The influence of prior oral contraceptive use on risk of endometriosis is conditional on parity

Abstract

To estimate the influence of prior oral contraceptive pill (OCP) use on future diagnosis of endometriosis in young women. Prospective cohort study, the Australian Longitudinal Study on Women's Health. Community-based sample. 9,585 women age 18-23 at study onset. None. Risk of self-reported endometriosis estimated with Cox proportional-hazards regression with time-dependent covariates. Compared with never users, endometriosis hazard ratios in nulliparous women with <5 years and ≥ 5 years of OCP use (preceding diagnosis) were 1.8 (95% CI, 1.30-2.53) and 2.3 (95% CI, 1.59-3.40), respectively. Similar risk was seen in both women reporting infertility and unsure fertility. In parous women with <5 years of use, the hazard ratio for endometriosis was 0.41 (95% CI, 0.15-0.56) and for ≥ 5 years of use was 0.45 (95% CI, 0.16-1.23). Women reporting early noncontraceptive OCP use had a twofold higher risk (odds ratio 2.07; 95% CI, 1.72-2.51). Prior OCP exposure reduces the risk of diagnosis of endometriosis in parous women but increases it among nulliparous women; these associations appear unaffected by fertility status. An increased risk of endometriosis diagnosis seen in women reporting early noncontraceptive OCP use may explain some of the positive OCP risk seen in nulliparous women.