Diet and Cancer

Professor Christine Edwards

 Cancer mortality in the UK
(2005)

20 most common causes of death from cancer in the UK

http://info.cancerresearchuk.org/
 Stages of carcinogenesis
Permanent change in DNA
Initiation
Very common

Environmental agent increases chance of expression of

Promotion damaged DNA reversible

Progression Further genetic change into cancer cells

 NORMAL CELL
DNA removal DNA
repair
Initiation DNA damage

Initiated cell
Promotion Spontaneous remission
Growth inhibitors
Premalignant cell
Anti-promoters

Progression
CANCER
 Problems in studying diet and
cancer
• Cancer can occur at many different site
• Many different causes
• Aetiology is multistage
• Development of cancer is slow
• Diet and other environmental agents could act at
any stage
• May act differently at each stage
• May enhance or inhibit cancer
 What measurements?

• Dietary intake when

• Disease occurrence – severity?

 At what point was diet important?

• Prenatal
• Weaning
• Childhood
• Accumulated lifelong
• At critical period
– How can we assess these?
How can we assess the evidence?
• Many different cancers
• Many different diets
• Different foods and nutrients
• Different eating patterns
• Interactions with other lifestyle factors
• Different genetic influences
World Cancer Research Fund and
IoM
• Carried out extensive systematic review
• Decided whether evidence was
– convincing
– Probable
– Limited suggestive
– Limited no conclusion
– Substantial effect unlikely
• Based on strict criteria
 WCRF 2007
convincing
• Evidence from more than one study type
• At least 2 independent cohort studies
• No substantial un explained heterogeneity
• Good quality studies to exclude random or
systematic error
• Plausible biological gradient – dose response
• Strong and plausible experimental evidence
– Human or animal studies
 WRCF 2007
Probable
• Evidence from at least 2 independent cohort
studies or at least 5 case control studies
• No substantial un explained heterogeneity
• Good quality studies to exclude random or
systematic error, confounding
• Evidence for biological plausibility
 Limited suggestive
WRCF 2007
• Evidence from at least 2 independent cohort
studies or at least 5 case control studies
• The direction of effect is generally
consistent though may be some unexplained
heterogeneity
• Evidence for biological plausibility
 Limited no conclusion
WRCF 2007
• Evidence is so limied no firm conclusion
may be made
• Few good studies
• Inconsistent effects or directions of effects
• Lack of adjustment for confounders
Substantial effect on risk unlikely
WRCF 2007
• Evidence from more than one study type
• At least 2 independent cohorts
• Summary estimate of effect close to 1 for high
versus low
• No substantial unexplained heterogeneity
• Good quality studies
• Absence of demonstrable dose response
• Absence of strong and plausible experimental
evidence
 Body fatness WRCF 2007
Decreases risk Increase risk Cancer site
convincing Body fatness Oesphagus
Pancreas, colorectum,
breast (post meno)
endometrium, kidney

Abdominal fat colorectum

probable Body fatness- Body fat Gallbladder

breast pre- Abdominal fat Pancreas, breast (post
menopausal meno)
Endometrium
Adult weight gain Breast (post meno)
Limited Body fatness Liver
suggestive Low body fatness lung
 Physical activity 2007 WRCF
Decreases risk Increases risk
Convincing colon

Probable Breast (post

menopausal)
endometrial

Limited suggestive Lung

Pancreas
Breast
(premenopausal)
Substantial effect none
inlikely
 Mechanisms for fat and cancer
• Sex hormones
• Immune function
• Cytokines
• Growth factors
• Prostaglandins
• Lipid peroxidation
• Membrane fluidity
• Expression of ocogenes
• Increased faecal bile acids
• Increased secondary bile acids
 Fat, Cholesterol and colorectal cancer

• Difficult to separate fat, energy and BMI

• Case control studies
– Increased fat increased colorectal cancer
– But later case control studies and meta-analysis
and cohort studies do not support effects.
 Fat and colorectal cancer
• Japanese migrants
– Increased fat increase colorectal cancer

• Animal studies very suggestive of role of fat

• Meta-analysis of 13 prospective cohort studies

– (Lui et al 2011 Eur J Nutrition 50 173-84)
– No relation between dietary fat and colorectal cancer
(Lui et al 2011 Eur J Nutrition 50 173-84)
 Fat and oils WRCF 2007
Evidence Decreases risk Increases risk Cancer site

convincing
probable

Limited Total fat lung,, breast , (post

suggestive menopausal)
Foods containing Colorectal
animal fat
butter lung

Substantial None identified

effect
unlikely
 Protein and Meat
• Case control studies
– No association with bowel cancer (Bingham 1990)
• Boston Nurses study
– Red meat increases colorectal cancer (Willett 1990, Giovanucci
1994)
• Prospective cohort
– 6,000 non meat eaters
– 5000 meat eaters
– 40% reduction in cancer deaths in vegetarians and fish eaters
– But confounding factors increased vegetables and fruit.
 Meat and Cancer
• WCRF Increased risk convincing for red
meat and processed meat
– Not white meat
– What is red meat??
– Many studies on processed meats
• Sausages– what is a sausage
– Is pork a red meat?
– Is chicken a fish?
 Meat, poultry, Fish, Eggs
Evidence Decreases risk site Increases Risk site

Convincing Red Meat Colorectal

Processed meat Colorectal
Probable Cantonese style Nasopharynx
salted fish
Limited Fish Colorectum Red Meat, Oesophagus
suggestive lung,
Foods
pancreas
containing vit D colorectum Endometrium
Processed meat Oesophagus
lung, stomach
prostate
Foods cont iron Colorectal
Smoked foods Stomach
Grilled or stomach
barbecued animal
foods
 Diet and cancer
Red meat and colorectal cancer; cohort studies
 Possible carcinogens associated
with red meat
• Heterocyclic amines
• N nitroso compounds (NOC)
• Iron
• Secondary bile acids
• Stearic acid
 Heterocyclic amines (HCA’s)
• Produced at high temperature during
cooking
• IQ compounds
– Amino imidazo-quinolines
– Amino imidazo quinoxalines
• PhIP compunds
– Amino imidazo pyridines
 Heterocyclic amines (HCA’s)
• Formed from creatine or creatinine, certain amino acids
and sugars during frying and grilling
• Brief microwaving reduces yield by removing creatine
• Some people more susceptible than others
• Different polymorphic enzymes
– Cytochorme P4501A (CYP1A2) and And N acetyltransferase
Type 2 (NAT2)
• Fast CYP1A2 and NAT2 phenotypes may be more
susceptible to colonic cancer because they rapidly activate
HCAs
 But no link with breast cancer
Delfino RJ et al 2000 Carcinogenesis 21 607 -615
• Case control study
• 114 cases 280 controls (benign breast
disease)
• Looked at NAT2 and heterocyclic amines
• Found no association with red meat but
protection with white meat.
 WRCF 2007 Cereals, starchy roots
tubers plantains and cancer
Evidence Decreases site Increases site
risk risk

Convincing aflatoxins liver

probable Foods containing colorectum

dietary fibre

Limited suggestive Foods containing Wholegrain

dietary fibre oesophagus

Substantial effect on None identified

risk unlikely
 Colorectal cancer
• Starch
– Cross-cultural studies show strong correlation.
• High starch low cancer risk
– Case control studies
• No relationship with cancer
• Inverse association with adenomas
– Cohort studies
• Unclear
– Animal and in vitro studies
• Resistant starch protective
 Colorectal cancer
• Dietary fibre
– Cross cultural/ correlation studies
• High dietary fibre low cancer risk
– Case control studies
• Meta-analysis (Howe et al 1992) dietary fibre protective
– Cohort studies
• Not supportive (Willet et al 1992, 1999)
• More protective (EPIC study) Bingham et al 2003 Lancet
– 40% reduction in risk for doubling in fibre intake
 Diet and cancer
Dietary fibre and colorectal cancer; cohort studies

World Cancer Research Fund

Diet and cancer report 2007
http://www.dietandcancerreport.org/
 Problems with studies in dietary fibre

• How do we measure fibre intake

– Vegetables
– Cereals
– Pulses
– Crude fibre, detergent fibre,
– Dietary fibre
– Non starch polysaccharides
– Insoluble fibre
– Soluble fibvre
– What about resistant starch, oligosaccharides, etc
 Fibre and animals studies
• No good models
• When do you give the fibre in relation to the
carcinogen?
• Soluble fibres increase tumour yield
• Insoluble fibres had little or protective
effects
• SCFA increase tumour yield.- feeds tumour.
 Vegetables and fruits etc and cancer risk
Evidence Decreases risk Cancer site Increases
risk
Convincing

Probable Mouth pharynx oesaphagus

Non starchy vegetables stomach

Allium garlic fruits Stomach, colorectum,

Mouth pharynx oesaphagus
stomach lung
Foods containing folate Pancreas
Foods containing carotenoids Mouth pharynx larynx lung
Foods containing lycopene prostate
Foods containing vit c Oesphagus
Foods containing selenium prostate
Substantial
effect unlikely
 Vitamin intervention trials
• Doses often much above that possible by diet alone

• Physicians health study - 22,000 US males

• Alpha Tocopherol Beta carotene prevention study (ATBCCPS)
29,000 Finnish Male smokers
• Beta Carotene and Retinol Efficacy Trial (CARET).
– 18,000 US male and female smokers and asbestos workers/
• Womens Health study
– 40,000 US female Health professionals
• Chinese – Linxian study
– 30,000 rural Chinese men and women.
• In well nourished populations no benefit of
beta carotene or Vit E
• Beta carotene may be harmful

• In less nourished populations

– Supplements containing beta-carotene, Vit E
and selenium but not vit C decreased cancer
risk
• Vit E
– ATBCCPS
• 50mg alpha tocopherol 34% decrease in prostate
cancer; no effect lung cancer
– Linxian study
• 15mg betacarotene, 30m alpha tocopherol 50mg
selenium or 120 mg vit C 30ug molybdenum
• Overall mortality all causes, cancer stomach cancer
decreased with first mixture no effect of second
mixture.
 Beta carotene

• Four large trials with supplements

– No benefits
– Two increased risks and overall mortality
• ATBCCPS 1994
– 20mg beta-carotene aqfter 5-8 years 18% higher lung
cancer, 8% higher mortality
• Caret 1996
– 30mg beta carotene 25,000 IU retinol
– 28% increase in lung cancer 17% increase in death
• WHD
– Terminated beta carotene trial after above results
 With precancerous conditions
• Beta carotene, vit C and vit E
• No effect in recurrence of coloractal
adenoma
• Vit E no effect on benign breast disease
• Beta carotene and Vit E may decrease oral
leucoplakia
 Flavonoids
• Don’t really know enough yet about
bioavailability to test

• Potentially useful

• Several mechanisms
– Anti oxidants, and direct effects on enzymes
and cells.
 Dietary patterns
• Difficult to see clear effects of single
nutrients
• May be interactions between nutrients and
food
• Therefore try and see the effect of dietary
patterns
P E. Miller J E. Muscat P Lazarus T J. Hartman
Dietary Patterns and Colorectal Adenoma and
Cancer Risk: A Review of the Epidemiological
Evidence
Nutrition and Cancer, 62(4), 413–424 2010

• 16 papers identified
• Food patterns analysed in different ways by
principle component analysis
2010
Diet, body size, physical activity and risk of prostate cancer: An umbrella review
of the evidence Eur J Cancer 2016 69 61-9
G Markozannes , I Tzoulaki D Karli E Evangelou E Ntzani MJ. Gunter T Norat J P. Ioannidis K K. Tsilidis

• Literature of diet, adiposity and physical activity with risk of prostate cancer
critically appraised.
•No associations were graded with strong evidence.
•Large and unbiased effects of these factors on prostate cancer are unlikely.
• Existing literature on relationship between diet, body size, physical activity and
prostate cancer risk summarised by the World Cancer Research Fund Continuous
Update Project (CUP).

• Robustness of evidence evaluated using several criteria addressing evidence

strength and validity, statistical significance of the random effects summary
estimate and of the largest study in a meta-analysis, number of prostate cancer
cases, between-study heterogeneity, 95% prediction intervals, small-study effects
bias, excess significance bias and sensitivity analyses with credibility ceilings.

• 248 meta-analyses extracted from CUP, studied associations of 23 foods, 31

nutrients, 8 indices of body size and 3 indices of physical activity with risk of
total prostate cancer development, mortality or cancer development by stage and
grade. .
• Of 176 meta-analyses using a continuous scale to measure the exposures,
no association presented strong evidence by satisfying all the criteria

• Only the association of height with total prostate cancer incidence and
mortality presented highly suggestive evidence with a 4% higher risk per 5
cm greater height (95% confidence interval, 1.03, 1.05).

• Associations for body mass index, weight, height, dietary calcium and
spirits intake were supported by suggestive evidence.

• Overall, the association of diet, body size, physical activity and prostate
cancer has been extensively studied, but no association was graded with
strong evidence.