Skip to main content

Table 3 Coefficients (standard errors) for covariates in multi-variable linear model predicting the square root of percent mammographic density, overall and by study site

From: A genome-wide linkage study of mammographic density, a risk factor for breast cancer

Covariate All sites together (n= 3,253) Australia (n= 1,534) Northern California
(n= 578)
Ontario
(n= 1,166)
Intercept 6.17 (0.72) 5.56 (1.00) 6.62 (1.69) 6.01 (1.33)
Age at mammogram -0.057 (0.0042) -0.056 (0.0056) -0.040 (0.011) -0.061 (0.0076)
Weight, kg -0.065 (0.0024) -0.058 (0.0034) -0.063 (0.0054) -0.074 (0.0044)
Height, cm 0.041 (0.0044) 0.041 (0.0060) 0.033 (0.103) 0.047 (0.0081)
Weight squared (centered at 70 kg) 0.00050 (6.9e-5) 0.00049 (1.1e-4) 0.00049 (1.1e-4) 0.00048 (1.4e-4)
Number of live births -0.103 (0.021) -0.100 (0.028) -0.109 (0.047) -0.096 (0.039)
Postmenopausal status -0.183 (0.076) -0.191 (0.109) -0.056 (0.185) -0.250 (0.131)
History of hormone therapy use 0.151 (0.0657) 0.133 (0.088) 0.111 (0.172) 0.090 (0.115)
  1. Non-significant covariates (age at menarche, square of mammogram age, square of height) were removed. For the analysis by site, only the covariates in the corresponding overall model were considered.