Case-control study of borderline ovarian tumors: reproductive history and exposure to exogenous female hormones.

Women residing in 3 urban counties of western Washington State diagnosed with serous and mucinous borderline ovarian tumors or grade 1 malignancies, between 1980 and 1985 were interviewed to assess risk factors related to reproductive history. 121 subjects were interviewed in person, and compared to 158 controls ascertained and interviewed by random sampling via telephone. The stratified analysis technique, adjusting for parity, revealed a 60% reduction of risk in users of oral contraceptives, although neither duration, nor age at 1st use or years since last use of pills had any effect. Similarly, lactation reduced risk 50%, without any influence of duration. The reduction in risk with pill use was largely confined to women who had their 1st child after age 21. Having had 1 or 2 children conferred 0.7 of the risk of nulliparas, and 3 or more children 0.4. A history of miscarriage, but not of induced abortion, was more common in cases than in controls. History of tubal ligation was negatively correlated with borderline tumors, in women whose sterilization was less than 12 years before diagnosis. In general, borderline tumors have similar epidemiological patterns to those of malignant ovarian tumors, with-increased risk for nulliparas, women reporting infertility, those who have used oral contraceptives, and those who have not lactated.