Risk factors for epithelial ovarian tumours of borderline malignancy.

A case-control study was conducted on 91 cases with histologically-confirmed borderline ovarian tumours and 237 control subjects in hospital for acute non-gynaecological, hormonal or neoplastic disease. Women reporting three or more births, compared to nulliparae, had a relative risk (RR) estimate of 0.6, but this finding was not statistically significant (95% confidence interval (CI): 0.2-1.4). The risk of borderline tumours increased, although not significantly, with later age at first birth: compared to women reporting first birth at age 24 or before, the RRs were 1.3 and 1.7 in those reporting respectively their first birth at age 25-29 and 30 years or more. No significant relationship emerged between borderline ovarian cancer and age at menarche, menopausal status and lifelong menstrual pattern. Cases tended to report a later age at menopause than controls, but the trend in risk was not statistically significant. Nine cases (9.9%) and 68 controls (24.9%) reported oral contraceptive use: compared with never users the multivariate RR for ever users was 0.3, and the risk dropped with duration of use to 0.2 in users for two years or more (chi 2 (1) trend = 12.70, p less than 0.001). This study provides epidemiological evidence of a pathogenetic continuum between borderline and invasive ovarian tumours.