We all know that a good diet is key to good health, but it's now clear that certain foods we eat can unmask underlying susceptibilities to cancer.

Genetic factors are known to be important in determining propensity to develop cancer but there is strong evidence that the worldwide variation in cancer incidence is mainly due to our lifestyle and environmental factors such as the food we eat. A deeper understanding of this relationship is being afforded by the largest study of diet and health ever undertaken, involving over half a million people in ten countries. Researchers at the Medical Research Council (MRC) Centre for Nutritional Epidemiology in Cancer Prevention and Survival (CNC), in the University of Cambridge's Department of Public Health and Primary Care, are building on these findings under the directorship of Professor Sheila Bingham.

Cancer incidence worldwide

Geographically, when age is taken into account using world-standardised rates, there are up to 100-fold variations in the incidence of particular cancers in different regions of the globe. The incidence of colon cancer, as an example, varies 20-fold between the highest (USA) and the lowest (India) incidence. Even across Europe there are marked differences in cancer rates, with incidence rates in Greece being about half those of Germany.

  • Worldwide, about 10 million people are diagnosed annually with cancer and more than 6 million people die of the disease.
  • The most common cancers worldwide (excluding non-melanoma skin cancer) are lung (12.3%), breast (10.4%) and colorectal (9.4%).
  • In developed countries, the main cancers are of the lung, large bowel, breast, prostate and bladder.
  • In developing countries, the main cancers are of the cervix, liver, stomach and mouth.
  • In sub-Saharan Africa, cancer of the large bowel, breast and lung are virtually absent.
  • 'Hot spots' exist for cancers at certain sites: oesophageal cancer in parts of Iran and nasopharyngeal cancer in parts of South-East Asia.

Studies of migrants moving from a low- to a high-risk area have shown that the migrants acquire the cancer pattern of the host country within a relatively short period of time – as swiftly as within a single generation.

Over the past 40 years in Japan, colon cancer has increased enormously and the incidence of bowel cancer in Japanese men is now twice that of men in the UK, despite being extremely rare only 40 years ago. It's clear that changes in the gene pool cannot account for such rapid changes in Japan. A more likely explanation is that the Japanese population have a susceptibility to this cancer that has been unmasked by their rapidly changing diet.

What is the link with nutrition?

There are very strong links between the amount of certain dietary items consumed, together with the dietary practices used in various populations, and the incidence rates for common cancers. On the basis of such comparisons, it has been estimated that 32–35% of cancers could be attributed to nutrition, although the contribution of diet to specific types of cancer varies from as low as 10% for lung cancer to as high as 80% for cancer of the large bowel.

Even so, there is a complex interaction between genetic factors, individual metabolic characteristics and diet. By themselves, common variants in genes that regulate the metabolism of food constituents are unlikely to confer large cancer risks, but they could do so in individuals who smoke, drink or have a particular dietary pattern. Links between diet and cancer can therefore only be found by measuring not only gene variants but also dietary exposure in populations. To do this, you need to study as large a number of people as possible – which can only be made possible through significant collaboration.

An EPIC study

The European Prospective Investigation into Cancer and Nutrition (EPIC) is the largest study ever to investigate specifically the relationship between diet, metabolic and genetic factors, and certain types of cancer. EPIC includes 519,978 volunteers (366,521 women and 153,457 men) who have been recruited to 23 regional or national centres in 10 countries across Europe. Blood was collected from approximately 400,000 subjects at enrolment and the different fractions were stored under special conditions as straws in liquid nitrogen pending further analysis. By studying people in different countries with differing diets, it is anticipated that specific information about the effect of diet on long-term health will be discovered.

EPIC is an enormous resource and, since its inception in 1992, numerous papers have been published by the study collaborators, at a fast increasing rate as more and more cancers develop within the cohort; 75 papers were published in 2006. Of these, two key published findings have been that an approximate doubling of dietary fibre intake is associated with a 40% reduction in colorectal cancer incidence and that red and processed (but not white) meat is associated with an increase in risk of colorectal cancer, whereas fish is protective. A further finding was that eating vegetables and fruits did not appear to protect against breast cancer.

Collaborating centres in the UK

In the UK, there are two EPIC collaborating centres, one based in Oxford and the other at the University of Cambridge's Institute of Public Health (EPIC Norfolk;, which has recruited 25,000 men and women from Norfolk. EPIC Norfolk has taken an important lead in the biostatistical handling of data, in collaboration with the MRC Biostatistics Unit, and in developing nutritional methods for measuring diet on such large population samples. One of the important findings of EPIC Norfolk has been the demonstration that fat could be an important risk factor in breast cancer, a finding that has been replicated in a subsequent study in the USA.

The search continues

CNC, a new MRC Centre awarded to the Institute of Public Health, builds on the success of the Europe-wide EPIC collaboration, and is the focus of a unique consortium of cohorts of approximately 100,000 participants from several UK universities investigating how diet can prevent cancer. Cancer survival in relation to diet is being assessed in the SEARCH (Studies of Epidemiology And Risk Factors in Cancer Heredity) study of cancer genetics in breast, bowel, prostate and ovarian cancer survival and prognosis. CNC also fosters epidemiological biomarker studies into the mechanisms whereby nutritional factors are involved in DNA alterations leading to mutation and ultimately tumour development.

Cancer is still one of the most common causes of death in western countries, including the UK. As a significant number of cancers may be the result of diet, studies at EPIC Norfolk and the CNC will provide a crucial focus for nutritional epidemiology and our future understanding of the link between nutrition and cancer.

'32–35% of cancers could be attributed to nutrition.... from as low as 10% for lung cancer to as high as 80% for cancer of the large bowel.'

For more information, please contact the author Professor Sheila Bingham ( at the CNC (

This work is licensed under a Creative Commons Licence. If you use this content on your site please link back to this page.