Comparison of breast density in Asian women and Caucasian women

Abstract

Background: Breast cancer screening trials using mammography show a 30% reduction in breast cancer mortality, but dense breast tissue obscures small breast tumors, increasing the rate of missed cancers and contributing to breast cancer mortality. California has one of the largest Asian population in United States. Previous studies have reported the variation of breast density in different races with confl icting results. If Asian women have mostly dense breast tissue, the effi cacy of breast cancer screening with mammography will be diminished. The aim of this study is to determine breast density and factors affecting it in Asian and Caucasian women. Methods:100 Caucasian and 100 Asians women were randomly selected for mammogram breast density readings from women who completed questionnaires for a parallel breast screening study. These women had a mammogram taken and were used to grade breast density. Two radiologists blinded to the patients’ race and name graded breast density. Breast density was correlated to the women’s age, race, parity and status of exogenous hormone therapy. Fitted multiple regressions and F-tests were performed for exclusion of terms in the model. Results: The average age was similar in both groups (55 and 54 years old). Breast density was slightly increased in Asian compared to Caucasian: 57% of Asian had dense breasts compared to 44% of Caucasian. Race alone was found to have a marginal effect on breast density with a p value of .087. Analysis on additional data that could account for the slightly denser tissue including age difference, exogenous hormonal therapy, parity, breastfeeding, breast size, and body mass index (BMI) was performed. Only 18%. Parity was similar. Breast size was markedly discrepant in Caucasian compared to Asian. 78% of Asian have bra cup size B or below, whereas 58% of Caucasian have bra cup size C and above. This may be a refl ection of body size and habitus which would infl uence breast size and composition. Half of the Caucasian were overweight compared with only 35% of Asian. Several series of modeling steps were performed adding different confounders and the fi nal model included body mass index and age as a variable with increased rates and found that these confounders were not statistically signifi cant. Interestingly, there is a signifi cant negative age effect and the BMI effect. Conclusions: Our study found that the background breast density of Caucasian was less than that seen in Asian women, but when the confounders of age and BMI are taken into account, race is no longer signifi cant with a p-value of 0.248. Further studies comparing breast density of Asian population in Hong Kong with the population in North America is underway.