Diet, nutrition and the prevention of cancer

Citation
Key, Timothy J, Arthur Schatzkin, Walter C Willett, Naomi E Allen, Elizabeth A Spencer, and Ruth
C Travis. 2004. “Diet, Nutrition and the Prevention of Cancer.” Public Health Nutrition 7 (1a):
187–200. https://doi.org/10.1079/phn2003588.

Permanent link
http://nrs.harvard.edu/urn-3:HUL.InstRepos:41426781

Terms of Use
This article was downloaded from Harvard University’s DASH repository, and is made available
under the terms and conditions applicable to Other Posted Material, as set forth at http://
nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of-use#LAA

Share Your Story

The Harvard community has made this article openly available.
Please share how this access benefits you. Submit a story .

Accessibility
 Public Health Nutrition: 7(1A), 187–200 DOI: 10.1079/PHN2003588

Diet, nutrition and the prevention of cancer

Timothy J Key1,*, Arthur Schatzkin2, Walter C Willett3, Naomi E Allen1,
Elizabeth A Spencer1 and Ruth C Travis1
1
Cancer Research UK Epidemiology Unit, University of Oxford, Oxford, UK: 2Nutritional Epidemiology Branch,
Division of Cancer, Epidemiology and Genetics, National Cancer Institute, Bethesda, USA: 3Departments of
Epidemiology and Nutrition, Harvard School of Public Health, Boston, USA

Abstract
Objective: To assess the epidemiological evidence on diet and cancer and make
public health recommendations.
Design: Review of published studies, concentrating on recent systematic reviews,
meta-analyses and large prospective studies.
Conclusions and recommendations: Overweight/obesity increases the risk for
cancers of the oesophagus (adenocarcinoma), colorectum, breast (postmenopausal),
endometrium and kidney; body weight should be maintained in the body mass index
range of 18.5-25 kg/m2, and weight gain in adulthood avoided. Alcohol causes
cancers of the oral cavity, pharynx, oesophagus and liver, and a small increase in the
risk for breast cancer; if consumed, alcohol intake should not exceed 2 units/d.
Aflatoxin in foods causes liver cancer, although its importance in the absence of
hepatitis virus infections is not clear; exposure to aflatoxin in foods should be
minimised. Chinese-style salted fish increases the risk for nasopharyngeal cancer,
particularly if eaten during childhood, and should be eaten only in moderation. Fruits
and vegetables probably reduce the risk for cancers of the oral cavity, oesophagus,
stomach and colorectum, and diets should include at least 400 g/d of total fruits and
vegetables. Preserved meat and red meat probably increase the risk for colorectal
cancer; if eaten, consumption of these foods should be moderate. Salt preserved
foods and high salt intake probably increase the risk for stomach cancer; overall
consumption of salt preserved foods and salt should be moderate. Very hot drinks and
foods probably increase the risk for cancers of the oral cavity, pharynx and Keywords
oesophagus; drinks and foods should not be consumed when they are scalding hot. Diet
Physical activity, the main determinant of energy expenditure, reduces the risk for Nutrition
colorectal cancer and probably reduces the risk for breast cancer; regular physical Cancer
activity should be taken. Review

Dietary factors have been thought to account for about have, therefore, based our review on the material
30% of cancers in Western countries1, making diet second summarised in these reports and on more recent studies,
only to tobacco as a preventable cause of cancer. The giving particular credence to the results of large
contribution of diet to cancer risk in developing countries prospective studies and to the few results from
has been considered to be lower, perhaps around 20%2. randomised controlled trials. We start by briefly discussing
Unravelling the effects of diet on cancer risk is, therefore, the types of evidence available for formulating and testing
of great public health importance, but research to date has hypotheses.
uncovered few definite effects and left frustratingly large
areas of uncertainty. International comparisons, migrants and time
In this paper, we summarise our view of the current trends
state of knowledge on diet and cancer. This paper is not a Many of the prominent hypotheses for effects of diet on
systematic review of diet and cancer. Other organisations cancer risk have been derived from examination of the
have recently published detailed reviews of this subject, in associations between dietary patterns and cancer rates in
particular the World Cancer Research Fund/American different populations around the world. It was noted in the
Institute for Cancer Research3, and the Department of 1970s that developed Western countries have diets high in
Health in the UK4. These publications provide good animal products, fat and sugar, and high rates of cancers
reviews of research published up until the mid 1990s; we of the colorectum, breast and prostate. In contrast,

*Corresponding author: Email [email protected] q The Authors 2004

188 TJ Key et al.
developing countries typically have diets based on one or products, sugar and other refined carbohydrates, and
two starchy staple foods, low intakes of animal products, reduced intake of relatively unrefined starchy staple foods.
fat and sugar, low rates of these ‘Western’ cancers, and In terms of nutrients, Western diets are characterised by
sometimes high rates of other types of cancer such as adequate or excessive energy intake, together with high
cancers of the oesophagus, stomach and liver5. Other intakes of protein and fat, whereas micronutrient
studies have shown that cancer rates often change in deficiency (i.e. deficiency of vitamins or trace minerals)
populations which migrate from one country to another, is much more common in developing countries. Thus, the
and change over time within countries. For example, the search for the nutritional causes of the typical Western
formerly low rates of colorectal cancer among Japanese cancers has focussed mainly on excess consumption of
people have increased both on migration to the USA and, macronutrients, whereas the search for the nutritional
more recently, with the increasing Westernisation of the causes of the typical cancers of developing countries has
diet in Japan6. As discussed below, some of the main focussed on deficient intake of micronutrients.
hypotheses that were derived from these ecological
observations have not been supported by the results of Case-control studies
detailed studies of the diets of individuals. However, the During the last 30 years, hundreds of studies have been
international variations in diet and cancer rates continue to published that have examined the association between the
suggest that diet is an important risk factor for many diets of individuals and their risk for developing cancer.
common cancers, and therefore that cancer may be partly Most of the earlier studies used a case-control design, in
preventable by dietary changes. which people who already have cancer are asked what
Figure 1 shows estimated incidence rates for the most they used to eat before they were diagnosed with cancer,
common cancers world-wide in 20007. Lung cancer is the and their diets are compared with those reported by
most common cancer in the world, and the most common people without cancer (controls). These case-control
cancer among men in both developed and developing studies are useful for searching for possible dietary effects,
countries, whereas breast cancer is the most common but cannot be relied on to establish moderate dietary
cancer among women. The archetypal Western cancers associations because they are susceptible to both recall
are those of the colorectum, breast and prostate. and selection biases: people with cancer may recall their
Westernisation encompasses many changes in diet and diet differently from healthy people, and healthy controls
lifestyle, including increased consumption of meat, dairy are rarely fully representative of the base population and
may report a relatively ‘healthy’ diet8,9. The impact of these
biases varies between studies, especially in relation to the
participation rate among eligible controls, but in general
relative risks in case-control studies of up to at least 1.3
may reflect bias rather than a true association.

Prospective studies
In prospective studies, dietary intakes are measured at
recruitment and people are followed-up for cancer
incidence, then the cancer incidence rates are compared
between groups with different diets at baseline. This
design eliminates the recall and selection biases to which
case-control studies are susceptible, but other factors such
as measurement error and confounding must be
considered when interpreting the results of prospective
studies (as well as case-control studies) of diet and cancer.
Another limitation of most prospective studies is that the
baseline dietary intake is measured at one point during
adult life, which may not be the most relevant period in
relation to the development of cancer over many years.

Interpretation of observational studies

Large prospective studies can establish whether or not
there is any association between reported diet and cancer
risk within the population studied. The main factors that
Fig. 1 Age-adjusted incidence rates of common cancers among need to be considered when interpreting results from such
men and women in developed and developing countries studies are measurement error and confounding.
Diet, nutrition and cancer 189
Measurement error that only a small number of nutritional factors can be
Epidemiological studies generally use a relatively short tested in each trial, usually for a short period. The results of
and simple dietary questionnaire. The validity of such the few large trials that have been completed are
questionnaires has been extensively investigated, and important, but where trials do not show an effect it
while it is clear that current assessment methods are remains possible that an effect would have been seen at a
moderately precise and can detect some associations of different dose, at a different time in life, or if the duration
diet with disease risk, it is also possible that the of the trial had been longer. Another point which should
measurement error is sometimes large enough to obscure be considered when interpreting the results from
some potentially important associations of diet with cancer randomised controlled trials is the possibility that the
risk. A further problem is that there are often strong effect of a dietary component on cancer risk may differ
correlations between different foods and nutrients, according to the characteristics of the population studied;
making it hard to attribute associations with risk to for example, the effects of a multinutrient supplement
particular dietary factors, especially when the precision of could be more marked in a population with a low dietary
the measurements is variable. intake of micronutrients than in a population with a high
Body mass index (the weight in kilograms divided by dietary intake of micronutrients.
the square of the height in metres: kg/m2 ) and alcohol Within the field of diet and cancer, randomised
present special cases. Body mass index is not diet per se, controlled trials are most suitable for testing hypotheses
but is determined by the balance of energy intake and for protective effects of specific micronutrients. An
energy expenditure; body mass index therefore serves as extensive review of these hypotheses and of the design
an indicator of chronic energy balance, and can be and interpretation of such trials is given in the first IARC
measured much more accurately and precisely in Handbook of Cancer Prevention10.
epidemiological studies than either energy intake or
energy expenditure. Alcohol is not always included under
Review of the role of diet in the aetiology of the
the term diet, but alcoholic drinks do contribute a
major cancers
significant supply of energy and some nutrients in many
populations. Although estimates of alcohol intake
commonly suffer from underreporting, they generally Cancers of the oral cavity, pharynx and
rank individuals much more precisely than estimates of oesophagus
intakes of other foods and nutrients. Cancers of the oral cavity, pharynx and oesophagus were
estimated to account for 867,000 cases and 582,000 deaths
Confounding in 200011. Incidence rates of these cancers vary widely
Observed associations of dietary factors with cancer risk between populations; for example, oesophageal cancer is
can be confounded by other risk factors for cancer, such as over a hundred times more common in parts of Central
smoking and physical activity, which are associated with Asia, China and Southern Africa than in most parts of
dietary habits. In theory, confounding can be allowed for Europe, North America and West Africa7. In developed
by statistical adjustments, but in practice this adjustment is countries, the main risk factors are alcohol and tobacco,
never perfect because the non-dietary risk factors and up to 75% of these cancers are attributable to these
themselves are measured with some error. The possibility two lifestyle factors6. The mechanism of the effect of
that apparent dietary associations with cancer risk are alcohol on these cancers is not known, but may involve
confounded by other risk factors, therefore, needs to be direct effects on the epithelium12. Overweight/obesity is
examined very carefully. Allowing for confounding is an established risk factor specifically for adenocarcinoma
extremely important for cancers such as lung cancer, where (but not squamous cell carcinoma) of the oesophagus13 –
15
smoking causes a very large increase in risk and is known . In developing countries, around 60% of cancers of the
to be associated with diet. The relationship between diet oral cavity, pharynx and oesophagus are thought to be
and other cancers needs to take into account potential due to micronutrient deficiencies related to a restricted
confounding factors, such as Helicobacter pylori for diet that is low in fruits and vegetables and animal
stomach cancer, physical activity for colorectal cancer, products3,16; it should be noted, however, that the
and human papillomavirus for cervical cancer (see below). evidence for a protective effect of fruits and vegetables is
largely derived from case-control studies and there are few
Randomised controlled trials data yet from prospective studies17. The relative roles of
Randomised controlled trials eliminate both the biases and various micronutrients are not yet clear, but deficiencies of
the confounding which can affect observational studies riboflavin, folate, vitamin C and zinc may all be
and the results can, therefore, be confidently interpreted in important3,16. There is also consistent evidence that
terms of cause and effect. Within the field of diet and consuming drinks and foods at a very high temperature
cancer, however, trials are limited by the difficulty of increases the risk for these cancers18. The results of trials
randomising at the level of foods, and by the constraints in Linxian, China, aimed at reducing oesophageal cancer
190 TJ Key et al.
11
rates with micronutrient supplements, have been prom- 492,000 deaths in 2000 . Incidence rates are approximately
ising but not definitive19,20. 10-fold higher in developed than in developing countries7.
It has been suggested that diet-related factors may account
Nasopharyngeal cancer for up to 80% of the between-country differences in rates30.
This is particularly common in Southeast Asia7, and has The best established dietary-related risk factor is over-
been consistently associated with a high intake of Chinese- weight/obesity15. Alcohol probably causes a small increase
style salted fish, especially during early childhood21,22, as in risk3. Adult height, which is partly determined by the
well as with infection with the Epstein–Barr virus23. adequacy of nutrition in childhood and adolescence, is
Chinese-style salted fish is a special product which is weakly associated with increased risk, and physical activity
usually softened by partial decomposition before or has been consistently associated with a reduced risk15,31.
during salting; other types of salted fish have been studied These factors together, however, do not explain the large
and not found to be convincingly associated with the risk variation between populations, and there is almost
for developing nasopharyngeal cancer22. universal agreement that some aspects of a Western diet
are a major determinant of risk.
Stomach cancer
Stomach cancer was estimated to account for 876,000 Meat
cases and 647,000 deaths in 200011. Until about 20 years International correlation studies show a strong association
ago stomach cancer was the most common cancer in the between per capita consumption of meat and colorectal
world, but mortality rates have been falling in all Western cancer mortality5, and several mechanisms have been
countries24 and stomach cancer is now much more proposed through which meat may increase cancer risk.
common in Asia than in Europe or North America7. Mutagenic heterocyclic amines and polycyclic aromatic
Infection with the bacterium H. pylori is an established hydrocarbons can be formed during the cooking of meat
risk factor, but not a sufficient cause, for the develop- at high temperatures32,33, and nitrites and their related
ment of stomach cancer25. Diet is also thought to be compounds found in smoked, salted and some processed
important in the aetiology of this disease, and dietary meat products may be converted to carcinogenic N-nitroso
changes are implicated in the recent decline in stomach compounds in the colon34. In addition, high iron levels in
cancer incidence and mortality rates in many countries. the colon may increase the formation of mutagenic free
Substantial evidence, mainly from case-control studies, radicals35. The results of observational studies of meat and
suggests that risk is increased by high intakes of some colorectal cancer have varied3; a recent systematic review
traditionally preserved salted foods, especially meats and concluded that preserved meat is associated with an
pickles, and with salt per se, and that risk is decreased by increased risk for colorectal cancer but that fresh meat is
high intakes of fruits and vegetables26, perhaps due to not36 and most studies have not observed positive
their vitamin C content. However, evidence from associations with poultry or fish3. However, mortality
prospective studies does not clearly support a protective rates for colorectal cancer are similar in Western
effect for fruits and vegetables27,28. The introduction of vegetarians and comparable non-vegetarians37. Overall,
refrigeration has also been associated with decreased the evidence is not conclusive but suggests that high
risk, probably through reducing intakes of salted foods consumption of preserved and red meat probably
and facilitating year-round fruit and vegetable availability3. increases the risk for colorectal cancer.
The results of micronutrient supplementation trials in
developing countries have been encouraging but not Fat
definitive. In Linxian, China, combined supplementation As with meat, international correlation studies show a
with b-carotene, selenium and a-tocopherol resulted in a strong association between per capita consumption of fat
significant reduction in stomach cancer mortality, but no and colorectal cancer mortality5. Possible mechanisms
significant benefit was obtained from vitamin C19. A recent proposed to explain such an association are that a high fat
trial in Colombia showed increased regression of intake may increase the levels of cytotoxic free fatty acids
precancerous gastric dysplasia both in subjects given or secondary bile acids in the lumen of the large intestine.
b-carotene and in subjects given vitamin C29. However, the results of observational studies of fat and
Further prospective data are needed, in particular to colorectal cancer have, overall, not been supportive of an
examine whether some of the dietary associations may be association with fat intake, especially after adjusting for
partly confounded by H. pylori infection and whether total energy intake3,38.
dietary factors may modify the association of H. pylori
with risk. Fruits, vegetables and fibre
Burkitt39 suggested that the low rates of colorectal cancer
Colorectal cancer in Africa were due to the high consumption of dietary
Colorectal cancer is the third most common cancer in the fibre, and there are several plausible mechanisms for a
world7 and was estimated to account for 945,000 cases and protective effect. Fibre increases stool bulk and speeds the
Diet, nutrition and cancer 191
transit of food through the colon, thus diluting the gut Cancer of the liver
contents and perhaps reducing the absorption of Liver cancer was estimated to account for 564,000 cases
carcinogens by the colonic mucosa40. Fermentation of and 549,000 deaths in 200011. Approximately 75% of cases
fibre (and resistant starch) in the large intestine produces of liver cancer occur in developing countries, and liver
short chain fatty acids such as butyrate, which may protect cancer rates vary over 20-fold between countries, being
against colorectal cancer through the ability to promote much higher in sub-Saharan Africa and Southeast Asia
differentiation, induce apoptosis and/or inhibit the than in Europe and North America7. The major risk factor
production of secondary bile acids by reducing luminal for hepatocellular carcinoma, the main type of liver
pH41,42. Many case-control studies of colorectal cancer cancer, is chronic infection with hepatitis B, and to a lesser
have observed moderately lower risk in association with extent, hepatitis C virus63. Ingestion of foods contaminated
high consumption of dietary fibre, and/or fruits and with the mycotoxin aflatoxin22,64 is an important risk factor
vegetables43,44, but the results of recent large prospective among people in developing countries with active
studies have been inconsistent45 – 47. Furthermore, results hepatitis virus infection. Excessive alcohol consumption
from randomised controlled trials have not shown that is the main diet-related risk factor for liver cancer in
intervention over a 3–4 year period with supplemental Western countries, probably via the development of
fibre or a diet low in fat and high in fibre and fruits and cirrhosis and alcoholic hepatitis6. Little is known about
vegetables can reduce the recurrence of colorectal possible nutritional cofactors for viral carcinogenesis, but
adenomas48 – 50. It is possible that some of the incon- this may be an important area for research3.
sistencies are due to differences between studies in the
types of fibre eaten and in the methods for classifying fibre Cancer of the pancreas
in food tables. Other possibilities are that the association Cancer of the pancreas was estimated to account for
with fruits and vegetables is principally due to an increase 216,000 cases and 214,000 deaths in 2000 and is more
in risk at very low levels of consumption51, or that high common in Western countries than in developing
intakes of refined flour or sugar (rather than low intakes of countries7,11. Time trends suggest that both incidence
fibre) increase risk through chronic hyperinsulinaemia or and mortality for cancer of the pancreas are increasing in
other mechanisms52,53. At present, the hypothesis that most parts of the world, although some of this apparent
fruits, vegetables and fibre may reduce the risk for increase may be due to improvements in diagnostic
colorectal cancer has not been firmly established. methods6. Overweight/obesity possibly increases the
risk3,65. Some studies have suggested that risk is increased
Folate by high intakes of meat, and reduced by high intakes of
Some recent prospective studies have suggested that a vegetables, but these data are not consistent and come
methyl-deplete diet (i.e. a diet low in folate and mostly from case-control studies3. Over the next few years
methionine and high in alcohol) is associated with an there will be substantially more prospective data on diet
increased risk of colon cancer54,55. Also, use of folic acid- and cancer of the pancreas, and it is possible that more
containing multiple vitamin supplements has been clear-cut associations with dietary factors will emerge.
associated with lower risk of colon cancer56. A diminished
folate status may contribute to carcinogenesis by alteration Lung cancer
of gene expression and increased DNA damage57,58 and Lung cancer is the most common cancer in the world7 and
chromosome breakage59. The finding that a common was estimated to account for 1,239,000 cases and 1,103,000
polymorphism in the methylenetetrahydrofolate reductase deaths in 200011. Heavy smoking increases the risk by
gene involved in folic acid metabolism may also be around 30-fold, and smoking causes over 80% of lung
associated with colorectal cancer60 strengthens the cancers in Western countries6. The possibility that diet
hypothesis that dietary folate may be an important factor might also have an effect on lung cancer risk was raised in
in colorectal carcinogenesis. the 1970s following the observation that, after allowing for
smoking, increased lung cancer risk was associated with a
Calcium low dietary intake of vitamin A66. Since then, numerous
Another promising hypothesis is that relatively high observational studies have found that lung cancer patients
intakes of calcium may reduce the risk for colorectal generally report a lower intake of fruits, vegetables and
cancer, perhaps by forming complexes with secondary related nutrients (such as b-carotene) than controls3,4. The
bile acids in the intestinal lumen3 or by inhibiting the only one of these factors to have been tested in controlled
hyperproliferative effects of dietary haem61. Several trials, namely b-carotene, has however, failed to produce
observational studies have supported this hypothesis3,4, any benefit when given as a supplement for up to 12
and two trials have suggested that supplemental calcium years67 – 69.
may have a modest protective effect on the recurrence of The possible effect of diet on lung cancer risk remains
colorectal adenomas50,62. More data are needed to controversial. Several recent observational studies have
evaluate this hypothesis. continued to observe an association of fruits and
192 TJ Key et al.
77
vegetables with reduced risk, but this association has been support this hypothesis , and only limited data are
weak in prospective studies70,71. This apparent relation- available to evaluate whether dietary fat alters circulating
ship may be partly due to residual confounding by oestrogen levels78,79.
smoking, since smokers generally consume less fruits and
vegetables than non-smokers, but there may also be some Other dietary factors
protective effect of these foods. In public health terms, The results of studies of other dietary factors including
however, the overriding priority is to reduce the meat, dairy products, fruits and vegetables, fibre and
prevalence of smoking. phyto-oestrogens are inconsistent3,4,80,81.

Breast cancer Cancer of the endometrium

Breast cancer is the second most common cancer in the Endometrial cancer was estimated to account for 189,000
world and the most common cancer among women. cases and 45,000 deaths in women in 2000, with the
Breast cancer was estimated to account for 1,105,000 cases highest incidence rates occurring in Western countries7,11.
and 373,000 deaths in women in 200011. Incidence rates Endometrial cancer risk is about 3-fold higher in obese
are about five times higher in Western countries than in women than lean women15,82. As with breast cancer, the
less developed countries and Japan7. Much of this effect of obesity in postmenopausal women on the risk for
international variation is due to differences in established endometrial cancer is probably mediated by the increase
reproductive risk factors such as age at menarche, parity in serum concentrations of oestradiol and the reduction in
and age at births, and breastfeeding72,73, but differences in serum concentrations of sex hormone-binding globulin; in
dietary habits and physical activity may also contribute. In premenopausal women, the mechanism probably
fact, age at menarche is partly determined by dietary involves the increase in anovulation and consequent
factors, in that restricted dietary intake during childhood increased exposure to oestradiol unopposed by pro-
and adolescence leads to delayed menarche. Adult height, gesterone83. Some case-control studies have suggested
also, is weakly positively associated with risk, and is partly that diets high in fruits and vegetables may reduce risk and
determined by dietary factors during childhood and that diets high in saturated or total fat may increase risk,
adolescence72. Oestradiol and perhaps other hormones but the data are limited3.
play a key role in the aetiology of breast cancer72, and it is
possible that any further dietary effects on risk are Cancer of the cervix
mediated by hormonal mechanisms. Cancer of the cervix was estimated to account for 471,000
cases and 233,00 deaths in women in 200011. The highest
Overweight/obesity rates are in sub-Saharan Africa, Central and South America,
Obesity increases breast cancer risk in postmenopausal and south-east Asia7. The major cause of cervical cancer is
women by around 50%, probably by increasing serum infection with certain subtypes of the human papilloma-
concentrations of free oestradiol72. Obesity does not virus84. Fruits, vegetables and related nutrients such as
increase risk among premenopausal women (perhaps carotenoids and folate tend to be inversely related with
because it frequently leads to anovular menstrual cycles), risk3,4, but these associations may be largely due to
but obesity in premenopausal women is likely to lead to confounding by papillomavirus infections, smoking and
obesity throughout life and therefore to an eventual other factors. Further research is needed, particularly on
increase in breast cancer risk. the possible role of folate deficiency3,4.

Alcohol Cancer of the ovary

The only other established dietary risk factor for breast Cancer of the ovary was estimated to account for 192,000
cancer is alcohol. There is now a large amount of data cases and 114,000 deaths in women in 200011, with the
from well-designed studies which consistently shows a highest incidence rates occurring in Western countries7.
small increase in risk with increasing consumption, with Risk is reduced by high parity and by long-term use of
about a 7% increase in risk for an average of one alcoholic combined oral contraceptives85. Some studies have
drink every day74. The mechanism for this association is suggested that risk is increased by high intakes of fat or
not known, but may involve increases in oestrogen dairy products, and reduced by high intakes of vegetables,
levels75; alternatively, some recent studies suggest that the but the data are not consistent and more prospective data
adverse effect of alcohol may be exacerbated by a low are required to examine these possible associations3.
folate intake76.
Prostate cancer
Fat Prostate cancer was estimated to account for 543,000
Much research and controversy has surrounded the cases and 204,000 deaths in 200011. Prostate cancer
hypothesis that a high fat intake increases breast cancer incidence rates are strongly affected by diagnostic
risk. The best data currently available, however, do not practices and, therefore, difficult to interpret, but mortality
Diet, nutrition and cancer 193
rates show that death from prostate cancer is about ten Conclusions on the effects of diet on cancer risk
times more common in North America and Europe than in
Asia7. Strengths and weaknesses of the evidence
Little is known about the aetiology of prostate cancer, Attaining definitive evidence to confirm or refute effects of
although ecological studies suggest that it is positively specific dietary factors on risks of human cancers is
associated with a Western-style diet5. The data from challenging and for many relationships may be imposs-
prospective studies have not established causal or ible9. Ideally, each potential relationship would be tested
protective associations for specific nutrients or dietary in multiple randomised trials to achieve a clear conclusion.
factors3,4. Diets high in red meat, dairy products and However, this is not feasible for many reasons, including
animal fat have frequently been implicated in the the large number of dietary constituents that could be
development of prostate cancer, although the data are tested and the many different human cancers. In addition,
not entirely consistent3,86 – 88. Randomised controlled trials uncertainty about the time in life and number of years
have provided substantial, consistent evidence that before diagnosis that a specific aspect of diet may act
supplements of b-carotene do not alter the risk for hinders the design and interpretation of randomised trials.
prostate cancer67,68,89 but have suggested that vitamin E89 Practical problems with compliance in long-term studies
and selenium90 might have a protective effect. Lycopene, and the need for many thousands of subjects create further
primarily from tomatoes, has been associated with a obstacles. Finally, many dietary factors may not act in
reduced risk in some observational studies, but the data isolation and it may be their interaction with other dietary,
are not consistent91. lifestyle and/or genetic factors that may alter cell growth
Hormones control the growth of the prostate, and and affect cancer risk. For these and other reasons,
interventions that lower androgen levels are moderately randomised trials have contributed only modestly to
effective in treating this disease. Prospective studies which present knowledge about diet and cancer, and this is likely
have examined the possible associations between serum to continue to be true for many years to come.
hormone concentrations and prostate cancer risk have The primary alternative to randomised trials is observa-
suggested that risk may be increased by high levels of tional studies of human cancers or their precursors,
bioavailable androgens92,93 and of insulin-like growth interpreted in the light of other evidence including
factor-I (IGF-I)94,95, although there are not sufficient data metabolic studies, animal experiments and mechanistic
to consider either of these associations as established. Diet investigations. Studies comparing rates of cancer in various
might affect prostate cancer risk by affecting hormone populations and among migrants have been of prime
levels, and recent data suggest that animal protein may importance in documenting the major role of environ-
increase levels of IGF-I96,97. mental factors in the aetiology of nearly all major human
cancers. However, from the beginning, epidemiologists
have recognised that firm conclusions about specific
Bladder cancer aetiological factors cannot be based on comparisons of
Cancer of the urinary bladder was estimated to account cancer rates among countries due to potential confounding
for 336,000 cases and 132,000 deaths in 200011. The by the multitude of lifestyle and other environmental
geographic variation in incidence is about 10-fold, with factors that vary geographically. Case-control and pro-
relatively high rates in Western countries7. Smoking spective cohort studies within countries can provide better
increases the risk for bladder cancer6. Studies suggest control of potential confounding variables, because these
that high intakes of fruits and vegetables may reduce risk, factors usually vary less within a geographic region and
but this is not established and more prospective data are they can be measured and controlled for in statistical
needed3,98,99. analyses. Until the last few years, case-control studies
provided the large majority of data on diet and cancer.
Concerns have existed that methodological biases, related
Kidney cancer to both the selection of study participants and the recall of
Cancer of the kidney was estimated to account for 189,000 diet after the diagnosis of cancer, could in some
cases and 91,000 deaths in 200011. The range of circumstances seriously distort the results of case-control
geographic variation in incidence is moderate, with the studies. Now that a number of prospective studies are
highest incidence in Scandinavia and among the Inuit6. providing data on diet and cancer incidence, these
Overweight/obesity is an established risk factor for cancer concerns about the potential for bias in case-control
of the kidney, and may account for up to 30% of kidney studies have been supported because different results have
cancers in both men and women100. There are only limited often been observed, even within the same study
data on the possible role of diet in the aetiology of kidney population8,9. At the time of the 1997 WCRF/AICR review3
cancer, but some studies have observed an increase in risk it was recognised that associations between dietary fat and
with high intakes of meat and dairy products and a risk of breast cancer seen in case-control studies had not
reduced risk with high intakes of vegetables3. been confirmed in prospective studies with substantial
194 TJ Key et al.
statistical power. Similar differences in results have now predict cardiovascular disease incidence, strongly suggest
been observed for fat intake in relation to incidence of that major associations between dietary factors and cancer
colon and lung cancers. A major conclusion of the 1997 risk can be detected. In addition, prospective studies offer
review was that a greater intake of fruits and vegetables the opportunity for repeated measurements of dietary
would substantially reduce risks of a broad range of intake, which can reduce the error in measurement of
cancers and decrease total cancer incidence by approxi- long-term diet. However, because dietary assessments are
mately 20%. However, most of the evidence was derived inevitably imperfect and the size of study populations is
from case-control studies and, again, prospective studies finite, modest but still potentially important associations
have often found no or only weak support for the earlier are usually impossible to exclude entirely. For some
findings. In particular, prospective studies have not specific dietary factors, biochemical measurements can
supported earlier suggestions for strong inverse associ- improve assessments of intake, but for many aspects of
ations between overall intakes of fruits and vegetables and diet such measurements do not exist or enhance precision.
risks of lung and colon cancer, and little relation has been A fundamental challenge in nutritional epidemiology is
observed with breast cancer. Because rates of oral and that foods are complex combinations of thousands of
oesophageal cancer are low in affluent populations, these chemical constituents, thus isolation of the active factors
have not been adequately evaluated in prospective studies. can be difficult or impossible. For this reason, conclusions
Thus, conclusions regarding protective effects of fruits and will be most reliable for foods or food groups, although
vegetables for these cancer sites should be viewed with data on supplement use can assist in evaluating
some caution until confirmed in prospective studies. hypotheses related to specific nutrients. Because dietary
Although prospective cohort studies will often provide behaviours are often associated with other aspects of
the best available evidence regarding diet and cancer lifestyle that could affect cancer risk, studies also need to
relationships, like any study they also have potential be evaluated for the degree to which confounding by such
limitations that should be considered in the interpretation variables has been addressed.
of findings. As in almost all dietary studies, only a part of Because all forms of studies have constraints, in most
the range of possible human intakes can be investigated. situations no single form of evidence will provide
Thus, conclusions need to be limited to the range of diets definitive conclusions regarding diet and cancer relation-
investigated, in part because dose– response relationships ships. Thus, the best conclusions will be based on careful
may be non-linear. As with randomised trials, conclusions and critical evaluation of all forms of evidence.
about findings need to be limited to the period in life or
interval between dietary assessment and cancer diagnosis Summary of the evidence for dietary factors and
that was studied. Adequate precision in measurement of human cancer
dietary intakes is necessary to detect true associations; the Dietary components that we believe to be convincingly or
extensive literature on validity of dietary questionnaires, probably related to the incidence of specific cancers are
and the ability of current dietary assessment methods to summarised in Table 1, together with other dietary factors

Table 1 Diet, nutrition and cancer: levels of evidence

Level of evidence Decrease risk Increase risk

Convincing Physical activity (colon) Overweight and obesity (oesophagus, colorectum,

breast in postmenopausal women, endometrium,
kidney)
Alcohol (oral cavity, pharynx, larynx,
oesophagus, liver, breast)
Aflatoxin (liver)
Chinese-style salted fish (nasopharynx)

Probable Fruits and vegetables (oral cavity, Preserved meat and red meat
oesophagus, stomach, colorectum*) (colorectum)
Physical activity (breast) Salt preserved foods and salt
(stomach)
Very hot (thermally) drinks and
food (oral cavity, pharynx, oesophagus)

Insufficient Fibre, soya, fish, n-3 fatty acids, carotenoids, vitamins Animal fats, heterocyclic amines, polycyclic
B2, B6, folate, B12, C, D, E, calcium, aromatic hydrocarbons, nitrosamines
zinc, selenium, non-nutrient plant constituents
(e.g. allium compounds, flavonoids, isoflavones,
lignans)

* A protective effect of fruit and vegetable intake has been suggested by many case-control studies but has not been supported in several large prospective
studies, suggesting that if a benefit does exist it is likely to be modest.
Diet, nutrition and cancer 195
which were considered to be possibly related to cancer may potentially influence the accumulation of body fat
risk but for which the evidence was considered include high consumption of refined carbohydrate, highly
insufficient. Physical activity is also listed in Table 1 energy-dense food (i.e. high energy content relative to
because it is related to energy balance and overweight/- volume or weight of food), and low fibre intake. However,
obesity; a full discussion of the effects of physical activity evidence based on long-term studies for these effects of
on cancer risk is outside the scope of this paper, therefore, dietary composition is inadequate at present. Thus, at this
we have based our evaluation on the conclusions in the time the appropriate emphasis for weight control appears
IARC Handbook on Weight Control and Physical to be limitation of excessive energy intake from any source
Activity101. and the adoption of adequate daily physical activity.

Dietary factors which convincingly increase risk Alcoholic beverages

Another aspect of diet clearly related to cancer incidence is
Overweight/obesity consumption of alcoholic beverages, which convincingly
The nutritional factor for which the evidence was increases the risk of cancers of the oral cavity, pharynx,
considered most convincing, and for which the quantitative larynx, oesophagus, liver and breast (and probably
impact on overall cancer rates is most important in colorectum). The increase in risk appears to be primarily
populations with Western cancer incidence patterns, is due to alcohol per se rather than specific alcoholic
overweight/obesity. Overweight/obesity is convincingly beverages. Whereas most of the excess risks occurs with
related to risks for cancers of the oesophagus (adenocarci- high alcohol consumption, a small (about 7%) increase in
noma), colorectum, breast (postmenopausal), endome- risk of breast cancer has been observed with approxi-
trium and kidney. Importantly, excess risk of these cancers mately one drink per day. Recent studies suggest that the
increases continuously with greater adiposity and is not excess risk of breast and colon cancer associated with
limited to clinical obesity (BMI over 30 kg/m2). The large alcohol consumption may be concentrated in persons with
increases in endogenous estrogen levels caused by excess low folate intake.
body fat among postmenopausal women probably explain
the higher risks of postmenopausal breast and endometrial Aflatoxin
cancer. The mechanisms for other cancers are less clear, but Food contaminated with aflatoxin convincingly increases
it has been suggested that hyperinsulinaemia may increase the risk of liver cancer. However, this contamination
the risk for colon cancer52,102. The WHO/IARC working occurs mainly in areas where hepatitis viruses are a major
group on weight control and physical activity estimated cause of liver cancer, and the importance of aflatoxin in
that in countries with high rates of cancers related to the absence of hepatitis virus infections (for example, after
overweight, excess body weight (BMI over 25 kg/m2) immunisation) is not clear.
accounts for approximately 39% of endometrial, 25% of
kidney, 11% of colon, 9% of postmenopausal breast cancer
Chinese-style salted fish
and 5% of total cancer incidence82,101. Although these
High intake of Chinese-style salted fish, predominantly
percentages will be lower in populations in some
consumed in some Asian populations, convincingly
developing countries where virally related cancers are
increases the risk of nasopharyngeal cancer.
more important, the rapid increase in overweight/obesity
in developing countries means that cancers due to
Dietary factors which probably reduce risk
overweight/obesity will become increasingly important
world-wide. Fruits and vegetables
We recognise that overweight/obesity, a reflection of Overall, a high intake of fruits and vegetables probably
excessive energy intake, can result from both over- reduces the risks of cancers of the oral cavity, oesophagus,
consumption of energy from food and low expenditure of stomach and colorectum. Previous reviews of diet and
energy as physical activity; the relative importance of these cancer, including the 1997 WCRF/AICR review, have given
two factors can vary among individuals and populations. greater emphasis to increasing fruit and vegetable
Considerable attention has been given to the possibility consumption for cancer prevention, and have included
that the composition of the diet influences the probability cancers of the larynx, lung, pancreas, breast and bladder.
of body fat accumulation and thus, indirectly, risk of At that time, however, the available literature was based
cancer. Although the percentage of energy from dietary fat largely on case-control studies, and subsequent prospec-
has been hypothesised to be an important determinant of tive studies have not supported important protective
body fat and the topic has been controversial, an effects for cancers of the lung and breast, and have
important effect of dietary fat has not been supported in suggested that the reduction in risk for colorectal cancer
randomised trials lasting 1 year or more, and populations may be modest. These discordant results, which add to
consuming low fat/high carbohydrate diets can clearly concerns about the potential for bias in case-control
develop high rates of obesity103. Other dietary factors that studies, also suggest the need for some caution regarding
196 TJ Key et al.
conclusions about intake of fruits and vegetables and the evidence that this relationship was not confounded by H.
risks of oral, oesophageal and stomach cancers, which pylori infection. Notably, stomach cancer rates in the US
have not been adequately examined in large prospective are now very low even though salt intake is not.
studies. Furthermore, none of the dietary studies of
stomach cancer has controlled adequately for infection by Very hot drinks and foods
H. pylori, which is a potential confounding variable. Consumption of very hot drinks and foods typically
Although support for a broad and strong protective consumed in some cultures probably increases risk of
effect of higher fruit and vegetable intake against cancer cancers of the oral cavity, pharynx and oesophagus.
incidence has weakened with the results from recent
studies, modest benefits of increasing fruit and vegetable Cancers not included in table
intake have not been excluded and probably do exist. The For cancers of the pancreas, lung, cervix and ovary, we did
issue of dose –response is important, and some evidence not consider these to be convincing or probable evidence
suggests that a very low intake of fruits and vegetables, of a dietary relationship. In earlier reviews, evidence had
e.g. less than 2 servings or 200 g/d, is related to increases in been considered stronger for a protective effect of fruit
risk compared with higher intakes, but that there may be and vegetable consumption against lung cancer risk.
little additional benefit for intakes higher than about However, more recent prospective studies have found
400 g/d51,81. Also, fruits and vegetables are extremely weaker associations, and residual confounding by
heterogeneous, and it is possible that only specific foods cigarette smoking remains a concern. Although a weak
are related to risk for specific cancers. As examples, some relation with fruit and vegetable intake is possible,
studies have suggested that intake of tomato products, avoidance of smoking is the only effective way to
high in lycopene104, is inversely related to risk of prostate substantially reduce lung cancer rates.
cancer and that cruciferous vegetable intake is inversely Diets very high in starch and low in overall
associated with bladder cancer incidence99. In these same micronutrient intake, consumed by poor populations in
studies, overall fruit and vegetable consumption was not many countries, have been associated with increased risks
associated with cancer risk. Should this be considered as of oral and oesophageal cancers. Isolation of the specific
general support for a protective effect of fruits and micronutrients responsible has been elusive, but an
vegetables, or only for more specific relationships? overall improvement of these poverty-related diets is
Another complexity can arise when a micronutrient in clearly warranted for health in general.
the form of a supplement is shown to be related to cancer
risk. This might be construed as evidence that fruits and Conclusions on dietary factors and cancer
vegetables containing this factor are protective against Since the 1981 Doll and Peto review on diet and cancer
cancer. However, this does not necessarily follow because mortality1, about one third of cancers have generally been
bioavailability of micronutrients in fruits and vegetables thought to be related to dietary factors. More recent
might be low or antagonistic factors might be present. evidence suggests that this number may be too high, but a
Folate may provide such an example as there is now revised quantitative estimate is beyond the scope of this
considerable evidence that higher intake, mainly due to review. Among the diet-related factors, overweight/obesity
use of multiple vitamins, may be related to lower risks of convincingly increases the risks of several common
colorectal and breast cancers. Also, in this situation, use of cancers. After tobacco, overweight/obesity appears to be
supplements and fortified foods could mask a beneficial the most important avoidable cause of cancer in
effect of fruits and vegetables if folate were an important populations with Western patterns of cancer incidence.
protective component of these foods. Among non-smoking individuals in these populations,
avoidance of overweight is the most important strategy for
Preserved meat and red meat cancer prevention.
In many studies, high intakes of preserved meat or red
meat have been associated with increased risk of Policy implications
colorectal cancer, whereas the total fat content of the
diet does not appear to be related to risk. The components Public health policy with respect to nutrition and cancer
of preserved and red meat that might increase colorectal should be based on the best available scientific research.
cancer risk are not established; heterocyclic amines In the previous section, we concluded that few dietary
created by cooking, haeme iron, and specific fatty acids effects on cancer risk have been established. Avoiding
have been proposed as explanations. overweight/obesity, limiting alcohol intake and increasing
physical activity will reduce cancer risk, as will limiting
Salt preserved foods and salt consumption of Chinese-style salted fish and minimising
High intakes of salt-preserved foods and of salt probably dietary exposure to aflatoxin in populations where these
increases the risk of stomach cancer. Convincing evidence dietary factors are important. Risk will probably be
would require confirmation in prospective studies and decreased by increasing the average intake of fruits and
Diet, nutrition and cancer 197
Table 2 Diet, nutrition and cancer: recommendations manuscript. The authors are particularly thankful to Dr
1. Maintain BMI in range of Elio Riboli, International Agency for Research on Cancer
18.5– 25 kg/m2, and avoid weight gain (IARC), Lyon, France, for his valuable help and
in adulthood
2. Engage in regular physical activity contributions during the drafting of the manuscript.
3. Consumption of alcoholic beverages is
not recommended: if consumed, do
References
not exceed 2 units/d (1 unit is equivalent to approximately
10 g of alcohol and is
provided by one glass of 1 Doll R, Peto R. The causes of cancer: quantitative estimates
beer, wine or spirits) of avoidable risks of cancer in the United States today.
4. Minimise exposure to aflatoxin in Journal of the National Cancer Institute 1981; 66:
foods 1191– 308.
5. Chinese-style salted fish should only 2 Miller AB. Diet in cancer prevention. http://www.who.int/
be eaten in moderation, especially ncd/cancer/publications/abstracts/abs9810_05 (accessed
during childhood. Overall consumption of 2001).
salt-preserved foods and salt should 3 World Cancer Research Fund. Food, Nutrition, and the
be moderate
Prevention of Cancer: A Global Perspective. Washington,
6. Have a diet which includes
at least 400 g/d of total fruit and DC: American Institute for Cancer Research, 1997.
vegetables 4 COMA. Nutritional Aspects of the Development of Cancer
7. Meat: those who are not (Report of the Working Group on Diet and Cancer of the
vegetarian are advised to moderate Committee on Medical Aspects of Food and Nutrition
consumption of preserved meat (e.g. Policy). London: The Stationery Office, 1998.
sausages, salami, bacon, ham etc.) 5 Armstrong B, Doll R. Environmental factors and cancer
and red meat (e.g. beef, incidence and mortality in different countries, with special
pork, lamb). Poultry and fish reference to dietary practices. International Journal of
(except Chinese-style salted fish, see Cancer 1975; 15: 617 –31.
5. above) have been studied 6 International Agency for Research on Cancer. Cancer:
and found not to be Causes, Occurrence and Control. IARC Scientific Publi-
associated with increased cancer risk cations No. 100. Lyon: IARC, 1990.
8. Do not consume foods or 7 FerIay J, Bray P, Pisani P, Parkin DM. Globocan 2000:
drinks when they are at Cancer incidence, mortality and prevalence worldwide,
a very hot (scalding hot)
Version 1.0. IARC CancerBase No. 5. http://www-dep.iarc.
temperature
fr/globocan/globocan.html (accessed Nov 3, 2001).
8 Giovannucci E, Stampfer MJ, Colditz GA, et al. A
comparison of prospective and retrospective assessments
vegetables, and by limiting intake of preserved and red of diet in the study of breast cancer. American Journal of
meat, salt preserved foods and salt, and very hot drinks Epidemiology 1993; 137: 502– 11.
and food. Public health policy should, therefore, be 9 Willett WC. Nutritional Epidemiology. New York: Oxford
University Press, 1998.
focussed on these factors (Table 2). If other dietary factors 10 International Agency for Research on Cancer. Principles of
are established to affect cancer risk then these can be Chemoprevention. IARC Scientific Publications No. 139.
added to existing policy frameworks, but at present the Lyon: IARC, 1996.
other potentially important dietary factors are not 11 Parkin DM, Bray F, Ferlay J, Pisani P. Estimating the world
cancer burden: Globocan 2000. International Journal of
sufficiently well supported by evidence, and including Cancer 2001; 94: 153– 6.
them in policy would risk confusing priorities and 12 International Agency for Research on Cancer. IARC
diverting attention from the really important issues. Monographs on the Evaluation of Carcinogenic Risks to
Humans, vol. 44. Alcohol drinking. Lyon: IARC, 1988.
13 Brown LM, Swanson CA, Gridley G, et al. Adenocarcinoma
Acknowledgements of the esophagus: role of obesity and diet. Journal of the
National Cancer Institute 1995; 87: 104 –9.
14 Cheng KK, Sharp L, McKinney PA, et al. A case-control
An earlier version of this paper was prepared as a study of oesophageal adenocarcinoma in women: a
background paper for the Joint WHO/FAO Expert preventable disease. British Journal of Cancer 2000; 83:
Consultation on diet, nutrition and the porevention of 127– 32.
15 International Agency for Research on Cancer. Overweight
chronic diseases (Geneva, 28 January –1 February 2002).
and lack of exercise linked to increased cancer risk. IARC
The authors wish to thank Professor Tony McMichael, Handbooks of Cancer Prevention. vol. 6. Geneva: World
National Centre for Epidemiology and Population Health, Health Organisation, 2002.
Australia National University, Canberra, Australia, Dr Larry 16 World Health Organisation: International Agency for
Research on Cancer. Cancer: Causes, Occurrence and
N.Kolonel, Cancer Research Center of Hawaii, University
Control. Lyon: IARC, 1990.
of Hawaii, Honolulu, USA, Dr Shaw Watanabe, Depart- 17 Steinmetz KA, Potter JD. Vegetables, fruit, and cancer
ment of Nutrition and Epidemiology, Tokyo University of prevention: a review. Journal of the American Dietetic
Agriculture, Tokyo, Japan, and Dr Shoichiro Tsugane, Association 1996; 96: 1027 –39.
18 Sharp L, Chilvers CE, Cheng KK, et al. Risk factors for
Epidemiology & Biostatistics Division, National Cancer squamous cell carcinoma of the oesophagus in women: a
Center Research Institute, Kashiwa-City, Japan, for their case-control study. British Journal of Cancer 2001; 85:
thoughtful and constructive comments on the earlier 1667 – 70.
198 TJ Key et al.
19 Blot WJ, Li JY, Taylor PR, et al. Nutrition intervention trials 8300 deaths among 76,000 men and women in five
in Linxian, China: supplementation with specific vitamin/ prospective studies. Public Health Nutrition 1998; 1:
mineral combinations, cancer incidence, and disease- 33– 41.
specific mortality in the general population. Journal of 38 Howe GR, Aronson KJ, Benito E, et al. The relationship
the National Cancer Institute 1993; 85: 1483 – 92. between dietary fat intake and risk of colorectal cancer:
20 Li JY, Taylor PR, Li B, et al. Nutrition intervention trials in evidence from the combined analysis of 13 case-control
Linxian, China: multiple vitamin/mineral supplementation, studies. Cancer Causes & Control 1997; 8: 215– 28.
cancer incidence, and disease-specific mortality among 39 Burkitt DP. Related disease—related cause? Lancet 1969; 2:
adults with esophageal dysplasia. Journal of the National 1229 –31.
Cancer Institute 1993; 85: 1492 – 8. 40 La Vecchia C. Diet and human cancer: a review. European
21 Yu MC. Nasopharyngeal carcinoma: epidemiology and Journal of Cancer Prevention 2001; 10: 177–81.
dietary factors. In: O’Neill IK, Chen J, Bartsch H, eds. 41 Hague A, Elder DJ, Hicks DJ, Paraskeva C. Apoptosis in
Relevance to Human Cancer of N-nitroso Compounds, colorectal tumour cells: induction by the short chain fatty
Tobacco Smoke and Mycotoxins. IARC Scientific Publi- acids butyrate, propionate and acetate and by the bile salt
cations No. 105. Lyon: 1ARC, 1991, 39– 47. deoxycholate. International Journal of Cancer 1995; 60:
22 International Agency for Research on Cancer. IARC 400– 6.
Monographs on the Evaluation of Carcinogenic Risks to 42 Nagengast FM, Grubben MJ, van Munster IP. Role of bile
Humans, vol 56. Some Naturally Occurring Substances: acids in colorectal carcinogenesis. European Journal of
Food Items and Constituents, Heterocyclic Aromatic Cancer 1995; 31A: 1067 –70.
Amines and Mycotoxins. Lyon: IARC, 1993. 43 Potter JD, Steinmetz K. Vegetables, fruit and phytoestro-
23 Doll R, Peto R. Epidemiology of cancer. In: Weatherall DJ, gens as preventive agents. In: Stewart BW, McGregor D,
Ledingham JGG, Warrell DA, eds. Oxford Textbook of eds. Principles of Chemoprevention. IARC Scientific
Medicine. Oxford: Oxford University Press, 1996, 197–221. Publication No. 139. Lyon: 1ARC, 1996, 61– 90.
24 World Health Organisation. World Health Statistics Annual. 44 Jacobs DRJ, Marquart L, Slavin J, Kushi LH. Whole-grain
http://www.who.int/whosis/ (accessed Nov 10, 2001). intake and cancer: an expanded review and meta-analysis.
25 Helicobacter and Cancer Collaborative Group. Gastric Nutrition and Cancer 1998; 30: 85– 96.
cancer and Helicobacter pylori: a combined analysis of 12 45 Bueno De Mesquita HB, Ferrari P, Riboli E, on behalf of
case control studies nested within prospective cohorts. Gut EPIC. Plant foods and the risk of colorectal cancer in
2001; 49: 347– 53. Europe: preliminary findings. IARC Scientific Publications
26 Palli D. Epidemiology of gastric cancer: an evaluation of Series. 2002; 156: 89–95.
available evidence. Journal of Gastroenterology 2000; 46 Fuchs CS, Giovannucci EL, Colditz GA, et al. Dietary fiber
35(Suppl. 12): 84 –9. and the risk of colorectal cancer and adenoma in women.
27 McCullough ML, Robertson AS, Jacobs EJ, Chao A, Calle EE, New England Journal of Medicine 1999; 340: 169– 76.
Thun MJ. A prospective study of diet and stomach cancer 47 Michels KB, Edward G, Joshipura KJ, et al. Prospective
mortality in United States men and women. Cancer study of fruit and vegetable consumption and incidence of
Epidemiology Biomarkers & Prevention 2001; 10: 1201 – 5. colon and rectal cancers. Journal of the National Cancer
28 Botterweck AA, van den Brandt PA, Goldbohm RA. A Institute 2000; 92: 1740 – 52.
prospective cohort study on vegetable and fruit consump- 48 Schatzkin A, Lanza E, Corle D, et al. Lack of effect of a low-
tion and stomach cancer risk in The Netherlands. American fat, high-fiber diet on the recurrence of colorectal
Journal of Epidemiology 1998; 148: 842 –53. adenomas. Polyp Prevention Trial Study Group. New
29 Correa P, Fontham ET, Bravo JC, et al. Chemoprevention of England Journal of Medicine 2000; 342: 1149 –55.
gastric dysplasia: randomized trial of antioxidant sup- 49 Alberts DS, Martinez ME, Roe DJ, et al. Lack of effect of a
plements and anti-Helicobacter pylori therapy. Journal of high-fiber cereal supplement on the recurrence of color-
the National Cancer Institute 2000; 92: 1881 – 8. ectal adenomas. Phoenix Colon Cancer Prevention
30 Cummings JH, Bingham SA. Diet and the prevention of Physicians’ Network. New England Journal of Medicine
cancer. British Medical Journal 1998; 317: 1636– 40. 2000; 342: 1156 – 62.
31 Hardman AE. Physical activity and cancer risk. Proceeding 50 Bonithon-Kopp C, Kronborg O, Giacosa A, Rath U, Faivre J.
of the Nutrition Society 2001; 60: 107 – 13. Calcium and fibre supplementation in prevention of
32 Gooderham NJ, Murray S, Lynch AM, et al. Assessing colorectal adenoma recurrence: a randomised intervention
human risk to heterocyclic amines. Mutation Research trial. European Cancer Prevention Organisation Study
1997; 376: 53– 60. Group. Lancet 2000; 356: 1300 – 6.
33 Kazerouni N, Sinha R, Hsu CH, Greenberg A, Rothman N. 51 Terry P, Giovannucci E, Michels KB, et al. Fruit, vegetables,
Analysis of 200 food items for benzo[a]pyrene and dietary fiber, and risk of colorectal cancer. Journal of the
estimation of its intake in an epidemiologic study. Food National Cancer Institute 2001; 93: 525– 33.
Chemistry and Toxicology 2001; 39: 423 – 36. 52 McKeown-Eyssen G. Epidemiology of colorectal cancer
34 Bingham SA, Pignatelli B, Pollock JR, et al. Does increased revisited: are serum triglycerides and/or plasma glucose
endogenous formation of N-nitroso compounds in the associated with risk? Cancer Epidemiology Biomarkers &
human colon explain the association between red meat and Prevention 1994; 3: 687–95.
colon cancer? Carcinogenesis 1996; 17: 515 – 23. 53 Franceschi S, Dal Maso L, Augustin L, et al. Dietary glycemic
35 Lund EK, Wharf SG, Fairweather-Tait SJ, Johnson IT. load and colorectal cancer risk. Annals of Oncology 2001;
Oral ferrous sulfate supplements increase the free 12: 173– 8.
radical-generating capacity of feces from healthy volun- 54 Giovannucci E, Rimm EB, Ascherio A, Stampfer MJ, Colditz
teers. American Journal of Clinical Nutrition 1999; 69: GA, Willett WC. Alcohol, low-methionine, low-folate diets,
250–5. and risk of colon cancer in men. Journal of the National
36 Norat T, Lukanova A, Ferrari P, Riboli E. Meat consumption Cancer Institute 1995; 87: 265– 73.
and colorectal cancer risk: a dose– response meta-analysis 55 Glynn SA, Albanes D, Pietinen P, et al. Alcohol consump-
of epidemiological studies. International Journal of tion and risk of colorectal cancer in a cohort of Finnish
Cancer 2002; 98: 241 – 56. men. Cancer Causes & Control 1996; 7: 214– 23.
37 Key TJ, Fraser GE, Thorogood M, et al. Mortality in 56 Giovannucci E, Stampfer MJ, Colditz GA, et al. Multivitamin
vegetarians and non-vegetarians: a collaborative analysis of use, folate, and colon cancer in women in the Nurses’
Diet, nutrition and cancer 199
Health Study. Annals of Internal Medicine 1998; 129: of individual data from 53 epidemiological studies,
517– 24. including 58,515 women with breast cancer and 95,067
57 Choi SW, Mason JB. Folate and carcinogenesis: an women without the disease. British Journal of Cancer
integrated scheme. Journal of Nutrition 2000; 130: 129 –32. 2002; 87: 1234– 45.
58 Potter JD. Colorectal cancer: molecules and populations. 75 Dorgan JF, Baer DJ, Albert PS, et al. Serum hormones and
Journal of the National Cancer Institute 1999; 91: 916–32. the alcohol-breast cancer association in postmenopausal
59 Blount BC, Mack MM, Wehr CM, et al. Folate deficiency women. Journal of the National Cancer Institute 2001; 93:
causes uracil misincorporation into human DNA and 710– 5.
chromosome breakage: implications for cancer and 76 Sellers TA, Kushi LH, Cerhan JR, et al. Dietary folate intake,
neuronal damage. Proceedings of the National Academy alcohol, and risk of breast cancer in a prospective study of
of Sciences of the United States of America 1997; 94: postmenopausal women. Epidemiology 2001; 12: 420– 8.
3290 –5. 77 Smith-Warner SA, Spiegelman D, Adami HO, et al. Types
60 Ma J, Stampfer MJ, Christensen B, et al. A polymorphism of of dietary fat and breast cancer: a pooled analysis of
the methionine synthase gene: association with plasma cohort studies. International Journal of Cancer 2001; 92:
folate, vitamin B12, homocyst(e)ine, and colorectal cancer 767– 74.
risk. Cancer Epidemiology Biomarkers & Prevention 1999; 78 Wu AH, Pike MC, Stram DO. Meta-analysis: dietary fat
8: 825–9. intake, serum estrogen levels, and the risk of breast cancer.
61 Sesink AL, Termont DS, Kleibeuker JH, Van Der Meer R. Journal of the National Cancer Institute 1999; 91: 529– 34.
Red meat and colon cancer: dietary haem-induced colonic 79 Holmes MD, Spiegelman D, Willett WC, et al. Dietary fat
cytotoxicity and epithelial hyperproliferation are inhibited intake and endogenous sex steroid hormone levels in
by calcium. Carcinogenesis 2001; 22: 1653 – 9. postmenopausal women. Journal of Clinical Oncology
62 Baron JA, Beach M, Mandel JS, et al. Calcium supplements 2000; 18: 3668– 76.
and colorectal adenomas. Polyp Prevention Study Group. 80 Key TJ, Allen NE. Nutrition and breast cancer. The Breast
Annals of the New York Academy of Sciences 1999; 889: 2001; 10(Suppl. 3): 9– 13.
138– 45. 81 Smith-Warner SA, Spiegelman D, Yaun SS, et al. Intake of
63 International Agency for Research on Cancer. IARC fruits and vegetables and risk of breast cancer: a pooled
Monographs on the Evaluation of Carcinogenic Risks to analysis of cohort studies. Journal of the American Medical
Humans, vol 59. Hepatitis Viruses. Lyon: IARC, 1994. Association 2001; 285: 769– 76.
64 Saracco G. Primary liver cancer is of multifactorial origin: 82 Bergstrom A, Pisani P, Tenet V, Wolk A, Adami HO.
importance of hepatitis B virus infection and dietary Overweight as an avoidable cause of cancer in Europe.
aflatoxin. Journal of Gastroenterology and Hepatology International Journal of Cancer 2001; 91: 421– 30.
1995; 10: 604 – 8. 83 Key TJ, Pike MC. The dose –effect relationship between
65 Michaud DS, Giovannucci E, Willett WC, Colditz GA, ‘unopposed’ oestrogens and endometrial mitotic rate: its
Stampfer MJ, Fuchs CS. Physical activity, obesity, height, central role in explaining and predicting endometrial
and the risk of pancreatic cancer. Journal of the American cancer risk. British Journal of Cancer 1988; 57: 205– 12.
Medical Association 2001; 286: 921 – 9. 84 International Agency for Research on Cancer. IARC
66 Bjelke E. Dietary vitamin A and human lung cancer. Monographs on the Evaluation of Carcinogenic Risks to
International Journal of Cancer 1975; 15: 561 –5. Humans, vol 64. Human Papillomaviruses. Lyon: IARC,
67 Hennekens CH, Buring JE, Manson JE, et al. Lack of effect 1995.
of long-term supplementation with beta carotene on the 85 Banks E, Beral V, Reeves G. The epidemiology of epithelial
incidence of malignant neoplasms and cardiovascular ovarian cancer: a review. International Journal of
disease. New England Journal of Medicine 1996; 334: Gynecological Cancer 1997: 425– 38.
1145 –9. 86 Schuurman AG, van den Brandt PA, Dorant E, Goldbohm
68 Omenn GS, Goodman GE, Thornquist MD, et al. Effects of a RA. Animal products, calcium and protein and prostate
combination of beta carotene and vitamin A on lung cancer cancer risk in The Netherlands Cohort Study. British
and cardiovascular disease. New England Journal of Journal of Cancer 1999; 80: 1107– 13.
Medicine 1996; 334: 1150 –5. 87 Chan JM, Stampfer MJ, Ma J, Gann PH, Gaziano JM,
69 The Alpha-Tocopherol Beta Carotene Cancer Prevention Giovannucci EL. Dairy products, calcium, and prostate
Study Group. The effect of vitamin E and beta carotene on cancer risk in the Physicians’ Health Study. American
the incidence of lung cancer and other cancers in male Journal of Clinical Nutrition 2001; 74: 549– 54.
smokers. New England Journal of Medicine 1994; 330: 88 Michaud DS, Augustsson K, Rimm EB, Stampfer MJ, Willett
1029 –35. WC, Giovannucci E. A prospective study on intake of
70 Feskanich D, Ziegler RG, Michaud DS, et al. Prospective animal products and risk of prostate cancer. Cancer Causes
study of fruit and vegetable consumption and risk of lung & Control 2001; 12: 557– 67.
cancer among men and women. Journal of the National 89 Heinonen OP, Albanes D, Virtamo J, et al. Prostate
Cancer Institute 2000; 92: 1812– 23. cancer and supplementation with alpha-tocopherol
71 Voorrips LE, Goldbohm RA, Verhoeven DT, et al. Vegetable and beta-carotene: incidence and mortality in a controlled
and fruit consumption and lung cancer risk in the trial. Journal of the National Cancer Institute 1998; 90:
Netherlands Cohort Study on diet and cancer. Cancer 440– 6.
Causes & Control 2000; 11: 101 – 15. 90 Clark LC, Dalkin B, Krongrad A, et al. Decreased incidence
72 Key TJ, Verkasalo PK, Banks E. Epidemiology of breast of prostate cancer with selenium supplementation: results
cancer. Lancet Oncology 2001; 2: 133 – 40. of a double-blind cancer prevention trial. British Journal of
73 Collaborative Group on Hormonal Factors in Breast Cancer. Urology 1998; 81: 730– 4.
Breastfeeding and breast cancer: collaborative reanalysis of 91 Kristal AR, Cohen JH. Invited commentary: tomatoes,
individual data from 47 epidemiological studies in 30 lycopene, and prostate cancer. How strong is the evidence?
countries, including 50,302 women with breast cancer and American Journal of Epidemiology 2000; 151: 124– 7.
96,973 women without the disease. Lancet 2002; 360: 92 Eaton NE, Reeves GK, Appleby PN, Key TJ. Endogenous
187– 95. sex hormones and prostate cancer: a quantitative review of
74 Collaborative group on hormonal factors in breast cancer. prospective studies. British Journal of Cancer 1999; 80:
Alcohol, tobacco and breast cancer-collaborative reanalysis 930– 4.
200 TJ Key et al.
93 Gann PH, Hennekens CH, Ma J, Longcope C, Stampfer MJ. Swedish food survey reported that common foods such as
Prospective study of sex hormone levels and risk of
prostate cancer. Journal of the National Cancer Institute
fried potatoes, potato crisps (chips), bread and breakfast
1996; 88: 1118 –26. cereals contain moderately high levels of acrylamide2.
94 Chan JM, Stampfer MJ, Giovannucci E, et al. Plasma insulin- Subsequent studies showed that acrylamide is generated
like growth factor-I and prostate cancer risk: a prospective when the amino acid asparagine is heated above 1008C in
study. Science 1998; 279: 563 –6.
95 Stattin P, Bylund A, Rinaldi S, et al. Plasma insulin-like the presence of sugars; potatoes and cereals are both rich in
growth factor-I, insulin-like growth factor-binding proteins asparagine3,4. Moderate acrylamide levels in common
and prostate cancer risk: a prospective study. Journal of the foods have now been reported from several countries, and
National Cancer Institute 2000; 92: 1910 –7. these findings have raised concern that dietary acrylamide
96 Allen NE, Appleby PN, Davey GK, Key TJ. Hormones and
diet: low insulin-like growth factor-I but normal bioavail- may increase cancer risk in humans5. The intake of
able androgens in vegan men. British Journal of Cancer acrylamide from food may be several hundred-fold lower
2000; 83: 95– 7. than the lowest dose at which adverse effects have been
97 Heaney RP, McCarron DA, Dawson-Hughes B, et al. Dietary observed in rats6, and at the time of writing ( January 2003)
changes favorably affect bone remodeling in older adults.
Journal of the American Dietetic Association 1999; 99: the only information concerning cancer risk in humans
1228– 33. comes from a case-control study in Sweden7. This study
98 Zeegers MP, Goldbohm RA, van den Brandt PA. Consump- estimated acrylamide intake from the reported intake of 10
tion of vegetables and fruits and urothelial cancer
foods, and observed no evidence that increasing intake of
incidence: a prospective study. Cancer Epidemiology
Biomarkers & Prevention 2001; 10: 1121 – 8. food sources of acrylamide was associated with the risk for
99 Michaud DS, Spiegelman D, Clinton SK, Rimm EB, Willett cancers of the colorectum, kidney or bladder. Although
WC, Giovannucci EL. Fruit and vegetable intake and these first results are moderately reassuring, more data are
incidence of bladder cancer in a male prospective
cohort. Journal of the National Cancer Institute 1999; 91:
urgently needed from studies using questionnaires
605–13. validated for estimating acrylamide intake, or biomarkers
100 Bergstrom A, Hsieh CC, Lindblad P, Lu CM, Cook NR, Wolk of acrylamide exposure.
A. Obesity and renal cell cancer—a quantitative review.
British Journal of Cancer 2001; 85: 984 – 90.
101 International Agency for Research on Cancer. IARC Further reading
Handbooks of Cancer Prevention. Vol 6. Weight Control
and Physical Activity. Lyon: IARC, 2002. 1 International Agency for Research in Cancer. Acrylamide.
102 Giovannucci E. Insulin and colon cancer. Cancer Causes & IARC Monographs on the Evaluation of Carcinogenic Risks to
Control 1995; 6: 164 – 79. Humans 1994; 60: 389–433.
103 Willett WC. Is dietary fat a major determinant of body fat? 2 Tarake E, Rydberg P, Karlsson P, Eriksson S, Törnqvist M.
American Journal of Clinical Nutrition 1998; 67: Analysis of acrylamide, a carcinogen formed in heated
556S– 62S. foodstuffs. Journal of Agricultural and Food Chemistry 2002;
104 Giovannucci E, Clinton SK. Tomatoes, lycopene, and 50: 4998 – 5006.
prostate cancer. Proceedings of the Society for Experimental 3 Mottram DS, Wedzicha BL, Dodson AT. Food chemistry:
Biology and Medicine 1998; 218: 129 – 39. acrylamide is formed in the Maillard reaction. Nature 2002;
419: 448–9.
4 Stadler RH, Blank I, Varga N, et al. Food chemistry: acrylamide
Note added in press—acrylamide from Maillard reaction products. Nature 2002; 419: 449– 50.
5 Food and Agriculture Organization/ World Health Organiz-
ation. FAO/WHO Consultation on the Health Implications or
Acrylamide is a chemical classified by the International Acrylamide in Food: Summary Report. Geneva: WHO, 2002.
Agency for Research on Cancer as a probable human 6 Shaw I, Thomson B. Acrylamide food risk. Lancet 2003; 361:
carcinogen1. This classification is based on experiments in 434.
7 Mucci LA, Dickman PW, Steineck G, Adami H-O, Augustsson
rats and mice, which show that high doses of acrylamide K. Dietary acrylamide and cancer of the large bowel, kidney,
increase the risk for several cancers, including cancers of and bladder: absence of an association in a population-based
the lung, mammary gland, thyroid and oral cavity. In 2002, a study in Sweden. British Journal of Cancer 2003; 88: 84– 9.