Induced abortion and anxiety, mood, and substance abuse disorders: Isolating the effects of abortion in the national comorbidity survey

https://doi.org/10.1016/j.jpsychires.2008.10.009Get rights and content

Abstract

The purpose of this study was to examine associations between abortion history and a wide range of anxiety (panic disorder, panic attacks, PTSD, Agoraphobia), mood (bipolar disorder, mania, major depression), and substance abuse disorders (alcohol and drug abuse and dependence) using a nationally representative US sample, the national comorbidity survey. Abortion was found to be related to an increased risk for a variety of mental health problems (panic attacks, panic disorder, agoraphobia, PTSD, bipolar disorder, major depression with and without hierarchy), and substance abuse disorders after statistical controls were instituted for a wide range of personal, situational, and demographic variables. Calculation of population attributable risks indicated that abortion was implicated in between 4.3% and 16.6% of the incidence of these disorders. Future research is needed to identify mediating mechanisms linking abortion to various disorders and to understand individual difference factors associated with vulnerability to developing a particular mental health problem after abortion.

Introduction

Does induced abortion carry the potential to adversely affect the psychological well-being of women? This seemingly straight-forward question is complicated by a number of characteristics inherent in the variables of interest as well as external factors surrounding investigative efforts. Diverse personal, relational, situational, and cultural forces converge in every woman’s decision to abort and adjustment afterwards is likewise embedded in a multifaceted context rendering it difficult to tease out effects of the procedure. The private, sensitive, and frequently distressing nature of the abortion experience also introduces challenges to data collection with many women declining to participate or dropping out mid-study resulting in potentially skewed results. Finally, as a topic of academic study with bearing on a divisive social issue that engenders strong emotion, the socio-political views of researchers, reviewers, and journal editors may compromise objectivity in data collection, analysis, interpretation, and publication.

Despite these obstacles, the international literature pertaining to abortion as a predictor of adverse mental health outcomes has grown considerably in the past several decades and the rigor of the published studies has increased. Bradshaw and Slade (2003), authors of an extensive review of published studies on abortion and emotional experiences concluded “There has been increasing understanding of abortion as a potential trauma” (p. 929) and “The quality of studies has improved, although there are still some methodological weaknesses” (p. 929). In a review by Thorp et al. (2003) employing strict inclusion criteria related to sample size and length of time before follow-up, the researchers concluded that induced abortion increased the risk for “mood disorders substantial enough to provoke attempts of self-harm.” (p. 67).

Employment of national data sets with reproductive history and mental health variables collected for broad investigative purposes greatly minimizes the potential for bias in data collection and low consent-to-participate rates which might otherwise compromise research on abortion. Large government funded data collection efforts have the benefit of employing professionally trained researchers or clinicians who are blind to the hypotheses of potential studies generated from the data. Further, the integrity and utility of data is maximized when trained professionals interview respondents to determine if they have experienced the symptoms of various disorders. Large-scale, national data sets also typically contain numerous personal and family history background variables that can be conveniently used as control variables.

Unfortunately the number of studies employing large representative samples with controls for third variables likely to be related to both the choice to abort and to the development of mental health problems remains rather small or non-existent for some disorders. Nevertheless, there are studies with nationally representative samples and a variety of controls for extraneous variables indicating an induced abortion puts women at risk for depression (Cougle et al., 2003, Fergusson et al., 2006, Pedersen, 2008, Rees and Sabia, 2007), anxiety (Cougle et al., 2005, Fergusson et al., 2006), and substance abuse (Coleman, 2006, Fergusson et al., 2006, Pedersen, 2007, Reardon et al., 2004). Only one of these studies incorporated a comprehensive measure of mental health problems, leading to insight regarding the likelihood that women who have an abortion will develop an actual diagnosable psychological disorder (Fergusson et al., 2006).

There are a few studies employing national samples that have failed to detect significant associations between abortion and subsequent mental health (Gilchrist et al., 1995, Schmiege and Russo, 2005). However, in the Gilchrist et al. study, very few controls were applied for confounding third variables. As a result, the comparison groups may very well have differed systematically with regard to income, relationship quality including exposure to domestic violence, social support, and other potentially critical factors. The attrition rate in this study was very high and there were additional methodological shortcomings. In the Schmiege and Russo (2005) study central analyzes lacked controls for variables identified as significant predictors of abortion (higher education, income, and smaller family size). Without the controls, the delivery group, which was associated with lower education and income and larger families, had more depression variance erroneously attributed to pregnancy resolution.

In a recently published qualitative paper by Goodwin and Ogden (2007), the authors concluded that “women’s responses to their abortion do not always follow the suggested reactions of grief, but are varied and located within the personal and social context” (p. 231). This reality underscores the necessity of employing sufficient controls for confounding variables. All the large-scale studies described above controlled for an assortment of basic demographic variables including age, marital status, social support, number of children, and education. Many of these studies also included control variables indicative of pre-abortion mental health. However, a handful of very recent studies have gone a step beyond and included experiential variables that may be related to the choice to abort and to mental health outcomes. Among the variables in this latter category are relationship problems and childhood or adult history of physical and/or sexual abuse (Fergusson et al., 2006, Pedersen, 2007, Pedersen, 2008).

There is ample evidence indicating adverse interpersonal experiences, particularly abuse of various forms, predisposes individuals to emotional problems and mental illness (Adams and Bukowski, 2007, Fergusson et al., 1996, Ferraro and Johnson, 1983, Neria et al., 2008, Schilling et al., 2007, Zlotnick et al., 2000). Women who experience intimate partner violence are also more likely to abort compared to women who were not victimized (Silverman et al., 2007), necessitating the advent of controls for these personal history variables in studies of abortion and mental health.

No existing studies of abortion and mental health have included all the above categories of potential third variables in addition to incorporating variables suggestive of other sources of significant stress in women’s lives. One obvious factor that should be controlled is history of miscarriage or stillbirth as non-voluntary forms of perinatal loss have been linked with mental health problems including anxiety and depression (Broen et al., 2005). Miscarriages are common, with estimates ranging from 25% to 43% of women experiencing at least one in their lifetime (Cote-Arsenault and Dombeck, 2001), underscoring the need to collect data on involuntary perinatal loss and control for it in research on the mental health effects of abortion. Serious accidents or life threatening illnesses, chronic health problems, heavy familial demands, and difficulty paying bills are relatively common stressors that should be controlled as well. Social support is another variable that may differ systematically based on abortion choice and/or mental health status and there is research indicating that women who have a strong support system are less likely to be harmed by an abortion (Major et al., 1990).

The purpose of the current study was to explore associations between abortion history and a wide range of anxiety (panic disorder, panic attacks, PTSD, agoraphobia), mood (bipolar disorder, mania, major depression), and substance abuse disorders (alcohol and drug abuse and dependence) using a nationally representative sample. In line with current research trends, the present study incorporates controls for 22 personal history and socio-demographic characteristics. Data from the national comorbidity survey were selected because the data base provides the most comprehensive epidemiological data on the prevalence of psychological disorders in the US. Given that most of the previously reviewed large-scale studies employing a variety of controls have detected an independent contribution of abortion to a variety of mental health concerns, abortion was hypothesized to have a similar effect with the present survey data, which employed more comprehensive assessments and a more expansive list of controls.

A few of the diagnoses examined herein have not been actively explored in the previous literature on abortion and mental health and inclusion will expand the range of outcomes that have been investigated. Although only one study has identified an association between abortion history and bipolar disorder (Coleman et al., 2002) an extensive literature review conducted by Alloy et al. (2005) revealed that individuals with bipolar disorder often experience an increase in stressful events before the onset and recurrences of mood episodes. Similarly no studies to date have examined a potential link between abortion and panic attacks or panic disorder, yet panic disorder is twice as common in women compared to men and research indicates a history of psychosocial stressors including trauma in many who experience panic episodes (Sansone et al., 1998).

Most of the diagnoses examined in this report have been identified as significant correlates of abortion; however the effects have not been isolated effectively due to insufficient controls for third variables. In the context of surveying and controlling for potential third variables, this study has the added benefit of providing useful data regarding the magnitude of a large number of individual and situational predictors of several different mental disorders. Oftentimes when the available evidence pertaining to abortion and mental health is debated, there is an assumption that the correlational evidence could likely be explained away by uncontrolled third variables. For example, some may argue it is not the abortion per se, but exposure to intimate partner violence that is behind both the abortion choice and ensuing mental health struggles. Quantification of these risks should bring some clarity to the debate.

Section snippets

Data source

The national comorbidity survey (NCS) is widely recognized as the first nationally representative survey of mental health in the United States. The general purpose of the NCS was to study the prevalence and correlates of DSM III-R disorders and service utilization trends for these disorders (Kessler, 2008). The structured psychiatric interviews were administered by the Survey Research Center at the University of Michigan (UM), Ann Arbor, between September 14, 1990 and February 6, 1992.

Participants

The NCS

Results

The control variables employed in this study are listed in Table 1. Significance tests (chi-square for dichotomous variables and t-tests for continuous dependent variables) revealed differences between women with and without abortion experience relative to marital status, race, number of residents in the respondent’s household, employment status, educational attainment, feelings of being worthy/equal to others, history of miscarriage/stillbirth, rape, having been sexually molested in childhood,

Discussion

The results of this study revealed that women who have aborted are at a higher risk for a variety of mental health problems including anxiety (panic attacks, panic disorder, agoraphobia, PTSD), mood (bipolar disorder, major depression with and without hierarchy), and substance abuse disorders when compared to women without a history of abortion after controls were instituted for a wide range of personal, situational, and demographic factors. As noted above there were a number of demographic and

Contributors

Priscilla K. Coleman, PhD: Conceptualization of study design, data analysis, interpretation of results, writing and editing of manuscript. Catherine T. Coyle, PhD: Consultation, interpretation of results, and assisted with writing and editing of manuscript. Martha Shuping, MD: Consultation, interpretation of results, and assisted with writing and editing of manuscript. Vincent M. Rue, PhD: Consultation, interpretation of results, and assisted with writing and editing of manuscript.

Role of funding source

None

References (38)

  • P.K. Coleman et al.

    State-funded abortions versus deliveries: a comparison of outpatient mental health claims over four years

    The American Journal of Orthopsychiatry

    (2002)
  • D. Cote-Arsenault et al.

    Maternal assignment of fetal personhood to a previous pregnancy loss: relationship to anxiety in the current pregnancy

    Health Care for Women International

    (2001)
  • Cougle J, Reardon DC, Coleman PK. Depression associated with abortion and childbirth: a long-term analysis of the NLSY...
  • M. Farrell et al.

    Nicotine, alcohol, and drug dependence and psychiatric comorbidity: results of a national household survey

    British Journal of Psychiatry

    (2001)
  • D.M. Fergusson et al.

    Abortion in young women and subsequent mental health

    Journal of Child Psychology and Psychiatry

    (2006)
  • K. Ferraro et al.

    How women experience battering: the process of victimization

    Social Problems

    (1983)
  • L.B. Finer et al.

    Disparities in rates of unintended pregnancy in the United States, 1994 and 2001

    Perspectives on Sexual and Reproductive Health

    (2006)
  • A. Gilchrist et al.

    Termination of pregnancy and psychiatric morbidity

    British Journal of Psychiatry

    (1995)
  • P. Goodwin et al.

    Women’s reflections about their past abortions: an exploration of how emotional reactions change over time

    Psychology and Health

    (2007)
  • Cited by (62)

    • Update on Reproductive Rights and Women's Mental Health

      2019, Medical Clinics of North America
    • Reproductive Rights and Women's Mental Health

      2017, Psychiatric Clinics of North America
    • Psychological aspects of abortion

      2016, Journal de Gynecologie Obstetrique et Biologie de la Reproduction
    View all citing articles on Scopus
    View full text