Invited paper
Later Abortions and Mental Health: Psychological Experiences of Women Having Later Abortions—A Critical Review of Research

https://doi.org/10.1016/j.whi.2011.02.002Get rights and content

Abstract

Background

Some abortion policies in the U.S. are based on the notion that abortion harms women’s mental health. The American Psychological Association (APA) Task Force on Abortion and Mental Health concluded that first-trimester abortions do not harm women’s mental health. However, the APA task force does not make conclusions regarding later abortions (second trimester or beyond) and mental health. This paper critically evaluates studies on later abortion and mental health in order to inform both policy and practice.

Method

Using guidelines outlined by Steinberg and Russo (2009), post 1989 quantitative studies on later abortion and mental health were evaluated on the following qualities: 1) composition of comparison groups, 2) how prior mental health was assessed, and 3) whether common risk factors were controlled for in analyses if a significant relationship between abortion and mental health was found. Studies were evaluated with respect to the claim that later abortions harm women’s mental health.

Results

Eleven quantitative studies that compared the mental health of women having later abortions (for reasons of fetal anomaly) with other groups were evaluated. Findings differed depending on the comparison group. No studies considered the role of prepregnancy mental health, and one study considered whether factors common among women having later abortions and mental health problems drove the association between later abortion and mental health.

Conclusion

Policies based on the notion that later abortions (because of fetal anomaly) harm women's mental health are unwarranted. Because research suggests that most women who have later abortions do so for reasons other than fetal anomaly, future investigations should examine women’s psychological experiences around later abortions.

Introduction

Mental health aspects of abortion have been a topic in scholarly work, policy, and clinical care. The published version of the American Psychological Association (APA, 2008) task force report on abortion and mental health (Major et al., 2009) concluded that abortion in the first trimester does not harm women’s mental health. Major et al. (2009) made this conclusion because most research on abortion and mental health has either focused on first-trimester abortion or has not ascertained gestational age at the time of abortion. Studies that have not identified the gestational age are assumed to address how first-trimester abortions relate to mental health because most abortions occur in the first trimester (Jones, Kost, Singh, Henshaw, & Finer, 2009). Here, I review and critically examine the research on later abortions (defined as second trimester and beyond) and mental health. Following Major et al. (2009), who examined post 1989 quantitative research on first-trimester abortion and mental health, I examine post 1989 quantitative research on later abortions and mental health.

The number of women experiencing later abortions each year is not trivial. Jones et al. (2009) report that 11.3% of abortions were performed in the second trimester or later in 2004, and this proportion has been fairly consistent since 1974. This means that of the 1.22 million abortions performed in 2004, 137,860 were later abortions. Consequently, it is necessary to understand women’s experiences of later abortion, including the psychological aspects.

It is also important to understand women’s psychological experiences of later abortions for clinical and policy reasons. For at least 40 years, mental health has been present in the abortion policy arena. Before Roe v. Wade, mental health was a common legal reason for justifying abortion in the United States (Payne, Kravitz, Notman, & Anderson, 1976) and is currently a reason for allowing abortion in 23 countries (Singh, Wulf, Hussain, Bankole, & Sedgh, 2009). These policies highlight the importance of allowing abortion to promote women’s mental health. In contrast, other policies are aimed at preventing women from having abortions with the ostensible reason of protecting their mental health (Siegel, 2008). For instance, seven U.S. states have informed consent policies requiring providers to notify women of only the negative psychological effects of abortion (Guttmacher Institute, 2010). In addition, one state (Nebraska) recently passed a law outlawing abortion after 20 weeks, with at least one state legislators claiming that there is evidence that abortions harm women psychologically (LB 1103, 2010). Policy makers have relied on published studies suggesting that abortion harms women’s mental health in crafting these statutes (e.g., LB 1103 (2010), SD Codified Law 34-23A-10.1 (2005)).

While recent reviews have concluded that these published studies suggesting that first-trimester abortions harm women's mental health are methodologically flawed and cannot be used to set policy (e.g., Major et al., 2009, Charles et al., 2008, Robinson et al., 2009), less has been concluded regarding the research on the relationship of later abortions and mental health.

From a clinical perspective, it is important to understand women’s psychological experiences of later abortion so that services are better equipped to help them cope with this experience. Social, psychological, or other factors associated with having later abortions may also put women at risk of psychological problems after later abortions. Factors that may be associated with both having later abortions and poor mental health include reasons for the delay, level of wantedness of the pregnancy, experiences of adverse life circumstances, mental health, demographic and economic factors, type of procedure, experiences of fetal movement, and stigma (Drey et al., 2006, Finer et al., 2005, Finer et al., 2006; Foster et al.; Major et al., 2009, Mentula et al., 2010, Statham, 2002). Just as some women - those with histories of mental health problems or experiences of violence - following a delivery or first-trimester abortion (Steinberg et al., 2011, Steinberg and Finer, 2011, Steinberg and Russo, 2008), some women may be more likely to have mental health problems after a later abortion.

Following Steinberg and Russo’s (2009) recommendation that studies must be evaluated with respect to a specific directional hypothesis, here studies are evaluated with respect to the claim that later abortions harm women’s mental health. Evaluating studies from this perspective allows us to ascertain whether a study is biased toward confirming or rejecting this specific directional claim. If a study’s design is biased toward finding this and it does, this does not provide strong evidence that later abortion harms women’s mental health; however, if a study is biased toward finding this and it does not, then there is evidence that later abortion does not harm women’s mental health. If a study is less biased toward finding that abortion harms women’s mental health, then its findings, regardless of whether they support or disconfirm this claim, are taken as evidence for whatever is found.

Three guidelines described by Steinberg and Russo (2009) for evaluating abortion and mental health research are followed here: with whom women having abortions were compared, how prior mental health was controlled for in analyses, and whether alternative explanations were considered. Women who have later abortions must be compared with other groups of women, and certain comparison groups, such as all other women or women who have never been pregnant, are inappropriate and create bias toward finding evidence that abortion harms women’s mental health (Steinberg & Russo, 2009). Consequently, Steinberg and Russo (2009) recommend comparing women who have had abortions to women with other pregnancy outcomes. For the case of later abortions, the appropriate comparison group depends on the circumstances of the women having later abortions. For instance, if the women are having later abortions because of a severe physical or lethal fetal anomaly, the most appropriate comparison group may be women who have other types of perinatal losses such as later miscarriages, stillbirths, or neonatal deaths.

It is also important to consider how prior mental health was controlled for in analyses, because it is a strong predictive factor of mental health postpregnancy (American Psychiatric Association, 2000, Kessler and Magee, 1994). Controlling for this is important for two reasons. First, it is important to rule out pre-existing mental health differences between groups (abortion versus the comparison group) as the reason for an association. Second, controlling for prior mental health eliminates the variance in post-pregnancy mental health that is due to prior mental health, leaving less variance to be explained by the pregnancy outcome (e.g., abortion versus delivery or miscarriage), and increasing the chances of finding an association if one truly exists (Tabachnick & Fidell, 2007).

The third guideline, whether alternative explanations were considered, is relevant only if a significant association between abortion and mental health is found. This guideline recognizes that a significant association between abortion and mental health may occur for a reason other than an actual relationship. Common risk factors associated both with abortion and with mental health problems, for example, may cause an apparent association between abortion and mental health (Steinberg & Finer, 2011). Because similar circumstances, such as emotional problems, experience of violence, or lower socioeconomic status, are associated both with having an abortion and with mental health problems (Fisher et al., 2005, García-Moreno and Stöckl, 2009, Hudson, 2005, Jones et al., 2002, Kessler et al., 1994, Steinberg et al., 2011, Steinberg and Finer, 2011, Steinberg and Russo, 2008), there may be a relationship between abortion and mental health when such factors are not considered in analyses. According to this perspective, when these common factors are controlled for in analyses, the relationship between abortion and mental health should be nonsignificant or at least reduced (Steinberg & Finer, 2011).

Section snippets

Methods

I used two methods to examine the quantitative research on later abortions and mental health published in English. First, I examined the reference sections of papers that reviewed published quantitative research on abortion and mental health (APA, 2008; Bradshaw and Slade, 2003, Cameron, 2010, Casey, 2010, Charles et al., 2008, Coleman et al., 2005, Major et al., 2009, Robinson et al., 2009, Thorp et al., 2002) and on later abortions for reasons of fetal anomaly and mental health (Statham, 2002

Results

I found 19 post-1989 articles that used quantitative methods to examine the relationship of later abortion and mental health. All articles focused on women who had abortions because of fetal anomalies. Eleven articles compared women who had later abortions with another group (Davies et al., 2005, Iles, 1993, Kersting et al., 2005, Kersting et al., 2009, Kersting et al., 2007, Korenromp et al., 2005, Korenromp et al., 2007, Lilford et al., 1994, Rona et al., 1998, Salvesen et al., 1997, Zeanah

Discussion

Most comparison groups (9) in the 11 studies that I reviewed herein included women who had other pregnancy events, such as later miscarriages, stillbirths, or delivery of preterm or healthy infants. The four studies comparing women who had later abortions because of fetal anomaly with women having other perinatal losses did not find worse mental health among the former group, suggesting that policies based on the notion that later abortions (for reasons of fetal anomaly) harm women’s mental

Acknowledgments

The author thanks the Charlotte Ellertson Social Science Postdoctoral Fellowship in Abortion and Reproductive Health for the generous support while writing this, as well as Tracy Weitz for sparking my interest and engaging in discussions the topic of later abortion and mental health.

Julia Steinberg is currently an Assistant Professor in the Department of Psychiatry at UCSF. She received her PhD in Social Psychology from Arizona State University and completed the Ellertson Postdoctoral Fellowship in Abortion and Reproductive Health in the Department of OB/GYN and Reproductive Sciences at UCSF. Her research interests are at the intersection of psychology and reproductive health.

References1 (46)

  • J.R. Steinberg et al.

    Evaluating research on abortion and mental health

    Contraception

    (2009)
  • Diagnostic and statistical manual of mental disorders—text revision

    (2000)
  • Report of the Task Force on Mental Health and Abortion

    (2008)
  • P.K. Coleman et al.

    The psychology of abortion: A review and suggestions for future research

    Psychology and Health

    (2005)
  • V. *Davies et al.

    Psychological outcome in women undergoing termination of pregnancy for ultrasound-detected fetal anomaly in the first and second trimesters: A pilot study

    Ultrasound in Obstetrics and Gynecology

    (2005)
  • E.A. Drey et al.

    Risk factors associated with presenting for abortion in the second trimester

    Obstetrics & Gynecology

    (2006)
  • L.B. Finer et al.

    Reasons U.S. women have abortions: Quantitative and qualitative perspectives

    Perspectives on Sexual and Reproductive Health

    (2005)
  • W.A. Fisher et al.

    Characteristics of women undergoing repeat induced abortions

    CMAJ

    (2005)
  • An overview of abortion laws. State policies in brief

    (2010)
  • C.G. Hudson

    Socioeconomic status and mental illness: Tests of the social causation and selection hypotheses

    American Journal of Orthopsychiatry

    (2005)
  • S. *Iles et al.

    Psychiatric outcome of termination of pregnancy for foetal Abnormality

    Psychological Medicine

    (1993)
  • R.K. Jones et al.

    Patterns in the socioeconomic characteristics of women obtaining abortions in 2000–2001

    Perspectives on Sexual and Reproductive Health

    (2002)
  • R.K. Jones et al.

    Trends in abortion in the United States

    Clinical Obstetrics and Gynecology

    (2009)
  • Cited by (25)

    • CA125, progesterone & β-hCG in prediction of first trimester abortion

      2017, Reproducao e Climaterio
      Citation Excerpt :

      Pregnancy loss is a common medical problem in reproductive-aged females, with around 25 percent of all women attempting pregnancy experiencing at least one spontaneous abortion. Medically, all pregnancy losses before 20 weeks’ gestation termed as abortion.1,2 More than fifty percent of human pregnancies are aborted before term.

    • Acceptability and clinical outcomes of first- and second-trimester surgical abortion by suction aspiration in Colombia

      2014, Contraception
      Citation Excerpt :

      Our findings of a significantly higher initial satisfaction in women who presented in the second trimester compared to first trimester may be due to the lack of second-trimester abortion services in Colombia, so women are happy to be able to access care. The majority of women reported no emotional discomfort after the abortion, which is consistent with studies that show a lack of mental health consequences after an abortion, including in the second trimester [26]. The American Psychological Association has reported that perceived stigma and need for secrecy about the abortion, both common in restrictive settings [27], may be predictive of psychological responses after an abortion, such as guilt [28].

    • Receiving versus being denied an abortion and subsequent drug use

      2014, Drug and Alcohol Dependence
      Citation Excerpt :

      In addition, the American Psychological Association systematic review of the literature on abortion and mental health limited their conclusion that the best scientific evidence indicates that abortion is not associated with increased risk of subsequent mental health problems to the first trimester (American Psychological Association Task Force on Mental Health and Abortion, 2008). The limited research on effects of later abortion has primarily been conducted with women seeking abortion because of fetal anomaly (Steinberg, 2011). As women having abortions because of fetal anomalies may differ in factors, such as wantedness of pregnancy, from women having abortions for other reasons, it is not clear that this research applies to women receiving later abortions for reasons other than fetal anomalies.

    View all citing articles on Scopus

    Julia Steinberg is currently an Assistant Professor in the Department of Psychiatry at UCSF. She received her PhD in Social Psychology from Arizona State University and completed the Ellertson Postdoctoral Fellowship in Abortion and Reproductive Health in the Department of OB/GYN and Reproductive Sciences at UCSF. Her research interests are at the intersection of psychology and reproductive health.

    1

    References denoted by ‘*’ indicate a study critically examined in the review.

    View full text