Table 3.

Risk of breast cancer according to quintiles of percent and absolute density for case-control triplets (n = 813)

Quintile of percent density
P*Quintile of absolute density
P*
1234512345
Percent dense volume (%)Dense volume (cm3)
No. of cases44455659674549466566
No. of controls119117107103961181131179797
OR unadjusted (95% CI)Reference1.04 (0.6, 1.7)1.42 (0.9, 2.3)1.55 (1.0, 2.5)1.89 (1.2, 3.0)0.002Reference1.14 (0.7, 1.8)1.03 (0.6, 1.7)1.76 (1.1, 2.8)1.78 (1.1, 2.8)0.002
OR adjusted for risk factors (95% CI)Reference1.06 (0.6, 1.7)1.48 (0.9, 2.4)1.72 (1.0, 2.8)2.21 (1.3, 3.7)0.001Reference1.15 (0.7, 1.9)1.04 (0.6, 1.7)1.82 (1.1, 2.9)1.86 (1.1, 3.0)0.003
Percent dense area (%)Dense area (cm2)
No. of cases41456856614046556763
No. of controls1221179510610212311610895100
OR unadjusted (95% CI)Reference1.14 (0.7, 1.9)2.13 (1.3, 3.4)1.57 (1.0, 2.5)1.78 (1.1, 2.9)0.01Reference1.22 (0.7, 2.0)1.57 (1.0, 2.5)2.17 (1.3, 3.5)1.94 (1.2, 3.1)0.0005
OR adjusted for risk factors (95% CI)Reference1.34 (0.8, 2.2)2.60 (1.6, 4.3)2.04 (1.2, 3.5)2.40 (1.3, 4.3)0.002Reference1.35 (0.8, 2.3)1.74 (1.0, 2.9)2.35 (1.4, 3.9)2.11 (1.3, 3.5)0.0005
  • NOTE: Shown are mean (SD) and percentage for, respectively, continuous and categorical variables.

  • * Unconditional logistic regression analysis. P value is from a test of linear trend with quintiles as an ordinary variable in the model.

  • Age at mammogram (y), age at first birth (y), weight (kg), height (cm), menopausal status (premenopausal, postmenopausal), and parity (parous, nonparous).