Caballero B. (ed.) Encyclopaedia of Food Science, Food Technology and Nutrition. Ten-Volume Set

Подождите немного. Документ загружается.

(1) antioxidant nutrient components; (2) absorption

and bioavailability; (3) food processing and storage;

(4) food additives and nutritional supplements;

(5) chiral form and other chemical characteristics of

nutrients. A dietary antioxidant is a substance that

significantly decreases the adverse effects of reactive

oxygen species, reactive nitrogen species, or both on

normal physiological functions in humans. The diet

contains antioxidants recognized as essential nutri-

ents or nonnutrients, and such nutrient antioxidants

include: vitamin E (tocopherols and tocotrienols),

vitamin A, ascorbic acid, and nutrients essential for

the normal function of endogenous enzymatic anti-

oxidant systems, such as Cu, Mn, Zn (cofactors of

superoxide dismutase), Se (cofactor of glutathione

peroxidase) and Fe (cofactor of catalase and peroxid-

ases). The diet also contains nonnutrient antioxi-

dants, including a large number of phytochemicals.

Dietary Antioxidants

Mechanisms of Antioxidant Action and Relation

with CHD Pathophysiology

0004 Vitamin E (tocopherols and tocotrienols) Vitamin E

is the generic term used to describe at least eight

naturally occurring compounds that exhibit the bio-

logical activity of a-tocopherol. It includes a-, b-, g-,

and d-tocopherol and a-, b-, g-, and d-tocotrienol.

The naturally occurring RRR stereoisomer and the

synthetic all-rac-a-tocopherol are equally well absorb-

ed, yet levels of a-tocopherol in the blood and tissues

increase significantly more than all-rac. In the liver,

the a-tocopherol transfer protein preferentially

selects RRR-a-tocopherol over g- and other toco-

pherols for incorporating into very-low-density lipo-

protein (VLDL), which then enters into systemic

circulation. Tocotrienols appear in the blood and

tissues at significantly lower levels than tocopherols,

even when ingested at equivalent or higher amounts.

0005 The antioxidant activity of the tocopherols and

tocotrienols is mainly due to their ability to donate

their phenolic hydrogens to lipid free radicals.

a-Tocopherol acts as a chain-breaking antioxidant

by donating its phenolic hydrogen to the chain-

propagating lipid peroxyl radical (LOO



) in mem-

branes and lipoproteins and replacing the latter with

the less reactive a-tocopheroxyl radical (a-TO



). Al-

ternatively, a-tocopherol may react directly with the

initiating radical to prevent LOO



formation. LOO



may also be eliminated via a radical–radical reaction

with a-TO



. Ascorbate, ubiquinone 10 and other co-

antioxidants, donates electrons to the a-TO



reducing

it back to a-tocopherol. Both, a- and g-tocopherol,

the principal forms of vitamin E in the diets of

Europe and the USA, respectively, protect against

peroxynitrite-induced lipid peroxidation. This pro-

cess is inhibited more effectively by g-tocopherol

than a-tocopherol. These results suggest that

g-tocopherol can act as a trap for membrane-soluble

electrophilic nitrogen oxides, forming stable carbon-

centered adducts through its nucleophilic 5-position,

which is blocked in a-tocopherol. Besides the reac-

tions with free radicals, tocopherols also react with

singlet oxygen either by physical quenching or by

chemical reactions. In physical quenching, excited

state singlet oxygen (O

) is deactivated to

ground-state triplet oxygen (O

) through a

charge-transfer mechanism. The tocopherols also

react chemically with singlet oxygen and become des-

troyed after the reaction.

0006LDL particles may contain approximately

12 mol of vitamin E per mol of LDL. The oxidative

resistance of LDL is increased in vitamin E-supple-

mented individuals, especially at intakes 269–

336 mg of a-tocopherol equivalents, and there are

strong correlations between oxidative resistance and

LDL-a-tocopherol concentrations in these subjects.

In vitamin E-supplemented subjects, the rate of oxi-

dation was significantly decreased at 269 and 537 mg

of a-tocopherol equivalents per day, leading to the

conclusion that vitamin E was the most important

variable that determined oxidative resistance of

LDL. However, this relationship becomes weaker in

unsupplemented subjects. In contrast, some studies

on LDL oxidation have suggested that a-tocopherol

accelerates the peroxidation of LDL lipids under mild

free-radical generation and in the absence of co-

antioxidants, such as ascorbate and coenzyme Q

0007The bioactivity of



NO is particularly sensitive to

oxidative stress. The superoxide radical combines

readily with



NO in a diffusion-limited reaction to

form peroxynitrite. In atherosclerosis, hypercholes-

terolemia, hypertension, and diabetes, the superoxide

flux in the arterial wall is increased, leading to

reduced



NO bioavailability/activity. a-Tocopherol

localizes mainly in lipoproteins and membranes,

where it serves to scavenge lipid peroxyl radicals.

Despite this action, atherosclerosis is characterized

by lipid peroxidation within the arterial wall, even

in the presence of a-tocopherol. Experimental studies

have shown that arteries deficient in a-tocopherol

showed a dose-dependent impairment of



NO-

mediated arterial relaxation upon exposure to

oxLDL. In contrast, vessels containing a high amount

of a-tocopherol were resistant to this effect of oxLDL.

This effect may be accounted for by inhibition of

protein kinase C-a phosphorylation by a-tocopherol,

which will prevent protein kinase C stimulation by

oxLDL. One hallmark of atherosclerotic plaque is

1656 CORONARY HEART DISEASE/Antioxidant Status

proliferarion of vascular smooth muscle. Consider-

able in-vitro data indicate that a-tocopherol inhibits

the proliferation of smooth muscle cells. This effect

was also associated with the inhibitory effect of a-

tocopherol on protein kinase C. Thrombus formation

within the vessel lumen, an important event in patho-

physiology of cardiovascular disease, is often precipi-

tated by the adhesion and aggregation of platelets to a

ruptured atherosclerotic plaque. It has been reported

that platelet a-tocopherol content is an important

determinant in platelet responsiveness toward protein

kinase C-dependent stimuli leading to inhibition of

platelet aggregation. Moreover, platelet a-tocopherol

status appears to be important in regulating



NO and

superoxide radical production, which could also con-

tribute to inhibit platelet aggregation. Recent data

indicating that patients treated with vitamin E

showed a high risk for hemorrhagic stroke could be

related to its modulation on platelet function.

0008 The antioxidant efficacy of tocotrienols in mem-

branes is higher than that of tocopherols, in spite of

their uptake and biodistribution after oral ingestion

being lower than that of a-tocopherol. Cell-culture

studies indicate that tocotrienols inhibit cholesterol

synthesis by directly regulating the expression of

3-hydroxy-3-methylglutaryl-coenzyme A reductase

(HMGR), through a posttranscriptional process in-

volving accelerated degradation of the reductase

protein. The molecular mechanism for this suppres-

sion by tocotrienols was ascribed to their side-

chain’s unique ability to increase cellular farnesol,

a mevalonate-derived product, which signals the

proteolytic degradation of HMGR. Some studies

in humans have shown that tocotrienol supplements

decreased LDL-cholesterol, apolipoprotein B (apoB),

lipoprotein (a) (Lp(a)), thromboxane B

, and platelet

factor 4, suggesting hypolipidemic and antithrombo-

tic effects. However, supplementation studies with a-

tocotrienyl acetate, which is hydrolyzed, absorbed,

and detectable in human plasma, did not lower chol-

esterol in hypercholesterolemic subjects, although it

was potent in decreasing LDL oxidizability. In-vitro

studies with HepG2 cells suggest that a-tocotrienol is

effective at levels of 10 mmol l

1

. Interestingly, g- and

d-tocotrienol, which lack the 5-methyl substituents

present in a-tocotrienol, show a higher HMGR sup-

pression. This structure–activity relationship indicates

that in addition to the requirements of the prenyl side-

chain for HMGR supression, changes in the methyl

substitution on the chromanol ring may also lead to a

divergent effect on HMGR activity. The conflicting

results obtained in human studies might also be related

to differences in the effective concentrations in cells

necessary to inhibit HMGR. Moreover, it has recently

been found that humans do not respond uniformily to

the cholesterol-lowering action of tocotrienols, par-

ticularly when cholesterol and alcohol intakes are

not controlled. Tocotrienols, as well as other HMGR

inhibitors, lower apoB levels partly by upregulating

LDL receptors in the liver. This facilitates the liver

uptake of circulating apoB-LDL. In addition, tocotrie-

nol has been shown to increase the intracellular pro-

teolytic degradation of apoB and alter the assembly

process of VLDL. Thus, it seems that the ability of

tocotrienol to reduce apoB levels in blood plasma

depends on both the clearance rate of LDL and the

production of VLDL. A novel tocotrienol fraction

from specially processed rice bran oil enriched with

didesmethyl-tocotrienol (with no methyl group on the

chromanol ring) has been shown to decrease plasma

Lp(a). Additionally, long-term prevention studies in

humans are needed to validate this effect, which may

have important implications in the prevention of ath-

erosclerosis and thromboembolism.

0009Carotenoids Carotenoids are pigments found only

in plants and microorganisms. Most carotenoids in

the diet are provided by deeply pigmented vegetables

and fruits. Nearly 600 of these compounds have been

identified in nature. Less than 10% of the carotenoids

can be metabolized to retinol and act as vitamin A

precursors. The predominant carotenoids found in

blood plasma (i.e., about 90%) are b-carotene, lyco-

pene, lutein, b-cryptoxanthin and a-carotene. The

structure of carotenoids is a key determinant of their

physical properties, chemical reactivity, and bio-

logical functions. The unique chemical features of

each carotenoid, such as size, shape, hydrophobicity,

and polarity, determine its ability to be incorporated

into the molecular microenvironment and its bio-

logical function. Thus, when carotenoids are con-

sumed by humans from the diet, these structural

properties influence the absorption, bioavailability,

and biodistribution, and may affect their actions

at the subcellular level and on biochemical pathways.

Carotenoids exist in different geometric forms (cis-

and trans-isomers), which can be interconverted

by light, thermal energy, or chemical reactions. After

passive absorption by the enterocyte, unmetabo-

lized carotenoids are incorporated into chylo-

micra followed by the uptake of remnants by liver.

Cis-b-carotene appears in blood and tissues at signifi-

cantly lower concentrations than the corresponding

trans form, even when ingested at equivalent or

higher amounts. After absorption, the cis form of b-

carotene is converted to the trans form. This conver-

sion seems to increase the bioavailability of the trans

form at the expense of the cis form. The more non-

polar carotenoids (e.g., b-carotene, a-carotene, lyco-

pene) are predominatly within LDL. The more polar

CORONARY HEART DISEASE/Antioxidant Status 1657

carotenoids (e.g., lutein) are associated with high-

density lipoprotein (HDL). Approximately four ca-

rotenoid molecules are associated with each VLDL

and one with each LDL particle, whereas for every

1000 HDL particles, only one contains carotenoids.

An inverse relation between body mass and plasma

concentrations of carotenoids, irrespective of dietary

carotenoid intake, has been found in both epidemi-

ological and clinical studies. High concentrations of

carotenoids are found in tissues that are rich in LDL

receptors, such as the corpus luteum, adrenal gland,

and testes.

0010 Carotenoids are efficient quenchers of singlet

oxygen and directly scavenge free radicals. In com-

parison, vitamin A is a poor antioxidant. Lycopene

exhibits a better antioxidant capability than b-

carotene and lutein. The in-vivo antioxidant actions

of carotenoids is mainly due to inhibition of lipid

peroxidation. Moreover, cell-culture studies and

very small clinical trials suggest that lycopene may

inhibit macrophage cholesterol synthesis, secondary

to the inhibition of HMGR, and increases macro-

phage LDL receptors. It is worth noting that lipid-

lowering pharmacological agents can reduce serum

concentrations of lycopene, via a reduction in lipo-

protein particle size and competition between the

drug and carotenoids for incorporation into VLDL.

The biological link between high levels of carotenoids

and a reduced risk of cardiovascular diseases has been

suggested to be the antioxidant effect of carotenoids

in LDL. Moreover, carotenoids have been detected in

lipid-rich atherosclerotic plaques, although concen-

trations are very low. The serum concentrations of

b-carotene achieved with supplementation are much

higher than those achieved with the consumption of a

carotenoid-rich diet. This fact may have contributed

to the adverse effects observed in some intervention

trials (ATBC, Table 1). Thus, a diet intervention strat-

egy is advantageous considering that: (1) it is risk-free

because, until now, no adverse effects have been

associated with consuming more carotenoids from

food; and (2) benefits may be gained, even if other

biologically active compounds, other than carote-

noids, are the truly active agents or act in synergism

with carotenoids.

0011 Ascorbic acid Ascorbic acid (vitamin C) is an essen-

tial micronutrient required for normal metabolic func-

tions in the organism. Like other primates, humans

lost the ability to synthesize ascorbic acid, as a result of

a mutation in the gene coding for l-gulonolactone

oxidase, required for the biosynthesis of ascorbic

acid via the glucuronic acid pathway. A lack of ascor-

bic acid in the diet causes the disease scurvy, which can

be prevented with 10 mg per day, an amount easily

obtained by consumption of fruit and vegetables.

Tissue saturation in healthy men occurs at ascorbic

acid intakes of approximately 0.1 g per day. Thus,

further increases over this intake may have minimal

or no additional effect on tissue ascorbic acid concen-

tration, and hence disease risk. Ascorbic acid is a

cofactor for several enzymes involved in the biosyn-

thesis of neurotransmitters, collagen, and carnitine.

Ascorbic acid has also been implicated in the catabol-

ism of cholesterol to bile acids by the enzyme choles-

terol 7a-monooxygenase and in the steroid

metabolism in the adrenals. The role of ascorbic acid

in these metabolic pathways is basically to reduce the

central metal ion of the mono- and dioxygenases,

acting as a cosubstrate in these reactions. Ascorbic

acid is an important water-soluble antioxidant in bio-

logical fluids. Ascorbic acid scavenges reactive oxygen

and nitrogen species, such as superoxide, hydroxyl,

nitroxide and aqueous peroxyl radicals, singlet

oxygen, ozone, peroxynitrite, nitrogen dioxide, and

hypochlorous acid. Two major properties of ascorbic

acid make it a strong antioxidant: (1) the low one-

electron reduction potentials of both ascorbate and its

one-electron oxidation product, the ascorbyl radical,

which allows both forms to react with and reduce

basically all physiologically relevant oxidants; (2) sta-

bility and low reactivity of the ascorbyl radical. The

latter readily dismutates to form ascorbate and dehy-

droascorbic acid, or is reduced back to ascorbate by an

NADH-dependent semidehydroascorbate reductase.

Ascorbic acid can also act as a co-antioxidant by

regenerating a-tocopherol from a-TO



, although

in vivo, this interaction is not clear.

0012The effect of ascorbic acid on hypercholesterolemia

has been investigated in numerous studies, although

results are still controversial. In one supplementation

study, consumption of 1.0 g of ascorbic acid per day

for 4 weeks resulted in a reduction in total choles-

terol, whereas in another study, supplementation

with (0.060–6.0) g per day for 2 weeks had no effect.

The positive effect of ascorbate may be related to its

role as a cofactor for cholesterol 7a-monooxygenase,

or, its modulating effect on HMGR. Several observa-

tional studies have found a significant association

between elevated plasma ascorbic acid and increased

concentrations of HDL-cholesterol and reduced con-

centrations of LDL-cholesterol. In relation to throm-

bosis, two studies found an inverse association

between serum ascorbate concentrations and coagu-

lation factors, as well as a positive association be-

tween low serum ascorbate and elevated coagulation

activation markers. However, these effects were not

confirmed by further studies. In-vitro studies have

shown that physiologic concentrations of ascorbic

acid increase PGE

(protaglandin E

) and PGI

1658 CORONARY HEART DISEASE/Antioxidant Status

(prostacyclin) production, reducing platelet aggrega-

tion and thrombus formation. Low concentrations of

ascorbate have also been associated with increased

concentration of plasminogen activator inhibitor 1,

a protein that inhibits fibrinolysis. High doses of

ascorbate, administered either orally or by intra-

arterial infusion, have shown beneficial effects on

vasodilation. Four studies investigated vasodilation

in patients with CVD and found increases of

45–220% in vasodilation after administration of

ascorbate (1.0–2.0 g oral or 0.025 g min

1

infusion).

A 100% reversal of epicardial artery vasoconstriction

was observed in coronary spastic angina patients in-

fused with 0.010 g of ascorbate per minute. In the

studies reporting an inverse association between

plasma ascorbate and angina pectoris and CHD, the

association was reduced after adjusting for smoking,

suggesting that smokers may need additional ascorbic

acid intakes. The possible mechanisms to explain

the positive effect of ascorbate on vasodilation are

related to its antioxidant activity and are suggested

as the following: (1) ascorbate may spare



NO by

scavenging superoxide radicals or preventing the for-

mation of oxLDL; (2) maintenance of intracellular

concentrations of glutathione by a sparing effect or

regeneration of thiols from thyil radicals, which may

enhance the bioavailability of



NO or increase

the stabilization of



NO through the formation of

S-nitrosothiols.

0013Phytochemicals (polyphenols) The secondary me-

tabolism of plants (e.g., shikimate and acetate path-

ways) generates products that are grouped under the

term phytochemicals or polyphenols. The term poly-

phenols includes several classes of compounds that

share a common structure; among them, flavonoids

constitute the most important group, including more

than 5000 compounds already identified. Subclasses

of flavonoids are flavonols, flavones, catechins, fla-

vanones, anthocyanidins, and isoflavonoids. Pheno-

lics are derivatives of benzene with one or more

hydroxyl groups associated with the aromatic rings.

Several in-vitro studies have shown that flavonoids

can modulate a variety of mammalian enzyme

systems. Some of these enzymes are involved in

important pathways that regulate cell division and

proliferation, platelet aggregation, and inflammatory

and immune responses. Soybean isoflavones, in

particular genistein and daidzein, have a weak proes-

trogenic or antiestrogenic activity, which enable them

to interact with estrogen receptors and decrease

serum cholesterol concentrations. In vitro studies

tbl0001 Table 1 Randomized trials of antioxidants in CHD

Study Participants Follow-up

(years)

Antioxidants Results

Primary

prevention trials

ATBC 29 000

male smokers

(50 – 60 years)

5 – 8 VE (50 mg per day) No effect on

coronary mortality (increased

incidence of lung cancer)b-Carotene (20 mg per day)

PHS 22 071 healthy males

(40 – 84 years)

12 b-Carotene

(50 mg every other day)

Supplementation confers

neither benefit nor harm with

respect to CHD or cancer

CARET 18 314 men and women

heavy smokers/asbestos

workers

4 b-Carotene

(30 mg per day)

Trend towards excess

cardiovascular deaths

(relative risk ¼1,26; 95%

CI ¼0.99–1.61)

Retinol (25 000 IU

per day)

Women’s

Health Study

40 000 healthy US health

professionals

(40 – 84 years)

Ongoing VE (600 mg every other day) Cancer and

cardiovascular endpointsb-Carotene (50 mg every

other day)

Su. Vi.MAX 12 735 healthy men and

women

Ongoing VC (120mg per day) Premature deaths, cancers,

cardiovascular diseases end pointsVE (30 mg per day)

b-Carotene (6 mg per day)

Selenium (100 mg per day)

Zinc (20 mg per day)

ATBC, a-tocopherol b Carotene Cancer Intervention Trial; CARET, b-carotene and Retinol Efficacy Trial; CI, confidence interval; PHS, Physicians’ Health

Study; Su. Vi.MAX, Supplementation en Vitamines et Mineraux Study; VC, vitamin C; VE, vitamin E.

CORONARY HEART DISEASE/Antioxidant Status 1659

have shown that genistein, a specific inhibitor for

tyrosine kinases, prevents the development of ather-

oma by inhibiting cell adhesion and proliferation, by

inhibiting LDL oxidation and by altering growth

factor activity. In addition, soy isoflavones improve

coronary vascular reactivity in female nonhuman pri-

mates, as well as, the endothelium-dependent flow-

mediated vasodilation and arterial compliance in

postmenopausal women. The phenolic compounds

found in tea (e.g., quercetin, kaempferol, myrecitin,

and epigallocatechin gallate) may be protective

against CHD. Some epidemiological studies have

correlated a high tea flavonoid intake with a lower

incidence of CHD (Table 2). Since the ‘French para-

dox’ was revealed, it has been demonstrated that

red wine, as well as dark grapes, contains many phen-

olic antioxidant compounds. In particular, trans-

resveratrol has been shown in vitro to act as: (1) a

strong antioxidant, (2) a vasorelaxant, (3) a phytoes-

trogen, and (4) an inhibitor of platelet aggregation,

cyclooxygenase-2, and polymorphonuclear leuko-

cytes activity. However, human studies are still

needed to confirm the effects of trans-resveratrol in

the protection against CHD. The antioxidant proper-

ties of olive oil phenolics, particularly oleuropein

and hydroxytyrosol, have also been demonstrated in

cellular and animals models. Among biological activ-

ities demonstrated by olive oil phenolics, the

following can be mentioned: inhibition of platelet

aggregation and human neutrophil respiratory burst;

reduced formation of thromboxane B

and leuko-

triene B

by human leukocytes; and increased pro-

duction of



NO by mouse macrophages. Cocoa

(Theobroma cacao L.) also contains appreciable

amounts of phenolic substances, and in-vitro and

ex-vivo studies indicate that they are endowed with

certain biological activities such as an antioxidant

capacity and immunoregulatory effects.

Epidemiological Studies on Antioxidants

and Coronary Heart Disease

0014In the last two decades, observational studies around

the world have suggested an inverse association be-

tween the consumption of fresh fruits and vegetables,

important sources of antioxidants, and the risk of

tbl0002 Table 2 Prospective studies on flavonoid intake and CHD

Population Age

(years)

Follow-

up (years)

Relative risk

(95% confidence

interval)

Cross-cultural study

Seven-countries Study 40–59 25 r ¼0.5; P ¼ 0.01

12 763 men with CHD

Cohort studies

Zutphen Elderly Study

(The Netherlands)

50–69 15 0.27 (0.11–0.70)

552 men

with stroke

Caerphilly Study (UK) 49–59 14 1.6 (0.9–2.9)

1 900 men with CHD

Finish Study

5 133 men þ women

with CHD

30–69 20 Men: 0.67

(0.44–1.00)

Women: 0.73

(0.41–0.32)

Relative risk of highest versus lowest flavonoid intake group, adjusted for

age, diet and other risk factors for CHD.

tbl0003 Table 3 Case-control studies relating antioxidants and CVD

Study Comparison groups Comments

Scotland, Finland,

and Southern

Italy men

Plasma VE, VC, b-carotene

were determined in angina

patients and healthy controls

The relative risk of angina

between the lowest and highest

quintiles of lipid-standardized

VE levels was 2.98 (95%

CI ¼1.07–6.70); after adjustment for

cigarette smoking, the relative risks for

VC and b-carotene were, respectively,

1.63 (0.76–3.49) and 1.41 (0.63–3.13); antioxidants

did not reflect regional CVD mortality rates

EURAMIC Adipose tissue levels of

a-tocopherol and b-carotene

were obtained in men from 10

European countries with first acute

MI and healthy controls

The multivariate odds ratio in the lowest

quintile b-carotene level

compared to the highest was 1.78 (95%

CI ¼1.17–2.71); the associations were

strongest among current and former

smokers. a-Tocopherol levels were

not associated with an apparent

increased risk, with an odds ratio

of MI in the lowest compared to the

highest category of 0.83 (95% CI ¼ 0.57–1.21)

MI, myocardial infarction; VC, vitamin C; VE, vitamin E.

1660 CORONARY HEART DISEASE/Antioxidant Status

CHD. Several case-control studies (Table 3) and

prospective cohort studies (Tables 2 and 4), compar-

ing rates of CHD with different plasma levels or

intake of antioxidants, have reported a significantly

reduced CVD incidence and mortality in subjects

or populations presenting a high intake or plasma

levels of vitamin E, vitamin C, and b-carotene, while

others have not supported these associations. How-

ever from the results of these studies, it was not

possible to establish a causal relationship and several

limitations. Some of the restrictions to the case-con-

trol studies are sample size, biases in cases and control

selection, quality of information on exposures, and

recall bias. Moreover, blood levels of antioxidants are

assessed with a single measurement some time after

the index event, thus precluding any distinctions be-

tween short- and long-term exposures. In addition,

storage of samples for long periods may decrease the

concentrations of antioxidants. The prospective

cohort studies are also limited by several factors: (1)

subjects taking vitamin supplements may also have a

healthier lifestyle than nontakers, such as practice of

physical activities, equilibrated diet, and no smoking;

(2) lack of precision and reliability of nutrient intake

tbl0004 Table 4 Prospective cohort studies of dietary intake and CVD

Study Participants Follow-up (years) Comparison groups Results

(relative risk; 95% CI)

Nurses’ Health

Survey (NHS)

121 000 female nurses

(USA) (34–59 years)

8 Upper vs. lower quintiles

of dietary and supplemental

vitamin E, vitamin C or

b-carotene intake

b-Carotene (0.78)

Vitamin E (0.66)

Vitamin C (0.80;

nonsignificant after

controlling for

vitamin E intake)

Health Professional

Follow-up Study (HPFS)

39 910 male health

professionals

(USA) (40–75 years)

4 Upper vs. lower quintiles of

dietary and supplemental

vitamin E, vitamin C or

b-carotene intake

b-Carotene (0.71)

Vitamin E (0.60)

Vitamin C (1.29)

First National Health and

Nutrition Examination Survey

(NHANES-1)

11 349 adults

(USA) (25–74 years)

10 Upper vs. lower dietary

and supplemental

vitamin C intake

Vitamin C (0.66)

Intake of other vitamin

supplements was not

taken into account

Iowa Women Health Study 34 486 postmenopausal

women (USA)

(55–69 years)

7 Upper vs. lower quintiles of

dietary and supplemental

vitamin E, vitamin A,

retinol, and

carotenoid intake

Vitamin E (0.96)

No benefit for Vitamin A,

retinol, and

carotenoids

2 748 men (Finland)

(30–69 years)

14 Upper vs. lower tertiles

of dietary and

supplemental

vitamins or

b-carotene intake

Vitamins (0.66)

b-Carotene (1.02)

2 385 women (Finland)

(30–69 years)

14 Upper vs. lower tertiles

of dietary and

supplemental

vitamins or

b-carotene intake

Vitamins (0.35)

b-Carotene (0.62)

1 899 men with

hyperlipidemia (USA)

(40–59 years)

13 Upper vs. lower

quartiles of serum

b-carotene intake

b-Carotene (0.64)

CORONARY HEART DISEASE/Antioxidant Status 1661

ascertained by food-frequency questionnaires; and (3)

confounding factors that were neither measured nor

controlled for in multivariate analysis.

0015 The randomized intervention trials (Tables 1 and 5)

have been considered as the golden standard to test

the hypothesis that antioxidant micronutrients

effectively have a causal relationship with the de-

crease of CHD morbidity/mortality. Until now, the

results of these trials have raised the possibility that

some benefits from observational epidemiology may

have been overestimated. b-Carotene and vitamin C

appear to confer no overall benefit in the primary

prevention of CVD among well-nourished subjects.

However, whether risk can be reduced among those

with disease or lower baseline levels of these micro-

nutrients remains unclear. For vitamin E, the results

are still controversial; while the results of the Cam-

bridge Heart Antioxidant Study (CHAOS) show a

tbl0005 Table 5 Randomized trials of antioxidants in CHD

Study Participants Follow-up (years) Antioxidants Results

Secondary prevention trials

CHAOS 2002 men and women with

angiographically proven CHD

510 days (3–981) VE (400 or 800 mg per day) 75% reduction in nonfatal MI;

nonsignificant increase in

cardiovascular death

HOPE 9 541 patients with CHD, stroke or pad 4.5 VE (400 mg per day) No effect on death from CVD

þ MI þ stroke

GISSI 11 324 men

and women with CHD

3.5 VE (300 mg per day) 4.7% (NS) reduction of total

mortality þ nonfatal MI

þ CVA; 35% reduction of

sudden death

WACS 8 000 women with CHD Ongoing VE (400 mg per day) MI, stroke, coronary

revascularization, and death

from CVD end points

VC (1 g per day)

b-Carotene (20 mg per day)

HPFS 20 000 patients with previous angina,

stroke, claudication, or diabetes

Ongoing VE (600 mg per day) Incidence of coronary

mortality and all-cause

mortality end points

VC (250 mg per day)

b-Carotene (20 mg per day)

CVA, cerebrovascular accidents; GISSI, Gruppo Italiano per lo Studio della Sopravvivenza nell’infarto; HOPE, Heart Outcome Prevention Evaluation;

HPFS, Heart Protection Study; MI, myocardial infarction; VC, vitamin C; VE, vitamin E: NS, nonsignificant; WACS, Women’s Antioxidant Cardiovascular

Study.

Lipid phase

Aqueous phase

ROOH

ROH

RNS

ROS

Lipids

Lipoproteins

Carotenoid

Dehydroascorbate

Carotenoid*

ROO /RO

Tocopherol

cycle

Asc

oxThiols

NAD(P)H

NAD(P)

+ H

TRX reductase

GSSG reductase

Lipoamida

Dehydrogenase

GSH-dependent

dehydroascorbate reductase

Thioredoxin

Thiol transferase

Thiol

cycle

redThiol

Asc

Ascorbate

cycle

Carotenoid

fig0002 Figure 2 Integrated action of antioxidants. Asc, ascorbate; Asc



, ascorbyl radical; GSSG, oxidized glutathione; carotenoid



, carot-

enoid-derived free radical; carotenoid

, excited-state carotenoids; oxThiols, oxidized thiols (glutathione, lipoic acid, thioredoxin); RO



alcoxyl; ROH, lipid-hydroxyl derived; ROO



, peroxyl radical; ROOH, lipid hydroperoxide; ROS, reactive oxygen species; RNS, reactive

nitrogen species; T, tocopherols; T



, tocopheroxyl radical; TT, tocotrienols; TT



, tocotrienoxyl radical; TRX, thioredoxin.

1662 CORONARY HEART DISEASE/Antioxidant Status

lower incidence of nonfatal infarction and major car-

diovascular events in supplemented subjects, HOPE

results indicate no benefit on chosen endpoints. There

are also criticisms of the randomized trials. Supple-

mentation of a diet with individual antioxidants

could not exert optimal effects because the concerted

mechanism of action of the antioxidant system

(Figure 1) could require the availability of adequate

and harmonized levels of several antioxidant com-

ponents in the diverse microenvironments in the or-

ganism. Furthermore, it may be possible that the

antioxidants chosen (vitamin C, vitamin E, and b-

carotene) could only be markers for the intake of

other, as yet unidentified, substance(s) with the true

biological effect. The rationale behind the approach

of antioxidant supplementation in order to prevent

the development and progression of CVD has not yet

been firmly established. More basic research is re-

quired to ascertain the appropriate antioxidant(s)

and doses, oxidation biomarkers, intermediate and

final end points, among other factors, before plann-

ing further large intervention trials. Thus, until the

results of these studies produce the final answers,

the consumption of fruits and vegetables and a

healthy life style would be advisable.

See also: Antioxidants: Natural Antioxidants; Synthetic

Antioxidants; Role of Antioxidant Nutrients in Defense

Systems; Ascorbic Acid: Properties and Determination;

Physiology; Carotenoids: Occurrence, Properties, and

Determination; Physiology; Soy (Soya) Beans: Dietary

Importance; Tocopherols: Properties and Determination;

Physiology

Further Reading

Antony MS (2000) Soy and cardiovascular disease: choles-

terol lowering and beyond. Journal of Nutrition 130:

662S–663S.

Carr AC and Frei B (1999) Toward a new recommended

dietary allowance for vitamin C based on antioxidant

and health effects in humans. American Journal of Clin-

ical Nutrition 69: 1006–1007.

Chisolm GM and Steinberg D (2000) The oxidative modifi-

cation hypothesis of atherogenesis: an overview. Free

Radical Biology and Medicine 28: 1815–1826.

Clinton SK (1998) Lycopene: Chemistry, biology and impli-

cations for human health and disease. Nutrition Reviews

56: 35–51.

Diplock AT, Charteux JL, Crozier-Willi G et al. (1998) Func-

tional food science and defence against reactive oxidative

species. British Journal of Nutrition 80: S77–S112.

Gaziano JM (2000) Dietary antioxidants and cardiovascu-

lar disease. Vitamins and Hormones 58: 299–320.

Hollman PCH and Katan MB (1999) Health effects and

bioavailability of dietary flavonols. Free Radical

Research 31: S75–S80.

Kamal-Eldin A and Appelqvist LA (1996) The chemistry

and antioxidant properties of tocopherols and tocotrie-

nols. Lipids 31: 671–701.

Keaney JF, Simon DI and Freedman JE (1999) Vitamin E

and vascular homeostasis: implications for atheroscler-

osis. FASEB Journal 13: 965–976.

Krinsky NI (1998) The antioxidant and biological proper-

ties of the carotenoids. Annals New York Academy of

Sciences 854: 443–447.

Marchioli R (1999) Antioxidant vitamins and prevention of

cardiovascular disease: laboratory, epidemiological and

clinical trial data. Pharmacological Research 40: 227–

238.

Packer L, Weber SU and Rimbach G (2001) Molecular

aspects of a-tocotrienol antioxidant action and cell sig-

nalling. Journal of Nutrition 131: 369S–373S.

Theriault A, Chao JT, Wang Q et al. (1999) Tocotrienol: a

review of its therapeutic properties. Clinical Biochemis-

try 32: 309–319.

Visioli F, Borsani L and Galli C (2000) Diet and prevention

of coronary heart disease: the potential role of phyto-

chemicals. Cardiovascular Research 47: 419–425.

Intervention Studies

P C Elwood, Llandough Hospital, Penarth, South

Glamorgan, UK

Definition and Aims

0001In the section on epidemiology a number of research

strategies are discussed. These fall into two obvious

groups: observation and intervention. In general, ob-

servation studies seek to ascertain more about the

natural course of a disease process, whereas interven-

tion studies test the effect of a possible therapeutic or

preventive measure. (See Epidemiology.)

0002There is a further division of intervention studies,

and this division is of particular importance in nutri-

tion research. First, there are trials in which a food-

stuff or a nutrient is given to subjects, and the effect

upon a mechanism or a metabolic process, or the

effect on a biochemical or hematologic variant is

observed. Measuring the change in blood lipids

following changes in dietary fat intake, or measuring

the blood pressure response to coffee prepared in

different ways, illustrates this kind of study. In these,

the numbers of subjects can be relatively small, be-

cause a measurement of the outcome variable (serum

cholesterol, blood pressure, or platelet aggregation,

etc.) is obtained for every subject. Furthermore, such

trials can usually be relatively short-term, and often it

is acceptable to use a cross-over design, in which each

CORONARY HEART DISEASE/Intervention Studies 1663

patient has a period on the intervention diet or sup-

plement being evaluated, and a period on a normal

diet without the intervention, thus serving as his or

her own control. This last can be a most efficient

design, provided the order of the intervention and

control periods is random for each subject, and pro-

vided an adequate washout time is allowed between

the two periods.

0003 In the second kind of intervention trial, the effect of

a dietary change, or a nutritional supplement, on the

incidence of a disease is studied. This might be the

reduction in the mortality of patients who have had a

heart attack and who are advised to eat fatty fish, or

the effect of a magnesium supplement given to patients

judged to be at risk of a heart attack or stroke. Trials

such as these are very different in that the outcome is a

disease event, and this is likely to occur in only a very

small proportion of persons in the trial. As a conse-

quence, the numbers involved have to be very large

and/or the period of observation very prolonged.

0004 Both kinds of trials involve intervention and the

level of confidence that one can put on the results is,

in general, much higher than in studies from which the

evidence is purely observational. Most of the design

features are the same in the two kinds of intervention

trial, but the problems which arise in those which aim

to change the progress or reduce the incidence of a

disease are substantial and so most of the following

discussion will focus on the second.

Design

0005 The only design for intervention studies which is now

acceptable, fulfills the basic criteria for randomized

controlled trial (RCT). These criteria include the

following:

0006 The allocation of subjects to the group which will

receive the intervention, or to the control group

which will not, must be random. Only allocation

at random is likely to insure that there are no

systematic differences or bias between the groups,

which might account for some of the effects found

and which might wrongly be attributed to the

intervention. In fact, the comparability between

the groups should later be checked by comparing

the distributions of age, social class, and perhaps

smoking, etc. in the two groups to give some as-

surance that by chance some systematic difference

or bias has not occurred.

0007 The advice, handling, and monitoring of the two

subgroups of subjects – those receiving the interven-

tion and those not – must be identical in every

respect other than the intervention being tested. As

far as possible the subjects in the trial should be

‘blind’ with regard to whether or not they are receiv-

ing the intervention under test. While this last is easy

in a drug trial by the use of placebo tablets, it is

unlikely to be possible in a trial of a foodstuff.

0008The outcome, whether this is a change in a bio-

chemical or other variant or a difference in disease

incidence or disease progression, must be assessed

by an observer who is blind as to whether or not

the subject has or has not been receiving the inter-

vention.

0009The setting-up and conduct of an RCT is a major

undertaking, and this is particularly true of a dietary

trial. It may be appropriate therefore to consider the

use of a factorial design. In this, two or more interven-

tions are separately randomized. Equal numbers of

subjects therefore receive each possible combination

of the interventions, with some receiving all and some

none. This design is acceptable provided that there is

no reason to expect strong interactions between the

different interventions. It is highly efficient as the main

effects are all tested simultaneously without loss of

power. For example, a trial by Burr and colleagues

evaluated the effects of three dietary changes on mor-

tality after a heart attack: a reduction in saturated fat,

an increase in fatty fish, and an increase in cereal fiber

consumption. Advice on each of these dietary changes

was given to half of the total patient group, selected for

each intervention at random.

0010The compliance of subjects with the intervention

being tested is of great importance and should be

taken into account when drawing conclusions from

a trial. In fact, it is important to monitor both the

degree of compliance by those who have been given

advice about the dietary factor being tested and the

extent to which the control subjects may spontan-

eously change their consumption of that factor.

Ideally, compliance should be monitored in an object-

ive way; for example, the estimation of plasma eico-

sapentaenoic acid will indicate how much fatty fish

has been consumed. If an appropriate objective meas-

ure is not available, then at intervals the subjects

(both those advised and the controls) can be asked

to keep a diary of the food they consume.

0011Every reasonable effort should be made to achieve

and maintain a high level of compliance. At the same

time, whether or not a subject complies with the

intervention, and whether or not a control subject

spontaneously adopts the measure under test, all the

subjects must be retained within their originally ran-

domized group, and their results analyzed accord-

ingly. This leads to what is known as an ‘intention

to treat’ analysis. Omission of patients from one or

other group because of poor compliance or other

reason can unbalance the comparability obtained by

1664 CORONARY HEART DISEASE/Intervention Studies

the original randomization, and is highly likely to

introduce bias and invalidate any conclusions drawn

from the results of the trial.

0012 In planning a trial, careful attention should always

be given to the likely power of a study, that is, the

likelihood that the planned trial will be large enough,

and the outcome measures sensitive enough to detect

a statistically significant beneficial effect, should one

occur. A number of things affect the power of a study.

The most obvious are the number of outcome events

and the likely difference in this between the interven-

tion and the control group. Clearly, this last will not

be known, otherwise the trial would be unnecessary.

An estimate usually has to be made. If the interven-

tion involves the consumption of a foodstuff, such as

fish, or fruit and vegetables, the reproducibility of

measurement of consumption should be assessed,

and any changes in the consumption of other foods

or nutrients monitored.

0013 All these estimates should be realistic. It is all too

easy to overestimate the likely effect of advice. For

example, dietary intakes are relatively fixed, particu-

larly in older subjects, and advice to increase the

consumption of, say, fruit and vegetables, or to

reduce fat intakes, has relatively little effect.

Primary and Secondary Trials in Coronary

Heart Disease

0014 The terms ‘primary’ and ‘secondary’ are used to distin-

guish trials which test a preventive measure in patients

who have not already had a clinical episode of a disease

(primary) from those based on patients who have al-

ready had a myocardial infarct or other clinical event

(secondary trials). A distinction of this kind is most

often made in coronary heart disease (CHD).

0015 In fact, the concept of primary and secondary trials

is not helpful. The terms relate to past history, and not

to future risk. In relation to risk of a future vascular

event, subjects simply form a continuum, ranging

from subjects at exceedingly low risk, such as

young, healthy subjects, to patients who have recently

had a vascular event. Ranged between these are sub-

jects who smoke, who have raised blood pressure,

raised cholesterol, etc., whose risk of a vascular

event is somewhere between those of subjects at the

two extremes. In the evaluation of the risk of any

individual, the occurrence of a past vascular event is

only one of a number of risk factors which determine

the likelihood of a future event. Figure 1 displays the

continuum of risk.

0016 Ideally, all possible preventive measures should be

tested in an RCT based upon younger healthy sub-

jects. The numbers required for a trial in such indi-

viduals are however formidable. Further difficulties

arise in persuading healthy subjects to comply with

the intervention to a sufficient degree and for a

sufficient time.

0017Intervention trials are therefore often conducted in

subjects who have evidence of an increased risk, be it

raised levels of risk factors, or some evidence of vas-

cular disease, such as angina pectoris. Trials can even

be based upon patients who have already had a CHD

event, such as a myocardial infarct. All such patients

know that they are at an increased risk of a further

event, and they are likely to be more easily persuaded

to comply with a dietary change for long periods. It

can be argued that the underlying disease of survivors

of a myocardial infarction is at an advanced stage and

may not be reversible within the time span of a trial of

a reasonable duration. There may well be some truth

in this, particularly if the aim of the intervention is to

reduce atherosclerosis.

0018Nevertheless, a myocardial or a cerebral infarct

results when a thrombus develops on top of an ather-

omatous plaque on the wall of a blood vessel. Dietary

factors are undoubtedly relevant to the risk of throm-

bosis. For example, the head of an intravascular

thrombus is usually a mass of aggregated platelets,

and alcohol reduces platelet activity. A myocardial

infarct can precipitate ventricular fibrillation and

death, and there is evidence that fish oil reduces the

electrical stability of the myocardium. Free radicals

from an ischemic lesion can damage the myocardium,

and antioxidants from fruit and vegetables may

reduce this damage to the heart muscle. There are

therefore mechanisms in vascular disease, other than

atherosclerotic disease in the vessel wall, which are

affected by dietary items and which are appropriately

tested in dietary intervention RCTs.

Young, active,

healthy subjects

Patients judged to be at

increased risk because of

age, smoking, raised blood

pressure, etc.

Low risk

High risk

Patients with a

recent vascular event

Risk of a vascular event

fig0001Figure 1 Low verus high risk of a vascular events

CORONARY HEART DISEASE/Intervention Studies 1665