Skip to main content
. 2013 Jan 12;28(5):1406–1417. doi: 10.1093/humrep/des466

Table II.

Adjusted ORs to estimate risk of type I/II invasive ovarian cancer for reproductive factors.

Reproductive factors Controls (N = 2100) Type I (N = 358)
Type II (N = 1108)
Phet
n (%) n (%) ORa (95% CI) n (%) ORa (95% CI)
Parity
 Nulliparous 378 (18.0) 160 (44.7) 1.00 (Ref) 287 (25.9) 1.00 (Ref) <0.001b
 1 267 (12.7) 53 (14.8) 0.44 (0.31–0.63) 137 (12.4) 0.64 (0.49–0.83)
 2 667 (31.8) 80 (22.4) 0.27 (0.20–0.36) 337 (30.4) 0.63 (0.50–0.78)
 ≥3 788 (37.5) 65 (18.2) 0.15 (0.11–0.21) 347 (31.3) 0.44 (0.35–0.55)
Breastfeeding (parous)
 0 months 699 (40.6) 104 (52.5) 1.00 (Ref) 447 (54.5) 1.00 (Ref) 0.006b
 >0 to ≤4 months 290 (16.8) 42 (21.2) 1.03 (0.70–1.51) 122 (14.9) 0.69 (0.54–0.89)
 >4 to ≤10 months 242 (14.1) 15 (7.6) 0.43 (0.24–0.76) 98 (11.9) 0.65 (0.50–0.86)
 >10 to ≤19 months 244 (14.2) 13 (6.6) 0.38 (0.21–0.69) 93 (11.3) 0.63 (0.48–0.83)
 >19 months 247 (14.3) 24 (12.1) 0.79 (0.49–1.27) 61 (7.4) 0.47 (0.34–0.64)
Ptrendc = 0.09 Ptrendc = 0.001 0.85d
Age at menarche
 <12 years 423 (20.1) 72 (20.1) 0.99 (0.71–1.38) 246 (22.2) 1.16 (0.94–1.45) 0.65
 12 years 572 (27.2) 100 (27.9) 1.00 (Ref) 290 (26.2) 1.00 (Ref)
 >12 years 1105 (52.6) 186 (52.0) 0.98 (0.75–1.28) 572 (51.6) 1.04 (0.87–1.25)
Ptrendc = 0.67 Ptrendc = 0.03 0.32d
Menstrual cycle lengthe
 <28 days 245 (11.7) 47 (13.4) 1.00 (0.70–1.42) 142 (12.9) 1.06 (0.83–1.34) 0.42b
 28 days 1025 (49.1) 196 (55.7) 1.00 (Ref) 561 (51.0) 1.00 (Ref)
 >28 to ≤30 days 497 (23.8) 69 (19.6) 0.73 (0.54–0.99) 267 (24.3) 0.99 (0.82–1.20)
 >30 days 170 (8.2) 19 (5.4) 0.68 (0.41–1.13) 66 (6.0) 0.83 (0.61–1.13)
 Irregular cycles 149 (7.1) 21 (6.0) 0.76 (0.47–1.24) 65 (5.9) 0.82 (0.60–1.13)
Ptrendc = 0.05 Ptrendc = 0.15 0.32d
Age at natural menopausef
 <45 years 32 (2.9) 9 (5.7) 1.76 (0.77–4.00) 21 (3.0) 0.96 (0.51–1.81) 0.02b
 45 to <50 years 58 (5.2) 11 (6.9) 1.25 (0.60–2.61) 49 (7.0) 1.30 (0.80–2.13)
 50 to <53 years 991 (88.7) 129 (81.1) 1.00 (Ref) 600 (85.6) 1.00 (Ref)
 ≥53 years 36 (3.2) 10 (6.3) 1.99 (0.91–4.36) 31 (4.4) 1.44 (0.81–2.56)
Ptrendc = 0.67 Ptrendc = 0.43 0.007d
OC pill use
 Never or <3 months 766 (36.5) 161 (45.0) 1.00 (Ref) 581 (52.4) 1.00 (Ref) 0.06b
 3 to <12 months 161 (7.7) 33 (9.2) 1.14 (0.75–1.73) 95 (8.6) 0.91 (0.68–1.21)
 12 to <24 months 164 (7.8) 25 (7.0) 0.81 (0.51–1.28) 92 (8.3) 0.82 (0.62–1.10)
 24 to <60 months 378 (18.0) 58 (16.2) 0.85 (0.61–1.18) 149 (13.5) 0.60 (0.48–0.76)
 ≥60 months 631 (30.1) 81 (22.6) 0.66 (0.49–0.89) 191 (17.2) 0.43 (0.35–0.53)
 Including non-users Ptrendc = 0.007 Ptrendc = <0.001 0.003d
IUD use 353 (16.8) 39 (10.9) 0.71 (0.49–1.01) 164 (14.8) 1.00 (0.81–1.24) 0.06b
Infertility
 None 1664 (79.2) 274 (76.5) 1.00 (Ref) 879 (79.3) 1.00 (Ref) 0.53b
 Male 41 (2.0) 10 (2.8) 1.24 (0.60–2.55) 22 (2.0) 0.85 (0.49–1.47)
 Female 161 (7.7) 38 (10.6) 1.35 (0.92–1.97) 95 (8.6) 1.05 (0.80–1.38)
 Cause not found 234 (11.1) 36 (10.1) 0.89 (0.61–1.30) 112 (10.1) 0.87 (0.67–1.11)
Endometriosis 165 (7.9) 51 (14.3) 1.92 (1.36–2.71) 95 (8.6) 1.09 (0.83–1.43) 0.003b
Painful periodse 677 (32.4) 153 (43.3) 1.55 (1.22–1.95) 413 (37.3) 1.20 (1.03–1.41) 0.04b
Hysterectomyg 183 (8.7) 23 (6.4) 0.71 (0.45–1.13) 111 (10.0) 1.16 (0.89–1.51) 0.03b
Tubal ligation 419 (20.0) 26 (7.3) 0.40 (0.26–0.60) 168 (15.2) 0.91 (0.74–1.12) <0.001b

aModels were adjusted for age (continuous), study center (Massachusetts, New Hampshire), study phase (1992–1997, 1998–2003, 2003–2008), parity (0, 1, 2, >2), OC pill use (0, 3 months to <5 years, ≥ 5 years), family history of ovarian cancer (yes/no), family history of breast cancer (yes/no) and tubal ligation (yes/no) unless noted otherwise.

bThe P-value for heterogeneity (Phet) is from the likelihood-ratio test that compares a model with the same estimate for the association with the exposure of interest (e.g. categories of parity) across type I/II categories to a model, which allows the association of interest to vary across type I/II categories. It indicates if there were statistically significant differences in the association of interest between the different outcome categories.

cThe Ptrend is from the Wald statistic using a continuous variable.

dThe Phet is from the likelihood-ratio test that compares a model with the same estimate for the association with the exposure of interest (trend variable) across type I/II categories to a model that allows the exposure of interest to vary across type I/II categories.

eNumbers may not add up to total due to missing data.

fAnalyses were restricted to post-menopausal women and were additionally adjusted for an indicator if a participant had not had a natural menopause.

gModels were additionally adjusted for ever use of post-menopausal hormones (yes/no).

HHS Vulnerability Disclosure