| Although parity and OC use have been associated consistently with a decreased risk of ovarian cancer, the association of other menstrual and reproductive factors is still inconsistent across published studies. Given the disadvantages derived from hospital-based case-control design, controls could be misclassified on disease status and, thereby, introduce bias into the analysis. This may be a particular concern for ovarian cancer given the silent nature of ovarian cancer.; The purpose of this study was to assess misclassification bias associated with disease status among control women in a hospital-based case-control study as well as to assess whether exclusion of control women developing ovarian cancer subsequently would affect the consistency of the relation between reproductive and menstrual factors and ovarian cancer.; Case women comprised 434 patients aged 18 to 95 years (54.9 +/- 14.1) who were treated with ovarian epithelial cancer at Roswell Park Cancer Institute (PRCI) between 1982 and 1998 and also completed a comprehensive epidemiologic questionnaire. Control women were 3,429 age-matched patients with non-malignancy diagnoses and completed a comprehensive epidemiologic questionnaire. Information related to clinical features of malignancy and Information about menstrual, reproductive and other factors of interest were obtained through self-administered questionnaire. Unconditional logistic regression was used to estimate crude and adjusted odds ratios (ORs) along with 95% confidence intervals (CIs). Identifying information of 3,429 control women was linked to the New York State Cancer Registry during 1983--1999 to detect any incident cancer during the follow-up. The expected numbers of cancer among control women were compared to the observed ones using the Poisson distribution for rates. Multivariate logistic regression analyses were carried out to evaluate the association of reproductive risk factors after removing control women who developed cancer subsequently. |
