Content deleted Content added

Inline

Revision as of 18:26, 7 December 2004

The Abortion-Breast Cancer (ABC) debate centers around the fact that during early pregnancy hormone levels increase significantly. [1] This initiates cellular differentiation (growth) in the breast preparing for lactation. The ABC theory is if the pregnancy is aborted prior to full differentiation in the third trimester this could leave more "vulnerable" cells than prior to the pregnancy; resulting in an elevated risk of breast cancer. The Planned Parenthood website notes that "attempts to prove this theory, however, have failed." [2]

At the same time it has yet to be invalidated. There is anecdotal evidence to suggest a connection with steroidal estrogen on the federal carcinogen list [3] and both the American Cancer Society (ACS) [4] and the National Cancer Institute (NCI) [5] specifying reproductive hormones can elevate breast cancer risk. It is a polarized debate that has been ongoing for decades with both sides maintaining valid scientific evidence supports their position.

Dr. Joel Brind

Dr. Brind is the leading advocate for an ABC link. This sets him apart from the scientific mainstream which is highly skeptical of an ABC link. Dr. Brind became an anti-choice (pro-life from his perspective) born again Christian. He is arguably the most passionate and controversial intellectuals on the ABC issue. [6] However he is a professor of biology and endocrinology at Baruch College. Dr. Brind is also a leading critic of ABC studies and was an invitee to the National Cancer Institute's conference on the ABC issue [7] where he filed the minority dissenting comment. [8]

Additionally Dr. Brind participated in one of the two large meta-analyzes on the ABC issue [9], and was the subject of a Discover article [10] entitled The Scientist Who Hated Abortion. However controversial Dr. Brind is — his position and criticisms provide the counter point to the ABC debate. Furthermore how Dr. Brind is mischaracterized by pro-choice advocates [11] and publications [12] illustrates the lack of scientific rigor on both sides of the public debate.

Spontaneous Abortion

Pro-choice advocates have routinely stated that spontaneous abortions (miscarriage) consistently show no ABC correlation. [13] Hence this is compelling evidence of no ABC link because miscarriages usually are not caused by low hormone levels. While this is true it is also disingenuous since miscarriages are characterized by low hormone levels. The difference being that low hormones predominantly do not initiate the abortion; they are lower as a result of a miscarriage in progress. This is common knowledge for reproductive specialists. [14]

One of the first studies on hormone levels and spontaneous abortion by Kunz & Keller (1976) [15] shows that 89% percent of the time progesterone is abnormally low a miscarriage occurs. It is also reflected in studies published by Hertz et al (1979) [16] and in more detail with Stewart et al (1993) [17] conducted at a fertility clinic. This focuses the ABC debate on induced abortion.

Scientific Studies

There are several different kinds of scientific studies involved in the ABC debate. Each with strengths and weaknesses that are based on rats, interviews (case-control), meta-analysis and cohorts. These are listed roughly in order of scientific significance and cover the recent history starting in 1980 when the first rat study was published. Since the debate has focused on recent studies that have been scrutinized, are generally larger, and utilize modern methodologies that will be the focus of this article; even though there are ABC studies conducted as far back as 1957. [18]

The majority of the results in epidemiologic scientific studies are calculated as a relative risk with 1.0 being 0% and a (95% confidence interval). This means a relative risk of 1.51 (0.93-1.87) is a 51% increased risk with a 95% chance that the actual risk is within the range given. With more data the confidence interval becomes smaller; making it an indicator of the statistical reliability of the result.

Confounding Factors

There are many confounding factors when it comes to breast cancer such as genetics and the age of menarche. As western society has modernized confounding factors such as, environmental carcinogens, delayed child rearing, hormone replacement therapy, oral contraceptives, and obesity have increased. If these were not accounted for they would obscure any effect an individual factor would have. Examining the ABC issue is all the more difficult because induced abortions have increased during the same period. [19]

In the late 20th century there was some concern in an increase of breast cancer incidence. This was found to be partly due to women living longer and better detection methods finding breast cancer earlier. [20] It should be noted the overall incidence does not effect studies with proper controls because the case and control subjects would be equally effected.

Rats

In the Russo & Russo (1980) study [21] of the Fox Chase Cancer Center in Philadelphia found that rats who received abortions had a "similar or even higher incidence of benign lesions" than virgin rats of matching age.

Russo & Russo conducted a more thorough examination of the phenomenon in 1982 [22], which confirmed the results and a later study in 1987 [23] helped to explain their previous findings. After differentiation of the mammary gland at the time of first full-term pregnancy of the rat, the rate of cell division decreases and length of the cell cycle increases, allowing more time for DNA repair. [24]

In the Discover article [25] sidebar entitled Humans Are Not Rats, Dr. Gil Mor the director of the reproductive immunology at the Yale University School of Medicine, disagrees with Dr. Brind and Drs. Russo & Russo on the importance of the rat studies findings. Dr. Mor emphasizes that rat studies are ideal for understanding basic processes but it would be scientifically "wobbly" to extrapolate that to humans. Conversely world-renowned epidemiologist, feminist and ABC skeptic Dr. Devra Davis wrote the following in When Smoke Ran Like Water:

"Of course, science has made impressive progress in understanding how cancer arises, and people certainly do differ from lab rats. Still, everything proven to cause cancer in humans has been shown to do so in animals. So how can we be sure that we are different enough?"

Interviews

Interview (case-control) based studies have been inconsistent on the ABC link. Also with the small numbers involved in each individual study and the possibility recall bias skewed the results; recent focus has switched to much larger meta-analysis and record based studies. [26] Here are a few studies of note.

Daling

Dr. Daling from the Fred Hutchinson Research Center headed two studies on the ABC issue looking at women in Washington state. The 1994 study results indicated a 50% (1.2-1.9) increased risk. [27] The risk did not vary according to number of induced abortions or children (completed pregnancies), but it did change according to age of the women at abortion and the duration of that pregnancy. This was reflected in higher risks for women younger than 18 or older than 30 years of age and had abortions after 8 weeks' gestation.

Their conclusion emphasized that although this study supported the ABC link the overall results from epidemiologic studies are inconsistent. The 1994 study appears to be the first to point out a serious flaw in a key study Lindefors-Harris (1991) [28] supporting the statistical significance of recall bias.

The Daling study in 1996 resulted in a relative risk of 1.2 (1.0-1.5). [29] The risk was highest among women without children who had abortions prior to 9 weeks' gestation. Dr. Daling et al. examined the possibility of recall bias by comparing results from two recent studies on invasive cervical cancer and ovarian cancer and the results argued against significant recall bias. The Rookus (1996) study [30] noted that patients with cervical cancer may report differently than breast cancer patients.

On September 28th, 1997 an interview with Dr. Daling was published by the L.A. Daily News. In it she made the following statement which is often quoted by anti-choice and ignored by pro-choice advocates:

"I have three sisters with breast cancer and I resent people messing with the scientific data to further their own agenda, be they pro-choice or pro-life. I would have loved to have found no association between breast cancer and abortion, but our research is rock solid and our data is accurate."

Sanderson

There was a 2001 study [31] conducted in Shanghai, China and the lead scientist was Dr. Sanderson from the University of South Carolina and South Carolina Cancer Center at Columbia. Since induced abortion is common, legal, and even mandated by the government recall bias would be minimized. The conclusion was there was no ABC link and that multiple abortions did not put one at greater risk. [32]

The problem with Chinese studies is that the same factors that make them ideal for reducing recall bias; also makes them inappropriate for comparison to the West. [33] [34] With the wide availability of abortion services over 80% of them were done within the first eight weeks of gestation. In comparison only 55% of American women had an abortion by the ninth week. [35] In line with China’s strict population control the vast majority of the abortions in the Chinese study were done after the first full-term pregnancy which had been relatively early. Once again this is not reflected in North America.

Recall Bias

Recall or response bias occurs when women are asked to provide their abortion history. Because of the personal, and in some places controversial, nature of abortion some women may not provide full disclosure. If this afflicted women in the control group, who have no serious illness and hence have less motivation to be truthful than those trying to diagnose their problem; then it would artificially create an ABC link where none existed. Two major studies were published regarding response bias.

Lindefors-Harris Study

The Lindefors-Harris (1991) study [36] was the first major study to examine recall bias. It looked at the data of two independent Swedish induced abortion studies and concluded there was a 50% (1.1-2.1) error due to recall bias. However eight women (seven cases, one control) included in this error apparently "overreported" their abortions. Meaning the women reported having an abortion but it was not reflected in the records; so it was decided these women did not have abortions.

Dr. Daling found it "reasonable to assume that virtually no women who truly did not have an abortion would claim to have had one," [37] and that the lack of corresponding records could have occurred for a variety of reasons. With these eight women (7 cases, 1 control) removed the error is reduced to 16% which severely limits its statistical significance. Subsequently the authors obliquely retracted the 50% conclusion in 1998 [38] but reasserted based on no ABC link being found in the Denmark (Melbye 1997) cohort study [39] that Dr. Brind's et al meta-analysis [40] result of 30% increased risk must be the accumulative result of recall bias.

Dr. Brind believes the remaining 16% in the Lindefors-Harris study is not beyond further scrutiny as many of the women in the control group under 40 years of age who underreported their abortion history could have resulted from the Swedish fertility registry. [41] These women were interviewed as mothers, which could increase the tendency to underreport given a mother wouldn’t want to appear unfit. [42]

Rookus Study

The Rookus (1996) study [43] compared two regions in the Netherlands to assess the effect of religion on ABC results based on interviews. The secular (western) and conservative (southeastern) regions showed ABC relative risks of 1.3 and 14.6 respectively. Although this was a large variance Dr. Brind et al. pointed out in correspondence [44] that it was attained with an extremely small sample size. (12 cases and 1 control)

So Rookus supported their finding with an analysis of how much recall bias there was with oral contraceptive use that could be verified through records. It did corroborate the bias but it only indicated recall bias between the two regions; not between case and control subjects within regions which could effect ABC findings. Dr. Brind et al. meta-analysis [45] noted that both of the results were positive and the difference between regions may have resulted from any number of confounding factors.

Dr. Rookus et al. responded [46] to Dr. Brind et al. criticism by noting there was 4.5 month underreporting difference between control and case subjects in the conservative region. With this Rookus et al. maintain this is indirect evidence for reporting bias since the willingness to report oral contraception should be higher than induced abortion. They also acknowledged the weakness in the Lindefors-Harris study but emphasized more controls (16/59 = 27.1%) than case patients (5/24 = 20.8%) did not report registered induced abortions. In conclusion they believed finding a causal link between abortion-breast cancer would be a disservice to the public and epidemiologic research if "bias has not been ruled out convincingly."

Meta-analysis

Brind study

A meta-analysis was conducted by Dr. Brind et al. (1997) with pro-choice and anti-choice scientists examining 28 published studies. [47] The conclusion was that there was on average a 30% (1.2-1.4) increased risk of breast cancer. The meta-analysis was criticized for using studies with widely varying results, not working with the raw data from several studies and including some studies that have alleged methodological weaknesses. [48]

The Royal College of Obstetrics and Gynaecology in March 2000 published evidence based guidelines on women requesting induced abortion. The review of the available evidence at the time was "inconclusive" regarding the ABC link. They also noted "Brind's paper had no methodological shortcomings and could not be disregarded." [49]

Beral study

Recently Dr. Beral et al. (2004) published the "Collaborative Group on Hormonal Factors in Breast." [50] This meta-analysis of 53 epidemologic studies undertaken in 16 countries did not find evidence of a relationship between induced abortion and breast cancer with a relative risk of 0.93 (0.89-0.96). Many organizations and media outlets have referenced it as the latest and most comprehensive overview of the ABC evidence. Dr. Brind criticizes the meta-analysis for its selection process, how it presents itself as "collaborative" and questions the publication history of the authors. [51]

Cohorts

Howe

The 1989 study [52] by Holly Howe et al. at the New York State Department of Health examined young women in New York State. The results indicated a significant 90% (1.2-3.0) increased risk for induced abortion, an insignificant 50% (0.7-3.7) increase for spontaneous abortion and 300% (1.5-13.6) increase for multiple abortions with no intervening births. However, the cohort consisted only of women under age 40 at the time of breast cancer diagnosis; and the follow-back search was restricted to events that occurred since 1971. The authors believed that the study was inconclusive but did raise new questions for continuing research as women's recorded contraceptive histories grew.

It's also strong evidence that the positive ABC results in interview studies were not purely a result of recall bias. According to Scott Somerville of Accuracy in Media it took a long time for Howe's study to be published as a result of American journals rejecting the article. Eventually in Britain the International Journal of Epidemiology published it in 1989. [53]

Lindefors-Harris

A much larger cohort study by Dr. Lindefors-Harris et al. (1989) [54] was done looking at 49,000 women who had received abortions before the age of 30 in Sweden. The risk for women with a birth previous to the abortion was 0.58 (0.38-0.84), whereas women with no births had an insignificant risk increase of 1.09 (0.71-1.56). Overall the relative risk was 0.77 (0.58-0.99) making for a 23% reduced risk in comparison to "contemporary Swedish population with due consideration to age."

The problems with the Lindefors-Harris study as Scott Somerville sees it is their control group isn't well defined. The study combines women with and without children. It also makes no attempt to compensate for the fact most women in Sweden, unlike America, already have children at the time of abortion. Thus the protective effect seen in the study could be from how abortion is used in Sweden rather than from abortion itself. They then compare the combined results to the total population which has a high number of women with abortions. Additionally a large part of the funding for the study came from "Family Health, International," which is apparently a research arm of the abortion industry. [55]

Melbye

The largest and highly regarded ABC study was published by Dr. Melbye et al. (1997) [56] of the Statens Serum Institute in Copenhagen, which had 1.5 million Danish women in the database. Of those 280,965 women had induced abortions recorded in the computerized registry started in 1973 when having an induced abortion through 12 weeks was fully legalized. The relative risk after statistical adjustment came to 1.00 (0.94 to 1.06). This led to the conclusion that "induced abortions have no overall effect on the risk of breast cancer." The Melbye study conclusions garnered great attention from the media and many organizations such as the NCI and Planned Parenthood; who use it as a foundation to argue that the best scientific evidence does not support an ABC link.

However Drs. Brind and Chinchilli had some concerns about the Melbye study database and how they statistically adjusted their overall relative risk.

The Melbye study used women born from 1935 to 1978 but the computerized registry of induced abortions only started in 1973 when having an induced abortion through 12 weeks was fully legalized. The history of legal abortion in Denmark goes back to 1939 and was expanded several times leading up to 1973. [57] It therefore isn’t surprising that Dr. Brind found more than 30,000 women had been misclassified as having no abortion because the induced abortion occurred prior to 1973. [58]

Dr. Melbye et al. responded by noting if the misclassified older women had their risk underestimated it would be expected that the younger groups would have a higher risk. Their adjusted relative risks indicated that was not the case with the relative risks for these age cohorts at 1973: 35 years or older, 1.09 (0.93 to 1.27); 30 to 34 years, 1.04 (0.94 to 1.16); 25 to 29 years, 0.92 (0.83 to 1.02); 20 to 24 years, 0.97 (0.84 to 1.13); and less than 20 years, 1.06 (0.87 to 1.29).

In a large cohort study it is necessary to account for confounding factors that may have increased over time. For example if the pill effected breast cancer rates younger women may have a higher risk than older women who did not have that option. As a result 40 year old women in 1990 (young birth-cohort) would have a higher incidence of breast cancer than 40 year olds in 1970 (older birth-cohort). Typically this is corrected by having controls who have similar confounding factors but no abortion history matched to those who had an abortion. In the Melbye study it was statistically adjusted out instead of minimizing for observed birth-cohort increases with the more expensive and time consuming case-control matching.

Since the exact reasons for the steady increase in breast cancer incidence are unknown and induced abortion have increased over the same period; Dr. Brind argues Dr. Melbye et al. adjusted out the very thing they were looking for from the overall results and the finding of exactly 1.00 indicates that is what happened. Dr. Melbye et al. found the point to be "self-contradictory" considering Dr. Brind wanted birth-cohort matching then argued "against taking birth-cohort differences into account." It is unclear how this is a contradiction since Dr. Brind is against the use of statistical adjustment, and for the use of case-control cohort matching to account for birth-cohort differences.

Drs. Senghas and Dolan did not understand why a statistically significant result for induced abortions done after 18 weeks gestation was not in the Results section of the Melbye study abstract. Even though they acknowledge the data in the category is smaller than most other categories within the study; they considered 14,000 patient years to be "hardly a small number." Dr. Melbye et al. explained that even though that result was "in line with the hypothesis of Russo and Russo," they deemed the number of cancer cases small and did not want to "overstate the finding."

Here is the first section of Table 1 in the Melbye study:

Week of gestation	No. of Cancers	Person-Years (Thousands)	Relative Risk (95% CI)*	Multivariate Relative Risk (95% CI)†
<7	36	82	0.81 (0.58-1.13)	0.81 (0.58-1.13)
7-8	526	1012	1.01 (0.89-1.14)	1.01 (0.89-1.14)
9-10‡	534	1118	1	1
11-12	205	422	1.12 (0.95-1.31)	1.12 (0.95-1.31)
13-14	6	14	1.13 (0.50-2.52)	1.13 (0.51-2.53)
15-18	17	35	1.24 (0.76-2.01)	1.23 (0.76-2.00)
>18	14	14	1.92 (1.13-3.26)	1.89 (1.11-3.22)

*The relative risks were calculated separately for each of the five variables, with adjustment for women's age, calendar period, parity, and age at delivery of a first child. CI denotes confidence interval.
†Values were adjusted for women's age, calendar period, parity, age at delivery of a first child, and the other variables shown in the table.
‡The women with this characteristic served as the reference group.

Other sections listed age at induced abortion, number of induced abortions, time since induced abortion, and time of induced abortion and live-birth history. There was an indication of an elevated risk of 1.29 (0.80-2.08) for 12-19 year olds (relative to 20-24 subcohort), and a protective effect 0.74 (0.41-1.33) for women with an induced abortion before and after their first live birth (relative to induced abortion after 1st live birth subcohort).

NCI Workshop

The National Cancer Institute conducted a workshop to evaluate the scientific evidence regarding the ABC link. This was done in response to alterations to the NCI website by the Bush administration in November 2002. [59] The workshop concluded that the evidence was well established against an abortion increasing a woman's risk of breast cancer. [60]

Dr. Brind was an invitee to this workshop and he did not like what he observed. [61] The chairperson exercised "major" control over who was to be invited, and many of the invited scientists were dependant on the NCI or other federal agencies for grants. The main expert who made the formal ABC link presentation primarily had expertise in obesity and exercise in relation to breast cancer. Dr. Daling who had published on the ABC subject was asked to present on another topic. No minority or dissenting report was requested but Dr. Brind did file a dissenting opinion. [62]

Anti-Abortion Bias

When anti-abortion advocates link abortion-breast cancer it is transparent the goal is to stop women from having an induced abortion. Because breast cancer illicits disproportionate fear [63] in women it seems accurate to say they use it as scare tactic. To this end anti-choice advocates focus on positive and/or averaged results, ignoring caveats, and low risk subgroups within scientific studies that could help minimize the risk. "Pro-life" advocates claim their ABC information is for the benefit of women's health and to provide "informed consent". At the same time their efforts to increase obstacles (ie. counseling, waiting periods, notification) and reducing the availability of abortion services increases the time it takes to get an abortion. [64] Thus compounding abortion risk factors and undermining pro-life claims of helping women.

Media Bias

The book What Liberal Bias? [65] by former CBS producer Bernard Goldberg examines media bias. In it Goldberg makes a detailed case against conservative assertions that the media has a liberal bias, and that actually the opposite is true. However he also points out that on hot button social issues such as the death penalty and abortion the media is quite liberal.

Conclusion

Despite claims from both sides that the ABC issue has been resolved the scientific consensus was that the evidence is inconclusive. This is due to conflicting studies and even contradictory results within studies. When a correlation is indicated the relative risk or the data is usually statistically insufficient (according to epidemiological standards) to conclude a causal link. Recently some organizations have changed their position to the evidence being well-established [66] against an ABC link.

"Because bias impedes our vision and is subject to sound inquiry, we are far from reaching a scientific ‘limit.’ Indeed, after this excursion into the issue of abortion, bias, and breast cancer, it seems our future has as much to do with human behavior as with human biology."
- Drs. Weed & Kramer (1996) J National Cancer Inst. 88(23):1698-700

For an extended argument from the anti-choice perspective, see Breast Cancer : Its Link to Abortion and the Birth Control Pill by Chris Kahlenborn, MD (ISBN 0966977734). As of November 2004, women seeking abortions in Mississippi must first sign a form indicating they've been told abortion can increase their risk of breast cancer. In Texas, Louisiana, and Kansas, state law requires women receive a pamphlet that suggests a cancer link with abortion. Similar legislation requiring notification has also been introduced, and is pending, in 14 other states. [67]

With indications from several studies, including Melbye (1997), that breast cancer risk increases marginally as the pregnancy progresses; it is yet one more reason to abort a pregnancy as soon as possible when a decision has been reached. [68]

Links

Scientific

Pro-Choice

Anti-Choice