Abortion Does Not Add to Breast Cancer Risk

BOSTON, April 23 -- Neither abortion nor miscarriage contributes to breast cancer risk, according to a prospective observational study of more than 100,000 women.

In the Nurses' Health Study II, premenopausal women were not at an elevated breast cancer risk after induced (hazard ratio 1.01, 95% confidence interval 0.88 to 1.17) or spontaneous abortion (HR 0.89, 95% CI 0.78 to 1.01) found Karin B. Michels, Sc.D., Ph.D., of Brigham and Women's Hospital here, and colleagues.

The study, published in the April 23 issue of the Archives of Internal Medicine, echoed the conclusion of a National Cancer Institute panel in 2003 that found no association of induced abortion with breast cancer risk.

The nurses' study was the latest addition to a debate begun more than 20 years ago on whether uncompleted pregnancy would lead to a greater risk of breast cancer by raising pregnancy hormones. The hypothesis was that raising hormones would foster tumor cell proliferation, without the protective effect from a completed pregnancy

The study also supports two retrospective registry studies from Scandinavia that found no increased breast cancer risk with induced abortion, though these did not control for other variables. Case-control studies have been inconsistent for induced abortion, but most have come down against a link between breast cancer and miscarriage.

However, reporting bias has challenged both types of studies, Dr. Michels and colleagues noted.

"When a sensitive personal issue such as induced abortion is studied, underreporting is probable," they wrote.

While registry studies might draw from the most objective source of information on induced abortion, they added, abortion registries are scarce, seldom have data on potential confounding variables, and usually do not capture women's entire reproductive history.

To address some of these issues, the researchers examined biennial survey responses from 105,716 women participating in the ongoing Nurses' Health Study II.

All the women were free of cancer and ages 29 to 46 at baseline for the analysis in 1993 when questions on miscarriage and abortion were first included on the questionnaire. The first survey asked about any prior pregnancies lasting less than six months and thereafter women were asked only about pregnancies in the intervening period.

At baseline, 93% of the women were premenopausal, and 66% remained so throughout follow-up. Nearly all were non-Hispanic white (92%).

Of the 99.1% of the total cohort who answered the question on spontaneous and induced abortion in the baseline survey, 15.2% reported one or more induced abortions and 20.6% reported at least one spontaneous abortion. Both findings were the same among the women included in the analysis.

Among those who provided information on abortion or miscarriage at least once in follow-up questionnaires between 1993 and 2003, 14.6% reported ever having had an induced abortion and 22.3% reported ever having had a miscarriage.

During 973,437 person-years of follow-up, participants reported 1,458 newly diagnosed cases of invasive breast cancer.

After adjustment for family history of breast cancer, body mass index (current and at age 18), oral contraceptive use, menopause status, and other breast cancer risk factors, the hazard ratio findings were:

• 1.01 for one or more induced abortions (95% CI 0.88 to 1.17).
• 0.95 for two or more induced abortions (95% CI 0.68 to 1.31, P for trend=0.98).
• 0.89 for one or more miscarriages (95% CI 0.78 to 1.01).
• 0.82 for two or more miscarriages (95% CI 0.62 to 1.08, P for trend=0.06)

Women who had a first spontaneous abortion before age 20 were less likely to have breast cancer than those who had never had a miscarriage. This result, though, was based on only 13 cases and no trend emerged for age (P=0.27 for trend).

No trend was observed with age at first induced abortion (P=0.41).

Whether a woman had a completed pregnancy in her life did not affect the association between breast cancer and either induced abortion (P for interaction=0.54) or miscarriage (P for interaction=0.10). Likewise, breast cancer incidence did not significantly differ between women who had an induced abortion before their first birth, had an induced abortion after their first birth, or had no abortion (P for interaction=0.10).

There were no differences in the associations by breast cancer subtype, except that parous women who reported an induced abortion were significantly more likely to have progesterone receptor-negative breast cancer (HR 1.58, 95% CI 1.13 to 2.20).

However, the researchers cautioned, "no obvious mechanisms can be provided for these subgroup findings; thus, chance has to be considered as a possible explanation."

They added, though, "we cannot exclude a modest association in subgroups defined by known breast cancer risk factors, timing of abortion, or parity."

Because most of the women were still premenopausal at the end of follow-up, "their reproductive history may still have been incomplete." Also, the oldest participants were well into their reproductive years--age 27--in 1973 when abortion was legalized by Roe v. Wade.

Yet the nurses in the study "may not be representative of the social classes most likely to use abortion services" and the breast cancer cases in the study population were almost exclusively premenopausal limiting generalizability of the findings to postmenopausal women.