Confounder-adjusted Cox model estimates for various adiposity measures and their association with breast cancer risk
| Measures of adiposity | Effective sample size | Number of events | HR | Lower 95% CI | Upper 95% CI | p Value |
|---|---|---|---|---|---|---|
| Full model with individual adiposity measure | ||||||
| CSS | 87 334 | 1013 | 1.33 | 1.121 | 1.579 | 0.001 |
| Skirt size at 25 | 87 783 | 1017 | 1.006 | 0.958 | 1.056 | 0.829 |
| Skirt size at study entry | 87 661 | 1018 | 1.051 | 1.014 | 1.089 | 0.006 |
| BMI at recruitment | 87 615 | 1021 | 1.076 | 1.012 | 1.136 | 0.009 |
| Full model with specific adiposity measure combinations | ||||||
| CSS | 86 678 | 1008 | 1.266 | 1.041 | 1.538 | 0.018 |
| BMI | 1.037 | 0.97 | 1.109 | 0.287 | ||
| Skirt size at study entry | 87 003 | 1013 | 1.037 | 0.992 | 1.085 | 0.109 |
| BMI | 1.037 | 0.964 | 1.116 | 0.329 | ||
| Full model with one year time lag | ||||||
| CSS | 86 284 | 650 | 1.422 | 1.149 | 1.758 | 0.001 |
| Full model with multiple imputation model | ||||||
| CSS | 92 834 | 1090 | 1.3 | 1.102 | 1.534 | 0.002 |
Full model means adjusted for all potential confounders: Index of Multiple Deprivation (IMD), age at first and last period, number of pregnancies, hysterectomy, sterilisation, infertility, breast and ovarian cancer family history, oral contraceptive use (pill) use, hormone replacement therapy (HRT) use (from recruitment questionnaire); education, skirt size, continuing use of HRT, smoking, alcohol use, health status and cancer diagnosis (from study-entry questionnaire).
BMI, body mass index at recruitment per 5 kg/m2; CSS, change of skirt size per 10 years.