Diet, nutrition and cancer: public, media and scientific confusion

Abstract

Ten years ago, steps to prevent cancer were known. Intake of vegetables and fruits was clearly associated with reduced cancer risk, antioxidants were a crucial ingredient in lowering cancer risk and overweight and obesity was the most important risk factor for cancer in nonsmokers. What were clear messages then have been confused by subsequent findings and media reports. It is useful to stop and take stock of the current situation.The World Cancer Research Fund (WCRF) launched a second substantial compendium on ‘Food, Nutrition, Physical Activity, and the Prevention of Cancer: a Global perspective’ [1.World Cancer Research Fund Report Food, Nutrition and the Prevention of Cancer: a Global Perspective.2007Google Scholar]. According to the accompanying press release [2.WCRF Press ReleaseLandmark Report: Excess Body Fat Causes Cancer http://www.wcrf-uk.org/press_media/word_docs/report.doc (8 November 2007, date last accessed)Google Scholar], the Report ‘includes 10 recommendations … that represent the most definitive and authoritative advice that has ever been available on how the general public can prevent cancer’. The words ‘tobacco’ or ‘smoking’ are not mentioned once in the press release.In presenting its summary and recommendations, the report implicitly downplays the key importance of tobacco smoking in cancer causation. Contrary to that stated in the press release (the best advice for cancer prevention is to avoid weight gain [2.WCRF Press ReleaseLandmark Report: Excess Body Fat Causes Cancer http://www.wcrf-uk.org/press_media/word_docs/report.doc (8 November 2007, date last accessed)Google Scholar]), avoiding tobacco smoking and use of tobacco in other forms is the single best advice to reducing cancer risk as one-third of cancer deaths in high-income countries is attributable to tobacco use [3.Danaei G. Vander Hoorn S. Lopez A.D. et al.Comparative Risk Assessment Collaborating Group (Cancers). Causes of cancer in the world: comparative risk assessment of nine behavioural and environmental risk factors.Lancet. 2005; 366: 1784-1793Abstract Full Text Full Text PDF PubMed Scopus (952) Google Scholar, 4.International Agency for Research on Cancer (IARC) Monographs on the Evaluation of Carcinogenic Risks to Humans. Volume 83. Tobacco Smoke and Involuntary Smoking. IARC, Lyon2004Google Scholar, 5.Doll R. Peto R. The Causes of Cancer. Oxford University Press, Oxford1981Google Scholar]. Failure to include ‘stop smoking’ and ‘avoid exposure and exposing others to second-hand smoke’ among the 10 key recommendations undermines the most important message in cancer control. The ‘best advice’ also fails to mention the importance of a variety of established cancer risk factors including sun behaviour, occupational exposures, chronic infections and use of exogenous hormones.A decade after its original report [6.World Cancer Research Fund/American Institute for Cancer Research Food, Nutrition, Physical Activity, and the Prevention of Cancer: a Global Perspective. AICR, Washington, DC1997Google Scholar], the current report [1.World Cancer Research Fund Report Food, Nutrition and the Prevention of Cancer: a Global Perspective.2007Google Scholar] presents considerably weaker conclusions on the basis of the dilution of the strength of the evidence linking specific components of diet and nutrition to cancer risk. This is particularly true for the fruits and vegetables for which the strength of evidence for a protective effect of high intakes against several common epithelial cancers was downgraded from ‘convincing’ in the first WCRF report [6.World Cancer Research Fund/American Institute for Cancer Research Food, Nutrition, Physical Activity, and the Prevention of Cancer: a Global Perspective. AICR, Washington, DC1997Google Scholar] to ‘probable’ in this second report [1.World Cancer Research Fund Report Food, Nutrition and the Prevention of Cancer: a Global Perspective.2007Google Scholar].When considering the importance and significance of diet and, more generally, nutrition on cancer risk, it is essential to separate ‘what we know’ from ‘what we think we know’ from ‘what we believe’ and to understand ‘what we do not know’. The WCRF report represents a comprehensive summary of the information necessary to evaluate the current state-of-knowledge, and the associated certainty, of this situation.‘We believe’ that diet and nutrition play a key role in determining the risk of cancer. It is difficult to ignore the rapid changes in the risk of cancers at sites such as the colorectal and breast among groups of migrants from low-risk countries to high-risk countries in this regard. Other known risk factors of cancer are unable to explain such rapid changes, and diet and other nutritional factors are the most plausible determinants of such changes.‘We think we know’ or, more accurately, ‘we thought we knew’ that a high-fat diet and low consumption of fruits, vegetables and fibres were associated with increased risks of common cancers. However, faith in the cancer prevention properties of fruits and vegetables began to crack when all the available evidence was critically reviewed by an International Agency for Research on Cancer (IARC) Working Group [7.IARC Handbooks on Cancer Prevention. Vol. 8. Fruit and Vegetables. IARC, Lyon, France2003Google Scholar]. Subsequently, it has crumbled as major analyses of prospective studies have continued to demonstrate consistently a lack of association between intake of fruits and vegetables and risk of several cancers. This major change in classification of one the few agents classified by WCRF in the category of strongest evidence in 1997 casts doubt on the rationale to classify ‘convincing’ to the evidence linking high meat intake to colorectal cancer risk in the current report [1.World Cancer Research Fund Report Food, Nutrition and the Prevention of Cancer: a Global Perspective.2007Google Scholar]. This also raises questions about the evaluation process and about the robustness of the classification system.The IARC Monographs Programme, which adopts comparable categories to evaluate the strength of evidence of carcinogenicity of different classes of agents, has evaluated >900 agents with ∼100 classified as having ‘sufficient’ evidence of carcinogenicity in humans [8.IARC. Overall Evaluations of Carcinogenicity to Humans http://monographs.iarc.fr/ENG/Classification/crthgr01.php (8 November 2007, date last accessed)Google Scholar], somewhat equivalent to convincing evidence in WCRF terminology. Despite an extensive work of reassessment of many sufficient carcinogens by IARC as the amount of evidence increases, none has ever been downgraded to a lower category of evidence.In view of the fragile grounds on which the conclusions of WCRF report on diet and cancer are based on, the information to the media should have been more cautious, which would probably have prevented confusion of messages disseminated to the public which have left a clear indication that cancer prevention depends on stopping eating bacon and avoiding drinking sodas [9.Blame it on the bacon. The Guardian. http://www.guardian.co.uk/commentisfree/story/0, 2204021,00.html (8 November 2007, date last accessed)Google Scholar, 10.Is anything safe to eat? Cancer report adds bacon, ham and drink to danger listDaily Mail online http://www.dailymail.co.uk/pages/text/print.html?in_article_id=490845&in_page_id= (8 November 2007, date last accessed)Google Scholar]. Clearly there is confusion in the media which is being transmitted to the general public.Evidence from animal and physiological experiments, although with conflicting evidence from observational studies, led to antioxidants being widely considered as one of the key protective elements in the diet, being found commonly in vegetables and fruits, and gave rise to a series of chemopreventive trials. A recent Cochrane Review has thrown light on the use of antioxidant supplementation on mortality and found no evidence to support antioxidant supplementation (e.g. vitamin A, beta-carotene, vitamin C, vitamin E, selenium) for either primary or secondary prevention [11.Bjelakovic G. Nikolov D. Gluud L.L. et al.Antioxidant supplements for prevention of mortality in healthy participants and patients with various diseases.Cochrane Database Syst Rev. 2008; (Art. No. CD007176)Crossref Scopus (385) Google Scholar]. Indeed, vitamin A, beta-carotene and vitamin E may increase mortality [11.Bjelakovic G. Nikolov D. Gluud L.L. et al.Antioxidant supplements for prevention of mortality in healthy participants and patients with various diseases.Cochrane Database Syst Rev. 2008; (Art. No. CD007176)Crossref Scopus (385) Google Scholar].The value of this systematic review is that the conclusions are on the basis of the results of 67 randomised trials involving 232 500 participants randomised to antioxidant supplementation or placebo. The conclusions include the recommendation that more research needs to be done, perhaps focusing on vitamin C and selenium, rather than on vitamin A, beta-carotene and vitamin E.Vitamin D supplementation has been shown, in a meta-analysis, to reduce mortality in the arm randomised to the active intervention [12.Autier P. Gandini S. Vitamin D supplementation and total mortality: a meta-analysis of randomized controlled trials.Arch Intern Med. 2007; 167: 1730-1737Crossref PubMed Scopus (979) Google Scholar]. Again, more work in this area appears justified although it is premature to take action at this point.Renehan et al. [13.Renehan A.G. Tyson M. Egger M. et al.Body-mass index and incidence of cancer: a systematic review and meta-analysis of prospective observational studies.Lancet. 2008; 371: 536-537Abstract Full Text Full Text PDF PubMed Scopus (3595) Google Scholar] recently produced a systematic review and a meta-analysis to shed light on the strength of the association between excess bodyweight, expressed as body mass index (BMI) and the incidence of different types of cancer and to investigate differences in these associations between sex and ethnic groups. Renehan et al [13.Renehan A.G. Tyson M. Egger M. et al.Body-mass index and incidence of cancer: a systematic review and meta-analysis of prospective observational studies.Lancet. 2008; 371: 536-537Abstract Full Text Full Text PDF PubMed Scopus (3595) Google Scholar] carried out random-effects meta-analyses and meta-regressions of study-specific incremental estimates to determine the risk of cancer associated with a 5-kg/m2 increase in BMI.In all, 282 137 incident cases of cancer were included in this analysis. In men, a 5-kg/m2 increase in BMI was strongly associated with oesophageal adenocarcinoma (RR 1.52, P < 0.0001) and with thyroid (RR 1.33, P = 0.02), colon (RR 1.24, P < 0.0001) and renal (RR 1.24, P < 0.0001) cancers. In women, strong associations were recorded between a 5-kg/m2 increase in BMI and endometrial (RR 1.59, P < 0.0001), gallbladder (RR 1.59, P = 0.04), oesophageal adenocarcinoma (RR 1.51, P < 0.0001) and renal (RR 1.34, P < 0.0001) cancers. Weaker positive associations (RR < 1.20) were noted between increased BMI and rectal cancer and malignant melanoma in men; postmenopausal breast, pancreatic, thyroid and colon cancers in women and leukaemia, multiple myeloma and non-Hodgkin’s lymphoma in both sexes. Associations were stronger in men than in women for colon cancer (P < 0.0001). Associations were generally similar in studies from North America, Europe and Australia, and the Asia-Pacific region, but stronger associations were recorded in Asia-Pacific populations between increased BMI and premenopausal (P = 0.009) and postmenopausal (P = 0.06) breast cancers.While the authors concluded that an increased BMI is associated with increased risk of common and less common malignancies, the relative risks are small and the increment (5 kg or 11 pounds per m2) is large.In France, it is estimated that the attributable fraction for cancer death in the year 2000 was 24% for tobacco smoking and 2% for obesity and overweight [14.IARC Working GroupAttributable Causes of Cancer in France in the Year 2000. IARC Working Group Reports Volume 3. Lyon: IARC 2007Google Scholar]. These figures are consistent with other estimates of cancer deaths in high-income countries of 29% attributed to tobacco smoking and 3% to obesity and overweight [3.Danaei G. Vander Hoorn S. Lopez A.D. et al.Comparative Risk Assessment Collaborating Group (Cancers). Causes of cancer in the world: comparative risk assessment of nine behavioural and environmental risk factors.Lancet. 2005; 366: 1784-1793Abstract Full Text Full Text PDF PubMed Scopus (952) Google Scholar]. Most recently, it has been estimated that each year, 6000 cancer cases out of 120 000 in women in UK are related to overweight and obesity [15.Reeves G. Pirie K. Beral V. et al.Cancer incidence and mortality in relation to body mass index in the million women study.BMJ. 2007; 335 (Epub 2007, Nov 6): 1134Crossref PubMed Scopus (1029) Google Scholar]. A large follow-up study in the United States reported little or no association of excess all-cancer mortality with any of the BMI categories and that 11% of cancer deaths were from cancers potentially considered to be obesity related [16.Flegal K.M. Graubaud B.I. Williamson D.F. Gail M.H. Cause-specific excess deaths associated with underweight, overweight and obesity.JAMA. 2007; 298: 2028-2039Crossref PubMed Scopus (1170) Google Scholar].More than 30 years ago, Armstrong and Doll [17.Armstrong B.K. Doll R. Environmental factors and cancer incidence and mortality in different countries, with special reference to dietary practices.Int J Cancer. 1975; 15: 617-631Crossref PubMed Scopus (1871) Google Scholar] suggested strong correlations between elements of diet and cancer risk or more precisely between national food disappearance data and cancer mortality rates. Doll and Peto [5.Doll R. Peto R. The Causes of Cancer. Oxford University Press, Oxford1981Google Scholar] estimated that ∼30% of cancer mortality in the United States could be ascribed to nutritional factors: the range of certainty for this estimate was between 10% and 70%. The National Academy of Science in the United States established a Working Group to examine ‘after tobacco smoking, what is the commonest cause of cancer in the United States’. They concluded that it was nutrition [18.Committee on Diet, Nutrition and Cancer, National Research Council, National Academy of Sciences Diet. Nutrition and Cancer. National Academy Press, Washington, DC1982Google Scholar].Like the National Academy of Science report [18.Committee on Diet, Nutrition and Cancer, National Research Council, National Academy of Sciences Diet. Nutrition and Cancer. National Academy Press, Washington, DC1982Google Scholar] in 1982 led us down a clear path, the WCRF report [1.World Cancer Research Fund Report Food, Nutrition and the Prevention of Cancer: a Global Perspective.2007Google Scholar] now clearly tells us that it is time to retreat and try another route. Diet is remarkably difficult to measure and separation of the effects of individual components is extremely complicated given the multiple correlations which exist between the elements. It is past time to give up on reliance of questionnaire or sample measurement of diet and basing the assessment of nutritional components on instruments such as food composition tables. The way forward has to be through the development of meaningful biomarkers of cancer risk which may reflect the nutritional factors relevant to carcinogenesis. Such ‘biomarkers of nutritional status’ would encompass and integrate dietary intake, metabolism, diet-dependent endogenous formation of carcinogens, as well as hormonal or other (e.g. immunological) responses. Some elucidation of the role of nutrition and diet in human cancer will likely come from studies in populations with small intraindividual and large interindividual variation in consumption patterns as it experienced today in countries in transition (China, India, South East Asia, Latin America) rather than from further analyses of European, North American and other ‘Western’ population groups.After all the initial excitement and enthusiasm about the importance of nutrition [17.Armstrong B.K. Doll R. Environmental factors and cancer incidence and mortality in different countries, with special reference to dietary practices.Int J Cancer. 1975; 15: 617-631Crossref PubMed Scopus (1871) Google Scholar, 18.Committee on Diet, Nutrition and Cancer, National Research Council, National Academy of Sciences Diet. Nutrition and Cancer. National Academy Press, Washington, DC1982Google Scholar, 19.Tannenbaum A. Relationship of body weight to cancer incidence.Arch. Pathol. 1940; 30: 509-517Google Scholar], we appear to be moving towards a fairly depressing phase, and the weak conclusions of the WCRF report echo this trend. The substantial review of the evidence in the WCRF report demonstrates that there is no discernible association between many forms of cancer and specific dietary practices. There are still some very interesting hypotheses to pursue, such as the value of an approach on the basis of the food patterns (e.g. the Mediterranean diet score [20.Trichopoulou A. Lagiou P. Kuper H. Trichopoulos D. Cancer and Mediterranean dietary traditions.Cancer Epidemiol Biomarkers Prev. 2000; 9: 869-873PubMed Google Scholar]) rather than individual foods and nutrients, but the cupboard is remarkably bare.After decades of research activity, we still do not know how we need to change what we eat to reduce our cancer risk. It is a clear priority to establish a strategy to understand the causes of those cancers not related to tobacco smoking or other established risk factors. The way forward has to be through the development of meaningful biomarkers which may reflect the nutritional factors relevant to carcinogenesis and which also will allow randomised trials to be of shorter duration.Overweight and obesity are major issues for global health and are a public health priority at the present time. However, placing weight gain ahead of tobacco smoking as a cause of cancer appears difficult to justify. The public are frequently receiving confused and confusing messages: so too are the media. The scientific community has a collective responsibility to transmit clear messages to the general community and not those which favour their position or findings.The media and the general public need to be told clearly what are the facts and how important they are in the broad scheme of things. While there is an urgent need for more innovative research into the causes of cancer in men and women around the world, much could be achieved in cancer control within our current knowledge by tobacco control worldwide and an increasing move towards vaccination against common infective causes of cancer particularly in low-resource countries. Ten years ago, steps to prevent cancer were known. Intake of vegetables and fruits was clearly associated with reduced cancer risk, antioxidants were a crucial ingredient in lowering cancer risk and overweight and obesity was the most important risk factor for cancer in nonsmokers. What were clear messages then have been confused by subsequent findings and media reports. It is useful to stop and take stock of the current situation. The World Cancer Research Fund (WCRF) launched a second substantial compendium on ‘Food, Nutrition, Physical Activity, and the Prevention of Cancer: a Global perspective’ [1.World Cancer Research Fund Report Food, Nutrition and the Prevention of Cancer: a Global Perspective.2007Google Scholar]. According to the accompanying press release [2.WCRF Press ReleaseLandmark Report: Excess Body Fat Causes Cancer http://www.wcrf-uk.org/press_media/word_docs/report.doc (8 November 2007, date last accessed)Google Scholar], the Report ‘includes 10 recommendations … that represent the most definitive and authoritative advice that has ever been available on how the general public can prevent cancer’. The words ‘tobacco’ or ‘smoking’ are not mentioned once in the press release. In presenting its summary and recommendations, the report implicitly downplays the key importance of tobacco smoking in cancer causation. Contrary to that stated in the press release (the best advice for cancer prevention is to avoid weight gain [2.WCRF Press ReleaseLandmark Report: Excess Body Fat Causes Cancer http://www.wcrf-uk.org/press_media/word_docs/report.doc (8 November 2007, date last accessed)Google Scholar]), avoiding tobacco smoking and use of tobacco in other forms is the single best advice to reducing cancer risk as one-third of cancer deaths in high-income countries is attributable to tobacco use [3.Danaei G. Vander Hoorn S. Lopez A.D. et al.Comparative Risk Assessment Collaborating Group (Cancers). Causes of cancer in the world: comparative risk assessment of nine behavioural and environmental risk factors.Lancet. 2005; 366: 1784-1793Abstract Full Text Full Text PDF PubMed Scopus (952) Google Scholar, 4.International Agency for Research on Cancer (IARC) Monographs on the Evaluation of Carcinogenic Risks to Humans. Volume 83. Tobacco Smoke and Involuntary Smoking. IARC, Lyon2004Google Scholar, 5.Doll R. Peto R. The Causes of Cancer. Oxford University Press, Oxford1981Google Scholar]. Failure to include ‘stop smoking’ and ‘avoid exposure and exposing others to second-hand smoke’ among the 10 key recommendations undermines the most important message in cancer control. The ‘best advice’ also fails to mention the importance of a variety of established cancer risk factors including sun behaviour, occupational exposures, chronic infections and use of exogenous hormones. A decade after its original report [6.World Cancer Research Fund/American Institute for Cancer Research Food, Nutrition, Physical Activity, and the Prevention of Cancer: a Global Perspective. AICR, Washington, DC1997Google Scholar], the current report [1.World Cancer Research Fund Report Food, Nutrition and the Prevention of Cancer: a Global Perspective.2007Google Scholar] presents considerably weaker conclusions on the basis of the dilution of the strength of the evidence linking specific components of diet and nutrition to cancer risk. This is particularly true for the fruits and vegetables for which the strength of evidence for a protective effect of high intakes against several common epithelial cancers was downgraded from ‘convincing’ in the first WCRF report [6.World Cancer Research Fund/American Institute for Cancer Research Food, Nutrition, Physical Activity, and the Prevention of Cancer: a Global Perspective. AICR, Washington, DC1997Google Scholar] to ‘probable’ in this second report [1.World Cancer Research Fund Report Food, Nutrition and the Prevention of Cancer: a Global Perspective.2007Google Scholar]. When considering the importance and significance of diet and, more generally, nutrition on cancer risk, it is essential to separate ‘what we know’ from ‘what we think we know’ from ‘what we believe’ and to understand ‘what we do not know’. The WCRF report represents a comprehensive summary of the information necessary to evaluate the current state-of-knowledge, and the associated certainty, of this situation. ‘We believe’ that diet and nutrition play a key role in determining the risk of cancer. It is difficult to ignore the rapid changes in the risk of cancers at sites such as the colorectal and breast among groups of migrants from low-risk countries to high-risk countries in this regard. Other known risk factors of cancer are unable to explain such rapid changes, and diet and other nutritional factors are the most plausible determinants of such changes. ‘We think we know’ or, more accurately, ‘we thought we knew’ that a high-fat diet and low consumption of fruits, vegetables and fibres were associated with increased risks of common cancers. However, faith in the cancer prevention properties of fruits and vegetables began to crack when all the available evidence was critically reviewed by an International Agency for Research on Cancer (IARC) Working Group [7.IARC Handbooks on Cancer Prevention. Vol. 8. Fruit and Vegetables. IARC, Lyon, France2003Google Scholar]. Subsequently, it has crumbled as major analyses of prospective studies have continued to demonstrate consistently a lack of association between intake of fruits and vegetables and risk of several cancers. This major change in classification of one the few agents classified by WCRF in the category of strongest evidence in 1997 casts doubt on the rationale to classify ‘convincing’ to the evidence linking high meat intake to colorectal cancer risk in the current report [1.World Cancer Research Fund Report Food, Nutrition and the Prevention of Cancer: a Global Perspective.2007Google Scholar]. This also raises questions about the evaluation process and about the robustness of the classification system. The IARC Monographs Programme, which adopts comparable categories to evaluate the strength of evidence of carcinogenicity of different classes of agents, has evaluated >900 agents with ∼100 classified as having ‘sufficient’ evidence of carcinogenicity in humans [8.IARC. Overall Evaluations of Carcinogenicity to Humans http://monographs.iarc.fr/ENG/Classification/crthgr01.php (8 November 2007, date last accessed)Google Scholar], somewhat equivalent to convincing evidence in WCRF terminology. Despite an extensive work of reassessment of many sufficient carcinogens by IARC as the amount of evidence increases, none has ever been downgraded to a lower category of evidence. In view of the fragile grounds on which the conclusions of WCRF report on diet and cancer are based on, the information to the media should have been more cautious, which would probably have prevented confusion of messages disseminated to the public which have left a clear indication that cancer prevention depends on stopping eating bacon and avoiding drinking sodas [9.Blame it on the bacon. The Guardian. http://www.guardian.co.uk/commentisfree/story/0, 2204021,00.html (8 November 2007, date last accessed)Google Scholar, 10.Is anything safe to eat? Cancer report adds bacon, ham and drink to danger listDaily Mail online http://www.dailymail.co.uk/pages/text/print.html?in_article_id=490845&in_page_id= (8 November 2007, date last accessed)Google Scholar]. Clearly there is confusion in the media which is being transmitted to the general public. Evidence from animal and physiological experiments, although with conflicting evidence from observational studies, led to antioxidants being widely considered as one of the key protective elements in the diet, being found commonly in vegetables and fruits, and gave rise to a series of chemopreventive trials. A recent Cochrane Review has thrown light on the use of antioxidant supplementation on mortality and found no evidence to support antioxidant supplementation (e.g. vitamin A, beta-carotene, vitamin C, vitamin E, selenium) for either primary or secondary prevention [11.Bjelakovic G. Nikolov D. Gluud L.L. et al.Antioxidant supplements for prevention of mortality in healthy participants and patients with various diseases.Cochrane Database Syst Rev. 2008; (Art. No. CD007176)Crossref Scopus (385) Google Scholar]. Indeed, vitamin A, beta-carotene and vitamin E may increase mortality [11.Bjelakovic G. Nikolov D. Gluud L.L. et al.Antioxidant supplements for prevention of mortality in healthy participants and patients with various diseases.Cochrane Database Syst Rev. 2008; (Art. No. CD007176)Crossref Scopus (385) Google Scholar]. The value of this systematic review is that the conclusions are on the basis of the results of 67 randomised trials involving 232 500 participants randomised to antioxidant supplementation or placebo. The conclusions include the recommendation that more research needs to be done, perhaps focusing on vitamin C and selenium, rather than on vitamin A, beta-carotene and vitamin E. Vitamin D supplementation has been shown, in a meta-analysis, to reduce mortality in the arm randomised to the active intervention [12.Autier P. Gandini S. Vitamin D supplementation and total mortality: a meta-analysis of randomized controlled trials.Arch Intern Med. 2007; 167: 1730-1737Crossref PubMed Scopus (979) Google Scholar]. Again, more work in this area appears justified although it is premature to take action at this point. Renehan et al. [13.Renehan A.G. Tyson M. Egger M. et al.Body-mass index and incidence of cancer: a systematic review and meta-analysis of prospective observational studies.Lancet. 2008; 371: 536-537Abstract Full Text Full Text PDF PubMed Scopus (3595) Google Scholar] recently produced a systematic review and a meta-analysis to shed light on the strength of the association between excess bodyweight, expressed as body mass index (BMI) and the incidence of different types of cancer and to investigate differences in these associations between sex and ethnic groups. Renehan et al [13.Renehan A.G. Tyson M. Egger M. et al.Body-mass index and incidence of cancer: a systematic review and meta-analysis of prospective observational studies.Lancet. 2008; 371: 536-537Abstract Full Text Full Text PDF PubMed Scopus (3595) Google Scholar] carried out random-effects meta-analyses and meta-regressions of study-specific incremental estimates to determine the risk of cancer associated with a 5-kg/m2 increase in BMI. In all, 282 137 incident cases of cancer were included in this analysis. In men, a 5-kg/m2 increase in BMI was strongly associated with oesophageal adenocarcinoma (RR 1.52, P < 0.0001) and with thyroid (RR 1.33, P = 0.02), colon (RR 1.24, P < 0.0001) and renal (RR 1.24, P < 0.0001) cancers. In women, strong associations were recorded between a 5-kg/m2 increase in BMI and endometrial (RR 1.59, P < 0.0001), gallbladder (RR 1.59, P = 0.04), oesophageal adenocarcinoma (RR 1.51, P < 0.0001) and renal (RR 1.34, P < 0.0001) cancers. Weaker positive associations (RR < 1.20) were noted between increased BMI and rectal cancer and malignant melanoma in men; postmenopausal breast, pancreatic, thyroid and colon cancers in women and leukaemia, multiple myeloma and non-Hodgkin’s lymphoma in both sexes. Associations were stronger in men than in women for colon cancer (P < 0.0001). Associations were generally similar in studies from North America, Europe and Australia, and the Asia-Pacific region, but stronger associations were recorded in Asia-Pacific populations between increased BMI and premenopausal (P = 0.009) and postmenopausal (P = 0.06) breast cancers. While the authors concluded that an increased BMI is associated with increased risk of common and less common malignancies, the relative risks are small and the increment (5 kg or 11 pounds per m2) is large. In France, it is estimated that the attributable fraction for cancer death in the year 2000 was 24% for tobacco smoking and 2% for obesity and overweight [14.IARC Working GroupAttributable Causes of Cancer in France in the Year 2000. IARC Working Group Reports Volume 3. Lyon: IARC 2007Google Scholar]. These figures are consistent with other estimates of cancer deaths in high-income countries of 29% attributed to tobacco smoking and 3% to obesity and overweight [3.Danaei G. Vander Hoorn S. Lopez A.D. et al.Comparative Risk Assessment Collaborating Group (Cancers). Causes of cancer in the world: comparative risk assessment of nine behavioural and environmental risk factors.Lancet. 2005; 366: 1784-1793Abstract Full Text Full Text PDF PubMed Scopus (952) Google Scholar]. Most recently, it has been estimated that each year, 6000 cancer cases out of 120 000 in women in UK are related to overweight and obesity [15.Reeves G. Pirie K. Beral V. et al.Cancer incidence and mortality in relation to body mass index in the million women study.BMJ. 2007; 335 (Epub 2007, Nov 6): 1134Crossref PubMed Scopus (1029) Google Scholar]. A large follow-up study in the United States reported little or no association of excess all-cancer mortality with any of the BMI categories and that 11% of cancer deaths were from cancers potentially considered to be obesity related [16.Flegal K.M. Graubaud B.I. Williamson D.F. Gail M.H. Cause-specific excess deaths associated with underweight, overweight and obesity.JAMA. 2007; 298: 2028-2039Crossref PubMed Scopus (1170) Google Scholar]. More than 30 years ago, Armstrong and Doll [17.Armstrong B.K. Doll R. Environmental factors and cancer incidence and mortality in different countries, with special reference to dietary practices.Int J Cancer. 1975; 15: 617-631Crossref PubMed Scopus (1871) Google Scholar] suggested strong correlations between elements of diet and cancer risk or more precisely between national food disappearance data and cancer mortality rates. Doll and Peto [5.Doll R. Peto R. The Causes of Cancer. Oxford University Press, Oxford1981Google Scholar] estimated that ∼30% of cancer mortality in the United States could be ascribed to nutritional factors: the range of certainty for this estimate was between 10% and 70%. The National Academy of Science in the United States established a Working Group to examine ‘after tobacco smoking, what is the commonest cause of cancer in the United States’. They concluded that it was nutrition [18.Committee on Diet, Nutrition and Cancer, National Research Council, National Academy of Sciences Diet. Nutrition and Cancer. National Academy Press, Washington, DC1982Google Scholar]. Like the National Academy of Science report [18.Committee on Diet, Nutrition and Cancer, National Research Council, National Academy of Sciences Diet. Nutrition and Cancer. National Academy Press, Washington, DC1982Google Scholar] in 1982 led us down a clear path, the WCRF report [1.World Cancer Research Fund Report Food, Nutrition and the Prevention of Cancer: a Global Perspective.2007Google Scholar] now clearly tells us that it is time to retreat and try another route. Diet is remarkably difficult to measure and separation of the effects of individual components is extremely complicated given the multiple correlations which exist between the elements. It is past time to give up on reliance of questionnaire or sample measurement of diet and basing the assessment of nutritional components on instruments such as food composition tables. The way forward has to be through the development of meaningful biomarkers of cancer risk which may reflect the nutritional factors relevant to carcinogenesis. Such ‘biomarkers of nutritional status’ would encompass and integrate dietary intake, metabolism, diet-dependent endogenous formation of carcinogens, as well as hormonal or other (e.g. immunological) responses. Some elucidation of the role of nutrition and diet in human cancer will likely come from studies in populations with small intraindividual and large interindividual variation in consumption patterns as it experienced today in countries in transition (China, India, South East Asia, Latin America) rather than from further analyses of European, North American and other ‘Western’ population groups. After all the initial excitement and enthusiasm about the importance of nutrition [17.Armstrong B.K. Doll R. Environmental factors and cancer incidence and mortality in different countries, with special reference to dietary practices.Int J Cancer. 1975; 15: 617-631Crossref PubMed Scopus (1871) Google Scholar, 18.Committee on Diet, Nutrition and Cancer, National Research Council, National Academy of Sciences Diet. Nutrition and Cancer. National Academy Press, Washington, DC1982Google Scholar, 19.Tannenbaum A. Relationship of body weight to cancer incidence.Arch. Pathol. 1940; 30: 509-517Google Scholar], we appear to be moving towards a fairly depressing phase, and the weak conclusions of the WCRF report echo this trend. The substantial review of the evidence in the WCRF report demonstrates that there is no discernible association between many forms of cancer and specific dietary practices. There are still some very interesting hypotheses to pursue, such as the value of an approach on the basis of the food patterns (e.g. the Mediterranean diet score [20.Trichopoulou A. Lagiou P. Kuper H. Trichopoulos D. Cancer and Mediterranean dietary traditions.Cancer Epidemiol Biomarkers Prev. 2000; 9: 869-873PubMed Google Scholar]) rather than individual foods and nutrients, but the cupboard is remarkably bare. After decades of research activity, we still do not know how we need to change what we eat to reduce our cancer risk. It is a clear priority to establish a strategy to understand the causes of those cancers not related to tobacco smoking or other established risk factors. The way forward has to be through the development of meaningful biomarkers which may reflect the nutritional factors relevant to carcinogenesis and which also will allow randomised trials to be of shorter duration. Overweight and obesity are major issues for global health and are a public health priority at the present time. However, placing weight gain ahead of tobacco smoking as a cause of cancer appears difficult to justify. The public are frequently receiving confused and confusing messages: so too are the media. The scientific community has a collective responsibility to transmit clear messages to the general community and not those which favour their position or findings. The media and the general public need to be told clearly what are the facts and how important they are in the broad scheme of things. While there is an urgent need for more innovative research into the causes of cancer in men and women around the world, much could be achieved in cancer control within our current knowledge by tobacco control worldwide and an increasing move towards vaccination against common infective causes of cancer particularly in low-resource countries.