Skip to main content
. Author manuscript; available in PMC: 2012 Jul 1.
Published in final edited form as: Cancer Epidemiol Biomarkers Prev. 2011 May 6;20(7):1391–1397. doi: 10.1158/1055-9965.EPI-11-0022

Table 2.

Relative risk (RR) estimates and 95% confidence intervals (CIs) for the association between oral contraceptive (OC) use and risk of mortality in the Women's Contraceptive and Reproductive Experiences (CARE) Study

All-cause deaths (N=1064) Breast cancer deaths (N=828)
Baseline variable No. RR(95% CI) No. RR(95% CI)
OC use status
 Never 255 1.00 172 1.00
 Ever 809 1.01(0.86–1.19) 656 1.03(0.85–1.25)
 Former 767 1.01(0.86–1.19) 616 1.03(0.85–1.24)
 Current 41 0.89(0.62–1.28) 39 0.95(0.65–1.40)
OC use duration (years)
 <1 187 0.94(0.77–1.15) 142 0.92(0.73–1.17)
 1–<5 295 1.17(0.97–1.41) 230 1.15(0.93–1.43)
 5–<10 180 0.97(0.78–1.19) 154 1.02(0.80–1.29)
 ≥10 147 0.93(0.74–1.16) 130 1.00(0.78–1.28)
P-trend 0.52 0.35
Age at first OC use (years)
 <20 324 1.01(0.82–1.24) 283 1.06(0.84–1.35)
 20–24 264 0.97(0.80–1.18) 220 1.01(0.81–1.27)
 25–29 136 1.03(0.83–1.28) 92 0.93(0.71–1.22)
 ≥30 84 1.06(0.82–1.36) 60 1.09(0.81–1.47)
Years between last OC use and breast cancer diagnosis
 ≥25 236 1.11(0.92–1.34) 175 1.12(0.90–1.41)
 20–24 191 1.01(0.82–1.23) 145 0.97(0.77–1.24)
 10–<20 250 0.94(0.76–1.15) 216 1.00(0.80–1.27)
 <10 131 0.82(0.64–1.06) 119 0.87(0.65–1.15)
Years between menarche and first OC use
 <9 408 1.01(0.83–1.23) 355 1.06(0.85–1.33)
 9–<15 253 0.95(0.79–1.15) 198 0.97(0.77–1.21)
 ≥15 147 1.08(0.88–1.33) 102 1.05(0.82–1.35)
Years of OC use before first full-term pregnancy (FFTP)
 1st OC use after FFTP 396 1.01(0.86–1.20) 314 1.05(0.86–1.29)
 No FFTP 125 1.06(0.84–1.35) 89 0.90(0.68–1.20)
 <3 61 0.89(0.66–1.21) 51 0.89(0.63–1.25)
 3–<6 98 1.09(0.84–1.41) 88 1.15(0.86–1.52)
 ≥6 125 0.94(0.74–1.21) 110 0.96(0.73–1.27)
Calendar year of first OC use
 <1972 544 1.05(0.89–1.24) 421 1.07(0.88–1.31)
 ≥1973 264 0.87(0.69–1.09) 234 0.87(0.67–1.13)

Total number of deaths may vary due to missing values in corresponding variables. All models use age (in days) as the time metric and stratify on age (in years) at breast cancer diagnosis. Covariates included in the models are study site, race, estrogen receptor status, tumor stage, education level, smoking status, alcohol consumption, number of comorbidities, number of mammograms within the 5 years before breast cancer diagnosis, body mass index 5 years before cancer diagnosis, and age at menarche.