Abstract 3249: Differential impact of body mass index (BMI) on absolute and percent breast density; implications regarding their use as breast cancer risk (BCR) biomarkers

Percent breast density (PBD) is a commonly used biomarker of BCR. However, its use is greatly confounded by the strong influence of non-dense breast tissue on its measurement and factors, such as BMI, which have a direct impact on this non-dense tissue. Consequently, BMI, a potent BCR factor, is actually negatively correlated with PBD. We propose that absolute breast density (ABD), a more direct determinant of BCR which can be easily and accurately measured on a routine basis, is a more accurate and valid biomarker of BCR. To address this issue, we compared the correlation between PBD and ABD with baseline demographics (e.g., age, BMI, waist:hip ratio, reproductive parameters, family history of BC and personal history of breast biopsies) and dietary and physical activity variables (measured using a modified version of the NCI Diet History Questionnaire and the International Physical Activity Questionnaire, respectively) in a group of 169 healthy postmenopausal women enrolled in a chemoprevention trial testing the individual and combined effects of the antiestrogen Raloxifene and the omega-3 preparation Lovaza on BD used as a biomarker of BCR (NCT00723398). A volumetric breast density software (Volpara, Matakina, Wellington, New Zealand) was used for determination of ABD and PBD since it may be superior to the two-dimensional method in predicting BCR. As observed previously, a strong negative correlation between PBD and BMI was observed (Rho = -0.5, P