Clinical Question: Is there a link between abortion, induced or spontaneous, and breast cancer?
Study Design: Case-control
Synopsis: This study was a collaborative pooled analysis of primary data from 53 studies of 83,000 women with breast cancer from 16 countries. To be eligible, the studies had to occur in countries in which abortion is legal and had to enroll at least 100 women newly diagnosed with breast cancer, and the authors of those studies must have inquired systematically about the women’s reproductive history, including previous spontaneous and induced abortions.
Thirteen of the studies were prospective (44,000 women with breast cancer), and in 40 studies, data were collected retrospectively (39,000 women with breast cancer). The authors of this analysis tried to avoid some of the problems of variability across studies (e.g., different matching strategies, effect of individual centers) by stratifying the analyses within studies before pooling the data.
In the prospective studies, the risk of breast cancer in women with spontaneous abortions was the same as in women without spontaneous abortions (relative risk [RR] = 0.98; 95 percent confidence interval [CI], 0.92 to 1.04). A slightly decreased risk of breast cancer occurred in women with induced abortions (RR = 0.93; 95 percent CI, 0.89 to 0.96).
The findings in the retrospective studies, however, differed slightly. Although these studies showed no increase in cancer risk for spontaneous abortions, the risk associated with induced abortions was increased slightly (RR = 1.11). The differences probably are related to many of the biases inherent in retrospective studies. When possible, the authors of this analysis adjusted for socioeconomic status, education, oral contraceptive use, breastfeeding, age at menarche, weight, alcohol consumption, menopausal status, and family history of breast cancer. These adjustments did not alter the results.
Bottom Line: No association exists between induced or spontaneous abortions and the development of breast cancer. (Level of Evidence: 1a)