Abstract

The aim of our study was to examine how the use of oral contraceptives (OCs) interact with alcohol on breast cancer risk within the large prospective follow-up study, Norwegian Women and Cancer Study. Between 1991 and 1997, women aged 30 to 70 years were drawn at random from the central person register and mailed an invitation. Follow-up information was collected throughout 2001 by linkage to national registries. Only women (n = 86,948) with complete information on alcohol consumption and duration of OC use were included in the present analysis. A total of 1,130 invasive breast cancers were diagnosed during 618,638 person-years of follow-up. Consumption of > or =10.0 g/d alcohol was associated with a breast cancer relative risk (95% confidence interval) of 1.69 (1.32-2.15), consistent with a linear relationship (P for trend < 0.0001). Among alcohol consumers, an excess risk of breast cancer was observed for total duration of OC use only among women who consumed <5 g/d alcohol (P for trend = 0.0009). We observed a negative interaction between duration of OC use and alcohol consumption effects (P for interaction = 0.01). After stratification on menopausal status, the association between high alcohol intake and breast cancer was more prominent among postmenopausal women than among premenopausal women (P for heterogeneity = 0.01). No interaction between alcohol and duration of OC use were significant after stratification on menopausal status. Our findings in conjunction with biological data imply that alcohol and OCs have antagonistic effects on breast cancer risk through a common pathway. Whether the interactive effect differs according to menopausal status remains unclear and needs further investigations.