THE CASE FOR INCREASING THE 
POPULATION CONSUMPTION OF FRUIT 
AND VEGETABLES AND THE EVIDENCE 
FOR THE EFFECTIVENESS OF 
INTERVENTIONS 
 
 
Dr Joyce Hughes 
April 2000 
 
 
 
 
1

---

Contents 
 
 
 
 
 
 
 
 
 
Page 
Summary 
 
 
 
 
 
 
 
Chapters 
1. 
 
Introduction 
       
2.  
Definition of fruit and vegetables 
3. 
Components and properties of fruit and vegetables with  
possible protective effects 
4. 
Biological and Experimental Evidence 
5. 
Cancer 
6. 
Cardiovascular Disease 
7. 
Other chronic diseases 
8. 
Vegetarian diets 
9. 
Mediterranean-type diet 
10. 
Genetic factors 
11. 
Burden of diseases that have an inverse association with the  
consumption of fruits and vegetables 
12. 
Current and further research 
13. 
Recommendations regarding intake of fruit and vegetables 
14. 
Current intake of fruit and vegetables 
15. 
Which ‘fruit and vegetables’ and how much is a ‘portion’? 
16. 
Dietary Interventions 
17. 
Conclusion 
18. 
Appendix 1 -  
19. 
Table 1 – Cancer Studies  
20. 
Table 2 – Cardiovascular disease studies 
21. 
Table 3 – Fruit and Vegetable Dietary Intervention Studies 
22. 
Annex 1 – Data from the NDNS survey of 4 to 18 year olds 
23. 
Annex 2 – MAFF funded studies on increasing fruit and  
vegetable consumption 
24. 
Annex 3 – DH and MAFF Current Research Projects on  
Dietary Interventions 
25. 
Annex 4 - HEA Paper Spring 2000 
26. 
Annex 5 – Reference list of papers covering USA 5A Day program 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2

---

Summary 
 
S.1. Introduction 
 
S.1.1  Fruit and vegetables contain many biologically active substances, including 
nutrients, and fibre which not only help to reduce the likelihood of developing chronic 
deficiency diseases, such as scurvy, but they may also help to reduce the likelihood of 
developing chronic diseases, such as coronary heart disease, stroke and some cancers.  
In addition they tend to have a low energy density which can help reduce the 
likelihood of developing obesity and non-insulin dependent diabetes mellitus 
(NIDDM).  Convincing evidence is also emerging about the positive role of fruit and 
vegetable consumption and a reduced incidence of cataracts, chronic obstructive 
pulmonary disease (COPD), and hypertension (see 1.1). 
 
S.1.2  This report reviews the evidence for the effectiveness of fruit and vegetables 
in promoting health and preventing disease, particularly cancer, coronary heart 
disease, stroke, and the case for increasing the population consumption of fruit and 
vegetables to at least five portions (400g or more) per day. It also looks at the 
effectiveness of campaigns and interventions to change dietary behaviour, taking into 
account not only community interventions, but also the effect of national policies, and 
national campaigns (e.g. the American 5 A Day campaign).   
 
S.1.3  The aim is to provide the scientific rationale for supporting enhanced 
intervention in the promotion of increased fruit and vegetable consumption as part of 
chronic disease risk reduction and healthy eating. 
 
S.2. 
Definition of fruit and vegetables 
 
S.2.1  The term “fruit(s) and vegetables” includes all fresh, frozen, canned and dried 
fruits and vegetables except potatoes. Beans and pulses are included (see 2.1). The 
exclusion of potatoes results from the fact that health advice consistently makes 
recommendations for starchy foods such as potatoes. COMA recommended a 50% 
increase in “bread, potatoes, and fruit and vegetables”.  In the UK potatoes are not 
generally used in the same way as other vegetables.  Most people would not consider 
having pasta with sauce and potatoes (see 2.2). 
 
S.3. 
Components and properties of fruit and vegetables with possible 
protective effects 
 
 
S.3.1  Antioxidants e.g. vitamin C, vitamin E, carotenoids and flavonoids 
Folate 
 Potassium 
and 
magnesium 
Phytoestrogens (lignans, isoflavones) 
 Plant 
sterols 
 
Sulphur containing compounds  
Carbohydrates (starch and fructose) and dietary fibre (NSP)  
Other phytochemicals e.g. glucosinolates, indoles, thiocyanates and 
isothiocyanates, protease inhibitors and limonene 
Low energy density 
 
 
3

---

S.3.2  These phytochemicals can have complementary and overlapping mechanisms 
of action, including modulation of detoxification enzymes, stimulation of the immune 
system, reduction of blood pressure, and antioxidant, antibacterial, and antiviral 
effects (see 3.2 and Chapter 4). 
 
S.4. 
Biological and Experimental Evidence 
 
S.4.1  There are many biologically plausible reasons why consumption of vegetables 
and fruits might slow or prevent the onset of chronic diseases, such as cancer, 
cardiovascular diseases, cataracts, obesity, NIDDM, COPD and hypertension. These 
include the presence of dietary antioxidants, vitamins C and E, the carotenoids, such 
as beta-carotene, lycopene and lutein, and the flavonoids, such as quercetin, luteolin 
and kaempferol. It is hypothesised, and there is some evidence, that these antioxidants 
lead to less oxidative damage, one of the mechanisms leading to chronic diseases such 
as cancers and coronary heart disease, usually resulting in lowered risk (see 4.1). 
However, the association of individual vitamin intake as opposed to fruit and 
vegetable consumption, and the risk of the development of various cancers is 
inconsistent (see 4.1.5). 
 
S.4.2  In addition fruit and vegetables contain many other classes of biologically 
active compounds and dietary fibre. Most of the data for the observations on the 
protective potential of all of these compounds have come from animal and in vitro 
studies.  At almost every one of the stages of the cancer process, identified 
phytochemicals are known to be able to alter the likelihood of carcinogenesis – 
occasionally in a way that enhances risk but usually in a favourable direction.  For 
example, glucosinolates and indoles (see 4.7.2), thiocyanates and isothiocyanates, 
phenols, and coumarins can induce a multiplicity of phase II  (solubilising and usually 
inactivating) enzymes (see 4.7.3); vitamin C and phenols block the formation of 
carcinogens such as nitrosamines (see 4.1.9); flavonoids and carotenoids act as 
antioxidants, essentially disabling the carcinogenic potential of specific compounds 
(see 4.1.18); lipid soluble compounds such as carotenoids and sterols may alter 
membrane structure or integrity; some sulphur-containing compounds suppress DNA 
and protein synthesis (see 4.6); carotenoids can suppress DNA synthesis and enhance 
differentiation (see 4.1); and phytoestrogens compete with oestradiol for oestrogen 
receptors in a way that is generally antiproliferative (see 4.4).   
 
S.4.3  The main components of fruit and vegetables likely to decrease the risk of 
cardiovascular disease are (i) the antioxidants, which are capable of reducing the 
oxidation of LDL cholesterol, and (ii) folate (and vitamin B6 and B12) which lowers 
plasma homocysteine levels. Even mild to moderate elevation of plasma 
homocysteine level is a strong risk factor for arteriosclerosis of the coronary, cerebral, 
and peripheral arteries and increased folate intake reduces this strong risk factor (see 
4.2).  Potassium and magnesium (and fibre) have been identified as significant 
modulators of stroke risk in men (see 4.3.1).  The protective effects were particularly 
pronounced in hypertensive subjects.  The observed effects may be due to direct and 
indirect effects of these nutrients on blood pressure and regulatory functions, such as 
endothelial function. 
 
S.4.4  Fruits and vegetables are relatively high in fibre and water content and tend to 
be lower in energy density (lower in calories) and fat content.  Diets which are lower 
 
4

---

in energy density and fat have been shown to reduce the risk of weight gain and hence 
obesity and NIDDM (see 4.9). 
 
S.5 Cancer 
 
S.5.1  The body of evidence that indicates that high intakes of fruit and vegetables 
are associated with a reduced risk of cancer at several sites is growing all the time. 
Overall, the epidemiological evidence on the link between diet and cancer provides 
consistent evidence for an inverse association between risk of certain cancers and the 
consumption of fruit and vegetables.  The association is generally most marked for 
epithelial cancers, apparently stronger for those of the digestive and respiratory tracts 
(oral, pharynx, oesophagus, stomach, larynx, large bowel/colon and rectum) as well as 
of the urinary tract, and somewhat weaker for hormone-related cancers (breast, 
endometrium) (see Chapter 5 and Table 1) 
 
S.5.2  Several reviews of epidemiologic studies have been carried out.  COMA’s 
Working Group on Diet and Cancer commissioned a review of the epidemiology 
relating diet to the development of cancers and considered the data derived from 
human experimental studies, both in vitro and in vivo, and reached the following 
conclusion for fruit and vegetable consumption and the risk of developing cancer (see 
5.3):   
 
Overall, there is moderate evidence that higher vegetable consumption would reduce 
the risk of colorectal cancer, and that higher fruit and vegetable consumption would 
reduce the risk of gastric cancer.  There is weak evidence, based on fewer data, that 
higher fruit and vegetable consumption would reduce the risk of breast cancer.  These 
cancers combined represent about 18% of the cancer burden in men and about 39% 
of the cancer burden in women in the UK.  Even a small reduction in relative risk 
would have important public health benefits in terms of the reduction in the absolute 
numbers of people affected.  In addition, the data are generally consistent with a 
graded reduction in risk for higher fruit and vegetable consumption and no cancer 
consistently shows a higher risk with higher fruit and vegetable consumption.  The 
overall picture, therefore, is consistent and supports the hypothesis that the 
consumption of fruits and vegetables protects against the development of some 
cancers.  The working Group recommends that fruit and vegetable consumption in 
the UK should increase. 
 
S.5.3  The evidence since COMA’s Working Group on Diet and Cancer published its 
report has grown and is reported in 5.3.2 to 5.3.21 (see also Table 1 of the Report).  
 
 
S.5.4  From a public health viewpoint, epidemiological evidence indicates that a 
substantial reduction in epithelial cancer risk can be obtained by increasing fruit and 
vegetable consumption in adults. Evidence is lacking from observational studies that 
increasing fruit and vegetable consumption in childhood will reduce the risk of cancer 
in adult life. The multi-stage processes involved in carcinogenesis and the wide 
variety of dietary factors which could influence each stage of the process means that 
the dietary contributors to the stages of carcinogenesis a decade or more before a 
cancer becomes apparent are difficult to identify even in detailed prospective studies 
(see 5.3.27).   
 
 
5

---

S.5.5  Recommendations to both adults and young persons (children) to make dietary 
changes, particularly to increase the consumption of vegetables and fruits, should be 
continued.  Compliance, even by a small proportion, could result in large numbers 
benefiting due to the burden from these chronic diseases (see 5.3.27). 
 
S.5.6  Understanding the association between fruit and vegetable intake and other 
health behaviours is important for properly interpreting the rapidly growing number 
of studies that link low intakes of fruits and vegetables to the risk of developing 
cancers. Because fruit and vegetable intake covaries with several other chronic 
disease risk factors, it is important to account for possible confounding between fruit 
and vegetable intake and other behaviours in etiologic studies of the risk of cancers. 
However, issues of confounding should not cause anyone to discount the value of 
eating more fruits and vegetables (see 5.3.28). 
 
S.6 Cardiovascular 
Disease 
 
S.6.1  Observational studies have found that persons who consume large amounts of 
fruit and vegetables have lower rates of coronary heart disease (see 6.1). In 1994, 
COMA’s Cardiovascular Review Group reported on the evidence for the roles of 
dietary factors in the development of cardiovascular disease.  At that time they 
reported fewer data on diet and cardiovascular disease are available for women, 
children and the elderly than for middle-aged men. More evidence has become 
available since that report, but there is still limited evidence for women, children and 
the elderly.  However, the Review Group concluded that the recommendations made 
in the report, which included an approximate 50% increase in the levels of vegetables 
and fruit (and also potatoes and bread) compared with current intakes, should apply to 
the population as a whole.  Many processes linked with the development of 
cardiovascular disease have their origins in childhood and COMA recommended that 
by the age of 5 years children should be consuming a diet consistent with the 
recommendations for adults.  However, they state that the dietary and other 
recommendations for adults do not apply to children below the age of 2 years, for 
whom adequate energy intake for growth remains paramount.  Between the ages of 2 
and 5 years a flexible approach to the timing and extent of dietary change should be 
taken.  
 
S.6.2  For this report, scientific papers reporting on observational and 
experimental/intervention studies which considered fruit and vegetable consumption 
and risk of stroke and/or coronary heart disease risk were identified from a review of 
the literature published since 1995.  This review was not systematic but used Medline 
and covered a large number of journals and included all studies identified whether 
positive or negative associations were described. Table 2 lists the studies identified.   
The results are consistent with a strong protective effect of fruit and vegetables for 
stroke and a weaker protective effect on coronary heart disease. 
 
S.6.3  Fruit and vegetable intake is generally associated with behavioural risk 
factors. In one study intake was found to be lowest among those who also reported 
that they were sedentary, heavy smokers, heavy drinkers, or had never had their blood 
cholesterol checked. However, adjustment for behavioural and other risk factors do 
not explain the benefits of higher fruit and vegetable consumption, although 
adjustments for these factors often attenuate the observed associations. There appears 
 
6

---

to be a dose-response relationship which provides support for the recommendation to 
consume at least five portions of fruit and vegetables a day (see 6.19). 
 
S.7 
Other chronic diseases  
 
S.7.1  Cataracts - The growing data base of evidence for the protective role of fruit 
and vegetable consumption and their components, e.g. carotenoids, for cataracts has 
lead one group of researchers to conclude that diets rich in fruit and vegetables may 
provide the least costly and most practicable means to delay cataract (see 7.1 and 7.2). 
  
S.7.2  Diabetes Mellitus - The fibre in fruit and vegetables may help in the 
management of diabetes.  It is known that selected soluble fibres can delay blood 
glucose absorption from the small intestine.  However, the long-term effect of dietary 
fibre on blood glucose control in individuals with diabetes is probably not significant.  
The data are more convincing for a protective effect of soluble fibre on the serum 
total and LDL-cholesterol levels in individuals with diabetes, particularly non-insulin 
dependent diabetes (NIDDM).  The low energy density of most fruit and vegetables 
can play a role in the prevention of obesity, which is a risk factor for the development 
of NIDDM (see 7.4). 
 
S.7.3  Chronic Obstructive Pulmonary Disease (COPD) - Beneficial effects on lung 
function were seen in British school aged children who consumed fruit more than 
once a day compared to those who did not. A high intake of fruit and vegetables was 
positively associated with pulmonary function in three European cohort studies of 
men. In addition, low intake of dietary vitamin C has been associated with increased 
bronchitis and wheezing and with pulmonary problems (see 7.5). 
 
S.7.4  Fruit intake was associated with reduced risk of developing COPD, including 
asthma, after controlling for age, smoking habits, body mass index and energy intake 
in men in the Netherlands. In another European study a high vegetable intake was 
associated with a decreased risk of developing bronchitis and bronchial asthma (see 
7.6). 
 
 
S.7.5  Obesity - Obesity is one of the most important avoidable risk factors for 
diseases such as CHD, stroke, hypertension, hypercholesterolaemia and NIDDM. 
Fruits and vegetables can play an important role in the prevention of and dietary 
treatment of obesity because of their low fat, low energy contributions to healthy 
eating. Fruits and vegetables represent a food group which can be consumed in 
relative abundance even during a weight-loss phase of weight management (see 7.7). 
 
S.7.6  Bone health - Intakes of nutrients found in abundance in fruit and vegetables 
have been found in a recent study to be positively associated with bone health. After 
controlling for present energy intake, higher intakes of calcium, magnesium, 
phosphorus and potassium were associated with higher bone mass.  Femoral neck 
bone mass density was higher in women consuming a high intake of fruit in their 
childhood (see 7.8). This finding needs to be verified.  
 
S.8 Vegetarian 
Diets 
 
 
7

---

S.8.1  It is likely that vegetarians eat more portions or greater quantities of plant 
foods, including fruits and vegetables than non-vegetarians do. A pooled analysis of 
the data from five large prospective studies found that vegetarians had a 24% lower 
mortality from ischaemic heart disease than non-vegetarians did.  This reduction in 
mortality was greater at younger ages, with a 45% reduction in risk of death from 
ischaemic heart disease before the age of 65 years (see 8.3).  Although observational 
studies of vegetarians cannot provide conclusive evidence, they do support the 
hypothesis that a diet higher in fruit and vegetables might lower the risk of coronary 
heart disease and some cancers (see Chapter 8). 
 
S.9  
Mediterranean-type diet 
 
S.9.1  The Mediterranean diet, high in fresh fruit and vegetables and fish, is high in 
antioxidants and polyunsaturated fatty acids and has been shown to produce 
favourable effects on blood lipids. It is hypothesised and there is some evidence that 
such a diet will lead to less oxidative damage, one of the mechanisms leading to 
chronic diseases such as atherosclerosis and cancer.  A more recent hypothesis is that 
the higher intake of folate found in fruit and vegetables is largely responsible for the 
protective effect for CHD risk of fruit and vegetables in the Mediterranean diet (see 
Chapter 9). 
 
S.10 Genetic 
factors 
 
S.10.1  Important diet-gene interactions may exist, as illustrated by differential 
responses to variation in folate status in those with methylenetetrahydrofolate 
reductase (MTHFR) polymorphisms (see 10.1).  In addition genetic changes in the 
development of cancers and the environmental mutagenic factors which produce these 
changes may determine which susceptible individuals develop cancers. The complex 
interrelations between food components make it difficult to define the precise role of 
specific food factors and diet-gene interactions (see Chapter 10).   
 
S.11 
Burden of diseases that have an inverse association with the consumption 
of fruits and vegetables. 
 
S.11.1  Cardiovascular disease (stroke and coronary heart disease) and cancer are the 
major causes of morbidity and mortality in the UK. A reduction in the incidence of 
these diseases and in the incidence of hypertension, obesity, diabetes mellitus, 
cataracts, and COPD should reduce the number of premature deaths and the cost of ill 
health (see 11.1). 
 
S.11.2  In July 1999 the government set targets: “to reduce the death rate from cancer 
in people under 75 years by at least a fifth (20%) by 2010 [compared with 1997] – 
saving up to 100,000 lives.”  and  “to reduce the death rate from coronary heart 
disease and stroke in people under 75 years by at least two fifths (40%) by 2010”, set 
out in the White Paper Saving lives: Our Healthier Nation (see 11.2 and 11.4).  
 
S.11.3  It has been estimated that diet might contribute to the development of one 
third of all cancers.  The cancers which cause most deaths are lung, breast, colorectal, 
stomach and prostate.  Since the risk of developing all of these cancers has been 
shown, in epidemiological studies, to have some inverse association with the 
 
8

---

consumption of fruit and vegetables, an increase in the consumption of fruit and 
vegetables is likely to reduce the incidence of these cancers (see 11.2). COMA’s 
Working Group on Diet and Cancer concluded that even a small reduction in relative 
risk of developing colorectal, gastric and breast cancers would have important public 
health benefits in terms of the absolute numbers affected. The World Cancer Research 
Fund (WCRF) estimated that increasing fruit and vegetable consumption could 
prevent 20% or more of all cases of cancer (see 11.3). 
 
S.11.4  It has been estimated that consuming one extra portion of fruit and vegetables 
per day resulted in a 6% lower risk of ischaemic stroke so that the consumption of at 
least five portions per day should have a substantial impact on the reduction of 
ischaemic stroke. Similar risk reductions have also been seen for the risk of coronary 
heart disease and the increased consumption of fruit and vegetables (see 11.4). 
 
S.11.5  Obesity and overweight generate significant health service costs in England.   
The Office of Health Economics estimated in 1994 that obesity directly costs the NHS 
over £29M per year and indirectly costs the NHS over £165M if a portion of the cost 
of treating some of the conditions for which obesity is a risk factor is also included.  
Any reduction in the incidence of obesity would reduce the burden of ill health 
resulting from obesity, including coronary heart disease and diabetes (see 11.5). 
 
S.12 
Current and Further Research 
 
S.12.1  An important area for future study is to identify the particular dietary 
components which predispose or protect individuals against cancer, coronary heart 
disease and stroke. This will necessitate: 1) overcoming the limitations in the 
precision and validity of traditional dietary intake measurements, 2) the use of 
biomarkers, and 3) the use of populations with differing dietary habits (see 12.1).   
 
S.12.2  This is in part being addressed by the European Prospective Investigation into 
Cancer and Nutrition (EPIC) which is a multi-centre prospective cohort study 
designed to investigate the relation between diet, nutritional and metabolic 
characteristics, various lifestyle factors and the risk of cancer. The study is also being 
used to investigate the relation between diet and the risk of other diseases e.g. 
coronary heart disease and osteoporosis.  This is an ongoing study with UK/English 
cohorts and will produce results relevant to the UK as long as funds continue to be 
made available to continue collecting and analysing the samples and data (see 12.3). 
 
S.12.3  Interactions among lifestyle behaviours suggest the need for further research 
on the effects of overall dietary patterns on chronic disease risk, as well as on finding 
ways to encourage people to follow dietary recommendations as a whole rather than 
focusing on just one nutrient or food group (see 12.4).  
 
S.13 
Recommendations regarding intake of fruit and vegetables 
 
S.13.1  The Government’s advisory committee on food and nutrition, the Committee 
on Medical Aspects of Food and Nutrition Policy (COMA) has made several 
recommendations in the past decade on the consumption of fruit and vegetables.  The 
report of COMA’s Cardiovascular Review Group in 1994 recommended an 
approximate 50% increase in the levels of vegetables and fruit.  This was based on the 
 
9

---

average household consumption in 1992 (National Food Survey), which was 
estimated as an average of 3½ portions of fruit and vegetables per person per day. 
This recommendation equates to at least five portions of vegetables and fruit per 
person per day on average i.e. an average increase of one and a half portions per day. 
In addition they recommended approximate 50% increase in consumption of potatoes 
(see 13.1). 
 
S.13.2  The Report of COMA’s Working Group on Diet and Cancer in 1998 
recommended that fruit and vegetable consumption in the UK should be increased.  
No target was set by this Working Group, since it found insufficient evidence to 
quantify the optimum level of fruit and vegetable consumption associated with the 
lowest risk of cancer.  It cited the advice of the Working Group on Nutritional 
Aspects of Cardiovascular Disease to increase the consumption by 50%, to at least 5 
portions per person per day on average, as a potentially achievable goal and likely to 
be conducive to better health in general and a lower risk of cancer in particular (see 
13.2.1).  
 
S.13.3  The official guidance on diet and health is set out in the booklet, Eight 
Guidelines for a Healthy Diet, published by the Health Education Authority in 
association with MAFF and the Department of Health (DH). Guideline 5 states “Eat 
plenty of fruit and vegetables” and the accompanying information states “Currently 
many people would benefit from increasing the amount of fruit and vegetables that 
they eat.  A balanced diet contains at least five portions of fruit and vegetables a day.” 
(see 13.4).  The Chief Medical Officer’s Ten Tips – your guide to better health 
includes “Follow a balanced diet with plenty of fruit and vegetables.” (see 13.5). 
 
S.13.4  Internationally an intake of at least five portions (400g) of fruit and vegetables 
per day has become an established ‘healthy eating’ message.  In the USA, the ‘5 A 
Day—for Better Health’ Campaign including its simple, positive, behaviour-specific 
message to eat 5 servings of fruits and vegetables every day as part of a low-fat, high 
fibre diet, was adopted as a national initiative by the National Cancer Institute (NCI) 
and the Produce for Better Health Foundation in 1991 (see 13.6).  
 
S.13.5  In the UK health promotion agencies, producers and retailers have promoted 
‘at least 5-a-day’ message as a mechanism for helping to reduce the risk of coronary 
heart disease, strokes and many cancers. The same message has been promoted in a 
campaign by the Europe Against Cancer programme (see 13.7).   
 
S.13.8  The scientific basis for recommending an increase in fruit and vegetable 
consumption is now widely accepted, and there is a broad agreement on the target, i.e. 
‘at least five portions a day’ (or roughly five 80g portions).  However, it is necessary 
to clarify exactly which fruits and vegetables are included in the advice, and how 
much a portion is.  It is also helpful to know what the current level of consumption is 
and how this varies with age, sex, socio-economic group and region (see 13.8). 
 
S.14 
Current intake of fruit and vegetables 
 
S.14.1  The National Food Survey (NFS), an annual survey of household purchases 
measured over one week, shows that over the decade, 1988 to 1998, there has been a 
 
10

---

decline in the consumption of vegetables (excluding potatoes) by 8% and the 
consumption of fresh fruit has increased by 20% (see 14.1).   
 
S.14.2  Total average vegetable consumption per person was about 156g/day in 1998 
(see 14.2) and total average consumption of fruit including fresh fruit, fruit products 
including fruit juice, was 155g/day (see 14.3).  The NFS overestimates the weight of 
vegetables and fruit consumed in the home because it does not take into account non-
edible parts and thus the edible weight for the average daily consumption of 
vegetables and fruit combined is about 250g in 1998 (see 14.4).  Since a portion is 
roughly equivalent to 80g, this equates to about 3 portions a day.  However, this does 
not cover food eaten outside the home including school meals, and it also includes 
very young children and the elderly who do not eat a lot of fruit (see 14.4). 
 
S.14.3  The NFS data shows considerable variation in consumption with income and 
region. In 1998, the average daily consumption of fresh fruit was 156g/day in income 
group A (highest income) and 58g/day in income group D (lowest income).  The 
consumption of fruit and vegetables is considerably lower in Scotland than England 
and Wales and in the North and North West regions of England compared to the other 
regions (see 14.5). Because it is a survey of households (not individuals) the NFS 
cannot determine trends by sex or age group. 
 
S.14.4  The Dietary and Nutritional Survey of Adults, carried out in 1986/87, is the 
only national survey of British adults (sample n=2197) for which data on fruit and 
vegetable consumption by individual adults is available.  This survey preceded the 
National Diet and Nutrition Survey (NDNS) programme (a Government programme 
of national dietary surveys of individuals from four different population age groups). 
The survey of adults was used to obtain estimates of the average frequency of 
consumption by the adult population (aged 16 to 64 years) per week. Men and women 
ate vegetables excluding potatoes eight times during the 7-day recording period.  Men 
ate five servings and women ate seven servings of fruit (including fruit juice) during 
the 7-day recording period. We know from the NFS that there has been increased 
purchases of fruit (20%) and decreased purchases of vegetables (8%) since the adult 
survey but it is unlikely that the number of portions eaten weekly will have increased 
greatly in that time. [Further analyses are being carried out on this survey data to 
determine the variation in average daily consumption of fruit and vegetables by age, 
sex, socio-economic group and region and will be available soon.] The NDNS 
programme has a new survey of adults aged 18 to 64 years beginning later in 2000 
which will take place over 12 months (see 14.6 and 14.7). 
 
S.14.5  Estimates of the number of portions of fruit and vegetables consumed daily or 
over the 4- or 7-day recording periods are being made using data from the other three 
age groups from the NDNS programme, children aged 1½ - 4½ years, young people 
aged 4 to 18 years, and adults aged 65 years or over (see 14.8).  
 
S.14.6 Information on the consumption of fruit, fruit juice and carrots by children 
participating in the NDNS of young people aged 4 to 18 years and an assessment of 
the increased tonnage for apples to enable each child to eat an apple a day is included 
at Annex 1 of the main report. 
 
 
11

---

S.14.6 The report Health in England 1998: Investigating the links between social 
inequalities and health explored the eating habits and knowledge of the population 
about what constitutes a healthy diet and found that lower proportions of men than 
women usually ate fruit or vegetables daily (59% compared with 74%).  Men were 
less likely than women to mention the need to ‘eat lots of fruit, vegetables or salad’ 
(62% compared with 74%).  Age was the characteristic most closely associated with 
diet quality, those in the younger and older age-groups were more likely than those in 
the middle age-groups to have a ‘less healthy diet’. Diet quality decreased with 
decreasing levels of household income, qualifications, and social class.  In addition, it 
was relatively lower among the widowed, divorced and separated, the unemployed, 
and those in local authority housing.  It was also relatively lower among men in 
privately rented accommodation, among lone mothers and, to a lesser extent, among 
women who were economically inactive and women living alone (see 14.9). 
 
S.15 
Which ‘fruit and vegetables’ and how much is a ‘portion’? 
 
S.15.1  The term “fruit(s) and vegetables” includes all fresh, frozen, canned and dried 
fruits and vegetables except potatoes (for explanation see paragraph 2.2). Beans and 
pulses are included but not nuts. One glass of fruit juice per day, but not fruit 
squashes, may also be included as one portion (see 15.1). 
 
S.15.2  The advice to eat at least five portions of fruit and vegetables per day is based 
on the broad range of health and nutritional properties of fruits and vegetables, and it 
is important to emphasise the use of a wide variety of both fruits and vegetables. It 
should also take into account the availability, social and cultural uses of fruit and 
vegetables, and should encourage everyone, but especially young people, to try ‘new’ 
vegetables and fruits (see 15.2). This target is seen as moderate and achievable (see 
15.3). 
 
S.15.3  A portion is a predefined ‘discrete’ amount of food, e.g. an apple.  A serving is 
the amount an individual serves himself or herself, or is served, and is variable.  
Information from GB dietary surveys suggests that a mean portion size of around 80g 
ties in well with average serving sizes used by households in Britain.  Some practical 
examples of adult portions for consumers are as follows: 2 tablespoons vegetables 
(raw, cooked, frozen or canned); a dessert bowlful salad; an apple, banana, orange or 
other citrus fruit and a cupful of grapes or berries. Further information can be found in 
a table in paragraph 15.4. 
 
S.15.4  The use of food frequency questionnaires (FFQ) to measure the consumption 
of fruits and vegetables has been shown to vary the estimated intake depending on the 
number of questions asked about fruit and vegetable and reported consumption.  More 
questions led to exaggerate intakes, therefore caution must be exercised when using 
this method to estimate fruit and vegetable consumption (see 15.5). 
 
S.16 Dietary 
Interventions 
 
S.16.1  National average intakes equate roughly to about 3 portions of fruit and 
vegetables each day compared to the recommended amount of at least 5 portions each 
day. There is a need for the development of dietary interventions which are effective 
in increasing intakes among the population as a whole, and also among specific 
 
12

---

groups within the population for whom there are particular concerns, for example 
school aged children, young men and women (18 to 30 years), people on low 
incomes, and people in some regions of England. (See 16.1)   
 
S.16.2  Table 3 (see paragraph 16.2.1) lists recent publications which describe dietary 
interventions or studies to identify opportunities and barriers to increasing the 
consumption of fruit and vegetables along with the main findings summarised here 
and in 16.3): 
 
Barriers: 
•  Over estimation of current consumption of fruit and vegetables by children and 
adults; 
•  Lack of experience in cooking vegetables led to unwillingness to try more and 
new vegetables especially in families who just ‘coped’ with feeding rather than 
planned meals; 
•  Time constraints and inconvenience especially with respect to preparing and 
cooking vegetables;  
•  Shopping is more difficult and, for some cost, was seen as a barrier 
•  Lack of availability at work and when eating out 
•  Lack of social pressure to increase intake (women feel more social pressure than 
men) 
 
Opportunities: 
•  More advanced stages of change were more likely to be associated with 
knowledge of the 5-a-day recommendation, self-efficacy, and greater fruit and 
vegetable consumption. 
 
Some intervention methods found to be successful in increasing intake: 
•  Practical advice; live demonstrations; hands-on practice; tasting; build on familiar 
ingredients; ‘convenience’ preparation i.e. quick recipes or use of frozen and 
canned vegetables and fruits; and targeting specific groups.   
•  Emphasis on portions, e.g. what is ‘a portion’; practical means of increasing the 
number of portions per day tailored to intervention group; and self-monitoring 
using diaries. (Fruits were seen as more easily incorporated into daily eating 
patterns than vegetables.) 
•  Video presentations and telephone counselling were almost as successful in 
increasing intake as personal contact. 
 
Implications for interventions: 
•  Need education to increase knowledge of the 5-a-day fruit and vegetable 
consumption message and the evidence that increased fruit and vegetable 
consumption leads to reduced risk of several diseases, especially coronary heart 
disease and cancer; 
•  Consumers need to become more aware of their actual consumption rather than 
their perceived consumption; 
•  Portions sizes need defining more clearly; 
•  Vegetables need greater targeting than fruits; 
•  Collaboration between retailers, caterers, employers to ensure better accessibility. 
 
 
13

---

S.16.3  MAFF and DH funded UK studies are establishing whether fruit tuck shops in 
primary schools increase fruit consumption and DH funded projects are establishing 
whether counselling in primary health care is effective in increasing consumption of 
fruit and vegetables.  These projects will be completed by the end of 2001 (See Annex 
3). 
 
S.16.4  The Health Education Authority (HEA) commissioned a series of reviews of 
effectiveness of healthy eating promotions, which were published in 1997 and 1998. 
A summary of the conclusions is attached in Appendix 1 (See 16.2.5).  In addition the 
HEA, now the Health Development Agency (HDA) have produced a paper (see 
Annex 4) for the Department of Health that provides a brief overview of the current 
(2000) information held by the HDA, concerning the effectiveness of interventions to 
increase fruit and vegetable intakes (see 16.3).  The final section of the HEA paper 
makes recommendations of steps, which could be taken to increase fruit and vegetable 
intakes, and these are summarised below and produced in full in paragraph 16.3: 
 
To address circumstantial barriers: 
 
•  It is recommended that the Government and FSA make it an early priority to press 
the European Union for a health equity assessment of the Common Agricultural 
Policy (CAP) and that means of ensuring access to affordable supplies of fruit and 
vegetables, be made a priority in the current reforms of the CAP.          
 
•  It is suggested that fruit and vegetables should be made available to those eligible 
for the Welfare Food Scheme using a  token system in the same way that milk and 
infant formula are currently made available. This would help to address issues of 
affordability and improved access by creating a ‘supply and demand’ situation. 
  
•  It is recommended that schools be given information about accessing the free fruit  
available under the CAP intervention scheme and that ways are identified of  
developing a system of  distributing the fruit between schools so that it is received 
in manageable amounts.  
 
•  Initiatives such as Sure Start, Education Action Zones, Employment Action 
Zones, the New Deal and Neighbourhood Renewal schemes should be considered 
when investigating ways in which Government initiatives other than those 
primarily concerned with health, can support initiatives to increase consumption. 
For example the development of community led local retail strategies might form 
part of a neighbourhood renewal scheme with a particular emphasis on increasing 
access to fruit and vegetables.  
 
•  In order to improve shopping access to fruit and vegetables for people living in 
deprived neighbourhoods consideration should be given to improving support for 
smaller retailers and to how the larger retailers may be able to contribute to 
improving access to fruit and vegetables in these areas.  However it should be 
recognised that in some cases it may be more appropriate to take ‘people to the 
shops’, rather than ‘shops to the people’ and as such the provision of affordable 
transport may be a more important factor in some communities.   
 
 
14

---

•  Methods of increasing access to affordable fruit and vegetables with local 
communities could include the initiation of small grant schemes to assist in the set 
up costs of establishing local food co-operatives. Healthy Living Centres might be 
involved in the development of such food co-operative schemes.  In rural areas 
Local Authorities could develop Village Shop assistance schemes, which help to 
ensure the continuing existence of such shops and that they provide fruit and 
vegetables.  
 
•  It is recommended that increasing fruit and vegetable consumption should be 
reflected as a priority in the Department of Health’s National Priorities Guidance,  
the CHD National Service Framework, Our Healthier Nation key area’s contracts 
and the policy direction for Healthy Living Centres.      
 
 
To address attitudinal barriers:  
 
•  It is recommended that The Food Standards Agency (FSA) should develop, in 
partnership with the Department of Health, the HDA and industry, information 
campaigns for the public and professionals which give clear messages concerning:  
i)  the benefits of eating more fruit and vegetables;  
ii)  the recommended intake and the fact that most people are not eating enough 
and need to virtually double their intake;  
iii) the fact that frozen and canned fruit and vegetables count;  
iv)  a clear guide as to what constitutes a portion of fruit and vegetables;  
v)  and opportunities for incorporating these into the diet.  
This should be done in the context of an overall balanced diet using the Balance of 
Good Health.      
  
•  To address ‘lack of cooking skills’ and other barriers for adults, it has been 
suggested that a programme similar to the Expanded Food and Nutrition 
Education (EFNEP) programme in the USA could be developed in the UK. Such a 
programme would train and use people from the local community to work with 
clients to develop their cooking skills together with shopping and menu planning 
skills.  
 
•  Home economists should be used to ‘train the trainers’ because they possess both 
practical skills in these areas together with nutrition knowledge.  
 
•  To help address lack of cooking skills in future generations, it is strongly 
recommended that cooking and budgeting skills are reintroduced into the national 
curriculum. 
 
S.16.6  USA 5 A Day campaign  
 
S.16.6.1 
In 1988, the Californian Department of Health Services embarked 
upon an innovative social marketing program to increase fruit and vegetable 
consumption. The ‘5 A Day—for Better Health’ Campaign had several distinctive 
features, including its simple, positive, behaviour-specific message to eat 5 servings 
of fruits and vegetables every day as part of a low-fat, high fibre diet. It used the mass 
media; formed  partnerships between the state health department and the produce and 
 
15

---

supermarket industries; and made extensive use of point-of-purchase messages (see 
16.5.1).  
 
S.16.6.2 
In 1991 the campaign was adopted as a national initiative by the 
National Cancer Institute (NCI) and the Produce for Better Health Foundation. The ‘5 
a Day’ Campaign appears to have raised public awareness that fruits and vegetables 
help reduce cancer risk, increased fruit and vegetable consumption in major 
population segments, and created an ongoing partnership between public health and 
agribusiness that has allowed extension of the campaign to other populations 
segments, namely children and ethnic groups (see 16.5.1).  
 
S.16.6.3 
Various methods of intervention with school children, low income 
communities and work site populations have been described along with outcomes in 
the scientific press.  Paragraphs 16.5.2 to 16.5.5 in the main report summarise some of 
these interventions and outcomes, and there is a complete publication list in Annex 5. 
Individual interventions have been evaluated but the whole programme has not, so it 
is not known which methods both for intervention and evaluation are the most 
successful and reliable (see 16.5.6).  
 
S.17 Conclusion 
 
S.17.1  The benefits of eating fruits and vegetables appear to be beyond dispute (see 
Chapters 5, 6, 7, 8 and 9), especially since these foods provide essential nutrients not 
readily available from other sources.  But eating more fruits and vegetables should be 
linked to other healthy dietary and lifestyle patterns (see 17.1).  
 
S.17.2  The current mean consumption of fruit and vegetables per person per day in 
the UK is approximately three portions (Chapter 14), although this varies by age, sex, 
socio-economic group, educational level, and region.   
 
S.17.3  The message to ‘Eat 5- portions-a-day of fruits and vegetables’ is recognised 
by many (but not all) and appears to be one that consumers can implement with the 
removal of circumstantial and attitudinal barriers and with tailored advice and 
practical tips and experience.  Opportunities have been identified and these should be 
incorporated into educational programs and interventions (see 17.5).  
 
S.17.4  It is important to tailor the message to the population sub-groups most in need 
of changing.  It is advisable to spend more time understanding the food consumption 
habits of the population sub-group at which the message is aimed so that appropriate 
messages can be developed to foster behaviour change (see 17.6). 
 
S.17.5  There is much scope for the development of an integrated strategy to increase 
fruit and vegetable intakes. Any strategy should be developed in consultation with the 
Health Development Agency (HDA) and the Food Standards Agency (FSA). This 
strategy needs both a continuing commitment to carrying out further research which 
will develop the evidence base as to what is effective and the most reliable tools to 
evaluate the outcome. It also needs action, in particular to tackle the circumstantial 
barriers such as increasing access to affordable supplies of fruit and vegetables, 
especially for those on low income (see 17.7). 
 
 
16

---

THE CASE FOR FRUIT AND VEGETABLES  
 
1 Introduction 
 
1.1 
For over a century fruit and vegetables have been known as “protective foods” 
because people who ate more of them were protected from many of the common 
deficiency diseases.  Now we know that, not only are they nutrient-dense, but they 
contain many biologically active substances and fibre which not only help to reduce 
the likelihood of developing chronic deficiency diseases, such as scurvy, but they may 
also help to reduce the likelihood of developing chronic diseases, such as coronary 
heart disease, stroke and some cancers.  In addition they tend to have a low energy 
density which can help reduce the likelihood of developing obesity and non-insulin 
dependent diabetes mellitus (NIDDM).  Convincing evidence is also emerging about 
the positive role of fruit and vegetable consumption and a reduced incidence of 
cataracts, chronic obstructive pulmonary disease, and hypertension. 
 
1.2 
This paper reviews the evidence for the effectiveness of fruit and vegetables in 
promoting health and preventing disease, particularly cancer, coronary heart disease, 
stroke, and the case for increasing the population consumption of fruit and vegetables 
to at least five portions (400g or more) per day. It also looks at the effectiveness of 
campaigns and interventions to change dietary behaviour, taking into account not only 
community interventions, but also the effect of national policies, and national 
campaigns (e.g. the American 5 A Day campaign).   
 
1.3 
The aim is to provide the scientific rationale for supporting enhanced 
intervention in the promotion of increased fruit and vegetable consumption as part of 
chronic disease risk reduction and healthy eating. 
 
2 
Definition of fruit and vegetables 
 
2.1 
The term “fruit(s) and vegetables” includes all fresh, frozen, canned and dried 
fruits and vegetables except potatoes. Beans and pulses are included.  
 
2.2 
The exclusion of potatoes results from the fact that health advice consistently 
makes recommendations for starchy foods such as potatoes. COMA recommended a 
50% increase in “bread, potatoes, and fruit and vegetables” [1].  In the UK potatoes 
are not generally used in the same way as other vegetables.  Most people would not 
consider having pasta with sauce and potatoes.  It is worth noting that, although 
potatoes are included in the 5 A Day programme in the United States, crisps and chips 
(French fries) are excluded.  A key difference is the high consumption of sweet 
potatoes in the US and these are included under root vegetables.  
 
2.3 
Several terms are used to describe types of vegetables which include 
brassicas/cruciferous vegetables (cabbage, cauliflower, broccoli and Brussels sprouts), 
yellow/orange vegetables (carrots, peppers, tomatoes, swedes), salad vegetables 
(mainly green leafy salad vegetables), root vegetables (swedes, turnips, carrots); 
allium vegetables (onions, shallots, scallions, leeks and garlic).  Some studies also 
distinguish between citrus and non-citrus fruits. 
 
 
 
17

---

3 
Components and properties of fruit and vegetables with possible 
protective effects 
 
 
3.1 
Antioxidants e.g. vitamin C, vitamin E, carotenoids and flavonoids 
Folate 
 Potassium 
and 
magnesium 
Phytoestrogens (lignans, isoflavones) 
 Plant 
sterols 
 
Sulphur containing compounds  
Carbohydrates (starch and fructose) and dietary fibre (NSP)  
Other phytochemicals e.g. glucosinolates, indoles, thiocyanates and 
isothiocyanates, protease inhibitors and limonene. 
Low energy density 
 
3.2 
These phytochemicals can have complementary and overlapping mechanisms 
of action, including modulation of detoxification enzymes, stimulation of the immune 
system, reduction of blood pressure, and antioxidant, antibacterial, and antiviral 
effects.  Although these effects have been examined primarily in animal and cell-
culture models, experimental dietary studies in humans have also shown the capacity 
of vegetables and fruit and their constituents to modulate some of these potential 
disease-preventive mechanisms.  The human studies have relied on intermediate 
endpoints related to disease risk.  Rigorously conducted experimental dietary studies 
in humans are an important link between population- and laboratory-based research 
[2]. 
 
4 
Biological and Experimental Evidence 
 
4.1 
Antioxidants e.g. vitamin C, vitamin E, carotenoids and flavonoids 
There is substantial metabolic and experimental evidence to implicate antioxidant 
micronutrients in the risk of developing chronic diseases such as cancers and coronary 
heart disease.   The dietary antioxidants  include  vitamins C and E, the carotenoids, 
such as beta-carotene, lycopene and lutein, and the flavonoids, such as quercetin, 
luteolin and kaempferol.   
 
4.1.1  One aspect involved in initiation and development of both cardiovascular 
diseases and cancers (stomach, colon. breast, prostate, ovary, and endometrium) are 
abnormal oxidative processes leading to the generation of hydroxy radicals and 
peroxy compounds.  Normal oxidative metabolism produces huge amounts of 
potentially dangerous oxidants (free radicals), controlled by a variety of antioxidant 
systems.  An imbalance between the generated and exogenously inflicted oxidants and 
the oxidant system is termed oxidative stress.  Even without oxidative stress, i.e. 
under normal physiological conditions, the damage to DNA is extensive.  More than 
one hundred different oxidative modifications in DNA have been described.  The 
hydroxylation of guanine in the 8-position is the most frequent and most mutagenic 
lesion described.  Numerous publications from epidemiology and intervention studies 
with antioxidants point at oxidative modifications as an important factor in cancer 
development at certain sites [3]. 
 
 
 
18

---

4.1.2.  The ‘antioxidant hypothesis’ proposes that vitamin C, vitamin E, carotenoids, 
and other antioxidants occurring in fruit and vegetables afford protection against heart 
disease and cancer by decreasing oxidative damage (free radical mediated damage in 
lipids, proteins and to DNA, respectively, but the fine details of the causal mechanism 
have still to be elucidated. Whether dietary factors can modulate DNA repair – a 
crucial element in the avoidance of carcinogenesis – is an intriguing question that has 
not yet been satisfactorily answered.  A study at the Rowett Research Institute, 
Aberdeen, investigated the effects of beta-carotene on oxidative DNA damage and its 
repair [4].   The researchers found that, in cases where the baseline plasma beta-
carotene concentration was high, or where supplementation increased the plasma 
concentration, recovery from oxidative damage was relatively rapid.  However, the 
researchers concluded that the methodology used could have influenced the findings 
and further work is required. 
 
4.1.3  To test elements of the same hypothesis, Collins et al [5] measured blood 
levels of dietary antioxidants, and 8-oxodeoxyguanosine (8-oxo-dG) concentrations in 
lymphocyte DNA, in healthy men and women (aged 25-45 years, non-smokers) from 
five European countries: France, Ireland, The Netherlands, Spain, and the UK.  The 
subjects were measured before and after a 12-week carotenoid supplementation 
regime.  They found that 8-oxo-dG levels in lymphocyte DNA vary significantly 
according to sex and country: it was significantly higher (up to about three fold) in 
men in Ireland and the UK.  They also found that oxidative DNA damage is not 
significantly affected by carotenoid supplementation; nor is there any association 
between mean baseline levels of antioxidants, which are generally similar in the 5 
countries.  They found that there is a strong association between premature coronary 
heart disease mortality in men and the mean levels of 8-oxo-dG for the five countries 
(r = 0.95, p<0.01).  Women have low coronary heart disease mortality rates which do 
not correlate with 8-oxo-dG.  In terms of cancer deaths, only colorectal cancer in men 
shows a significant positive correlation (r+0.91, p<0.05), and stomach cancer in 
women is negatively correlated with DNA oxidation (r=-0.92, p+0.01).  
 
4.1.4  The relationship between intakes of vegetables and fruits with levels of 
oxidative DNA damage in women consuming their own usual diet or a diet low in fat 
has been investigated by Djuric et al [6].  Preliminary results are suggestive of a 
negative association of DNA damage with cooked vegetable intake. 
 
4.1.5  Epidemiological studies as well as laboratory experimentation have yielded 
sound data and evidence in support of the fact that vegetables and fruits and 
antioxidants therein account mechanistically for inhibition in oxidative reactions [7]. 
Some of these nutrients can act synergistically, and high concentrations are often 
found in tissues, such as the leucocytes and mucosal cells, that are particularly prone 
to oxidative attack.  However, the association of individual vitamin intake, as opposed 
to fruit and vegetable consumption, and the risk of the development of various cancers 
is inconsistent. COMA concluded, “Overall there is not enough evidence to conclude 
that vitamins A, C, E or β-carotene protect against the development of various 
cancers.  Higher intakes of the antioxidant vitamins, β-carotene, vitamin C and 
vitamin E have been variously associated with lower risks of breast cancer, colorectal 
cancer, lung cancer, gastric cancer, and cervical cancer in case-control and 
prospective studies.  Most of the intervention trials that have been carried out so far 
with supplements of these vitamins have failed to confirm a hypothesised protective 
 
19

---

effect of these vitamins on cancer.  The lack of effect in the intervention trials raises 
questions about the capability of such (relatively) short-term trials to demonstrate a 
protective effect.  If these vitamins exert a protective effect at an early stage of the 
carcinogenic process, for example by protecting against free-radical induced DNA 
damage, the relatively short-term trials reported so far would be unable to 
demonstrate a protective effect even if one existed.  Alternatively, the observed 
associations may relate to a substance or mixture of substances in the diet for which 
intakes of these nutrients are acting as a marker” [8].  
 
4.1.6  The intervention studies also highlight the lack of information on the long-
term safety of sustained intakes of moderate to high doses of micronutrient 
supplements.  In particular, the finding of an increased incidence of lung cancer in 
those taking B-carotene supplements in two intervention trials in people at high risk 
(ATBC, CARET trials) raises the possibility that a change in the usual balance of 
carotenoids in the diet (for instance by high dose purified supplements) might lead to 
potentially adverse perturbations in their absorption, metabolism or function.  Such 
findings caution against the widespread use of moderate to high dose micronutrient 
supplements, which cannot be assumed to be without adverse effects [8].  Results of 
other clinical trials are still awaited, but as yet, there is no evidence to recommend 
supplements of these particular micronutrients for the prevention of cancer  [9]. 
 
4.1.7  Antioxidants have differing solubilities which partition across the phases of 
tissues, cells and macro molecular structures: water-soluble ascorbate (vitamin C); 
lipid-soluble tocopherols (vitamin E) and carotenoids, and intermediatory-soluble 
flavonoids.  The health protection provided by fruit and vegetables could arise 
through an integrated reductive environment delivered by plant antioxidants of 
differing solubility in each of the tissue, cellular and macromolecular phases [10]. 
 
4.1.8  Under certain circumstances, many antioxidants can have pro-oxidant 
properties.  For example. In the presence of free iron, ascorbic acid can induce 
oxidative damage.  The ultimate effect of a particular compound will therefore depend 
on a number of factors, including the level of intake, and cannot easily be predicted.  
It is not always clear under what conditions an antioxidant will develop pro-oxidant 
properties and thereby induce potentially adverse effects [8]. 
 
4.1.9  Vitamin C (ascorbic acid) acts as a first line of defence against water-soluble 
free radicals, reacting rapidly with superoxide and hydroxyl radicals, but only poorly 
with hydrogen peroxide.  Vitamin C is also thought to be involved in the regeneration 
of other reducing agents, such as alpha-tocopherol (vitamin E) depleted from 
membranes by oxidation.  Vitamin C may also have a role as a chemical defence 
against endogenous genotoxins.  For example, at acid pH ascorbic acid prevents the 
formation of N-nitroso mutagens produced in the presence of nitrite [8]. 
 
4.1.10  Vitamin E (alpha-tocopherol) in membranes is the major radical trap, reacting 
rapidly with peroxyl radicals.  It also quenches singlet oxygen and scavenges 
superoxide radicals [8].  A randomised placebo-controlled trial in smokers has been 
used to investigate the effect of dietary supplementation with antioxidants (vitamin E, 
ascorbic acid or co-enzyme Q10) on a biomarker of oxidative DNA damage (8-oxo-
dG).  Two months of supplementation did not result in significant changes in the 
urinary excretion rate of 8-oxo-dG in any group.  The lack of effect of antioxidant 
 
20

---

supplementation on the excretion rate of 8-oxo-dG, despite substantial increases in 
plasma antioxidant concentrations, agrees with the results from recent large 
intervention studies with cancer as an endpoint.  The cancer-protective effect of fruit 
and vegetables seems to rely not on the effect of single antioxidants but rather on 
other anticarcinogenic compounds or on a concerted action of several micronutrients 
present in these foods [11]. 
 
4.1.11  Flavonoids (quercetin, kaempferol, myricetin, luteolin, and apigenin) are 
effective antioxidants and, in theory, may provide protection against cancer, although 
direct human evidence of this is scarce [12]. Quercetin is a strong antioxidant and a 
major dietary flavonoid. Epidemiological studies suggest that consumption of 
quercetin protects against cardiovascular disease, although Hollman states that a 
protective effect of dietary flavones and flavonols against cardiovascular disease is 
not conclusive [13].  
 
4.1.12  Whilst the data from epidemiological studies do not support a strong inverse 
association between intake of flavonoids and total coronary heart disease, they do not 
exclude the possibility that flavonoids have protective effect in men with established 
coronary heart disease [14].  Total flavonoid intake was associated with a decreased 
risk of CHD death in the Iowa postmenopausal women’s cohort study after adjusting 
for age and energy intake (p for trend = 0.04) [15].  However, there was no 
association between total flavonoid intake and stroke mortality. 
  
4.1.13  The Department of Health is currently funding under the Nutrition Research 
programme a project to examine the use of plasma flavonol measurement as a 
biomarker of dietary flavonoid intake in a clinical trial of a primary care intervention 
to increase fruit and vegetable consumption.  The data from this study is due in June 
2000. 
 
4.1.14 Franke et al [16] investigated the effects of the predominant dietary flavonoids 
and isoflavonoids at inhibiting neoplastic transformation induced by 3-
methylcholanthrene in C3H 10T1/2 murine fibroblasts.  They found that most 
phenolic agents tested were equal to or superior to known chemopreventive agents 
such as carotenoids or vitamins in effectiveness.  Hesperetin, hesperidin and catechin 
were the most potent agents among the flavonoids tested.  The soy isoflavonoids were 
weakly active except when applied in combination suggesting a synergistic effect.   
 
4.1.15  In addition to their antioxidant properties, flavonoids can modify the 
expression of phase I and phase II enzymes, and can therefore function as blocking 
agents, and potentially have an anticarcinogenic effect [8].   
 
4.1.16  The relation between the intake of antioxidant flavonoids and subsequent risk 
of cancer was studied among 9,959 Finnish men and women aged 15-99 years and 
initially cancer free.  The results are in line with the hypothesis that flavonoid intake 
in some circumstances may be involved in the cancer process, resulting in lowered 
risks [17].  Of the major dietary flavonoid sources, the consumption of apples showed 
an inverse association with lung cancer incidence, with RR = 0.42 (95% CI 0.23-0.76) 
after adjustment for the intake of other fruits and vegetables. However, no association 
between the intake of 5 major flavonoids and mortality from total cancer, lung cancer, 
 
21

---

colorectal cancer or stomach cancer was observed on an analysis of the data from the 
Seven Countries Study after 25 years of follow-up [18] 
 
4.1.17  Birt et al [19] studied the prevention of skin tumour promotion by apigenin, a 
plant flavonoid.  Apigenin may block several points in the process of tumour 
promotion, including inhibiting kinases, reducing transcription factors and regulating 
cell cycle. 
 
4.1.18  Carotenoids  (alpha-carotene, beta-carotene, lutein, and lycopene) have been 
shown to be highly effective radical trapping agents and thus potentially to be capable 
of reducing oxidative damage to DNA [8]. However, human intervention studies with 
beta-carotene have given equivocal results in terms of cancer incidence.  In an 
alternative molecular epidemiological approach, Collins et al [20] employed the 
‘comet assay’ (single cell alkaline gel electrophoresis) to measure strand breaks, 
oxidized pyrimidines and altered purines in the DNA of lymphocytes from volunteers 
supplemented with alpha/beta-carotene, lutein, lycopene or placebo.  In addition, they 
measured concentrations of main serum carotenoids, and vitamins E and C.  They 
reported a significant negative correlation between basal concentrations of total serum 
carotenoids and oxidized pyrimidines.  A similar correlation was seen between 
individual carotenoids (notably lutein and beta-carotene) and oxidized pyrimidines.  
However, carotenoid supplementation did not have a significant effect on endogenous 
oxidative damage. The authors state that “this suggests that there are some factors in 
the basal diet, probably found in fruit and vegetables, that decrease oxidative damage 
to DNA.  In this case, basal serum carotenoids may simply be markers of 
consumption of fruit and vegetables, they, themselves, having little or no protective 
value.”  In a study at the UK International Antioxidant Research Centre on the effect 
of tomato consumption on DNA oxidative damage, consumption of a single serving 
of tomatoes by healthy volunteers was sufficient to alter levels of oxidative DNA base 
damage in white cell DNA within 24 hours.  Levels of the mutagenic oxidised purine 
base 8-hydroxyguanine decreased, especially in those subjects whose initial levels of 
this base were higher than the mean.  However, total DNA base damage remained 
unchanged since levels of 8-hydroxyadenine rose.  The research group suggest that 
the ability of tomato consumption to modulate oxidative DNA damage in the short 
term may indicate why daily consumption of fruits and vegetables is beneficial in 
decreasing cancer incidence [21]. Lycopene is a more effective radical trapping agent 
than beta-carotene [8]. 
 
4.1.19  Inverse correlations have been found in most studies on the relationship 
between dietary intake and plasma concentrations of carotenoids on one side and 
degenerative diseases such as cancer and cardiovascular diseases on the other side. 
This has initiated trials with relatively high doses of carotenoid supplements.  In the 
study in Linxian (China) in a rural population with poor nutritional status, 
supplementation with beta-carotene, zinc, selenium and vitamin E lowered total 
mortality and mortality from stomach cancer.  Other intervention trials (Alpha-
tocopherol, Beta-carotene/ATBC, CARET) on well-fed subjects did not show 
beneficial effects on mortality from cancer and cardiovascular diseases.  On the 
contrary, higher mortality and lung cancer incidence was found in supplemented 
subjects that were also exposed to asbestos and cigarette smoke.  In these studies, 
doses of supplemental beta-carotene were high and varied from 20 to 50 mg/day.  One 
still ongoing study, is supplementing subjects for eight years with a cocktail of 
 
22

---

vitamins and minerals including 6mg per day of beta-carotene.  This supplementation 
which is regarded as being  physiologically  more “normal”  might give clarity on the 
question if beta-carotene is the protective factor in fruits and vegetables [22]. 
 
4.2 
Folate 
 
4.2.1  Fruits and vegetables are rich sources of folate.  Folate lowers plasma 
homocysteine levels.  Even mild to moderate elevation of plasma homocysteine level 
is a strong risk factor for arteriosclerosis of the coronary, cerebral, and peripheral 
arteries. Experimental evidence suggests that the atherogenic propensity associated 
with raised homocysteine levels results from endothelial dysfunction and injury 
followed by platelet activation and thrombus formation.  Plausible biological 
mechanisms have been proposed.  Homocysteine has been shown to promote 
oxidative modification of LDL cholesterol, which can promote the formation of 
atherosclerotic lesions.  It can also have a direct effect on vascular endothelium by 
stimulating the proliferation of vascular smooth muscle cells.  Homocysteine can alter 
the normal antithrombotic mechanisms by enhancing the activities of factor XII and 
factor V and depressing the activation of protein C.  In addition, homocysteine 
reduces the production of endothelial-derived nitric oxide, which has vasodilatory 
effects [23]. 
 
4.2.2.  The British Regional Heart Study examined prospectively the relationship of 
serum homocysteine levels to risk of stroke.  The mean level of homocysteine (13.7 
micromol/l) was significantly higher in stroke cases (107 men) when compared with 
the 118 matched controls selected for this study (11.9 micromol/l).  This relationship 
was unchanged after adjusting for known risk factors for stroke [24]. 
 
4.2.3  The Department of Health is currently funding a project under the Nutrition 
Research programme of a RCT on the effect of increasing dietary folate (through 
increasing fruit and vegetable intake) on plasma homocysteine, blood pressure and 
lipids in a high-risk population in primary care.  The results will be available after 
November 2001. 
 
4.3 
Potassium and magnesium 
 
4.3.1  A recent study has identified potassium, magnesium, and fibre as significant 
modulators of stroke risk in men [25].  The protective effects were particularly 
pronounced in hypertensive subjects.  The observed effects may be due to direct and 
indirect effects of these nutrients on blood pressure and regulatory functions, such as 
endothelial function.   The best strategy to achieve a high intake of these nutrients is a 
diet rich in fruits and vegetables. 
 
4.4 
Phytoestrogens  
 
4.4.1  Phytoestrogens are an important class of phenolic compounds.  The principle 
compounds of phytoestrogens are the isoflavones and lignans. Phytoestrogens are 
structurally similar to the mammalian oestrogen, oestradiol and possess weakly 
oestrogenic activity.  Certain foods, such as soy and broccoli,  contain relatively large 
amounts of phytoestrogens.  Oestrogenic compounds can be agonistic or antagonistic 
to oestradiol when they act simultaneously at target tissues.  In animal models and in 
 
23

---

in vitro experimental systems, the isoflavones appear to act as anti-oestrogens, i.e. 
they compete with oestradiol for oestrogen receptors in a way that is generally 
antiproliferative.  Phytoestrogens are postulated to play a preventive role in hormone 
dependent cancers [26].  However, despite plausible mechanisms, COMA concluded 
that there is insufficient evidence to draw conclusions on the effect of phytoestrogens 
on the risk of breast cancer [8].   
 
4.4.2   Lignans are present in many cereals, grains, fruits and vegetables, and give 
rise to the mammalian lignans, enterodiol and enterolactone; however, the richest 
source is linseed and other oil seeds.  In addition to their oestrogenic activity, many of 
these plant compounds can interfere with steroid metabolism and bioavailability, and 
also inhibit enzymes, such as tyrosine kinase and topoisomerase, which are crucial to 
cellular proliferation. 
 
4.5 
Plant sterols 
 
4.5.1  Plant sterols (phytosterols) are substances present in vegetable oils, corn and 
some fruits.  They have been proposed as being protective in colon carcinogenesis 
[27].  Beta-sitosterol (BSS) and its glycoside (BSSG) are sterol molecules synthesised 
by plants.  In animals, BSS and BSSG have been shown to exhibit anti-inflammatory, 
anti-neoplastic, anti-pyretic, and immune-modulating activity.  A proprietary 
BSS:BSSG mixture has demonstrated promising results in a number of studies, 
including in vitro studies, animal models, and human clinical trials.  This phytosterol 
complex seems to target specific T-helper lymphocytes, the Th1 and Th2 cells, 
helping normalise their functioning and resulting in improved T-lymphocyte and 
natural killer cell activity.  A dampening effect on overactive antibody responses has 
also been seen, as well as normalisation of the DHEA:cortisol ratio.  The re-
establishment of these immune parameters may be of help in numerous disease 
processes relating to chronic immune-mediated abnormalities, including chronic viral 
infections, tuberculosis, rheumatoid arthritis, allergies, cancer, and auto-immune 
diseases [28].   
 
4.5.2  In a case control study carried out to establish a possible protective role of 
plant sterols in lung carcinogenesis, total plant sterol intake was associated with a 
reduction in risk of 50% when contrasting the upper exposure quartile with the lower, 
after controlling for major confounders, including tobacco smoking and total energy 
intake.  The researchers concluded that plant sterol intake appears to be an important 
variable in lung carcinogenesis and recommend that further studies are needed to 
replicate the findings [27]. 
 
4.6  
Sulphur containing compounds  
 
4.6.1  Allium plants, for example onions, garlic, leeks, are a major source of sulphur 
compounds in the diet.  The most important compounds are the sulphur substituted 
cysteine sulphoxides. Some sulphur-containing compounds suppress DNA and 
protein synthesis. 
 
4.6.2  The incidence of stomach cancer is lower in individuals and populations with 
high Allium vegetable intakes.  Allium vegetables, particularly garlic, have antibiotic 
activity [29].  Helicobacter pylori are a bacterium implicated in the aetiology of 
 
24

---

stomach cancer.  The antibacterial activity of Allium vegetables against H.pylori has 
been investigated .  Minimum inhibitory concentration of thiosulfinate (a component 
of garlic) was 40 micrograms per ml.  It is plausible that the sensitivity of H. pylori to 
garlic extract at such low concentrations may be related to the reported lower risk of 
stomach cancer in those with a high Allium vegetable intake.  
 
4.7 
Other components  
 
4.7.1  Fruits and vegetables contain an abundance of phenolic compounds, 
terpenoids, pigments, and other natural antioxidants that have been associated with 
protection from and/or treatment of chronic disease such as heart disease, cancer and 
diabetes, and hypertension as well as other medical conditions [30]. 
 
4.7.2  A summary of case control studies found that vegetables with a high content 
of glucosinolates, brassica vegetables, were associated with a lower risk of cancer, 
particularly colon cancer, which was independent of fibre [31]. 
The mechanism by which brassica vegetables, including all types of cabbages, 
broccoli, cauliflower and Brussels sprouts, might decrease the risk of cancer have 
been reviewed at the TNO Nutrition and Food Research Institute, Zeist, The 
Netherlands [32].  Brassicas may be protective against cancer due to their relatively 
high glucosinolate content.  Glucosinolates are usually broken down through 
hydrolysis catalysed by myrosinase, an enzyme that is released from damaged plant 
cells.  Some of the hydrolysis products, such as indoles and isothiocyanates, are able 
to influence phase I and phase II biotransformation enzyme activities, thereby 
possibly influencing several processes related to chemical carcinogenesis, e.g. the 
metabolism, DNA-binding and mutagenic activity of promutagens.  A reducing effect 
on tumour formation has been shown in rats and mice [8].  The anticarcinogenic 
action of isothiocyanates and indoles depends upon many factors, such as the test 
system, the target tissue, the type of carcinogen challenge and the anticarcinogenic 
compound, their dosage, as well as the timing of the treatment.  Most evidence 
concerning anticarcinogenic effects of glucosinolate hydrolysis products and brassica 
vegetables has come from studies in animals.  In addition, studies carried out in 
humans using high but still realistic human consumption levels of indoles and brassica 
vegetables have shown putative positive effects on health. 
 
4.7.3  Lin et al [33] tested the hypothesis that broccoli consumption in combination 
with glutathione transferase M1 (GSTM1) null genotype would be associated with a 
lower prevalence of colorectal adenomas because of higher isothiocyanate levels.  
Subjects with the highest quartile of broccoli intake (an average of 3.7 servings per 
week) had an odds ratio of 0.47 (95% CI 0.30-0.73) for colorectal adenomas, 
compared with subjects who reportedly never ate broccoli.  When stratified by 
GSTM1 genotype, a protective effect of broccoli was observed only among subjects 
with the GSTM1 null genotype (p for trend 0.001; p for interaction 0.01).  Lin claims 
that the observed broccoli-GSTM1 interaction is compatible with an isothiocyanate 
mechanism. 
 
4.7.4 Brassica 
(cruciferous) 
vegetables 
may also induce cytochrome P4501A2 
(CYP1A2), which catalyses the metabolic activation of various procarcinogens, 
including aromatic amines in tobacco.  Thus frequent intake of brassica vegetables 
could result in cancer-enhancing effects.  Probst-Hensch et al [34] found an 
 
25

---

interactive effect of GSTM1 genotype and vegetable intake on CYP1A2 activity.  
They concluded that their results suggest that Brassica vegetables contain CYP1A2 
inducers, which are deactivated in the presence of GSTM1. 
 
4.7.5 Indole-3-carbinol 
(I3C), a naturally occurring component of Brassica 
vegetables, has been shown to reduce the incidence of spontaneous and carcinogen-
induced mammary tumours [35]. 
 
4.7.6  Resveratrol, a phytoalexin found in grapes and other food products has been 
shown to have cancer chemopreventive activity in assays representing three major 
stages of carcinogenesis.  Resveratrol was found to act as an antioxidant and 
antimutagen and to induce phase II drug-metabolizing enzymes (anti-initiation 
activity); it mediated anti-inflammatory effects and inhibited cyclooxygenase and 
hydroperoxidase functions (antipromotion activity); and it induced human 
promyelocytic leukemia cell differentiation (antiprogression activity).  These data 
suggest that resveratrol,  a constituent of grapes and other food products, merits 
investigation as a potential cancer chemopreventive agent in humans [36]. 
 
4.7.7  Vegetables also contain salicylates.  Intake of acetylsalicylic acid reduces the 
risk of cardiovascular disease and is associated with a reduced risk for colorectal 
cancer. Vegetable foods were found to be the main sources of salicylates in a study of 
volunteers from 14 countries and 4 continents to estimate the bioavailable salicylate 
contents of diets [37].  However, amounts in a variety of diets are evidently low and 
probably insufficient to affect disease risk. 
 
4.8 
Carbohydrates (starch and fructose) and non-starch polysaccharides (dietary 
fibre/NSP) 
 
4.8.1  Ingestion of fructose reduces plasma phosphate transiently, and 
hypophosphatemia stimulates 1,25(OH)2D production.  Laboratory and clinical data 
indicate an antitumour effect of 1,25(OH)2D on prostate cancer [38]. 
 
4.8.2  Carbohydrate such as starch and non-starch polysaccharide (NSP) entering the 
large bowel stimulates anaerobic fermentation, leading to an increase in microbial cell 
mass.  The stimulation of bacterial growth, together with water binding to residual 
unfermented NSP, leads to an increase in stool weight, dilution of colonic contents 
and faster transit time through the large gut.  Average stool weight is inversely 
associated with colorectal cancer incidence in different communities.  Low stool 
weights lead to constipation which is a risk factor for colorectal cancer [8]. A 
protective effect of starch and NSP probably arises from their marked effect on 
bacterial metabolism in the large bowel, which can be postulated to affect gene 
expression and DNA repair via increased butyrate production. The pH is reduced 
leading to alterations in secondary bile acid production, and mucosal cell proliferation 
[39].  In rodent models, ‘insoluble’ sources of NSP are generally protective, although 
butyrate, resistant starch and soluble NSP may not reduce tumorigenisis. 
 
4.8.3  Data from a population-based case-control study in Hawaii to evaluate the role 
of various types and components of fibre, as well as micronutrients and foods of plant 
origin, on the risk of colorectal cancer support a protective role of fibre from 
vegetables against colorectal cancer, which appears independent of its water solubility 
 
26

---

property and of the effects of other phytochemicals [40]. Data from other recently 
reported case-control studies also show strong or moderately strong inverse 
associations for dietary fibre and the risk of colorectal cancer or adenomatous polyps 
[41], [42] and [43].  However, the data from two US cohort studies suggest that there 
is only a weak inverse association for fibre from fruit (soluble fibre) and no 
significant association with insoluble fibre and risk of adenomatous polyps of the 
distal colon in men [44] and no significant association between the intake of dietary 
fibre and risk of colorectal cancer in women [45]. 
 
4.8.4  The strength of the hypothesis that fibre reduces breast cancer is its biological 
plausibility, which is supported by experimental and interventional findings and by 
the coherence of observational studies. Three dietary intervention studies taking the 
plasma concentration of oestrogens as an end-point in pre-menopausal women on 
high fibre diets have found a significant reduction in circulating oestradiol or oestrone 
levels [46], [47],  [48]. However, whilst most of the case-control studies show a 
decreased risk in breast cancer for an increased fibre intake, a large cohort study in the 
Netherlands did not show a significant association for dietary fibre and breast cancer 
risk [49]. The apparently inconsistent results may be due to chance or bias, to 
different approaches to data analysis and interpretation, but may also reflect 
heterogeneity in the dietary sources of fibres (cereals or vegetables and fruit) in 
various populations, and different correlates of fibre intake [50] [51].  
 
4.9 Low energy density 
 
4.9.1  Diets high in fruit and vegetables tend to be lower in energy density (lower in 
calories) and fat content because of the relatively high fibre and water content.  Diets 
which are lower in energy density and fat have been shown to reduce the risk of 
weight gain [52]. Overweight and obesity are serious medical problems.  In particular, 
obesity is associated with the development of diabetes mellitus, coronary heart 
disease, hypertension, an increased incidence of some forms of cancer, obstructive 
sleep apnoea, and osteoarthritis of large and small joints. COMA concluded “Overall, 
there is moderate to strong evidence that maintaining a healthy weight would reduce 
the risk of post-menopausal breast cancer and endometrial cancer.  There is weak 
evidence that it would reduce the risk of colon cancer.  There is no evidence that 
increasing obesity protects against cancers.” [8]. 
 
4.9.2  A recent UK intervention study [53] reported that an increase in fruit and 
vegetable consumption amongst people who were consuming more than 35% of their 
energy from fat, resulted in a mean decrease in fat energy percentage from 39% to 
35%.  A Canadian intervention with a diet consisting largely of leafy and other low-
calorie vegetables, fruits and nuts markedly reduced lipid risk factors for 
cardiovascular disease in healthy Canadian men and women [54].   
 
4.10  Conclusion. As described above, there are many biologically plausible reasons 
why consumption of vegetables and fruits might slow or prevent the onset of chronic 
diseases, such as cancer and cardiovascular diseases.  These include the presence in 
plant foods of such potentially anticarcinogenic substances as carotenoids, vitamin C, 
vitamin E, flavonoids, dietary fibre (and its components), and many other classes of 
biologically active compounds. Most of the data for the observations on the 
anticarcinogenic potential of all of these compounds have come from animal and in 
 
27

---

vitro studies.  At almost every one of the stages of the cancer process, identified 
phytochemicals are known to be able to alter the likelihood of carcinogenesis – 
occasionally in a way that enhances risk but usually in a favourable direction.  For 
example, glucosinolates and indoles, thiocyanates and isothiocyanates, phenols, and 
coumarins can induce a multiplicity of phase II  (solubilising and usually inactivating) 
enzymes; ascorbate and phenols block the formation of carcinogens such as 
nitrosamines; flavonoids and carotenoids act as antioxidants, essentially disabling the 
carcinogenic potential of specific compounds; lipid soluble compounds such as 
carotenoids and sterols may alter membrane structure or integrity; some sulphur-
containing compounds suppress DNA and protein synthesis; carotenoids can suppress 
DNA synthesis and enhance differentiation; and phytoestrogens compete with 
oestradiol for oestrogen receptors in a way that is generally antiproliferative.  For 
cardiovascular disease, the main components likely to decrease the risk of disease are 
the antioxidants, which are capable of reducing the oxidation of LDL cholesterol and 
folate which lowers plasma homocysteine levels and hence reduces the strong risk 
factor for arteriosclerosis of the coronary, cerebral, and peripheral arteries. 
 
5. Cancer 
 
5.1 
The body of evidence that indicates that high intakes of fruit and vegetables 
are associated with a reduced risk of cancer at several sites is growing all the time. 
Overall, the epidemiological evidence on the link between diet and cancer provides 
consistent evidence for an inverse association between risk of certain cancers and the 
consumption of fruit and vegetables [55], [8], [56], [57].  The association is generally 
most marked for epithelial cancers, apparently stronger for those of the digestive and 
respiratory tracts (oral, pharynx, oesophagus, stomach, larynx, large bowel/colon and 
rectum) as well as of the urinary tract, and somewhat weaker for hormone-related 
cancers (breast, endometrium) (see Table 1).  Fruits and vegetables are a rich source 
of micronutrients and other metabolically active substances.  Many of the 
micronutrients appear to be good candidates for anticarcinogens because of their 
involvement in physiological processes associated with cell proliferation and the 
maintenance of normal function (see Chapter 4 for more detailed information).  The 
fact that there are so many of these micronutrients and metabolically active substances 
lends biological plausibility to the findings of the human observational studies. 
 
5.2 
The tools to determine the dietary pattern that is associated with the lowest 
overall disease (cancer, CHD, etc.) risk are epidemiology, animal models, in vitro 
studies, and dietary intervention studies.   These all have their strengths and 
weaknesses.  While epidemiological studies may produce associations from which 
causal inferences can be drawn, establishing causal links is much more difficult.  The 
strength of causal inference that can be drawn from epidemiological studies differs 
with different types of study design.  Broadly, epidemiological studies can be divided 
into observational (the researcher observes but does not determine exposure) or 
experimental/interventional designs (the researcher determines the exposure).  
Individual-based observational studies may be divided into case-control 
(retrospective) or cohort (prospective) studies. Population-based observational studies 
are known as ecological studies.  When drawing causal inferences the most powerful 
studies are experimental, followed by individual based observational studies, 
particularly cohort studies.   
 
 
28

---

5.3 
Several reviews of epidemiologic studies have been carried out.  COMA’s 
Working Group on Diet and Cancer commissioned a review of the epidemiology 
relating diet to the development of cancers and considered the data derived from 
human experimental studies, both in vitro and in vivo, and reached the following 
conclusion for fruit and vegetable consumption and the risk of developing cancer:   
 
Overall, there is moderate evidence that higher vegetable consumption would reduce 
the risk of colorectal cancer, and that higher fruit and vegetable consumption would 
reduce the risk of gastric cancer.  There is weak evidence, based on fewer data, that 
higher fruit and vegetable consumption would reduce the risk of breast cancer.  These 
cancers combined represent about 18% of the cancer burden in men and about 39% 
of the cancer burden in women in the UK.  Even a small reduction in relative risk 
would have important public health benefits in terms of the reduction in the absolute 
numbers of people affected.  In addition, the data are generally consistent with a 
graded reduction in risk for higher fruit and vegetable consumption and no cancer 
consistently shows a higher risk with higher fruit and vegetable consumption.  The 
overall picture, therefore, is consistent and supports the hypothesis that the 
consumption of fruits and vegetables protects against the development of some 
cancers.  The working Group recommends that fruit and vegetable consumption in 
the UK should increase. 
 
5.3.1  Other reviews show consistently a cancer protective effect of fruit and 
vegetable consumption [55], [56], but there is little understanding of which 
phytochemicals account for this observation. Table 1 sets out studies published since 
the end of COMA’s review in 1995 for individual cancers. [It is acknowledged that 
whilst this was a fairly thorough review it was not comprehensive and other studies 
may well have been published].  Few studies were based in the UK.  The findings are 
summarised below: 
 
5.3.2 Bladder 
cancer: A large male cohort study [58] found a weak, non-significant 
inverse association between total vegetable intake and bladder cancer risk, and a 
significant inverse association between cruciferous vegetables with risk (RR=0.49; 
95% CI, 0.32-0.75, for highest cf. lowest intake).  A review of 7 case-control and 
cohort studies [59] found that in six studies there was a reduced risk with increasing 
consumption of fruit and vegetables, with RR estimates between 0.5 and 0.7 for the 
highest cf. lowest consumption. 
 
5.3.3.  Breast cancer:  All 4 of the case-control studies [60], [61], [62], [63] which 
reported on fruit and vegetable intake and risk of breast cancer found an inverse 
association, although not all the associations were statistically significant.  In a cohort 
study in the Netherlands [49], fruit consumption, although not vegetable consumption, 
showed a non-significant inverse association with breast cancer risk (RR = 0.76 for 
highest cf. lowest quintile of intake).  Of the two reviews identified, one [51] 
concluded that any reduction in breast cancer risk appears to be significantly 
dependent on the level of fibre intake and recommended a fibre-rich diet both from 
cereals and from fruit and vegetables.  The second review [64] concluded that the 
considerable evidence suggests that low intakes of vegetables modestly increase the 
risk of breast cancer. 
 
 
29

---

5.3.4  Cervical cancer:  One case-control study [65] found a weak protective effect 
of increased fruit consumption (P=0.03), but this became non-significant after 
adjustment for sexual behaviour. 
 
5.3.5  Colorectal cancer and adenomatous polyps:  All seven papers of case-
control studies [42], [66], [41], [43], [67], [40], [68] reporting on vegetable and fruit 
consumption or vegetable fibre consumption and risk of colorectal cancer found an 
inverse association, although three of the reports were considering different fruit and 
vegetable components in the same large study [68], [67], [43].  The latter authors 
concluded that ‘for the addition of one daily serving of vegetables, there was more 
than 20% reduction in risk of colorectal cancer’.   
 
5.3.6  The two cohort studies which reported on fibre from fruit and vegetables, 
rather than the food groups, did not find a strong association. The Health 
Professionals Follow-up cohort study reported a modest reduced risk of distal colon 
adenoma with increasing intake of fibre from fruit (P-trend = 0.03) but not vegetables 
[44]. The Nurses’ Health Cohort Study reported no association between the intake of 
dietary fibre and the risk of colorectal cancer or colorectal adenoma after adjustment 
for age, established risk factors, and total energy intake [45].   
 
5.3.7  An ecological study using data from the ‘Seven Countries Study’ found that 
fibre intake was inversely associated with colorectal cancer mortality (an increase of 
10g fibre/day was associated with 33% lower 25-year colorectal cancer mortality risk) 
[69].  An analysis of FAO data and mortality data for colorectal and breast cancers 
[70] found that vegetables (and cereals) were protective against both cancers, fruit had 
no effect and starchy roots had a weak, non-significant promoting effect.  Only the 
current intake of vegetables was protective, intake in early life seemed to offer no 
protection. International comparisons show strong inverse protective associations with 
starch, dietary fibre (NSP) and vegetable intake, and positive associations with meat 
consumption in large bowel cancer [39].  
 
5.3.8  The Polyp Prevention Trial [71] [72], a multicentre randomised controlled trial 
(RCT), is examining the effect of a low-fat (20% of total energy intake), high fibre 
(18g/1000kcal), high vegetable and fruit (5-8 daily servings) dietary pattern on the 
recurrence of adenomatous polyps of the large bowel, but the results were not 
identified in this review.  In a review of epidemiology of colorectal cancer, Wilmink 
(1997) [73] concluded that “conclusive evidence of the effectiveness of primary 
prevention of colorectal cancer via dietary measures is lacking, but that secondary 
prevention by interrupting the adenoma carcinoma sequence is an actual possibility, 
its effectiveness, however, needs to be determined”. 
    
5.3.9  Endometrium:  Only one case-control study [74] was identified.  This study 
compared cases from five ethnic groups living in Hawaii with population based 
ethnically matched controls and found sources of fibre, including cereal and vegetable 
and fruit fibre, were associated with a 29-46% reduction in risk for women in the 
highest quartiles of consumption.  High consumption of soy products and other 
legumes was associated with a decreased risk of endometrial cancer (P for trend = 
0.01; OR = 0.46, 95% CI 0.26-0.83) for the highest cf. lowest quartile of soy intake.  
The effects of soy and legumes were limited to women who were never pregnant or 
never used unopposed oestrogens. 
 
30

---

 
5.3.10  Gastric/stomach: All four of the case-control studies [75], [76], [77], [78] 
which reported on the consumption of fruit and vegetables or their components as a 
risk factor for gastric cancer found some inverse association between vegetables 
and/or fruit or components of fruit and vegetables and risk of gastric cancer.  
However, pickled vegetables increased the risk in one study [76] and no association 
was found for carotenoids in another study [77] although there was a significant 
inverse association with kaempferol and total flavonoid intake.    
 
5.3.11  One of the few cohort studies in the UK from which data is available, the 
Caerphilly Study [79], found that fruit and vegetable consumption was independently 
from other risk factors, inversely related to mortality from cancer of the digestive tract 
(P for trend +0.021), mainly due to an inverse association with fruit consumption (RR 
highest cf. lowest quartile = 0.3, 95% CI: 0.1-0.8).  The Netherlands Cohort Study 
[80] found age- and sex- adjusted rate ratios of stomach cancer in increasing quintiles 
of combined fruit and vegetable consumption were 1.00, 0.70, 0.65, 0.76 and 0.64 (p 
trend = 0.04).  However, inverse associations for total vegetables, pulses, raw leafy 
vegetables, total fruit, citrus fruit, and apples and pears that were observed in crude 
analyses became weaker or disappeared in multivariate analyses.  In a Swedish 
Cohort study ([81] the risk of stomach cancer was inversely related to fruit and 
vegetable consumption, RR = 5.5 among subjects with lowest cf. highest intake with a 
statistically significant dose risk trend P <0.05.  The European ‘Seven Countries’ 
Cohort Study [82] reported that adherence to the recommendation of the European 
Code Against Cancer on consumption of vegetables, fruits and fibre-rich cereals was 
inversely related to stomach cancer mortality.   
 
5.3.12  An ecological study which examined the relationship between the 
consumption of different foods in each of the 50 Spanish provinces and the mortality 
rate of stomach cancer in these same provinces 20 years later [83] found a significant 
inverse association between gastric cancer mortality and consumption of vegetables, 
fruits and vegetable oil.  Another ecological study of 18 districts in Portugal [84] 
found that the highest negative correlation coefficients between dietary consumption 
and gastric cancer mortality were obtained for vegetables, fruit, vitamin A and 
carotene consumption.  Two reviews by the same authors [85], [86] found 
consumption of vegetables and fruit is associated with decreased risk of stomach 
cancer and that both epidemiologic and experimental data suggest that vitamin C and 
carotenoids lower risk of stomach cancer. 
 
5.3.13  Kidney: All five case-control studies identified [87], [88], [89], [90], [91] 
which reported on fruit and vegetable consumption and risk of renal cell carcinoma 
found inverse associations between fruit, vegetables or their components e.g. 
carotenoids and risk of renal cell carcinoma.  The International Renal Cell Cancer 
study [90] which is operating in five countries, Australia, Denmark, Germany, 
Sweden and the USA, reported low intake of fruit and vegetables was associated with 
an increased risk of renal-cell cancer (RCC).  A USA study [88] found strong inverse 
associations between cruciferous and dark green vegetable and RCC (P trend for both 
<0.001) and between carotenoids and RCC risk.  A significantly protective effect on 
risk of RCC was observed with increasing consumption of fruit (P for trend, 0.05) 
which was strongest among non-smokers and a suggestion of a protective effect of 
high total vegetable consumption was seen in the Swedish study [91]. 
 
31

---

 
5.3.14  Lung:  All six of the case-control studies identified [27], [92], [93], [94], [95], 
[96] found an inverse association between vegetables and/or fruit and/or components 
e.g. carotenoids and sterols and risk of lung cancer.  Statistically significant inverse 
associations were found for consumption of yellow orange vegetables (mainly carrots) 
and tomatoes in a study in women in Spain [94], for frequent usual consumption of 
boiled vegetables and risk of lung cancer in men in Poland [93], and for high fruit 
consumption and lung cancer risk in African Americans and Mexican Americans in 
the USA [92] and finally for total plant sterol intake and lung cancer risk in Uruguay 
[27].  The other inverse associations were not statistically significant.   
 
5.3.15  Three cohort studies, one each in the Netherlands [97], Finland [17] and the 
USA [98], reported mainly on the antioxidant (vitamin C, E, carotenoid and 
flavonoid) with few references to actual fruit and vegetable intakes and risk.   In the 
Netherlands cases with low stable intakes (i.e. low at all three points of measurement 
– 1960, 1965 and 1970) of vegetables, fruits, and beta-carotene experienced more 
than two-fold increased RR of lung cancer than those with high stable intakes. The 
Finnish study reported that consumption of apples showed a statistically significant 
inverse association with risk of lung cancer (RR=0.42 (95% CI, 0.23-0.76) after 
adjustment for the intake of other fruit and vegetables. The authors of the Netherlands 
report conclude that there is no apparent relation of vitamin E to lung cancer risk, 
however, for beta-carotene, vitamin C, vegetables, and fruit there is a weak inverse 
association. The USA group report a significant inverse association for vitamin C and 
lung cancer risk (RR=0.66, 95% CI, 0.45-0.96), and for carotenoids among current 
smokers (RR=0.49, 95% CI, 0.29-0.84), which was modified by the intensity of 
smoking (RR= 0.33, 95% CI, 0.13-0.84) only for those in the lowest tertile of pack-
years of smoking), and for vitamin E in current smokers in the lowest tertile of pack-
years of smoking (RR=0.36, 95% CI, 0.16-0.83).  When vitamin E, vitamin C and 
carotenoid intakes were examined in combination in the USA study, a strong 
protective effect was observed for those in the highest compared with the lowest 
quartile of all three intakes (RR=0.32, 95% CI, 0.14-0.74).  Flavonoids were the focus 
of attention in the Finnish study, where a statistically significant inverse association 
for total flavonoid intake and risk of lung cancer was reported (RR=0.54, 95% CI, 
0.34-0.87) for highest cf. lowest quartile of intake.    
 
5.3.16  The two reviews of epidemiologic studies and lung cancer included in Table 1, 
appear to have reached different conclusions. Koo (1997) [99] in a review of studies 
over 20 or more years concluded that the data have not provided overwhelming 
evidence that vegetables and fruit or such micronutrients as carotenoids, selenium, 
vitamins A, C or E is associated with reduced lung cancer risk, whereas Ziegler 
(1996) [100] concluded that prospective and retrospective observational studies 
suggest strongly that increased vegetable and fruit intake is associated with reduced 
risk in men and women; in various countries; in smokers, ex-smokers, and never 
smokers; and for all histologic types of lung cancer.  In addition that prospective 
studies of blood beta-carotene levels (biomarker of fruit and vegetable intake) indicate 
that low levels are predictive of increased lung cancer incidence.  However, in two 
RCT intervention studies, lung cancer incidence, and total mortality were increased 
significantly among the men receiving beta-carotene supplements.  
 
 
32

---

5.3.17  Upper aerodigestive tract:  Seven papers from case-control studies were 
identified [101], [102], [103], [104], [105], [106], [107] which covered oesophageal, 
laryngeal, oral, pharyngeal and hypopharyngeal cancers.  Six studies were from 
Europe and the seventh was a nested case-control study based in the USA [107].  The 
latter study found that alpha- and beta-carotene, beta-cryptoxanthin, total carotenoids 
and gamma-tocopherol levels were significantly lower in the upper aerodigestive tract 
cancer patients than in the controls. Both alpha- and beta-carotene were consistently 
and strongly associated with reduced risk at each site.  The two case-control studies 
dealing with oesophageal cancer patients only [101], [102] found that low 
consumption of fruit and vegetables in one study were associated strongly with an 
increase in oesophageal cancer risk.  A review [108] of ecological, case-control, 
cohort and intervention studies of oesophageal cancer found a large body of evidence, 
especially from case-control studies,  for the protective effect of fruit and vegetable 
consumption.   
 
5.3.18  The four European case-control studies which reported on cancers of the oral 
cavity, pharynx, larynx and hypopharynx [103] , [104], [105], [106] found that high 
intake or frequent consumption of fruit and vegetables was associated with reduced 
risk of these cancers.  The only cohort study identified was a UK study, the Caerphilly 
Study [79], which found that fruit consumption and intake of dietary fibre were 
inversely associated with respiratory tract cancer, but after adjustment for potential 
confounders including age, smoking, and social class, the association with fruit 
consumption became non-significant. In the same paper the researchers reported that 
vegetable and fruit consumption was inversely related to all-cause cancer mortality, 
and the strongest association was observed for fruit consumption (RR highest cf. 
lowest quartile was 0.5; 95% CI, 0.3 – 1.0).  In a review of the epidemiologic 
evidence for  laryngeal cancer, Riboli et al (1996) [109] found consistent evidence 
that low consumption of fruit and vegetables is associated with higher risk of 
laryngeal cancer, after statistical adjustment for alcohol and tobacco.  The evidence 
indicates that, in the presence of tobacco and/or alcohol, low intake of fruit and 
vegetables may account for 25-50% of the cases among men.  La Vecchia et al (1997) 
[110] reviewed the epidemiology for oral and pharyngeal cancers with a specific 
focus on Europe and concluded that the influence of dietary factors, in particular a 
diet poor in fresh fruit and vegetables, may account for 10-15% of oral cancer cases in 
Europe, and that alcohol and tobacco account for about 75% of oral cancers. 
 
5.3.19  Ovary:  Only one study, a women’s cohort study [111] was identified in this 
review.  Total vegetable intake was modestly and inversely associated with the risk of 
ovarian cancer (p for trend = 0.21), however, green leafy vegetable intake was more 
strongly associated with a decreased risk, highest cf. lowest intake RR+ 0.44 
(p=0.01). 
 
5.3.20  Pancreas:  A nested case-control study in Finland [112] found that serum 
folate and pyridoxal-5’-phosphate concentrations showed statistically significant 
inverse dose-response relationships with pancreatic cancer risk, highest serum tertiles 
having approximately half the risk of the lowest.  A summary of data on the 
occurrence, the trends, and the life-style, environmental, occupational and genetic 
determinants of pancreatic cancer in Sweden found that fruit and vegetable 
consumption appear to be protective [113], and a review of international 
 
33

---

epidemiologic data [114] found consistent inverse relationships with fruit and 
vegetable intakes and with two markers of such foods, fibre and vitamin C. 
 
5.3.21  Prostate: Three case-control studies [115], [116], [117] reported on fruit and 
vegetable consumption or their components and risk of prostate cancer.  The largest of 
the three studies [116] found no association between fruit intake and prostate cancer 
risk but a significant inverse trend with frequency of consumption of vegetables and 
risk of prostate cancer. A study in Uruguay [117] looked at vegetables and fruit 
combined and found an OR of 0.5, 95% CI, 0.3-0.9 for the highest quartile of intake 
cf. lowest quartile.  The third paper (study) reported only on intake of retinol and 
vegetable protein for which there was a non-significant inverse relationship with risk 
of prostate cancer.  A cohort study [118] also reported inverse associations between 
risk of prostate cancer and total vegetables RR 0.80, 95% CI, 0.57-1.12; prepared 
vegetables RR 0.85, 95% CI, 0.61-1,19; raw vegetables RR 0.96, 95% CI, 0.69-1.34; 
pulses RR 0.71, 95% CI 0.51-0.98 (all except pulses were non-significant), and 
positive associations with total fruit RR 1.31, 95% CI, 0.96-1.79 and citrus fruit RR 
1.27, 95% CI, 0.93-1.73.  Individual vegetables and fruit were evaluated as  
continuous variables (g/day).  Non-significant inverse associations (RRs per 
increment of 25g/day) were found for consumption of kale (RR 0.74), raw endive (RR 
0.72), mandarins (RR 0.75), and raisins or other dried fruit (RR 0.49).  Observed 
positive associations were significant for leeks (RR 1.60) and mushrooms (RR 1.49).   
A second cohort study [38] reported on intake of fructose and fruit and risk of prostate 
cancer.  High fructose intake was related to a lower risk of advanced prostate cancer 
(multivariate RR 0.51, 95% CI, 0.33-0.80 for intakes <70 versus ≤40g/day; p for trend 
0.007).  Fruit intake was inversely associated with risk of advanced prostate cancer 
(RR 0.63, 95% CI 0.43-0.93; for >5 cf. ≤1 serving/day) and this association was 
accounted for by fructose intake.  Two reviews of the epidemiology [119], [120] 
reported that evidence of a protective effect of fruit and vegetables is weak and 
inconsistent.  Interest appears to be growing in phytoestrogens and the extent to which 
dietary manipulation with these and other phytochemicals might influence prostate 
cancer by modifying male sex hormone levels or actions. 
 
5.3.22  It would appear from the above review of the scientific literature that the 
association between fruit and vegetables and cancers is generally most marked for 
epithelial cancers, apparently stronger for those of the digestive and respiratory tracts, 
and somewhat weaker for hormone-related cancers. 
 
 
5.3.23  Stienmetz and Potter (1996) [55] in their review of the scientific literature on 
the relationship between vegetable and fruit consumption and risk of cancer state: 
 
 The evidence for a protective effect of greater vegetable and fruit consumption is 
consistent for cancers of the stomach, oesophagus, lung, oral cavity and pharynx, 
endometrium, pancreas, and colon.  The types of vegetables and fruit that most often 
appear to be protective against cancer are raw vegetables, followed by allium 
vegetables, carrots, green vegetables, cruciferous vegetables, and tomatoes. 
 
5.3.24  The conclusion and recommendation of the World Cancer Research Fund
 
(WCRF) and the American Institute for Cancer Research in their report ‘Food, 
 
34

---

Nutrition and the Prevention of Cancer: a global perspective’ [56] for vegetables and 
fruit follows: 
 
There is a strong and consistent pattern showing that diets high in vegetables and 
fruits decrease the risk of many cancers, and perhaps cancer in general.  The 
evidence that such diets decrease the risk of mouth and pharyngeal, oesophageal, 
lung and stomach cancers, is convincing, and they probably also protect against 
laryngeal, pancreatic, breast, and bladder cancers.  The evidence that diets high in 
vegetables decrease the risk of colorectal cancer is convincing.  The panel notes that 
such diets possibly protect against ovarian, cervical, endometrial and thyroid 
cancers, and that diets high in vegetables possibly protect against primary liver, 
prostate and renal cancers.  On the whole, the evidence for a protective effect of 
vegetables is rather stronger than that for fruits, perhaps reflecting the fact that 
vegetables are generally consumed in greater quantities than fruits, and thus in more 
variable quantities within populations.  
 
5.3.25  A review by Verhoeven et al [121] of 7 cohort and 87 case-control studies on 
the association between brassica consumption and cancer risk found that the cohort 
studies showed inverse associations between the consumption of cabbage, 
cauliflower, and broccoli and risk of lung cancer; between the consumption of 
brassicas and risk of stomach cancer; between broccoli consumption and risk of all 
cancers taken together; and between brassica consumption and the occurrence of 
secondary primary cancers.  Of the case-control studies, 67% showed inverse 
association between consumption of total brassica vegetables and risk of cancer at 
various sites.  For cabbage, broccoli, cauliflower, and Brussels sprouts, these 
percentages were 70, 56, 67 and 29%, respectively.  Although, the measured effects 
might have been distorted by various types of bias, the reviewers concluded that a 
high consumption of brassica vegetables is associated with a decreased risk of cancer.  
This association appears to be the most consistent for lung, stomach, colon, and rectal 
cancer and least consistent for prostatic, endometrial, and ovarian cancer.  It is not yet 
possible to resolve whether associations are to be attributed to brassica vegetables per 
se or to vegetables in general.  
 
5.3.26  Consumption of diets low in plant foods results in a reduced intake of a wide 
variety of those substances that can plausibly lower cancer risk.  In the presence of a 
diet and lifestyle high in potential carcinogens (whether derived from fungal 
contamination, cooking or tobacco) or high in promoters (such as salt and alcohol), 
overall risk of cancer at many epithelial sites is elevated.  Plant-foods appear to exert 
a generally risk lowering effect; the patterns of exposure to cancer initiators and 
promoters and of genetic susceptibility may determine the variations in the site 
specific risks of cancer seen across populations [122]. Although a large emphasis has 
been given to different components of the diet attention has recently shifted to the diet 
as a whole.   
 
5.3.27  From a public health viewpoint, epidemiological evidence indicates that a 
substantial reduction in epithelial cancer risk can be obtained by increasing fruit and 
vegetable consumption in adults. Evidence is lacking from observational studies that 
increasing fruit and vegetable consumption in childhood will reduce the risk of cancer 
in adult life. COMA’s Working Group on Diet and Cancer recognised the emerging 
evidence on the multi-stage processes involved in carcinogenesis and the wide variety 
 
35

---

of dietary factors which could influence each stage of the process means that the 
dietary contributors to the stages of carcinogenesis a decade or more before a cancer 
becomes apparent are difficult to identify even in detailed prospective studies [8].  
Recommendations to both adults and young persons (children) to make dietary 
changes, particularly to increase the consumption of vegetables and fruits, must be 
continued, especially with regard to persons familially susceptible and, additionally, 
with the recognition that the measures recommended also serve to lessen 
susceptibility to cardiovascular diseases and other diseases.  Compliance, even by a 
small proportion, could result in large numbers benefiting due to the burden from 
these chronic diseases. 
 
5.3.28  Understanding the association between fruit and vegetable intake and other 
health behaviours is important for properly interpreting the rapidly growing number 
of studies that link low intakes of fruits and vegetables to the risk of cancer and 
cardiovascular disease.  To examine the association between fruit and vegetable 
intake and behavioural risk factors for chronic diseases, Serdula et al [123] analyzed 
data from a population-based behavioural risk factor survey which took place in 1990 
among 21,892 adults in the USA.  Respondents answered questions about behaviours 
related to chronic disease risk, including their frequency of intake of fruits and 
vegetables, using a six-item questionnaire.  Consumption of fruit and vegetables was 
lowest among those who also reported that they were sedentary, heavy smokers, 
heavy drinkers, or had never had their blood cholesterol checked.  Because fruit and 
vegetable intake covaries with several other chronic disease risk factors, it is 
important to account for possible confounding between fruit and vegetable intake and 
other behaviours in etiologic studies of the risk of cancer and cardiovascular disease. 
However, issues of confounding should not cause anyone to discount the value of 
eating more fruits and vegetables. 
 
6 Cardiovascular 
Disease 
 
6.1 
Observational studies have found that persons who consume large amounts of 
fruit and vegetables have lower rates of coronary heart disease. Scientific papers 
reporting on observational and experimental/intervention studies which considered 
fruit and vegetable consumption and risk of stroke and/or coronary heart disease risk 
were identified from a review of the literature since 1995 for this report.  This review 
was not systematic but covered a large number of journals and it included all studies 
identified whether positive or negative associations were described. Table 2 lists the 
studies identified.   
 
6.2 
A systematic review [124] of all ecological, case-control, cohort studies, and 
all unconfounded trials in humans which had outcomes of coronary heart disease, 
stroke and total circulatory disease and reported on fresh fruit and vegetables or a 
nutrient which could serve as a proxy has been carried out by Ness and Powles (1997) 
at the Institute of Public Health, Cambridge.  For coronary heart disease, 9 of 10 
ecological studies, 2 of 3 case-control studies and 6 of 16 cohort studies  found a 
significant protective association with consumption of fruit and vegetables or 
surrogate nutrients.  For stroke, 3 of 5 ecological studies, none (of one) case-control 
study and 6 of 8 cohort studies found a significant protective association with 
consumption of fruit and vegetables or surrogate nutrients.  For total circulatory 
disease, one of two cohort studies reported a significant protective association.   No 
 
36

---

attempt was made to arrive at a summary measure of the association because of the 
differences in study type, study quality and the different exposure measures used,  
The authors concluded that although null findings may be underreported the results 
are consistent with a strong protective effect of fruit and vegetables for stroke and a 
weaker protective effect on coronary heart disease. 
 
6.3 
Several cohort studies have reported since the above review. Two papers 
based on cohorts of men and women including a large proportion who were 
vegetarians looked at all cause mortality, ischaemic heart disease and cerebrovascular 
disease.  The first one reported on a UK cohort of men and women [125] which had a 
mean follow-up time of 16.8 years.  Within the cohort, daily consumption of fresh 
fruit was associated with significantly reduced mortality from ischaemic heart disease 
(RR, adjusted for smoking, 0.76; 95% CI, 0.47-0.98) and from cerebrovascular 
disease (RR 0.68, 95% CI, 0.47-0.98) and for all causes of mortality combined (0.79., 
95% CI, 0.70-0.90).  The second was a collaborative analysis using original data from 
5 prospective studies in the UK, USA and Germany with a mean follow-up of 10.6 
years [126].  In comparison with non-vegetarians, vegetarians had a 24% reduction in 
mortality from ischaemic heart disease (death rate ratio 0.76, 95% CI, 0.62-0.94).  
The reduction in mortality was little affected by adjustment for alcohol intake, 
education, exercise and BMI, suggesting that it cannot be explained by confounding 
of these variables.  There was no significant difference between vegetarians and non-
vegetarians in mortality from other causes of death examined.  
 
6.4 
Joshipura et al (1999) [127] examined intakes of specific fruits and vegetables 
as well as overall fruit and vegetable consumption, in relation to the incidence of 
ischemic stroke in two large cohorts of women and men, to understand better the 
relationship between diet and stroke.  All individuals were free of cardiovascular 
disease, cancer, and diabetes at baseline.  The follow-up for women was 14 years and 
for men eight years.  After controlling for standard cardiovascular risk factors, 
persons in the highest quintile of fruit and vegetable intake (median of 5.1 servings 
per day among men and 5.8 servings per day among women) had a RR of 0.69 (95% 
CI 0.52-0.92) compared with those in the lowest quintile.  An increment of 1 serving 
per day of fruits or vegetables was associated with a 6% lower risk of ischemic stroke 
(RR, 0.94; 95% CI, 0.90-0.99).  Cruciferous vegetables (RR, 0.68 for an increment of 
1 serving per day; 95% CI, 0.49-0.94), green leafy vegetables (RR, 0.79; 95% CI, 
0.62-0.99), citrus fruit including juice (RR, 0.81; 95% CI, 0.68-0.96), and citrus fruit 
juice (RR 0.75; 95% CI, 0.61-0.93) contributed most to the apparent protective effect 
of total fruits and vegetables.  Legumes were not associated with lower ischemic 
stroke risk.  The authors conclude that these data support a protective relationship 
between consumption of fruit and vegetables – particularly cruciferous and green 
leafy vegetables and citrus fruit and juice- and ischemic stroke risk. 
 
6.5 
The Nurses’ Health cohort study [128] reported that intakes of fruits and 
vegetables and magnesium, potassium and dietary fibre were inversely associated 
with self-reported systolic and diastolic blood pressures, but that magnesium, 
potassium and dietary fibre were not significantly associated with risk of diagnosed 
hypertension.  The Iowa postmenopausal women’s cohort study [15] reported total 
flavonoid intake was associated with a decreased risk of CHD death after adjusting 
for age and energy intake (p for trend = 0.04). Of the foods that contributed the most 
 
37

---

to flavonoid intake in this cohort, only broccoli was strongly associated with reduced 
risk of CHD.   
 
6.6 
The Male Health Professionals cohort study [14] reported that a modest but 
non-significant inverse association between intake of flavonols and flavones and 
subsequent coronary mortality rates (RR=0.63; 95% CI, 0.33-1.20) for the highest 
compared with the lowest quintile for intake.  However, for non-fatal myocardial 
infarction the RR was 1.08 (95% CI, 0.81-1.43) for the highest compared with the 
lowest quintiles for intake of flavonols and flavones after adjustment for risk factors 
for cardiovascular disease. Hollman and Katan [13] investigated the association of 
flavones with risk of CHD in a review of the literature and found that the intake of 
flavonols and flavones was inversely associated with subsequent CHD in most but not 
all prospective epidemiological studies. 
 
6.7 
The Massachusetts Health Care Panel cohort study [129] found that, for total 
CVD deaths and fatal MIs, risks were lower among those in the highest quartile for 
consumption of carotene containing fruits and vegetables as compared with the 
lowest.  For death due to CVD, RR=0.54 (95% CI, 0.34-0.86) and for MI RR= 0.25 
(95% CI, 0.09-0.67). 
 
6.8 
Rexrode (1996) [130] writing about antioxidant vitamins and coronary heart 
disease said “The strength of the epidemiological evidence differs for each of the 
antioxidant vitamins.  High intake of vitamin E from food or supplements has 
generally been associated with a lower incidence of coronary heart disease.  The 
evidence for beta-carotene intake is inconsistent, with several studies finding a modest 
reduction in risks among persons with high intake, and others failing to find an 
association.  Although many studies have examined the relationship between vitamin 
C intake and CVD, no significant benefit was seen in any of the large studies that 
were able to control for other antioxidant intake or multivitamin use.  Observational 
studies cannot discern whether the decreased risk observed is caused by the 
antioxidants themselves or other characteristics of the individuals who consume them.  
The epidemiologic associations may be due to other nutrients in antioxidant rich 
foods, or other dietary or lifestyle factors.  RCTs are necessary to confirm or refute 
the observational data.  On the basis of available evidence, we should recommend a 
healthy diet, rich in fruit and vegetables, but should not endorse vitamin 
supplementation unless conclusive evidence of benefit is demonstrated in clinical 
trials.” 
 
6.9 
Randomised controlled trial data will be essential in fully assessing whether or 
not there is a causal effect of antioxidants in reducing the risk of cardiovascular 
disease.  A report of a review of observational and intervention studies was published 
in 1999 [131].  The authors found six cohort studies reported relative risks of CHD 
between 0.27 and 0.78 for high beta-carotene levels (plasma/serum levels and dietary 
intake), but four more recent ones reported RR around unity (range 0.84 to 1.19).  The 
evidence from case-control studies supports a role of beta-carotene in the prevention 
of CHD (ORs between 0.37 and 0.71), with a possible stronger protection for current 
smokers.  The four published randomised clinical trials of beta-carotene 
supplementation found RR close to unity (range 0.96 to 1.26) for the relation between 
beta-carotene and CHD.  The authors concluded that the benefit reported in some 
observational studies may be related to consumption of foods rich in beta-carotene 
 
38

---

rather than  beta-carotene itself, as foods rich in beta-carotene are usually rich in other 
antioxidants, or to time related factors, i.e., longer supplementation in intervention 
studies. Other reviews have been carried out for vitamin E and coronary artery disease 
[132] and vitamins C, E, A, and carotenoids [133].  Gey (1998) concludes that 
optimal health requires synchronously optimised vitamins C,E, A carotenoids and 
vegetable co-nutrients.  
 
6.10   This has stimulated a new generation of laboratory research aimed at better 
defining the role of antioxidants in the primary and secondary prevention of 
atherosclerotic disease, and information concerning the amount of these nutrients 
which may be required. COMA’s Cardiovascular Review Group, reporting in 1994, 
stated “Our recommendations for an increase in starchy foods and fruits and 
vegetables and substitution of unsaturated vegetable oils for more saturated animal 
fats would have the effect of increasing the consumption of various antioxidants.  We 
do not recommend supplementation with concentrated or purified preparations as a 
widespread policy for CHD prevention, as long term safety and efficacy in a variety 
of population groups has not been demonstrated.  In contrast a diet rich in vegetables 
and fruits is generally regarded as conducive to long term health [1].  In addition the 
Review Group recommended that research is carried out to identify the range and 
effects of various antioxidants in foods and to quantify desirable intakes of dietary 
antioxidants and/or their food sources. 
 
6.11  A comparative cohort study ‘Seven Countries Study’ [134] found large 
differences in food group consumption between the seven countries (USA, Finland, 
The Netherlands, Italy, Yugoslavia, Greece and Japan) and large differences in 
population death rates from CHD.  Vegetable food groups (except potatoes) as well as 
fish and alcohol were inversely associated with CHD mortality. Other comparative 
population studies [135], [136],[137] have also found inverse correlations between 
incidence of coronary heart disease and consumption of fruit and vegetables, but one 
study comparing healthy young persons from southern Italy and Bristol [138] found 
that the intake of fresh fruit and vegetables did not differ, although the southern 
Italians consumed more tomatoes and tomato juice and had a higher level of plasma 
lipid antioxidant vitamin E and of beta-carotene than the Bristol group, who had a 
higher level of plasma lipid peroxidation. 
 
6.12  Epidemiological data suggest that the ordinary diets consumed in parts of 
Europe (particularly Mediterranean regions) are associated with reduced risk of 
cardiovascular disease (see Chapter 9 for further information on Mediterranean diets). 
 
6.13  Two cohort studies in the USA [139], [140] and one in Finnish men [141] 
have shown inverse correlations for dietary fibre and coronary heart disease.  For 
women in the Nurses’ Health study [139], those with the highest quintile of dietary 
fibre intake had an RR for major CHD events of 0.53 (95% CI, 0.40-0.69) compared 
with those with the lowest quintile of intake.  After controlling for cardiovascular risk 
factors, age, dietary factors and multivitamin supplement use, the RR was 0.77 (95% 
CI, 0.57-1.04).  However, only cereal fibre among the different sources of dietary 
fibre was strongly associated with a reduced risk of CHD.  Age adjusted RR for total 
MI in the Male Health Professionals cohort study [140] for those with intakes of 
dietary fibre in the highest quintile compared with the lowest quintile of intake was 
0.59 (95% CI, 0.46-0.76).  The association was strongest for fatal CHD.  As before 
 
39

---

cereal fibre was most strongly associated with a reduced risk of total MI (RR=0.71; 
95% CI, 0.55 – 0.91) for each 10g increase in fibre per day.  Men in the Finnish study 
[141] with intakes of total dietary fibre in the highest quintile had an RR for coronary 
death risk of 0.69 (95% CI, 0.54-0.88) compared with men in the lowest quintile.  
Water soluble fibre was slightly more strongly associated with reduced coronary 
death than water-insoluble fibre, and cereal fibre also had a stronger association than 
vegetable or fruit fibre.  These findings suggest that greater intake of foods rich in 
fibre (especially cereal fibre but also fruit and vegetable fibre) can substantially 
reduce the risk of CHD, and particularly coronary death).  
 
6.14  Two intervention studies looked at the effect of increased fruit and vegetable 
consumption on CHD risk factors [142], [54].  In the first study the intervention of a 
fat modified and fruit and vegetable enriched diet in conjunction with moderate 
physical activity was carried out over 3 years in subjects with coronary artery disease. 
The intervention group was found to have greater reduction in central obesity and 
greater decline in cardiac event rates and total mortality compared to the control 
group.  The second intervention was an RCT with a cross-over design in healthy 
Canadian adults.  The intervention consisted of a diet high in leafy and other low-
calorie vegetables, fruit and nuts for two week period.  After two weeks on the 
vegetable diet, lipid risk factors for CVD were significantly reduced by comparison 
with the control diet (see Table 2).  The reduction in total serum cholesterol was 34% 
to 49% greater than would be predicted by differences in dietary fat and cholesterol. 
The researchers concluded that a diet consisting largely of low-calorie vegetables and 
fruit and nuts markedly reduced lipid risk factors for cardiovascular disease. 
 
6.15  Other recent papers considered cardiovascular risk factors in young people, 
and characteristics of people with low CVD incidence [143], [144].  Children with a 
family history of hypertension had a greater cardiovascular reactivity to a video game 
stressor than those without, but this was not greater in black children than white 
although the latter had a higher vegetable intake at baseline and a higher potassium 
excretion [145]. In a comparison of relations among dietary pattern, nutrient intake, 
fitness, serum antioxidants and cardiolipoprotein profiles in young women (17.1±0.5 
years), fruit consumption is positively correlated with estimated VO2max and 
predicted VO2max is positively correlated with circulating beta-carotene and alpha-
tocopherol [146]. The researchers conclude that this study provides evidence that the 
positive associations of exercise and fruit consumption with cardiovascular health 
apply to female adolescents as well as to adults. 
 
6.16  Epidemiological studies suggest that significant reductions in the incidence of 
stroke, as with coronary heart disease, can be expected by reducing the prevalence or 
shifting the distribution of risk factors across the entire population.  Thus, identifying 
risk factors and intervention to control or modify them remains the most important 
means of further reducing the incidence and case fatality of stroke and coronary heart 
disease in developed countries. 
 
6.17  European and American recommendations for coronary heart disease 
prevention were discussed in a paper from the Department of Epidemiology, Harvard 
School of Public Health, Boston, USA [147].  In this paper they stated “Coronary 
patients should have professional support to stop smoking, eat a healthier diet (reduce 
the dietary intake of fat to 30% or less of total energy; saturated fat to no more than 
 
40

---

one third of total fat intake, cholesterol to less than 300mg per day; increase 
monounsaturated fat from both vegetables and marine sources; increase fruit and 
vegetables; achieve optimal weight, and become physically fitter through regular 
aerobic exercise.”  
 
6.18  In 1994, COMA’s Cardiovascular Review Group reported on the evidence for 
the roles of dietary factors in the development of cardiovascular disease [1].  At that 
time they reported fewer data on diet and cardiovascular disease are available for 
women, children and the elderly than for middle-aged men. More evidence has 
become available since that report, but there is still limited evidence for women, 
children and the elderly.  However, the Review Group concluded that the 
recommendations made in the report, which included an approximate 50% increase in 
the levels of vegetables and fruit (and also potatoes and bread) compared with current 
intakes, should apply to the population as a whole.  Many processes linked with the 
development of cardiovascular disease have their origins in childhood and COMA 
recommended that by the age of 5 years children should be consuming a diet 
consistent with the recommendations for adults.  However, they state that the dietary 
and other recommendations for adults do not apply to children below the age of 2 
years, for whom adequate energy intake for growth remains paramount.  Between the 
ages of 2 and 5 years a flexible approach to the timing and extent of dietary change 
should be taken.  
 
6.19  Conclusion: The data relating fruit and vegetable intake and CHD and CVD 
are growing and support a protective relationship between consumption of fruit and 
vegetables and reduced risk of CHD and CVD, especially ischemic stroke risk. Fruit 
and vegetable intake is generally associated with behavioural risk factors such as 
smoking and exercise, however, adjustment for behavioural and other risk factors do 
not explain the benefits of higher fruit and vegetable consumption, although 
adjustments for these factors often attenuate the observed associations. There appears 
to be a dose-response relationship which provides support for the recommendation to 
consume at least five portions of fruit and vegetables a day. 
 
 7 
Other chronic diseases  
 
7.1 
Cataracts occur when the lens of the eye is unable to function due to opacities.  
Lens opacities develop when proteins in the eye are damaged by photo-oxidation; 
these damaged proteins build-up, clump and precipitate.  First line defence systems 
protecting the lens from the initial oxidative stress are believed to be antioxidants 
such as vitamin C and carotenoids, found widely in fruit and vegetables. Three US 
cohort studies [148], [149], [150] have investigated relative risks for developing 
cataracts with consumption of fruit and vegetables or their micronutrient and fibre 
components. In the first study dietary sources of fibre and carotenoids were associated 
with lower risk for cataracts, particularly in men [148]. For the cohort of women aged 
45 to 67 years, high dietary carotenoid intake was found to be associated with lower 
risk for cataract extraction [149], and for the cohort of US male health professionals, a 
significantly lower risk of cataracts was found with higher intakes of beta-carotene 
rich foods [150].   
 
7.2 
The growing data base of evidence for the protective role of fruit and 
vegetable consumption and their components, e.g. carotenoids, for cataracts has lead 
 
41

---

one group of researchers to conclude that diets rich in fruit and vegetables may 
provide the least costly and most practicable means to delay cataract [151].  
 
7.3 
Protective effects of carotenoids have also been found for pathologies of the 
skin [152]. 
 
7.4 
Diabetes Mellitus.  The fibre in fruit and vegetables may also help in the 
management of diabetes.  It is known that selected soluble fibres can delay blood 
glucose absorption from the small intestine.  However, the long-term effect of dietary 
fibre on blood glucose control in individuals with diabetes is probably not significant.  
The data are more convincing for a protective effect of soluble fibre on the serum 
total and LDL-cholesterol levels in individuals with diabetes, particularly non-insulin 
dependent diabetes (NIDDM) [153]. 
 
7.5 
Chronic Obstructive Pulmonary Disease (COPD). Common examples of 
COPD are asthma and bronchitis, each of which affects large numbers of children and 
adults.  Cook et al [154] reported a beneficial effect on lung function in British school 
aged children who consumed fruit more than once a day compared to those who did 
not.  Salad and green vegetable consumption in this group of children (aged 8 to 11 
years) was also associated with a beneficial effect on lung function but the 
relationship was weaker than for fresh fruit. Data from three European cohort studies 
of men were used to evaluate the cross sectional association of dietary factors with 
pulmonary function [155].  A high intake of fruit and vegetables was positively 
associated with pulmonary function.  A high intake of vitamin C, vitamin E and beta-
carotene tended to be positively associated with pulmonary function before, but not 
after, adjustment for energy intake.  Associations of individual antioxidants with 
pulmonary function were not consistent across countries.  In a study of adults (18-69 
years), Strachan et al [156] found that winter fruit consumption had a protective effect 
on lung function compared with those who never drank fruit juice and only ate fresh 
fruit less than once a week.  In addition, low intake of dietary vitamin C have been 
associated with increased bronchitis and wheezing [157] and with pulmonary 
problems [158]. 
 
7.6 
In the Zutphen study cohort of men in The Netherlands, fruit intake, after 
controlling for age, smoking habits, body mass index and energy intake, was 
associated with reduced risk of developing COPD, including asthma [159]. A high 
vegetable intake was associated with a decreased risk of developing bronchitis and 
bronchial asthma in a study by La Vecchia et al [160]. 
 
7.7 
Obesity. Obesity is one of the most important avoidable risk factors for 
diseases such as CHD, stroke, hypertension, hypercholesterolaemia, NIDDM, 
gallstones, degenerative joint disease and sleep apnoea. The prevalence of obesity in 
the UK has reached epidemic proportions and is continuing to increase [161].  Fruits 
and vegetables can play an important role in the prevention of  and dietary treatment 
of obesity because of their low fat, low energy contributions to healthy eating. Despite 
the health risks associated with obesity, it has not proved easy to treat effectively 
[162]. Fruits and vegetables represent a food group which can be consumed in relative 
abundance even during a weight-loss phase of weight management [163]. 
 
 
42

---

7.8 
Bone health. New et al [164] recently identified intakes of nutrients found in 
abundance in fruit and vegetables to be positively associated with bone health.  They 
examined this finding further by considering axial and peripheral bone mass and 
markers of bone metabolism.  After controlling for present energy intake, higher 
intakes of calcium, magnesium, phophorus, potassium and alcohol were associated 
with higher bone mass.  Femoral neck bone mass density was higher in women 
consuming a high intake of fruit in their childhood (P<0.01).  They conclude that the 
bone density results confirm their previous research work (but at peripheral bone mass 
sites) and the findings associating bone resorption with dietary factors provide further 
intriguing evidence of a possible link between fruit and vegetable consumption and 
bone health. 
 
8 Vegetarian 
Diets 
 
8.1 
It is likely that vegetarians eat more portions or greater quantities of plant 
foods, including fruits and vegetables than non-vegetarians do. 
 
8.2 
Rimm et al [165] reviewed the beneficial and adverse effects of vegetarian 
diets in various medical conditions.  Soybean-protein diet, legumes, nuts and soluble 
fibre significantly decrease total cholesterol, low-density lipoprotein cholesterol and 
triglycerides.  Diets rich in fibre and complex carbohydrate, and restricted in fat, 
improve control of blood glucose concentration, lower insulin requirement and aid in 
weight control in diabetic patients.  An inverse association has been reported between 
nut, fruit, vegetable and fibre consumption, and risk of CHD.  Patients eating a 
vegetarian diet, with comprehensive lifestyle changes, have had reduced frequency, 
duration and severity of angina as well as regression of coronary atherosclerosis and 
improved coronary perfusion.  An inverse association between fruit and vegetables 
and stroke has been suggested.  Consumption of fruits and vegetables, especially 
spinach and greens, was associated with a lower risk of age-related ocular macular 
degeneration.  A decreased breast cancer risk has been associated with high intake of 
soy bean products.  The beneficial effects could be due to the diet (monounsaturated 
and polyunsaturated fatty acids, minerals, fibre, complex carbohydrate, antioxidant 
vitamins, flavonoids, folic acid and phytoestrogens) as well as the associated healthy 
lifestyle in vegetarians.  There are few adverse effects, mainly increased intestinal gas 
production and a small risk of vitamin B12 deficiency. 
 
8.3 
A pooled analysis of the data from five large prospective studies found that 
vegetarians had a 24% lower mortality from ischaemic heart disease than non-
vegetarians did.  This reduction in mortality was greater at younger ages, with a 45% 
reduction in risk of death from ischaemic heart disease before the age of 65 years.  
Vegetarian diets can differ in many ways from non-vegetarian groups, therefore it is 
impossible to draw any conclusions as to which aspect of the diet is protective.  One 
possible explanation for the lower mortality from ischaemic heart disease is that they 
have lower serum total cholesterol concentrations than non-vegetarians, largely 
because meat is a major source of saturated fatty acids but probably augmented by the 
hypocholestrolaemic effects of some plant foods.  It is also possible that some of the 
reduction in mortality from ischaemic heart disease in vegetarians is due to other 
mechanisms such as reduced oxidation of low density lipoprotein cholesterol or 
changes in blood clotting.  
 
 
43

---

8.4 
A comparison of vegetarians with non-vegetarians in a cohort of 34,192 
Californian Seventh-day Adventists found that intake of legumes was negatively 
associated with risk of colon cancer in non-vegetarians and risk of pancreatic cancer.  
Higher consumption of all fruit or dried fruit was associated with lower risks of lung, 
prostate, and pancreatic cancers.  Cross-sectional data suggest vegetarian Seventh-day 
Adventists have lower risks of diabetes mellitus, hypertension, and arthritis than non-
vegetarians [166].  This cannot be ascribed only to the absence of meat. 
 
8.5 
In summary, although observational studies of vegetarians cannot provide 
conclusive evidence, they do support the hypothesis that such a diet might lower the 
risk of coronary heart disease and some cancers. 
 
9 Mediterranean-type 
diet 
 
9.1 
The Mediterranean diet, high in fresh fruit and vegetables and fish, has been 
shown to produce favourable effects on blood lipids and also protects against 
oxidative stress. The preference of fresh fruit and vegetables in the Mediterranean diet 
will result in a higher consumption of raw foods, a higher intake of antioxidant 
micronutrients and a lower production of cooking-related oxidants. It is hypothesised 
and there is some evidence that such a diet will lead to less oxidative damage, one of 
the mechanisms leading to chronic diseases such as atherosclerosis and cancer.  
However, Parodi [167] has hypothesised that it is folate which is largely responsible 
for the protective effect for CHD risk of fruit and vegetables in the Mediterranean 
diet. 
 
9.2 
Although a large emphasis has been given to different components of the diet, 
attention has recently shifted to the diet as a whole.  The Mediterranean diet is able to 
modulate oxidative stress through complex mechanisms and not just the high 
antioxidant compound content. An intervention trial carried out on 605 coronary heart 
disease patients in the Lyon  Diet Heart Study, in which the patients were randomized 
to follow either a cardioprotective Mediterranean-type diet or a control diet close to 
the Step 1 American Heart Association prudent diet, found that the intakes of fruits, 
vegetables and cereals were significantly higher on the Mediterranean-type diet.  The 
results of the trial led the researchers to conclude that patients following a 
cardioprotective Mediterranean diet have a prolonged survival and may also be 
protected against cancer [168].  
 
10 Genetic 
factors 
 
10.1  Important diet-gene interactions may exist, as illustrated by differential 
responses to variation in folate status in those with methylenetetrahydrofolate 
reductase (MTHFR) polymorphisms.  
 
10.2  Molecular genetics has shown that accumulation of genetic changes is 
important in the development of colorectal cancer.  Mutations of at least four to five 
genes are required for the formation of a malignant tumour.  Environmental 
mutagenic factors may determine which susceptible individuals grow carcinomas.  
Environmental risk factors for colorectal cancer are found in a western diet, rich in 
fat, meat, and animal protein and low in fibre, fruit and vegetables.  The complex 
 
44

---

interrelations between food components make it difficult to define the precise role of 
specific food factors and diet-gene interactions.  
 
11 
Burden of diseases that have an inverse association with the consumption 
of fruits and vegetables. 
 
11.1  Cardiovascular disease (stroke and coronary heart disease) and cancer are the  
major causes of morbidity and mortality in the UK. A reduction in the incidence of 
these diseases and in the incidence of hypertension, obesity, diabetes mellitus, and 
cataracts should reduce the number of premature deaths and the cost of ill health to 
the NHS. 
 
11.2  Cancer is among the three leading causes of death in England and Wales at all 
ages except for pre-school children [169].  In July 1999 the government set a target: 
“to reduce the death rate from cancer in people under 75 years by at least a fifth 
(20%) by 2010 [compared with 1997] – saving up to 100,000 lives.” [170]. It has been 
estimated that diet might contribute to the development of one third of all cancers [8].  
The cancers which cause most deaths are lung, breast, colorectal, stomach and 
prostate.  Since the risk of developing all of these cancers has been shown, in 
epidemiological studies, to have some inverse association with the consumption of 
fruit and vegetables, an increase in the consumption of fruit and vegetables is likely to 
reduce the incidence of these cancers.   
 
11.3  COMA’s Working Group on Diet and Cancer concluded that, overall, the 
evidence is moderately consistent that higher vegetable consumption would reduce 
the risk of colorectal cancer, and that higher fruit and vegetable consumption would 
reduce the risk of gastric cancer.  There is weakly consistent evidence, based on fewer 
data, that higher fruit and vegetable consumption would reduce the risk of breast 
cancer.  These cancers combined represent about 18% of the cancer burden in men 
and about 30% of the cancer burden in women in the UK.  Even a small reduction in 
relative risk would have important public health benefits in terms of the absolute 
numbers affected. The World Cancer Research Fund (WCRF) estimated that 
increasing fruit and vegetable consumption could prevent 20% or more of all cases of 
cancer [56]. 
 
11.4  Every year heart disease and stroke kill 214,000 people. The mortality rates in 
the UK for cardiovascular diseases exceed those of other Western nations, and they 
remain a major cause of morbidity and premature death in men and women. In July 
1999 the government set a target: “to reduce the death rate from coronary heart 
disease and stroke in people under 75 years by at least two fifths (40%) by 2010.” 
[170].  If consuming one extra portion of fruit and vegetables per day lead to a 6% 
lower risk of ischaemic stroke [127], then the consumption of at least five portions per 
day should have a substantial impact on the reduction of ischaemic stroke. Similar 
risk reductions have also been seen for the risk of coronary heart disease and the 
increased consumption of fruit and vegetables. 
 
11.5  Obesity and overweight generate significant health service costs in England.   
The Office of Health Economics estimated in 1994 that obesity directly costs the NHS 
over £29M per year and indirectly costs the NHS over £165M if a portion of the cost 
of treating some of the conditions for which obesity is a risk factor is also included  
 
45

---

[171].  Any reduction in the incidence of obesity would reduce the burden of ill health 
resulting from obesity, including coronary heart disease and diabetes. 
 
11.6  Data from the USA show that the national health care expenditure for 1990 
totalled $666 billion of which 30% is said to be related to inappropriate diet.  
Identification of external factors that contribute to premature death would aid 
preventive efforts, improve the quality of life, and reduce health care costs. Dietary 
factors are associated with 5 of the 10 leading causes of death, including coronary 
heart disease, certain types of cancer, stroke, non-insulin dependent diabetes mellitus 
and atherosclerosis.   Even though genetic predisposition increases susceptible 
people’s risk for many of these chronic diseases, these conditions may be diminished 
or prevented by improvements in the American diet.  Recommendations for each of 
these chronic diseases include an increase in fruit and vegetable consumption which 
should lead to an enhancement of nutrient density.  This recommendation along with 
the recommendation to decrease dietary fat also impact on obesity and diminish the 
compounding of other disease states affected by excessive body weight. 
 
11.7  Policies to improve health require integration of nutrition needs with economic 
growth and development, agriculture and food production, processing, marketing, 
health care and education, and includes changing life styles and food choices.  
 
12 
Current and Further Research 
 
12.1  A major problem has been identifying the particular dietary components which 
predispose or protect individuals against cancer, coronary heart disease and stroke. 
The complexity of our diets and the multitude of potential dietary effects which may 
be important in disease development make this a fertile area for future study. 
Limitations in the precision and validity of traditional dietary intake measurements 
and limited use of biomarkers combined with narrow ranges of variations in dietary 
habits within single populations, have been the main reasons for the limited success in 
identifying more specific diet and disease links.   
 
12.2  For the development of cancer, more data are required from human 
experimental studies linking alterations in diet with known stages in carcinogenesis in 
the large bowel, and from large cohort studies which have collected biological 
samples in order that interaction between diet, biomarkers of diet, e.g. serum 
concentrations of beta-carotene and vitamin C, and different genotypes that may 
determine risk can be examined. Dietary data must include more detail regarding the 
types and amounts of fruits and vegetables either as individually itemised or grouped 
according to plant families.   
 
12.3  The European Prospective Investigation into Cancer and Nutrition (EPIC) is a 
multi-centre prospective cohort study designed to investigate the relation between 
diet, nutritional and metabolic characteristics, various lifestyle factors and the risk of 
cancer and is also being used to investigate the relation between diet and the risk of 
other diseases e.g. coronary heart disease and osteoporosis.  The study is based in 22 
collaborating centres in nine European countries (including the England) and includes 
populations characterised by large variations in dietary habits and cancer risk.  Data 
are collected on diet, physical activity, sexual maturation and reproductive history, 
lifetime consumption of alcohol and tobacco, previous and current illnesses and 
 
46

---

current medication.  It is planned to include about 400,000 middle-aged men and 
women by the end of 1997.  It is expected that about 23,000 cancer cases will be 
identified during the first 10 years follow-up [172]. Results from this study will make 
a major contribution to our knowledge about the relation between diet, cancer and 
other diseases in the UK and Europe. 
 
12.4  Interactions among lifestyle behaviours suggest the need for further research 
on the effects of overall dietary patterns on chronic disease risk, as well as on finding 
ways to encourage people to follow dietary recommendations as a whole rather than 
focusing on just one nutrient or food group.  
 
13 
Recommendations regarding intake of fruit and vegetables 
 
13.1  The Government’s advisory committee on food and nutrition, the Committee 
on Medical Aspects of Food and Nutrition Policy (COMA), has made several 
recommendations in the past decade on the consumption of fruit and vegetables.  The 
report of COMA’s Cardiovascular Review Group in 1994 recommended an 
approximate 50% increase in the levels of vegetables and fruit.  This was based on the 
average household consumption in 1992 (National Food Survey), which was an 
average of 1130g/person/week (161g/day or approximately 2 portions per day) for 
vegetables and 930g/person/week (133g/day or 1½ pieces fruit each day) for fruit and 
products. This recommendation equates to at least five portions of vegetables and fruit 
per person per day on average i.e. an average increase of one and a half portions per 
day. In addition they recommended approximate 50% increase in potatoes [1]. 
 
13.2  The Report of COMA’s Working Group on Diet and Cancer in 1998 
recommended that fruit and vegetable consumption in the UK should be increased [8].  
No target was set by this Working Group, since it found insufficient evidence to 
quantify the optimum level of fruit and vegetable consumption associated with the 
lowest risk of cancer.  It cited the advice of the Working Group on Nutritional 
Aspects of Cardiovascular Disease to increase the consumption by 50%, to at least 5 
portions per person per day on average, as a potentially achievable goal and likely to 
be conducive to better health in general and a lower risk of cancer in particular.  
 
13.3  The Report of COMA’s Working Group on the Nutrition of Elderly People 
published in 1992 recommended “Elderly people, in common with those of all ages, 
should be advised to eat more fresh vegetables, fruit, and whole grain cereals.” [173]. 
 
13.4  The official guidance on diet and health is set out in the booklet, Eight 
Guidelines for a Healthy Diet, published by the Health Education Authority in 
association with MAFF and the Department of Health (DH) [174]. Guideline 5 states 
“Eat plenty of fruit and vegetables” and the accompanying information states 
“Currently many people would benefit from increasing the amount of fruit and 
vegetables that they eat.  A balanced diet contains at least five portions of fruit and 
vegetables a day.” 
 
13.5  The Chief Medical Officers Ten Tips – your guide to better health includes 
“Follow a balanced diet with plenty of fruit and vegetables.” [170]. 
 
 
47

---

13.6  Internationally an intake of at least five portions (400g) of fruit and vegetables 
per day has become an established ‘healthy eating’ message [175].  In the USA, 
national health objectives for the year 2000 included recommendations to decrease 
intake of dietary fat and alcohol and increase intake of fruits, vegetables, and grains, 
in an effort to decrease cancer risk among the population [176] [177].  Rational public 
health recommendations in the USA include: 1. Five-A-Day serving of fruit and 
vegetables, a doubling of mean intake; 2. Systematic investigation of the covariates of 
extremes of fruit and vegetable intake; 3. Discouragement of beta-carotene 
supplement use, due to adverse effects in smokers and no evidence of benefit in non-
smokers [178]. 
 
13.7  In the UK health promotion agencies, producers and retailers have promoted 
‘at least 5-a-day’ message as a mechanism for helping to reduce the risk of coronary 
heart disease, strokes and many cancers [179] [174, 180] [56]. The same message has 
been promoted in a campaign by the Europe Against Cancer programme [181].   
 
13.8  The scientific basis for recommending an increase in fruit and vegetable 
consumption is now widely accepted, and there is a broad agreement on the target, i.e. 
‘at least five portions a day’ (or roughly five 80g portions).  However, for this to be a 
truly useful public health communication, it is necessary to clarify exactly which 
fruits and vegetables are included in the advice, and how much a portion is.  It is also 
helpful to know what the current level of consumption is and how this varies with age, 
sex, socio-economic group and region. 
 
 14 
Current intake of fruit and vegetables 
 
14.1  The National Food Survey (NFS) shows that over the decade, 1988 to 1998, 
there has been a decline in the consumption of vegetables (excluding potatoes) by 8% 
and the consumption of fresh fruit has increased by 20% [182].   
 
14.2  Total average vegetable consumption per person rose from 1950 (120g/day), 
to a high of 174g/day in 1986, but recently has gradually fallen to about 156g/day in 
1998.  This amounted to about 1090g/person/week, made up (weekly) of 246g fresh 
green vegetables (23% of total purchased), 486g other fresh vegetables (46%), 88g 
frozen vegetables (8%), 271g other processed non-frozen vegetables (mostly canned 
vegetables) (25%) [182]. Some vegetables will not be included in these figures, as 
they are ‘hidden’ in other foods and classified elsewhere, e.g. ready meals that contain 
vegetables but also contain meat or fish will be classified as meat or fish products.  
 
14.3  Total consumption of fruit has more than doubled since 1950 from 73g/day to 
155g/day for fresh fruit, fruit products including fruit juice, which is very similar to 
the total daily consumption of vegetables (156g/day). Average consumption of fresh 
fruit and fruit products (excluding juices) in 1998 was 716g (595g in 1988) and 70g 
(98g in 1988) per person per week, respectively.  And average consumption of fruit 
juice was 304ml per person per week in 1998 (211ml/person/week in 1988). Bananas 
were consumed in the largest amounts (197g/person/week), followed closely by 
apples (172g/person/week) and then by oranges and other citrus fruits 
(63g/person/week and 75g/person/week respectively).   
 
 
48

---

14.4  The figures from the NFS data include the non-edible parts of the fruit and 
vegetables and waste, which is said to constitute about 20% of the weight [183], 
hence the NFS overestimates the weight of vegetables and fruit consumed in the home 
by about 20% which gives an edible weight for the average daily consumption of 
vegetables and fruit of 250g in 1998.  Since a portion is roughly equivalent to 80g, 
this equates to about 3 portions a day.  However, this data from the NFS does not 
cover food eaten outside the home including school meals, and the NFS national 
average also includes very young children and the elderly who do not eat a lot of fruit. 
 
14.5  The NFS data averaged over years 1993-1995 showed a trend to lower 
consumption of fruit and vegetables from income group A (highest income group) to 
income group D (lowest income).  The consumption of fruit and vegetables is 
considerably lower in Scotland than England and Wales and in the North and North 
West regions of England compared to the other regions [8].   
 
Fruit consumption by income group (g/person/week) in 1998 [182] 
 
A1 A2 B  C  D  OAP 
Fresh 
fruit  156.0 133.5 92.5 72.8 58.2 101.2 
 
14.6  Whilst the NFS, an annual survey of household purchases measured over one 
week, can be used to estimate the average consumption of vegetables and fruit per 
person per day or week and to detect variations between different income groups and 
regions, and to follow trends over time, other surveys which measure individuals’ 
consumption of foods, e.g. the National Diet and Nutrition Survey (NDNS) (a 
Government programme of dietary surveys of four different population age groups), 
have to be used to estimate the number of portions of vegetables and fruit eaten daily 
by people of different ages and gender.  The survey of adults, The Dietary and 
Nutritional Survey of Adults [184], carried out in 1986/87, is the only national survey 
of British adults for which data on fruit and vegetable consumption by individual 
adults is available. [Note: The NDNS programme has a new survey of adults aged 18 
to 64 years beginning later in 2000 and will take place over 12 months.]  [Further 
analyses are being carried out on this survey data to determine the average daily 
consumption of total fruit and vegetables.] The most commonly consumed vegetables 
(based on a 7-day recording period) were peas, eaten by 72% of the sample (sample 
n=2197), with carrots and leafy green vegetables each eaten by at least 60% of the 
sample.  The mean intake over the 7-day recording period of peas by consumers was 
155g, of carrots was 120g and of leafy green vegetables was179g.   However, a 
smaller proportion of younger people than older adults ate carrots (52% people aged 
16-24 years cf. 65% people aged 50-64 years) and leafy green vegetables (49% people 
aged 16 to 24 years cf. 71% people aged 50 to 64 years) during the seven-day 
recording period, whereas a larger proportion of younger people ate baked beans 
(56% 16-24 years cf. 38% 50-64 years). 
 
14.7  The survey of adults was used to obtain estimates of the average frequency of 
consumption by the adult population (aged 16 to 64 years) per week.  Men and 
women ate vegetables excluding potatoes eight times during the 7-day recording 
period.  Men ate five servings and women ate seven servings of fruit (including fruit 
juice) during the 7-day recording period [8].  Whilst we know that there have been 
some changes in dietary habits since 1986/87, we also know from the NFS that the 
 
49

---

consumption of vegetables has declined by 8% and the consumption of fruit has 
increased by 20% but it is unlikely that the number of portions eaten weekly will have 
increased greatly and certainly not by the 50%  recommended by COMA in the 
Report of the Cardiovascular Review Group [1].  The latest survey, the NDNS of 
adults aged 18 to 64 years, will be carried out over one year from 2000/1 and the data 
from this survey will form the basis for evaluating the effectiveness of population 
dietary interventions to increase the consumption of fruit and vegetables in adults.  
  
14.8  Estimates of the number of portions of fruit and vegetables consumed daily or 
over the 4- or 7-day recording periods are being made using data from the other three 
age groups from the NDNS, children aged 1½ - 4½ years, young people aged 4 to 18 
years and adults aged 65 years or over, and will be added as Annex  6. Information on 
the consumption of fruit, fruit juice and carrots by children participating in the NDNS 
of young people aged 4 to 18 years and an assessment of the increased tonnage for 
apples to enable each child to eat an apple a day is included at Annex 1. 
 
14.9  The report Health in England 1998: Investigating the links between social 
inequalities and health [185] explored the eating habits and knowledge of the 
population about what constitutes a healthy diet.  Lower proportions of men than 
women usually ate fruit or vegetables daily (59% compared with 74%).  Men were 
less likely than women to mention the need to ‘eat lots of fruit, vegetables or salad’ 
(62% compared with 74%).  Age was the characteristic most closely associated with 
diet quality.  Those in the younger and older age-groups were more likely than those 
in the middle age-groups to have a ‘less healthy diet’, e.g. among those aged 16-24, 
nearly a third of the men and nearly a quarter of the women, had a ‘less healthy diet’.  
Diet quality improved with increasing age until the 55-64 age-group, but declined in 
the two oldest age-groups.  Diet quality decreased with decreasing levels of household 
income, qualifications, and social class.  In addition, it was relatively lower among the 
widowed, divorced and separated, the unemployed, and those in local authority 
housing.  It was also relatively lower among men in privately rented accommodation, 
among lone mothers and, to a lesser extent, among women who were economically 
inactive and women living alone. 
  
14.10  The Welsh cohort study, “The Caerphilly Study”, of men aged 45-69 years 
included baseline (1979-1983) measurements of vegetable and fruit consumption. 
Mean consumption of vegetables and fruit at baseline was 118g/day and 83g/day 
respectively [79] which equates to less than 3 portions per day. 
 
14.11  A survey into Scottish eating patterns and attitudes towards fruit and 
vegetables found that less than half of the sample were meeting the WHO 
recommendation of 400g, or five portions of fruit and vegetables (excluding potatoes) 
per day [186]. 
 
14.12  Recent information from the Spanish Group of EPIC shows that the average 
consumption of fruit and vegetables in healthy adults in Spain is considerably higher 
(men: mean daily consumption of vegetables and fruits was 274g (3.4 servings) and 
348g (4.4 servings) respectively; women: 244g (3.1 servings) and 349g (4.4 servings) 
vegetables and fruits, respectively) than in most European countries and the United 
States [187]. This complies with what is considered to be the Mediterranean diet. 
 
 
50

---

14.13  Information on the number of portions of fruit and vegetables consumed 
among US adults prior to participation in the 5 A Day research trials in seven study 
centres was published recently by Thompson et al [188].  Results indicate an overall 
mean intake of 3.6 daily servings of fruits and vegetables.  Significant differences in 
mean daily servings were found among the regional study centres (low of 3.0 to high 
of 4.1).  Only 17% of respondents ate 5 or more servings of fruits and vegetables per 
day.   
 
15 
Which ‘fruit and vegetables’ and how much is a ‘portion’? 
 
15.1 
Which fruit and vegetables? 
The term “fruit(s) and vegetables” includes all fresh, frozen, canned and dried fruits 
and vegetables except potatoes (for explanation see paragraph 2.2). Beans and pulses 
are included but not nuts. One glass of fruit juice per day, but not fruit squashes, may 
also be included as one portion.  
 
Fruits and vegetables which count towards the 5 portions a day. 
Fruits Vegetables 
All fresh, frozen and canned  All fresh, frozen and 
fruits. 
canned vegetables.  
Fruit juice is included but 
Beans and pulses. 
only one glass 
Not potatoes 
(approximately 150ml) 
Composite dishes such as 
should be counted towards 
meat and vegetable stews 
the ‘5 portions a day’. 
count as long as they 
Dried fruit is included but 
contain enough vegetables 
preferable only one portion 
to constitute a portion. 
should be included in the 5-
a-day. 
Made up fruit dishes e.g. 
apple crumble can count as 
half or one portion 
depending on the amount of 
fruit used. 
 
 
 
15.2  The advice to eat at least five portions of fruit and vegetables per day is based 
on the broad range of health and nutritional properties of fruits and vegetables.  It is 
important to emphasise the use of a wide variety of both fruits and vegetables, and in 
particular to include more than once a week vegetables from the group known as 
Brassicas or cruciferous vegetables, e.g. cabbage, broccoli, cauliflower, Brussels 
sprouts; vegetables rich in carotenoids (yellow/orange/red vegetables), e.g. carrots, 
tomatoes, peppers; and citrus fruits e.g. oranges, satsumas, and mandarins..  Any 
advice on the types of fruit and vegetables to eat, should take into account the 
availability, social and cultural uses of fruit and vegetables, and should encourage 
everyone, but especially young people, to try ‘new’ vegetables and fruits. 
 
15.3  Precise targets e.g. 5 portions fruit and vegetables a day, might be 
misunderstood as biologically optimal, but targets are usually a compromise between 
 
51

---

biological advantage and social, cultural or economic acceptability [1].  This target is 
seen as moderate and achievable. 
 
15.4  A distinction can be made between a ‘portion’ and a ‘serving’.  A portion is a 
predefined ‘discrete’ amount of food, e.g. an apple.  A serving is the amount an 
individual serves himself or herself, or is served, and is variable [183],[189]. Surveys 
record either the weight of individual fruit and vegetables eaten (i.e. servings), or the 
frequency of consumption (the number of times each day or week), but rarely both. 
MAFF publishes average serving sizes recorded in national dietary surveys [190].  By 
comparing these average servings with the total weights of foods consumed, it is 
possible to estimate the number of portions eaten.  This calculation suggests that 
nutrition information to ‘eat five’, which assumes a mean portion size of around 80g 
(2½- 3 oz), ties in well with average serving sizes used by households in Britain.  
 
Some practical examples of adult portions for consumers: 
Fruit and Vegetable Type 
Portion/Size 
Vegetables, raw, cooked, frozen or 
2 tablespoons 
canned 
Salad 
1 dessert bowlful 
Grapefruit/avocado pear 
½ fruit 
Apples, bananas, oranges, and other 
1 fruit 
citrus fruit 
Plums and similar sized fruit 
2 fruit 
Grapes, cherries and berries 
1 cupful 
Fresh fruit salad, stewed or canned fruit 
2-3 tablespoonfuls (incl. A little juice or 
syrup) 
Dried fruit (raisins, apricots etc.) 
½-1 tablespoonful 
Fruit juice 
1 glass (150 ml) 
 
Note: Children aged 5 years and over should also eat at least five portions per day, but 
portions may be smaller. 
 
15.5  The use of food frequency questionnaires (FFQ) to measure the consumption 
of fruits and vegetables has been shown to vary the estimated intake depending on the 
number of questions asked about fruit and vegetable and reported consumption.  More 
questions led to exaggerate intakes [191]. A recent UK study by Cox et al [192]; [193]  
using a FFQ listing 104 items also found that such a FFQ appeared to exaggerate 
consumption relative to other data but provided a reasonable measure of variation in 
intakes.  Therefore caution must be exercised when using this method to estimate fruit 
and vegetable consumption.   
 
15.6  The same study [193] was able to compare the number of portions of fruit and 
vegetables recorded with the measured weights of all fruits and vegetables consumed,  
and found (assuming an average portion weight of 80g) lower values were recorded 
by the portion measures.  There were highly significant correlations between the two 
measures. Data from the same study did suggest that the defined portion measure and 
five fruit and vegetable portions/day target is a reasonable way of correctly classifying 
individuals with regard to meeting the ‘greater than 400g/d’ recommendation, and is 
particularly capable of highlighting those consuming less than this target amount.  
This may be due to the fact that the ‘5-a-day’ concept measures discrete portions 
 
52

---

whilst ‘hidden’ fruits and vegetables within composite meals, e.g. vegetables added to 
meat stews, may account for the ‘extra’ quantity recorded by the weighed method. 
 
16 Dietary 
Interventions 
 
16.1  As seen above (para. 14.4), national average intakes equate roughly to about 3 
portions of fruit and vegetables each day compared to the recommended amount of at 
least 5 portions each day. It is therefore important that interventions are developed 
which are effective in increasing intakes among the population as a whole. However 
there are also specific groups within the population for whom there are particular 
concerns, for example school aged children, young men and women (18 to 30 years) 
and people on low incomes, as well as people in some regions of England.   
 
16.2  Summary of interventions to increase fruit and vegetable consumption 
 
16.2.1  MAFF and DH, along with other funding bodies, for example, the British 
Heart Foundation and the WCRF are currently funding some UK research projects to 
determine effective interventions to increase the consumption of fruit and vegetables 
or to explore the opportunities and barriers to increasing the consumption of fruit and 
vegetables.  Some of the MAFF funded projects have reported and are summarised in 
Annex 2.  Those being funded by the Department of Health under the Nutrition 
Research programme are not completed and are listed in Annex 3 along with ongoing 
MAFF/FSA projects.   
 
16.2.2  Table 3 lists recent publications which describe dietary interventions or studies 
to identify opportunities and barriers to increasing the consumption of fruit and 
vegetables along with the main findings. The studies tend to fall into a broad range of 
categories which are listed below. 
 
1.    Baseline data collection or validation of assessment tools. 
2.  Assessment of whether socio-demographic variables predict consumption. 
3.  Assessment of the influence of psychosocial variables. 
4.  Identifying behavioural models which can be used to influence dietary change. 
5.  Identifying suitable approaches to increase fruit and vegetable consumption. 
6.  Establishing the feasibility of increasing and maintaining fruit and vegetable 
consumption. 
7.  Assessing other influences on dietary habits. 
 
16.2.3  Dietary interventions have been carried out in school settings with primary 
and secondary age children, in workplace settings, in supermarkets and cafeterias, in 
primary health care settings, in churches and community centres and by Cancer 
Information telephone centres. Depending on the setting of the intervention and 
sometimes using behavioural models, investigators have assessed the most effective 
means of increasing fruit and vegetable consumption.  
 
16.2.4  The main findings from the MAFF funded and other recent studies are 
summarised here: 
Barriers: 
•  Over estimation of current consumption of fruit and vegetables by children 
and adults; 
 
53

---

•  Lack of understanding in young children as to what counts as vegetables 
and fruits;  
•  Lack of experience in cooking vegetables and unwillingness to try more 
and new vegetables especially in families who just ‘coped’ with feeding 
rather than planned meals; 
•  Time constraints and inconvenience especially with respect to preparing 
and cooking vegetables;  
•  Shopping is more difficult and, for some cost, was seen as a barrier; 
•  Lack of availability at work, at school and when eating out; 
•  Lack of social pressure to increase intake (women feel more social 
pressure than men). 
Opportunities: 
•  More advanced stages of change were more likely to be associated with 
knowledge of the 5-a-day recommendation, self-efficacy, and greater fruit 
and vegetable consumption. 
 
Some intervention methods found to be successful in increasing intake: 
•  Practical advice; live demonstrations; hands-on practice; tastings; build on 
familiar ingredients; ‘convenience’ preparation i.e. quick recipes or use of 
frozen and canned vegetables and fruits; and targeting specific groups.   
•  Emphasis on portions, e.g. what is ‘a portion’; practical means of 
increasing the number of portions per day tailored to intervention group; 
and self-monitoring using diaries. (Fruits were seen as more easily 
incorporated into daily eating patterns than vegetables.) 
•  Video presentations and telephone counselling were almost as successful 
in increasing adults’ intake as personal contact. 
•  A series of incentive videos using peer models known as the “Food 
Dudes” and rewards have been shown to increase fruit and vegetable 
intakes in very young and primary school children. 
 
Implications for interventions: 
•  Need education to increase knowledge of the 5-a-day fruit and vegetable 
consumption message and the evidence that increased fruit and vegetable 
consumption leads to reduced risk of several diseases, especially coronary 
heart disease and cancer; 
•  Consumers need to become more aware of their actual consumption rather 
than their perceived consumption; 
•  Portions sizes need defining more clearly; 
•  Vegetables need greater targeting than fruits; 
•  Collaboration between producers, retailers, caterers and employers to 
ensure better accessibility.  
 
16.2.4  MAFF funded UK studies are taking place to establish whether a novel whole-
school intervention, fruit tuck shops, peer modelling videos and reward system and art 
therapy in primary schools increase fruit and vegetable consumption, DH are funding 
an after school ‘Food Club’ project in secondary schools.  Some of these are due to 
report later this year.  DH funded projects to establish whether counselling in primary 
health care is effective in increasing consumption of fruit and vegetables will be 
completed by the end of 2001 (See Annex 3). 
 
54

---

 
16.2.5  The Health Education Authority (HEA) commissioned a series of reviews of 
effectiveness of healthy eating promotions which were published in 1997 and 1998. 
These included the effectiveness of health promotion dietary interventions in the 
general population [194], interventions to promote healthy eating in elderly people 
living in the community [195]; in pre-school children aged 1-5 years [196]; in 
pregnant women and women of childbearing age [197]; in infants under one year of 
age [198]: and in people from minority ethnic groups [199].  A summary of the 
conclusions is attached in Appendix 1.   
 
16.3  The HEA have produced a paper (see Annex 4) for the Department of Health 
that provides a brief overview of the current (2000) information held by the HEA 
(now the HDA), concerning the effectiveness of interventions to increase fruit and 
vegetable intakes. It largely draws on the following sources: 
 
i) 
The initial draft of the content for the ‘Social Exclusion and Food’ section of  
Evidence Base 2000, produced by Lynn Stockley on behalf of the HEA.   
ii) 
A previous paper, also produced by Lynn Stockley, which made proposals for 
ways in which the HEA could support initiatives in the short and long term 
which may increase fruit and vegetable intakes, as part of its food and 
nutrition contract with the Department of Health (Stockley, unpublished).  
iii) 
The HEA’s review of effectiveness of interventions to promote healthy eating 
in the general population [194]. 
 
16.3.1  Since the second of the papers referred to above have been produced, various 
developments have taken place. These include: the announcement of the replacement 
of the Health Education Authority  (HEA) by the Health Development Agency 
(HDA); the development of the Food Standards Agency (FSA);   the  publication of 
the Social Exclusion Unit  report on improving shopping access for people living in 
deprived neighbourhoods; and the consultation on  proposals  to review the Welfare 
Food Scheme.  
 
16.3.2  The final section of the HEA paper makes recommendations of steps which 
could be taken to increase fruit and vegetable intakes and these are summarised 
below: 
 
To address circumstantial barriers: 
 
•  Given that people on lower incomes tend to be disproportionately affected by the 
Common Agricultural Policy’s (CAP) influence on the price of fruit and 
vegetables, we recommend that the Government and FSA make it an early priority 
to press the European Union  for a health equity assessment of the CAP and  that 
means of ensuring access to affordable supplies of fruit and vegetables, be made a 
priority in the current reforms of the Common Agricultural Policy.          
 
•  In response to the review of the Welfare Food Scheme, the HEA has suggested 
that fruit and vegetables should be made available to those eligible for the scheme 
using a  token system in the same way that milk and infant formula are currently 
made available. This would help to address issues of affordability for families on 
low income. It may also improve access in areas where good quality fruit and 
 
55

---

vegetables are not currently available in local shops by creating a ‘supply and 
demand’ situation. Retailers would have some incentive to stock fruit and 
vegetables if they were assured some turnover from members of the local 
community eligible for the scheme. Food co-operatives would need to be able to 
register for such a scheme in order to ensure that they do not lose trade to other 
local retailers.   We would strongly encourage the piloting of such a scheme by the 
Department of Health Welfare Foods Unit.    
.  
•  We recommend that schools be given information about accessing the free fruit  
available under the CAP intervention scheme and that ways are identified of  
developing a system of  distributing the fruit between schools so that it is received 
in manageable amounts.  There is currently much support for such a scheme. A 
similar recommendation was made by the Education and Employment Committee 
in the report of its recent inquiry into school lunches, and following letters 
proposing such a scheme to all MP’s from the National Heart Forum,  an Early 
Day Motion has been tabled in the House of Commons.             
  
•  To address the broader determinants of fruit and vegetable consumption,  
consideration should be given to ways in which Government initiatives other than 
those primarily concerned with health, can support initiatives to increase 
consumption. Within this, initiatives such as Sure Start, Education Action Zones, 
Employment Action Zones, the New Deal and Neighbourhood Renewal schemes 
should be considered.  For example the development of community led local retail 
strategies might form part of a neighbourhood renewal scheme with a  particular 
emphasis on increasing access to fruit and vegetables.  
 
•  The findings of the Social Exclusion Unit’s Report  on improving shopping access 
for people living in deprived neighbourhoods should be considered with reference 
to increasing access to fruit and vegetables. In particular consideration should be 
given to improving support for smaller retailers and to how the larger retailers 
may be able to contribute to improving access to fruit and vegetables in these 
areas.  However it should be recognised that in some cases it may be more 
appropriate to take ‘people to the shops’, rather than ‘shops to the people’ and as 
such the provision of affordable transport may be a more important factor in some 
communities.   
 
•  In areas where it is appropriate to take ‘shops to the people’ methods of increasing 
access to affordable fruit and vegetables with local communities could include the 
initiation of small grant schemes to assist in the set up costs of establishing local 
food co-operatives. This could be funded by the Public Health Development Fund 
and could be developed as part of local Health Improvement Programmes and 
Local Authority Community Plans. Healthy Living Centres be involved in the 
development of such food co-operative schemes.  In rural areas Local Authorities 
could develop Village Shop assistance schemes, which help to ensure the 
continuing existence of such shops and that they provide fruit and vegetables.  
 
•  Increasing fruit and vegetable consumption should be reflected as a priority in the 
Department of Health’s National Priorities Guidance,  the CHD National Service 
Framework, Our Healthier Nation key area’s contracts and the policy direction for 
Healthy Living Centres.      
 
56

---

 
 
To address attitudinal barriers:  
 
•  We believe an early priority for the Food Standards Agency should be to develop, 
in partnership with the Department of Health, the HDA and industry, information 
campaigns for the public and professionals which give clear messages concerning: 
the benefits of eating more fruit and vegetables; the recommended intake, the fact 
that most people are not eating enough and need to virtually double their intake; 
the fact that frozen and canned fruit and vegetables count; a clear guide as to what 
constitutes a portion of fruit and vegetables; and opportunities for incorporating 
these into the diet. This should be done in the context of an overall balanced diet 
and use the Balance of Good Health to illustrate the proportion of the diet which 
should be made up by fruit and vegetables.      
  
•  To address barriers such as cooking skills for adults, we suggest that consideration 
should be given as to how a programme similar to the Expanded Food and 
Nutrition Education (EFNEP) programme [200] in the USA could be developed in 
the UK with a particular focus on increasing fruit and vegetable intakes in low 
income communities. (Note: EFNEP aims to assist low-income families to acquire 
the knowledge, attitudes, and skills needed to adopt nutritionally sound diets, thus 
contributing to dietary improvements for the family.)  People from the local 
community could be trained to work with clients to develop their cooking skills 
together with shopping and menu planning skills. A programme of this type could 
be piloted within Health Action Zones, with the evaluation of its effectiveness 
disseminated on a national basis.   
 
•  Home economists should be used to ‘train the trainers’ rather than chefs as have 
been used in some other interventions. This is because home economists possess 
both practical skills in these areas together with nutrition knowledge and could be 
considered  to be an under utilised resource. The feasibility of developing such an 
initiative should be discussed with professional bodies such as the Institute of 
Home Economics, the UK Federation of  Home Economics and the National 
Association of  Teachers of Home Economics (NATHE).   
 
•  To help address lack of cooking skills in future generations, we strongly 
recommend that cooking and budgeting skills are reintroduced into the national 
curriculum. This should be urgently addressed by the Department for Education  
and Employment (DFEE).  Cooking skills and supporting studies such as nutrition 
are an essential component of a whole school approach to promoting healthy 
eating and while initiatives such as after school cooking and holiday cooking 
clubs are to be applauded, it remains to be seen how effective they are in 
improving children’s ability to cook everyday healthy meals, using good food 
hygiene practices and to learn to budget successfully. Such clubs should be 
evaluated for their effectiveness. [Note: DH are evaluating an after school food 
club.] 
 
16.4  The National Heart Forum published a report titled ‘At Least Five a Day. 
Strategies to increase vegetable and fruit consumption’ [201] which included a paper 
by Foster and McColl on interventions to change attitudes and improve access [202].  
 
57

---

They concluded, “Research indicates that it is possible to increase fruit and vegetable 
intakes to achieve the current recommendations.  Clear and simple strategies need to 
be identified and promoted, preferably with quantitative guidelines.  In addition, 
research suggests that strategies should be carefully targeted and should address 
beliefs, readiness to change and barriers to change of the target group.”  
 
16.5. Models of behavioural change 
 
16.5.1  Several intervention studies have used models of behavioural change.  
Two commonly used models include the Health Belief Model and the Trans-
theoretical Model for health behaviour.  The Health belief model suggests that 
individuals may be more ready to eat fruit and vegetables if they are aware of the 
benefits of eating them, and consider themselves to be at risk of, say CHD. One study 
which used this model is from Washington USA [203]. Analysis revealed that 
attitudes only accounted for 16% of variance in fruit and vegetable consumption. 
Barriers to fruit and vegetable consumption were the largest component of the 
variability. The Trans-theoretical model for health behaviour suggests that people 
experience different stages of readiness for change. These include pre-contemplation, 
preparation, action and maintenance.  
 
16.5.2  People who are in the pre-contemplation phase – that is have no intention to 
make any changes, need to be differently targeted than those who are ready to change.  
A study in North Carolina [204] illustrated this by the failure of a nutrition education 
program to increase fruit and vegetable consumption, because it was not tailored to 
the individual’s stage of change and health beliefs.  
 
16.5.3  Anderson et al (1998) [205] uses the model of the theory of planned 
behaviour. This is an extension of the theory of reasoned action which suggests that 
people form intentions to perform or not perform behaviours, depending on the 
perceived benefits and costs. These are themselves based on their own attitude to the 
behaviour, and social norms. The theory of planned behaviour incorporates ‘perceived 
behavioural control’ as a predictor in behavioural intention.  
 
16.5.4  Baranowski et al (1999) provides a review of these models [206] and 
Campbell et al (1998 and 1999) look at the stages of change model in a study on 
African American church members [207], [208].  
 
16.6  USA 5 A Day campaign 
 
16.6.1  In 1988, the Californian Department of Health Services embarked upon an 
innovative social marketing program to increase fruit and vegetable consumption. The 
‘5 A Day—for Better Health’ Campaign had several distinctive features, including its 
simple, positive, behaviour-specific message to eat 5 servings of fruits and vegetables 
every day as part of a low-fat, high fibre diet. It used the mass media; formed  
partnerships between the state health department and the produce and supermarket 
industries; and made extensive use of point-of-purchase messages. In 1991 the 
campaign was adopted as a national initiative by the National Cancer Institute (NCI) 
and the Produce for Better Health Foundation.  By 1994, over 700 industry 
organisations and 48 states, territories, and the District of Columbia were licensed to 
participate. The ‘5 a Day’ Campaign appears to have raised public awareness that 
 
58

---

fruits and vegetables help reduce cancer risk, increased fruit and vegetable 
consumption in major population segments, and created an ongoing partnership 
between public health and agribusiness that has allowed extension of the campaign to 
other populations segments, namely children and ethnic groups. The program is being 
evaluated nationally but, whilst the evaluation of some of the individual projects have 
been published, the outcome of the evaluation of the program as a whole has not yet 
been published (Farrell, personal communication).   
 
16.6.2  The outcomes of two of the programmes with children to increase fruit and 
vegetable consumption: Gimme 5 [209] and 5-a-Day Power Plus [210], both 
multicomponent school-based programmes resulted in increased fruit consumption.  
The intervention in the elementary schools (fourth and fifth grades) consisted of 
behavioural curricula in classrooms, parental involvement, school food service 
changes, and industry support and involvement.  Changes in consumption were 
monitored using lunchtime observations and 24-hour food recalls and psychosocial 
factors were measured using parent telephone survey and a health behaviour 
questionnaire. This intervention increased lunchtime fruit consumption and combined 
fruit and vegetable consumption, lunchtime vegetable consumption among girls, and 
daily fruit consumption as well as the proportion of total daily calories attributable to 
fruits and vegetables.  The researchers concluded that greater involvement of parents 
and more attention to increasing vegetable consumption, especially among boys, 
remain challenges in future intervention research [210].  The intervention in the High 
schools followed the students for 3 years from 9th to 12th Grades and comprised a 
media campaign, classroom workshops, school meal modification, and parental 
support.  Three primary end-points were evaluated at the school level: 1) increased 
awareness; 2) increased positive attitudes and knowledge toward eating at least five 
daily servings of fruit and vegetables; and 3) increased daily consumption of fruit and 
vegetables. Usual daily servings of fruit and vegetables increased by 14% in the 
intervention compared to the control group in the first 3 years.  At follow-up 
consumption within the control group also increased resulting in no significant 
difference between groups.  Intervention group knowledge scores and awareness 
indicators were significantly higher than those of the control group.  The researchers 
concluded that high school students can be influenced by positive media messages 
relative to that age group, increased exposure to tasty products and minimal classroom 
activity [209]. 
 
 
16.6.3  Two work place programmes from the USA 5-a-day program have also 
reported.  In the first, a randomized trial, peer education was tested for effectiveness 
at increasing fruit and vegetable intake among lower socio-economic, multicultural 
labour and trades employees [211].  Following a baseline survey the employees 
received an 18-month intervention program through standard communication 
channels (e.g. workplace mail, cafeteria promotions, and speakers).  Forty-one 
matched social networks (cliques) of employees were pair matched on intake and one 
clique per pair was randomly assigned to the peer education intervention. Employees 
who were central in the communication flow of the peer intervention cliques served as 
peer educators during the last 9 months of the intervention program.  Fruit and 
vegetable intake was measured with 24-hour intake recall and with FFQ at baseline, 
after 18 months intervention and again after 6-month follow-up.  The intervention 
group increased their fruit and vegetable intake by 0.77 total servings/day based on 
 
59

---

24-hour recall and 0.46 servings/d based on the FFQ after 18 months intervention.  
The effect on the total number of servings persisted at the 6-month follow-up when 
measured by the 24-hour recall (increase of 0.41 servings/d) but not when measured 
by the FFQ (decrease of 0.04 servings/d).  The researchers conclude that peer 
education appears to be an effective means of achieving an increase in fruit and 
vegetable intake among lower socio-economic adult employees. 
 
16.6.4  The second work place study, the Treatwell 5-a-Day study, involved families 
in promoting increased consumption of fruit and vegetables [212].  Twenty-two 
worksites were randomly assigned to three groups: 1) a minimal intervention control 
group; 2) a work site intervention, and 3) a work site-plus-family intervention.  The 
interventions used community-organizing strategies and were structured to target 
multiple levels of influence, following socio-ecological model.  Data were collected 
by self-administered employee surveys before and after the intervention.  A process 
tracking system was used to document intervention delivery.  Total fruit and 
vegetable consumption increased by 19% in the work site-plus-family group, 7% in 
the work site intervention group, and 0% in the control group.  These changes reflect 
a one half serving increase among the work site-plus-family group compared with the 
control group.  The researchers concluded that the work site-plus-family intervention 
was more successful in increasing fruit and vegetable consumption than was the work 
site intervention and suggest that work site interventions involving family members 
appear to be a promising strategy for influencing workers’ dietary habits. 
 
16.6.5  Other interventions evaluated in the above 5-a-day program are community-
based interventions aimed at low-income groups, African Americans and women, 
infants and children served by the Special Supplemental Nutrition Program (WIC).  
The North Carolina Black Churches United for Better Health Project was funded 
under this programme because of the high percentage of African Americans who 
attend church and the importance of this institution within the African American 
community.  50 churches were pair matched and randomly assigned to either 
intervention or delayed intervention.  A multi-component intervention was conducted 
over approximately 20 months.  Data were collected at baseline and at 2-year follow-
up interviews [213].  At the 2-year follow-up the intervention group consumed 0.85 
servings/d more than the delayed intervention group (the target was 0.5 servings/d 
increase).  The largest increases were observed in persons 66 years or older (1 
serving/d increase), those with education beyond high school (0.92 servings/d 
increase), those widowed or divorced (0.96 servings/d) and those attending church 
frequently (1.3 servings/d).  The least improvement occurred among those aged 18 to 
37 years and those who were single.  The WIC intervention comprised a multi-
component intervention program using a randomized crossover design.  Data was 
collected at baseline, 2 months post intervention, and 1 year later.  At the two-month 
post-intervention, mean daily consumption had increased by 0.56 servings in 
intervention participants and 0.13 in control participants.  Intervention participants 
also showed greater change in stages of change, knowledge, attitudes, and self-
efficacy.  Changes in consumption were closely related to the number of nutrition 
sessions attended, baseline stage of change, race, and education.  One year later, mean 
consumption had increased by an additional 0.27 servings/d in both intervention and 
control participants.  The researchers concluded that dietary change can be achieved 
and sustained in this hard-to-reach, low-income population.  However, many 
obstacles must be overcome to achieve such changes [214]. 
 
60

---

 
16.6.6  The USA ‘5 A Day’ campaign uses a variety of intervention methods, usually 
multi-component, and a variety of evaluation tools to measure changes in knowledge 
and intake.  Since there has not yet been an evaluation of the whole programme it is 
not possible to determine the best method(s)/tool(s) to evalute the outcome of 
individual interventions. A reference list of the publications arising out of the USA 5-
a-Day programme to date is attached at Annex 5 and some of the individual projects 
are included in Table 3. 
 
17 
Conclusion 
 
17.1  The benefits of eating fruits and vegetables appear to be beyond dispute (see 
Chapters 5, 6, 7, 8 and 9), especially since these foods provide essential nutrients not 
readily available from other sources.  But eating more fruits and vegetables should be 
linked to other healthy dietary and lifestyle patterns.  
 
17.2  The official guidance on diet and health is set out in the booklet, Eight 
Guidelines for a Healthy Diet, published by the Health Education Authority in 
association with MAFF and the Department of Health (DH) [174]. Guideline 5 states 
“Eat plenty of fruit and vegetables” and the accompanying information states 
“Currently many people would benefit from increasing the amount of fruit and 
vegetables that they eat.  A balanced diet contains at least five portions of fruit and 
vegetables a day.” 
 
17.3  The Chief Medical Officers Ten Tips – your guide to better health includes 
“Follow a balanced diet with plenty of fruit and vegetables.” [170]. 
 
17.4  The current mean consumption of fruit and vegetables per person per day in 
the UK is three portions (Chapter 14), although this varies by age, sex, socio-
economic group, educational level, and region.   
 
17.5  The message to ‘Eat 5- portions-a-day of fruits and vegetables’ is recognised 
by many (but not all) and appears to be one that consumers can implement with the 
removal of circumstantial and attitudinal barriers and with tailored advice and 
practical tips and experience.  Opportunities have been identified and these should be 
incorporated into educational programs and interventions.  
 
17.6  A single message to increase fruit and vegetable consumption may not reach 
the population segments most in need of changing.  It is advisable to spend more time 
understanding the food consumption habits of the population under investigation to 
develop messages to foster behaviour change.  It is also important to tailor the 
message to the sub-group at which the message is aimed. 
 
17.7  There is much scope for the development of an integrated strategy to increase 
fruit and vegetable intakes. Any strategy should be developed in consultation with the 
Health Development Agency (HDA) and the Food Standards Agency (FSA). This 
strategy needs both a continuing commitment to carrying out further research which 
will identify the most reliable tools to evaluate the outcome of interventions and 
develop the evidence base as to what is effective.  It also needs action, in particular to 
 
61

---

tackle the circumstantial barriers such as increasing access to affordable supplies of 
fruit and vegetables, especially for those on low income. 
 
 
62

---

Appendix 1    
 
Characteristics of interventions which are effective in promoting healthy eating 
taken from the HEA Effectiveness Reviews 
 
In schools, workplaces, primary care settings and the community, interventions are 
more likely to be effective if they:   
 
•  Focus on diet only or diet and physical activity 
•  Are based on behavioural theories and goals rather than on the provision of  
information alone 
•  Emphasise personal contact with individuals or  small groups, actively involve 
people and use specific behavioural change strategies 
•  Show some degree of personalisation of the intervention to the individual, by 
either trained personnel or individualised materials 
•  Provide feedback on individual changes in behaviour and risk factors 
•  Provide multiple contact over a sustained period of time 
•  Encourage support in dietary change for the individual, by involving family 
members or local leaders in the intervention, or by promoting changes in the 
supporting environment, e.g. in catering provision.   
 
In supermarkets and catering, the following types of interventions seem to be most 
effective: 
 
•  Simple signs to identify healthier choices accompanied by explanatory leaflets and 
information, but only for as long as the intervention is in place 
•  Covert changes to ingredients and cooking methods, throughout the menu as a 
whole 
•  Changes in the availability of  healthier food choices and the accessibility of less 
healthy choices, but only for the duration of the intervention.  
 
63

---

    
References: 
 
1. 
Department of Health, Nutritional Aspects of Cardiovascular Disease, . 1994, 
The Stationery Office: London. 
2. Lampe, 
J., 
Health effects of vegetables and fruit: assessing mechanisms of 
action in human experimental studies. american Journal of Clinical Nutrition, 
1999. 70(3 Suppl): p. 475S-490S. 
3. 
Poulsen, H., H. Prieme, and S. Loft, Role of oxidative DNA damage in cancer 
initiation and promotion. European Journal of Cancer Prevention, 1998. 7(1): 
p. 9-16. 
4. 
Fillion, L., A. Collins, and S. Southon, Beta-carotene enhances the recovery of 
lymphocytes from oxidative DNA damage. Acta Biochimica Polonica, 1998. 
45(1): p. 183-190. 
5. Collins, 
A., 
et al., Oxidative DNA damage measured in human lymphocytes: 
large differences between sexes and between countries, and correlations with 
heart disease mortality rates, in FASEB Journal. 1998. p. 1397-1400. 
6. Djuric, 
Z., 
et al., Oxidative DNA damage levels in blood from women at high 
risk for breast cancer associated with dietary intakes of meats, vegetables, and 
fruits. Journal of the American Dietetic Association, 1998. 98(5): p. 524-528. 
7. Weisburger, 
J., 
Evaluation of the evidence on the role of tomato products in 
disease prevention. Proceedings of the Society for Experimental Biology & 
Medicine, 1998. 218(2): p. 140-143. 
8. 
Department of Health, Nutritional Aspects of the Development of Cancer, . 
1998, The Stationery Office: London. p. 184-188. 
9. Fraser, 
G., 
Micronutrients, vitamins, and cancer risk. Vitam Horm, 1999. 70(3 
Suppl): p. 1-23. 
10. Eastwood, 
M., 
Interaction of dietary antioxidants in vivo: how fruit and 
vegetables prevent disease? QJM, 1999. 66(2): p. 527-530. 
11. Prieme, 
H., 
et al., No effect of supplementation with vitamin E, ascorbic acid, 
or coenzyme Q10 on oxidative DNA damage estimated by 8-oxo-7,8-dihydro-
2'-deoxyguanosine excretion in smokers. American Journal of Clinical 
Nutrition, 1997. 65(2): p. 503-507. 
12. 
Denis, L., M. Morton, and K. Griffiths, Diet and its preventive role in 
prostatic disease. Eur Urol, 1999. 57(5-6): p. 377-387. 
13. 
Hollman, P. and M. Katan, Dietary flavonoids: intake, health effects and 
bioavailability. Food Chem Toxicol, 1999. 37: p. 937-942. 
14. Rimm, 
E., 
et al., Relation between intake of flavonoids and risk of coronary 
heart disease in male health professionals. Ann Intern Med, 1996. 66(2): p. 
384-389. 
15. Yochum, 
L., 
et al., Dietary flavonoid intake and risk of cardiovascular disease 
in postmenopausal women. Am J Epidemiol, 1999. 149: p. 943-949. 
16. Franke, 
A., 
et al., Inhibition of neoplastic transformation and bioavailability 
of dietary flavonoid agents. Advances in Experimental Medicine & Biology, 
1998. 439: p. 237-248. 
17. Knekt, 
P., 
et al., Dietary flavonoids and the risk of lung cancer and other 
malignant neoplasms. American Journal of Epidemiology, 1997. 146(3): p. 
223-230. 
 
64

---

18. Hertog, 
M., 
et al., Flavonoid intake and long-term risk of coronary heart 
disease and cancer in the Seven Countries Study. Arch Intern Med, 1995. 155: 
p. 381-386. 
19. Birt, 
D., 
et al., Diet intervention for modifying cancer risk. Progress in Clinical 
and Biological Reseaarch, 1996. 395: p. 223-234. 
20. Collins, 
A., 
et al., Serum carotenoids and oxidative DNA damage in human 
lymphocytes. Carcinogenesis, 1998. 19(12): p. 2159-2162. 
21. Rehman, 
A., 
et al., Tomato consumption modulates oxidative DNA damage in 
humans. Biochem biophys Res Commun, 1999. 262: p. 828-831. 
22. Faure, 
H., 
et al., Carotenoids: 2. Diseases and supplementation studies. Ann 
Biol Clin (Paris), 1999. 57(3): p. 273-282. 
23. 
Department of Health, Folic Acid and the Prevention of Disease. Report of the 
Committee on Medical Aspects of Food and Nutrition Policy. Report on 
Health and Social Subjects 50, . 2000, The Stationery Office: London. p. 48-
49. 
24. Perry, 
I., 
et al., Prospective study of serum total homocysteine concentration 
and risk of stroke  in middle-aged men. Lancet, 1995. 346: p. 1395-1398. 
25. Suter, 
P., 
The effects of potassium, magnesium, calcium, and fiber on risk of 
stroke. Nutrition Reviews, 1999. 57: p. 84-88. 
26. Davis, 
D., 
et al., Rethinking breast cancer risk and the environment: the case 
for the precautionary principle. Environmental Health Perspectives, 1998. 
106(9): p. 523-529. 
27. Mendilaharsu, 
M., 
et al., Phytosterols and risk of lung cancer: a case-control 
study in Uruguay. Lung Cancer, 1998. 21(1): p. 37-45. 
28. 
Bouic, P. and J. Lamprecht, Plant sterols and sterolins: a review of their 
immune-modulating properties. Altern Med Rev, 1999. 4(170-177). 
29. Sivam, 
G., 
et al., Helicobacter pylori--in vitro susceptibility to garlic (Allium 
sativum) extract. Nutrition & Cancer, 1997. 27(2): p. 118-121. 
30. Craig, 
W., 
Phytochemicals: guardians of our heatlh. Journal of the American 
Dietetic Association, 1997. 97: p. S199-S204. 
31. 
Steinmetz, K. and J. Potter, Vegetables, fruit and cancer. I. Epidemiology. 
Cancer Causes Control, 1991. 2: p. 325-357. 
32. Verhoeven, 
D., 
et al., A review of the mechanisms underlying 
anticarcinogenicity by brassica vegetables. Chemico-Biological Interactions, 
1997. 103(2): p. 79-129. 
33. Lin, 
H., 
et al., Glutathione transferase null genotype, broccoli, and lower 
prevalence of colorectal adenomas. Cancer Epidemiology, Biomarkers & 
Prevention, 1998. 7(8): p. 647-652. 
34. Probst-Hensch, 
N., 
et al., Absence of the glutathione S-transferase M1 gene 
increases cytochrome P4501A2 activity among frequent consumers of 
cruciferous vegetables in a Caucasian population. Cancer Epidemiology, 
Biomarkers & Prevention, 1998. 7(7): p. 635-638. 
35. Cover, 
C., 
et al., Indole-3-carbinol inhibits the expression of cyclin-dependent 
kinase-6 and induces a G1 cell cycle arrest of human breast cancer cells 
independent of estrogen receptor signaling. Journal of biological Chemistry, 
1998. 273(7): p. 3838-3847. 
36. Jang, 
M., 
et al., Cancer chemopreventive activity of resveratrol, a natural 
product derived from grapes. Science, 1997. 275: p. 218-220. 
 
65

---

37. Janssen, 
P., 
et al., Urinary salicylate excretion in subjects eating a variety of 
diets shows the amounts of bioavailable salicylates in foods are low. American 
Journal of Clinical Nutrition, 1996. 64(5): p. 743-747. 
38. Giovannucci, 
E., 
et al., Calcium and fructose intake in relation to risk of 
prostate cancer. Cancer Research, 1998. 68(3): p. 442-447. 
39. Bingham, 
S., 
Meat, starch and non-starch polysaccharides, are 
epidemiological and experimental findings consistent with acquired genetic 
alterations in sporadic colorectal cancer? Cancer Letters, 1997. 114(1): p. 25-
34. 
40. 
Le Marchand, L., et al., Dietary fibre and colorectal cancer risk. 
Epidemiology, 1997. 8(6): p. 658-665. 
41. Ghadirian, 
P., 
et al., Nutritional factors and colon carcinoma: a case-control 
study involving French Canadians in Montreal, Quebec, Canada. Cancer, 
1997. 80(5): p. 858-864. 
42. Slattery, 
M., 
et al., Plant foods and colon cancer: an assessment of specific 
foods and their related nutrients (United States). Cancer Causes & Control, 
1997. 8(4): p. 575-590. 
43. Negri, 
E., 
et al., Fiber intake and risk of colorectal cancer. Cancer 
Epidemiology, Biomarkers & Prevention, 1998. 7(8): p. 667-671. 
44. Platz, 
E., 
et al., Dietary fiber and distal colorectal adenoma in men. Cancer 
Epidemiology, Biomarkers & Prevention, 1997. 6(9): p. 661-670. 
45. Fuchs, 
C., 
et al., Dietary fiber and the risk of colorectal cancer and adenoma 
in women. N Engl J Med, 1999. 8(3): p. 169-176. 
46. 
Goldin, B. and S. Gorbach, Hormone studies and the diet and breast cancer 
connection. Adv Exp Med Biol, 1994. 364: p. 35-46. 
47. 
Bennett, F. and D. Ingram, Diet and female sex hormone concentrations: an 
intervention study for the type of fat consumed. Am J Clin Nutr, 1990. 52: p. 
808-812. 
48. Rose, 
D., 
et al., High-fibre diet reduces serum oestrogen concentrations in 
pre-menopausal women. Am J Clin Nutr, 1991. 54: p. 520. 
49. Verhoeven, 
D., 
et al., Vitamins C and E, retinol, beta-carotene and dietary 
fibre in relation to breast cancer risk: a prospective cohort study. British 
Journal of Cancer, 1997. 75(1): p. 149-155. 
50. 
La Vecchia, C. and L. Chatenoud, Fibres, whole-grain foods and breast and 
other cancers. European Journal of Cancer Prevention, 1998. 7: p. S25-S28. 
51. Gerber, 
M., 
Fibre and breast cancer. European Journal of Cancer Prevention, 
1998. 7: p. S63-S57. 
52. 
Stubbs, R. and A. Prentice, The effect of covertly manipulating the dietary fat 
to carbohydrate ratio of isoenergetically dense diets on ad libitum food 
intakes in free living humans. Proc Nut Soc, 1993. 52: p. 351A. 
53. Cox, 
D.N., 
et al., Take Five, a nutrition education intervention to increase 
fruit and vegetable intakes:impact on consumer choice and nutrient intakes. 
British Journal of Nutrition, 1998. 80: p. 123-31. 
54. Jenkins, 
D., 
et al., Effect of a diet high in vegetables, fruit, and nuts on serum 
lipids. Metabolism: Clinical & Experimental, 1997. 46(5): p. 530-537. 
55. 
Steinmetz, K. and J. Potter, Vegetables, fruit, and cancer prevention: A 
review. Journal of american Dietetic Association, 1996. 96(10): p. 1027-1039. 
56. 
World Cancer Research Fund and A.I.f.C. Research, Food, Nutrition and the 
Prevention of Cancer: a global perspective, . 1997, American Institute for 
Cancer Research: Washington DC. 
 
66

---

57. 
La Vecchia, C. and A. Tavani, Fruit and vegetables, and human cancer. Eur J 
Cancer Prev., 1998. 7: p. 3-8. 
58. Michaud, 
D., 
et al., Fruit and vegetable intake and incidence of bladder 
cancer in a male prospective cohort. Journal National Cancer Institute, 1999. 
91: p. 605-613. 
59. 
La Vecchia, C. and E. Negri, Nutrition and bladder cancer. Cancer Causes 
and Control, 1996. 7(1): p. 95-100. 
60. Franceschi, 
S., 
Micronutrients and breast cancer. European Journal of Cancer 
Prevention, 1997. 6(6): p. 535-539. 
61. Potischman, 
N., 
et al., Intake of food groups and associated micronutrients in 
relation to risk of early-stage breast cancer. International Journal of Cancer, 
1999. 82(3): p. 315-321. 
62. Potischman, 
N., 
et al., Diet during adolescence and risk of breast cancer 
among young women. Journal of the National Cancer Institute, 1998. 90(3): p. 
226-233. 
63. Freudenheim, 
J., 
et al., Premenopausal breast cancer risk and intake of 
vegetables, fruits, and related nutrients. Journal of National Cancer Institute, 
1996. 88(6): p. 340-348. 
64. 
Hunter, D. and W. Willett, Nutrition and breast cancer. Cancer Causes & 
Control, 1996. 7(1): p. 56-68. 
65. 
Cuzick, J., P. Sasieni, and A. Singer, Risk factors for invasive cervix cancer in 
young women. European Journal of Cancer, 1996. 32A(5): p. 836-841. 
66. Witte, 
J., 
et al., Relation of vegetable, fruit, and grain consumption to 
colrectal adenomatous polyps. American Journal of Epidemiology, 1996. 
144(11): p. 1015-1025. 
67. Franceschi, 
S., 
et al., Role of different types of vegetables and fruit in the 
prevention of cancer of the colon, rectum , and breast. Epidemiology, 1998. 
9(3): p. 338-341. 
68. Franceschi, 
S., 
et al., Food groups and risk of colorectal cancer in Italy. 
International Journal of Cancer, 1997. 72(1): p. 56-61. 
69. Jansen, 
M., 
et al., Dietary fiber and plant foods in relation to colorectal 
cancer mortaility: the Seven Countries Study. International Journal of Cancer, 
1999. 53(2): p. 174-179. 
70. 
Caygill, G., A. Charlett, and M. Hill, Relationship between intake of high-fibre 
foods and energy and the risk of cancer of the large bowel and breast. 
European Journal of Cancer Prevention, 1998. 7: p. S11-S17. 
71. Schatzkin, 
A., 
et al., The polyp prevention trial I: rationale, design, 
recruitment, and baseline participant characteristics. Cancer, Epidemiology, 
Biomarkers & Prevention, 1996. 5(5): p. 375-383. 
72. Lanza, 
E., 
et al., The polyp prevention trial II: dietary intervention program 
and participant baseline dietary characteristics. Cancer Epidemiology, 
Biomarkers & Prevention, 1996. 5(5): p. 385-392. 
73. Wilmink, 
A., 
Overview of the epidemiology of colorectal cancer. Diseases of 
the Colon & Rectum, 1997. 40(4): p. 483-493. 
74. Goodman, 
M., 
et al., Association of soy and fiber consumption with the risk of 
endometrial cancer. American Journal of Epidemiology, 1997. 146(4): p. 294-
306. 
75. Harrison, 
L., 
et al., The role of dietary factors in the intestinal and diffuse 
histolgic subtypes of gastric adenocarcinoma: a case-control study in the U.S. 
Cancer, 1997. 80(6): p. 1021-1028. 
 
67

---

76. Ahn, 
Y., 
Diet and stomach cancer in Korea. International Journal of Cancer, 
1997. Suppl 10: p. 7-9. 
77. Garcia-Closas, 
R., 
et al., Intake of specific carotenoids and flavonoids and the 
risk of gastric cancer in Spain. Cancer Causes Control, 1999. 10(1): p. 71-75. 
78. 
Ward, M. and L. L'Opez-Carillo, Dietary factors and the risk of gastric cancer 
in Mexico City. Am J Epidemiology, 1999. 26: p. 925-932. 
79. Hertog, 
M., 
et al., Fruit and vegetable consumption and cancer mortality in 
the Caerphilly Study. Cancer Epidemiology, Biomarkers & Prevention, 1996. 
5(9): p. 673-677. 
80. 
Botterweck, A., P. van den Brandt, and R. Goldbohm, A prospective cohort 
study on vegetable and fruit consumption and stomach cancer risk in the 
Netherlands. American Journal of Epidemiology, 1998. 148(9): p. 842-853. 
81. 
Terry, P., O. Nyren, and J. Yuen, Protective effect of fruits and vegetables on 
stomach cancer in a cohort of Swedish twins. International Journal of Cancer, 
1998. 76(1): p. 35-37. 
82. Ocke, 
M., 
et al., Adherence to the European Code Against Cancer in relation 
to long-term cancer mortality: intercohort comparisons from the Seven 
Countries Study. Nutrition & Cancer, 1998. 30(1): p. 14-20. 
83. Corella, 
D., 
et al., Dietary habits and geographic variation in stomach cancer 
mortailty in Spain. European Journal of Cancer Prevention, 1996. 5(4): p. 249-
257. 
84. Azevedo, 
L., 
et al., Diet and gastric cancer in Portugal - a multivariate 
model. Eur J Cancer Prev, 1999. 8(1): p. 41-48. 
85. 
Hirohata, T. and S. Kono, Diet/nutrition and stomach cancer in Japan. 
International Journal of Cancer, 1997. Suppl 10: p. 34-36. 
86. 
Kono, S. and T. Hirohata, Nutrition and stomach cancer. 1996, 1996. Cancer 
Causes & Control(7): p. 1. 
87. Tavani, 
A., 
et al., Attributable risks for kidney cancer in northern Italy. 
European Journal of Cancer Prevention, 1997. 6(2): p. 195-199. 
88. Yuan, 
J., 
et al., Cruciferous vegetables in relation to renal cell carcinoma. 
International Journal of Cancer, 1998. 77(2): p. 211-216. 
89. 
Boeing, H., B. Schlehofer, and J. Wahrendorf, Diet, obesity and risk for renal 
cell carcinoma: results from a case-control study in Germany. Zeitschrift fur 
Ernahrungswissenschaft, 1997. 36(1): p. 3-11. 
90. 
Tavani, A. and C. La Vecchia, Epidemiology of renal-cell carcinoma. Journal 
of Nephrology, 1997. 10(2): p. 93-106. 
91. Lindblad, 
P., 
et al., Diet and risk of renal cell cancer: a population-based 
case-control study. Cancer Epidemiology, Biomarkers & Orevention, 1997. 
6(4): p. 215-223. 
92. Pillow, 
P., 
et al., Case-control assessment of diet and lung cancer risk in 
African Americans and Mexican Americans. Nutrition & Cancer, 1997. 29(2): 
p. 169-173. 
93. 
Pawlega, J., J. Rachtan, and T. Dyba, Evaluation of certain risk factors for 
lung cancer in Cracow, Poland. Acta Oncologica, 1997. 36(5): p. 471-476. 
94. Agudo, 
A., 
et al., Vegetables and fruit intake and the risk of lung cancer in 
women in Barcelona, Spain. European Journal of Cancer, 1997. 33(8): p. 
1256-1261. 
95. Hu, 
J., 
et al., A case-control study of diet and lung cancer in northeast China. 
International Journal of Cancer, 1997. 71(6): p. 924-931. 
 
68

---

96. Nyberg, 
F., 
et al., Dietary factors and risk of lung cancer in never-smokers. 
International Journal of Cancer, 1998. 78(4): p. 430-436. 
97. Ocke, 
M., 
et al., Repeated measurements of vegetables, fruits, beta-carotene, 
and vitamins C and E in relation to lung cancer.  The Zutphen Study. 
American Journal of Epidemiology, 1997. 145(4): p. 358-365. 
98. Yong, 
L., 
et al., Intake of vitamins E, C, and A and risk of lung cancer.  The 
NHANES 1 epidemiologic follow-up study.  First National Health and 
Nutrition Examination Survey. American Journal of Epidemiology, 1997. 
146(3): p. 231-243. 
99. Koo, 
L., 
Diet and lung cancer 20+ years later: more questions than answers? 
International Journal of Cancer, 1997. Suppl 10: p. 22-29. 
100.  Ziegler, R., S. Mayne, and C. Swanson, Nutrition and lung cancer. Cancer 
Causes & Control, 1996. 7(1): p. 157-177. 
101. Launoy, 
G., 
et al., Diet and squamous-cell cancer of the oesophagus: a 
French multi-centre case-control study. International Journal of Cancer, 1998. 
76(1): p. 7-12. 
102. Tavani, 
A., 
et al., Tobacco and other risk factors for oesphageal cancer in 
alcohol non-drinkers. European Journal of Cancer Prevention, 1996. 5(5): p. 
313-1-318. 
103. Franceschi, 
S., 
et al., Food groups, oils and butter, and cancer of the oral 
cavity and pharynx. British Journal of Cancer, 1999. 80: p. 614-620. 
104. Bosseti, 
C., 
et al., Risk factors for oral and pharyngeal cancer in women: a 
study from Italy and Switzerland. British Journal of Cancer, 2000. 82: p. 204-
207. 
105. Esteve, 
J., 
et al., Diet and cancers of the larynx and hypopharynx: the IARC 
multi-center study in southwestern Europe. Cancer Causes & Control, 1996. 
7(2): p. 240-252. 
106. Levi, 
F., 
et al., Food groups and risk of oral and pharyngeal cancer. 
International Journal of Cancer, 1998. 77(5): p. 705-709. 
107. Normura, 
A., 
et al., Serum micronutrients and upper aerodigestive tract 
cancer. Cancer, Epidemiology, Biomarkers & Prevention, 1997. 6(6): p. 407-
412. 
108.  Cheng, K. and N. Day, Nutrition and oesophageal cancer. Cancer Causes & 
Control, 1996. 7(1): p. 33-40. 
109.  Riboli, E., N. slimani, and R. Kaaks, Identifiability of food components for 
cancer chemoprevention. IARC Scientific Publications (Lyon), 1996(139): p. 
23-31. 
110. La 
Vecchia, 
C., 
et al., Epidemiology and prevention of oral cancer. Oral 
Oncology, 1997. 33(5): p. 302-312. 
111. Kushi, 
L., 
et al., Prospective study of diet and ovarian cancer. American 
Journal of Epidemiology, 1999. 149(1): p. 21-31. 
112. Stolzenberg-Solomon, 
R., 
et al., Pancreatic cancer risk and nutrition-related 
methyl-group availability indicators in male smokers. J Natl Cancer Inst, 
1999. 
113. Weiderpass, 
E., 
et al., Occurrence, trends and environment etiology of 
pancreatic cancer. Scandinavian Journal of Work, Environment & Health, 
1998. 24(3): p. 165-174. 
114.  Howe, G. and J. Burch, Nutrition and pancreatic cancer. Cancer Causes& 
Control, 1996. 7(1): p. 69-82. 
 
69

---

115. Ghadirian, 
P., 
et al., Nutritional factors and prostate cancer: a case-control 
study of French Canadians in Montreal, Canada. Cancer Causes & Control, 
1996. 7(4): p. 428-436. 
116.  Cohen, J., A. Kristal, and J. Stanford, Fruit and vegetable intakes and prostate 
cancer risk. Journal National Cancer Institute, 2000. 92(1): p. 61-68. 
117. Deneo-Pellegrini, 
H., 
et al., Foods, nutrients and prostate cancer: a case-
control study in Uruguay. British Journal of Cancer, 1999. 80: p. 591-597. 
118. Schuurman, 
A., 
et al., Vegetable and fruit consumption and prostate cancer 
risk: a cohort study in the Netherlands. Cancer Epidemiology, Biomarkers & 
Prevention, 1998. 7(8): p. 673-680. 
119. Kolonel, 
L., 
Nutrition and prostate cancer. Cancer Causes & Control, 1996. 
7(1): p. 83-94. 
120.  Giles, G. and P. Ireland, Diet, nutrition and prostate cancer. International 
Journal of Cancer, 1997. Suppl 10: p. 13-17. 
121. Verhoeven, 
D., 
et al., Epidemiological studies on brassica vegetables and 
cancer risk. Cancer epidemiology, Biomarkers & Prevention, 1996. 5(9): p. 
733-748. 
122.  Potter, J. and K. Steinmetz, Vegetables, fruit and phytoestrogens as preventive 
agents. IARC Scientific Publications (Lyon), 1996(139): p. 61-90. 
123. Serdula, 
M., 
et al., The association between fruit and vegetable intake and 
chronic disease risk factors. Epidemiology, 1996. 7(2): p. 161-165. 
124.  Ness, A. and J. Powles, Fruit and vegetables, and cardiovascular disease: a 
review. International Journal of Epidemiology, 1997. 26(1): p. 1-13. 
125. Key, 
T., 
et al., Dietary habits and mortality in 11,000 vegetarians and health 
conscious people: results of a 17 year follow up. BMJ, 1996. 313(7060): p. 
775-779. 
126. Key, 
T., 
et al., Mortality in vegetarians and non-vegetarians: a collaborative 
analysis of 8300 deaths among 76,000 men and women in five prospective 
studies. Public Health Nutrition, 1998. 1(1): p. 33-41. 
127. Joshipura, 
K., 
et al., Fruit and vegetable intake in relation to risk of ishemic 
stroke. JAMA, 1999. 282(13): p. 1233-1239. 
128. Ascherio, 
A., 
et al., Prospective study of nutritional factors, blood pressure, 
and hypertension among US women. Hypertension, 1996. 27(5): p. 1065-1072. 
129. Manson, 
J., 
et al., Antioxidants and cardiovascular disease: a review. Journal 
of American College of Nutrition, 1993. 12: p. 426-432. 
130.  Rexrode, K. and J. Manson, Antioxidants and coronary heart disease: 
observational studies. Journal of Cardiovascular Risk, 1996. 3(4): p. 363-367. 
131.  Tavani, A. and C. La Vecchia, Beta-carotene and risk of coronary heart 
disease.  a review of observational and intervention studies. Biomed 
Pharmacother, 1999. 53: p. 409-416. 
132.  Spencer, A., D. Carson, and M. Crouch, Vitamin E and coronary artery 
disease. Arch Intern Med, 1999. 28: p. 1313-1320. 
133. Gey, 
K., 
Vitamins E plus C and interacting conutrients required for optimal 
health.  A critical and constructive review of epidemiology and 
supplementation data regarding cardiovascular disease and cancer. 
Biofactors, 1998. 7: p. 113-174. 
134. Menotti, 
A., 
et al., Food intake patterns and 25-year mortality from coronary 
heart disease: cross-cultural correlations in the Seven Countries Study.  the 
Seven Countries Research Group. Eur J Epidemiol, 1999. 275(6): p. 507-515. 
 
70

---

135. Howard, 
A., 
et al., Do hydroxy-carotenoids prevent coronary heart disease?  
A comparison between Belfast and Toulouse. Int J Vitam Nutr Res, 1996. 
66(2): p. 113-118. 
136.  Campbell, T., B. Parpia, and J. Chen, Diet, lifestyle, and the etiology of 
coronary artery disease: the Cornell China study. Am J Cardiol, 1998. 82: p. 
18T-21T. 
137. Ginter, 
E., 
Different trends in cardiovascular mortaility and food consumption 
in Austria and in former Czechoslovakia. Wiener Klinische Wochenschrift, 
1997. 109(17): p. 663-668. 
138.  Mancini, M., V. Parfitt, and P. Rubba, Antioxidants in the Mediterranean diet. 
Canadian Journal of Cardiology, 1995. 11(Suppl G): p. 105G-109G. 
139. Wolk, 
A., 
et al., Long-term intake of dietary fibre and decreased risk of 
coronary heart disease among women. JAMA, 1999. 281: p. 1998-2004. 
140. Rimm, 
E., 
et al., Vegetable, fruit, and cereal fiber intake and risk of coronary 
heart disease among men. JAMA, 1996. 275(6): p. 447-451. 
141. Pietinen, 
P., 
et al., Intake of dietary fiber and risk of coronary heart disease in 
a cohort of Finnish men.  The Alpha-Tocopherol, Beta-Carotene Cancer 
Prevention Study. Circulation, 1996. 94(11): p. 2720-2727. 
142. Singh, 
R., 
et al., Effect of diet and moderate exercise on central obesity and 
associated disturbances, myocardial infarction and mortality in patients with 
and without coronary artery disease. J Am Coll Nutr, 1996. 15(6): p. 592-501. 
143. Lindberg, 
S., 
et al., Age relations of cardiovascular risk factors in a 
traditional Melanesian society: the Kitava Study. American Journal of Clinical 
Nutrition, 1997. 66(4): p. 845-852. 
144. Bobak, 
M., 
et al., An ecological study of determinants of coronary heart 
disease rates: a comparison of Czech, Bavarian and Israeli men. Int J 
Epidemiol, 1999. 28: p. 437-444. 
145. Sorof, 
J., 
et al., Potassium intake and cardiovascular reactivity in children 
with risk factors for essential hypertension. Journal of Pediatrics, 1997. 131: p. 
87-94. 
146. Lloyd, 
T., 
et al., Fruit consumption, fitness, and cardiovascular health in 
female adolescents: the Penn State Young women's Health Study. American 
Journal of Clinical Nutrition, 1998. 67(4): p. 624-630. 
147. Menotti, 
A., 
et al., European and American recommendations for coronary 
heart disease prevention. European Heart Journal, 1998. 275(6): p. A12-A19. 
148. Mares-Perlman, 
J., 
et al., Siet and nuclear lens opacities. Am J Epidemiol, 
1995. 141: p. 322-334. 
149. Hankinson, 
S., 
et al., Nutrient intake and cataract extraction in women, a 
prospective study. BMJ, 1992. 305: p. 335-339. 
150. Brown, 
L., 
et al., A prospective study of carotenoid intake and cataracts 
among US men. Am J Epidemiol, 1998. 147: p. 213A. 
151.  Taylor, A., P. Jacques, and E. Epstein, Relations among aging, antioxidant 
status, and cataract. An J Clin Nutr, 1995. 62(suppl): p. 1439S-1447S. 
152.  Cooper, D., A. Eldridge, and J. Peters, Dietary carotenoids and certain 
cancers, heart disease, and age-related macular degeneration: a review of 
recent research. Nutrition Reviews, 1999. 57(7): p. 201-214. 
153. Nuttall, 
F., 
Dietary fiber in the management of diabetes mellitus. Diabetes, 
1993. 42: p. 503-508. 
154. Cook, 
D., 
et al., Effect of fresh fruit consumption on lung function and wheeze 
in children. Thorax, 1997. 52: p. 628-633. 
 
71

---

155. Tabak, 
C., 
et al., Dietary factors and pulmonary function: a cross sectional 
study in middle aged men from three European countries. Thorax, 1999. 54: p. 
1021-1026. 
156. Strachan, 
D., 
et al., Ventilatory function and winter fresh fruit consumption in 
a random sample of British adults. Thorax, 1991. 46: p. 624-629. 
157.  Schwartz, J. and S. Weiss, Dietary factors and their relationship to 
respiratory symptoms: NHANES II. Am J Epidemiol, 1990. 132: p. 67-76. 
158.  Schwartz, J. and S. Weiss, Relationship between dietary vitamin C intake and 
pulmonary function in the First National Health and Nutrition Examination 
Survey (NHANES I). Am J Clin Nutr, 1994. 59: p. 110-114. 
159. Miedema, 
I., 
et al., Dietary determinants of long-term incidence of chronic 
non-specific lung diseases.  The Zutphen Study. Am J Epidemiol, 1993. 138: p. 
37-45. 
160.  La Vecchia, C., A. Decarli, and R. Pagano, Vegetable consumption and risk of 
chronic disease. Epidemiol, 1998. 9: p. 208-210. 
161.  Hughes, J. and S. Martin, The Department of Health's project to evaluate 
weight management services. Journal of Human Nutrition and Dietetics, 1999. 
12(suppl): p. 1-8. 
162.  Glenny, A. and S. O'Meara, systematic Review of Interventions in the 
Treatment and Prevention of obesity. NHS Centre for Reviews and 
Dissemination Report 10, . 1997, Centre for Reviews and Dissemination: 
York. 
163. Hughes, 
J., 
Factors influencing successful weight loss and weight loss 
maintenance, in St George's Hospital Medical School,. 1992, University of 
London: London. 
164. New, 
S., 
et al., Dietary influences on bone mass and bone metabolism - 
further evidence of a positive link between fruit and vegetable consumption 
and bone health? In press. 
165. Rimm, 
E., 
et al., Vegetarian diet: panacea for modern lifestyle disease. QJM, 
1999. 275(6): p. 531-544. 
166. Fraser, 
G., 
Associations between diet and cancer, ischaemic heart disease, and 
all-cause mortality in non-Hispanic white Seventh-day Adventists. American 
Journal of Clinical Nutrition, 1999. 70(3 Suppl): p. 532S-538S. 
167. Parodi, 
P., 
The French parodox unmasked: the role of folate. Medical 
Hypotheses, 1997. 49: p. 313-318. 
168. de 
Lorgeril, 
M., 
et al., Mediterranean dietary pattern in a randomized trial: 
prolonged survival and possible reduced cancer rate. Arch Intern Med, 1998. 
158: p. 1181-1187. 
169. Dickinson, 
H., 
Cancer trends in England and Wales. BMJ, 2000. 320: p. 884-
885. 
170.  Department of Health, Saving lives: our healthier nation, . 1999, The 
Stationery Office: London. 
171. West, 
R., 
Obesity, . 1994, Office of Health Economics: London. p. 38-41. 
172.  Riboli, E. and R. Kaaks, The EPIC Project: rationale and study design.  
European Prospective Investigation into Cancer and Nutrition. International 
Journal of Epidemiology, 1997. 26(Suppl 1): p. S6-S14. 
173.  Department of Health, The Nutrition of elderly people. Report on Health and 
Social Subjects 43, . 1992, HMSO: London. 
174.  Health Education Authority, Eight Guidelines for a Healthy Diet, . 1997, 
Health Education Authority: London. 
 
72

---

175. WHO, 
Diet, nutrition and the prevention of chronic diseases. Technical 
Report Series 797, . 1991, World Health Organisation: Geneva. 
176. DHHS, 
Healthy people 2000 National health promotion and disease 
prevention objectives, . 1991, US Department of Health and Human Services: 
Washington DC. 
177. Krebs-Smith, 
S., 
Progress in improving diet to reduce cancer risk. Cancer, 
1998. 83(7): p. 1425-1432. 
178. Omenn, 
G., 
Chemoprevention of lung cancer: the rise and demise of beta-
carotene. Annual Review of Public health, 1998. 19: p. 73-99. 
179.  Health Education Authority, Enjoy fruit and vegetables campaign, . 1992, 
Health Education Authroity: London. 
180.  Health Education Board for Scotland, Enjoy fruit and vegetables campaign, . 
1992, Health Education Board for Scotland: Edinburgh. 
181.  Cancer Education Coordinating Group, A Key to Health: Eat Fruit and 
Vegetables every Day, . 1994, Europe Against Cancer: London. 
182.  Ministry of Agriculture Fisheries and Food, National Food Survey 1998, . 
1999, The Stationery Office: London. 
183. Williams, 
C., 
Healthy Eating: clarifying advice about fruit and vegetables. 
BMJ, 1995. 310: p. 1453-1455. 
184. Gregory, 
J., 
et al., The Dietary and Nutritional Survey of British Adults. 1990, 
London: HMSO. 
185.  ONS and HEA, Health in England 1998: investigating the links between 
social inequalities and health, . 1999, The Stationery Office: London. 
186. Anderson, 
A., 
et al., Ripe for change: Fruit and vegetables in Scotland - 
current patterns and potential for change. Health Bulletin, 1994. 52(1): p. 51-
64. 
187. Agudo, 
A., 
et al., Dietary intake of vegetables and fruits among adults in five 
regions of Spain. EPIC Group of Spain.  European Prospective Investigation 
into Cancer and Nutrition. Eur. J. Clin. Nutr., 1999. 53(3): p. 174-180. 
188. Thompson, 
B., 
et al., Baseline fruit and vegetable intake among adults in 
seven 5 a day study centers located in diverse geographic areas. Journal 
American Dietetic Association, 1999. 99(10): p. 1241-1248. 
189. Williams, 
C., 
Recommendations and current consumption patterns: how big is 
the gap?, in At Least Five a Day. Strategies to increase vegetable and fruit 
consumption, I. Sharp, Editor. 1997, The Stationery Office: London. 
190.  Ministry of Agriculture Fisheries and Food, Food Portion sizes. Second 
edition, . 1993, HMSO: London. 
191. Krebs-Smith, 
S., 
et al., Using food frequency questionnaires to estimate fruit 
and vegetable intake: association between the number of questions and total 
intakes. J. Nutr. Edu., 1995. 27: p. 80-88. 
192. Cox, 
D., 
et al., Consumer strategies for increasing fruit and vegetable 
consumption in the UK. Proceed. Nutr. Soc., 1997. 56(1A): p. 62A. 
193. Cox, 
D.N., 
et al., Measuring fruit and vegetable intake; is five a day enough? 
European Journal of Clinical Nutrition, 1997. 51(177-80). 
194. Roe, 
L., 
et al., Health promotion interventions to promote healthy eating in 
the general population - a review. 1997, London: HEA. 
195.  Health Education Authority, Effectiveness of interventions to promote healthy 
eating in elderly people living in the community, . 1998, Health Education 
Authority: London. 
 
73

---

196.  Health Education Authority, Effectiveness of interventions to promote healthy 
eating in preschool children aged 1-5 years: a review, . 1998, Health 
Education Authority: London. 
197.  Health Education Authority, Effectiveness of interventions to promote healthy 
eating in pregnant women and women of childbearing age: a review, . 1998, 
Health Education Authority: London. 
198.  Health Education Authority, Effectiveness of interventions to promote healthy 
feeding of infants under one year of age: a review, . 1998, Health Education 
Authroity: London. 
199.  Health Education Authority, Effectiveness of interventions to promote healthy 
eating in people from minority ethnic groups: a review, . 1998, Health 
Education Authority: London. 
200.  Kennedy, L., C. Hunt, and P. Hodgson, Nutrition Education Program Based 
on EFNEP 
for Low-Income Women in the United Kingdom: "Friends with Food". Journal of 
Nutrition Education, 1998. 30(2): p. 89-99. 
201.  National Heart Forum, At Least Five a Day. Strategies to increase vegetable 
and fruit consumption, . 1997, The Stationery Office: London. 
202.  Foster, A. and K. McColl, Increasing vegetable and fruit consumption: 
interventions to change attitudes and improve access, in At least five a day. 
Strategies to increase vegetable and fruit consumption., I. Sharp, Editor. 1997, 
The Stationery office: London. 
203.  Dittus, K.L., V.A. Hiller, and K.A. Beerman, Benefits and barriers to fruit and 
vegetable intake: relationship between attitudes and consumption. Journal of 
Nutrition Education, 1995. 27: p. 120-26. 
204. Campbell, 
M.K., 
et al., Improving dietary behaviour: the effectiveness of 
tailored messages in primary care settings. American Journal of Public 
Health, 1994. 84(5): p. 783-87. 
205. Anderson, 
A.S., 
et al., Take Five, a nutrition education intervention to 
increase fruit and vegetable intakes:impact on attitudes to dietary change. 
British Journal of Nutrition, 1998. 80: p. 133-40. 
206.  Baranowski, T., K.W. Cullen, and J. Baranowski, Psychosocial correlates of 
dietary intake; advancing dietary intervention. Annual Review of Nutrition, 
1999. 262: p. 17-40. 
207. Campbell, 
M.K., 
et al., Stages of Change and Psychosocial Correlates of 
Fruit and Vegetable Consumption among Rural African-American Church 
members. American Journal of Health Promotion, 1998. 12(3): p. 185-91. 
208. Campbell, 
M.K., 
et al., Stages of change for increasing fruit and vegetable 
consumption among adults and young adults participating in the national 5-a-
day for Better Health community studies. Health Education Behaviour, 1999. 
26: p. 513-34. 
209. Nicklas, 
T., 
et al., Outcomes of a High School Program to increase Fruit and 
Vegetable Consumption: Gimme 5 - A fresh Nutrition Concept for Students. 
Journal of School Health, 1998. 68(6): p. 248-253. 
210. Perry, 
C.L., 
et al., Changing fruit and vegetable consumption among children; 
The 5 a day Power Plus Program in St. Paul, Minnesota. American Journal of 
Public Health., 1998. 88(4): p. 603-609. 
211. Buller, 
D.B., 
et al.,  Randomised trial testing the effect of peer education at 
increasing fruit and vegetable intake. Journal National Cancer Institute, 1999. 
91: p. 1491-500. 
 
74

---

212. Sorenson, 
G., 
et al., Increasing fruit and vegetable consumption through 
worksites and families in the Treatwell 5-a day study. American Journal of 
Public Health, 1999. 89(1): p. 54-60. 
213. Campbell, 
M., 
et al., Fruit and Vegetable Consumption and Prevention of 
Cancer: The Black Churches United for Better Health Project. American 
Journal of Public Health, 1999. 89(9): p. 1390-1396. 
214. Havas, 
S., 
et al., Final results of the Maryland WIC 5 A Day Promotion 
Program. American Journal of Public Health., 1998. 88(8): p. 1161-7. 
 
75

---

Document Outline

• ��
  • ��
    • ��
    • ��
  • ��
• ��
• ��
  • ��
  • ��
  • ��
  • ��
  • ��
  • ��
  • ��
• ��
• ��
  • ��
  • ��
• ��
• ��
• ��
• ��
  • ��
  • ��
  • ��
  • ��
  • ��
    • ��
  • ��
• ��
• ��
• ��
  • ��
  • ��

---