The Complexity and Distortions of Post-Abortion Research

David C. Reardon, Ph.D.

The Distortions of Scientific Research

Is Abortion Better Than Having an 'Unwanted' Child?

Notes

Efforts to scientifically measure psychological problems after abortion are very difficult, highly politicized, and frequently distorted in reports to both individual patients and to the media. In fact, while there have been many studies regarding the emotional aftermath of abortion, very little has been firmly established. This is because the very nature of abortion is such that it would be extremely difficult, if not impossible, to complete a study that would be generally conclusive and above reproach.

Researchers are faced with four major obstacles:

First, in longitudinal and retrospective studies, approximately 50 percent of women who have had an abortion will conceal their past abortion[s] from interviewers¹. Even in short-term follow-up studies, there are high dropout rates, typically in the range of 20 to 60 percent. Demographic comparisons of those who initially consent to participate in a follow-up study and then subsequently refuse to be interviewed indicate that those who exclude themselves from the final sample are more likely to match the profiles of women who report the greatest post-abortion distress².

Second, women can experience a wide variety of psychological reactions related to a previous abortion, which are not always covered in a given study or easily categorized. As will be described in the case studies throughout this book, in my own practice I have treated women with symptoms of depression, anger, impacted grief, chronic guilt, anxiety, intrusive memories, self-destructive behaviors, eating disorders, substance abuse, sexual maladjustment, and personality orders, to name only a few. Some abortion reactions may fit into the model of complicated bereavement or pathological grief³. In other cases, some women exhibit symptoms that fall within the diagnostic criteria for post-traumatic stress disorder (PTSD)⁴. Still other women report symptoms that are not placed in any existing category for mental disease.

In addition, researchers who support abortion may have an entirely different definition of what symptoms are considered 'significant' than do researchers who are anti-abortion. As a result, it is practically impossible for researchers to examine the entire range of every possible post-abortion reaction in a single study. While hundreds of studies have been done, all of them provide only a limited look at a small, specific group of reactions.

Third, reactions to abortion also vary over time. Women who are initially filled with grief and self-reproach may subsequently find emotional healing, whereas women who were initially coping well may subsequently find themselves emotionally shattered. In one study of 260 women who reported negative post-abortion reactions, between 63 and 76 percent claimed there was a period of time during which they would have denied any negative feelings connected to their abortions. The average period of denial reported by the survey population was 63 months⁵.

This can impact both short and long-term studies. A woman facing a delayed reaction may score 'normal' on some scale three weeks after the abortion and be a wreck three years later. Ten years later, the same woman may score 'normal' because she has worked through her negative feelings in the intervening years. If surveyed at this time, if the researcher fails to ask how the abortion affected her previously, she will be tallied as 'normal' and unaffected by her abortion. Without a properly designed research instrument, the researcher would have no way of knowing that the abortion previously led to problems like substance abuse or a nervous breakdown.

Fourth, standardized questionnaires have been shown to be inadequate for uncovering deeply repressed feelings related to abortion⁶. One psychiatrist has reported that in the course of psychotherapy for 50 women, none of whom were originally seeking treatment for abortion-related problems, deep feelings of pain and bereavement about a prior abortion emerged during the time the patients were receiving therapy for seemingly unrelated problems⁷. In subsequent research with women who were not in psychotherapy, he found that an initial reaction of emotional numbness may distort questionnaire-based studies. He concluded that an underlying sense of loss and pain can only be reliably identified in a clinical setting⁸.

In summary, drawing conclusions from research regarding the emotional aftereffects of abortion is exceptionally difficult because (1) the cooperation of the study population is inconsistent and unreliable, (2) the variety of negative reactions reported by women is so broad that it may be impossible to examine every claimed dysfunction in a single study, (3) the intensity of reactions appears to vary over time, with many women reporting delayed reactions, and (4) the use of questionnaires and other standardized survey instruments may be inadequate for uncovering deep-seated reactions.

Given such complexities, it is understandable that Surgeon General C. Everett Koop concluded, in 1987, that the research in this field is entirely inadequate for drawing any general conclusions about either the efficacy or the dangers of induced abortion⁹. The same criticism holds to this day.

The Distortions of Scientific Research

Nonetheless, abortion proponents are often quick to make sweeping statements regarding the safety of abortion. For example, on February 10, 1988, a front page story in USA Today reported that a study by Brenda Major of 600 women found that 'most women who choose abortion don't suffer physical or emotional distress.' A more careful reading revealed that 'only' 15 percent of the women (a rather significant minority) did report depressive symptoms.

It is only in reading the actual study that one discovers other important qualifying facts. First, the researchers were only looking for signs of depression. Second, this evaluation took place just thirty minutes after the abortion, with a second follow-up evaluation three weeks after the abortion. Third, six percent of the patients refused to participate in the 30 minute follow-up, and there and a full sixty percent dropped out at the three week follow-up. Finally, in comparing patients who did exhibit depression to those who did not, Major identified six predictive risk factors for depression in the short term, each of which encompassed 33 to 52 percent of the patients studied¹⁰.

As a therapist who has found that most of my clients suffering from post-abortion problems had long periods of denial in which they coped well, I would interpret a 15 percent depression rate (with a 60 percent study dropout rate) as being very worrisomely high. While it is true that 15% is less than 50%, but to characterize this study as evidence that 'most' women never experience emotional distress after an abortion is a gross exaggeration.

Another example is a study published by Nancy Felipe Russo in 1992, and again in 1997¹¹. Russo, who is director of the women's studies program at Arizona State University, used data from the National Longitudinal Study of Youth, which is funded by the U.S. Department of Labor. NLSY has tracked the education and employment patterns of several thousand people yearly over the last two decades. At various intervals, NLSY researchers have asked women about their reproductive histories. On two occasions, in 1980 and 1987, women also answered ten questions from the Rosenberg Self-Esteem Scale. Analyzing this subset of data, Russo found that on average women reporting a single abortion did not have lower Rosenberg Self-Esteem scores than women who did not report a history of abortion. Major newspaper articles quoted Russo as claiming that her study 'proves' that abortion does not have any psychological risk. She insisted that talk about 'post-abortion syndrome' was a 'dangerous myth,' and she used her findings to attack informed consent laws that require that women be told about the physical and mental health risks of abortion¹².

The numerous flaws and deceptions involved in Russo's study have been documented at length¹³. For space purposes, I will review only a few of the most glaring problems.

First, Russo went to great lengths to represent her findings as being applicable to the general population of women having abortions. After all, the NLSY data set included 5,300 women. What she failed to tell the media, or even to mention in her paper, is that the Alan Guttmacher Institute (Planned Parenthood's research arm) had previously looked at the NLSY data set and concluded that with only 773 of those women reporting a history of abortion, more than 60 percent of the women in the NLSY data set were lying to researchers to conceal their past abortions¹⁴. In other words, the women who were most ashamed of their abortions were not included in Russo's sample of women who had abortions. It is well known that those who conceal their abortions are more likely to demographically match those women who are most likely to complain of post-abortion maladjustments¹⁵.

Second, Russo found that women with a history of multiple abortions had lower self-esteem scores than both those women who did not report a history of abortion and those who reported only one abortion. But oddly, although this finding is noted in the data, it is almost completely ignored in Russo's discussion of the findings and was never mentioned in her press releases or interviews. Since approximately half of women having abortions have had a previous abortion, this is a very disturbing finding.

Third, even when Russo's data is confined to women admitting a history of a single abortion, the self-esteem scores are clearly skewed from the 'normal' sample. While the average score is about the same, women with a history of abortion were more likely to score at the extreme ends of the scale, either very low or very high. What she fails to note is that extremely high scores may be as indicative of poor adjustment as extremely low scores. Some women will attempt to compensate for an injured psyche by projecting a grandiose or inflated sense of their self-worth. Attempting to boost their image in their own eyes and in the eyes of others, they would naturally answer Rosenberg Self Esteem questions such as 'I am a person of great worth,' at the highest end of the scale. Indeed, one in-depth study of women three to five years after their abortion found that 68 percent exhibited significant histrionic characteristics and, 33 percent had narcissistic characteristics¹⁶. Either of these personality disorders could inflate Rosenberg Self Esteem scores in such a way that a high score may actually be an indicator of ill-health rather than of 'well-being,' as Russo asserts.

Finally, self-esteem can vary from moment to moment. A woman who scores high self-esteem at one moment might suddenly score very low self-esteem when a researcher begins to ask about any number of stressful issues, such as abortion, divorce, or her relationship with her parents. The Rosenberg self-esteem scale used in the NLSY provides only a 'snap shot' of a woman's self-esteem (as defined by the limitations of that scale) at one moment in time. It cannot distinguish between women who were devastated by their abortions for several years but are now recovered and those women who are still in denial but will have a nervous breakdown next year. Russo's wholesale dismissal of 'post-abortion syndrome' based on this 'snap-shot' measure of self-esteem is like taking the temperature of AIDS patients and declaring, based on a finding that the average temperature is normal, that there is no such thing as AIDS.

In short, neither the NLSY data set nor the Rosenberg self-esteem scale were ever designed to examine post-abortion problems. They are the wrong tool for the job. Examining this data may certainly be useful for the purpose of gathering insights to guide future research. But using it as 'proof' that abortion has no mental health risks . . . that is either sloppy thinking or gravely dishonest.

A final example of how the medical discussion of abortion risks has been deeply politicized can be seen in a commentary published in the prestigious Journal of the American Medical Association, entitled 'The Myth of Abortion Trauma Syndrome' by Dr. Nada Stotland. Stotland begins and ends her piece with the emphatic pronouncement that post-abortion trauma "does not exist" and that "there is no evidence of an abortion trauma syndrome"^.17 Ironically, however, her broad declarations of absolute truth were inconsistent with the evidence she herself presented to bolster her argument.

As just one example, Stotland cited a study by Bryan Lask that found that 11 percent of the women in the study reported adverse psychological effects six months after their abortions^.18 Rather than expressing concern about these women's post-abortion problems, or worrying that this short-term study may have revealed just the tip of the iceberg, Stotland insists that this minority (less than 50 percent) reaction supports her conclusion that psychiatric illness after abortion is 'rare.'

In fact, Stotland only reported part of Lask's findings. Lask actually found that 32 percent of those studied had an "unfavorable" outcome to the abortion. This 32 percent included not only patients who had suffered post-abortion mental illness, but also patients who regretted having the abortion, and patients who had moderate to severe feelings of guilt, loss, or self-reproach which was not classified as mental illness at the time of the six month follow up.

Another example of Stotland's selective reading of the literature occurred in her summary of a study by Elizabeth Belsey. Stotland cited this study as proving that "the majority" of women having abortions feel relief and regain their pre-abortion mental health status. But once again, a reading of all of this study's findings reveals a different picture¹⁹.

Belsey's main finding is that 49 percent of the group (still a minority!) had experienced one or more maladjustments within 3 months after the abortion. Most importantly, Belsey found that the women most at risk of experiencing negative reactions could be pre-identified during pre-abortion screening. Belsey broadly summarized these high-risk screening criteria as: 1) a history of psycho-social instability, 2) a poor or unstable relationship with her male partner, 3) few friends, 4) a poor work pattern, or 5) failure to take contraceptive precautions. Using these factors, Belsey found that 64 percent of the abortion patients she studied should have been referred for more extensive counseling. Of this high risk group, 72 percent actually did develop negative post-abortion reactions, compared to the low risk group, of whom 28 percent experienced one or more negative reactions.

It is notable that in 1998 Stotland published a case study which would indicate a retreat from her previously unequivocal view that 'abortion does not cause emotional problems or mental illness.²⁰ In this case study, Stotland describes a patient whose miscarriage precipitated an unexpected release of grief over a prior abortion that shook both the patient and Stotland.²¹ Her experience with this patient inspired Stotland to question her own preconceived views about how abortion does not involve any psychological risks. In this latter, less circulated article, she attempts to call attention to 'the psychological complexities of induced abortion.' She observes that no matter what a woman's political perspective may be, 'an abortion is experienced by that woman as both the mastery of a difficult life situation and as the loss of a potential life. There is the danger that the political, sociological context can overshadow a woman's authentic, multilayered emotional experience.' The failure to address this loss, Stotland writes, 'leaves the person vulnerable to reminders and reenactments, to difficulties that may surface in life and in subsequent psychotherapy.'

Is Abortion Better Than Having an 'Unwanted' Child?

As abortion proponents have been forced to retreat from the position that abortion does not have psychological risks, they are beginning to move to the argument that 'even if abortion causes emotional problems, giving birth to an unwanted child is worse.' For example, after publishing her case study acknowledging her treatment of a case of post-abortion psychological problems, Nada Stotland has now taken the position that 'women who are at high risk [of psychological problems] after abortion are at equally high or higher risk if they continue their pregnancies.'²² She has refused, however, to identify any research that substantiates this assertion.²³ Nor has she been willing to clarify precisely what emotional problems, in her view, women who carry to term are at 'higher risk' of suffering.

No matter how many medical degrees an abortion proponent like Stotland may have, it is important to recognize that this presumption that abortion is beneficial to women, or is in any measurable way better for women than carrying to term, is no more than that - an unsubstantiated presumption. Not only do abortion proponents lack any studies to support this conclusion, but the existing conclusion clearly contradicts their position.

Clearly, both childbirth and abortion will forever change a woman's life. The question before us, however, is how to measure in some meaningful way how these two experiences affect the psychological and physical health of women. Obviously, there are some risks which are unique to abortion. Certainly, even Stotland would not claim that women who carry to term are not at 'equally high or higher risk' of experiencing the particular symptoms of guilt, shame, or remorse uniquely associated with abortion.

In fact, regarding more general psychological problems several studies have been done comparing women who abort to women who carry to term. In many, but not all of these studies, the comparison has been to women who carry an unplanned pregnancy to term. An exhaustive review of these studies clearly show that women who carry to term are less likely than those who abort to attempt suicide²⁴ to require subsequent psychiatric care²⁵ to experience sexual dysfunctions,²⁶ or to engage in alcohol abuse, drug abuse, or smoking.²⁷

In addition, while there is a long list of risk factors that reliably predict greater risk of experiencing psychological or physical problems after an abortion, there is simply no corresponding list of characteristics that predict any benefit of abortion compared to childbirth. It is actually rather shocking to discover that abortion providers have not published any research that identifies situations or characteristics wherein abortion is most likely to improve a woman's life or well-being, much less any research that has quantified such improvements.

Even in such cases where women may not be hurt physically or emotionally by an abortion, there is no logical basis for assuming that lack of harm correlates to positive benefit. Humans are extremely adaptable. Some mothers who have been denied abortion will subsequently claim, after having bonded with their children, that they never wanted an abortion in the first place.²⁸

Speaking to a reporter of the London Express in 1967, a British physician, Aleck Bourne, expressed his opposition to legalized abortion saying easy access to abortion would be a 'calamity' for women: 'I've had so many women come to my surgery and pleading with me to end their pregnancies and being very upset when I have refused. But I have never known a woman who, when the baby was born, was not overjoyed that I had not killed it." A similar sentiment is frequently reported by crisis pregnancy counselors and physicians who have successfully encouraged abortion-minded clients to choose birth.

It should never be presumed, then, that abortion automatically confers some benefit upon women. It certainly changes the courses of their lives, as does childbirth, but it has never been scientifically established when, if ever, an abortion is likely to be beneficial. Therefore, just as it is impossible to accurately estimate how many women are adversely affected by abortion, it is also impossible to estimate how many women benefit from abortion. The only difference is that while there are well-established risk factors that predict post-abortion sequelae, there is not a similar body of identified factors that are helpful for predicting when abortion is likely to benefit women.

Since abortion is sought for a wide variety of reasons, it would seem essential to know in which cases abortion best fulfills the hopes and expectations of patients. Are women who seek abortions because of relationship problems likely to report that their relationships were improved? Or were their relationships hindered? Or did the abortion not make a difference? Are women who abort to protect their educational or career plans more likely to finish school or advance in their careers than women who carry to term and resume their education or career at a later date? Do women who abort in order to avoid embarrassing themselves or their families achieve higher levels of emotional security or family harmony?

In the absence of any research demonstrating when abortion is beneficial, it is difficult to understand how physicians can fulfill their obligation to give women considering abortions sound medical advise. Just as the risks of abortion vary by the characteristics of the individual, it is likely that research into any benefits that may be attached to abortion would also indicate that these benefits are most likely to be attained in certain situations or for women meeting certain physical and psycho-social criteria. Until this research is done, proper screening for known and suspected risk factors is even more important to safeguard patients' health.

This challenge to abortion proponents has been well articulated by Philip Ney: 'We should remember that in the science of medicine, the onus of proof lies with those who perform or support any medical or surgical procedure to show beyond reasonable doubt that the procedure is both safe and therapeutic. There are no proven psychiatric indications for abortion. The best evidence shows abortion is contraindicated in major psychiatric illness. There is no good evidence that abortion is therapeutic for any medical conditions with possible rare exceptions. In fact, there are no proven medical, psychological, or social benefits . . . . If abortion was a drug or any other surgical procedure about which so many doubts have been raised regarding its safety and therapeutic effectiveness, it would have been taken off the market long ago.'²⁹

Notes

1. Jones, E.F. & Forrest, J.D., 'Under reporting of Abortion in Surveys of U.S. Women: 1976 to 1988," Demography, 29(1):113-126 (1992).

2. Sûderberg H, Andersson C, Janzon L, Sjûberg N-O, 'Selection bias in a study on how women experienced induced abortion' Eur J Obstet Gynecol Reprod Biol 77(1):67-70 (1998); Adler, N., 'Sample Attrition in Studies of Psycho social Sequelae of Abortion: How Great A Problem?' J Applied Soc Psych, 6(3):240-259 (1976).

3. Angelo, E.J., "Psychiatric Sequelae of Abortion: The Many Faces of Post-Abortion Grief" Linacre Quarterly, 59(2):69-80, 1992; Brown, D., Elkins, T.E., Lardson, D.B., 'Prolonged Grieving After Abortion,' J Clinical Ethics, 4(2):118-123 (1993).

4. Speckhard, A. & Rue, V., "Postabortion Syndrome: An Emerging Public Health Concern," J Social Issues 42(3):95-119, 1992; Barnard, C.A., The Long-Term Psycho social Effects of Abortion (Portsmouth, NH: Institute for Pregnancy Loss, 1990).

5. Reardon, D., 'Psychological Reactions Reported After Abortion,' The Post-Abortion Review, 2(3):4-8 (1994).

6. Lazarus, A. & Stern, R., 'Psychiatric Aspects of Pregnancy Termination,' Clin Obstet Gynaecol, 13:125-134 (1986).

7. Kent, I., et.al., 'Emotional Sequelae of elective Abortion,' BC Med J, 20:118-9 (1978).

8. Kent, I. & Nicholls, W., 'Bereavement in Post-Abortion Women: A Clinical Report,' World J Psychosyn 13:14-17 (1981).

9. Koop, C.E., Letter to President Reagan, January 9, 1989.

10. Brenda Major, Pallas Mueller and Katherine Hildebrandt, 'Attributions, Expectations, and Coping With Abortion,' Personality and Social Psychology, 48(3):585-599 (1985).

11. Russo, N.F. and Zierk, K.L., 'Abortion, Childbearing, and Women's Well-Being,' Professional Psychology, 23(4):296-280 (1992); Russo, N.F. and Dabul, A.J., 'The Relationship of Abortion to Well-Being: Do Race and Religion Make A Difference?' Professional Psychology, 28(1) (1997).

12. Kathy Nixon, 'Study refutes claim that abortion threatens mental health,' Mesa Tribune, Oct 4, 1992 I-1; Jane E. Brody, 'Study Disputes Abortion Trauma,' New York Times, Feb. 12, 1997, B12; Karl Bland, 'Is Politics Tainting Research? Post-abortion syndrome disputed by ASU professor,' The Phoenix Gazette, Oct. 6, 1992, B1.

13. David C. Reardon, 'A Study of Deception: Feminist Researcher 'Proves' Abortion Increases Self-Esteem', The Post-Abortion Review, 3(4):4-7 (1995).

14. Jones, E.F. & Forrest, J.D., 'Under reporting of Abortion in Surveys of U.S. Women: 1976 to 1988," Demography, 29(1):113-126 (1992).

15. Sûderberg H, Andersson C, Janzon L, Sjûberg N-O, 'Selection bias in a study on how women experienced induced abortion' Eur J Obstet Gynecol Reprod Biol 77(1):67-70 (1998); Adler, N., 'Sample Attrition in Studies of Psycho social Sequelae of Abortion: How Great A Problem?' J Applied Soc Psych, 6(3):240-259 (1976).

16. Catherine A. Barnard, 'The Long Term Psychological Effects of Abortion,' (Portsmouth, NH: Institute for Pregnancy Loss, 1990).

17. Nada L. Stotland, "The Myth of Abortion Trauma Syndrome," JAMA, 268(15):2078-2079, Oct. 21, 1992.

18. Bryan Lask, "Short-term psychiatric sequelae to therapeutic termination of pregnancy," Br J Psychiatry. 1975; 126:173-177 (1975).

19. Greer, Belsey, et al., "Psycho social Consequences of Therapeutic Abortion: Kings Therapeutic Study III," Br. J. Psychiatry, 128:74-79 (1976); and Belsey, Greer, et al., "Predictive Factors in Emotional Response to Abortion: King's Termination Study - IV," Soc. Sci. & Med. 11:71-82 (1977).

20. Nada L. Stotland, Abortion: Facts and Feelings (Washington, DC: American Psychiatric Press, 1998) 106.

21. Nada L. Stotland, 'Abortion: Social Context, Psychodynamic Implications', Am J Psychiatry, 155(7):964-967, 1998.

22. Nada L. Stotland, letter to David Reardon, Ph.D., Feb. 16, 1999.

23. David C. Reardon, letter to Nada L. Stotland, March 9, 1999.

24. Mika Gissler, et. al., 'Suicides After Pregnancy in Finland, 1987-94: Register Linkage Study,' British Medical Journal 313:1431-4 (1996).

25. See the following review article that examines 23 studies comparing women who carried to term to women who aborted: Thomas W. Strahan, 'Childbirth as Protective of the Health of Women in Contrast to Induced Abortion-III: Mental Health and Well-Being,' Research Bulletin 12(4):1-8, May/June 1998.

26. Thomas W. Strahan, 'Sexual Dysfunction Related to Induced Abortion,' Research Bulletin 11(4):1-8, September/October 1997.

27. See the following review article that examines 25 studies comparing women who carried to term to women who aborted: Thomas W. Strahan, 'Childbirth as Protective of the Health of Women in Contrast to Induced Abortion-II: Smoking, Alcohol and Drug Use,' Research Bulletin 12(3):1-7, March/April 1998.

28. David, H., et.al., Born Unwanted: Developmental Effects of Denied Abortion (New York: Springer Pub. Co., 1988).

29. Ney, P.G., 'Some Real Issues Surrounding Abortion, or, the Current Practice of Abortion is Unscientific,' The Journal of Clinical Ethics, 4(2):179-180 (1993).