The major causes of a disease should be able to explain its dominant epidemiological patterns. Yet, with respect to breast cancer, there appears to be a major paradox. The role of endogenous estrogens in the etiology of breast cancer in women is well defined. However, differences between Caucasian and Asian women, with respect to estrogen production and blood levels, are not sufficiently large to explain the 5-fold difference in breast cancer incidence between these groups. Given the obligate role of estrogen receptors in estrogen response expression of these receptors in normal breast tissue could be a major determinant of breast cancer risk and an important explanatory factor of the marked variability in breast cancer incidence worldwide. A number of studies have indicated that breast cancer risk may be elevated when estrogen receptor expression is high, but exogenous determinants of estrogen receptor expression have not been identified. Diet in adult life does not appear to be a powerful determinant of breast cancer risk; however, consumption of vegetables, fruits, and olive oil are inversely associated with risk, whereas alcohol intake and perhaps saturated fat of animal origin may increase risk. Because food groups and nutrients may affect breast cancer risk, and expression of estrogen receptor-a (ER-a) in normal mammary tissue may modify this risk, we wish to investigate whether diet may affect the expression of ER-a in normal mammary epithelium. Specifically, we wish to investigate whether 1-) a diet high in fruits and vegetables or high in olive oil intake reduces ER-a expression in the mammary tissue, and 2-) whether a diet high in animal fat, in n-6 polyunsaturated fats, or alcohol consumption increases ER-a expression in mammary tissue. In the breast clinic of a major university hospital in Athens, Greece, some 4,000 women undergo annual mammography screening. Every year, more than 300 non-malignant breast biopsies are taken from these women. Thus, over a period of 18 months we estimate that biopsies from 440 women without breast cancer will be available. These women will be interviewed in person for demographic, socioeconomic, reproductive and biomedical variables, and they will complete a validated extensive food frequency questionnaire. The percent of ER-a positive cells (ER-oc+ percent) will be determined through immunohistochemistry and through semi-quantitative PCR. All statistical analyses will conducted using linear models to determine the distribution of ER-a percent, by diet and other covariates.
