Abortion, Mental Health, and Politicized Science

The news site Mercatornet.com conducted an important interview with psychologist Priscilla Coleman, who has published contrarian research results on the link between abortion and mental health problems. Priscilla K. Coleman is a Professor of Human Development and Family Studies at Bowling Green State University in Ohio. Dr Coleman has nearly 50 peer-reviewed journal articles published, including 33 on abortion and mental health. In recognition of her strong publication record, she has been called to serve as an expert in several state and civil court cases, has spoken at the UN, and in 2007 she testified before U.S. Congress. Dr. Coleman is currently on the editorial boards for five international psychology and medical journals. Crisis reprints that interview with permission.

MercatorNet: In September you had a study published in the British Journal of Psychiatry showing an association between induced abortion and mental health problems. What were your main findings and how significant is this study?

Priscilla Coleman: This review offers the largest estimate of mental health risks associated with abortion available in the world literature. The results revealed moderate to high increased risk of mental health problems after abortion. Consistent with evidence-based medicine, this information should be used by health care professionals.

These are the basic results and what should women know:

Overall, women with an abortion history experienced an 81% increased risk for mental health problems. The results showed that the level of increased risk associated with abortion varies from 34% to 230% depending on the nature of the outcome. Separate effects were calculated based on the type of mental health outcome with the results revealing the following: the increased risk for anxiety disorders was 34%; for depression it was 37%; for alcohol use/abuse it was 110%, for marijuana use/abuse it was 220%, and for suicide behaviors it was 155%.

When compared to unintended pregnancy delivered, women who terminated had a 55% increased risk of experiencing any mental health problem.

Finally, nearly 10% of the incidence of all mental health problems was shown to be directly attributable to abortion.

Why was this meta-analysis conducted?

Recently published, less systematic reviews of the scientific literature on abortion and mental health, including the American Psychological Association report of 2008 and one by Johns Hopkins researchers among others, are prone to bias, and as a result actively mislead the public.

Practitioners need an accurate synopsis of the best available evidence in order to provide women with valid information in order to make informed health care decisions.

How is a meta-analysis different from the many other papers and reviews published in recent years? Why should we believe the information from a meta-analysis is any more accurate and reliable or less biased?

A meta-analysis is a quantitative or numerical synthesis of data from many previously published studies.

In a meta-analysis all studies are not treated equally. Contributions of individual study effects to the overall results are weighted statistically based on sample size.

Only studies that meet very stringent methodologically-based criteria are entered into the analysis; whereas in other types of reviews authors may not reveal the criteria employed or the criteria may be too restrictive (missing valuable studies) or too general (including weak studies in conclusions).

The bottom line is the results are far more reliable than the results of a single study or a qualitative review, because of the wealth of data incorporated and the objective methods for combining effects.

Isn’t it possible for meta-analyses to differ in strength and reliability? What distinguishes this one?

In order to avoid any allegations of bias, very stringent inclusion criteria were employed. This means every strong study was included and weaker studies were excluded. Specifically, among the rules for inclusion were sample size of 100 or more participants, use of a comparison group, and employment of controls for variables that may confound the effects such as demographics, exposure to violence, prior history of mental health problems.

This meta-analysis is based on 22 published studies, 36 effects, and it brings together data on 877,181 participants, 163,831 of whom experienced an abortion.

The paper has been published in a very prestigious journal, the British Journal of Psychiatry, which is considered one of the top psychiatry journals in the world. This means the paper has been extensively scrutinized by well-respected scientists and the results of studies are trusted by practitioners throughout the world.

The study attracted a lot of criticism — from the American Psychological Association among others — and you were accused of everything from professional incompetence (virtually) to personal bias. Were you surprised at this barrage?

No not at all. It is much easier for them to attack me than to accept the reality of negative mental health consequences of abortion, a reality that goes against the “civil right” the APA has been advocating for over the last four decades. They are extremely biased on this issue and it wouldn’t be realistic for me to expect to be easily recognized by the APA or similarly minded groups as an objective well-trained scientist.

Were any of the criticisms valid, in your view? What about the charge that you did not control for pre-existing mental health issues in the women who had such problems after an abortion? And the fact that half the studies you reviewed were authored or co-authored by you?

None that I’ve heard so far. We selected only studies that included controls for 3rd variables and prior psychological history was controlled in 11 out of the 22 studies included.

A Danish study published in the New England Journal of Medicine while yours was awaiting publication has been compared favourably with your study. What does it contribute to knowledge about this issue?

Very little, actually. The authors were able to publish in the NEJM because they had politically correct findings. In reality there are many problems with this study.

The researchers focus on the fact that there is not a statistically significant difference in first-time inpatient admissions and outpatient psychiatric visits before and after an abortion, concluding that it is unlikely that the abortion procedure causes mental health problems.

However there are some major problems with this conclusion.

First, the measure of pre-abortion mental health is likely high (more than 3 times greater than prior to birth, 14.6% vs. 3.9%).

This may be because many of the women were probably in the midst of abortion decision making when they experienced their first psychiatric visit or they were involved in unstable or possibly violent relationships.

This high rate of pre-abortion mental health problems is construed to indicate that women who choose abortion will often experience post-abortion mental health problems based on factors other than the procedure. In fact, the women in the sample are quite unlikely to fall into this “vulnerable” category since none of the women included in the study had any history of psychological diagnoses prior to 9 months before the abortion.

These researchers used a window of 0-9 months to measure pre-abortion mental health; however, the assessment should instead have been before the pregnancies were detected.

Even with these shortcomings, the data do indicate that rates of mental health problems are significantly higher after abortion compared to after childbirth (15.2% vs. 6.7%) and compared to not having been pregnant (8.2%).

The bottom line is: the fact that they found comparable rates before and after abortion does not negate a possible causal link between abortion and mental health.

This is true because many women were likely disturbed to the point of seeking help, because they were pregnant and contemplating an abortion, or had already chosen one and were awaiting the procedure, or they were involved in troubled relationships. There are numerous published studies indicating high levels of stress among women facing an unplanned pregnancy and considering an abortion and many women who seek abortion are in abusive relationships.

Second, the authors note in the beginning of their article that previous studies lack controls for third variables, but the only third variables they consider are age and parity. They include no controls for pregnancy wantedness, coercion by others to abort, marital status, income, education, exposure to violence and other traumas, etc. Many studies have been deemed inadequate based on only one of these variables not being accounted for (see APA Task Force Report, 2008), yet the study design was considered adequate to merit publication in the NEJM.

Third, all women who had psychiatric histories more than 9 months prior to the abortion were not included in the study and there are many studies showing that these women are at heightened risk for post-abortion mental health problems. In this study, the researchers have narrowed the participant pool to only the healthiest of women and yet there are high rates before and after abortion…Imagine if all women had been included! Women who experience repeat abortions are likewise not considered at all and they are more likely to be at risk for mental health problems post-dating the procedure.

Fourth, the results follow women for only one year post-abortion or childbirth and there is plenty of evidence suggesting that the negative effects of abortion may not surface for several years. There is also data indicating that women are most likely to experience postpartum psychological problems soon after birth with the benefits of motherhood often manifesting later than the first year wherein many life-style adjustments are necessary.

The Danish Civil Registration System (data source) contains over 40 years of data, but the researchers compressed the study period to 12 yrs.

A more appropriate analytic strategy would have been to include all women experiencing an abortion, a birth, or no pregnancy and then compare pre and post-pregnancy mental health visits with statistic al controls for all psychiatric visits pre-dating conception and all other relevant third variables described above. I am confident that the data would then be quite consistent with the dozens of studies published in recent years in high impact journals indicating that abortion increases risk for a variety of mental health problems.

However, even without appropriate improvements to the design, the data reported does indicate increased rates of particular diagnoses at specific points in the first year. Relative risk for psychiatric visits involving neurotic, stress-related, or somatoform disorders was 47% and 37% higher post-abortion compared to pre-abortion at 2 and 3 months respectively. In addition, psychiatric contact for personality or behavioral disorders was 56%, 45%, 31%, and 55% higher at 3, 4-6, 7-9, and 10-12 months respectively.

On the subject of personal bias — a couple of critics called you “an anti-abortion campaigner”. Are you? Do you receive funding from pro-life groups?

No. I do not hold membership in any political organizations and my work has never been funded by any pro-life group. As a professor at a public university, what motivates me is simply the desire to foster high quality research and reach as many people as one individual can with an accurate appraisal of the literature, given the biases that permeate the study of abortion and dissemination of information through the usual channels.

Have you ever been asked to address a pro-choice group?

Not beyond my own university, but I have had pro-choice colleagues who teach health and women’s studies courses invite me to their classes for guest lectures on the research related to abortion and mental health.

Is it difficult for you to do research on this subject and get it published in mainstream journals?

It is never very easy when the results are so counter to what the dominant political groups want to hear, but we usually persevere and continue resubmitting until the papers find a home. I have published over 30 papers on the topic and sometimes it is very difficult and takes years, other times the papers are accepted the first time they’re submitted. We are always very careful about our methodology since one or two shortcomings and the papers are rejected. These papers seem to be held to a much higher standard than papers that show not serious complications of abortion.

Since abortion is so politicised, won’t it always be like this: allegations of bias, methodological nit-picking, and even pressure to keep information from being published? Are there any signs that the wider research community and/or healthcare professionals are opening up to evidence on this issue?

The general push towards evidence-based practice in medicine is raising the bar for all services to be backed by legitimate current research data. Influential professionals not associated with pro-life groups have recognized the scientific evidence. These include Dingle in Australia, Pedersen in Norway and Fergusson in New Zealand.

Fergusson and colleagues (2006) concluded: “the present research raises the possibility that for some young women, exposure to abortion is a traumatic life event which increases longer-term susceptibility to common mental disorders.” After the Fergusson study was published a group of psychiatrists and obstetricians wrote a letter to the London Times stating: “Since women having abortions can no longer be said to have a low risk of suffering from psychiatric conditions such as depression, doctors have a duty to advise about long-term adverse psychological consequences of abortion.”

There is also more willingness on the part of mainstream news media to describe the research. Fair reports have appeared in the Wall Street Journal and the London Times, among others.

Recent years have brought more legal challenges wherein women’s abortion-related injuries have received more public attention. Women are more outspoken and the topic of abortion is finding its way into the popular media. A good example is Bruce Issacson’s movie South Dakota: A Woman’s Right to Choose, in which the lives of two teens are followed — one chooses abortion and one chooses birth.

How important is it for women that a truly scientific approach should prevail?

Extremely important. This is a very common medical procedure and at least 20 per cent of women who undergo an abortion are at risk for serious psychological problems. They have a right to know if they are in a high risk group and what the real risks are afterwards, just as with any medical procedure.

For the published meta-analysis see: “Abortion and mental health: quantitative synthesis and analysis of research published 1995-2009,” The British Journal of Psychiatry, September 2011. Dr Coleman’s reply in BJP to correspondence is here.