Breast cancer risk in women who have had children with different partners

The fetal antigen hypothesis suggests that the lowered risk of breast cance r in parous women may be afforded by the development of antibodies to fetal antigens that are structurally similar to mammary cancer antigens. It has previously been hypothesized that the likelihood of developing such antibod ies may be higher among women who have had children with more than 1 partne r. Utilizing information on parenthood and breast cancer available in natio nwide Swedish registers, we undertook a case-control study nested within a nation-wide cohort to address this issue. Number of partners fathering a ch ild was categorized as 1, 2 and 3 or more. All analyses were restricted to subjects with 2 or more births and encompassed a total of 20,881 women with breast cancer and 111,989 control women. In an unadjusted analysis, the ri sk of breast cancer was somewhat lower (odds ratio [OR] = 0.94, 95% confide nce interval [CI] 0.89-0.99) in women who had had children with 2 different partners compared with women who had had children conceived with the same partner. After adjustment for parity, age at first birth and educational le vel, however, the risk of breast cancer was slightly elevated (OR = 1.09, 9 5% Cl: 1.03-1.15). Among women who had had children with 3 or more differen t men, the pattern was similar. The present results provide no support for the hypothesis that greater antigenic exposure afforded by having children with more than I man may reduce the risk of breast cancer. It remains possi ble, however, that pregnancy may influence breast cancer risk through some immunologic mechanism; further testing of the fetal antigen hypothesis may require development of relevant laboratory measures. (C) 2001 Wiley-Liss, I nc.