Mammographic features and breast cancer risk: effects with time, age, and menopause status

J Natl Cancer Inst. 1995 Nov 1;87(21):1622-9. doi: 10.1093/jnci/87.21.1622.

Abstract

Background: Mammographic images from women with a high proportion of epithelial and stromal breast tissues are described as showing high-density parenchymal patterns. Most past studies that noted an increase in breast cancer risk associated with mammographic parenchymal patterns showing high density either 1) lacked information on other breast cancer risk factors, 2) were too small, or 3) included insufficient follow-up time to adequately resolve persisting doubts whether mammographic features are "independent" measures of breast cancer risk and not a detection artifact.

Purpose: The purpose of this study was twofold: 1) to evaluate the associations between mammographic features and other breast cancer risk factors and 2) to assess effects of mammographic features on breast cancer risk by time, age, and menopause status.

Methods: To address these questions, we analyzed detailed information from a large, nested case-control study with 16 years of follow-up. This study used information from both screening and follow-up phases of the Breast Cancer Detection Demonstration Project, a nationwide program that offered annual breast cancer screening for more than 280,000 women from 1973 to 1980. Mammographic features were assessed from the base-line screening mammographic examination for 1880 incident case subjects and 2152 control subjects. Control subjects were randomly selected from women of the same age and race as each case subject. Control subjects attended the same screening center as the case subject and were free of breast cancer at the case subject's date of diagnosis. Odds ratios (ORs) with 95% confidence intervals (CIs) provided estimates of the relative risk of breast cancer.

Results: Mammographic features were associated with known breast cancer risk factors. However, the high-density parenchymal pattern effects were independent of family history, age at first birth, alcohol consumption, and benign breast disease. The increase risk for women with Wolfe's two high-density parenchymal patterns, P2 (OR = 3.2; 95% CI = 2.5-4.0) and Dy (OR = 2.9; 95% CI = 2.2-3.9), was explained primarily by measured percent of the breast with dense mammographic appearance. Compared with women with no visible breast density, women who had a breast density of 75% or greater had an almost fivefold increased risk of breast cancer (95% CI = 3.6-7.1). These effects persisted for 10 or more years and were noted for both premenopausal and postmenopausal women of all ages.

Conclusions: Of the breast cancer risk factors assessed in the participants, high-density mammographic parenchymal patterns, as measured by the proportion of breast area composed of epithelial and stromal tissue, had the greatest impact on breast cancer risk. Of the breast cancers in this study, 28% were attributable to having 50% or greater breast density.

MeSH terms

  • Age Factors
  • Breast Neoplasms / diagnostic imaging*
  • Breast Neoplasms / prevention & control
  • Female
  • Humans
  • Mammography*
  • Mass Screening*
  • Menopause*
  • Middle Aged
  • Risk Factors
  • Time Factors
  • United States