| |
Return to Medicine & Science Page
Summary and Conclusions from:
Abortion and Breast Cancer: A Layperson’s Annotated Guide to Medical Research Reports
By
Joel Brind, Ph.D
Dec. 6, 1992
Epidemiological research on the relationship of abortion to breast cancer clearly entered a new phase with the study of Pike et al. in 1981. Previously, a history of abortion was more frequently associated with breast cancer cases in many, but not all, studies. This trend toward increased risk was also often small and not statistically significant. In addition, abortions were almost exclusively spontaneous. The “new era” of abortion/breast cancer research has been marked by:
- The presence of induced abortion in numbers sufficient for statistical analysis,
- More precise determination of reproductive history, in particular the number of abortions and their timing relative to full-term pregnancies,
- More emphasis on younger patients (less than 40 years old), among whom incidence of breast cancer has been rising most rapidly, and in whom the disease spreads (and kills) more quickly,
- Greater understanding of the biology of breast tissue growth and differentiation, particularly in regard to susceptibility to carcinogens and changes induced by pregnancy (due largely to the work of Russo and Russo, 1980), all resulting in
- A more precisely defined, worldwide trend of increased risk of breast cancer in women who have had any abortions before having a child or who have had multiple abortions, despite
- A strong political bias (among the medical and medical research professions in general) in favor of removal of all legal restrictions on induced abortion. Thus, most of the research in this field has been directed toward disproving the work of Pike et al. (1981), and in most studies, objective scientific principles of experimental design, data analysis and reportage of results have been compromised.
The case-control study of Pike et al. (1981) was striking in its finding of a 2.4-fold increase in the incidence of breast cancer among young California white women (less than 33 years old) only if they had had an abortion (spontaneous or induced) before (or in the absence of) a full-term pregnancy. The same finding was reproduced exactly in the subsequent study of Chinese women in Shanghai (Yuan et al., 1988). Although they claimed to refute these findings, Vessey et al.’s 1982 English study had too few young patients with a history of induced abortion (“only a handful”) for a serious comparison, and Brinton et al., substantially reproduced the findings (1.3-fold and 5.5-fold for induced abortion of first pregnancy among parous and nulliparous women, respectively) in their 1983 multicenter U.S. study, even though they also claimed to refute Pike et al.’s findings! Similarly, over in Denmark, Ewertz and Duffy (1988) found 1.2-fold and 3.8-fold risk increases among parous and nulliparous women, respectively. Back in the U.S., Rosenberg et al. (1988) found a 1.3-fold risk increase in nulliparous women, while parous women under 40 had a 1.4-fold risk increase after an induced abortion at any time (a 1.7-fold if they’d had more than one), similar to the 1.5-fold risk increase found among Japanese women by Hirohata et al. (1985), and the 1.2-fold risk elevation among parous or nulliparous Italian women whose first pregnancy had ended with an induced abortion (Parazzini et al. 1991). In the former Soviet Union, where induced abortion has been the primary means of birth control since 1955, a study originally published in 1978 (in Russian) found North Caucasian women who had had one or two induced abortions to be at a 2-fold higher risk for breast cancer, and those with 3 or more abortions to be at a 3.4-fold higher risk, according to Remennick (1990).
Meanwhile, in Sweden, Lindefors-Harris et al. (1991) have attempted to dismiss this whole international body of case-control literature with the idea that in such retrospective studies, wherein data are obtained from interviews with breast cancer patients and controls, the cancer patients are more likely to remember and report abortions than the controls, and more likely to imagine induced abortions that never happened! To back up their hypothesis, they compare two studies that they performed on the same patients, one by retrospective questionnaires (in which no effect of abortion on breast cancer risk was found-Adami et al., 1990) and the other by examination of computerized registry data (Lindefors-Harris et al., 1989), in which they reported a 20% decrease in breast cancer risk among women who had had any induced abortions. Among other problems in study design and analysis, as previously detailed herein, the authors misinterpreted the protective effect of having children as a protective effect of having an abortion, and actually found a 1.9-fold risk increase among those women who had had an induced abortion in the absence of a live birth! In the same year, back in the U.S., Howe et al. (1989) reported their results of a case-control study that they had assembled using only computerized registry data on upstate (and Long Island) New York women under 40. Their results echoed what most other researchers found in questionnaire-based studies: A 1.7-fold risk increase among those who had had any induced abortions, and a 4.0-fold increase among those with 2 or more abortions between live births.
A few recent studies still reflect only spontaneous abortions, the results, though less consistent, are still troubling. Thus, Hadjimichael et al. (1986) found that Connecticut women who had aborted their first pregnancy had a 3.5-fold elevation in breast cancer risk. Moreover, their risk increased more rapidly with age than those who had not aborted. Similarly high risk increases had been observed by Soini (1975) among Finnish women who presumably had only spontaneous abortions (1.6-fold risk increase for any abortions; 5.0-fold for 3 or more). There have been notable exceptions, however: Kvale et al. (1987) found that Norwegian women who had aborted more than 3 times enjoyed a 25% reduction in breast cancer risk, just as Rosenberg et al. (1988) found a 40% risk reduction in northeast American women who had had 3 or more spontaneous abortions, and as Paffenbarger et al. (1980) had found a 19% risk reduction among postmenopausal California women who had any history of spontaneous abortion. These apparent anomalies are easily explained, however, (as Paffenbarger et al., did) in any population in which any form of contraception is widespread: Contraception (especially in the early reproductive years, when miscarriage is more common) lowers the likelihood of aborted pregnancy as well as full-term pregnancy (parity). Lower parity, and especially, delayed first childbirth, enhance breast cancer risk. Thus, a breast cancer patient population is likely to have fewer children and fewer spontaneous abortions. The consensus finding is that an aborted pregnancy has the same effect on breast cancer risk, regardless of whether the abortion is spontaneous or induced. This follows from the ever improving understanding of how breast cancer develops.
A particularly lucid presentation of what is known of breast and breast cancer development has been given by Krieger (1989). Briefly, breast tissue undergoes its first major round of growth (cell proliferation) during puberty, but it still is relatively undifferentiated (i.e., far from being capable of lactation). As such, it is highly susceptible to the effects of carcinogens of all sorts. Carcinogens are substances which can damage DNA and thereby initiate the change of normal body cells into potential tumor cells. Substances which make abnormal cells grow into full blown cancer are known as “promoters.” The most potent promoters of cancers of the breast (and other female reproductive tissues) are the female sex hormones, estrogens. Most established risk factors for breast cancer relate to some form of estrogen excess. For example, women who enter the menopause at an older age are at higher risk of breast cancer because they are exposed for more years to cyclically high levels of estrogens. The biggest estrogen surge occurs in early pregnancy, stimulating the second major round of breast tissue growth. In late pregnancy, other hormones act to make the breast tissue differentiate into milk producing tissue. These terminally differentiated cells are much less susceptible to the initiating effects of carcinogens.
Knowing these basic facts, it is easy to see how events (or non-events) in a woman’s reproductive history influence breast cancer risk: The longer the time-span between puberty and first birth, the longer the relatively undifferentiated breast cells are susceptible to the initiating effects of environmental carcinogens; hence, the well established increased risk due to delaying childbirth by any means (including abortion). But abortion does not simply delay childbirth: Rather, it allows the estrogen surge of early pregnancy to promote the rapid proliferation of breast tissue cells and abnormal cells, but cuts off the influence of the late pregnancy hormones that induce full differentiation. The net result is increased tumor promotion of whatever abnormal cells may have formed earlier. After a woman has already had a full term pregnancy (regardless of the outcome as far as the child is concerned), her more mature breasts are much less susceptible to carcinogenic influences, hence the much lower (close to zero) risk enhancing effect of an aborted second pregnancy. Multiple events do add up, however. Thus, multiple births have been found to decrease overall risk, while multiple abortions, even after a live birth, tend to increase risk.
This concordance of biological theory and epidemiological evidence is not sufficient to establish the abortion-breast cancer link as fact, but it is complemented by elegant experimental work in rats. Thus, Russo and Russo (1980) were able to demonstrate that early childbirth prevents the development of cancer in rats exposed to a chemical carcinogen, while aborting the early pregnancy sentenced most of the rats to breast cancer.
The practical import of the increased risk of getting breast cancer after aborting a first pregnancy is best illustrated by a hypothetical example. Estimating most conservatively, an average 15-year-old American girls has at least a 10% lifetime risk (assuming she will go on to have at least one child, when in her 20’s). If she shows up at a crisis pregnancy center pregnant at 15, an honest, informed counselor would tell her that having the baby will reduce her lifetime risk of getting breast cancer to about 7% (this has been well known in the professional literature since 1970). If she has a first trimester abortion, her breast cancer risk will go up to about 15%, assuming she can and does have at least one child in her 20’s. If the abortion renders her infertile, however (which happens in a significant percentage of abortions), her lifetime risk will double again, to about 30%! Abortion providers (and the American Medical Association, for that matter) counsel that abortion is much safer than childbirth, which is even more risky for teenagers. However, the maternal death rate in the U.S. is only about 8 per 100,000 births (0.008%)-slight by any measure, while the increase in breast cancer risk-although relatively small-is staggering by any measure. In terms of public health in the US, where 800,000 girls and women abort (by induced abortion, that is) their first pregnancy every year, the low end estimate of a 1.5-fold risk increase from the abortion-independent of the risk elevation due to the postponement of pregnancy-would translate to 40,000 additional breast cancer cases every year. Legal abortion probably accounts for a significant fraction of new cases already, in what the last session of the U.S. Congress has already called a “growing epidemic.”
Even more ominous are the recent findings of Olsson et al. (1991), who studied actual breast cancer tissue from premenopausal patients with and without a history of oral contraceptive use or induced abortion. Patients with either history had tumors which were more abnormal and faster growing (and presumably more lethal) than those of patients with no such history (and premenopausal breast cancers in general are already more invasive and lethal than postmenopausal ones).
Perhaps the most disturbing aspect of this whole subject is that many (if not most) of the researchers in the field (whose professional associations are, for the most part, unabashedly “pro choice”) have slanted their study designs (e.g., through improper age matching), and/or data analysis (e.g., through separating out nulliparous cases and finding too few cases for statistical significance), and/or interpretation (e.g., through characterization of even twofold risk increases as “slight,” or claiming no risk increase when the number of cases they studied was insufficient for statistical significance). The group at the Karolinska Institute in Sweden has even been so audacious (as discussed above) as to manipulate their data so as to show a “protective effect” of abortion and to concoct a shabby study in an attempt to discredit the whole rest of the field!
The real danger here, is that since breast cancer is a high incidence disease, the relative risk values are small (though the impact is huge), which makes them easier to hide. So far, those who would hide the facts have succeeded: The U.S. is on the verge of eliminating even the right of states to restrict abortion, and evidence of any public awareness (even on the part of the major medical associations-e.g., Harris, et al.’s New England Journal of Medicine review article, 1992, which pretends that the problem does not exist) is totally absent. It is the purpose of the present report to bring the knowledge of the breast cancer risk increase inherent in abortion into the public eye, so that public policy can be debated and amended and formulated accordingly.
ProLifeForum.org
A Ministry of Proclamation Presbyterian Church
278 Bryn Mawr Avenue
Bryn Mawr, PA 19010
Voice: 610-520-9500
Fax: 610-520-5240
|
|
