The evidence of a protective effect of vegetables and fruit on breast cancer (BC) is mostly based on case-control studies while, so far, prospective studies have provided weaker or no evidence of this effect.

A prospective study in a Mediterranean Country

A large variety of specific vegetables and fruit are traditionally consumed by Mediterranean populations, often in large amounts, thus offering the opportunity to evaluate in a favourable setting the effects of specific types of these foods. We have evaluated the relationship between vegetables and fruit consumption and BC risk in the Italian cohort of the European Prospective Investigation into Cancer and Nutrition (EPIC) study. The Italian section of the EPIC study includes five cohorts (Turin, Varese, Florence, Naples and Ragusa) and enrolled, in the period 1993-1998, more than 32.000 adult women. According to the common EPIC protocol, information about diet and life-style habits, anthropometric measurements and a blood sample were collected for each participant, after signature of an informed consent form.

The collection of information on dietary habits

Usual diet was investigated through Food Frequency Questionnaires specifically developed to capture local dietary habits typical of the Italian population in different areas of the country. The absolute frequency of consumption of each item is assessed asking the number of times a given food item is consumed (per day, week, month or year). The quantity of the food consumed is assessed through the selection of an image of a food portion, or by selection of a predefined standard portion when no image is provided. For some types of fruits (e.g. citrus) and vegetables (such as raw tomatoes and cabbages), whose consumption in Italy is strongly dépendent on season, intake is assessed separately in and out of the main cropping season of typical consumption.

The total vegetables consumption category included all sorts of raw and cooked fresh vegetables. Specific subgroups considered were: leafy vegetables (salad greens, chard, spinach and other leafy greens); tomatoes (raw and cooked); other fruiting vegetables (peppers, artichokes, aubergine, courgette, green beans, fennel, celery); root vegetables (including carrots and red beetroot); cabbages (including broccoli, brussel sprouts, cauliflower, black cabbage and savoy cabbage); onions, garlic and leek; mushrooms. The total fruit consumption category included all sorts of fresh fruits (analysed also separately as citrus fruit and other fruit), nuts and seeds and dried and canned fruits as well.

Increasing consumption of vegetables was inversely associated with the risk to develop breast cancer

After a median follow up of approximately 11 years, 1,072 newly diagnosed BC cases were identified and included in the analyses. Analyses adjusted for known or potential risk factor for BC (education, anthropometry, reproductive history, Hormone Replacement Therapy, physical activity, alcohol consumption and smoking habits) showed an inverse association between consumption of “all vegetables” and BC risk. Women in the highest (>264.8 g/day) in comparison with women in the lowest quintile (<107.8 g/day) of consumption have a 35% significant reduction in risk. An inverse association also emerged for increasing consumption of “leafy vegetables”, either cooked or raw, (with a reduction of 30% in risk among women in the highest in comparison with women in the lowest quintile of consumption) and of fruiting vegetables. An inverse association also emerged with increasing consumption of raw tomatoes, a major component of mixed salads in this Italian population, together with lettuce and other raw leafy vegetables For other vegetables, the point estimates suggested a possible inverse association with BC risk, although no statistically significant associations emerged. No association for fruit, overall or by subtypes, with BC risk was found.

Our results support an inverse association between increased consumption of vegetables, overall and of specific subtypes (namely leafy vegetables, either cooked or raw), and BC risk. The public health implications of a beneficial effect of vegetables on breast cancer, for which only a few modifiable risk factors have been so far identified, are relevant. These results warrant further investigation in order to better understand the specific effects of a high consumption of vegetables in the context of the Mediterranean diet.