Mammographic Breast Density and Risk Factors of Breast Cancer in Korean Women Using Multicenter Study

Background: Density patterns on mammography have been related to the risk factors for breast cancer in the western countries. High mammographic density appears to confer a 4-fold risk of breast cancer. While the proportion of high-density mammography is higher in Korean women than in Caucasian women, the incidence of breast cancer in Korean women is considerably lower. Therefore, we examined if the mammographic breast density pattern correlates with the risk factors for breast cancer in Korean women. Methods: In a cross-sectional design, we recruited 1,758 pre-and post-menopause women without prior history of breast cancer and breast surgery who underwent screening mammogram and completed a self-administered questionnaire in 6 general hospitals. On the basis of ACR BI RADS breast composition, four density patterns were classified in caudocranial and mediolateral mammography by a designated radiologist in each hospital. Multiple linear logistic regression was used for statistical analyses. Results: Inter-rater reliability between the initial report and the report made by another blinded radiologist was high (Pearson's co-efficient=0.81). Overall, the age, body mass index, family history of breast cancer, and duration of hormone replacement therapy correlated with the mammographic density patterns. In pre-menopausal women, the high body mass index and parity (= 2) were associated with low mammographic density. In post-menopausal women, older age, high body mass index, negative family history of breast cancer, and duration of hormone replacement therapy were associated with low mammographic density. Conclusion: Our data showed that the mammographic breast density patterns correlated with risk factors for breast cancer in Korean women. While the proportion of high-density mammography is higher in Korean women, the incidence of breast cancer is lower than in the western population. This maybe dependent on other unknown factors.