Fruits and vegetables and the risk of epithelial cancer

In this thesis, prospective studies on fruit and vegetable consumption in relation to epithelial cancer risk were described. The main research question was whether higher intakes were related to lower risks of epithelial cancers, mainly of lung cancer.

In the Seven Countries Study, at the population level, consumption of fruits, vegetables and total plant foods was not related to colorectal cancer risk, whereas a difference of 10 g/d of fiber intake was associated with a 33% lower risk. Average population consumption of fruit was inversely and of refined grains positively related to population stomach cancer risk. Low consumption of fruits was however strongly correlated with high refined grain consumption.

Fruit but not vegetable consumption was inversely associated with 25-year lung cancer mortality among European smoking men. This association was confined to heavy cigarette smokers. In Dutch men and women aged 20-59, vegetable consumption was inversely associated with lung cancer incidence, particularly of adenocarcinomas. Fruit consumption was not related to lung cancer after adjustment for smoking.

Adherence to the dietary guidelines for fruits and vegetables was inversely associated with cancer incidence in Dutch elderly men. Consumption of the recommended amount of fruit was related to a 38% lower risk, while vegetables were not associated. Variety in vegetable intake was however inversely related to total and non-lung epithelial cancer.

Plasma carotenoid concentrations were only crude indicators of usual vegetable and fruit intake in Dutch men and women aged 20-59. Plasmaβ-cryptoxanthin indicated fruit intake and total intake of vegetables, fruits and juices, whereas lutein was a marker for vegetable intake. Concentrations of carotenoids coul d not differ between all four quartiles of intake.

During 1987/88-1997/98, the mean fruit and vegetable consumption (excluding juices) decreased with 34 g/d (12%) in Dutch men and 23 g/d (8%) in Dutch women. The consumption was lowest and decreased most in those aged 19-35 with a low level of education. Using a computer simulation model, the maximum theoretically reduction in cancer incidence, i.e., when all would consume the recommended 400 grams daily, was estimated to be 14 to 22% for this group.

Valid assessment of fruit and mainly of vegetable intake, residual confounding by smoking and enough power of the study are major methodological concerns. In recent cohort studies weaker associations were observed compared to earlier risk estimates. Taken all evidence together, an inverse association between fruit and vegetable intake and cancer of the lung, stomach and colon/rectum is still indicated. There is not enough evidence yet to point at specific fruits and vegetables or plant compounds as responsible actors.