Exogenous estrogens and other risk factors for breast cancer in women with benign breast diseases.

A group of 1,439 white women who were initially treated for biopsy-proved benign breast diseases from 1942 to 1975 in a single private surgery practice were followed through 1976 for development of breast cancer. Information on age, parity, and use of noncontraceptive estrogens was collected on all women at the time of their initial benign lesions. At follow-up, ages at menarche, menopause, and birth of first child were obtained, as was information on use of estrogens after the initial lesion. Slides from the initial lesion were reinterpreted and classified by a panel of pathologists. The influence of gross and microscopic features of the initial benign conditions on subsequent risk of breast cancer was ascertained and reported previously. This paper presents results of additional analyses based largely on Cox's proportional hazards model. Risk of breast cancer was related to nulliparity and ages at menarche and birth of first child and to a lesser extent to age at artificial menopause. Exogenous estrogen taken prior to the initial benign lesion did not alter risk of breast cancer. Subsequent use, primarily of conjugated estrogens, eliminated the protective effect of an artificial menopause and appeared to act synergistically with epithelial hyperplasia or papillomatosis in the initial lesion and calcification of that lesion to increase the risk of breast cancer. A marked increase in risk of breast cancer in succeeding birth cohorts was unexpectedly found that could not be explained by temporal changes in any of the other risk factors considered.