The relation between contraceptive methods and risk of pelvic endometr
iosis has been analyzed in a case-control study. Cases were 376 women
with laparoscopically or laparotomically confirmed pelvic endometriosi
s admitted to a network of Obstetrics and Gynecology Clinics in Lombar
dy, Northern Italy. Controls were 522 women admitted for acute non-gyn
ecological and non-obstetrics conditions to the same hospitals where c
ases had been identified. A total 153 women (40.3%) out of the 377 cas
es and 254 (29.7%) out of the 522 controls reported ever oral contrace
ptive (OC) use: the corresponding relative risk (RR) was 1.6 (95% conf
idence interval, CI, 1.2-2.2). The risk was restricted to ex-OC users
(RR 1.7, 95% CI 1.3-2.4), the estimated RR for current users being 0.8
(95% CI 0.5-2.9). No clear relation emerged with duration, recency an
d latency of OC use and risk of endometriosis. In comparison with neve
r IUD users, the risk for ever users was 1.3 (95% CI 0.6-2.8), and no
clear relation emerged with duration of use. Likewise, no association
was observed between barrier method of contraception and risk of endom
etriosis (RR ever vs never users 0.5, 95% CI 0.3-2.4). The role of sel
ection and other biases should be considered in the interpretation of
epidemiological data on the role of OC on the risk of endometriosis.