All studies included in this pooled analysis had approval from ethics committees, and written informed consent was obtained from all study participants. Study characteristics are reported in the appendix.
We used primary data from all studies in the OCAC at the time this analysis was initiated; the study questionnaires included questions about endometriosis. Data for endometriosis were reported in 13 case–control studies of ovarian cancer. One study was undertaken in Australia,9 (link) three in Europe,26 (link), 27 (link), 28 (link) and nine in the USA.5 (link), 8 (link), 29 (link), 30 (link), 31 (link), 32 (link), 33 (link), 34 (link), 35 (link), 36 (link) The characteristics of the 13 studies are presented in table 1. Data for endometriosis were self-reported in all studies. Women with missing endometriosis data and those with non-epithelial tumours were excluded. Data for origin of endometriosis (endometrioma, peritoneal, or deep infiltrating disease) were not available. Our analysis dataset consisted of data from 23 144 women (7911 with invasive ovarian cancer, 1907 with borderline ovarian cancer, and 13 326 controls). Subsets of data from five studies have been reported previously (Australian Cancer Study, Australian Ovarian Cancer Study [AUS],9 (link) Diseases of the Ovary and their Evaluation Study [DOV],5 (link) Hawaii Ovarian Cancer Study [HAW],7 (link) Malignant Ovarian Cancer Study [MAL],7 (link) and University of Southern California, Study of Lifestyle and Women's Health [USC]8 (link)). We excluded one OCAC study (from Poland37 (link)) from this analysis because the investigators thought that the endometriosis data were not reliable.

Description of studies included in the analysis

Study nameStudy abbreviationStudy typeMethod of data collectionAscertainment period
Asia-Pacific
AustraliaAustralian Cancer Study*, Australian Ovarian Cancer Study*9 (link)AUSPopulation basedSelf-completed questionnaire, checked by trained research nurse2002–06
Europe
GermanyGerman Ovarian Cancer Study26 (link)GERPopulation basedSelf-completed questionnaire1992–98
DenmarkMalignant Ovarian Cancer Study27 (link)MALPopulation basedIn-person or phone interview1994–99
UKUnited Kingdom Ovarian Cancer Population Study28 (link)UKOPopulation basedSelf-completed questionnaire2006–07
USA
CTConnecticut Ovary Study29 (link)CONPopulation basedIn-person interview1999–2003
WADiseases of the Ovary and their Evaluation Study5 (link)DOVPopulation basedIn-person interview2002–05
HIHawaii Ovarian Cancer Study30 (link)HAWPopulation basedIn-person interview1994–2007
Western PA, northeast OH, western NYHormones and Ovarian Cancer Prediction Study31 (link)HOPPopulation basedIn-person interview2003–08
North central states (MN, SD, ND, IL, IA, WI)Mayo Clinic Ovarian Cancer Study32 (link)MAYClinic basedIn-person interview2000–08
NCNorth Carolina Ovarian Cancer Study33 (link)NCOPopulation basedIn-person interview1999–2008
NH and eastern MANew England Case-Control Study of Ovarian Cancer34 (link)NECPopulation basedIn-person interview1999–2008
Orange County and San Diego County, CAUniversity of California, Irvine Ovarian Cancer Study35 (link)UCIPopulation basedSelf-completed questionnaire1995–2005
Los Angeles County, CAUniversity of Southern California, Study of Lifestyle and Women's Health8 (link), 36 (link)USCPopulation basedIn-person interview1993–2005

Combined for the purpose of the analysis.

Data for timing of endometriosis available.

In each study, information was provided about potential confounding variables that were previously noted to be related to ovarian cancer risk: age, ethnic origin, parity, breastfeeding, duration of oral contraceptive use, family history of ovarian cancer, weight, height, and history of tubal ligation. All data were cleaned and checked for internal consistency and clarifications were requested from the original investigators when needed.
Free full text: Click here