Before attempting to provide some notes on the complex
subject of fats/oils, permit to start of stating what I think are at least
3 good rules of thumb with respect to the eating of refined oils or fats.
1. Vegetable fats are better than animal fats
2. Less the better
3. Avoid burning or overcooking of fats
The subject of fats in one's diet is indeed a complex story, and I am unable to do any justice to this confusing story except to list a few points and some good references for the intellectually curious to pursue further.
Now, with respect to this general query about which kinds of fats are preferable, I think we must be careful about categorical statements with respect to saturated fat (hereafter SF), polyunsaturated fat (hereafter PF), and monounsaturated fats (hereafter MF) per se simply because there are many kinds of individual fatty acids and they do not fall neatly into these 3 general classes. For instance, the n-3 and the n-6 polyunsaturated fatty acids have different functions, produce different eiscosanoids and interfere with or modify the metabolism of one another. The traditional PF/SF ratio also begins to lose some of its meaning when we consider that hydrogenated PF fats begin to act like SF's (e.g. they lie between PF and SF fats in their effects on raising serum cholesterol).
Hence, we have to be careful about confusing the different fatty acids via this overly-simplifying ratio of 2 classes of fats (variables) in what is turning out to be a very complex interplay of many fatty acids (i.e., a multivariate system).
Not all fats are equal and there appear to be some important health related differences between the sources of the "same" fats. For example, we have heard it repeated said that there is more saturated fat in some vegetable oils like palm oil than in lard. This is true but misleading as it stands. It is not the case with most plant oils but more importantly, what is not taken into account is that while both lard and palm oil are 99% fat vegetable oils do NOT contain cholesterol. Consider the following chart.
Composition per 100g SF MF PS PS/SF Cholesterol vitamin E lard 40.8g 43.8g 9.6g .23 93mg 0.0mg butter 54.0 19.8 2.6 .03 230 2.0 refined oils coconut 85.2 6.6 1.7 .02 0 .66 palm 45.3 41.6 8.3 .18 0 33.12 cottonseed 25.5 21.3 48.1 1.9 0 42.77 wheatgerm 18.8 15.9 60.7 3.2 0 136.65 soya 14.5 23.2 56.5 3.9 0 16.29 olive 14.0 69.7 11.2 .8 0 5.10 corn 12.7 24.7 57.8 4.6 0 17.24 sunflower 11.9 20.2 63.0 5.3 0 49.0 safflower 10.2 12.6 72.1 7.1 0 40.68 rapeseed 5.3 64.3 24.8 4.7 0 22.21
(Source, McCance & Widdownson's THE COMPOSITION OF FOODS,
5th ed., 1991)
While 93mg/100g of cholesterol in lard doesn't seem high, we need to keep in mind that fried rump steak has 82mg/100g and that cooked bacon has 270mg/100g. Hence, lard consists of a lot of cholesterol, and unlike the vegetable oils, lard has no vitamin E.
The key advantage of vegetable oils, as with all plant foods, is that they are CHOLESTEROL FREE. Also not to be overlooked is that vegetable oils contain VITAMIN E which serves as an anti- oxidant preventing the formation of free radicals from the polyunsaturated fats breaking down (as occurs especially during frying). Hence, as one attempts to increase one's proportion of PF's, one also needs to consume more vitamin E which occurs automatically if the PF's are from plant sources. Animal fats like beef lard, however, contains no vitamin E.
Having warned you about the traditional PF/SF ratio being a crude simplification, let us nevertheless consider one good study that supports the practical implications of this ratio. In one study patients in a Finish Mental Hospital (Turpeinen, 1979) were placed on the normal Finnish diet while patients situated in another Finish Mental Hospital were given a modified diet containing less saturated fatty acids and cholesterol and more linoleic acid. After 6 years the diets were switched between the two hospitals and at the end of the second 6 year period the results were published. For both hospitals the age-adjusted death rate from coronary heart disease in males during the high PF/SF diet was about 50% lower than during control, low PF/SF diet. This is a pretty good reduction, but not as good as what vegans achieve (about 90%.lower). In this study we here have a trapped population consuming less SF and more PF as achieved by substituting many of the meat and animal fats an with plant-based dishes and plant oils.
It is well-known that SF raises cholesterol levels. As one reviewer put it:
"the over-all conclusions from rather abundant data from both experimental and epidemiologic studies, are consistent in showing that saturated fatty acids elevate serum cholesterol and are the primary dietary factor while the polyunsaturated fatty acids lower serum cholesterol. . .the data fail to demonstrate any specific role for the monounsaturated fatty acids" (Hegsted, 1990)
Hegsted also thinks that their is some beneficial effect of linoleic acid on CHD.
To return to our famous PF/SF ratio, other studies have attempted to simply increase the PF without at the same time lowering the SF, but here the most of the results are disappointing (Hornstra, 1990). The best way to the PF/SF increase this ratio is simply by reducing the amount of SF. Consider the following table of U.S. food sources contributing to fat intake (Park & Yetley, 1990; Harris & Welsh, 1989):
Total dietary fat Saturated fat meat, poultry, fish dairy and eggs 50-55% 60% fats & oils 33% 10% grains, vegetables, 12%-17% 30% fruits, all other.
We should also note current consumption of tropical oils
(palm, palm kernel and coconut oils) represents less than 4% of total fat
intake, less than 2% of daily energy intake, and 8% or less of the SF intake.
The real culprit, therefore, is the animal products in the diet.
Hence, by virtue of becoming a vegan one should, in theory, reduce one's total fat intake by about 50% and reduce intake of saturated fat by 60%, thereby decreasing total fat while increasing the PF/SF ratio at the same time. By consuming more refined oils and nuts, however, vegans may not achieve this level of reduction in practice. For example, one British sample of 26 vegan males were consuming 33% of their calories as fat. Even so, this group of vegans still had a lower risk than their omnivorous counterparts, which suggests that plant based fats are somehow better than animal fats.
To complicate our neat PF/SF ratio, we need to be therefore reminded that not all saturated fats are equal. For instance, some experimental studies (Hornstra, 1990) have shown that palm oil, which we can see is high in saturated fat, does not seem to promote atherosclerosis but even helped to reduce it, even when compared to highly unsaturated oils like linseed or olive oil.
Now, while fats for most people help to enhance the flavour,
very little is actually needed and there soon comes a point where additional
oils only makes the food greasy and no longer serves to enhance the flavour.
Refined oils have become so cheap and readily available that we are now
probably overdoing it in our foods. If I may be permitted a more personal
observation, good fresh vegetables have their own subtle flavours and too
much oil and spice simply overwhelms these interesting flavours.
OK, with respect to health, it seems pretty well agreed upon by all authorities that we need to eat less fat, not more. How much less is still an ongoing controversy but as time goes on the recommended levels do keep going down. Many authorities are now considering 20% of caloric intake as fats to be a better target to aim for than the previous 30%. Maybe in time Pritikin's early recommendation of 10% will be more widely accepted, but those stubborn hamstrings and habits of the general populace for meat and fatty foods will probably persist for some time and be reflected in recommended reductions that are deem minimally necessary but maximally acceptable for a populace that will persist in consuming plenty of fats.
In general, Canadians and Americans consume about 40% of their calories in the form of fat. Many Chinese take only 3% of their calories as fat with no signs of fatty acid deficiency. Getting enough fat in one's diet is certainly not our problem. Rather, it is simply eating too much fat that is killing us or lessing the quality of living a fit life. For instance, the concentration of cholesterol in human organs and serum generally increases with the level of fat in the diet.
With respect to age-adjusted cancer rates, below are partial correlations (corrected for confounding social and biological variables) with components of dietary fat (Prentice & Sheppard, 1990):
breast colon prostate Total fat .72 .62 .69 SF .58 .47 .55 MF -.01 .004 .02 PF .51 -.01 .46 linoleic (n-6p) .49 .05 .48 linolenic (n-3p) .16 -.22 -.03
(from Henderson, 1991)
Total number of calories from fat gives us the largest positive correlation between breast, prostate & colon cancers. We can see that both SF and PF entail an increased risk while MF doesn't.
Rates of breast cancer has been increasing to the point that now 1 in 9 women will be affected, while only 20 years ago it was 1 in 18. In is interesting to observe that except for colon cancer, consumption of SF and PF entailed about the same increased risk for breast and prostate cancers, with PF entailing a slightly lower risk. Linolenic acid (soybean oil being one good source) seems to have some anti-cancer relation to colon cancer, although there is a small correlation with breast cancer. With respect to some major cancers, therefore, it would seem from these results that the less fat you consume the better.
Other studies have uniformly arrived at similar results (see Henderson, 1990, for a review). Studies mapping the change of risk for immigrants to a different country due to their adopting a different dietary practice will within a certain period of residency reach the same level of risk. For instance, Italian women with half the risk for breast cancer as Australians, or immigrants from countries with a higher risk of colon cancer (e.g., Scotland) will achieve the same risk as the Australian citizens of their new host country. National dietary surveys of these Australian migrants show them to generally cook their usual dishes but add more red meat to them (Australians have the highest per capita consumption of beef).
Excess PF has been shown to impair immunity (Vitale, 1988) but this is complicated by such factors as the type of fat (e.g., increasing hydrated density), duration of intake, type of immunological function, etc.
Typical frying temperature is about 400 degrees F and can
reach up to 600-700 degrees F. When fats/oils are heated to such temperatures
the CIS fatty acids are converted to TRANS fatty acids. This simply pertains
to how the molecule is turned but the consequence is that unsaturated fats
begin to behave like saturated fats with respect to their effects in raising
instead of lowering serum cholesterol levels (about 50% of the cholesterol
increasing effect of saturated fat) and can raise LDL cholesterol by nearly
as much as the saturated fat (Mensink & katan, 1990). Besides the extra
fat consumed, this is another reason why fried foods are more likely to
contributes to a hardening of the arteries. Also, partially hydrogenated
vegetable oils can vary in their TRANS isomer content from 5% to 45%.
When fat is reheated to frying temperatures a second time (as in deep fryers) the fat is more likely to develop cancer the producing agent acrolein and even benzopyrene (one of the most potent carcinogens known).
Very hot temperature also destroy certain vitamins and may alter the major proteins.
Hence, baking is preferable to frying since here the commonly used temperature is 350 degrees F.
Burning produces charring products that are carcinogenic. If one is quick frying and burning occurs, temperatures up to 1000 degrees F could have been reached. Even if one does not burn the oil or fat, overcooking (esp. when one reuses oils) breaks down the polyunsaturated molecule and free radicals can form. These are fragments that have combined with any available oxygen to produce toxic peroxides. They are toxic because that act as strong oxidizing agents that damage and destroy cells.
Now, for a practical solution. One actually doesn't need to fry with oil, water will suffice as long as one uses a good stainless steel frying pan and stirs more frequently. This is a trick I picked up from one of Pritikin's books; a book useful for such advice as to how to use less oil. For instance, a very good recipe for making some very tasty chips produced without deep frying or any oils whatsoever is to be found there.
A. In general there are 3 kinds of fats:
1. NEUTRAL FATS which are compounds of fatty acids with glycerin
2. PHOSPHOLIPIDS which are compounds of fatty acids with a
phosphate part (e.g., lecithin)
3. CHOLESTEROL which is a sterol used in producing various hormones as what we find being produced by the ovaries and adrenals. Since the human liver is very efficient at producing cholesterol, we really don't need any exogenous source.
B. Saturation, Unsaturation And Hydrogenation Of Fats
All fats are composes of chains of hydrocarbons (CH2).
e.g., stearic acid (saturated) CH3(CH2)16COOH
H H H H H H H H H H H H H H H H H 0H | | | | | | | | | | | | | | | | | | H-C-C-C-C-C-C-C-C-C-C-C-C-C-C-C-C-C-C=O | | | | | | | | | | | | | | | | | H H H H H H H H H H H H H H H H H
Here we see some hydocarbons (H-C-H) connected in a chain
very much like a segmented centipede, each separate segment of centipede
being one carbon atom with two legs per segment being available to attach
to other hydrogen atoms (as a foot is attached to the end of a leg). In
effect, each carbon have 4 electrical legs which can attach to other atoms.
Two of the legs of these carbons are used to attach themselves to adjacent
carbons thereby forming the chain. Now, two of these legs per carbon are
free, as are the two legs per segment in our centipede. When every free
carbon leg is attached to a hydrogen the molecule is said to be SATURATED
as we see in the above representation of steric acid.
If instead some of the carbon arms of a line of carbons doesn't attach to hydrogen but instead doubles back on itself to attach to one of the other carbons in the same chain of carbons, the molecule is said to be unsaturated. It is MONOUNSATURATED if only one carbon is involved in this `doubling back' onto the molecule chain. It is POLYUNSATURATED when two or more carbons are involved in this doubling back.
e.g., linoleic acid (note the 2 double bonds, represented by an `=' sign, indicates that we have a polyunsaturate fatty acid):
H H H H H H H H H H H H H H H H H 0H | | | | | | | | | | | | | | | | | | H-C-C-C-C-C-C-C-C-C=C-C-C=C-C-C-C-C-C=O | | | | | | | | | | | | | H H H H H H H H H H H H H
HYDROGENATION simply involves making these back carbon
arms that are doubled back to other carbon atoms in the molecule to instead
attach to a hydrogen. Hence, hyrdogenation of polyunsaturated fats involves
making these fats carry more hydrogen atoms there by producing "hard"
fats like margarine out of "soft" unsaturated fats of vegetable
C. Linoleic and linolenic fatty acids
The next edition of McCance & Widdowson's THE COMPOSITION OF FOODS will provide a breakdown of values for the individual fatty acids. Until then, here are some figures from the USDA handbook #8; Scherz & Kloor, FOOD COMPOSITION AND NUTRITION TABLES 1981/82.
Linoleic & Linolenic Fatty Acids in 100 g Edible Portions of Foods: P/S SF linoleic linolenic total cal. w6 w3 cream .05 18 .8 .2 320 egg yolk 7/9 9 .6 1.1 380 cod liver oil ? 15+ 2 1 900 chicken .6 5 2.8 .2 230 beef (choice) .1 12 1 0.0 300 beef tallow .06 50+ 2.8 .5 900 apple pie .3 3 1 0.0 250 soybean oil 5 12.7 53 7.5 900 safflower oil 9 8.5 74 .5 900 corn oil 5 10 53 0.0 900 sesame seed oil 3+ 13 42 1 900 walnut oil 7+ 9 57 13 900 walnuts 7+ 6 38 6 580 wheat germ oil 4+ 16+ 42 9.5 900
There is some support for the effects of polyunsaturated fat of the w6 type helping to lower blood pressure (Iacono, 1991) and the w3 type to help provide an anti-inflamatory effect and maybe help suppress some autoimmune disease (Robinson et al., 1991) Holman (1960) thought that the requirement for linoleic acid should be 1% of dietary energy (3% for those who eat a high fat diet). This is still an open question but it does seem that these fatty acids are becoming more important, and as we can see vegetable oils are a good source.
>Isn't it best to avoid palm and Coconut oils due to >fat/cholesteral/saturation - don't exactly remember which of the . >culprits it is, take your pick. Because of all the bad press, lots of >(here in the States) packaged baked goods sport the label >"No tropical oils". If this is not the case, please inform. Thanks. > >Laura
In general, it is best to reduce or delimit your intake of ALL fats, particularly saturated fats (SF), as obtained from refined oils and especially animal products. The Western diet is too abundant with fats.
There appears to be some important health related differences
between the sources of these fats. For example, it is true that there is
more SF fat in a palm oil (45%) and coconut oil (85%) than in lard (41%).
This, however, is misleading as it stands. What also must be taken into
account is that while both lard and palm oil are 99% fat, vegetable oils
do NOT contain cholesterol. A key advantage of vegetable oils, as with all
plant foods, is that they are CHOLESTEROL FREE. Also not to be overlooked
is that vegetable oils contain VITAMIN E which serves as an anti- oxidant
preventing the formation of free radicals from the polyunsaturated fats
(PF) breaking down (as occurs especially during frying). Hence, as one attempts
to increase one's proportion of PF, one also needs to consume more vitamin
E, which occurs automatically if the PF's are from plant sources. Animalfats
like beef lard, however, contains NO vitamin E.
We should also note current consumption of tropical oils (palm, palm kernel and coconut oils) represents less than 4% of total fat intake, less than 2% of daily energy intake, and 8% or less of the SF intake. The real culprit of our excessive intake of SF is, therefore, the animal products in the diet.
While it is generally true that one should consume more PF than SF fats (the famous PF/SF ratio), to complicate our neat PF/SF ratio, we need to be therefore reminded that not all SF's. For instance, some experimental studies have shown that palm oil, which we can see is high in saturated fat, does NOT seem to promote atherosclerosis but even helped to reduce it, even when compared to highly unsaturated oils like linseed or olive oil! (see Hornstra, 1990 `Effects of dietary lipids on some aspects of the cardiovascular risk profile'. In G. Ziant [ed.], LIPIDS AND HEALTH).
If you must fry your foods, another advantage of these
heavy tropical oils is that they are less likely to burn or be transformed
during frying. This is not to be lightly discounted.
Typical frying temperature is about 400 degrees F and can reach up to 600-700 degrees F. When fats/oils are heated to such temperatures the CIS fatty acids are converted to TRANS fatty acids. This simply pertains to how the molecule is turned but the consequence is that unsaturated fats begin to behave like saturated fats with respect to their effects in raising instead of lowering serum cholesterol levels (about 50% of the cholesterol increasing effect of saturated fat) and can raise LDL cholesterol by nearly as much as the saturated fat (Mensink & Katan, 1990). So, the advantage pf PF in this respect becomes no greater than that of SF. Besides the extra fat consumed, this is another reason why fried foods are more likely to contributes to a hardening of the arteries. When fat is reheated to frying temperatures a second time (as in deep fryers) the fat is more likely to develop cancer the producing agent acrolein and even benzopyrene (one of the most potent carcinogens known).
Very hot temperature also destroy certain vitamins and may alter the major proteins.
Burning of fats produces charring products that are carcinogenic. If one is quick frying and burning occurs, temperatures up to 1000 degrees F could have been reached. Even if one does not burn the oil or fat, overcooking (esp. when one reuses oils) breaks down the polyunsaturated molecule and free radicals can form. These are fragments that have combined with any available oxygen to produce toxic peroxides. They are toxic because that act as strong oxidizing agents that damage and destroy cells.
In sum, palm oil has been unduly maligned in the public mind as a "bad oil" when in fact it may not be that at all. Remember, one serving of ice cream has several times the saturated fat content of any processed food made with palm oil (including doughnuts, cookies, crackers, etc.), yet how often is ice cream mentioned as a significant source of saturated fat that should be reduced?
Hence, if one must fry with oils, using tropical oils is
preferable, maybe palm oil being best (not to be confused with palm kernel
oil which is about 81% SF. palm oil has a good oxidative stability during
frying due to both its FA composition and its tocotrienols (vitamin E).
Also, if one uses palm oil for frying one should not be consuming cholesterol
as one recent report indicated that palm oil in a very high cholesterol
diet makes for high serum cholesterol levels, but when the cholesterol was
omitted from the diets, serum cholesterol level were normal. Another reason
why palm oil is preferable over coconut oil. We should also note in passing
that using, say, soybean oil because one is desirous of consuming less SF
is offset by the fact that soybean oil is usually partially hydrogenated
when marketed (Canola oil is unhydrogenated). Such a substitution of trans
fatty acids for SF is not the best solution.
Baking, of course, is even more preferable to frying since here the commonly used temperature is 350 degrees F. There are many dishes that one can cook though baking instead of frying. For example, one can produce some very tasty and completely fat free "French Fries" by simply steaming the potatoes until soft, then cutting them up into chips and simply baking them on a glass baking pan (the pan does NOT need to be oiled, and chips are simple turned over once during baking to very lightly brown both sides). Alternatively, you can freeze the cut chips first and then bake if one prefers an extra mushy interior as we find with fried chips.
Now, for a practical solution for those who insist upon fried foods. First of all, one actually doesn't need to fry with oil, water will suffice as long as one uses a good stainless steel frying pan and stirs more frequently. This is a trick I picked up from one of Pritikin's books; a book useful for such advice as to how to use less oil.
If you find that you can yet master the skill of frying with water instead of oil, then try a mixture of half water and half oil. At least this way you reduce the oil by 1/2 and you help keep the oil from burning since the water via steaming helps to prevent those excessively hot temperatures (you may need to add more water as it steams off).
As I suggested in my previous posting, palm oil has been
unduly maligned in the public mind as a "bad oil" when in fact
it may not be that at all. Again, note that one serving of ice cream has
SEVERAL TIMES the saturated fat content of any processed food made with
palm oil (e.g., doughnuts, cookies, crackers, etc.), yet how often is ice
cream mentioned as a significant source of saturated fat that should be
A little history is here illuminating.
"It was primarily the ASA and Center for Science in the public interest that took up this issue, but with little impact until a wealthy man, who started his own foundation called the National Heart Savers Foundation, ran full-page newspaper advertisements singling out companies that were `poisoning America'. This individual had neither credential nor nutrition expertise to make this judgement, but the publicity caused sales of specific food products to drop, and processors removed oils termed tropical from foods. Americans overindulge in total energy, fat and saturated fat, but the impact of tropical oils on our diet is minimal. It is unlikely that complete substitution of unhydrogenated seeds oils for these oils could have a measurable effect on serum cholesterol levels. However, the impact on nutrition in the developing countries is not to be taken lightly. palm oil experts to the U.S. constitute only 3% of the world market for palm oil, and a substantial portion of that goes to nonfood uses such as olechemicals. But adverse publicity from the U.S. has caused many countries to question the safety of palm oil, which is the number one oil of international trade and is second in production in the world. Nutritionists around the world are questioning the safety of palm oil due to the adverse publicity in this country. Even in countries with heart disease prevalence half that in the U.S. and where serum cholesterol concentrations are low by Western standards, such as Costa Rica, pressure is being exerted to reduce consumption of palm oil. These events can only lead to increased costs, a decreased quality of the food supply, and subsequent hunger in areas of the developing world where there is a serious shortage of calories and fat"
(D. Klurfeld, `Tropical Oil Turmoil', J. OF THE AMER. COLLEGE
OF NUTRITION, 1991, 10(6): 575.
In general, I think even vegetarians should not over consume
fats and oils and indeed, maybe most could do with some further reduction.
Epidemiological studies (see A. Keys, SEVEN COUNTRIES. Cambridge: Harvard
U. Press, 1980; Shekelle et al. , 1981,`Diet, serum cholesterol, and death
from coronary heart disease.' N. ENGL. J. MED. 304:65-70) have obtain high
correlations (r=.80 and r=.84) between the average percentage of dietary
energy from SF/cholesterol and mortality from heart disease.
To help put things into perspective, we should remember that most of the saturated fats (SF) in the U.S. diet (see table below) is obtained from meat, poultry, fish and diary products (60%). By eliminating meat, fish and dairy, you eliminate 60% of the saturated fat in one's diet, 45% of total fats, and 100% of the cholesterol, thereby ideally achieving a higher PF/SF ratio while at the same time beneficially reducing total calories by fat and completely eliminating dietary cholesterol (which is not an essential dietary nutrient)
Contribution of Selected Food groups to Intake of Total Fat and Saturated Fatty Acids by Women Aged 19-50 Years in 1985. Food Group Total Fat SF meat, poultry fish 26 27 diary products 19 33 eggs 4 3 legumes, nuts, seeds 4 2 grain products 9 6 vegetables 6 4 fats and oils 30 21 sugars, sweets 1 2
More than 30 years of research has prompted even conservative organizations like the NCEP to conclude that SF was the most potent single dietary component in increasing plasma cholesterol, and that NO MORE THAN 10% of the energy of one's diet be SF.
total fat/calories (men and women aged 30-39) 37% total consumption of SF/calories 13%
Given these figures, we can see that *in theory*:
total fat SF lacto-ovo vegetarian 20% 5% vegan 18 5%
In practice, of course, this is not achieved because the
intake of fats from other sources would naturally increase, but not by much
if the sources were simply more grains, legumes and vegetables.
Consider the trends from the following study of 137 Londoners who had radically changed their diet to become "vegetarians" for neither ethnic nor religious reasons. (from Draper & Wheeler, 1990, "What do `vegetarians' eat?" PROCEEDINGS OF THE NUTRITION SOCIETY, 49, 60A).
total Fat PF SF Male: vegans 34% 12% 7% lacto-ovo vegetarian 37% 8% 12% demi-vegetarians 38% 8% 14% Female vegans 33% 12% 8% lacto-ovo vegetarian 37% 9.7% 12% demi-vegetarians 39% 10% 7%
Note: demi-vegetarians are here defined as the 37 self-acclaimed
vegetarians who usually avoided fish meat or both.
A slightly larger sample of 52 vegans, vegetarians, fish eaters and meat eaters were assessed as follows (from Thorogood et al., 1990, `Dietary intake and plasma lipid levels: lessons from a study of health conscious groups' BMJ, 300 (May 19):1297-301):
DIET GROUP carbohydrate total fat PF/SF ratio SF Males vegan 52.5% 33.5% 1.85 6.2% vegetarian 47.7% 36.4% .73 12.1% fish eater 43.8% 38.2% .73 12.5% meat eater 43.0% 38.1% .56 13.2% Females vegan 51.4% 36.2% 1.77 7.4% vegetarian 46.4% 39.6% .63 14.3% fish eater 42.9% 40.5% .75 13.3% meat eater 43.2% 38.7% .49 14.2%
Now, it so be noted that our meat-eaters in this sample
are "health conscious" (as indicated by questionnaires), hence
the relatively low intake of SF compensated by a higher intake of PF (and
an higher intake of fiber) in our meat-eaters than is normally found in
the average British meat-eating diet. Still, everyone is eating too much
fat and ONLY the VEGANS are achieving the a SF intake below 10% and a PF/SF
ratio greater than 1.
In sum, only the vegans come close to achieving the conservative recommendations of no more than 10% of one's calories in the form of SF and total fats being no more than 30%.
Maybe it is these particular samples of Londoners who are still attached to the greasy, over-cooked liquidating of food that is one of the more notorious characteristics of British cuisine ;-)
Whatever is the problem, maybe even vegans should consider reducing any excessive use of refined oils. Lacto-ovo vegetarians, however, should definitely give serious thought to reducing their dairy/egg intake, as well as, the refined oils.
Ideally, everybody should be eating more carbohydrate to replace the energy from SF. Instead, the Thorogood study cited above showed that the non-vegan health conscious groups were achieving some reduction in SF (but not enough), but instead of substantively increasing their carbohydrate intake to maintain energy intake levels, they increased their PF intake. They did, however, increase their fiber intake by eating wholemeal breads and pasta.
Now, if I might make a more personal observation. Why is it that people insist upon sliming their bread with butter or margarine? Good wholewheat bread is very tasty on its own without greasing its upper surface. Maybe people need to be purchasing better quality wholewheat bread, among other things? ;-)
Remember, things like lard and refined vegetables oils are 99% fat, but at least vegetable oils, like all food plants, do NOT contain cholesterol. Consider the following chart.
Composition per 100g SF MF PS PS/SF Cholest. vitamin E lard 40.8g 43.8g 9.6g .23 93mg 0.0mg butter 54.0 19.8 2.6 .03 230 2.0 Refined Oils coconut 85.2 6.6 1.7 .02 0 .66 palm 45.3 41.6 8.3 .18 0 33.12 cottonseed 25.5 21.3 48.1 1.9 0 42.77 wheatgerm 18.8 15.9 60.7 3.2 0 136.65 soya 14.5 23.2 56.5 3.9 0 16.29 olive 14.0 69.7 11.2 .8 0 5.10 corn 12.7 24.7 57.8 4.6 0 17.24 sunflower 11.9 20.2 63.0 5.3 0 49.0 safflower 10.2 12.6 72.1 7.1 0 40.68 rapeseed 5.3 64.3 24.8 4.7 0 22.21
(Source, McCance & Widdownson's THE COMPOSITION OF FOODS,
5th ed., 1991)
Now, of all the refined oils, coconut oil unfortunately has the highest percentage of saturated fat (SF), and with respect to heart disease and cancer this is the fat that is most worrisome.
Having said that, we need to keep things in perspective because I do think that people get their health priorities frequently upside down at time and over-react by completely avoiding what CAN be a proper and judicious use of even coconut oil.
Todays' cheap, commercially produced refined vegetable
oils are NOT like yesterday's cold pressed unrefined vegetable oils. Instead,
the oils are produced first breaking down the seeds by cooking (up to 2
hours in temp. around 120 degrees C.) and then EXPELLER PRESSING out the
oil which again heats the oils (usually around 85 to 95 degrees C.). Alternatively,
the oil from seeds may be removed via a SOLVENT EXTRACTION (solvents like
hexane are used) leaving of course traces of the solvent behind in the extracted
oil. We now have "unrefined oil".
To refine them they are first "DEGUMMED" via water and phosphoric acid at 60 degrees C. so as to gummy protein-like compounds and some complex carbohydrates are removed. In the process, however, some nutrients like lecithin and phospholipids are also removed as well as chlorophyll, calcium, magnesium, iron, and copper.
"REFINING" the oils further will involve, at temperature around 75 degrees C, mixing them with caustic soda (Draino!) and maybe some sodium carbonate which is latter removed. The purpose here being to remove those "free" fatty acids from the oil. The presence of such free fatty acids indicates poor oil quality. In this process certain protein-like substances and minerals are further removed.
At this stage, the oil still contains red or yellow pigments (chlorophyll and beta-carotene). Heaven forbid that we don't have a clear, colourless liquid! Hence, the oil must undergo BLEACHING at 110 degrees C. using filters, Fuller's earth and/or acid-treated activation clays. Some toxic peroxides and conjugated fatty acids are formed.
Having got rid of those nasty colours, the next stage it to DEODORIZE via steam pressure and steam distillation. Besides removing some pungent odours and unpleasant tastes (which were *not present* in the natural oil before all this processing began!), we have here also removed some aromatic oils and free fatty acids. Here, some very high temperatures are used (240 to 270 degrees C.) for about 30 to 60 minutes. Remember, above 160 degrees C. we begin to form those nasty `trans' fatty acids. This stage will remove most of those toxic peroxides produce earlier in the refining, some pesticide residues but also some of the tocopherols (vitamin E) and phytosterols.
Now, what do we finally have? A orderless and tasteless oil that cannot be distinguished from other quite different oils similarly treated. The high temperatures have produced some unnatural isomers of the fatty acids (e.g., the "trans" configuration of the molecules). The resulting product is also now vitamin and mineral deficient compared to the original cold-pressed, unrefined oil. Indeed, we have to here be careful with these terms since "cold-pressed" can still refer to this refined product as long as no "external" heat was applied during the one stage of expeller pressing the oil!
If this isn't bad enough, supermarket oils now replace the natural antioxidants (the vitamin E and beta-carotene) that have been diminished in the refining process by now adding butylated hydoxytoluene (BHT), butylated hydroxyanisole (HGA), propyl gallate, tertiary butyhdroquinone (BHQ), citric acid, and methylsilicone. Finally, a bit of DEFOAMER is added and the oil is now bottled in those useless clear glass bottles. Maybe some "winterization" via a cooling and filtering so as to prevent it going cloudy in the fridge. Finally, this marvel of refining is sold to an uninformed consumer who insists upon such tasteless and nutrient deficient oils as long as it is clear and keeps for long periods of time.
To add insult upon injury, some of the very substances that are taken out of the oils are than sold as separate "health" products, like soy lecithin.
Now, you can, to some extent, buy in health food stores some oils not so overly refined, but they are more expensive and often will simply not keep as long (about 3 months). Still, we don't expect fruits and vegetables to last a year why should we expect the vegetable oils to last so long in our households? Now, unless we the consumer begins to demand quality foods by protesting with their pocket book, we really cannot expect the quality of our foodstuffs is be improved.
First of all, we need to consider which dietary components
are the leading contributors to our excessive intake of SF. No use eliminating
or reducing minor dietary components while ignoring that bulk of the SF
is still not being addressed.
Consider the following table of U.S. food sources contributing to fat intake (Park & Yetley, 1990; Harris & Welsh, 1989):
Total dietary fat Saturated fat meat, poultry, fish dairy and eggs 50-55% 60% fats & oils 33% 10% grains, vegetables, 12-17% 30% fruits, all other.
Now, consider the further breakdown per capital consumption of fat in the UK (from "Household Food Consumption and Expenditure, HMSO, London):
Food % total fat PF MF SF PF/SF milk/creme 15.6 .4 5.0 10.4 .04 cheese 4.1 .1 1.3 2.7 .04 meat 27.0 2.5 12.7 12.5 .2 fish 1.0 .4 0.4 0.3 1.3 eggs 3.0 .4 1.4 1.2 .33 butter 18.7 .5 5.6 11.6 .04 margarine 8.1 1.7 3.4 3.1 .5 cooking fats 10.3 1.7 4.8 3.9 .4 vegetable 1.5 .7 0.6 0.2 3.5 fruits 0.5 .1 0.2 0.2 .5 bread 1.4 .5 0.4 0.4 1.24 cakes, pastries 2.5 .5 1.1 1.1 .55 biscuits 4.3 .3 1.9 1.9 .15 other cereals 1.2 .3 0.5 0.5
The first source of worry should be animal products and
this also applies to lacto-ovo vegetarians. Meat being the first major source
of fat, then butter, then milk/creme and then the cooking fats (unfortunately,
this British breakdown probably includes lard among the cooking fats).
In the US diet, grains, vegetables, etc. is about equal to fats and oils being in volume the second main source of SF, but at least the PS/SF ratios in grains/vegetables are high and this is probably the best source for our intake of fatty acids. We don't really need the refined oils, but they do help to enhance the flavour of some foods (if used moderately, otherwise they simply make the food greasy and overwhelm the more subtle flavours of whole foods).
After the meat and diary products, the refined fats and oils then should be something we consider reducing or eliminating, especially where they are often used unnecessarily.
Just as meat was a once coveted luxury product that now is too easily affordable, I wonder if refined oils being so very cheap and readily available are not simply be being over-used. Just because a little is good we seem to overdo things in thinking that a lot must be better. My experience with Indian cooking here in Vancouver is that they do over-use the refined oils and like the British, like to aggress against fresh vegetables by over-cooking them ;-( To take another example, traditional Chinese cooking with a wok would use water before any oil was added, that does not seem to be the case here in North America where refined oils are used so liberally.
Now, if coconut oil were to be used sparing, in an otherwise, low fat diet, then this would probably be beneficial.
These things are simply a molecule or element with an unpaired
electron. This free electron is not tied up in a bond or an electron shell
with another electron, it is therefore loose and very active. Like any unattached
young adult, it is looking for a partner but unlike most considerate humans
this unpaired electron is willing to pair up with almost anybody and has
no compunction about splitting up other pairs of happy cohabiting pair electrons
to steal a mate. Normally, the body has ways of keeping in check the free
radicals by keeping their radical reactions and localized, after all, they
do serve as intermediates in many thousands of normal chemical reactions.
Light and oxygen produces free radical reactions and destroys the oil thereby. Bringing the oil to high temperatures accelerates this process. What happens is that photon of light gets caught by an electron on a carbon next to one of the double bonds of carbon in a molecule of unsaturated fatty acid. This election now carries more energy than before and in its excited state it can split off from the molecule (perhaps with a hydrogen nucleus) leaving behind a free electron not bonded with another electron and therefore quite hungry for a new partner. It will grab an electron partner wherever it can consequently leaving on that hapless molecule an free electron which will grab an electron from somewhere else, and so on. A typical chain reaction like this can take up to 60,000 cycles before it is stopped. Since there are billions of photons even on a cloudy day, it is understandable why unsaturated oils soon become altered, denatured and spoiled. The same happens to oxygen (hence, oxygen radicals) which then will steal from the fatty acid starting up the old chain reaction.
In natural oils (as well as the human body), there are molecules like vitamin E which are able to trap the free electrons and there are singlet oxygen "quenchers" like carotene. As long as the body is replete in its natural free radical protector substances (vitamin E, A, C, and some of the B complexes (B1, B5 and B6), sulphur- containing amino acid cysteine, zinc, selenium, some naturally produced cyclic (phenolic) compounds, and some other substances, the free radical damage to membranes can be prevented very efficiency. The rub is that these substances need to be provided in our foodstuffs in optimal quantities.
It doesn't make sense to decrease the intake of essential fatty acids, but it does make sense to increase the intake of these protective substances which ought to be already abundantly present in quality oils. The problem is that our over-refined or over-cooked foodstuffs are often depleted of some of these substances. Also, it is possible that the presence of so many synthetic substances in our foodstuffs, substances that in some ways may mimic the more natural substances can thereby produce "misfitted" products whose free radical reactions may not be as readily confined and controlled as the more natural substances. Now, the problems of frying result from the oxidation that takes place when oils are subjected to high temperatures in the presence of oxygen and light. This produces a great amount of free radical activity. Other chemical changes also take place in frying including the production of `trans' fatty acids, some polymer products whose effects on health are not yet known, and if burning occurs then some charring products that are carcinogenic will be produced. When fat is reheated to frying temperatures a second time (as in deep fryers) the fat is more likely to develop cancer the producing agent acrolein and even benzopyrene (one of the most potent carcinogens known).
Coconut oil is mostly saturated fat and even the little unsaturate fat is "saturated" since commercial coconut oil is often hydrogenated. Indeed, coconut butter is almost all saturated fat. The advantage of such a high saturated fat content is that light, air and heat only destroy it slowly. This is why such highly saturated fats are useful for frying.
With respect to heart disease, the worry here is that SF
raises serum cholesterol levels while PF can a cholesterol lowering effect.
Coconut oil, unfortunately, has the lowest PF/SF ratio. Ideally, these ratios
should be around 2 or greater since the cholesterol lowering effect of PF
is about half that of the cholesterol raising effect of SF.
Now, we must here thicken the discussion by noting that not all SF are equally hypercholesterolemic, especially those containing lots of stearic acid. It turns out that cocoa butter is less hypercholesterolemic than butter fat even though the saturation of the two fats is very similar. Cocoa butter is very rich in stearic acid (34.5% of total calories while coconut oil is only 2.5% total calories via stearic acid).
Indeed, SF is a crude differentiation of fatty acids and we should here consider a more analytic analysis. Now, I don't wish to get overly technical nor scare people with a bit of biochemistry (I am no biochemist myself), but a quick scan of the table below shows that indeed we are dealing with many different kinds of fatty acids and we should be careful about crude groupings like the SF and PF distinction.
common name # of # of melting typical carbon double point source atoms bonds SF ACIDS butyric 4 0 -7.9 butterfat capriic 6 0 -3.4 butterfat caprylic 8 0 16.7 coconut oil capric 10 0 31.6 coconut oil lauric 12 0 44.2 coconut oil myristic 14 0 54.4 butterfat/coconut palmitic 16 0 62.9 most fats/oils stearic 18 0 69.6 most fats/oils arachidic 20 0 75.4 peanut oil behenic 22 0 80.0 peanut oil PS ACIDS caproleic 10 1 -- butterfat lauroleic 12 1 -- butterfat myristoleic 14 1 18.5 futterfat palmitoleic 16 1 -- beef/some fish oleic 18 1 16.2 most fats/oils elaidic 18 1* 43.7 butterfat/HVO vaccenic 18 1* 44.0 butterfat/HVO linoleic 18 2 -6.5 most veg. oils linolenic 18 3 -12.8 soybean, canola gadoleic 20 1 -- some fish oils arachidonic eicosatetraenoic 20 4 -49.5 lard eicosapentaenoic 20 5 -- some fish oils erucic docosenoic 20 1 33.4 canola oil docosahexaenoic 22 6 -- some fish oils * these double bonds are in the `trans' configuration, all others are in the `cis' configuration. HVO = hydrogenated vegetable oils
Well, I hope this table didn't scare off too many readers, but the point is that there are many fats, oils containing mostly these simply fatty acids. Strictly speaking, only linoleic acid is essential and cannot be synthesized in the body. n-3 (older term being omega-3) fatty acids, like linolenic fatty acid (alpha-linolenate) may also be essential (see Neuringer & Conner, "n-3 fatty acids in the brain and retina: Evidence for their essentiality." NUTR. REV., 1986, 44:285-94). I believe that coconut oil is very low in these "essential" fatty acids, but I don't have any figures at hand. It is know that the best sources for linoleic acid is corn, safflower, soybean, cottonseed, sunflower seed, and peanut oils. Alpha-linolenic is most abundantly provided in linseed, soybean, and other seed oils, as indicated in the table below.
Linoleic and linolenic fatty acids in 100 g edible portions of foods: P/S SF linoleic linolenic total cal. w6 w3 cream .05 18 .8 .2 320 egg yolk 7/9 9 .6 1.1 380 cod liver oil ? 15+ 2 1 900 chicken .6 5 2.8 .2 230 beef (choice) .1 12 1 0.0 300 beef tallow .06 50+ 2.8 .5 900 apple pie .3 3 1 0.0 250 soybean oil 5 12.7 53 7.5 900 safflower oil 9 8.5 74 .5 900 corn oil 5 10 53 0.0 900 sesame seed oil 3+ 13 42 1 900 walnut oil 7+ 9 57 13 900 walnuts 7+ 6 38 6 580 wheat germ oil 4+ 16+ 42 9.5 900
There is some support for the effects of polyunsaturated
fat of the w6 type helping to lower blood pressure (Iacono, 1991) and the
w3 type to help provide an anti-inflamatory effect and maybe help suppress
some autoimmune disease (Robinson et al., 1991) Holman (1960) thought that
the requirement for linoleic acid should be 1% of dietary energy (3% for
those who eat a high fat diet). This is still an open question but it does
seem that these fatty acids are becoming more important, and as we can see
vegetable oils are a good source.
Now, to return to our concern about which fatty acids are shown to have the most cholesterol raising effect, the two saturates -- palmitic acid and myristic acid -- clearly raise LDL cholesterol levels. Maybe these are the two fatty acids we should be most concerned about.
Palmitic acid contents of common fats and oils Fat/Oil % of total calories palm oil 45.1 beef tallow 26.5 lard 24.8 chicken 23.2 cocoa butter 25.8 cottonseed oil 24.7 butter fat 26.2 coconut oil 8.4 palm kernel oil 8.0 corn oil 12.2 peanut oil 11.6 rapeseed oil 3.6 soybean oil 11.0
(from Grundy, "Which saturated fatty acids raise plasma
cholesterol levels" In G. Nelson (ed.), HEALTH EFFECTS OF DIETARY FATTY
Palmitic acid is the major saturated fatty acid consumed (about 2/3 to 3/4 of saturated fats in the average Amer. diet) and we can see that beef tallow, lard and butter fat would be the main sources in the average omnivorous diet. There is no doubt that palmitic acid raises serum LDL cholesterol compared to the unsaturated fatty acids. A well-known equation by Keys is to estimate that each 1% of dietary calories in the form of Palmitic acid raises serum total cholesterol by 2.7 mg/dl. Hegsted found the increase to be 2.16. Many other studies have consistently confirmed these figures.
Myristic Acid Contents Of Common Fats And Oils Fat/Oil % of total calories palm oil 0 beef tallow 3.3 lard 1.5 chicken 1.3 cocoa butter 0.1 cottonseed oil .9 butter fat 11.7 coconut oil 17.6 palm kernel oil 16.0 corn oil 0 peanut oil 0.1 rapeseed oil 0.1 soybean oil 0.1
According to Keys at al. (METABOLISM, 1965, 14:776) myristic
acid raises cholesterol levels by about the same amount as does palmitic
acid. The normal diet, however, contains much less myristic acid than palmitic
One other major fatty acid, lauric aid, seems to be less problematic than either palmitic or myristic acid. It would seem to still raise cholesterol levels, but the research here is less definitive and the average levels raised are lower than what occurs from palmitic or myristic acid.
Lauric acid contents of common fats and oils Fat/Oil % of total calories palm oil 1.1 beef tallow .01 lard .01 chicken 0.2 cocoa butter 0 cottonseed oil 0 butter fat 3.1 coconut oil 48.5 palm kernel oil 49.6 corn oil 0 peanut oil 0 rapeseed oil 0 soybean oil 0
Now, with respect to coconut oil, we can see that it is
high in both myristic acid and lauric acid but at least it is low in palmitic
acid. Lard is worst, but one study (Reiser et al., 1985, AM.J. CLIN. NUTR.,
42:190), however, found coconut oil to raise cholesterol levels at least
as much as beef fat itself (but it relatively high in stearic acid, about
21.6% of total calories). Of course, this is only one study but it may indicate
that even lauric acid under some circumstances have a serious cholesterol
Anyway, I hope the point is here taken that not all saturated fatty acids have the same effects on the metabolism of cholesterol and lipoproteins. Remember, some studies (Hornstra, 1990) have shown that palm oil, which we can have see to be high in saturated fat, does not seem to promote atherosclerosis but even helped to reduce it, even when compared to highly unsaturated oils like linseed or olive oil. The other point to observe is that things really are complex and one must weigh a number of factors both pro and con before arriving at any final comparison of the different fats. On the whole, I do think that the evidence supports the view that animal fats are overall much worst than vegetable fats and that we all could do with eating less fats whatever their source.
With respect to age-adjusted cancer rates, below are partial
correlations (corrected for confounding social and biological variables)
with components of dietary fat (Prentice & Sheppard, 1990):
breast colon prostate Total fat .72 .62 .69 SF .58 .47 .55 MF -.01 .004 .02 PF .51 -.01 .46 linoleic (n-6p) .49 .05 .48 linolenic (n-3p) .16 -.22 -.03
(from Henderson, 1991)
Total number of calories from fat gives us the largest positive correlation between breast, prostate & colon cancers. We can see that both SF and PF entail an increased risk while MF doesn't.
Rates of breast cancer has been increasing to the point that now 1 in 9 women will be affected, while only 20 years ago it was 1 in 18. In is interesting to observe that except for colon cancer, consumption of SF and PF entailed about the same increased risk for breast and prostate cancers, with PF entailing a slightly lower risk. Linolenic acid (soybean oil being one good source) seems to have some anti-cancer relation to colon cancer, although there is a small correlation with breast cancer. With respect to some major cancers, therefore, it would seem from these results that the less SF and even PF you consume the better. In general, we should note that it is fat en% rather than SF or PS en% that is more positively correlated with cancer. The Oct, 1992 edition of Consumer Reports reviewed some the literature and surveyed 94 nutrition professionals. Their conclusion was that we ought to "keep total fat intake at or below 20% to 25% of calories and SF at 7% calories". Certainly, this is better than the convention 30% recommendation and it comes close to the 15% to 20% figure that I personally think is the ideal.
Excess PF has been shown to impair immunity (Vitale, 1988 "Lipids, host defenses and immune function. In Beare-Rogers (ed.), DIETARY FAT REQUIREMENTS IN HEALTH AND DEVELOPMENT.) but this is complicated by such factors as the type of fat (e.g., increasing hydrated density), duration of intake, type of immunological function, etc.
In general, I think the less one consumes of these over-refined
oils, the better. In fact some authorities (see Harwood et al., "Medical
and agricultural aspects of lipids" in Junstone et al., THE LIPID HANDBOOK,
1986) argue that simply replacing saturated fats with unsaturated fats confers
no added protection unless there is also a reduction in total fat intake.
Things may work as a whole and this consideration does not run counter to
view that the relative dietary deficiency of essential fatty acids may contribute
to the pathogenesis of ischaemic heart disease (Oliver, 1979, BRIT. MED.
If you MUST fry in oil, then either palm or coconut oil is preferred. Do use them, however, in moderation. They contain almost no essential fatty acids. Ideally, you should be frying with oils that do contain essential fatty acids, but this takes more skill on our part than the usual quick and dirty frying that most people seem content with maintaining. Again, I would recommend that you try frying only with water, or at the very least, place your vegetables in the pan before adding the oil so as to a least help the oil a little from overheating and oxidation.
With respect to your query about raw coconut, this is not the same as the over-refined oils and is a much better substance to consume than would be refined coconut oil. A creamed block of dried coconut is 68.8% fat, desiccated coconut is 62% fat, while coconut oil, like all refined vegetable oils, is 99.9% fat. I wouldn't worry about raw coconut unless you are simply eating lots of it. :-) Its the unnatural, denatured over-refined oils that I think are to be reduced somewhat in most people's diet.
A easy and readable book I would recommend for one wishing
to first understand something about fats/oils, is:
Udo Erasmus FATS AND OILS: THE COMPLETE GUIDE TO FATS AND OILS IN HEALTH AND NUTRITION, Vancouver, Canada: Alive Books, 1986.
For the more technically inclined, there has been a series of papers during the last couple of year on the specific oils in the JOURNAL OF AMERICAN COLLEGE OF NUTRITION.
Hornstra, G. (1990). Effects of dietary lipids on some aspects of the cardiovascular risk profile. In G. Ziant (ed.), LIPIDS AND HEALTH, 1990.
Turpeinen et. al.(1979). INT. J. EPIDMIOL. 8:99-118.
Park & yetley (1990). AM. J. CLIN. NUTR., 51:738-48.
Harris & Welsh, 1989). NUTRITION TODAY 24(6):20-8.
Prentice & Sheppard (1990). Dietary fat and cancer: Consistency of the epidemiologic data, and disease prevention that may follow from a practical reduction in fat consumption.
CANCER, CAUSES AND CONTROL, 1:81-97):
Henderson, Maureen (1990). Correlations between fatty acid intake and cancer incidence. In G. Nelson (ed.), HEALTH EFFECTS OF DIETARY FATTY ACIDS.
Hegsted, d. (1990). Dietary fatty acids, serum cholesterol and coronary heart disease. (ibid.)
Robinson et al. (1991). Suppression of autoimmune disease by purified n-3 fatty acids. (ibid).
Iacono, J. (1990). The effect of w6 dietary fatty acids on blood pressure. (ibid) Mensink & Katan (1990). N. ENGL. J. MED.,
Vitale, J. (1988). Lipids, host defenses and immune function. In Beare-Rogers (ed.), DIETARY FAT REQUIREMENTS IN HEALTH AND DEVELOPMENT.