OBJECTIVE:To comparetwomethodsforcalculatinglifetimeovulatorycycles(LOC) to determine if more detailed menstrual cycle information results in stronger associations withovariancancer.METHODS:Using data from 232 cases and 242 controls in a population-based study ofovariancancer, we compared a standard method forcalculatingLOC with a second method that had more detailed information on menstrual characteristics. Odds ratios forovariancancerby number of LOC were estimated using unconditional logistic regression.RESULTS:The average number of LOC was 29 fewer for the second method that had more detailed menstrual cycle information, as compared to the standard method (p < 0.0001). The difference was due primarily to the second method considering episodes of missed/irregular periods. Associations between LOC andovariancancerwere weaker for the second method than the standard method. Further analyses suggested that a reduced number ofovulatorycyclesdue to menstrual irregularity was associated with increasedovariancancerrisk, in contrast to the protective effects observed for fewerovulatorycyclesdue to pregnancy or oral contraceptive use.CONCLUSION:Obtaining additional details on menstrual factors that affect LOC, particularly missed or irregularcycles, provides important information onovariancancerrisk. Our data suggest that episodes of anovulation due to menstrual disturbances should be evaluated separately from anovulation due to pregnancy or oral contraceptive use.
