Nutritional epidemiology and cancer: What do we know about eating to reduce cancer risk?

摘要

In 1981 Doll and Peto estimated that approximately 35% of cancer deathsin the US could potentially be avoided by dietary modification. A morerecent estimate (1) put the figure at 32%. These estimates have inspiredmany studies of diet and cancer aimed at determining exactly what foodsor nutrients constitute the ideal cancer-risk-minimizing diet; including theMelbourne Collaborative Cohort Study established by the Cancer CouncilVictoria (CCV) in the early 1990’s.In 2007 theWorld Cancer Research Fund and American Cancer Research Fundpublished ‘Food, Nutrition, Physical Activity, and the Prevention of Cancer: aGlobal Perspective’, a summary of the data to date (2). The recommendationsarising from this report: be lean, be physically active, limit consumption ofenergy dense foods, eat mostly plant foods, limit red meat intake and avoidprocessed meat, limit alcoholic drinks, limit salt, avoid mouldy cereals orgrains, aim to meet nutritional needs through diet, breastfeed, and for cancersurvivors, follow these same recommendations, are not inconsistent withgeneral dietary guidelines and unlikely to be controversial in most settings.While an earlier report in 1997 concluded that the evidence for fruit andvegetables reducing cancer riskwas ‘convincing’;based mainlyoncase-controlstudies, the 2007 report found only ‘probable’ or ‘limited suggestive’ evidencefor the benefits of fruit and vegetables based on new prospective studies.A recent study found that following the recommendations, especially forbody size, plant foods and alcohol, was associated with a 60% reduced riskfor post-menopausal breast cancer (3); while in EPIC following the recommendationswas associated with a 34% lower risk of death, and withreduced risk of dying from cancer (4).The Cancer Epidemiology Centre (CEC) established the Melbourne CollaborativeCohort study in the early 1990s to prospectively investigate associationsbetween diet and cancer (5). The cohort comprised largelyAustralian born men and women aged 40e69 years with 25% of southernEuropean migrants in order to broaden the range of dietary exposure. Afteraround 20 years of follow-up, linking the cohort to death and cancer registriesin Victoria and nationally,we have examined associations of diet with the more common cancers and contributed data to studies of less commoncancers. While many associations evaluated were not strong our findingscontribute to the overall evidence on which national and internationalrecommendations are based. We have also added to the understanding ofassociations between diet and cancer by examining differences in associationsbetween different cancer subgroups, such as by aggressiveness forprostate cancer or hormone receptor subtype for breast cancer. One areawhere reasonably consistent associations were found was for measures ofbody size. On the basis of this a TV campaign promoting waist circumferencesof under 100 cm for men and 85 cm for women (6) was developed.New topics of research in this area include studies we are undertakinginvestigating whether DNA methylation in peripheral blood can predictcancer risk, and whether methylation could be a mechanism linking dietwith cancer. We are also working with colleagues in the US to look at associationsbetween a dietary inflammatory score and different canceroutcomes. Despite what we already know, the recent Australian HealthSurvey (7) indicates that Australians are not following recommendationsthat could help reduce cancer risk. Among people aged 18 years or older,63% were obese or overweight, 48% met the Australian guidelines for fruitconsumption (2 or more serves per day for adults), 8% met the guidelinesfor vegetable consumption (5 or more serves per day), around 20%consumed more than 2 standard drinks per day and 67% of those aged 15years or older were sedentary or performed only a low level of exercise.Further research may refine recommendations and understanding ofmechanisms by which diet can modify cancer risk, but epidemiologists needtowork with others to bring the benefits of this to the population in general.