Supplemental Table 1.
The association between oral contraceptive use and risk of ovarian cancer, by attained age for 3989 BRCA1 mutation carriers
| Variable | Attained age of <50 y (n=3268) |
Attained age of >50 y (n=958) |
||||
|---|---|---|---|---|---|---|
| OvCa+, n (%)a | OvCa−, n (%)a | HR (95% CI)b,c | OvCa+, n (%)c | OvCa−, n (%)c | HR (95% CI)b,c | |
| n (%) | 150 (4.6) | 3118 (95.4) | 196 (20.5) | 762 (79.5) | ||
| Mean age at end of follow-up (SD), y | 43.6 (4.6) | 37.3 (8.5) | 57.9 (5.9) | 57.2 (6.3) | ||
| Age at end of follow-up, y | ||||||
| <45 | 77 (51.3) | 2402 (77.0) | 0 (0.0) | 0 (0.0) | ||
| 45–64 | 73 (48.7) | 716 (23.0) | 165 (84.2) | 663 (87.0) | ||
| >65 | 0 (0.0) | 0 (0.0) | 31 (15.8) | 99 (13.0) | ||
| Oral contraceptive use | ||||||
| Never (<6 mo) | 50 (34.7) | 467 (15.7) | 1.00 | 83 (46.9) | 251 (35.9) | 1.00 |
| Ever | 94 (65.3) | 2507 (84.3) | 94 (53.1) | 448 (64.1) | 0.63 (0.43–0.92) | |
| Ever, starting age unknown | 0 | 91 | 11 | 41 | ||
| Missing | 6 | 53 | 8 | 22 | ||
| Total duration of use | ||||||
| Never (<6 mo) | 50 (35.7) | 467 (16.7) | 1.00 | 83 (49.1) | 251 (37.0) | 1.00 |
| <5 y | 36 (25.7) | 524 (18.7) | 0.86 (0.52–1.44) | 31 (18.3) | 135 (19.9) | 0.62 (0.38–1.02) |
| 5–9 y | 28 (20.0) | 800 (28.5) | 0.48 (0.27–0.85) | 25 (14.8) | 119 (17.6) | 0.77 (0.45–1.33) |
| ≥10 y | 26 (18.6) | 1013 (36.1) | 0.27 (0.15–0.49) | 30 (17.8) | 173 (25.5) | 0.52 (0.31–0.86) |
| Ever, no period specific data | 4 | 261 | 19 | 62 | ||
| Missing | 6 | 53 | 8 | 22 | ||
| Trendd | P=2.8E-04 | P=.471 | ||||
Data are presented as number (percentage), unless otherwise indicated.
CI, confidence interval; EMBRACE, Epidemiological Study of Familial Breast Cancer; GENEPSO, Gene Etude Prospective Sein Ovaire; HR, hazard ratio; OvCa, ovarian cancer.
Schrijver et al. Oral contraceptive use and ovarian cancer risk for BRCA1/2 mutation carriers. Am J Obstet Gynecol 2021.
Distribution of variables at end of follow-up
Weighted: to account for the oversampling of affected individuals (breast and ovarian cancer)
Intrinsically stratified on study (EMBRACE, GENEPSO, other) and birth cohort (1920–1946, 1947–1954, 1955–1980). Clustered on family membership
Trend tests were based on the P value of the category-specific mean as a continuous variable of ever oral contraceptive users.