|
NRL News
Abortion
and Breast Cancer: New Study Marks 50 Years of Denial In April 1957, the Japanese Journal of Cancer Research published a nationwide study on cancer which found that there was a three-fold higher incidence of breast cancer among women who had induced abortions compared to those women who had not. Showing that those who do not remember the past are destined to repeat it, on the last day of the 50th anniversary month, a study by a Harvard research team headed by Karin Michels was published in the American Medical Association’s Annals of Internal Medicine. The National Cancer Institute-funded team studied more than 105,000 women over a 10-year period and found that induced abortion was not “associated with the incidence of breast cancer.” It joined a decade-long parade of flawed studies purporting to show there is no link between abortion and a heightened risk of breast cancer (ABC link). Such conclusions must ignore the results of many studies, including our 1996 collaborative meta-analysis, published in the British Medical Association’s Journal of Epidemiology and Community Health. That study reported a 30% overall average increased risk of breast cancer among women who’d had any induced abortions. (A meta-analysis is a compilation of all existing studies in the medical literature.) But then as now, these important results were challenged on the basis of a theoretical something called reporting bias (also known as response bias or recall bias). We’ve written about it many times in NRL News. Suffice it to say it is an unfounded speculation that breast cancer patients are more likely than are healthy women to honestly report any abortions that they might have—thus creating the illusion of a link between induced abortion and breast cancer where none exists. This “reporting bias” notion has been thoroughly debunked. Ironically, one of the studies that provided strong evidence against the reporting bias hypothesis—a 1995 study in Athens, Greece which reported a 51% increase in risk with abortion—was coauthored by Karin Michels herself! (Tellingly, Michels does not even mention or cite the Athens study in her latest paper!) Opponents of the ABC link turned to studies using “prospective data”—i.e., records such as medical records of abortion, made before any breast cancer diagnosis—which eliminate the possibility of reporting bias. The dozen studies over the last decade, they insist, show no increased breast cancer risk among women who’d had any abortions. But as I reported in my 2005 review of the prospective data-based studies published since our 1996 meta-analysis, these “studies, widely touted as refuting the abortion-breast cancer link, are found to embody many serious methodologic flaws sufficient to invalidate their findings.” In plain language, just because the data source is prospective, it doesn’t mean the study is valid. The latest study Michels et al. is no different. The irony is that had they processed and reported their own data properly, their study actually would have confirmed the ABC link. The Basic Biology The effect of terminating a normal pregnancy with abortion leaves a woman’s breasts with more undifferentiated Type 1 and Type 2 lobules, where almost all breast cancers start. But for cancers to develop, it takes at least 8 to 10 years. In the Michels study, participants completed questionnaires about their medical and reproductive histories in 1993 and every two years thereafter. This “Harvard Nurses Study II” study used data collected up through 2003. So far, so good, except for two critical facts: 1) Subjects were relatively young, aged 39–56 at the end of the study in 2003. Hence, only one-third of the population had reached menopause by the end of the study. This would not present a problem, except that 2) abortions occurring during the entire study—even including 2003—were included in the study. Thus, a woman in the study who had an abortion the day before she sent in the final questionnaire would be counted as having an abortion, but not having breast cancer. This has the effect of diluting the effect of abortion by including a significant number of women whose abortions were too recent to have resulted in subsequent breast cancers. (Researcher Karen Malec put it this way in an interview: “If you start smoking cigarettes today, are you going to develop lung cancer next month? Will you get it in six months? It’s going to take years before you develop lung cancer, if you’re going to get it at all.”) Importantly, this time-lapse flaw could have been eliminated from the Harvard study quite easily: simply exclude women who had abortions after 1993, and only include breast cancer cases reported in the most recent questionnaires. Instead, the real relative risk was unquestionably underestimated, although we cannot tell by how much. Including the women with recent abortions amounts to the proverbial thumb on the scale, in this case, to depress the relative risk result. A second major flaw in the Michels study can be quantified with precision but is too complicated to be readily summarized here. But coupled with the inadequate follow-up time for a significant portion of the study population, it is clear that the real result of the Michels study—properly calculated and presented—is in the range of the average 30% overall increased risk attributable to induced abortion. Finally, it is important to recall what was mentioned above: abortion leaves a woman’s breasts with more undifferentiated Type 1 and Type 2 lobules, where almost all breast cancers start. (Later in pregnancy these cells become specialized milk-producing cells.) There simply is no argument that a full-term pregnancy lowers the risk of future breast cancer, a protective effect eliminated when a woman has an induced abortion. Therefore, any reasonable standard of informed consent would include, at a minimum, a warning to women—especially teenagers and younger women—considering abortion, that having the abortion will leave them at higher long-term risk of breast cancer, compared to not having the abortion. But then, they couldn’t call abortion “safe,” could they? Joel Brind, Ph.D., is a professor of biology and endocrinology at Baruch College of the City University of New York, and editor and publisher of the Abortion-Breast Cancer Quarterly Update. |