Legal Issues
In a recent lawsuit filed by abortion providers in Indiana state court, the plaintiffs challenged Indiana’s law regulating abortion, SB-1, on the basis that it lacks an exception to permit abortion for women’s mental health.1 SB-1, which generally restricts abortion in Indiana, contains exceptions permitting abortion in cases of (1) rape and incest, (2) the diagnosis of a lethal fetal anomaly, and (3) when reasonable medical judgment dictates that performing the abortion is necessary to prevent death or a serious risk of substantial and irreversible physical impairment of a major bodily function (the “Health or Life Exception”). Plaintiffs in this lawsuit allege that this health or life exception is too narrowly tailored and thus violates the right to privacy in the Indiana Constitution. Their argument rests on the claim that abortion is often necessary to improve the mental health of pregnant women, particularly those with diagnosed mental health disorders. Here I examine whether this claim can be justified based on current empirical evidence on mental health and abortion.
Mental Health Effects of Abortion
To summarize the varied and disparate research literature on the relationship—if any—between abortion and mental health, I will focus on review articles, meta-analyses, medical record studies, and prospective longitudinal population-based studies. While the quality and results of some studies on these questions are mixed, I will argue that relevant conclusions can be drawn from the research as a whole.
As I will attempt to show, the current body of research suggests that a significant number of women suffer negative mental health consequences of abortion, with some identifiable risk factors. As the U.S. Supreme Court acknowledged in Gonzales v. Carhart, “It seems unexceptionable to conclude some women come to regret their choice to abort the infant life they once created and sustained. Severe depression and loss of esteem can follow” (IV.A).2 Evidence from clinical, epidemiological, and social science research supports this conclusion. Some women who obtain abortions subsequently suffer psychologically complex and distressing consequences, and in many of these cases, psychological harms are pronounced and measurable. Identifiable medical, psychological, and social factors in the patient’s history can also help predict which patients may be at elevated risk. By contrast, there are currently no available published studies suggesting that abortion improves mental health outcomes in some cases.
A. Reviews and Meta-Analyses
An early review article by Thorp and colleagues in 2003 found that abortion was associated with increased depression and could lead to self-harm behaviors. The authors concluded: “Induced abortion increased the risks for both a subsequent preterm delivery and mood disorders substantial enough to provoke attempts of self-harm. Preterm delivery and depression are important conditions in women’s health and avoidance of induced abortion has potential as a strategy to reduce their prevalence.”3
A 2011 meta-analysis by Coleman4 quantitatively synthesized research published between 1995 and 2009 on abortion and subsequent mental health outcomes. This analysis pooled 22 published studies, with over 800,000 participants and more than 161,000 women who had undergone abortions—the largest study to date of pooled data on the question of abortion and mental health. Coleman’s meta-analysis examined whether abortion is associated with a higher subsequent risk for well-defined mental health problems: (1) anxiety disorders, (2) depression, (3) alcohol abuse, (4) marijuana abuse, and (5) suicide behaviors.
The results showed that women with an abortion history experienced an 81% increased risk for mental health problems of various kinds compared to women who had not had an abortion. The study found statistically significant effects for all five areas measured: anxiety disorders increased by 34%, depression increased by 37%, alcohol abuse increased by 110%, marijuana abuse in-creased by 220%, and suicidal behaviors increased by 155%.
In addition to comparing women who had an abortion with women who did not, it is also instructive to compare women who had an abortion to women who had an unintended pregnancy but chose to carry the pregnancy to term. In Coleman’s meta-analysis, when compared to women with unintended pregnancy brought to term, women who had an abortion still had a 55% increased risk of mental health problems. Women in the unintended pregnancy carried to term group were closer to the results for the no abortion group than they were to the abortion group. So regardless of the type of comparison cohort used, abortion was associated with a significant increased risk of mental health problems in this analysis.
Following publication, some critics dismissed Coleman’s study, citing eight letters to the editor published in the same journal along with Coleman’s response.5 In her response to these letters, Coleman explained her methodological choices and the reasons these were suitable for analyzing the available literature.6 After considering the critiques, the journal editors maintained that Coleman’s paper utilized sound research methods and passed a thorough peer review process, and they appropriately refused to retract her paper in response to these critics.7 Again last year critics again pressured the journal for retraction and again the British Journal of Psychiatry declined to retract.8 Instead of considering Coleman’s findings in the context of other available research, and weighing the study’s particular strengths and limitations, many critics dismissed it wholesale due to methodological critiques. A better response would have involved publishing another meta-analysis using inclusion criteria or statistical methodology the authors believed to be more suitable; but Coleman’s critics have yet to publish their own meta-analysis.
In the same year (2011) that Coleman’s meta-analysis was published, another systemic review was published by the Academy of Medical Royal Colleges (AMRC) in the U.K.9 The results of this study were mixed: while this review did not find an overall increase in mental health problems following abortion as compared to live birth (when previous mental health problems were controlled for), it did identify a subset of women with particular risk factors who are likely at increased risk of mental health problems following abortion as compared to live birth. For example, women who show a negative emotional reaction immediately following an abortion are at elevated risk for worse mental health outcomes, as are women with previous mental health problems.10 The authors note, “Identifying these factors would enable health- care professionals to monitor and provide greater support for women identified as potentially ‘at risk’.”11 The AMRC concluded that there was no association with pregnancy outcome and mental health problems, thereby conceding that abortion did not improve mental health: “When a woman has an unwanted pregnancy, rates of mental health problems will be largely unaffected whether she has an abortion or goes on to give birth.”12
Fergusson and colleagues published in 201313 a reappraisal of the studies used in the Coleman and ARMC meta-analyses. The authors hypothesized that abortion may reduce adverse mental health consequences as compared to live birth. Instead, this study found that abortion was associated with statistically significant increases in the risks of alcohol misuse (2.3 times higher), illicit drug use/misuse (3.91 times higher), and suicidal behavior (1.69 times higher), as well as elevated risks of anxiety (though this was not statistically significant)—findings which supported a link between abortion and poor mental health outcomes.14 In disconfirming the authors’ hypothesis that women might benefit psychologically from abortion, this study also confirmed the finding of every other review or meta-analysis of the issue of abortion and mental health: abortion is not therapeutic from a mental health perspective. Research on this issue has never found abortion to confer mental health benefits and has often found it to confer mental health risks.
B. Medical Records Studies
One of the major challenges in doing research on abortion, including mental health outcomes research, is that abortion tends to be under-reported. Typically, respondents will report under half, and as few as 30%, of the number of abortions expected based on age-adjusted national data on abortion rates.15 Several studies have found that the cohort of women refusing to participate in follow-up studies are more likely to have experienced negative psychological reactions to their abortions.16 It’s not surprising that women who have had more negative reactions tend not to want to participate in abortion studies, since study questionnaires can trigger their negative thoughts and feelings.17 This results in a sampling selection bias for most research on abortion and mental health, skewing results toward finding lower rates of mental health problems associated with abortion than is actually the case.
One way for researchers to bypass this problem of selection and reporting bias is to examine medical records directly. Three well-designed studies using medical records have examined the connection between abortion and the risk for subsequent suicide. A Finnish record study showed that women who had an abortion were three times more likely to commit suicide within one year of the abortion than women in the general population, and more than six times more likely to commit suicide than women who carried their pregnancies to term.18 Another Dutch record study found increased risk of recurrence of mental health disorders following abortion in women with a history of mental illness, consistent with other studies that suggest this as a risk factor.19
A medical records study done in Denmark showed that women who had abortions were at higher risk for admission to psychiatric hospitals within three months than women who carried their pregnancies to term.20 (This finding is especially significant since the postpartum group is already at elevated risk for psychiatric hospitalization due to postpartum depression.) Finally, another study of abortion and suicide that utilized Medicaid claims for 173,000 women in California found that women who had abortion were 154% more likely to commit suicide compared to women who delivered.21
Findings from these studies are consistent with robust and consistent findings from other suicide research which shows that motherhood lowers the risk for suicide. One review of suicide risk factors noted: “Being pregnant and having young children in the home also are protective against suicide.”22 Another study on the impact of parental status on suicide risk found: “The presence of children is protective against suicide in parents in terms of having children and, to a higher degree, having a young child; these effects exist even when adjusted for marital, socioeconomic, and psychiatric status; and their influences are much stronger in women than in men.”23
C. Longitudinal Population-Based Studies
While there is evidence for an association between abortion and negative mental health outcomes, it is considerably more difficult to obtain definitive conclusions on whether abortion directly causes worse mental health outcomes. It is worth noting that Coleman, among others cited above, recognizes this limitation and discusses it when presenting the findings in her paper.24 The correlation/causation problem is a well-known challenge in social science and epidemiology. The easiest method to prove causation utilizes prospective studies where the independent variable—in this case, abortion—is manipulated. For this to happen, women would have to be randomly assigned to an abortion vs. no-abortion group and subsequently followed to see what happens (known as a randomized controlled trial or RCT). It would obviously be unethical and unacceptable ever to conduct such research. In the absence of such research, however, it is still possible to establish causation. Much valuable epidemiological research works without recourse to RCTs.
The next best kind of study for establishing causation is a population based, nationally representative longitudinal study with multiple temporal study points, while controlling for potential confounding factors. Longitudinal design also bypasses the difficulty of assessing and controlling for pre-abortion mental health status, since it includes multiple data points across time.
In 2016, Sullins published this kind of longitudinal study25 using data collected from a highly representative random sample of 8,005 American women (using The National Longitudinal Study of Adolescent to Adult Health) who were followed from ages 15 to 28 years. Results of this study showed that abortion is consistently associated with a moderate increase in risk of mental health disorders during late adolescence and early adulthood. After controlling for possible confounding variables (age, race, region of origin, parent education, and childhood poverty status) this study found increased rate of mental disorders associated with abortion (54%) and pregnancy loss (16%), but a decreased risk for live birth (19% less). The effect of repeated abortions was substantially additive, supporting the theory that emotional distress is associated with the abortions themselves, and not merely with accompanying conditions that may also be associated with the propensity to have an abortion. A similar randomly sampled population-based study was conducted by Canadian researchers in 2010 based on National Comorbidity Survey Replication data. Statistically significant associations were again found between abortion history and a wide range of mental health problems after controlling for the experience of interpersonal violence and demographic variables. When compared with women without an abortion history, women with a prior abortion experienced several statistically significant elevated risks, including a 61% increased risk of mood disorders, a 61% increased risk of social phobia, and 59% increased risk of suicidal ideation. Abortion was also linked to significant increased risks for alcohol misuse (261%), alcohol dependence (142%), drug misuse (313%), drug dependence (287%) and any substance use disorder (280%).26
D. Narrative Reviews and Negative Studies
Those who are skeptical of a link between abortion and poor mental health outcomes often cite a 2008 report from the American Psychological Association on abortion and mental health. However, to cite the APA report directly: “It is clear that some women do experience sadness, grief, and feelings of loss following termination of a pregnancy, and some experience clinically significant disorders, including depression and anxiety.”27 The somewhat misleading claim regarding this study is typically based upon a single widely reported finding highlighted in the report’s press release: “The relative risk of mental health problems among adult women who have a single, legal, first-trimester abortion of an unwanted pregnancy for nontherapeutic reasons is no greater than the risk among women who deliver an unwanted pregnancy” [italics added]. To draw from this the conclusion that abortion in general does not have a negative impact on women’s mental health is mistaken, because the many qualifiers included in this statement actually end up excluding the majority of women seeking abortion on any given day. As Reardon has cogently argued:
This reassuring conclusion was actually couched in nuances which make it applicable to only a minority of women undergoing abortions on any given day. It excludes the 48%–52% of women who already have a history of one or more abortions, the 18% of abortion patients who are minors, the 11% of patients beyond the first trimester, the 7% aborting for therapeutic reasons regarding their own health or concerns about the health of the fetus, and the 11%–64% whose pregnancies are wanted, were planned, or for which women developed an attachment despite their problematic circumstances.28
If we remove any one or any combination of these qualifiers, the risk of mental health problems following abortion significantly increases. In addition to the aforementioned qualifiers, the APA report identified several other risk factors for mental health problems after abortion, including, among others: perceived pressure from others to terminate a pregnancy; lack of perceived social support from others; low perceived or anticipated social support for the abortion decision; a prior history of mental health problems, personality factors such as low self-esteem and low perceived control over her life; use of avoidance and denial coping strategies; ambivalence about the abortion decision; low perceived ability to cope with the abortion prior to its occurrence; a history of prior abortion; abortion after the first trimester; terminating a pregnancy that is wanted or meaningful; and feelings of commitment to the pregnancy.
There are several narrative review articles that do not find a significant association between abortion and negative mental health outcomes—though as stated before, none of them found that abortion improved mental healthoutcomes. In examining the contribution of these studies it is important to understand the limitations of narrative reviews, in contrast to a meta-analysis like Coleman’s cited above. In a review article, studies with positive conclusions are individually summarized as well as studies with negative conclusions. But how these studies are weighted in a narrative review, and how overall conclusions are drawn, is largely left up to the discretion of the author.
A meta-analysis is a quantitative or numerical synthesis of data from many previously published studies. All studies are not treated equally; they are weighted statistically based upon their sample size and rigor. Data from several studies can be aggregated and analyzed together using statistical methods. The author of a meta-analysis must disclose clearly what the inclusion and exclusion criteria are, and how the data are being statistically analyzed, so that other researchers can replicate the study if they desire. Conclusions are presented according to aggregated quantitative findings, so as not to be unduly influenced by subtle biases of the authors. These methods make a meta-analysis like Coleman’s less subject to authorial bias than the narrative review studies we examine here.
For example, a review article by Charles and colleagues often cited by abortion advocates (and published in a journal funded by abortion advocates) shows some of these methodological limitations.29 The ranking system employed in this study ignored two central methodological considerations in prospective research designs: percentage of subjects consenting to participate, and retention of study participants over time. Abortion advocates also frequently cite the “Turnaway” study—a study of 956 total women, funded by private foundations with a long history of abortion advocacy30—to argue that abortion is not associated with poor mental health outcomes. However, the Turnaway study has serious methodological weaknesses which should be considered when interpreting its findings. We should note especially the low rates of initial enrollment and high rates of dropout in this study: despite a financial inducement (study participants were offered $50), over two-thirds (69%) of the women approached at the abortion clinics refused to participate in at least one interview, and half of those who agreed to enroll in the study later dropped out.31 This introduces a significant selection bias into the study sample, since we know that women who are ambivalent or struggling with the abortion decision are less likely to enroll in research studies on abortion.32 The selection bias may have been further amplified by the recruitment methods. According to the portion of study protocol that the researchers published: “It is up to the clinic staff at each recruitment site to keep track of when to recruit abortion clients to match to the turnaways [women who sought an abortion but were over the state’s gestational age limit] recruited.”33 This enrollment method allowed clinic staff to exercise considerable leeway in deciding which women to invite to participate in the study. The lack of random sampling could easily exclude women whom staff anticipated would have difficulties during or following their abortion.
There are additional methodological problems with this study that seriously undermine its purported findings. In many cases, women turned away at the original clinic later obtained an abortion elsewhere; however, in the study these women were classified as “non-abortive”. Furthermore, the study did not control for a history of previous abortions and did not control for repeat abortions obtained during the five-year follow up period. One study suggested that 40% of the turnaway group had a history of at least one previous abortion.34 Likewise, a large percentage of women in the no abortion group would have had a previous abortion (48 to 52% of women according to findings from other studies35), making it impossible to distinguish effect of a past abortion from the effect of being denied an abortion. Thus, there were postabortion women in both the “abortion” group and in the “turnaway” group—who obtained an abortion either later in the pregnancy or during a prior pregnancy. The study should have controlled for previous abortions and followed up to see which women in the turnaway group later obtained an abortion elsewhere. Absent this information, a significant number of women were simply misclassified, and the study’s conclusions were consequently badly compromised.
Given the nonrandom sampling, low rates of enrollment, high dropout rates, and misclassification of many participants, the Turnaway study suffers from serious methodological limitations that make it impossible to generalize to most women seeking abortion. Furthermore, for all its limitations, the Turnaway study did not find that access to abortion improved mental health outcomes for pregnant women.
E. Diagnosis of Fetal Anomalies
The research literature on mental health outcomes and abortion in the situations when a fetal anomaly/disability is detected show similar results. This research provides evidence that patients who opt for neonatal hospice fare better than those who choose abortion. While those who care for children with disabilities often deal with associated stress, they also frequently report that caring for such a child is immensely rewarding.
One study of 405 parents facing a life-limiting fetal condition who chose to carry the child to term found absence of regret in 97.5 percent of participants. The study authors noted: “Parents valued the baby as a part of their family and had opportunities to love, hold, meet, and cherish their child. Participants treasured the time together before and after the birth. Although emotionally difficult, parents articulated an empowering, transformative experience that lingers over time.”36 Another study found similar results for parents who chose to continue a pregnancy after a lethal fetal diagnosis. The authors noted that, “After the birth, and at the time of the baby’s death, parents expressed thank- fulness that they were able to spend as much time with their baby as possible.”37 Researchers in another study were “surprised to find that the majority of parents were so happy to meet their baby, even joyful and at peace, even if he/she was stillborn or died within a few hours. No obvious pattern of parent characteristics, such as their religiosity, were associated with this response.”38
We can contrast these findings with studies of women who choose abortion after a prenatal diagnosis of a fatal anomaly. One meta-analysis examined seventeen studies on the effects of abortion following prenatal diagnosis of fatal as well as non-fatal impairments. This study found that, “couples experienced selective termination as traumatic, regardless of the prenatal test revealing the fetal impairment or stage in pregnancy in which the termination occurred.”39 Moreover, the authors also noted that, “Women who terminated pregnancies following positive prenatal diagnosis… wanted to mourn but felt they did not deserve to mourn.” The study also found that “couples, health care providers, family, and friends underestimated the intensity and duration of feelings of loss following selective termination.”40 In contrast to the generally positive experiences of women who carried such pregnancies to term described in the above studies, women in this study who elected abortion often suffered adverse psychological reactions, including inner conflict, remorse, and complicated grief, as the authors explain:
The strategies women used to reconcile conflicts engendered by selective termination—denying the personhood of the baby, limiting the information they sought about the baby, transferring agency for choice to others, adopting a stance of moral relativity, avoiding disclosing or selectively disclosing the event to others—worked briefly but the women ultimately felt as if they were betraying themselves and their babies.41
Another study examining outcomes of prenatal anomaly situations, which directly compared both groups, found: “Women who terminated re- ported significantly more despair, avoidance and depression than women who continued the pregnancy.” The authors concluded, “There appears to be a psychological benefit to women to continue the pregnancy following a lethal fetal diagnosis. Following a lethal fetal diagnosis, the risks and benefits, including psychological effects, of termination and continuation of pregnancy should be discussed in detail with an effort to be as nondirective as possible.”42
Conclusion
In summary, while there remains disagreement among researchers regarding abortion and mental health, there are substantial areas of concurrence among the major studies: (1) abortion is consistently associated with elevated rates of mental health problems compared to women without a history of abortion; (2) the abortion experience contributes to mental health problems for at least some women; (3) there are risk factors, such as pre-existing mental illness, that identify women at elevated risk of mental health problems after an abortion; (4) it is challenging to conduct research in this field in a manner that can definitively identify the extent to which any mental illnesses following abortion can be causally attributed to abortion itself, however, available research is strongly suggestive of a causal link between abortion and poor mental health outcomes for some women; and (5) most importantly, no available research demonstrates that abortion improves mental health outcomes for pregnant women.43
In short, the research strongly suggests that abortion will worsen, not improve, women’s mental health overall—at least for a subset of women. On the other hand, research has not shown abortion to have any therapeutic benefit for women’s mental health. This undermines legal arguments that abortion is therapeutically necessary for addressing or treating women with mental health conditions, as in the recent case against Indiana’s abortion law cited at the beginning of this paper.
