Risk factors for primary breast cancer in Japan: 8-year follow-up of atomic bomb survivors

Background. Findings from the Life Span Study (LSS) of the health effects of exposure to atomic bomb radiation have documented a strong dose-response relation between radiation exposure and breast cancer incidence. Purpose. We analyzed data from the LSS cohort to identify nonradiation risk factors for breast cancer and to determine whether these factors were independent of the effects of radiation on breast cancer occurrence. Methods. Breast cancer incidence was ascertained among a cohort of 22,200 residents of Hiroshima and Nagasaki, Japan, who had completed a mail survey between 1979 and 1981 to study nonradiation risk factors for disease. During the subsequent follow-up period (average 8.31 years), 161 cases of primary breast cancer were identified through population-based tumor registries in the two cities. Results. The risk of breast cancer was inversely related to age at menarche and weakly positive in relation to age at menopause and years of menstruation. A significant negative association of full-term pregnancy against breast cancer was observed, although the number of pregnancies beyond the first was not related to the rate of breast cancer in the cohort. Women having their first full-term pregnancy before age 30 were at decreased risk of breast cancer relative to older women, but there was no trend. A nonsignificant, positive trend in risk was associated with increasing weight and body mass (kg/m(2)). The risk of breast cancer among women with a history of estrogen use was 1.64 (95% confidence interval 1.02-2.64) and with diabetes 2.06 (95% confidence interval 1.27-3.34). It was not possible to distinguish among additive and multiplicative models of the joint association of radiation dose and various non-radiation-related exposures (age at menarche, full-term pregnancy, female hormone preparations) identified in this analysis. Conclusions. Nonradiation risk factors for breast cancer among Japanese atomic bomb survivors were consistent with those identified among other populations of women, although the prevalence of common risk factors was low Reproductive factors and hormone use appear to act independently of radiation exposure on the risk of breast cancer among this population. (C) 1997 Academic Press.