We propose to continue to evaluate subcategories of benign breast disease in relation to risk of breast cancer. In addition, interactions between these characteristics and established and hypothesized risk factors for breast cancer will be examined using prospectively collected data. In particular, we will examine interactions between atypical hyperplasia and family history of breast cancer, menopause, use of postmenopausal estrogens, past use of oral contraceptives and relative weight (weight/height2), alcohol intake, dietary fat, intake of vitamin A (preformed and carotene), vitamin E, and caffeine. The proposed study is nested case-control within the Nurses' Health Study, a cohort of 121,700 US female registered nurses, currently aged 444-69 who were enrolled in a prospective study of risk factors for breast cancer in 1976 and who have been followed with biennial questionnaires since then. In addition, we add 50 cases and their controls from the Nurses' Health Study II. An extensively validated semi-quantitative food frequency questionnaire administered in 1980 enables us to assess the dietary variables. We propose to obtain and review histopathology slides from an additional 500 women with breast cancer diagnosed between return of the 1988 NHS questionnaire and the 1996 questionnaire who have a history of an earlier biopsy for benign breast disease and 2218 controls randomly selected from among women with a history of biopsy for benign breast disease in the cohort. Slides will be independently reviewed in a blinded fashion by two pathologists and graded according to the classification system based on that of Dr. Page. An histopathologic aim is to define objective criteria for the architecture and morphology of atypical hyperplasia. The relative risk of breast cancer associated with subcategories of benign breast disease (nonproliferative, proliferative, atypical) will be calculated using the women with no history of benign breast disease in the Nurses' Health Study cohort as a reference group. Potential interactions with epidemiologic and dietary risk factors will be assessed by stratified and multivariate analysis. We will also analyze data from the Nurses' Health Study II to examine the relation between long term use of oral contraceptives before first pregnancy and risk of atypical hyperplasia on breast biopsy.
Showing the most recent 10 out of 36 publications
