Overweight and Changes in Weight Throughout Adult Life in Breast Cancer Etiology

Numerous epidemiologic studies have found body size to be a significant risk factor in the etiology of breast cancer. In an Israeli study population of 1,065 breast cancer patients, 964 surgical controls, and 981 neighborhood controls, height and weight at three periods (age 18, "most of adult life," and recent) were ascertained. The authors analyzed these parameters and body mass index (weight/height) for each period, as well as body mass index changes throughout life, controlling for age, menstrual status, and ethnic origin. Odds ratios were determined for three body mass index categories: 19.1-23, 23.1-27, and 27.1+, with a relative risk of 1 for body mass index less than or equal to 19. Their results show an increase in risk for breast cancer with greater recent body mass index among postmenopausal women aged 60+ (n = 461 for breast cancer, n = 414 for surgical controls, n = 401 for neighborhood controls). Crude odds ratios for the breast cancer/surgical control comparison are 1.23, 1.58, and 2.20, respectively, for each body mass index category; for the breast cancer/neighborhood control comparison 2.16, 2.44, and 2.99, respectively. Odds ratios adjusted for confounding factors (ages at menarche, first birth, and menopause; number of births; years of education; previous benign breast disease; and family history of breast cancer) are 1.17, 1.44, and 2.38, respectively (breast cancer/surgical control); and 1.78, 1.92, and 2.53, respectively (breast cancer/neighborhood control). Overweight does not emerge as a risk factor for breast cancer among premenopausal or younger postmenopausal women. Weight loss from most of adult life to recent weight appears to be protective, since mean loss in the 60+ age category is greater in both control groups than in breast cancer patients. In addition, breast cancer patients aged 60+ gained more weight during adult life than controls, and premenopausal breast cancer patients gained less weight than controls (for both comparisons, p less than or equal to 0.05, breast cancer vs. all controls combined.