ArticlePDF AvailableLiterature Review

Omega-3 fatty acids in health and disease and in growth and development

  • Center for Genetics Nutrition and Health

Abstract and Figures

Several sources of information suggest that man evolved on a diet with a ratio of omega 6 to omega 3 fatty acids of approximately 1 whereas today this ratio is approximately 10:1 to 20-25:1, indicating that Western diets are deficient in omega 3 fatty acids compared with the diet on which humans evolved and their genetic patterns were established. Omega-3 fatty acids increase bleeding time; decrease platelet aggregation, blood viscosity, and fibrinogen; and increase erythrocyte deformability, thus decreasing the tendency to thrombus formation. In no clinical trial, including coronary artery graft surgery, has there been any evidence of increased blood loss due to ingestion of omega 3 fatty acids. Many studies show that the effects of omega 3 fatty acids on serum lipids depend on the type of patient and whether the amount of saturated fatty acids in the diet is held constant. In patients with hyperlipidemia, omega 3 fatty acids decrease low-density-lipoprotein (LDL) cholesterol if the saturated fatty acid content is decreased, otherwise there is a slight increase, but at high doses (32 g) they lower LDL cholesterol; furthermore, they consistently lower serum triglycerides in normal subjects and in patients with hypertriglyceridemia whereas the effect on high-density lipoprotein (HDL) varies from no effect to slight increases. The discrepancies between animal and human studies most likely are due to differences between animal and human metabolism. In clinical trials eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) in the form of fish oils along with antirheumatic drugs improve joint pain in patients with rheumatoid arthritis; have a beneficial effect in patients with ulcerative colitis; and in combination with drugs, improve the skin lesions, lower the hyperlipidemia from etretinates, and decrease the toxicity of cyclosporin in patients with psoriasis. In various animal models omega 3 fatty acids decrease the number and size of tumors and increase the time elapsed before appearance of tumors. Studies with nonhuman primates and human newborns indicate that DHA is essential for the normal functional development of the retina and brain, particularly in premature infants. Because omega 3 fatty acids are essential in growth and development throughout the life cycle, they should be included in the diets of all humans. Omega-3 and omega 6 fatty acids are not interconvertible in the human body and are important components of practically all cell membranes. Whereas cellular proteins are genetically determined, the polyunsaturated fatty acid (PUFA) composition of cell membranes is to a great extent dependent on the dietary intake.(ABSTRACT TRUNCATED AT 400 WORDS)
Content may be subject to copyright.
Review Article
Am J C/in Nutr 1991:54:438-63. Printed in USA. © 1991 American Society for Clinical Nutrition
Omega-3 fatty acids in health and disease
and in growth and development4
Arternis P Simopoulos
ABSTRACT Several sources of information suggest that
man evolved on a diet with a ratio ofw6 to w3 fatty acids of 1
whereas today this ratio is 10:1 to 20-25:1, indicating that
Western diets are deficient in w3 fatty acids compared with the
diet on which humans evolved and their genetic patterns were
established. Omega-3 fatty acids increase bleeding time; decrease
platelet aggregation, blood viscosity, and fibnnogen: and increase
erythrocyte deformability, thus decreasing the tendency to
thrombus formation. In no clinical trial, including coronary ar-
tery graft surgery, has there been any evidence of increased blood
loss due to ingestion of w3 fatty acids. Many studies show that
the effects of w3 fatty acids on serum lipids depend on the type
of patient and whether the amount of saturated fatty acids in
the diet is held constant. In patients with hyperlipidemia, w3
fatty acids decrease low-density-lipoprotein (LDL) cholesterol
if the saturated fatty acid content is decreased, otherwise there
is a slight increase, but at high doses (32 g) they lower LDL
cholesterol: furthermore, they consistently lower serum triglyc-
erides in normal subjects and in patients with hypertriglycer-
idemia whereas the effect on high-density lipoprotein (HDL)
varies from no effect to slight increases. The discrepancies be-
tween animal and human studies most likely are due to differ-
ences between animal and human metabolism. In clinical trials
eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA)
in the form of fish oils along with antirheumatic drugs improve
joint pain in patients with rheumatoid arthritis: have a beneficial
effect in patients with ulcerative colitis: and in combination with
drugs, improve the skin lesions, lower the hyperlipidemia from
etretinates, and decrease the toxicity of cyclosporin in patients
with psoriasis. In various animal models w3 fatty acids decrease
the number and size of tumors and increase the time elapsed
before appearance oftumors. Studies with nonhuman primates
and human newborns indicate that DNA is essential for the
normal functional development of the retina and brain, partic-
ularly in premature infants. Because w3 fatty acids are essential
in growth and development throughout the life cycle, they should
be included in the diets of all humans. Omega-3 and w6 fatty
acids are not interconvertible in the human body and are im-
portant components ofpractically all cell membranes. Whereas
cellular proteins are genetically determined, the polyunsaturated
fatty acid (PUFA) composition ofcell membranes is to a great
extent dependent on the dietary intake. Therefore appropriate
amounts ofdietary w6 and w3 fatty acids need to be considered
in making dietary recommendations, and these two classes of
PUFAs should be distinguished because they are metabolically
and functionally distinct and have opposing physiological func-
tions. Their balance is important for homeostasis and normal
development. Canada is the first country to provide separate
dietary recommendations for w6 and w3 fatty acids. Am J
Clin Nutr 199 1:54:438-63.
KEY WORDS Polyunsaturated fatty acids, w3 fatty acids,
w6 fatty acids, lipids, ct-linolenic acid, eicosapentaenoic acid,
docosahexaenoic acid, essentiality in growth and development,
cardiovascular disease, hypertension, inflammation, arthritis and
other autoim m une disorders, psoriasis, cancer, prostaglandins,
leukotrienes, interleukins, platelet-derived growth factor, en-
dothelium-derived relaxing factor
Omega-3 and w6 fatty acids: sources, elongation,
and desaturation
Evolutionary aspects: the w3 and u6 fatty acid balance
Large-scale production of vegetable oils
Agribusiness and modern agriculture
Imbalance of w6:3
Biological effects ofw3 fatty acids in relation to coronary
heart disease and hypertension
Eicosanoid metabolism
Molecular aspects and gene expression:
beyond the eicosanoids
Hypolipidemic effects
Effects on normal subjects
Effects on patients
Antiatheromatous actions
Antithrombotic effects
Vascular effects
Antiarrhythmic effects
Effects on restenosis
Effects on lipoprotein (a)
Additional effects
I From the Center for Genetics, Nutrition and Health, Washington,
2 Supported in part by The Council for Responsible Nutrition (CRN),
Washington, DC.
3 This review paper was prepared at the request of CRN and was
submitted to the Food and Drug Administration by the CRN on De-
cember 19. 1990.
4 Address reprint requests to AP Simopoulos, The Center for Genetics,
Nutrition and Health. 2001 S Street. NW, Suite 530, Washington, DC
Received February 27, 1991.
Accepted for publication March 20, 1991.
Coronary heart disease
Inflammatory and autoimmune disorders
Ulcerative colitis
Omega-3 fatty acids as an adjuvant to drug therapy
Essentiality: the role of w3 fatty acids in growth
and development
Animal studies
Human studies
Fetal development, human milk, and infant feeding
Children, adults, and elderly adults
Dietary implications
In the 1950s many investigators studied the effects ofcorn oil
and fish oil in humans primarily by determining their effects on
serum cholesterol concentrations in patients with atherosclerosis
(1-5). Corn oil (w6 fatty acid), an odorless, clear oil, was found
to lower cholesterol, particularly when it replaced butter or lard
in the diet. Although sardine oil (w3 fatty acid) had similar effects
and in addition lowered serum triglyceride concentrations, it
was not given the attention it deserved (2). Vegetable oils rich
in w6 fatty acids displaced other fats in the US diet and eventually
in the diet of Western Europeans, on the evidence of their hy-
pocholesterolemic properties. Omega-3 fatty acids were not
considered as important agents in the control of cardiovascular
disease (CVD) despite experimental and clinical work pointing
to their importance (6-9). With the emphasis on the lipid hy-
pothesis, the lowering ofserum cholesterol became the dominant
factor for the control of coronary heart disease (CHD). As a
result the primary contributions ofinflammation and thrombosis
in the development of CHD were not adequately investigated
until the late l970s and l980s and now.
In the 1970s Bang and Dyerberg ( 10- 12) reported their find-
ings that Eskimos had low rates ofCHD and cancer despite their
high-fat diet. Bang and Dyerberg (1 3, 14) emphasized the im-
portance of eicosapentaenoic acid (EPA) in the prevention of
heart attacks because of its antithrombotic effects, the increase
in bleeding time, and its effect in lowering serum cholesterol
concentrations. Subsequently other epidemiologic studies con-
firmed these findings and showed that fish-eating populations
other than the Eskimos had less CVD than did those who con-
sumed less fish ( 15-20). Even as little as 30-40 g of fish twice a
week made a difference (1 7). In two other studies this effect of
higher fish intake was not seen, most likely because of simul-
taneous high saturated fatty acid intake (2 1, 22).
Additional clinical investigations and experimental studies
confirmed the initial observations: when diets are supplemented
with o3 fatty acids, the latter partially replace the w6 fatty acids
in the membranes of practically all cells (ie, erythrocytes, plate-
lets, endothelial cells, monocytes, lymphocytes, granulocytes,
neuronal cells, fibroblasts, retinal cells, hepatic cells, and neu-
roblastoma cells); o3 fatty acids modulate prostaglandin metab-
olism and decrease triglycerides; and in high doses w3 fatty acids
lower cholesterol and have antithrombotic and anti-inflamma-
tory properties. These studies were extensively reviewed and re-
ported (23-28).
The 1980s were a period ofexpansion in our knowledge about
polyunsaturated fatty acids (PUFAs) in general and w3 fatty
acids in particular. Today we know that w3 fatty acids are es-
sential for normal growth and development and may play an
important role in the prevention and treatment ofcoronary artery
disease, hypertension, arthritis, other inflammatory and au-
toimmune disorders, and cancer. Research has been carried out
in animal models, tissue cultures. and humans. The original
observational studies have given way to controlled clinical trials.
A new arena for w3 fatty acids has emerged as adjuvants to drug
treatment leading to synergism (potentiating the effects of drugs)
or to decreasing their toxicity. This immense expansion in our
knowledge is shown by the increase in the number of publications
from 1 10 in 1984 to 319 in 1989 worldwide (based on January
10, 1990, NIH-MEDLINE search output), amounting to 1541
for the 5-year period (Fig I) (29). In September 1989 the US
National Library ofMedicine published a selective bibliography
on the health benefits offish oils that included publications from
January 1985 to May 1989 but was intentionally limited to hu-
man studies. Cited were 576 articles published in 155 journals
from around the world. The bibliography is indicative of the
explosive and expanding interest in the health benefits of fish
oils and w3 fatty acids.
Although a number of important conferences had been held
before 1985, such as the Reading conference held in 1984, the
expansion ofthis impressive growth in our knowledge can almost
be dated from the 1985 conference Health Effects of Polyun-
saturated Fatty Acids in Seafoods, held June 24-25, 1985, in
Washington, DC (24). The 1985 conference was the first major
international conference to establish the fact that w3 fatty acids
of marine origin, EPA and docosahexaenoic acid (DHA), play
important roles in prostaglandin metabolism, thrombosis and
atherosclerosis, immunology and inflammation, and membrane
function. The 1985 conference participants recommended 1)
the support of research on the role of w3 fatty acids in growth
and development and in health and disease and on the mech-
anisms involved and 2) the establishment of a test-materials
program to specifically define nutritional requirements through-
out the life cycle, and dose and type of w3 fatty acid in inter-
vention studies and in clinical trials.
After the conference the National Institutes of Health (NIH)
published a series of program announcements inviting appli-
cations for research on the role of w3 fatty acids in growth and
development and in health and disease (Table 1)(29). To support
the research, in December 1986 the US Department of Corn-
merce developed a special program, the Biomedical Test Ma-
terials (BTM) program, which provides standardized test ma-
terials of known composition of EPA and DHA nationally and
internationally. These test materials contain 0.2 mg tertiary bu-
tylhydroquinone (TBHQ)/g as an antioxidant and 2 mg to-
copherols/g (30).
The response of the scientific community made it obvious
from the very beginning that research with w3 fatty acids would
develop along two avenues: 1) studies on the essentiality of the
w3 fatty acids that would define their role in growth and devel-
opment throughout the life cycle based on the deficiency model
and improvement in various functions upon supplementation
with w3 fatty acids and 2) studies involving mechanisms in the
FIG I . Publications of marine oil and fish oil and w3 fatty acid studies retrieved from MEDLINE (National Library
of Medicine, National Institutes of Health) from 1984 to 1989. (Data as ofianuary 10, 1990, from MEDLINE. By
June 1989 the total number of publications for 1989 was 386.) Reproduced with permission from reference 29.
understanding of chronic diseases that would use the supple- medical and nutrition journals (27, 28, 35-37). The most recent
mentation approach by increasing the amount offish in the diet, research advances were extensively discussed at the Second In-
substituting fish for meat, or using fish oils. ternational Conference on the Health Effects ofOmega-3 Poly-
Since 1985 many conferences have been held in various parts unsaturated Fatty Acids in Seafoods, held March 20-23, 1990,
of the world to review progress in the field, define gaps in the in Washington, DC (25).
knowledge, and develop a research agenda (24, 25, 3 1-34). In This paper presents the state ofthe art in w3 fatty acid research,
addition, major reviews and commentaries appeared in leading drawn from the published literature, the NIH database Computer
Requests for applications (RFAs) and program announcements (PAs) by NIH and ADAMHA: December 6, 1985, to April 17, l987
Date Title Type Institute
December 6. 1985 Biological Mechanisms ofw3 Fatty Acids in Health and
Disease States
June 1986 Studies ofw3 Polyunsaturated Fatty Acids in
Thrombosis and Cardiovascular Disease
August 22. 1986
The Role ofw3 Polyunsaturated Fatty Acids in Cancer
April 17, 1987
The Role ofw3 Polyunsaturated Fatty Acid in Cancer
Prevention (reissued)
October 22, 1987 Fatty Acid Derived Mediators of Inflammation
S Reproduced with permission from reference 29.
NIH: National Institutes of Health: ADAMHA: Alcohol, Drug Abuse and Mental Health Administration: NCC: Nutrition Coordinating Committee:
NIADDK: National Institute of Diabetes and Digestive and Kidney Diseases: NINCDS: National Institute of Neurological and Communicative
Disorders and Stroke: NIAID: National Institute of Allergy and Infectious Diseases: NICHD: National Institute of Child Health and Human De-
velopment: NIGMS: National Institute of General Medical Sciences: NEI: National Eye Institute: NIEHS: National Institute of Environmental
Health Sciences: NIA: National Institute on Aging: NIAAA: National Institute on Alcohol Abuse and Alcoholism: NIMH: National Institute of
Mental Health: NHLBI: National Heart, Lung, and Blood Institute: NCI: National Cancer Institute: and NIAID: National Institute ofAllergy and
Infectious Diseases.
FIG 3. Essential fatty acid metabolism desaturation and elongation of w6 and w3.
Omega Carbons
a-tlnolenlc H, ““‘‘R.COOH
Unolslc H3
FIG 2. Structural formulas for w3 (a-linoleic), ,6 (linoleic), and w9
(oleic) fatty acids. The first number (before the colon) gives the number
of carbon atoms in the molecule and the second gives the number of
double bonds. w3, w6, and w9 indicate position ofthe first double bond
in a given fatty acid molecule.
Retrieval of Information on Scientific Projects (CRISP), and
presentations at the 1990 conference.
Omega-3 and w6 fatty acids: sources, elongation,
and desaturation
Unsaturated fatty acids consist ofmonounsaturates and poly-
unsaturates. There are two classes of PUFAs, w3 and w6. The
distinction between w3 and w6 fatty acids is based on the location
of the first double bond, counting from the methyl end of the
fatty acid molecule. Monounsaturates are represented by oleic
acid, which can be synthesized by all mammals including hu-
mans. Its double bond is between the 9th and 10th carbon atoms
(Fig 2).
Omega-3 and w6 fatty acids are also known as essential fatty
acids (EFAs) because humans, like all mammals, cannot make
them and must obtain them in their diet. Omega-6 fatty acids
are represented by linoleic acid (LA) and w3 fatty acids by
a-linolenic acid (LNA).
LA is plentiful in nature and is found in the seeds of most
plants except for coconut, cocoa, and palm. LNA on the other
hand is found in the chloroplast ofgreen leafy vegetables. Both
EFAs are metabolized to longer-chain fatty acids of 20 and 22
carbon atoms. LA is metabolized to arachidonic acid (AA) and
LNA, to EPA and DHA, increasing the chain length and degree
of unsaturation by adding extra double bonds to the carboxyl
group (Fig 3).
Humans and animals except for carnivores such as lions and
cats can convert LA to AA and LNA to EPA and DHA (38).
This conversion was shown by using deuterated LNA (39). There
is competition between w3 and w6 fatty acids for the desaturation
enzymes. However, both -4 and -6 desaturases prefer w3 to
w6 fatty acids (38, 40, 4 1). There is some evidence that -6
desaturase decreases with age (38). Premature infants (42), hy-
pertensive individuals (43), and some diabetics (44) are limited
in their ability to make EPA and DHA from LNA. These findings
are important and need to be considered when making dietary
recommendations. EPA and DHA are found in the oils of fish,
particularly fatty fish (Table 2) (24). AA is found predominantly
in the phospholipids of grain-fed animals.
LA, LNA, and their long-chain derivatives are important
components ofanimal and plant cell membranes. In mammals
and birds the w3 fatty acids are distributed selectively among
lipid classes. LNA is found in triglycerides, in cholesteryl esters,
and in very small amounts in phospholipids. EPA is found in
cholesteryl esters, triglycerides, and phospholipids. DHA is found
mostly in phospholipids. In mammals, including humans, the
cerebral cortex (45), retina (46), and testis and sperm (47) are
particularly rich in DHA. DHA is one of the most abundant
components ofthe brain’s structural lipids. DHA, like EPA, can
be derived only from direct ingestion or by synthesis from dietary
Evolutionary aspects: the w3 and w6 fatty acid balance
On the basis ofestimates from studies in paleolithic nutrition
and modern-day hunter-gatherer populations, humans evolved
on a diet that was much lower in saturated fatty acids than is
today’s diet. Furthermore, the diet contained small but roughly
equal amounts ofw6 and w3 PUFAs (Fig 4) (48-50).
Linoleate series Linolenat.e series
C18:2w6 Linoleic acid
6 desaturase
C18:3w6 Gamma-linolenic acid
C20:3w6 Dihomo-gamma-Linolenic Acid
& desaturase
C20:4w6 Arachidonic acid
C22:5w6 Docosapentaenoic acid
C18:3w3 Alpha-Iinolenic acid
6 desaturase
& desatwase
C20:5w3 Eicosapentaenoic acid
C22:5w3 Docosapentaenoic acid
C22:6w3 Docosahexaenoic acid
Content of w3 fatty acids and other fat components in selected fish
Fatty acids
Total Total Total
Fish fat
saturated monounsaturated polyunsaturated 18:3 20:5
22:6 Cholesterol
g/iOOg mg/lOOg
European 4.8 1.3 1.2 1.6 - 0.5 0.9 -
Bass, striped 2.3 0.5 0.7
0.8 Tr 0.2 0.6 80
Bluefish 6.5 1.4 2.9 1.6 - 0.4 0.8 59
Carp 5.6 1.1 2.3 1.4 0.3 0.2 0.1 67
Catfish, brown
bullhead 2.7 0.6
1.0 0.8 0.1 0.2 0.2 75
Catfish, channel 4.3 1.0 1.6 1.0 Tr 0.1 0.2 58
Cod, Atlantic 0.7 0. 1
0. 1 0.3 Tr 0. 1 0.2 43
Croaker, Atlantic 3.2 1.1
1.2 0.5 Tr 0.1 0.1 61
unspecified 1.0 0.2 0.3 0.3
Tr 0.1 0.1 46
Grouper, red 0.8 0.2 0.1
0.2 - Tr 0.2 -
Haddock 0.7 0. 1 0. 1 0.2 Tr 0. 1 0. 1 63
Greenland 13.8 2.4 8.4
1.4 Tr 0.5 0.4
Halibut, Pacific 2.3 0.3 0.8
0.7 0. 1 0. 1
0.3 32
Herring, Pacific 13.9
3.3 6.9 2.4 0.1 1.0 0.7 77
Herring, round 4.4
1.3 0.8 1.5 0.1 0.4 0.8 28
Mackerel,king 13.0 2.5 5.9
3.2 - 1.0 1.2 53
Mullet,striped 3.7 1.2 1.1 1.1 0.1 0.3 0.2 49
Ocean perch 1.6 0.3
0.6 0.5
Tr 0.1 0.1 42
Plaice, European 1.5 0.3 0.5 0.4 Tr 0.1
0.1 70
Pollock 1.0
0. 1 0. 1 0.5 - 0. 1
0.4 71
9.5 3.5 2.6 1.1
- 0.2 0.4 50
Salmon, Chinook 10.4 2.5 4.5
2.1 0.1 0.8 0.6 -
Salmon, pink 3.4 0.6
0.9 1.4 Tr
0.4 0.6 -
Snapper, red
1.2 0.2 0.2 0.4
Tr Tr 0.2 -
Sole, European
1.2 0.3 0.4
0.2 Tr
Tr 0. 1 50
Swordfish 2.1
0.6 0.8 0.2 - 0.1 0.1 39
Trout, rainbow 3.4 0.6
1.0 1.2
0. 1 0. 1 0.4 57
Tuna,albacore 4.9 1.2 1.2
1.8 0.2 0.3 1.0 54
Tuna, unspecified 2.5 0.9 0.6
0.5 - 0.1 0.4 -
S Per 100 g edible portion, raw. Dashes denote lack of reliable data for nutrient known to be present; Tr, trace (< 0.05 g/lOO g food). Adapted
from the United States Department ofAgriculture Provisional Table on the Content ofOmega-3 Fatty Acids and Other Fat Components in Seafoods
as presented by Simopoulos et al (24).
Large-scale production of vegetable oils
The increased consumption of w6 fatty acids in the last 100
y is due to the development of technology at the turn of the
century that marked the beginning of the modern vegetable-oil
industry and to modern agriculture with the emphasis on grain
feeds for domestic livestock (grains are rich in w6 fatty acids)
(5 1). The invention of the continuous screw press, named Ex-
pellet#{174}by VD Anderson, and the steam-vacuum deodorization
process by D Wesson made possible the industrial production
ofcottonseed oil and other vegetable oils for cooking(5 1). Solvent
extraction of oilseeds came into increased use after World War
I and the large-scale production of vegetable oils became more
efficient and more economic. Subsequently, hydrogenation was
applied to oils to solidify them. The partial selective hydrogen-
ation of soybean oil reduced the LNA content of the oil while
leaving a high concentration of LA. LNA content was reduced
because LNA in soybean oil caused many organoleptic problems.
It was recently documented that the hydrogenation process and
particularly the formation oftrans fatty acids has led to increases
in serum cholesterol concentrations whereas LA in its regular
state in oil is associated with a reduced serum cholesterol con-
centration (52, 53).
As stated in the introduction, since the 1950s, research on
the effects of 6 PUFAs in lowering serum cholesterol concen-
trations has dominated the research support on the role of
PUFAs in lipid metabolism. Although a number of investiga-
tom contributed extensively, the paper by Ahrens et al in 1954
(1) and subsequent work by Keys et al (2) firmly established
the a6 fatty acids as the important fatty acids in the field
of CVD. The availability of methods for the production of
vegetable oils and their use in lowering serum cholesterol
concentration led to an increase in both the fat content of the
____._.._._J2t!I±t___._ --
0 L I iT-i
(-4x10 y.wt)
5 Reproduced with permission from reference 58.
Hwiler-Gcf hirer
(10,000 ysort)
1800 1900 2000
TIME (yetars)
FIG 4. Scheme of the relative percentages of different dietary fatty acids (saturated fatty acids and w6 and w3
unsaturated fatty acids) and possible changes subsequent to industrial food processing, involving fattening of animal
husbandry and hydrogenation of fatty acids. Reproduced from reference 48.
diet and the greater increase in vegetable oils rich in w6 fatty
Agribusiness and modern agriculture
Agribusiness contributed further to the decrease in w3 fatty
acids in animal carcasses. Wild animals and birds who feed on
wild plants are very lean, with a carcass fat content ofonly 3.9%
(54), and contain about five times more PUFAs per gram than
is found in domestic livestock (55, 56). Most importantly, 4%
ofthe fat ofwild animals contains EPA. Domestic beef contains
very small or undetectable amounts of LNA because cattle are
fed grains rich in w6 fatty acids and poor in w3 fatty acids (57)
whereas deer that forage on ferns and mosses contain more w3
fatty acids (LNA) in their meat.
Modern agriculture with its emphasis on production has de-
creased the w3 fatty acid content in many foods: green leafy
vegetables, animal meats, eggs, and even fish (58-6 1). Foods
from edible wild plants contain a good balance of w6 and w3
fatty acids (Table 3) (58). Modern aquaculture produces fish
that contain less w3 fatty acid than do fish grown naturally in
the ocean, rivers, and lakes (Table 4) (60). As can be seen from
Table S comparing the fatty acid composition of egg yolk from
free-ranging chickens and the standard US Department of Agri-
culture (USDA) egg, the former has an w6-w3 ratio (w6:w3) of
1.3 whereas the USDA egg has an w6:w3 of 19.4 (59).
Imbalance of w6.w3
Before the l940s cod-liver oil was ingested mainly by children
as a source of vitamins A and D with the usual dose being a
teaspoon. Once these vitamins were synthesized consumption
of cod-liver oil was drastically decreased. Thus an absolute and
relative change of 6:w3 in the food supply of Western societies
has occurred over the last 100 y (Fig 4) (48). A balance existed
between w6 and w3 for millions of years during the long evo-
lutionary history of the genus Homo, and genetic changes oc-
curred partly in response to these dietary influences (49). How-
ever, rapid dietary changes over short periods of time as have
occurred over the past 100- 150 y is a totally new phenomenon
in human evolution.
Homo sapiens made his appearance ‘ 40 000 y ago and the
human genetic constitution has remained relatively unchanged.
Then, 10 000 y ago, agriculture began to bring changes slowly
in food consumption. It is only since the industrial revolution
that changes in food consumption have occurred rapidly. These
changes are reflected in increased consumption of animal fat
and in imbalances in w6:w3. The ratio that was 1 from veg-
etable and animal sources during the evolutionary period for
humans is now estimated by Hunter (6 1) to be 10-1 1: 1 from
vegetable sources. From evidence that the per capita consump-
tion of major foods in 1987 was 61.4 kg red meat, 28.6 kg
chicken, and 6.8 kg fish plus the increases in w6 fatty acids from
vegetable oils, the ratio is closer to 20-25: 1 from vegetable and
Fatty acid content of plants*
acid Purslane Spinach
Red leaf
lettuce Mustard
mg/g wet WI
14:0 0. 16 0.03 0.0 1
0.03 0.02
16:0 0.81 0.16 0.07
0.10 0.13
0.20 0.01 0.02 0.01
18:1w9 0.43 0.04
0.03 0.01 0.01
18:2w6 0.89 0.14
0.10 0.12 0.12
18:3w3 4.05
0.89 0.26 0.31
20:5w3 0.01 0.00
0.00 0.00 0.00
22:6w3 0.00 0.00
0.001 0.002 0.001
Other I .95 0.43 0. 1 1 0. I2
Total 8.50 1.70 0.60
0.702 1.101
Fat content and fatty acid composition of wild and cultured trout, eel, and
Trout (Salmo gairdneri and
Salmo trutta fario)
Eel (Angu i/la anguilla) Salmon (S a/mo salar)
Wild Cultured Wild
(n=2) (n=9) (n=4) (n=4)
(n=2) (n=2)
Fat (g/lOO g) 5 ± 3 6 ± 1 21 ± 6 30 ± 2t 10 ± 0.1 16 ± 0.64
Fatty acids
(g/ 100 g fatty acid)
l8:3w3 3±2 1 ±0.34 2±2 1 ±0.3 1 ±0.1
1 ±0.1
20:5w3 7±0.6 4 ± 14 4±2 3 ±0.6 5 ±0.2 5 ±0.1
22:6w3 15±2 13 ± lt
4±2 6 ±0.4 10 ±2 7 ±
Other w3 5 ± 0.6 2 ± 0.74
3 ± 1 2 ± 0.2t 3 ± 0.5 4 ± 0.1
l8:2w6 4±3 9 ±24 2±2
5 ±0.34 1 ±0.1 3 ±0.1
Other w611
I ± 0.4 0.6 ± 0.14 2 ± 0.3 0.4 ± 0.14 0.2 ± 0.1 0.5 ± 0.1
Total w3 30 ± 0.2 20 ± 34 14 ± 3
12 ± I 20 ± 2 17 ± 0.2
Total w6 5 ± 3 9 ± 2t
3 ± 1 6 ± 0.34 2 ± 0.1 3 ± 0.14
w3:w6 7±5 2 ±0.64
5±2 2 ±0.3t 11 ±2 6 ±
5 Reproduced from reference 60. ± SD; n, number of lots; each lot consisted of about six trout or eel or one or two salmon.
tt Significantly different from wild: tP < 0.05, 4P < 0.01.
§ 18:4w3 + 20:3w3 + 22:5w3.
II 20:4w6 + 22:4w6.
animal sources. From per capita quantities of foods available
for consumption in the US national food supply in 1985, the
amount of EPA is reported to be 50 mg - capita I d and
the amount of DHA is 80 mg - capita - d ‘. The two main
sources are fish and poultry (62).
Biological effects of w3 fatty acids in relation
to CHD and hypertension
Eicosanoid metabolism
AA and EPA are precursors of metabolic products that consist
of 20 carbon atoms and are known collectively as eicosanoids
(prostaglandins, thromboxanes, and leukotrienes) (Fig 5) (63,
64). The discovery of prostaglandins and subsequently the rec-
ognition that AA is the precursor of the 2-series of prostanoids
(prostaglandins and thromboxanes) and of leukotrienes of the
4-series expanded the horizons of research on w6 and w3 fatty
acids because LA, the precursor ofAA, is the predominant PUFA
in the Western diet. EPA and DHA are precursors ofthe pros-
tanoids of the 3-series and leukotrienes of the 5-series. The dis-
covery in 1979, by Needleman et al (65), that prostaglandins
derived from EPA have different biological properties than do
those derived from AA stimulated further research on fish oils
and on the nutritional aspects of prostaglandins.
Competition between the two different classes of PUFAs oc-
curs in prostaglandin formation: EPA competes with AA for
prostaglandin and leukotriene synthesis at the cyclooxygenase
and lipoxygenase level. When humans ingest fish or fish oil, the
EPA and DHA from the diet partially replace the w6 fatty acids,
especially AA, in the membranes of probably all cells but es-
pecially in the membranes of platelets, erythrocytes, neutrophils,
monocytes, and liver cells. As a result, ingestion of EPA and
DHA from fish or fish oil leads to 1) a decreased production of
prostaglandin E2 (POE2) metabolites; 2) a decrease in throm-
boxane A2 , a potent platelet aggregator and vasoconstrictor; 3)
a decrease in leukotriene B4 formation, an inducer of inflam-
mation and a powerful inducer of leukocyte chemotaxis and
adherence; 4 ) an increase in thromboxane A3 , a weak platelet
aggregator and a weak vasoconstrictor; 5) an increase in pros-
tacyclin PG!3 , leading to an overall increase in total prostacyclin
by increasing PGI3 without a decrease in PG!2 . Both P012 and
PG!3 are active vasodilators and inhibitors ofplatelet aggregation;
and 6) an increase in leukotriene B5 , a weak inducer of inflam-
mation and a weak chemotactic agent (63, 64).
Molecular aspects and gene expression:
beyond the eicosanoids
The phospholipid class and fatty acid composition and cho-
lesterol content of biomembranes are critical determinants of
physical properties of membranes and have been shown to in-
fluence a wide variety of membrane-dependent functions, such
as integral enzyme activity, membrane transport, and receptor
function. The ability to alter membrane lipid composition and
function in vivo by diet, even when EFAs are adequately sup-
plied, demonstrates the importance of diet in growth and me-
tabolism (66).
Complex interactions and displacements of the w3 and w6
fatty acids take place in plasma and cellular lipids after dietary
manipulations. Early steps ofcell activation, such as generation
of inositol phosphates, are induced by dietary fatty acids (67).
The effects ofdietary fatty acids on the inositol phosphate path-
way indicate that diet-induced modifications of PUFAs at the
cellular level affect the activity of the enzymes responsible for
the generation of lipid mediators in addition to the formation
of products (eicosanoids) directly derived from their fatty acid
precursors. This shows that dietary fats affect key processes in
cell function.
The role of w3 fatty acids in the control of gene expression is
an area that is expected to expand over the next 5 years as we
begin to understand the role of nutrients in gene expression. It
Supermarket egg
mg/g yolk
I .09
Fatty acid concentrations in chicken egg yolks*
Fatty acid Greek egg ______________
I 7:0
I 8:0
S Reproduced with permission from reference 59. Fatty acid com-
position and lipid content were determined in hard-boiled eggs.
t Ratio of polyunsaturated fatty acids to saturated fatty acids.
:1:Ratio of w6 to w3 fatty acids.
is known that nutrients, like hormones, influence and control
gene expression, and research is now providing more examples
(68). Omega-3 fatty acids in the form of menhaden oil lower
the enzyme fatty acid synthetase in the liver, presumably as a
consequence ofa large decrease in fatty acid synthetase mRNA
concentration (69).
Omega-3 fatty acids have been extensively studied in terms
of their hypolipidemic, antiatheromatous, anti-inflammatory,
antithrombotic, vascular, and other effects described below. In
fact, more is known about the effects ofw3 fatty acids in human
metabolism than any other class of fatty acids.
Hi’polipidemic effects
A review ofthe clinical investigations published up to February
1988 in peer-reviewed English-language journals was earned out
by Hams (37) on the effects offish oils on lipids and lipoproteins
in normal volunteers and in patients with primary hyperlipid-
emia, isolated hypercholesterolemia (type Ila), combined hy-
perlipidemia (type IIb), and isolated hypertriglyceridemia (types
IV and V). Hypercholesterolemia was defined as low-density-
lipoprotein (LDL)-cholesterol concentrations > 4. 14 mmol/L
and hypertriglyceridemia was defined as plasma triglyceride
concentrations > 2.26 mmol/L. There were marked variations
in the design of the studies. The amount of fish oil varied from
a low of 1.6 g/d to > 100 g/d and the w3 fatty acids varied from
0.5 to 25 g/d. The length of intervention varied from 2 wk to
> 2 y. The w3 fatty acid intake was in the form of whole fish,
cod-liver oil, fish-oil concentrate, fatty acid ethyl esters, or pu-
rifled EPA ethyl esters.
Effects on normal subjects. Harris (37) found that w3 fatty
acids did not influence LDL cholesterol concentration, but a
slight rise (‘-3%) occurred in high-density-lipoprotein (HDL)-
cholesterol concentrations and a 25% decrease occurred in tn-
glyceride concentrations. In other well-controlled studies using
relatively lower doses offish oil (< 20 g/d), similar findings were
reported by Sanders and Hochland (70), Zucker et al (7 1), and
Mortensen et al (72). Nagakawa et al (73) used purified EPA
with no other dietary change. They found modest decreases in
total cholesterol and LDL concentrations, no changes in HDL
concentrations, and a marked decrease in triglyceride concen-
Effects on patients. In patients with type ha hyperlipidemia,
dietary 0)3 fatty acids did not change total or LDL cholesterol,
slightly increased HDL, and lowered triglyceride concentrations.
In patients with combined hyperlipidemia, type IIb, total cho-
lesterol concentration did not change. LDL and HDL cholesterol
concentrations rose by 5-7% and triglyceride concentrations de-
creased by 38%. Similar results were reported in other well-con-
trolled, low-dose, crossover trials (7 1, 74-76). In patients with
isolated hypertriglyceridemia, total cholesterol and triglyceride
concentrations decreased by 8% and 52%, respectively, and LDL
and HDL increased by 30% and 10%, respectively. The decrease
in total cholesterol resulted from a fall in very-low-density Ii-
poprotein (VLDL). In patients with type IV hyperlipidemia, LDL
cholesterol concentration increased by 20% (75, 77).
These studies suggest that the type of patient studied deter-
mines the hypolipidemic response to w3 fatty acid supplemen-
tation. In patients with hypertriglyceridemia the fall in total cho-
lesterol is due to the decrease in VLDL. Sanders (78) in an up-
dated review reported similar findings. In addition, he found
that in type III patients fish oil lowers triglycerides and choles-
terol. Sanders also reviewed the studies in which w3 fatty acids
lowered triglycerides in patients with hypertriglyceridemia and
found that these effects of w3 fatty acids are indeed sustained,
contrary to reports by Schectman et al (79). Schectman et al
(79) suggested, on the basis ofa study on a small group of patients,
that triglyceride-lowering effects offish oils cannot be sustained.
This suggestion is not supported by other larger controlled trials.
Miller et al (80), in a randomized controlled trial for 3 mo,
showed that the triglyceride-lowering effect of 10 g MaxEPA#{174}/
d (3.2 g o,3 fatty acids) was sustained. Moreover, Saynor Ct al
(8 1) showed that the effect is sustained for years. The study of
Schectman et al (79) employed an ester concentrate. The failure
to sustain the triglyceride-lowering effect in that study could well
be related to poor patient compliance (78). Sanders found that
as little as 6 g fish oil/d (2 g w3 fatty acids) has a triglyceride-
lowering effect in hypertriglyceridemic patients. The more com-
monly used dose is 3 g/d for EPA and DHA.
In addition to the type of patient, another factor that affects
LDL concentration is whether saturated fatty acids are held
constant or decreased during supplementation. In normal sub-
jects when saturated fatty acids were held constant, LDL tended
Prostaglandtn G2
Leukotnienes 5
Eicosapentaenoic acid
Prostaglandin G3
FIG 5. Origin of w3 and w6 unsaturated fatty acids, biosynthesis of eicosanoids from arachidonic acid (C20:4w6)
and eicosapentaenoic acid (C20:5w3). Reproduced with permission from reference 63.
Oleic acid
1. retIculum
H3CCOOH vegetable Hf * *S## COOH
LinoIeic acid (w -6) foods Arachidonic acid
a-Linoienic acid (w-3)
Marine algae
flt(tOfl \44 O24
H3C,,,,,COoH foods
Eicosapentaenoic acid
l Marine algae
7 Plankton
j Fish
I-IC 7
3 cCOOH
Docosahexaenoic acid
to rise but when saturated fatty acids were reduced, the LDL
tended to decrease. In patients with hyperlipidemia in whom
saturated fatty acids were held constant, LDL increased except
in the study by Phillipson et al (82), who used a very high dose
offish oil. In general, a high dose offish oil (10 g w3 fatty acids!
d) may lower LDL whereas lower doses do not. Whether EPA
or DHA is more effective in lowering LDL is under investigation
with more-purified preparations of the individual fatty acids.
In summary, the effects ofw3 fatty acids on serum cholesterol
concentrations are similar to those of other PUFAs. When w3
fatty acids replace saturated fatty acids in the diet, they lower
serum cholesterol concentrations. Omega-3 fatty acids have the
added benefit of consistently lowering serum triglyceride con-
centrations whereas the w6 fatty acids do not and may even
increase them (82).
In considering these aspects ofw3 fatty acids on LDL-choles-
terol concentrations, the issue is whether this increase in LDL
is indeed significant in increasing the risk for atherosclerosis in
patients with type II and IV hyperlipidemia in view of the an-
tithrombotic, anti-inflammatory, and antivasorestrictive aspects
ofw3 fatty acids. Furthermore, the possibility that this new LDL
may not be atherogenic, or as atherogenic, needs to be considered
because fish-oil diets have produced changes in lipoprotein
composition in animal studies (83-85). Theoretically, EPA and
DHA may alter the rate or form of LDL oxidation in vivo and
thereby cause a reduced atherogenic potential not reflected in
an actual lowering of, or even despite an increase in, LDL con-
Another important consideration is the finding that during
chronic fish-oil feeding there is a decrease in postprandial tri-
glyceride concentrations. Furthermore, Nestel (86) reported that
fish-oil feeding blunted the expected rise in plasma cholesterol
concentrations when large amounts of cholesterol were fed to
humans. These findings are consistent with a reduced rate of
coronary artery disease in fish-eating populations. Studies in hu-
mans have shown that fish oils reduce the rate ofhepatic secretion
of VLDL triglyceride (77, 87-89). In normolipidemic subjects
w3 fatty acids prevent and reverse rapidly the carbohydrate-in-
duced hypertriglycenidemia (87). There is also evidence from
kinetic studies that fish oils increase the fractional catabolic rate
(FCR) ofVLDL (77, 88, 89).
.4 ntiat/n’ronatoii.s actions
The antiatheromatous actions ofw3 fatty acids are supported
by a number ofanimal studies. In dogs fed a diet high in saturated
fatty acids and cholesterol, supplementation with fish oils pre-
vented intimal hyperplasia that is induced on venous allografts
inserted into their arteries (90). In hyperlipidemic swine model,
dietary supplementation with cod-liver oil reduced the devel-
opment of coronary atherosclerosis without any significant
changes in plasma lipid concentrations between the supple-
mented animals and the controls (9 1). In the primate model,
dietary fat substitution with w3 fatty acids inhibited atherogenesis
in the aorta, carotid, and femoral arteries (92). Hollander et al
(93) confirmed Davis et al’s (92) findings in another primate
species without significant differences in serum lipid levels. Using
the rabbit atherogenesis model, Thiery and Seidel (94) found
that fish-oil feeding resulted in an enhancement of cholesterol-
induced atherogenesis whereas Zhu et al (95) found that ath-
erosclerosis was inhibited by fish oils in cholesterol-fed rabbits.
These conflicting results were observed in the rabbit atherogenesis
model whereas in all other models (dog, swine, and two primate
species) fish oils were found to have antiatherogenic effects even
if they did not lower serum lipids.
Antithromhotic effects
In addition to a prolongation of bleeding time, there is sub-
stantial agreement that platelet aggregation to epinephrine and
collagen is inhibited, thromboxane A2 production is decreased,
whole-blood viscosity is reduced, and erythrocyte membrane
fluidity is increased (24, 33, 34, 64, 96). Increased concentrations
of plasminogen activator and decreased concentrations of a
plasminogen inhibitor after fish-oil ingestion were reported (97).
Fibrinogen also decreases after w3 fatty acid ingestion. Although
some studies failed to show a decrease in fibrinogen concentra-
tions, a randomized, double-blind clinical trial did show a de-
crease after ingestion ofw3 fatty acids in adults with type lIb or
IV hyperlipoproteinemia (98). A decrease in fibrinogen was also
found in another double-blind trial with 64 men aged 35-40 y
randomly assigned to two groups (99). In the studies that failed
to show an effect, the study by Sanders et al (100) used a small
dose ofcod-liver oil and the study by Rogers et al (101) included
healthy volunteers and was of short duration.
Although a decrease in platelet count occurs with an increase
in platelet size after ingestion ofw3 fatty acids (especially in very
large quantities), clinical evidence of bleeding has yet to be re-
ported. Platelet survival has been found to be normal. Because
there is an increase in platelet size with a decrease in platelet
count, there is no overall decrease in platelet mass (102). When
fish oils are discontinued, the platelet count returns to normal.
In some cases platelet count rebounds to supernormal before
returning to normal. The mechanisms ofthese effects offish oil
on platelets and on megakaryocytes are unknown. These effects
of fish oils leading to an increase in bleeding time has raised
questions: What is the clinical significance of the prolonged
bleeding time? Are there any adverse effects?
The effects of different doses of fish oils on the prolongation
of bleeding time were investigated by Saynor et al (8 1). With
1.8 g EPA there was not any prolongation in bleeding time. At
4 g the bleeding time increased and the platelet count decreased
without any adverse effects. In studies in humans there has never
been a case of clinical bleeding, even in patients undergoing
angioplasty while they were on fish-oil supplements (103).
DeCaterina et al (104) recently reported on the preoperative
use offish oils in I 3 men and 2 women who underwent coronary-
artery-bypass graft surgery. The daily dose was 3 g EPA and 1.3
g DHA in purified fish oil that was taken for 28 d before surgery.
The control subjects were 14 men and 1 woman perfectly
matched for age and severity ofdisease who were scheduled for
surgery by the same surgeon. The control subjects did not receive
any fish oils. Despite changes in platelet function, increases in
bleeding time, and increases in vascular PG!2. the perioperative
blood loss was not increased in subjects receiving fish-oil sup-
plements. There is no evidence that the increase in bleeding
time is clinically significant or has any adverse effects.
Vascular effects
It was recently shown that w3 fatty acids inhibit the production
ofplatelet-denived growth factor (PDGF) ( 105) and increase en-
dothelium-derived relaxing factor (EDRF)(l06). Omega-3 fatty
acids reduce production of a PDGF-like protein in bovine en-
dothelial cells, which leads to inhibition in the migration and
proliferation of smooth muscle cells, fibroblasts, and macro-
phages in the arterial wall (105).
The endothelium releases an EDRF, presumably nitric oxide.
When animals are fed cod-liver oil or fish oils (EPA plus DHA),
they increase the release of relaxing factors, which facilitates
relaxation in large arteries and in resistance vessels (106). Also
in the presence of EPA, endothelial cells in culture increase the
release of relaxing factors indicating a direct effect of the fatty
acid on the cells. EDRF presumably contributes to antithrom-
botic and antiatherosclerotic effects ofw3 fatty acids by relaxing
vascular smooth muscle and inhibiting platelet aggregation.
Increases in PGI2 were shown in tissue fragments from the
atrium, aorta, and saphenous vein obtained at surgery in patients
treated with w3 fatty acids (104). This finding is very important
because it enhances our understanding ofthe effects ofw3 fatty
acids on vessel walls in humans and differs from the results of
some animal studies (107, 108). Rats do not form PGI3 after
dietary EPA (107, 108) whereas humans do (109). Therefore,
the importance of human studies is obvious.
Antiarrhvthmic effects
Sudden cardiac death is frequently a consequence of severe
ventricular fibrillation or terminal cardiac arrhythmia. Experi-
mental studies with isolated papillary muscles from either rats
or marmoset monkeys indicate much less susceptibility to cat-
echolamine-induced arrhythmia in the muscles from animals
fed fish-oil supplements than from those on w6 or low-fat diets
(1 10). Indomethacin abolishes these effects in vitro, suggesting
a mechanism operating via the eicosanoids (prostaglandins). In
studies with adult marmoset monkeys fed dietary fish oil for
several months, cardiac function improved, and the vulnerability
of the heart to develop cardiac arrhythmia was reduced when
subjected to ischemic stress. Burr et al (23) studied the effects
ofdietary intervention in the secondary prevention of myocardial
infarction. A modest intake of fatty fish two-to-three times per
week (or 3 g fish oils/d) reduced all-cause mortality by 29% over
a 2-y period, possibly by preventing sudden death from an-
Effects on restenosis
The antiatheromatous aspects of w3 fatty acids shown in an-
imal experiments suggested the use of w3 fatty acids to prevent
restenosis in patients undergoing angioplasty. The cause of re-
stenosis is unknown. However platelet aggregation, proliferation
ofsmooth muscle cells, and coronary vasospasm are considered
to be important contributors to restenosis. Although the success
rate of angioplasty is high, restenosis occurs in 25-40% in the
dilated lesions - 6 mo after the procedure. Most studies showed
a benefit when co3 fatty acids supplemented the standard regimen
before and after surgery (103, 1 1 1, 1 12). Dehmer et al (103)
provided evidence that when w3 fatty acids were given to the
patients along with aspirin and dipyridamole 7 d before angio-
plasty and continued for 6 mo afterward, there was a reduction
in the rate of restenosis on catheterization 3-4 mo after angio-
plasty. Others report no benefit ( 1 13, 1 14). There was no clinical
evidence of bleeding complications in any treated patient re-
ported in these studies. The role ofw3 fatty acids in the preven-
tion ofearly restenosis after coronary angioplasty is a major area
of research because percutaneous transluminal coronary angio-
plasty is an important treatment for selected patients with CHD.
Effects ofdietary w3 fatty acids on factors and mechanisms involved
in the development of inflammation, atherosclerosis, and immune
Reduce or inhibit risk and/or precipitating factors
Arachidonic acid
Platelet aggregation
Thromboxane A2 formation
Monocyte and/or macrophage function
Leukotriene formation (LTB4)
Formation of platelet activating factor (PAF)
Toxic oxygen metabolites
Interleukin I formation (IL- 1)
Formation of tumor necrosis factor (TNF)
Platelet-derived growth factor-like protein (PDGF)
Intimal hyperplasia
Blood pressure and/or blood pressure response
Very-low-density and low-density lipoproteins (VLDL, LDL)
Lipoprotein (a) [Lp(a)]
Blood viscosity
Increase beneficial and/or protective factors
Prostacyclin formation (PGI2 + PGI3)
Leukotriene B5 (LTB5)
Interleukin 2 (IL-2)
Endothelial-derived relaxing factor (EDRF)
Fibrinolytic activity
Red-cell deformability
High-density lipoprotein (HDL)
* Reproduced with permission from reference 29.
Effects on lipoprotein (a)
Lipoprotein (a) [Lp(a)] is a genetically determined protein
that has atherogenic and thrombogenic properties. The molecular
structure ofLp(a) apoprotein is strikingly similar to that of plas-
minogen. Omega-3 fatty acids were reported to inhibit the in-
hibitor of plasminogen activator and thus contribute to fibri-
nolysis (97). Thus it was only natural to test the effects of w3
fatty acids on Lp(a) concentrations ( 1 15). Herrmann et al (1 15)
reported on such a study at the poster session of the NATO
Advanced Research Workshop on Dietary w3 and w6 Fatty Ac-
ids: Biological Effects and Nutritional Essentiality. These inves-
tigatons studied 62 male patients who had myocardial infarction
6 mo before the study. Ingestion of fish oil reduced the concen-
tration of triglycerides, reduced blood pressure, and led to a
significant reduction in Lp(a). This study provided the first ev-
idence that w3 fatty acids lowered Lp(a). Recently, Schmidt et
al (1 16) showed that w3 fatty acids lowered serum Lp(a) con-
centrations when Lp(a) concentrations were > 200 mg/L but
had no effect < 200 mg/L.
More recently, Kostner and Herrmann (1 17) compared the
effects offish-oil concentrate (12 g FENICO#{174}/d, containing 70%
w3 PUFAs) in 35 patients with coronary disease and a control
group receiving an equivalent amount ofrapeseed oil. In addition
to measuring Lp(a), these investigators carried out standard
plasma lipid and lipoprotein determinations and hemostatic in-
dices. Plasma Lp(a) concentrations were reduced in the fish-oil
group but were unaffected in the rapeseed-oil group. The total
cholesterol, LDL cholesterol, and apolipoprotein B (apo B) con-
centrations fell significantly in both groups. HDL cholesterol
increased and triglycerides decreased significantly only in the
fish-oil group. Not everybody in the fish-oil group showed a
decrease in plasma Lp(a) concentrations. The investigators
therefore subdivided the participants in the study into two groups,
responders and nonresponders. Two-thirds of the people studied
were responders and they showed an average Lp(a) decrease of
24%. In this study, tissue plasminogen activator concentrations
were reduced significantly in both groups by 16%. There was
a concomitant but not significant increase of plasma activator
inhibitor, PAl5.
In a recent study by Seed et al (1 18) on the relation of serum
Lp(a) concentration and apolipoprotein A (apo A) phenotype
to CHD in patients with familial hypercholesterolemia, it was
shown that “the median lipoprotein(a) level in the 54 patients
with CHD was 57 mg/dl, which is significantly higher than the
corresponding value of 18 mg/dl in the 61 patients without CHD.
According to discriminant-function analysis, the lipopnotein(a)
level was the best discriminator between the two groups (as
compared with all other lipid and lipoprotein levels, age, sex,
and smoking status).” The authors conclude that “an elevated
level oflipoprotein(a) is a strong risk factor for CHD in patients
with familial hypercholesterolemia, and the increase in risk is
independent of age, sex, smoking status, and serum levels of
total cholesterol.” In another study on apo A and ischemic heart
disease in familial hypercholesterolemia, Wikiund et al (119)
also concluded that Lp(a) is a genetic trait that may be useful
in identifying patients with familial hypercholesterolemia at high
risk for CHD (1 19). Clinical investigations are urgently needed
to determine if lowering Lp(a) by w3 fatty acids lowers the risk
for CHD in these patients.
Additional effects
Omega-3 fatty acids have been shown in human monocytes
to inhibit the production of platelet activating factor (PAF). One
of the adverse effects of PAF is the activation of platelets, thus
contributing to atherogenesis (120). lnterleukin and tumor ne-
crosis factor (TNF) are reduced by feeding fish-oil supplements
to humans (121). Both interleukin 1 (IL-I) and TNF are con-
sidered atherogenic because they stimulate the synthesis of
adhesion molecules, thus causing monocytes to adhere to en-
dothelial cells. They also activate platelets, neutrophils, and
monocytes (121).
In conclusion, many studies indicate that w3 fatty acids appear
to decrease or inhibit risk and precipitating factors in the de-
velopment of CVDS. These factors are summarized in Table
The new findings in relation to interleukin metabolism and
gene expression indicate that, in addition to their major effects
on prostaglandin metabolism, w3 fatty acids have other far-
reaching effects on intracellular cell communication. These
findings indicate that it is very important to know and eventually
understand the numerous inter- and intracellular factors that
are influenced by w3 fatty acids as well as the specific mechanisms
involved. It is this type of information that will enable us to
design appropriate clinical trials to precisely define the dose of
(&,3 fatty acids to be utilized and the type of fatty acid and length
ofintervention required for effective therapy while avoiding any
possible adverse reactions.
Genetic determinants and environmental risk factors for CHD*
Genetic determinants
Family history ofCHD at an early age
Total serum cholesterol, LDL, and apo B concentrations
HDL cholesterol, apo A-I, and apo A-Il concentrations
LDL receptor activity
Thrombosis and coagulation variables
Triglycerides and VLDL concentrations
RFLPs in DNA at the apo A-I/apo C-Ill and apo B loci
Other DNA markers
Blood pressure
Insulin concentration and insulin response
Heterozygosity for homocystinuria
Environmental risk factors
Sedentary lifestyle (lack of aerobic exercise)
Diet (excess energy intake)
High saturated fatty acid intake
Low w3 fatty acid intake
Psychosocial factors
Type A personality
Social class
most-important paper on atherosclerosis entitled “Atheroscle-
rosis: A problem ofthe biology ofthe arterial wall cells and their
interaction with blood components,” in which the modern con-
cepts of atherosclerosis were presented. Ross in 1986 (126) and
Steinberg et al in 1989 (127) updated the concepts ofthe response
to injury hypothesis in the pathogenesis ofatherosclerosis, which
can be summarized as follows. The first step in the formation
of atherosclerosis is a nonspecific (functional) injury to endo-
thelium followed by an accumulation of monocytes and mac-
rophages, foam cell formation, and platelet aggregation. The
platelets release growth factor, which leads to smooth muscle
migration and proliferation. At this point cholesterol is deposited
in the smooth muscle cells and monocyte macrophages in the
vessel wall. These events further lead to the formation of ground
substance and eventually to plaque formation. As seen in Table
6, w3 fatty acid ingestion may be able to prevent the increase in
cellular components generated by these cells and interfere at
many steps in the development of the atherogenic process (Fig
6) (27, 48).
Vegetable oils rich in the w6 LA have been promoted as low-
ering blood cholesterol concentrations of people in the United
States. So much emphasis has been put on the lipid hypothesis
and on lowering serum cholesterol concentrations through diet
and drugs that the contributions ofinflammation and thrombosis
in the development of CHD have not been fully appreciated.
An increase in our understanding of the pathophysiology of
coronary artery thrombosis has led to the hypothesis that pre-
venting platelet activation and aggregation are essential steps in
the prevention of coronary thrombotic complications. Current
evidence suggests that the pathophysiology of unstable angina
involves platelet recruitment and thrombosis. Burr’s study re-
ferred to earlier (23) is the first prospective dietary-intervention
trial for secondary prevention ofCHD that demonstrates clinical
benefit in those given advice to eat fatty fish or fish-oil capsules.
Omega-3 fatty acids alone clearly will not lead to the universal
eradication of atherosclerosis. However, it is increasingly evident
that dietary fish-oil supplementation may help in the prevention
of atherosclerosis or its thrombotic complications. The favorable
effects shown by DeCaterina et al (104) provide further support
I Monocytes Thrombosis - - -* Fibrinol,sis
I/I (circulating) /X// _________1
I M . Smooth musde
- f
‘\ . L. . I
\ Earliest Atherosclerotic Lesson + LDL Iholesterol
\ \ \ Foam Cells + Connective Tissue + Thrombosis
------ Atheroma
FIG 6. Sites of potential interventions for preventing development of
atherosclerosis. Note that several of these possibilities would abort the
disease before the concentration of plasma LDL cholesterol could con-
tribute to the atherosclerotic process. Reproduced from reference 48.
C Reproduced with permission from reference I23.
Cornary heart disease
Much more is known about the effects of w3 fatty acids on
CVD than on any other disease entity. CHD is a multifactorial
disease with genetic determinants that interact with many en-
vironmental factors, including diet and other lifestyle changes
that contribute to its development. All the hyperlipoproteinemias
described so far have a significant genetic component. It has
been estimated that 50% of the variance in serum cholesterol
concentration and 15% of the variance in the fibrinogen con-
centration are due to genetic factors.
An extensive array of genes involved in normal regulation
and function of the cardiovascular system have been identified
with modern genetic techniques [DNA markers, restriction-
fragment-length polymorphisms (RFLPs) and in situ hybridiza-
tion studies]. These new genetic markers are being used to predict
risk for CHD, a most important aspect for developing strategies
for the prevention ofCHD. The early identification of individuals
in childhood and young adult life who are at high genetic risk
constitutes a very powerful health care strategy for the prevention
ofCHD. Common genetic lipoprotein disorders associated with
premature CHD include familial combined hyperlipidemia
(15%), familial hypentriglyceridemia (5%), and familial hyper-
cholesterolemia (5%) ( I 22) (Table 7) ( 123). In the United King-
dom, 5 g MaxEPA#{174} (1 8% EPA, I 2% DHA) twice a day has been
approved for the treatment of severe hypertriglyceridemia in
patients at risk of ischemic heart disease or pancreatitis.
Atherosclerosis is a complex disease of the arteries and the
arterial wall. Many cellular biochemical and physical compo-
nents interact at and within the arterial wall. The cellular dy-
namics in atherosclerosis were reviewed by Faggiotto ( 124) who
states that “atherosclerosis encompasses a number of pathological
processes and has many ofthe features ofdegenerative disorders,
inflammation and neoplasia.” In 198 1 Ross (125) published a
- - -*“Endothelial Injury”
Platelet Aggregation
PDG Factor
that w3 fatty acids could modify the occurrence and the course
of atherosclerosis.
Recently, Dolecek and Grandits (20) investigated the 24-h
dietary-recall data in the usual-care group of the Multiple Risk
Factor Intervention Trial and distinguished between w3 and t6
fatty acid intake and their relationship to four mortality cate-
gories: CHD, total CVD, all-cause mortality, and cancer. Analysis
of the combined fatty acids predominantly found in fish (EPA
and DHA) demonstrated significant inverse associations with
CHD, CVD, and all-cause mortality groups. The benefit ap-
peared to be in the highest intake quintile with a mean ingestion
of -664 mg/d of EPA and DHA. When compared with zero
intake, mortality from CHD, CVD, and-all cause mortality was
40%, 41%, and 24% lower, respectively. An inverse association
was noted between the ratio of 18:3w3 to 18:2w6 and cancer
mortality. Thirty-three percent fewer cancer deaths occurred in
the highest intake quintile when compared with the lowest.
Populations with high consumption of fish, such as the Es-
kimos and Japanese, have lower rates of myocardial infarction
(Table 8) (128). The epidemiologic studies of Kromhout et al
(1 7) and the intervention studies of Burr et al (23) showed a
decrease in CHD mortality in people consuming relatively small
amounts of fish (0.5 g w3 fatty acids/d or 1 .5 g fish oil/d) over
a long period of time (19 y and 2 y, respectively). This suggests
that small doses over long periods of time may have beneficial
effects, possibly by reducing blood pressure and other risk factors.
Certainly more studies are needed to define the precise dose of
w3 fatty acids based on the total long-term diet of the subjects
in terms of w3, w6, w9, and saturated fatty acids. The unique
pharmacokinetics of w3 fatty acids in terms of time and dose-
dependent accumulation in cell membranes should help define
the optimum amount ofw3 fatty acids and length oftime of the
The vast majority of survivors of myocardial infarction have
one or more of four lipoprotein abnormalities. These include
increased LDL-cholesterol concentrations, decreased HDL-
cholesterol concentrations usually accompanied by increased
triglyceride or VLDL concentrations, increased concentration
of chylomicron remnants and intermediate-density lipoprotein
(IDL), and the presence in plasma of increased concentrations
of Lp(a). The exact mechanism whereby each of these abnor-
malities causes CHD is an area of active investigation and the
genetic contribution to each ofthese abnormal lipoprotein phe-
notypes is coming into focus (129). As we begin to unravel the
genetics of atherosclerosis and identify individuals with genetic
susceptibility to CHD, modification ofdiet early in life and the
provision of increased amounts of w3 fatty acids should be ben-
eficial in the prevention of CHD.
Hypertension is also a multifactorial disorder involving gene-
nutrient interactions and other factors (1 30). Different mecha-
nisms appear to be involved, operating at variable proportions
based on the organ involved or the cause of hypertension.
Changes were reported in eicosanoid metabolism, renin con-
centrations, vascular reactivity, blood viscosity, loss of sodium,
increase in potassium in cells, and a decrease in intracellular
calcium, among others (130).
In 1983 two groups of investigators, Singer et al (1 3 1) and
Lorenz et al (1 32), were the first to show that adding mackerel
to the diet ofpatients with mild hypertension lowered the blood
Ethnic differences in fatty acid concentrations in thrombocyte
phospholipids and frequency of cardiovascular disorders
United States Japan
Arachidonic acid,
C20:4w6 (%) 26 21 8.3
Eicosapentaenoic acid,
C20:5w3 (%) 0.5 1.6 8.0
w6:w3 50 12 1
Cardiovascular mortality
45 12 7
a Adapted from reference I 28.
t Percent ofall deaths.
pressure. Many studies since then have used w3 fatty acids in
the form of fish oils with similar results in normal and hyper-
tensive subjects (72, 133-137) but not in all intervention trials
(138, 139).
More recently Knapp and FitzGerald (1 35) reported on a
controlled study of PUFA supplements in essential hypertension.
These investigators studied blood pressure and eicosanoid pro-
duction during supplementation of dietary fat for 4 wk in 32
men with mild essential hypertension. Groups ofeight subjects
received either 3 or 15 g w3 fatty acids/d in the form of 10 or
50 mL MaxEPA#{174}, 39 g w6 fatty acids/d in the form of 50 mL
safflower oil, or 50 mL/d of an oil mixture that approximated
the types offat present in the American diet. Urinary metabolites
were measured to assess biosynthesis of eicosanoids. The men
who received the high dose of fish oil had a mean decrease in
systolic blood pressure of6.5 mm Hg and a decrease of4.4 mm
Hg in diastolic blood pressure. The group receiving the low dose
of fish oil (10 mL/d) did not have any significant change in
blood pressure from baseline during the supplementation period
nor did the u,6-supplemented on the control groups. In the group
receiving the high dose offish oil, the formation of vasodilatory
prostacyclins (PG!2 and PGI3) increased initially but this increase
was not maintained as blood pressure fell. The concentration of
thromboxane A2 metabolites fell and metabolites of thrombox-
ane A3 were detected in the groups receiving fish oil. The for-
mation ofPGE2 increased during supplementation with safflower
oil and tended to decrease with fish oil but no POE3 metabolite
was detected. These data indicate that high doses of fish oil can
reduce blood pressure in men with essential hypertension.
Bonaa et al ( 140), in a population-based intervention trial
from Troms#{248},recently reported decreases of6 mm Hg in systolic
blood pressure and 3 mm Hg in diastolic blood pressure with
fish-oil supplementation. These investigators monitored diet and
assessed concentrations of plasma phospholipid fatty acids to
determine the relation among diet, fatty acids, and blood pres-
sure. Dietary supplementation with fish oil did not change mean
blood pressure in the subjects who ate fish three or more times
per week as part oftheir usual diet or in those who had a baseline
concentration of plasma phospholipid o3 fatty acids > 175.1
mg/L, suggesting that a relationship may exist between plasma
phospholipid w3 fatty acid concentration and blood pressure.
There was a lower blood pressure at baseline in subjects who
habitually consumed larger quantities of fish, suggesting that
supplementation with fish oils would be important from the
primary prevention standpoint. In another study, three cans of
mackerel per week (equivalent to 1.2 g w3 fatty acids/d or 1.2
x 3 = 3.6 g of fish oil/d) for 8 mo led to lowering of blood
pressure (1 33). This amount of fish oil could be considered ac-
ceptable for a daily intake by the general population.
It was suggested that these effects on blood pressure during
dietary supplementation with w3 fatty acids are due to changes
in the endogenous synthesis of vasoactive eicosanoids. Two re-
search groups showed that dietary EPA is converted to PGI3 in
man and does not suppress formation of PGI2 from AA (141,
142). Other possible mechanisms under consideration include
effects of w3 fatty acids on renal function, a lowering of blood
viscosity, and a reduction in vascular responsiveness to systemic
vasoconstrictors ( 143). Lorenz et al ( 132) observed an increase
in urinary sodium and a decrease in plasma renin activity at the
end ofthe fish-oil period in a group of men whose Western diet
was supplemented with cod-liver oil. Fish-oil supplements were
reported to have beneficial effects on the blood pressure of pa-
tients on hemodialysis with little residual function (144). Normal
subjects do not show any change in renal function even when
given pharmacologic doses offish oil, which is encouraging from
the safety standpoint (145).
Rats with reduced renal function (subjected to subtotal ne-
phrectomy) given dietary fish oil had reduced urinary PGE and
renal function along with increased proteinuria and mortality
(146). As indicated earlier, studies from rats cannot be extrap-
olated to humans because rats do not make PGI3 (107-109).
The effects of c,3 fatty acids on immune-mediated renal dys-
function are complex. Kelley et al (147) showed that fish oils
prolong the survival in mice that develop lupus nephritis whereas
Westberg et al (148) found less benefit. However, Kelley used a
larger dose of fish oil. In humans the clinical course of lupus
nephritis did not improve with fish-oil supplementation (149,
150). More clinical investigations are needed before any con-
clusions can be drawn.
Recent data on the additive effects offish-oil supplements and
propranolol were presented at the Second International Con-
ference on the Health Effects ofOmega-3 Polyunsaturated Fatty
Acids in Seafoods (1 5 1). The combination ofw3 fatty acids and
propranolol potentiated their blood-pressure-lowering effects.
In addition, the increase in plasma triglycerides often seen during
antihypertensive therapy did not occur. Therefore, c,3 fatty acid
supplementation has the potential for a beneficial modification
of several cardiovascular risk factors as adjuvants to the anti-
hypertensive regimen.
Inflammatory and autoimmune disorders
The effects of supplementing the diet with w3 fatty acids in
the form of fish oils on the function of the 5-lipoxygenase path-
ways of peripheral blood polymorphonuclear leukocytes and
monocytes has been investigated in normal subjects and in pa-
tients with diseases such as arthritis, psoriasis, ulcerative colitis,
lupus erythematosus, and asthma. These studies were stimulated
by the demonstration by Prickett et al (1 52) that the fatal spon-
taneous autoimmune renal disease in a genetic strain of NZB
mice can be largely prevented by changing the fat in the diet
from beeftallow to fish oil. In fact, the protective effect of marine
lipids on autoimmune renal disease is one ofthe most dramatic
effects of w3 fatty acids on any pathology (1 52). Supplementing
the diet with w3 fatty acids (3.2 g EPA and 2.2 g DHA) in normal
subjects increased the EPA content in neutrophils and monocytes
more than sevenfold without changing the quantities ofAA and
DHA. The anti-inflammatory effects of fish oils are partly me-
diated by inhibiting the 5-lipoxygenase pathway in neutrophils
and monocytes and inhibiting the leukotriene B4 (LTB4)-me-
diated function of neutrophils while increasing the production
of LTB5 (Fig 7) (153, 154). Studies since 1985 show that w3
fatty acids influence interleukin metabolism by decreasing IL-l
(121, 155, 156).
Many experimental studies have provided evidence that in-
corporation of alternative fatty acids into tissues may modify
inflammatory and immune reactions and that w3 fatty acids in
particular are potential therapeutic agents for inflammatory dis-
Advances in the understanding ofleukotriene metabolism and
its role in inflammation and autoimmune disorders began to
attract investigators who used fish oils in patients with arthritis
with promising results (1 54). In normal volunteers, marine lipids
suppress 5-lipoxygenase pathway products from both neutrophils
and monocytes and they also suppress production of PAF from
monocytes ( 1 20). In addition, the chemotactic response of neu-
trophils to transmembrane agonists is suppressed by dietary fish
oils. In patients with rheumatoid arthritis, inhibition of 5-lipox-
ygenase products are limited to LTB4 , suggesting inhibition of
the epoxide hydrolase step, and neutrophil chemotaxis is in-
creased by the fish-oil dose rather than suppressed as in normal
subjects. These differences between normal subjects and patients
with arthritis could be related to specific alterations ofthe disease
or to medications taken by patients (156). Omega-3 fatty acids
were shown to decrease IL- 1 in animals and in patients with
rheumatoid arthritis and normal volunteers (121, 155, 156).
Kremer et al ( 155) carried out a prospective randomized dou-
ble-blind, placebo-controlled parallel study. Three groups were
studied for 24 wk with two different doses of fish oil and one
dose of olive oil, and clinical and immunological indices were
measured. As in previous studies there was clinical improvement
and, in addition to the decrease in LTB4 and increase in LTB5,
there was a significant decrease in IL-l production and a non-
significant increase in IL-2. LTB4 exerts a positive modulating
effect on the genetic control ofIL-1 probably at the translational
level within the cytoplasm. These findings were confirmed by a
number of other investigators. Omega-3 fatty acids (fish oil), as
a dietary supplement, along with nonsteroid antirheumatic drugs
were shown to provide subjective relief to patients with rheu-
matoid arthritis.
The recognition that AA metabolism is altered in psoriasis
prompted attempts to inhibit the generation of proinflammatory
lipoxygenase products [LTB4 and l2-hydroxyeicosatetraenoic
acid (12-HETE)J, which are markedly elevated in the psoriatic
lesions (1 57). The addition of MaxEPA#{174} to the standard treat-
ment produced further improvement and a decrease in LTB4
and 12-HETE (Fig 7) (64). In other studies fish oil was success-
fully used in combination with etretinate to reduce the hyper-
lipidemia caused by that drug and with ultraviolet B (UVB),
where it prolongs the beneficial effects ofa course of phototherapy
(1 58). Studies of fish oil in combination with cyclosporin are in
progress in an attempt to reduce nephrotoxicity, which is the
major side effect of that drug (158).
[ UPOX[
__ __ _
6-tmns-LTB4 4 LTA4 LTA5- 6 tmns-LTB5
diastereoisomers ?:,(, diasteresorns
LTB Lit4 -‘LTD -LTE
L1C --#{248}’LTD5 -.‘tTE
FIG 7. Oxidative metabolism of arachidonic acid (AA), eicosapentaenoic acid (EPA), and docosahexaenoic acid
(DCHA) by the 5-lipoxygenase pathway. 5HPETE denotes 5-hydroperoxyeicosatetraenoic acid; 5HETE, 5-hydrox-
yeicosatetraenoic acid: 5HPEPE, 5-hydroperoxyeicosapentaenoic acid: 5HEPE, 5-hydroxyeicosapentaenoic acid:
7HPDCHA, 7-hydroperoxydocosahexaenoic acid: 4HPDCHA, 4-hydroperoxydocosahexaenoic acid; 7HDCHA,
7-hydroxydocosahexaenoic acid: 4HDCHA, 4-hydroxydocosahexanoic acid: and LT, leukotriene. Reproduced with
permission from reference 153.
LIcerativc’ cOlitiS
As indicated earlier, LTB4, a metabolite of AA, is produced
by activated neutrophils, and w3 fatty acids decrease its pro-
duction. LTB4 is an important mediator of inflammation in
ulcerative colitis and it is believed to recruit additional neutro-
phils from the bloodstream into the mucosa. Stenson et al (159)
conducted a study of the effects of fish-oil supplementation on
ulcerative colitis. Preliminary analysis showed statistically sig-
nificant improvement in sigmoidoscopy score and global clinical
assessment after 4 mo of fish-oil-supplemented diet compared
with placebo diet in active ulcerative colitis. This is the first
double-blind crossover trial of fish-oil supplementation in ul-
cerative colitis.
The number ofpublications from the use ofw3 fatty acids in
cancer studies in animals has increased exponentially over the
past 5 y (29). Animal tumor models in which the tumor was
induced by carcinogens and animal models with transplantable
tumors (breast, colon, pancreas, and prostate) have been inves-
tigated. The results have consistently shown that w3 fatty acids
delayed tumor appearance and decreased both the rate of growth
and the size and number of tumors. In these models calorie
restriction potentiated the effects of w3 fatty acids (160, 161)
whereas w6 fatty acids in the form of corn oil increased tumor
formation, size, and number (16 1, 162). It also was shown that
c(,3 fatty acids decrease POE2 production in animals fed w3 fatty
acids ( 162). As expected, fatty acid analysis of the transplanted
tumors reflects the specific composition ofthe dietary fat ingested
by the host ( 162). Furthermore, these studies indicate that the
composition of dietary lipids modifies lipid metabolism and that
high dietary intake of co3 fatty acids can prevent or delay the
expression ofthese neoplasms. In other studies involving human
breast-cancer cells in nude mice, the mice fed w3 fatty acids had
fewer pulmonary metastases, decreased serum estrogen and pro-
lactin concentrations, less PGE2 in the tumor, and reduced
c-mvc oncogene mRNA concentrations in the tumor-tissue cells
(160). The opposite occurred in the corn-oil-fed mice.
Animal studies in progress are using fish oils to elucidate the
mechanisms involved, including the changes in prostaglandin
production, immune function, free radical formation, membrane
fluidity changes, modulation of intracellular transport systems,
hormone secretion, calorie utilization, and gene expression ( I 63).
Diabetes is a chronic disorder with complications includ-
ing hypercholesterolemia, hypertriglyceridemia, atherosclerosis,
CHD, and hypertension. Many of these complications are at-
tributable to microvascular disease (164). Jensen et al (165),
using a double-blind crossover design, studied the effects on en-
dothelial permeability, blood pressure, and plasma lipids of 8-
wk supplementation of a diabetic diet with cod-liver oil rich in
(1)3 fatty acids compared with 8-wk supplementation with olive
oil in 1 8 insulin-treated diabetic patients with albuminuria with
> 30 mg/d. The patients receiving cod-liver oil showed a sig-
nificant fall in mean transcapillary escape rate ofalbumin corn-
pared with baseline and a reduction in mean blood pressure. No
changes occurred with olive oil. Cod-liver oil was associated with
a significant increase in plasma HDL cholesterol, a significant
decrease in the concentration of VLDL cholesterol and triglyc-
eride, and no change in the concentration of LDL cholesterol.
Jensen et al concluded that cod-liver oil may have a direct action
on vascular permeability that is independent of its beneficial
effect on blood pressure and postulated that this action results
from a decreased transfer of lipoproteins into the vascular wall.
In this study blood glucose concentrations were unchanged. In
other studies the use offish oils in non-insulin-dependent diabetes
mellitus (NIDDM) (166, 167) and in insulin-dependent diabetes
mellitus (IDDM) ( 168) increased blood glucose, glycosylated
hemoglobin, plasma total cholesterol, LDL cholesterol, and
serum apo B. The magnitude of these effects is generally small
and the changes in glucose metabolism are associated with in-
creased hepatic glucose output and impaired insulin secretion
but unaltered glucose disposal rates. Further studies are needed
ofthe fatty acid composition ofphospholipids in diabetic patients
during w3 supplementation under defined conditions of meta-
bolic control, diet, and type of NIDDM and IDDM.
Omega-3 fatty acids as an adjuvant to drug therapy
Omega-3 fatty acids in combination with drugs for the treat-
ment ofdiseases is an area ofimrnense interest because it opens
a new field in nutrition research, w3 fatty acids in the control of
metabolic and autoimmune disorders, that includes CVD, ar-
thritis, lupus erythernatosus, psoriasis, ulcerative colitis, and
cancer. Preliminary data from animal and human studies suggest
that the concurrent ingestion or administration ofw3 fatty acids
with drugs leads to potentiation of drug effects, as with pro-
pranolol (1 5 1), which may lead to a decrease both in the dose
of o3 fatty acids and in the drug dose or, as with cyclosporin
(1 58), to a decrease in toxicity ofthe drug. By partially replacing
the fatty acids of phospholipids in the cell membranes, w3 fatty
acids modify enzymes, receptors, and other proteins (29). Ad-
ditional studies suggest that the incorporation of w3 fatty acids
by cell membranes is enhanced in the presence of olive oil and
linseed oil, emphasizing once again the importance of nutrient-
nutrient interactions (169).
Cyclosporin is used widely in organ transplantation and in
many individuals its use leads to impairments in renal function
(1 70) and increased thromboxane formation (1 7 1). It was noted
that the use of fish oil instead of olive oil as the vehicle for its
administration in rats led to attenuation of the cyclosporin
nephrotoxicity ( 172) without affecting thromboxane synthesis
(173). These findings led to the use of 3.6 g EPA/d and 2.4 g
DHA/d in patients with psoriasis who were taking cyclosporin:
there was a decrease in the decline in renal function without
cyclosporin pharmacokinetics being affected ( 174). In another
patient who had received a renal transplant, a low dose of w3
fatty acids (1 .5 g EPA/d) did not lead to the expected increase
in blood pressure that is produced by cyclosporin and, in fact,
a favorable reduction in thromboxane A2 occurred ( 175). There
were not any adverse effects attributable to the supplements. In
a randomized controlled study van den Heide et al ( 176) inves-
tigated the effects offish-oil supplements on cyclosporin therapy
in renal-transplant patients. The fish-oil supplements caused a
significant decrease in renal vascular resistance, increased gb-
merular-filtration rates, and lowered mean arterial pressure. This
is another example of the beneficial effects of w3 fatty acids
combined with drugs in which w3 fatty acids decrease drug tox-
icity and also improve the hemoclynamic aspects of illness.
Essentiality: the role of w3 fatty acids
in growth and development
In parallel with the studies investigating the robe of w3 fatty
acids in disease states, an outstanding group ofscientists turned
their attention to the essentiality of w3 fatty acids throughout
the life cycle.
Animal studies
Studies in rats and rhesus monkeys showed that dietary re-
striction of to3 fatty acids (primarily LNA) during pregnancy
and lactation interferes with normal visual function and may
even impair learning ability in offspring (177). Connor et al (178)
concluded that w3 fatty acid deficiency in rhesus monkeys is
characterized by reduced vision, abnormal electroretinograrns
(ERGs), and polydipsia (29). In this model, abnormalities of the
ERG induced by c3 fatty acid deficiency during development
appear to be irreversible. In 1987 Rotstein et al (1 79) studied
the effects ofaging on the composition and metabolism of DHA-
containing lipids of the retina in rats. They concluded that an
impairment of the -4 desaturase enzyme system is probably
responsible for the decreased concentrations of 22:6w3 (and 22:
5w6) observed in retinal lipids as a consequence ofaging. Because
DHA is required for normal function of photoreceptors in rats
and primates, it is quite possible that a decrease in DHA plays
an important role in visual impairments that accompany old
age. Therefore, dietary DHA rather than LNA would be the
appropriate co3 fatty acid to use in studies aiming to influence
the development ofvisual impairments and even improve visual
function in elderly people.
Human studies
Pregnancy. The role of w6 and w3 fatty acids in pregnancy
has been examined only recently in humans. Over the past 50
y the influence of maternal nutrition on fetal growth has been
extensively studied in the context ofprotein-caborie malnutrition.
In 198 1 Crawford et al ( 180) calculated the LA and LNA re-
quirements for pregnancy to be ‘- 1% ofthe nonpregnant worn-
an’s dietary energy and the AA and DHA requirements to be
another 0.5%.
Fetal development, human milk, and infint /i’eding. As far
back as 1973, Crawford et al (1 8 1) analyzed 32 samples of human
milk and found that it contained LA, AA, LNA, EPA, and DHA
and recommended their inclusion in infant formula. Infant for-
mula does not contain AA, EPA, or DHA (Table 9) (35).
EPA and DHA are higher in the erythrocytes of breast-fed
infants than in those of bottle-fed infants (182). The milk of
vegan women contains lower concentrations of DHA than does
the milk of omnivores and this difference is reflected in the
Fatty acid composition of human milk and formulas (molar percent)*
Fatty acidt
Human milk (n = 1 1) Portagen#{174}4 Enfamil Premature#{174}4 Similac Special Care#{174}§
8:0 (caprylic acid) 0.3 60
24.5 24.1
10:0 (capric acid)
1.4 24 14.1 17.7
12:0 (lauric acid)
7.0 0.42 12.2
8.0 Trace
4.7 5.8
l6:0(palmiticacid) 19.8
0.19 7.5 6.8
16:1 (palmitoleic acid) 3.2
18:0(stearicacid) 5.9 0.47
1.7 2.3
18:1 (oleicacid) 34.8 4.1 12.4 10.0
I8:2w6 (linoleic acid) 16.0
8.1 22.4 17.4
I 8:3w3 (s-linolenic acid) 0.6
Trace 0.6 0.9
20:1 (gondoic acid)
1.1 0.3 0.1
20:2w6 0.6
20:3w6 (dihomo-y-Iinolenic acid)
20:4w6 (arachidonic acid) 0.6
2O:5w3 (eicosapentaenoic acid)
22:1 (docosenoicacid) 0.1
22:4w6 (docosatetraenoic acid)
22:5w6 (docosapentaenoic acid) 0.2
22:5w3 (docosapentaenoic acid)
22:6w3 (docosahexaenoic acid) 0.2
a Reproduced with permission from reference 35.
t Common name in parentheses.
4 Values from Pediatric Products Handbook, 1983 Edition, Mead Johnson Nutritional Division, Evanston, IN.
§ Ross Laboratories. Columbus, OH.
erythrocytes of their infants. After birth there is a decrease in
the DHA content of erythrocytes of full-term and premature
infants (42. 183). Infants born at term and fed mother’s milk
had approximately twice as much DHA in erythrocyte phos-
pholipids as did infants fed formula containing LNA but not
DHA. Because the greatest amount ofDHA accumulation occurs
during the last trimester of pregnancy, the amount of DHA
available to premature infants assumes critical importance. In
1987 Liu et al (184) determined that 11 mg DHA-kg’-d’
added to formula resulted in 0.2% DI-IA in the total dietary fatty
acids in the formula. which is within the range ofO. 1-0.3% found
in human milk. The inclusion of0.2% DHA in the formula did
not decrease plasma AA and appeared to be a physiological
amount that could prevent declines in membrane DHA of pre-
mature infants.
A major question remaining is to what degree the fatty acid
pattern in erythrocytes reflects the neural status of w6 and w3
metabolites in humans.
The effects of PUFA deficiency on the developing brain have
been widely documented in experimental animals whereas in-
formation from humans is scarce. However, work by Martinez
et al ( 185- 188), Innis ( 189), Carlson ( 190), Neuringer et al (191,
192), and Bourre et al ( I 93) has added considerably to our
knowledge, much of which was pioneered by Lamptey and
Walker ( 194), Walker ( 195), Wheeler et al ( 196), Crawford et al
(197), and Clandinin et al(198).
During l8:3w3 dietary deprivation, DHA is replaced by 22:
5w6 in the retina and brain of animals. Replacement with 22:
5w6, the fatty acid most closely resembling DHA, suggests ac-
tivation of a cellular compensatory mechanism. The w3 fatty
acids are required by the membranes ofphotoreceptor cells and
synapses for 1) synaptogenesis and photoreceptor membrane
biogenesis during the perinatal period, 2) normal functioning of
tissues, and 3) response to injury to the nervous system (ischemia
and convulsions) and also during retinal stimulation, both of
which trigger the release ofDHA from membrane phosphobipids;
some ofthis DHA may be peroxidated or lost through washout
to the blood and may need to be replenished (199).
Bourre et al (193) showed that the brain of the w3-deficient
rat is more susceptible to environmental toxins and alcohol.
It is now well recognized that nutrition during the first weeks
of life can have a decisive influence on brain development. Be-
cause fatty acid patterns ofall organs change during development,
it is necessary to know the normal profiles during the various
stages of development to understand the role of nutritional in-
Martinez ( 185) studied the composition of w3 and w6 fatty
acids in brain, liver, and retina in human fetuses during the last
trimester of pregnancy. After 30 wk gestation there is a prefer-
ential desaturation ofthe long-chain w3 fatty acids in the brain.
The liver shows a similar profile. In both tissues, 22:6w3 increases
quadratically and 20:4w6 and 18: lw9 decrease linearly in phos-
phatidylethanolamine (PE). In the retina, as in the forebrain and
the liver, the proportion ofw3 fatty acids increases whereas that
of w6 fatty acids decreases throughout development. These
changes can be clearly illustrated by using the ratio of 22:6w3
to 20:4w6. In the human retina this ratio doubles between 24
wk of gestation and term and continues to increase with age.
These findings should be the guidelines for the feeding of pre-
maturely born infants.
Martinez and Ballabriga (1 87) also investigated the liver and
forebrain of infants who had received total parenteral nutrition
high in linoleate (Intralipid#{174}) for 4-12 d. At autopsy a signifi-
cantly lower-than-normal proportion of 22:6w3 was found in
liver phosphoglycerides. There were a number of other changes
in long-chain PUFAs and high concentrations of 18:2w6 that
were not consistent with the values noted in normal fetal de-
Martinez’s ( I 85) findings in the retinas of two postnatally
malnourished infants were similar to those described in the liver
ofchildren receiving high intravenous doses of 18:2w6. One of
the malnourished children had mucoviscidosis. Both children
had unusually high concentrations of22:5w6 in retina and phos-
phatidylcholine as a sign of DHA deficiency. One premature
infant (25 wk gestation) had received commercial milk formulas
with w6:w3 varying between I 8: 1 and 66: 1 during the first 4 mo
oflife. The retina ofthis infant was very deficient in 22:6w3.
It can be concluded that diets with a high w6:w3 can be con-
sidered unbalanced relative to human breast milk and that these
diets are damaging to the PUFA composition ofthe developing
central nervous system in humans. Martinez ( 185) stated that
“when high doses of l8:2w6 are given intravenously, the inhib-
iting effect on the series is very strong. even with a theoretically
correct omega-6:omega-3 ratio, probably because substrate in-
hibition adds to competition between families of fatty acids for
the desaturase systems. This should serve as a warning against
manufacturing such unphysiobogical fatty acid mixtures for use
in pediatric nutrition.”
Carlson ( 190) showed that membrane 22:6w3 was highest at
birth and declined with time in formula-fed infants and that fish
oil could be used as a source of 22:6w3 in the range found in
human-milk-fed infants. One month after delivery, preterm in-
fants not fed human milk had plasma phospholipid 22:6w3 more
like that of monkeys fed safflower oil, and the bow concentrations
seen in premature infants are analogous to those at which de-
monstrable deficits in visual acuity occur in infant monkeys.
Uauy et al (200) showed that premature infants fed formulas
with a high ratio of LA to LNA (30: 1) have poorer ERG re-
sponsiveness early in infancy than do those fed human milk or
formula with a lower ratio of LA to LNA (9: 1). The addition of
fish oil further improved some ERG responses. Visual-acuity
development was improved by fish-oil supplementation of for-
mula during the first half of infancy compared with formula
containing 1 .5-2.5% ofenergy as LNA. These data strongly sug-
gest that DHA is essential for the functional development of the
eye and brain of premature infants.
Children, adults, and elderly adults. Omega-3 fatty acid de-
ficiency was originally reported by Holman et al (201) in a young
child. Subsequently Bjerve et al (202) reported w3 fatty acid
deficiency in a child and in a group ofelderly patients in nursing
homes who were fed orally for several years by gastric tube with
diets containing very low amounts of w3 fatty acids. Studies
based on clinical findings and determinations of plasma and
erythrocyte lipids after dietary supplementation with soya and
cod-liver oil strongly suggest that the patients had w3 fatty acid
deficiency. These patients represent the first adults and the second
child described with w3 fatty acid deficiency. The results indicate
that w3 fatty acids are essential for normal growth and cell func-
tion in humans in ways similar to those in several animal species.
Assuming linear relationships between dietary intake of w3
fatty acids and the measured concentrations ofw3 fatty acids in
plasma and erythrocyte lipids, the optimal intake of 18:3w3 has
been estimated to be 800-1 100 mg/d whereas the optimal intake
of very-long-chain w3 fatty acids was estimated to be 300-400
mg/d ( 1 15). It is suggested that the dietary requirements of w3
and w6 fatty acids should be stated in milligrams or grams per
day and not only as a percent of energy.
Dietary implications
Omega-3 and 6 PUFAs are two classes of EFAs that are not
interconvertible and that constitute a significant part of practi-
cally all cell membranes. Whereas cellular proteins are genetically
determined and control important cellular functions indepen-
dently ofdietary intake, the lipid composition ofcell membranes
is dependent to a great extent on the composition of the diet.
When ingested, fatty acids such as EPA and DHA are incor-
porated into the sn-2 position in cell membrane phospholipids,
displacing AA. Because the fatty acid composition ofcell mem-
branes modulates important cell functions and because the fatty
acids in membranes are dependent on dietary intake, it is obvious
that in referring to PUFAs it is essential to distinguish between
w3 and co6 fatty acids in making dietary recommendations. Sim-
ply using the P-S ratio ofpolyunsaturated fatty acids to saturated
fatty acids (P:S) ratio is inappropriate and inadequate on the
basis of the knowledge we have today.
Many dietary studies and interventions have been carried out
and dietary recommendations have been made in relation to
saturated fatty acids and cholesterol. However, the amount of
o,3 fatty acids in the diet and their effects on health and disease
have not yet been considered in the development of dietary
guidelines by national governments except for the most recent
Canadian Nutrition Recommendations (Table 10) (203). Be-
cause w3 fatty acids have different metabolic effects than do w6
fatty acids and because w3 fatty acids are essential for normal
growth and development and for overall health, accurate knowl-
edge ofthe amount and type ofw3 fatty acids in foods is essential.
Both terrestrial and marine sources ofw3 fatty acids are impor-
tant in this regard.
It is now accepted that it is important to consider the functions
ofthe different types offatty acids (w3, w6, and w9) rather than
simply total fat (percentage ofcalories from fat) or the amount
ofpolyunsaturates. The question is not simply about the P:S in
the diet but about the concentrations of the w3, w6 polyunsat-
urated, and w9 monounsaturated fatty acids relative to saturated
fatty acids in the diet. Fatty acids should be considered in terms
of their overall metabolic effects in growth and development
and for their effects on serum lipids, inflammation, thrombus
formation, and tumor development. Trans fatty acids were re-
cently shown to elevate serum cholesterol (52). Stearic acid
formed during hydrogenation does not raise cholesterol but it
increases the risk ofthrombosis (204). Clearly there is a need to
define precisely the functions of the various fatty acids.
Because ofthe increased amounts ofc,6 fatty acids in our diet,
the eicosanoid metabolic products from AA, specifically pros-
taglandins, thromboxanes, leukotrienes, hydroxy fatty acids, and
lipoxins, are formed in larger quantities than those formed from
fatty acids, specifically EPA. The eicosanoids from AA are
biologically active in very small quantities and ifthey are formed
in large amounts they contribute to the formation of thrombus
and atheroma; to allergic and inflammatory disorders, particu-
larly in those who are susceptible: and to proliferation of cells.
Thus a diet rich in w6 fatty acids shifts the physiological state
to one that is prothrombotic and proaggregatory with increases
in blood viscosity. vasospasm. and vasoconstriction and decreases
in bleeding time. Bleeding time is decreased in groups of patients
with hypercholesterolemia (205), hyperlipoproteinemia (206),
myocardial infarction (207, 208) and other forms of atheroscle-
rotic disease (209), and diabetes (obesity and hypertriglycerid-
emia). Bleeding time is longer in women than in men and longer
Summary ofexamples ofrecommended nutrients based on energy expressed as daily rates5
Age and sex Energy
Thiamin Riboflavin Niacin w3 PUFAs w6 PUFAs
Mi (keal) mg mg
NEf g g
0-4 mo (M, F) 2.51 (600) 0.3 0.3
4 0.5 3
5-12 mo (M. F) 3.77 (900) 0.4
0.5 7 0.5
I y (M. F) 4.60 (1 100) 0.5 0.6
8 0.6 4
2-3 y (M. F) 5.44 (1300)
0.6 0.7 9 0.7
7.53(1800) 0.7 0.9
13 1.0 6
7-9 y
9.20 (2200) 0.9 1.1 16 1.2
F 7.95(1900) 0.8
1.0 14 1.0
10-12 y
10.5 (2500) 1.0 1.3 18 1.4 8
9.20(2200) 0.9
1.1 16
1.1 7
13-15 y
M ll.7 (2800) 1.1 1.4
20 1.4 9
9.20 (2200) 0.9 1.1 16
1.2 7
16-18 y
M 13.4 (3200) 1.3 1.6 23 1.8 11
8.79 (2100) 0.8 1.1 15 1.2
19-24 y
M 12.6 (3000) 1.2 1.5 22 1.6
F 8.79 (2100) 0.8 1.1
15 1.2 7
25-49 y
M 11.3 (2700) 1.1 1.4 19 1.5 9
F 8.37(2000) 0.8
1.0 14 1.1 7
50-74 y
M 9.62 (2300) 0.9 1.3 16 1.3 8
F 7.53(1800)
0.84 1.04
144 1.14 74
75+ y
M 8.37 (2000) 0.8 1.0 14 1.0 7
7.11 (1700) 0.84 1.04 144 1.14 74
Pregnancy (additional)
1st trimester 0.42 ( 100) 0. 1 0. 1 0. 1
0.05 0.3
2nd trimester 1.26 (300) 0.1 0.3 0.2 0.16 0.9
3rd trimester 1.26 (300) 0.1
0.3 0.2 0.16 0.9
Lactation (additional) 1.88 (450) 0.2 0.4
0.3 0.25 1.5
S From reference 203.
t Niacin equivalents.
4 Value below which intake should not fall.
§ Assumes moderate physical activity.
in young than in old people. There are ethnic differences in
bleeding time that appear to be related to diet. The increase in
bleeding time brought on by the ingestion of fish or fish oils is
an attempt to return to a more physiological state.
Bleeding time is determined by platelet function, local tissue
factors, and components of the coagulation system. Rodgers and
Levin (2 10) carried out a critical reappraisal ofthe bleeding time
and concluded that there is no evidence that bleeding time is a
predictor of hemorrhage risk and summarized their findings as
follows: “The pathophysiobogy of an abnormal bleeding time
remains poorly understood. The bleeding time is affected by a
large number of diseases, drugs, physiologic factors, test con-
ditions, and therapeutic actions, not all ofthem platelet-related.
The test is likely to remain widely used for the diagnosis of
inherited disorders of platelet function, such as von Willebrand’s
syndrome. despite the lack of clear criteria for its use in this
context. There are no data that support use ofthe bleeding time
to predict bleeding: there is no evidence that the test changes
sufficiently in advance of serious bleeding to allow successful
intervention, that the risk of bleeding for a given bleeding time
is independent ofthe cause ofthe prolongation, or that bleeding
from the skin can predict bleeding elsewhere in the body (for
example, duration of bleeding from a skin wound does not cor-
relate with duration ofbleeding from a gastric biopsy site). There
is no evidence that the bleeding time is useful for monitoring
the effects of hemodialysis or transfusion therapy.”
Evidence that long-chain w3 fatty acids protect against the
development of CVD continues to accumulate from epidemi-
obogic surveys, animal-feeding studies, biochemical research, and
human clinical trials and intervention studies. Most investigators
advise that addition of fish to the diet several times weekly may
be ofbenefit in preventing CHD. There is insufficient evidence
to quantify the exact prophylactic benefit. Yet the epidemiologic
evidence from the Eskimo (1 3), the Japanese (15), the Oslo stud-
ies (16) and from population-intervention studies and clinical
trials are highly suggestive and support the hypothesis that ,3
fatty acids are contributing factors in the prevention of CHD
and in the control of blood pressure.
In considering dietary implications it is necessary to distinguish
among the various roles of w3 fatty acids:
1) Omega-3 fatty acids are essential for normal growth and
development throughout the life cycle and they must be included
in the diet of pregnant women, premature infants, full-term in-
fants. children, young adults, and elderly adults. As indicated
in the previous section, the optimal intake of 18:3w3 was esti-
mated to be 800-1 100 mg/d and that of very-long-chain w3
fatty acids to be 300-400 mg/d; the current amount in the US
population is 50 mg EPA and 80 mg DHA per capita per day,
indicating that the present diet does not provide adequate
2) Increased intake of fish or fish oils may be necessary over
and above the amount determined for their essentiality, partic-
ularly in those who have a family history or other evidence of
susceptibility to CHD, hypertension, arthritis, psoriasis, and
3) Omega-3 fatty acids are potentially valuable as adjuvants
to treatment ofsome ofthese diseases.
It is interesting to consider the progress that has taken place
in dietary recommendations from 1985 to 1990. At the Con-
ference on the Health Effects ofPolyunsaturated Fatty Acids in
Seafoods in 1985, it was noted that in the United States the per
capita consumption was 5.9 kg fish/y, which is equivalent to
about one fish meal per week (24). A recommendation was made
to increase this amount to two to three fish meals per week. It
appears that the American public is responding because the cur-
rent per capita consumption has increased to 6.8 kg fish per
It was further recommended at the 1985 conference that total
fat intake should be 30% of total calories, with 10% being sat-
urated fatty acids, 10% monounsaturated fatty acids, and 10%
PUFAs, the latter being equally divided between w6 and w3.
Therefore physicians should recommend the substitution of fish
for meat so that fish is eaten at least twice per week. Alternatively,
individuals who do not like fish and cannot get their w3 fatty
acids from their diet can get them from supplements.
An ideal supplement would contain high amounts of EPA
and DHA but little or no cholesterol or vitamins A and D. Vi-
tamin E should be added to prevent oxidation of these highly
unsaturated fatty acids. Products that conform to most of these
requirements are produced from fish oils pressed or extracted
from the flesh of the fish.
In the lay press there is confusion between fish-oil supplements
and cod-liver oil. Cod-liver oil extracted from the liver is rich
in vitamins A and D. Before the 1940s cod-liver oil was ingested
mainly by children as a source of vitamins A and D. In the
United Kingdom cod-liver oil is standardized by its vitamin
content rather than its fish-oil content. In contrast, fish-oil sup-
plements from the flesh offish contain only negligible amounts
of vitamins A and D.
At the most recent conference, the Second International Con-
ference on the Health Effects ofOmega-3 Polyunsaturated Fatty
Acids in Seafoods(March 20-23, 1990), the following statement
was released:
. Based upon clear evidence of an essential role for w3 fatty
acids in human development and health, the scientists rec-
ommended that all infant formula and diets for humans
should include w3 fatty acids, and they expressed concern
that steps be taken to stop marketing enteral and parenteral
formulas that fail to include any w3 fatty acids.
. The researchers urged that the appropriate government
agencies officially recognize the vitally important differences
between co3 and c,6 polyunsaturated fatty acids. Estimates
of the average w3 nutrition consumption in the U.S. pre-
sented by USDA scientists agreed with new nutrient mea-
surements reported from a NHLBI study, with both studies
indicating inadequate supplies ofw3 fatty acids in the typical
American diet.
. New evidence with an extremely high level of statistical
precision, from the National Heart, Lung and Blood Insti-
tute study, suggests that the daily dietary intake of 0.5 to
I .0 grams of long chain w3 fatty acids per day reduces the
risk ofcardiovascular death in middle aged American men
by about 40%, and some new data suggests that w3 fatty
acids may also decrease cancer mortality.
. The research reports make it increasingly evident that eating
o3 fatty acids can have beneficial effects on chronic inflam-
matory and cardiovascular diseases.
Recently Canada published its 1990 nutrition recommenda-
tions (203). As can be seen from Table 10, the Canadian nutrition
recommendations include separate recommendations for the two
classes of PUFAs. The amounts of w3 and u,6 fatty acids are
given in grams based on energy expressed as daily rates for the
various age groups from birth to 75+ y. For pregnancy additional
w3 and w6 fatty acids are recommended in amounts that increase
from the first to the second trimester. There is no increase be-
tween the second and third trimester. Additional w3 and w6
fatty acids are recommended during lactation.
Omega-3 fatty acids, LNA, EPA, and DHA, have been part
of the human diet throughout evolution. Modern agriculture
and aquaculture and the industrial revolution have led to changes
in the production of both plants and animals and to marked
changes in the composition of the food supply of Western so-
cieties. Specifically, the amount of w6 fatty acids in the food
supply has increased and that of w3 fatty acids has decreased
during human evolution from an estimated ratio of 1: 1 for w6:
w3 to 10: 1 or 20-25: 1, based on various estimates.
Omega-3 fatty acids are found in human milk. Earlier animal
studies with rodents and nonhuman primates and recent studies
with premature infants have shown that DHA is essential for
the normal function of retina and brain. Additional studies in-
dicate that w3 fatty acids are essential throughout the life cycle
and many scientific groups have recommended establishing rec-
ommended dietary allowance so that w3 fatty acids will be in-
eluded in infant formulas and in enteral and parenteral solutions.
Furthermore, current data support recommendations for the
general public. The 1990 Canadian nutrition recommendations
already include specific amounts for w3 and w6 fatty acids (in
g/d) for the various age groups, with additional amounts rec-
ommended during pregnancy and lactation. Thus, the great
progress that has been made in nutrition research on the role of
w3 fatty acids in health and disease is already incorporated into
nutrition policy in Canada.
Many investigators worldwide have examined the role of w3
fatty acids in disease states. About 2000 studies involving animal
models, tissue cultures, clinical investigations, and randomized
double-blind clinical trials have been reported in the past 6 y in
the world literature. Such studies include normal subjects and
patients with atherosclerosis; CHD: hypertension; inflammatory
and autoimmune disorders such as arthritis, psoriasis, and ul-
cerative colitis: and a number ofanimal models for research on
The majority ofstudies have been earned out in patients with
CVD. The role of w3 fatty acids, particularly EPA and DHA,
has been investigated in terms of their hypolipidemic, anti-
thrombotic, antiarrhythmic, antihypertensive, and anti-inflam-
matory aspects. Mechanisms include studies on eicosanoid pro-
duction and metabolism, cytokine production and suppression,
plasminogen activator production, and gene expression. Many
of these studies indicate that w3 fatty acids appear to decrease
or inhibit risk and precipitating factors in the development
of CVD.
Although an increase in consumption of w3 fatty acids alone
clearly will not eradicate CVD, it is increasingly evident that
increasing the amount of w3 fatty acids in the Western diet by
eating fish or supplementing the diet with fish oils may help in
the prevention of heart disease as well as in the prevention or
amelioration of other disease states.
Since this paper was completed in November 1990, a paper
was published in the December 1990 issue of The American
Journal of Clinical Nutrition entitled “Erythrocyte fatty acids,
plasma lipids, and cardiovascular disease in rural China” by
Wenxun et al (2 1 1). In this paper the authors concluded that
within China neither plasma total cholesterol nor LDL choles-
terol was associated with CVD. A strong inverse correlation be-
tween erythrocyte oleate concentrations and CVD was observed.
However, erythrocyte oleate concentrations were not associated
with plasma cholesterol but were strongly negatively associated
with arachidonate concentrations, suggesting potential dimi-
nution of CVD by oleate through reduced platelet aggregabil-
ity. This study then implicates arachidonate in contributing
toCVD. ci
I thank William Harris, Alexander Leaf, and Dwight Robinson for
their review and comments.
I . Ahrens EH, Blankenhorn DH, Tsaltas iT. Effect on human serum
lipids ofsubstituting plant for animal fat in the diet. Proc Soc Exp
Biol Med 1954;86:872-8.
2. Keys A, Anderson iT, Grande F. Serum cholesterol response to
dietary fat. Lancet l957;l:787(letter).
3. Malmros H, Wigand G. Report ofthe Minnesota Arteriosclerosis
Symposium. Minn Med 1955;38:864.
4. Keys