Breast cancer risk: effects of estrogen replacement therapy and body mass.

BACKGROUND: Epidemiologic studies have focused on the association between breast cancer risk and a variety of lifestyle and exogenous factors. PURPOSE: The purpose of this study was to clarify the effects of alcohol consumption, cigarette smoking, oral contraceptive (OC) use, estrogen replacement therapy (ERT), and body mass on risk of breast cancer. METHODS: These variables were examined in a case-control study of 604 patients with newly diagnosed breast cancer and 520 control subjects who did not have breast cancer and were frequency matched for age, hospital, and time of diagnosis. These case patients and control subjects were part of an ongoing study of breast cancer by the American Health Foundation and were selected for interview from hospitals in the New York City area from January 1987 through December 1989. The data were analyzed by computation of odds ratios (ORs) for potential risk factors, with adjustment for age at diagnosis and other potential confounding variables and with stratification by menopausal status. RESULTS: We observed positive effects of ERT and high body mass on the risk of postmenopausal breast cancer, particularly when each factor was examined in the absence of the other factor. In lean postmenopausal women, the adjusted summary OR associated with ERT was significantly elevated (OR = 2.0; 95% confidence interval [CI] = 1.1-3.5; P < .01), and there was a statistically significant dose response of breast cancer risk with ERT duration (adjusted ORs = 2.0 for < 5 years and 2.2 for > or = 5 years; positive trend, P < .02). Reciprocally, in women who did not receive ERT, high body mass (Quetelet index > 27) was a significant risk factor for postmenopausal breast cancer (OR = 2.1; 95% CI = 1.3-3.3; P < .02), and the linear trend in risk with increasing body mass was significant (positive trend, P < .02). The strongest effect of body mass occurred in women who were lean at age 18 and gained enough weight to place them in the upper tertile of body mass at the time of diagnosis (OR = 2.6; 95% CI = 1.5-4.6; P < .01). There was no evidence of significant positive associations between breast cancer risk and cigarette smoking, alcohol consumption, or OC use in any subgroup of these women. CONCLUSIONS: Our results support the hypothesis that excess adipose deposition heightens breast cancer risk in the postmenopausal years. Furthermore, they underscore the need for continuing investigation of the effects of exogenous estrogens on the development of this malignancy, particularly in lean postmenopausal women.