Abortion and breast cancer: The manufactroversy that won’t die

Editor’s note: Given the controversial nature of the topic, I think it’s a good time to point out my disclaimer before this post. Not that it’ll prevent any heated arguments or anything…

The Science-Based Medicine blog was started slightly over two years ago, and this is a post I’ve wanted to do since the very beginning. However, since January 2008, each and every time I approached this topic I chickened out. After all, the topic of abortion is such a hot button issue that I seriously questioned whether the grief it would be likely to cause is worth it. (Take the heat generated any time circumcision is discussed here and ramp it up by a factor of 10.) On the other hand, there is so much misinformation out there claiming a link between abortion and the subsequent development of breast cancer when the data simply don’t support such a link, and the name of this blog is Science-Based Medicine. Why should I continue to shy away from a topic just because it’s so religiously charged? More importantly, in my discussion how can I focus attention on the science rather than letting the discussion degenerate into the typical flamefest that any discussion of abortion on the Internet (or anywhere else, for that matter) will almost inevitably degenerate into. Indeed, such discussions have a depressing near-inevitability of validating Godwin’s law not once but many times — usually within mere hours, if not minutes.

My strategy to try to keep the discussion focused on the science will be to stay silent about my own personal opinions regarding abortion and, other than using it to introduce my trepidation about discussing the topic, the religious and moral arguments that fuel the controversy. That’s because the question of whether abortion is the murder of a human being, merely the removal of a lump of tissue, or somewhere in between is a moral issue that, at least as far as I’m concerned, can’t ever be definitively answered by science. That is why it is not my purpose to sway readers towards any specific opinion regarding the morality of abortion. Indeed, I highly doubt that any of our readers care much about my opinions on the matter. On the other hand, I would hope that I’ve built up enough trust over the last two years that our readers will be interested in my analysis of the existing data regarding something another related issue. It is my purpose to try to dispel a myth that is not supported by science, specifically the claim that elective abortion is causes breast cancer or is a very strong risk factor for its subsequent development. That is a claim that can be answered by science and, for the most part, has been answered by science with a fairly high degree of certainty. Despite the science against it, the medical myth that abortion causes breast cancer or vastly increases the risk of it is, like the myth that vaccines cause autism, a manufactroversy that won’t die, mainly because it is largely fueled by religious beliefs that are every bit as immune to science as the ideological beliefs that drive the antivaccine movement.

A bit of background

Let me step back a bit. Several years ago, I didn’t pay much, if any, attention to the ABC (“abortion-breast cancer”) claim. Then two years ago events conspired to force me to start paying attention. The first item was nothing more than a skeptic encountering bad science. The second thing that happened struck a lot closer to home. The first thing that happened was that in the fall of 2007 I saw this article in the Chicago Tribune entitled Snubbing cancer study will only hurt women: Research showing link to abortion ignored by media (also mirrored on John Byrne’s blog):

During National Breast Cancer Awareness Month, it is fitting and proper that women be informed about any newly discovered dangers, even as the public groans under the weight of all the warnings surrounding the mere act of living.

For example, a well-researched Chicago Tribune story last week disclosed that women who have just a couple of alcoholic drinks daily increase their breast cancer risk by 13 percent. Coincidentally, a new study reported that abortion is an important breast cancer risk factor, yet I couldn’t find a word describing the research in mainstream media.

How to explain this disparity? I’ll be vigorously advised that “most” studies disprove an abortion-breast cancer link. Or that the study in question appeared in a “conservative” scientific journal. Or that the study is bogus or unimportant. Or, more rudely, that the whole breast cancer argument has been concocted by anti-abortion rights advocates to make women afraid to have abortions. The issue is dead, I’ll be notified. Kaput. Here I would remind critics that in science it’s not who says it or how many say it that counts. What does count are the data and the rigor with which they are collected, analyzed and held up to a scientifically credible hypothesis.

Curious to find out what this study was, what it found, and whether or not it was evidence that I should change my mind, I decided to go straight to the source. That tends to be my reaction whenever I see such a veritable panoply of crank language, not unlike what we hear from “intelligent design” creationists when whining about why they are not taken seriously. The science that has failed to validate ABC may not be as well settled as the science supporting the theory of evolution, but the arguments against established science used by ABC advocates are disturbingly similar. As for the “disparity,” perhaps it could be because even the mainstream media has learned that the medical “journal” in which this dubious “study” appeared (and I use the terms “journal” and “study” very loosely) is a right wing propaganda rag masquerading as a medical journal. It’s also a “journal” with which regular readers of this blog should be very familiar. Yes, my friends, we’re talking about the Journal of American Physicians and Surgeons (JPANDS for short), and the study was entitled The Breast Cancer Epidemic: Modeling and Forecasts Based on Abortion and Other Risk Factors, by Patrick Carroll.

It’s hard not to note that one significant indication that the study is likely to be really, really, really bad is the very fact that it appeared in JPANDS. As you may recall, I’ve discussed JPANDS before, as has Kathleen Seidel. Of course, the fact that the study appeared in JPANDS does not necessarily mean it’s a bad study, although it does make it highly likely that it is, particularly given that a good study can get into a real scientific journal. In this case, it was a very bad study indeed. In fact, it was one of the most hilariously inept examples of confusing correlation with causation and abusing statistics that I had ever seen anywhere, any time. Truly, the Flying Spaghetti Monster’s example of correlating global warming with the decrease in the number of pirates seemed reasonable by comparison. Unfortunately, the Carroll “study” (if it can be called that) achieved its purpose. It got credulous pundits like Byrne presenting it as if it were actual evidence of an ABC link.

What surprised me, however, was not the incredibly inept attempt at “science” in this article. What actually surprised me was the second thing that made me stand up and take notice of the ABC activists, something that I came across while investigating whether or not there is a link between abortion and breast cancer. What I found was that someone I actually knew, someone who was actually connected to The Cancer Institute of New Jersey (where I was on the faculty from 1999 to 2008) due to her affiliation with one of the private hospitals affiliated with CINJ. I’m referring to Dr. Angela Lanfranchi, who is an anti-abortion activist, cofounder of the Breast Cancer Prevention Institute, which promotes the ABC link, and one of the foremost promoters of the link between abortion and breast cancer. Indeed, as far as I can tell, Dr. Lanfranchi is probably the foremost promoter of the link who actually takes care of breast cancer patients for a living. Worse, she has published arguments in favor of ABC in JPANDS, which, as regular readers of this blog know, is always a bad sign as far as pseudoscience or ideologically motivated anti-science goes. Sadly, after discovering her “other side,” in marked contrast to her generally strong competence as a surgeon, I soon found articles by Dr. Lanfranchi using classic crank language to make incompetent, ridiculously exaggerated, and scientifically unsupported statements:

It amounts to child abuse to take a teenager in a crisis pregnancy for an abortion. At best, it will give her a 30% risk of breast cancer in her lifetime. At worst, if she also has a family history of breast cancer, it will nearly guarantee this. As a mother, I need to be informed of this to protect my daughter. Medical professionals have an unfortunate history of continuing to harm women if it means admitting that they have injured or killed them with their treatments.

This is best illustrated through the well-known story of Ignaz Semmelweis, MD. He was an obstetrician-gynecologist in the 1840s who proved that hand washing would reduce mortality rates from childbed fever from 30% to 2% on maternity wards. His reward for this was ridicule from his professors and loss of his hospital appointments. Women continued to die needlessly for another 30 years until the germ theory proved Semmelweiss [sic] was correct. It must have been very embarrassing for a lowly resident to have told the greatest medical professors of his time they were responsible for many women’s deaths.

I’ve occasionally had the extreme chutzpah to propose “laws” and then to name them after myself, such as when I proposed Gorski’s Law of Testimonials at the SBM Conference during my talk at TAM7 or facetiously coined Gorski’s Law of the Pharma Shill Gambit right here on SBM. I’m seriously tempted to do so again here regarding the invocation of Semmelweis’s name. Whenever someone invokes the name of Ignaz Semmelweis in the context of overblown, hyperbolic language, that person is almost certain not to be basing her criticism of medicine on science. It applies to the anti-vaccine movement. It applies to Mike Adams of NaturalNews.com, and it applies here. In any case, the above is an incredible exaggeration that even the research of one of the foremost scientists making the ABC claim doesn’t support. (More on that later.)

Although I still miss CINJ from time to time, it was a really good thing, at least for my comfort level dealing with the local practitioners, that it was within a month after that I decided leave my old job to accept my current position. Less than four months later I was gone from CINJ. I can’t imagine the tension that would have occurred when I ran into Dr. Lanfranchi at her hospital’s tumor board, which I was occasionally assigned to attend because CINJ sent its faculty out as guests to the various tumor boards of its affiliates. In retrospect, I don’t know how I could have been right there in central New Jersey for so long and yet so oblivious to what was going on, but I was.

Back to the future

So what tweaked me to write about the question of whether abortion predisposes women to breast cancer after all this time? Last week, several readers forwarded me this article from WorldNet Daily by Janet Stanek, a prominent activist promoting claims of a link between abortion and breast cancer, in which she claims that a “Top scientist finally admits abortion-breast cancer link,” a sentiment echoed on her blog. Included in this article is an amazing claim about the NCI workshop in 2003 that concluded that there is no link between abortion and breast cancer:

At the time, 29 out of 38 studies conducted worldwide over 40 years showed an increased ABC risk, but the NCI workshop nevertheless concluded it was “well established” that “induced abortion is not associated with an increase in breast cancer risk.”

Brind went on to write a minority report NCI alludes to on its website without publishing or listing its author and did not even mention in its workshop summary report.

Life went on, except for post-abortive women inflicted with breast cancer anyway.

But six years later something happened. Dr. Brinton either flipped her lid, flipped ideologies, restudied the evidence and decided to recant, or couldn’t sleep at night – and she began righting her wrong.

In April 2009, Brinton co-authored a research paper published in the prestigious journal Cancer Epidemiology, Biomarkers and Prevention, which concluded that the risk of a particularly deadly form of breast cancer that attacks women under 40 raises 40 percent if a woman has had an abortion.

Stanek then gloats:

For nine months, that little bombshell of a disaster for pro-abortion ideology was published without the NCI acknowledging it or changing its stance.

Then this month, Brind spotted and wrote about Brinton’s concession and NCI’s hypocrisy.

In case you don’t know who he is, Joel Brind is a born-again Christian and professor of biology and endocrinology at Baruch College who has become one of the most prominent voices in the ABC movement. Arguably the most famous supporter of the idea that abortion causes breast cancer, he has campaigned tirelessly to promote his view that abortion predisposes to breast cancer. I also love how Stanek just counts studies without any consideration of quality or scientific rigor, as if all studies are equivalent. Truly, she’d fit right in with creationists or the alt-med movement. Be that as it may, one has to wonder about her claim that Dr. Brinton “flipped her lid.” To do that, I had to look up the actual study that Stanek is holding up as vindication, although I find it odd that it took the ABC supporters 9 months to discover this study, entitled Risk Factors for Triple-Negative Breast Cancer in Women Under the Age of 45 Years. Do they not keep up with the medical literature? It’s pretty easy to set up an automatic PubMed search for “abortion AND breast cancer risk”) that will send automatic updates every day.

In any case, I decided to see whether this study is such slam dunk evidence of a nefarious plot to hide the ABC link.

Back to the future, take two

Before I continue, it’s worth it to go over a bit of background. First of all, you need to know what “triple-negative” breast cancer is. Clinically, breast cancer is divided by whether or not it makes three different proteins: estrogen receptor (ER), progesterone receptor (PR), and HER2/neu. ER(+) tumors tend to be better differentiated and less aggressive, but, more importantly, they respond to therapies designed to block estrogen action. HER2/neu is an oncogene that a subset of breast cancers express and that tends to portend a worse prognosis. However, when it is present, the tumor can be treated with a targeted therapy designed to counteract its activity, specifically Herceptin. Triple-negative cancers, by contrast, make no ER, PR, or HER2/neu (hence the “triple-negative” moniker). More importantly, they tend to be more aggressive and more poorly differentiated. Paradoxically, they tend to respond well to chemotherapy but have a proclivity to relapse afterward. They’re nasty, nasty tumors and make up around 15% of all breast cancer. (Indeed, one of them killed my mother-in-law nearly one year ago.) More recently, breast cancers have been characterized on the basis of global gene expression into subgroups known as luminal A and B and basal-like breast cancer. Consistent with previous work, we know that the vast majority of triple-negative tumors fall into the basal-like classification, and basal tumors tend to be aggressive histologically, unresponsiveness to typical endocrine therapies, and poor prognosis.

Another critical piece of background that’s important is the known risk factors for breast cancer. Currently agreed-upon risk factors based on history for the development of breast cancer are:

1. Age (breast cancer becomes more common as women age)
2. Sex (men can get breast cancer, but it is very uncommon)
3. Personal history of breast cancer (a history of breast cancer in one breast is probably the strongest risk factor of all for the subsequent development of cancer in the contralateral breast)
4. Family history of breast cancer and genetic risk factors (BRCA mutations, for example)
5. Age at first menstrual period (earlier is worse)
6. Race (breast cancer is slightly more common in Caucasians but African Americans are more likely to die of the disease)
7. Age at menopause (older is worse)
8. Age at first live birth (older is worse)
9. Number of children (more is better, with nulliparity conferring the highest risk)
10. Previous chest irradiation (for childhood lymphoma in the chest, for example)
11. Oral contraceptive use (this is a very weak and somewhat controversial risk factor; women who have not taken OCPs for 10 years are at no higher risk than those who have not taken them)
12. Hormone replacement therapy
13. Certain pathological findings on breast biopsy (such as atypical ductal hyperplasia, lobular carcinoma in situ, etc.)
14. Breastfeeding (a weak protective effect)
15. Alcohol
16. Obesity

Also, before I get to the study, for purposes of my discussion of the alleged link between abortion and breast cancer, one theme that runs through the risk factors due to reproductive history is that the more time a woman spends having menstrual cycles uninterrupted by pregnancies, the higher the risk of breast cancer. That’s why more pregnancies are protective and why an earlier age at menarche and a later age of menopause are thought to be risk factors. Leaving aside family history, a personal history of breast cancer, or a biopsy with a high risk lesion (the three strongest risk factors of all), the woman at highest risk of developing breast cancer is one who has never been pregnant, had an early menarche, and has a late menopause. This relationship between reproductive history and breast cancer is thought to be due to a higher lifetime exposure to estrogen unopposed by progesterone and other pregnancy hormones. One consequence of these observations observed thus far is that the tumors that develop in women with these risk factors are more commonly ER(+), which makes sense from a biological standpoint. Indeed, strategies designed to decrease breast cancer risk in women at high risk by blocking estrogen action do decrease the risk of ER(+) breast cancer but have no effect on the risk of ER(-) breast cancer, which makes sense from a biological standpoint.

Triple-negative breast cancer, being ER(-), has become a hot research topic lately because it tends to strike younger women, especially African-Americans, and it tends to be deadlier than other forms. Worse, because it does not make ER or HER2/neu, there are no targeted treatments for it. Tamoxifen and aromatase inhibitors (which block the action of estrogen) don’t work; neither does Herceptin (which blocks the action of HER2/neu). All that leaves is chemotherapy. Paradoxically, triple-negative tumors tend to respond well initially to chemotherapy, but they have a high propensity to recur after an apparently good initial response. No doubt much of the impetus for this study were these well-known facts, coupled with how little is known about the risk factors for triple-negative breast cancer as opposed to other more common subtypes

Another thing that you need to know about the ABC claim is that the evidence is quite conclusive that spontaneous early miscarriages neither protect against nor decrease the risk of breast cancer. Other than possibly for women who have suffered more than three spontaneous miscarriages (the data are equivocal), pregnancy loss appears to be more or less neutral with respect to influencing breast cancer risk, neither increasing nor decreasing it. These observations are fairly strong suggestive evidence that an elective abortion would probably not behave any differently than a spontaneous miscarriage at the same point in pregnancy from a biological standpoint. These data are not enough to dismiss the ABC link in and of themselves, but they do lessen the biological plausibility of such a link. Not enough to reject further study, but enough to cast a skeptical eye on the retrospective studies that exist.

The final thing you need to know is about this study itself, specifically that it is not really a new study at all. Rather, it is a reanalysis of existing data from two studies of women from the 1980s and 1990s, as the methods section shows:

The cases included in this study were originally ascertained for two previous studies through the population-based Seattle–Puget Sound Surveillance, Epidemiology, and End Results cancer registry. Eligible cases from the first study population included all primary invasive breast cancers within the three-county Seattle metropolitan area, diagnosed between January 1, 1983, and April 30, 1990 (ages, 21-45 y). The methods for this study have been described elsewhere (17, 18). The study was confined to Caucasians because of the small representation of minorities in the region.

And:

The second population included the Seattle site participants of the multicenter Women’s Interview Study of Health, the methods for which have been described (19). Eligible cases included women in the Seattle area diagnosed with invasive breast cancer between May 1, 1990, and December 31, 1992 (ages 21-44 y).

References 17 and 18 are:

White E, Malone KE, Weiss NS, Daling JR. Breast cancer among young U.S. women in relation to oral contraceptive use. J Natl Cancer Inst. 1994 Apr 6;86(7):505-14.

Daling JR, Malone KE, Voigt LF, White E, Weiss NS. Risk of breast cancer among young women: relationship to induced abortion.J Natl Cancer Inst. 1994 Nov 2;86(21):1584-92.

And reference 19 is:

Brinton LA, Daling JR, Liff JM, Schoenberg JB, Malone KE, Stanford JL, Coates RJ, Gammon MD, Hanson L, Hoover RN. Oral contraceptives and breast cancer risk among younger women. J Natl Cancer Inst. 1995 Jun 7;87(11):827-35.

For two out of the three studies, the purpose was to ascertain whether oral contraceptives pose a risk for breast cancer. In the current study, pooling the two studies ended up with subjects that included 897 women who had developed breast cancer before age 45 and 1,569 controls. These studies were all retrospective, with all the potentials for confounding factors to which retrospective studies are prone. Moreover, it was a study based on interviews, in which women were interviewed about their health history and known and suspected causes of breast cancer. Recall bias is a well-known confounding factor that plagues studies of abortion. One reason is that, because of the social stigma associated with abortion, women tend not to tell everything about their history when it comes to abortions, either not admitting to the procedure or, if they’ve had more than one, not admitting to all of them. This may have been particularly true for older studies, when abortion had even more of a stigma attached to it. The other reason is that women with breast cancer who have had an abortion in the past tend to be more likely to admit to having had an abortion they think it might be a cause of their predicament. It’s very hard to evaluate the significance of recall or response bias and how much they might have affected the results of individual studies. In any case, such problems are why prospective studies are less likely to produce spurious associations. Also, many of the cases of breast cancer in these studies occurred before one of the commonly used markers that we test all new breast cancers for, HER2/neu, was routinely tested for, meaning that the investigators had to locate the pathology specimens and have them tested for HER2/neu, because separating triple-negative tumors from the rest depends upon it.

Now, on to the study. Here’s what Malek and Brind are crowing about. It’s part of the “money table” that summarizes the results:

Note the yellow at the bottom pointing out the line referring to abortion, thanks to Colby Cosh. According to the table, the odds ratio (OR) for breast cancer in women who have had one or more abortions is 1.4 (95% confidence interval 1.1 to 1.8), a barely statistically significant result. One thing that leapt right out at me and bothered me was not so much this result. For one thing, the original study (18) of the group that looked at this question found an OR = 1.5 (95% CI = 1.2 to 1.9). I would have been shocked if the odds ratio was significantly different in this study. In any case, it’s barely statistically significant and comes from a pooled retrospective study where the most recent cases date back to the early 1990s, both factors that make it very prone to bias or spurious results. More recent research done prospectively should have (and does have) greater weight.

Stranger still is an odd consistency in the data on abortion and breast cancer in this study when they broke down the numbers into triple-negative versus non-triple-negative breast cancers. For all potential breakdowns (all cancers, triple-negative cancers, non-triple-negative cancers), the OR is the same, 1.4. That struck me as quite odd. Under either of the common biological justifications posited by advocates of the ABC concept, there should be a difference. If it’s estrogen overexposure unopposed by progesterone, then we would predict that abortions would result in more ER(+) tumors than ER(-) (including triple-negative) tumors if abortion is truly a risk factor for breast cancer. If it’s the newer straws that the ABC movement is grasping at, namely stem cells, one might hypothesize that abortion would increase the risk of ER(-) or triple-negative tumors, since stem cells don’t make ER. True, there weren’t that many triple-negative tumors in this study (187); so who knows? But it’s still rather odd that the numbers are basically identical. It tends to suggest more of a statistical fluke than a specific biological mechanism.

What also makes this OR of 1.4 less convincing is that, if you look at the 1,569 controls, they broke down pretty similarly between having had one or more abortions and never having had one; i.e., 27.1% of controls had an abortion and between 32% and 35.5% of women with breast cancer had one or more. Nowhere is it analyzed whether that difference is statistically significant, but even if it is it’s not a huge difference. Moreover, this study only addressed breast cancer risk in women under 45. This is a relatively small fraction of the total cases of breast cancer; so even if this study did show a “40%” increased rate of breast cancer, it would only apply to a relatively small portion of the population. Given the problems with the study and in light of data gathered over the 15 years since the last of the three studies whose subjects make up this reanalysis was completed, I am completely underwhelmed with this study as any sort of strong evidence for an ABC link.

Finally, an OR = 1.4 is not a large odds ratio, particularly in a retrospective study. Colby Cash is not correct when he dismisses any OR under 2.0 as being insignificant, but small odds ratios do require multiple replications from different groups in different populations all converging around similar numbers before epidemiologists take them seriously. That’s what it took before secondhand smoke exposure, which routinely results in odds ratios around 1.3 for heart disease and lung cancer, was accepted as significant risk for health problems. Other aspects that might make ABC more credible would be if there were a “dose-response” effect, in which more abortions would increase the risk even more, or some apparently biological specificity for certain subtypes of cancer. Neither of these exist, and this study sure doesn’t provide such evidence. In light of that, more than likely this is simply a spurious result that has been refuted by later studies. Certainly it doesn’t even approach slam-dunk evidence that an ABC link exists, much less is being “covered up.”

The evidence

Here’s what the American Cancer Society says about the ABC link:

The Collaborative Group on Hormonal Factors in Breast Cancer, based out of Oxford University in England, recently put together the results from 53 separate studies conducted in 16 different countries. These studies included about 83,000 women with breast cancer. After combining and reviewing the results from these studies, the researchers concluded that “the totality of worldwide epidemiological evidence indicates that pregnancies ending as either spontaneous or induced abortions do not have adverse effects on women’s subsequent risk of developing breast cancer.”

And here’s what the National Cancer Institute says about it:

The relationship between induced and spontaneous abortion and breast cancer risk has been the subject of extensive research beginning in the late 1950s. Until the mid-1990s, the evidence was inconsistent. Findings from some studies suggested there was no increase in risk of breast cancer among women who had an abortion, while findings from other studies suggested there was an increased risk. Most of these studies, however, were flawed in a number of ways that can lead to unreliable results. Only a small number of women were included in many of these studies, and for most, the data were collected only after breast cancer had been diagnosed, and women’s histories of miscarriage and abortion were based on their “self-report” rather than on their medical records. Since then, better-designed studies have been conducted. These newer studies examined large numbers of women, collected data before breast cancer was found, and gathered medical history information from medical records rather than simply from self-reports, thereby generating more reliable findings. The newer studies consistently showed no association between induced and spontaneous abortions and breast cancer risk.

And here’s what the American College of Obstetrics and Gynecology said about it in 2009:

The relationship between induced abortion and the subsequent development of breast cancer has been the subject of a substantial amount of epidemiologic study. Early case-control studies that reported an association between induced abortion and subsequent development of breast cancer had significant methodological problems, most notably reliance on retrospective reporting of abortion history. A key methodological consideration in interpreting the evidence for any relationship between abortion and breast cancer risk is the sensitive nature of abortion, which could affect the accuracy in retrospective studies that rely on participant reports of having had an abortion.

[…]

Early studies of the relationship between prior induced abortion and breast cancer risk were methodologically flawed. More rigorous recent studies demonstrate no causal relationship between induced abortion and a subsequent increase in breast cancer risk.

A few of the most recent studies that have failed to find an ABC link include the California Teacher’s study, a large, prospective cohort study with detailed pregnancy history data; the Nurse’s Health Study II, which included over 100,000 women; and the EPIC Study, which included over 250,000 women. Numerous studies (including some of the above) that failed to support an ABC link are linked to here^{1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17}.

Given the preponderance of evidence, although it is still possible that there may be a link between abortion and breast cancer, it is unlikely that there is, and, if there is, it’s likely to be very, very small, given that numerous epidemiological studies have failed to uncover it. In this, the evidence for the ABC link is not unlike the state of evidence regarding vaccines and autism. Early studies, not as large, well-designed, or rigorous, suggested that there might be an association, but the larger and better-designed the study, the smaller the OR became until it converged on 1.0. Current evidence does not support the ABC link, and there are enough studies to allow us to conclude either that there probably is none or that it’s very small. That’s as good as it gets in epidemiological studies, which, unfortunately, can never entirely eliminate the possibility of a correlation. They can only conclude that the chance of a significant correlation is very, very low. Moreover, contrary to the inflated claims of some activists, even Joel Brind’s own infamous meta-analysis from 2003 does not show a 30% risk of breast cancer if a young woman has an abortion before the age of 18, much less a virtual certainty that she’ll develop breast cancer if she has a strong family history as well. In fact, Brind’s own work, which is held up as “proof” of an ABC link, only suggests at the most an OR = 1.3 to 1.5, which is nowhere near high enough to produce the 30% lifetime risk of breast cancers claimed by overwrought activists like Dr. Lanfranchi.

Despite the evidence, however, if you do a Google search for “abortion and breast cancer” what you will find is a preponderance of websites pushing a link that is not scientifically supported. The purpose of trying to push a link between breast cancer and abortion, as far as ideologues go, is clearly to frighten women about abortion. However, ABC is also being pushed to make physicians who perform abortion fearful of malpractice lawsuits by women who have abortions and later develop breast cancer, which, given how common breast cancer and abortion are, there are many by chance alone. The grounds for suing being advocated by ABC promoters is an alleged failure to inform women of the increased risk of breast cancer due to an abortion. In other words, if scaring women won’t work, then maybe threatening doctors with malpractice suits will.

Whatever you think of abortion, whether it’s murder, a necessary evil that’s not murder, a morally neutral surgical procedure, or a moral good (the last of which to me is going too far), I would hope that we could all manage to agree that attributing risks to the procedure that are not supported by strong science and epidemiology does not contribute anything to the debate other than confusion and fear. Of course, that’s exactly what such claims are meant to place in the minds of women of reproductive age: confusion and fear. They serve primarily as a means of frightening women with the specter of breast cancer if they consider the option of terminating a pregnancy. Indeed, dubious studies such as the Carroll study and putting dubious spins on studies that are not themselves dubious (like the study Stanek made such hay over) are a transparent attempt to abuse epidemiology to find a link between abortion and breast cancer that probably does not exist. They only serve to obfuscate the issue.

It’s one thing for anti-abortion activists to consider abortion to be evil and to lobby and demonstrate to have it outlawed. That is their religious belief, and the First Amendment guarantees that they can believe what they want and say nearly anything they want. However, it’s quite another to spread misinformation about an ABC link based on either horrifically bad science or on putting a deceptive spin on existing studies. If anti-abortion activists think that abortion is morally wrong because of their religion, then they should argue that. If they have any actual evidence that the procedure causes serious harm, they have every right to present that evidence. However, when they spin and distort science, behaving just like every other denialist, be it a creationist, a 9/11 Truther, an alt-med believer, or whatever, they should expect to be called on it.