In spite of several decades of research, few nutrition-related factors, other than obesity and alcohol consumption, have been unequivocally established as playing a causal role in human cancer ⁽¹⁾. Several methodological problems faced by epidemiological studies may explain this situation: dietary habits are difficult to assess accurately; biologically relevant dietary exposures may have occurred over many years and their role may be modified by other lifestyle factors. Casecontrol studies may be flawed by recall bias, and biological markers of diet or metabolism might be altered by the tumour presence. Prospective studies are not subject to these biases, but they may lack statistical power to study less common cancers and genetic interactions. Prospective studies have often been conducted within populations with relatively homogeneous lifestyles and dietary patterns. This homogeneity, combined with diet measurement errors, makes it difficult to demonstrate moderate associations.

What is EPIC ?

The European Prospective Investigation into Cancer and Nutrition (EPIC) – a multi-centre prospective cohort study – was devised in an attempt to overcome these limitations. Initiated in 1992, the study is a collaborative endeavour of 23 centres in ten European countries: Denmark, Germany, Greece, Italy, France, The Netherlands, Norway, Spain, Sweden and United Kingdom, including over 500,000 volunteer participants mostly aged from 35 to 70 years. Information of usual diet and anthropometric measurements were collected at enrolment. Blood samples were taken from most participants. EPIC represents the largest single resource available for prospective investigations in the aetiology of cancers that can integrate questionnaires on lifestyle and diet, biomarkers of diet, metabolism and genetic polymorphisms, with the additional advantage of the contrast in cancer rates and dietary habits between centres ⁽²⁾.

Testing the effect of fruit and vegetables against cancer

One of the hypotheses tested in the EPIC study is the potential protective effect of fruit and vegetable consumption against certain cancers. We briefly summarize the results here.

Between the first results of EPIC, the finding of the association between fibre intake and reduced colorectal cancer risk is of great scientific importance, since other cohorts questioned the potential protective effect of fibre ⁽³⁾. Two other results that support the potential beneficial effect of plant foods against colorectal cancer are the protective effect of fruit and vegetables observed in preliminary analyses ⁽⁴⁾ and the significant inverse association between nut and seed intake observed for colon cancer in women ⁽⁵⁾.

It is well established that the main risk factor of lung cancer is smoking. EPIC has confirmed previous findings of a potential protective effect of fruit intake against lung cancer in an analysis of 860 incident cases, while no association with vegetables was observed ⁽⁶⁾.

Although no association between total fruit and vegetable consumption and risk of renal cell carcinoma (306 incident cases) was demonstrated, the data do not exclude the possibility that very low consumption may be related to a higher risk, since an inverse association with root vegetables was observed ⁽⁷⁾. Gastric cancer (GC, 330 cases) and adenocarcinoma of oesophagus (ACO, 65 cases) were not related with fruit and vegetable in EPIC, although a potential beneficial effect of vegetables and of onion and garlic for the intestinal type of GC, and of citrus fruit against cardia GC and ACO, for which inverse associations were observed, cannot be discarded ⁽⁸⁾.

Finally, the data from 3659 invasive incident breast cancer cases in EPIC support previous evidence that vegetable and fruit intake is not associated with risk for breast cancer ⁽⁹⁾ and that prostate cancer is not associated with total fruit and vegetables consumption ⁽¹⁰⁾. Fruit and vegetables do not seem to protect from ovarian cancer (581 verified cases of epithelial cancer) ⁽¹¹⁾.

Overall, these results indicate that the potential beneficial effect of fruit and vegetables against certain cancers deserve further investigation, and that studies on specific types of vegetables and fruits, as well as studies using biological markers of diet are warranted. This research is actually ongoing in EPIC.