The pill and breast cancer: why the uncertainty?

The use of Oral contraceptive (OC) for family spacing is not associated with any change in the risk of breast cancer. This conclusion comes as a surprise because OCs reduce the risk of ovarian and endometrial cancer which like breast cancer are hormone dependent. Furthermore OCs may increase the risk of cervical abnormalities. Thus epidemiologically it is difficult to understand why the breast is spared any effect protective or otherwise of the longterm use of OCs. Current concern about OC and the breast focuses on the safety of using OCs in reproductive life particularly before the 1st pregnancy. Some epidemiological studies suggest an association between early use and later breast cancer; others do not. 1 possible reason for the uncertainty is that if use promotes precancerous changes that take time to progress into a tumor which in turn takes time to be diagnosed even recent studies may give misleading results. The New Zealand report includes most of the cases of breast cancer presenting in New Zealand in the 2 years up to June 1985 and shows no overall increased risk with over 10 years use of OCs and no increase in risk with longterm use before the 1st pregnancy and before age 25. The Swedish report finds an association between the longterm use of OCs and risk of breast cancer that does not appear to be particularly associated with use before 1st pregnancy. There is a 2-fold increase in risk for women who have ever used OCs for 8 or more years or for such a period before 1st pregnancy. Disregarding a latent effect of use of OCs on breast cancer there are other possible reasons for the discrepancy between these 2 studies. The Swedish study was conducted at a time of increasing consciousness of a possible association between OC and breast cancer and the accuracy of recall of the use of OCs going back 20 or more years may not have been comparable between cases and controls. Women with breast cancer have strong reasons for thinking hard about their history; controls have not. Yet the Swedish study matched the controls with cases very closely. In the New Zealand study the controls were not individually matched at all and this may be inadequate when one is concerned about a rapidly changing exposure rate. Finally the types of pill used by the women in these studies may have varied. The effect of OCs may be altered by hormone and dose as well as by duration of use and the womans age at use.