not affect plasma
cholesterol.(4)

The effect of fatty acids on serum cholesterol levels is dependent on the
original serum cholesterol levels:  high serum cholesterol decreases with
most fatty acids including all saturates; low serum cholesterol increases
with many of the fatty acids including saturates, monounsaturates and
sometimes the polyunsaturates. Over time, cholesterol levels tend to return
to baseline levels (that is, levels appropriate for the individual)
regardless of changes in amount or type of dietary fat.

We are among those arguing that a wide variety of risk factors contribute
to CHD, all of which may be summarized as the inappropriate application of
industrial techniques to farming, animal husbandry and food processing.

You cite a 1986 JAMA article by Grundy in which he states that the
relationship between serum cholesterol and risk of premature death from CHD
is continuous and graded.  Researchers in the MRFIT study did indeed find
that the relationship was continuous and graded, with deaths from CHD
rising in a straight line from the cholesterol levels between 140 and 299.
What Grundy failed to mention was that the relationship was also trivial,
with total CHD death at less than 1 per 1000 at 140, and slightly less than
2 per 1,000 at 299.  This was similar to findings of the Framingham study
which found virtually no difference in coronary -events- for cholesterol
levels between 180 and 300--the vast majority of the US population.  What
the MRFIT investigators did find was that the total deaths were 6 per 1000
for those with cholesterol levels less than 140, twice as high as total
deaths for those in the 140 to 260 range.

If the consumption of saturated fat within the context of the "average
American diet" is "deadly," why was heart disease so rare among Americans
at the turn of the century, who consumed large amounts of saturated fat
from dairy products, lard, tallows and coconut oil?  The answer is that,
like our hunter gatherer ancestors, -the context under which this was done
was much different than present day conditions,- namely the introduction of
altered foodstuffs into the food supply, including trans fats and the other
factors you mention such as large amounts of refined carbohydrates and a
high n6/n3 ratio.  The fault is not with the saturated fats, but with the
-context.-

You are arguing that the Paleolithic diet contained higher levels of
monounsaturated fats, stearic acid and n3 fatty acids than the modern diet,
but lower levels of overall saturates.  The crux of your argument seems to
be that the Paleolithic diet had far lower amounts of all saturates than
the modern diet, saturates that we obtain from meat and butterfat.  The
following table combines data from USDA and from your laboratories on elk
killed in the Fall 1996 hunting season:(5)

ELK            Brain    Muscle    Adipose    Marrow    Kidney

Saturates      41.3%    34.2%     53.4%      20.0%     66.0%

Mono           33.0%    28.8%     34.0%      75.5%     32.3%

Poly           26.0%    35.5%      8.6%       4.9%      1.7%

N6/N3 ratio     1.14    3.72       3.39       5.23       NA

P/S ratio       0.63    1.04       0.16       0.24     ~0.03


BEEF (USDA)

Saturates       44.3%   44.9%      53.3%       NA       44.9%

Mono            35.9%   50.8%      43.2%       NA       28.2%

Poly            19.8%    4.3%       3.2%       NA       26.9%

N6/N3 ratio      0.52   2.84%       3.6%       NA       66.0

P/S ratio        0.45    0.09        .06       NA       0.60

Your laboratories found that fat from elk muscle tissue was 2-4%, compared
to domesticated beef at 25-30%.  Beef muscle fat is more saturated than elk
muscle fat (and presumably contains a significant portion of
"non-atherogenic" stearic acid).  It contains far more "beneficial"
monounsaturated and a better N6/N3 ratio (2.84 vs 3.72).  Doesn't that
imply that the large amount of muscle fat is a plus for beef?

These tables indicate that elk adipose is just as saturated as beef adipose
fat.  Your laboratories did not provide a figure for adipose fat as a
percentage of total meat.  For domesticated beef, this ranges from about 9
to 27 percent.  We need this data before we can determine whether the elk
or the domesticated cow is a better overall source of saturated fat,
including the so-called atherogenic 14:0 and 16:0 fatty acids.  You found
that 14:0 was 5.1% and 16:0 was 33.1% in elk tallow for a total of 38.2%,
compared to 5.2% of 14:0 and 23.6 of 16:0 in beef tallow for a total of
28.8%.  In other words, beef has less of the so-called atherogenic fatty
acids than elk as a percentage of the total.  The question is, how do the
totals compare?

Unfortunately, we have no figures for beef marrow, but beef brain and elk
brain have similar fatty acid profiles, with beef brain slightly more
saturated, slightly higher in monounsaturated and slightly lower in
polyunsaturates.  Beef brain has a better N6/N3 ratio (0.52 vs 1.14).

We have evidence that Paleolithic man consumed the marrow and organ meats
by preference, (possibly leaving much of the muscle meat at the site of the
kill.)  Traditional cultures also ate these portions.  Modern man does not
eat these fatty organs and marrow portions of red meat (although he should)
so it can be argued that his intake of monounsaturates and polyunsaturates
is less, assuming other factors in the diet are similar.  But modern man
does not eat the kidney fat either--which probably totals 20 pounds on a
large ruminant--which the Paleolithic man almost certainly did,
particularly as this highly saturated fat has very good keeping qualities.
So it can be just as easily argued that modern man consumes less saturated
fat from beef than Paleolithic man consumed from wild game.

As for the lean meat in wild game, American Indian practices would indicate
that earlier man mixed this with tallow or bear fat to make more palatable
products, similar to pemmican, succatash or haggis.  After all, why would
anyone eat dry, tough lean meat when it tastes so much better mixed with
fat, unless the entire medical establishment and food industry were urging
him to do so?

Finally, regarding the purported atherogenicity of lauric, myristic and
palmitic acid, you may wish to note the recently published work of Tratwein
which found that olive oil produced the higher total cholesterol relative
to palm and coconut oil in the lipid sensitive hamster. (6)  In humans,
there have been numerous studies with added coconut oil to the diet that
resulted in lowering of cholesterol or no difference. (7)

Current recommendations to increase monounsaturated fats and reduce
saturated fats should, we believe, be viewed with caution.  While more
research is needed, we can cite a study by Kramer, et al which found that
feeding canola oil to rats caused increased incidence of myocardial
lesions.  When highly saturated cocoa butter was added to the diet, there
was a significant decline in the incidence of myocardial lesions. (8)
Researchers ascribed the adverse effects of the canola oil to its unique
fatty acid composition--high in oleic and n3 but low in saturated fatty
acids.  There is also some evidence that excess oleic acid in the diet
interferes with prostaglandin production in a similar fashion to excess
n6.(9)  We find it interesting that the monounsaturates became "politically
correct" just about the time that canola oil came on the market.

3.  The Framingham study found that higher levels of HDL were associated
with lower incidence of CHD.  In order to protect the lipid hypothesis,
they were obliged to pinpoint LDL as the villain.  We urge you to read the
discussion of this issue by Russell Smith, which we mentioned before and
which we will now repeat.  "In 40 years of several hundred Framingham
reports, this writer has found neither an actual correlation coefficient
published for either total or LDL cholesterol nor a figure showing the CHD
mortality and/or morbidity rate as a function of LDL level. . . Examination
of published Framingham data. . . indicate that the correlation between CHD
and total or LDL cholesterol is probably under 0.3.  The objective and
statistically sophisticated scientist would consider these correlations as
unrepresentative of cause and effect relationships and would analyze the
associated data base to find reasons why the observed correlations occurred
at all." (10)

We did not say that oxidized cholesterol in aged cheeses is ìa powerful
promoter of atherogenesis.  If this is true, why do the French, Italians,
and Greeks, who eat large quantities of aged cheeses, have generally lower
rates of CHD?  The type of damaging oxidation we are talking about occurs
when milk and eggs are "powdered" by blowing them out of a small hole at
high temperatures and pressures, thereby exposing a large surface area to
oxygen. Perhaps oxidized cholesterol was found in commercial "grated"
(actually powdered) Parmesan cheese, which remained for many months on the
grocers shelf.  If so, it is a gross deception to tar traditional aged
cheeses, which are a wholesome, nutrient-dense food, with the black brush
of commercial powdered--and probably rancid--Parmesan.

4.  You object to our statement that "one can point to populations with
relatively high protein consumption (30-40%) with little or no CHD, and to
populations with relatively low protein consumption (15-20%) with little or
no CHD."  Our analyses of diet descriptions and menus at the turn of the
century indicates that the American diet contained about 20% protein, 40%
fat and 40% carbohydrate. (11)  Higher amounts would be found in primitive
diets that contain no plant foods, such as the Eskimo, Canadian Indians and
groups that subsist exclusively on milk, meat and blood like the Masai,
although if one assumes a fat:protein ratio of 2:1, then the upper limit
for protein would be about 30%. As the prevailing viewpoint in this
discussion is that paleodiet was high in protein, but nonatherogenic, we
are at a loss to understand your objections.  You state that the
high-protein Paleolithic diet was nonatherogenic because the diet contained
a different fatty acid profile than the modern diet.  This is a different
issue than the protein issue, and is in fact the crux of our
debate--whether or not modern diets contain more saturated fats,
particularly those you contend are atherogenic (12:0, 14:0 and 16:0)--and
if so, whether these fats are contributing to the modern plague of heart
disease.  Can you show us statistics indicating that modern consumption of
12:0, 14:0 and 16:0 is much higher today than in 1900?

Regarding the ability of SFAs to lower serum Lp(a), we have already cited
one reference. (12)  We now have another, from the USDA laboratories. (13)

5.  There has been considerable discussion regarding your statement that
the hunter gatherer did not consume carbohydrates mixed with protein and
fat.  We cited succatash and pemmican as counter examples, and another
interesting example was given of the Yamani Indians in the Amazon basin who
never consume meat without plantains.  The consensus seems to be that the
paleodiet was mixed, and that our ancestors placed no strictures on
consuming carbohydrates with protein and fat.  Are you suggesting that by
consuming protein/fat and carbohydrates at separate meals, we can eliminate
CHD?

You state that it is "well established" that mixing fat with carbohydrate,
the glycemic response worsens.  In fact, studies of glycemic response show
that carbohydrates mixed with fat elicit a lower glycemic response.  For
example, Jenkins and colleagues found that a Mars bar had a glycemic index
of 68, compared to honey with an index of 87 and lucozade at 95; and potato
chips had an index of 51 compared to potatoes at 70. (14) The fact that
fats taken with meals regulate the too-rapid entry of glucose into the
blood stream is so well established as to be found in all the textbooks.
In fact, recent study by Rasmussen and colleagues (15) found that
carbohydrate (in the form of potato) mixed with largely saturated butter
provoked a lower glucose response than carbohydrate mixed with largely
monounsaturated olive oil.

6.  We did not state that overall PUFA in wild animal tissues is low.  We
stated that the PUFA in modern ruminant adipose tissue does not differ
significantly from the PUFA in wild ruminant adipose tissue.  We did not
include monogastric animals, and we were clearly referring only to the
adipose of ruminant animals.  We pointed out that modern diets generally
have much larger amounts of PUFA, particularly n6 PUFA, than traditional
diets because of the introduction of commercial vegetable oils, and not
because we are eating more animal fats..  The hunter/gatherer got
significant amounts of fat from bone marrow and organ meats--nutritious
foods that should be returned to American tables.  These provide moderate
amounts of both n6 and n3 PUFA, along with monunsaturated and saturated
fatty acids.  The Eskimo diet is unusual in the fact that it contains
relatively high amounts of PUFA, particularly n3 PUFA, which we have agreed
is very deficient in western diets. Some have argued that higher levels of
elongated n3's are necessary in cold climates, (16) but there is no reason
to assume that the levels of PUFA found in the Eskimo diet were
characteristic of hunter/gatherer diets from temperate climates based on
game, or on those from the tropics which would have included coconut and
palm oils.

You cite NHANES data for the fatty acid profiles of dairy products without
giving a reference.  In all likelihood, you are referring to the work of
Gladys Block.  Block reported on the NHANES II data in several papers, but
the fatty acid composition for all of NHANES II was and still is in great
error (17).  Much ice cream today, for example, is made with skim milk and
vegetable oil.

--continued next message--