To further understand the factors that contribute to interindividual variation in mammographic density, we evaluated the relationship between education level and each component of the mammographic density measures. Study participants included 535 women between 40 and 65 years of age who received a mammogram for a population-based twin and family study. Mammographic density was measured from digital mammograms using a computer-assisted thresholding method. To avoid negative confounding by obesity level, we calculated BMI-adjusted mammographic measures. Thereafter, each of the mammographic density measures was t-transformed using its mean value and SD for each age strata. The level of education was chosen as a marker representing socioeconomic status at the individual level. A linear mixed model considering familial correlations was used for analyses. In the unadjusted analysis for all women, the BMI-adjusted nondense area gradually decreased with an increase in education level (P for trend=0.017). This association persisted after adjusting for menstrual and reproductive factors. When we repeated the analysis according to menopausal status, an inverse association between education level and nondense area was evident in premenopausal women, whereas the inverse association attenuated to a statistically insignificant level after adjusting for menstrual and reproductive factors in postmenopausal women. Absolute dense area and percentage dense area were not associated with education level. The significant association between nondense area and education level after eliminating the effect of age and BMI suggests that socioeconomic factors may have an influential role in determining the amount of fat tissue in the breast.
aDepartment of Epidemiology, School of Public Health
bInstitute of Health Environment, Seoul National University
cDepartment of Family Medicine, Samsung Medical Center and Center for Clinical Research, Samsung Biomedical Research Institute, Sungkyunkwan University School of Medicine, Seoul, South Korea
Correspondence to Yun-Mi Song, MD, MPH, PhD, Department of Family Medicine, Samsung Medical Center and Center for Clinical Research, Samsung Biomedical Research Institute, Sungkyunkwan University School of Medicine, Irwon-ro 81, Gangnamgu, Seoul 135-710, South Korea Tel: +82 2 3410 2442; fax: +82 2 3410 0388; e-mail: email@example.com
Received July 21, 2014
Accepted December 3, 2014