Table 2.
Test set‐based hazard ratios per covariate‐adjusted standard deviation (SD) from univariable and multivariable analyses of mBCRS and polygenic risk score (n = 794).
| Full follow‐up a | Excluding first 2 years | |||||
|---|---|---|---|---|---|---|
| HR (95% CI) b | Z‐score | P‐value | HR (95% CI) b | Z‐score | P‐value | |
| Univariate models c | ||||||
| mBCRS | 1.78 (1.51, 2.11) | 6.70 | 2.1 × 10−11 | 1.83 (1.52, 2.19) | 6.45 | 1.1 × 10−10 |
| PRS | 1.57 (1.35, 1.83) | 5.84 | 5.1 × 10−9 | 1.62 (1.37, 1.92) | 5.69 | 1.3 × 10−8 |
| Multivariable model d | ||||||
| mBCRS | 1.74 (1.46, 2.06) | 6.23 | 4.6 × 10−10 | 1.76 (1.47, 2.12) | 6.03 | 1.6 × 10−9 |
| PRS | 1.55 (1.31, 1.83) | 5.14 | 2.8 × 10−7 | 1.59 (1.33, 1.92) | 4.98 | 6.3 × 10−7 |
Sample sizes: full follow‐up, n = 794 with 443 events; excluding first 2 years, n = 673 with 326 events.
Full follow‐up length: mean = 5.2 years, SD = 2.6.
Per covariate‐adjusted standard deviation increase in mBCRS score (or PRS), accounting for: age at blood draw, menopause status, body mass index, interaction term for BMI and menopause, physical activity, alcohol consumption, age at first birth (among parous), total number of births, age at menarche, menopause age (among postmenopausal), smoking pack‐years, previous number of breast biopsies, family history of breast cancer (number of affected sisters, youngest age of sister’s diagnosis), educational attainment, and durations of postmenopausal hormone use and breastfeeding, and standardized to the distribution of the noncases.
Results displayed are from two separate models including either mBCRS or PRS.
Results displayed are for both mBCRS and PRS that were included as covariates in a single model.