Abstract
BACKGROUND: Pregnancy may reduce breast cancer risk through induction of persistent changes of the mammary gland that make the breast less susceptible to carcinogenic factors. It is not known to what extent the effects of parity are independent of other breast cancer risk factors. METHODS: In a Norwegian cohort of 58 191 women (2890 breast cancers), we assessed whether the effects of parity on postmenopausal breast cancer risk may be modified by menstrual and anthropometric factors. We calculated attributable proportions due to interaction as a measure of synergism. RESULTS: Parity, height, body mass index (BMI), age at menarche and menopause were all associated with breast cancer risk in the expected directions. For BMI, follow-up was stratified into two age groups because of non-proportional hazards. We found that nulliparity and overweight may amplify each other's effect on breast cancer risk among women after 70 years of age (attributable proportion 0.21, 95% confidence interval 0.04-0.39). There was some indication that parity and age at menopause may antagonise each other's effect. Effects of parity were largely unaffected by age at menarche and height. CONCLUSION: Nulliparity and overweight may have a synergistic effect on breast cancer risk in elderly women. If confirmed by others, the findings may help disentangle the interplay of different causes of breast cancer.