Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2012 Apr 1.
Published in final edited form as: Am J Orthopsychiatry. 2011 Apr;81(2):193–201. doi: 10.1111/j.1939-0025.2011.01088.x

Does the Outcome of a First Pregnancy Predict Depression, Suicidal Ideation, or Lower Self-esteem? Data from the National Comorbidity Survey

Julia R Steinberg 1, Davida Becker 1, Jillian T Henderson 1
PMCID: PMC3076612  NIHMSID: NIHMS184490  PMID: 21486261

Abstract

The present study examines the risk of depression, suicidal ideation, and lower self-esteem following an abortion versus delivery, adjusting for important correlates. Using the National Comorbidity Survey, we examined how first pregnancy outcome (abortion versus delivery) related to subsequent major depression, suicidal ideation, and self-esteem. Models controlling for risk factors, such as background and economic factors, pre-pregnancy violence experience, and pre-pregnancy mental health, as well as a model with all risk factors, were examined. When no risk factors were entered in the model, women who had abortions were more likely to have subsequent depression (OR = 1.53, CI 1.05-2.22) and suicidal ideation (OR = 2.02, CI 1.40-2.92), but not more likely to have lower self-esteem (B = -.02). When all risk factors were entered, pregnancy outcome was not significantly related to later depression (OR = .87, CI .54-1.37) and suicidal ideation (OR = 1.19, CI .70-2.02); predictors of mental health following abortion and delivery included pre-pregnancy depression, suicidal ideation, and sexual violence. Policies and practices implemented in response to the claim that abortion hurts women are not supported by our findings. Efforts to support women’s mental health should focus on known risk factors, such as programs to address gender-based violence, rather than abortion history.

Keywords: Abortion, depression, suicidal ideation, self-esteem, control factors

Introduction

Compared to pregnancies ending in a delivery or miscarriage, pregnancies ending in abortion are much more likely to be the result of an unintended pregnancy (Finer & Henshaw, 2006). Unintended pregnancies are associated with life disadvantages such as lower socioeconomic status and violence experience (Dietz et al, 1999; Finer & Henshaw, 2006; Gazmararian et al., 1995, 2000; Goodwin et al., 2000; Russo & Denious, 1998, 2001), which are also common among women with depression, suicidal ideation, and lower self-esteem (Beitchman et al., 1992; Neumann, Houskamp, Pollock, & Briere, 1996; Taylor & Stanton, 2007). Therefore, a relationship between abortion and subsequent depression, suicidal ideation, or lower self-esteem may emerge when analyses do not adequately control for factors that capture the overlapping life circumstances of women with unintended pregnancies and poor mental health outcomes.

The claim that abortion causes mental health problems has been the basis for various state policies requiring that women seeking abortions be informed of the psychological risks, (Gold & Nash, 2007; Siegel, 2008, in press; Wilmoth, 1992; South Dakota, 2006). Psychological risks have also been cited as reasons to ban access to abortion services (Gonzales vs. Carhart, 2007). For example, in the decision to uphold the partial birth-abortion ban, Justice Kennedy stated (Gonzales vs. Carhart, 2007, p. 29) “…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.” Our study is an empirical test of this assertion. That is, we test how common severe depression, suicidal ideation, and lower self-esteem are following an abortion compared to a delivery, particularly in the context of other risk factors that have been shown to be consistent causes or correlates with mental health problems.

All research on abortion and mental health will have limitations for drawing causal inference because individuals cannot be randomly assigned to have an abortion or deliver (Adler, 2000; Steinberg & Russo, in press), but researchers can strive to minimize bias when conducting a study to test the claim that abortion leads to poor mental health. Unfortunately, much of the published research concluding that abortion causes depression or suicidal ideation is biased toward finding a relationship (e.g., Reardon & Cougle, 2002; Cougle et al., 2003; 2005; Coleman et al., 2009; Fergusson et al., 2006, 2008), in part, because inappropriate comparison groups are used or important risk factors are not taken into account. For instance, many studies compare the subsequent mental health outcomes of women who abort with women who deliver on the first pregnancy, but differentially exclude women with subsequent abortions only from the delivery group (e.g., Coleman, Reardon, Rue, & Cougle, 2002a, 200b; Cougle, Reardon, & Coleman, 2003, 2005; Reardon & Coleman, 2006; Reardon & Cougle, 2002; Reardon, Cougle, & Coleman, 2004; Reardon, Cougle, Rue, Shuping, Coleman, & Ney, 2003; Reardon, Ney, Scheuren, Cougle, Coleman, & Strahan, 2002). This creates bias towards finding worse mental health in the abortion group because it eliminates the most disadvantaged women, those from poor socioeconomic backgrounds or with histories of sexual abuse, only from the delivery group. Additionally, much research has failed to adequately control for risk factors, such as violence experience, associated with both having an abortion and having poor mental health outcomes (e.g., Coleman, 2006; Coleman et al., 2002a, 2002b; Coleman, Reardon, & Cougle, 2005; Coleman, Reardon, Rue, & Cougle, 2002a; Cougle et al., 2003; 2005; Reardon & Cougle, 2002; Reardon et al., 2002, 2003, 2004). In failing to control for important risk factors, plausible alternative explanations for a significant relationship between abortion and mental health cannot be ruled out. Indeed, analyses that control for risk factors such as violence experience demonstrate that the association of abortion and mental health is non-significant (Russo & Denious, 2001; Steinberg & Russo, 2008; Taft & Watson, 2008).

To test the claim that abortion leads to negative mental health in a nationally representative sample of women in the United States, we compare the mental health of women who have an abortion with women who deliver their first pregnancy. We examine clinically diagnosed major depression, suicidal ideation, and self-esteem. Some research has examined whether women who have an abortion are at an increased risk of subsequent depressive symptoms or disorders (e.g., Coleman et al., 2009; Reardon & Cougle, 2002; Russo & Denious, 2001; Schmiege & Russo, 2005; Taft & Watson, 2008; Pederson, 2008), having suicidal ideation or committing suicide (e.g., Fergusson et al., 2006, 2008; Gissler et al., 1996, 1997, 2004a, 2004b, 2005; Reardon et al., 2002), or lower self esteem (e.g., Bradshaw & Slade, 2003; Pope, Adler, & Tschann, 1999; Russo & Dabul, 1997; Russo & Zierk, 1992). Findings have varied depending on the outcome examined, the measurement of the outcome, the comparison group used, and what factors were controlled in the analyses. We are not aware of any studies examining all of these outcomes with the validated measures used here.

Our primary goal was to examine the prevalence of clinically diagnosed major depression and its most serious complication, suicidal ideation, and lower self-esteem in a U.S. nationally representative sample of women, after an abortion. For the purpose of informing policy, we compare women who had an abortion with those who delivered. We also present and discuss how the failure to include important risk factors that put women at risk for both abortion and poor mental health can influence findings.

Method

Sample

The data for this study are from the National Comorbidity Survey (NCS), an epidemiologic investigation designed to study the prevalence and correlates of DSM III-R disorders. The NCS was the first survey to administer a structured psychiatric interview to a nationally representative sample (see Zhao, Kessler, & Wittchen, 1994 for diagnostic criteria). The survey was fielded in 1990 to 1992 with a household-based sample of over 8,000 female and male respondents ages 15 to 54. In the NCS, data on previous events or experiences in the participants’ lives are collected retrospectively; and based on women’s responses the NCS staff computed participants’ first and most recent occurrence of each disorder. Because complex survey techniques were used to collect the data, survey design factors are applied to our analysis to account for the effects of stratification, clustering, and weighting on our estimates. The weighted n is used to describe our sample and is reported in our analyses. A subsample of the original respondents (n = 5,877) completed an NCS Part II survey that included, among other things, reproductive history information. Of these, 2,939 were women, and of these women, n = 2,062 reported having been pregnant at least once.

We compare women who aborted versus delivered on their first pregnancy event because we can most accurately ascertain when the first pregnancy occurred, and mental health subsequent to this pregnancy. We coded the outcome of each woman’s first pregnancy a abortion, delivery, or miscarriage/stillbirth according to women’s self-report. Data on when the first abortion occurred, when the first miscarriage occurred, and the age of the oldest child were used to compute the outcome of the first pregnancy. Based on the information, we were unable to determine the first pregnancy outcome for seventy women because (1) two pregnancy events occurred at the same earliest age of pregnancy (n = 37), (2) information necessary to compute the age at first abortion or miscarriage was missing (n = 6), (3) the age of a death of a child was not reported (n = 1), or (4) the age of the death of a child occurred prior to or at the same age as a first pregnancy outcome (n = 26). Women reporting miscarriages/stillbirths for the first pregnancy (n = 226) were excluded. Analyses were conducted on the (n = 1,765) women who aborted (n = 218) or delivered (n = 1,547) their first pregnancy.

Measures

Depression

Depression was assessed using the University of Michigan Composite International Diagnostic Interview (CIDI), modified from the World Health Organization CIDI (Zhao, Kessler, & Wittchen, 1994). Information regarding a woman’s first and most recent depressive episode was used to code when depression occurred relative to her first pregnancy outcome. If a depressive episode occurred at the same age as or after her first pregnancy outcome, she was coded as having the outcome subsequent to the first pregnancy in order to be conservative in our classification. If a depressive episode occurred before the age of a woman’s first pregnancy, she was coded as having depression before her first pregnancy. Three-hundred seventy-two women had at least one depressive episode after or at the same age as the first pregnancy event.

Suicidal ideation

Suicidal ideation was assessed with three questions asking if participants had ever: (1) seriously thought about committing suicide, (2) made a plan to commit suicide, or (3) attempted suicide. Women who answered yes to any of these items were coded as having suicidal ideations. Information regarding women’s first and most recent suicide thoughts, plans, or attempts was used to code when suicidal ideation occurred relative to a woman’s first pregnancy outcome. Similar to depression, if suicidal ideation occurred at the same age as or after a woman’s first pregnancy outcome, she was coded as having suicidal ideation subsequent to the first pregnancy. Two-hundred and three women had suicidal ideation after or at the same age as the first pregnancy outcome.

Self-esteem

Self-esteem was assessed at the time of the interview by the mean on four items from the Rosenberg (1965) self-esteem scale. Examples of items, rated on a four point scale, included “On the whole I am satisfied with myself” and “At times I think I am no good at all.” Scores ranged from 1 to 4 and the reliability for those who answered all items (n = 5688) was .78. Sixteen women were missing on all self-esteem items; therefore, the sample size for analyses in which self-esteem was the dependent measure is 1749, and the mean self-esteem of the sample was 3.44. Because self-esteem was assessed at the time of the interview, it was by definition after the first pregnancy event. A retrospective measure of self-esteem was not available in this survey.

Demographic and socioeconomic characteristics

We controlled for race, age at first pregnancy, marital status, education, and income at the time of the interview because previous research has found these characteristics to vary by mental health status (Cairney & Krause, 2005; Coyne & Downe, 1991; Kessler et al., 1994; Link & Phelan, 1995; Turner, Lloyd, & Roszell, 1999). Additionally, the distribution of these characteristics among women who have abortions differs from women in the general U.S. population (Jones, Darroch, & Henshaw, 2002). Information on marital status, education, and income was only assessed at the time of the interview and not at the time of the first pregnancy.

Violence experience

Information on age at first violence experience was used to classify individuals as having experienced sexual violence (rape or molestation) or other violence (including childhood physical abuse, being physically attacked, kidnapped, held captive, or threatened with a weapon) before the first pregnancy event.

Prior mental health

We coded whether women had experienced a depressive episode or suicidal ideation prior to the first pregnancy as described above. Similarly, alcohol and drug dependence before their first pregnancy were assessed by coding whether the first onset occurred prior to the first pregnancy. If the onset for the disorder was before the first pregnancy, women were coded as having the alcohol or drug dependence before the first pregnancy.

Data Analysis

For the dichotomous outcomes of depression and suicidal ideation, logistic regression models are analyzed; for the continuous outcome of self-esteem, linear regression models are analyzed. For each outcome, we present five models. First, unadjusted models are analyzed to examine the relationship of first pregnancy outcome and subsequent depression, suicidal ideation, or lower self-esteem without adjusting for any risk factors. Second, we conducted adjusted multiple logistic or linear regression to examine the independent relationship of pregnancy outcome and post-pregnancy depression, suicidal ideation, and lower self-esteem. The adjustment variables are entered in separate models to ascertain the individual influence of (1) demographic and socioeconomic factors, (2) violence experience, or (3) prior depression, suicidal ideation, alcohol dependence, or drug dependence. Finally, a fully adjusted model is analyzed. The survey design features of the NCS, which included weighting, stratification, and clustering, are accounted for in all of our analyses (Stata 10.1: College Station, TX)

Results

Means or proportions of each risk factor by first pregnancy outcome are presented in Table 1. Chi-squares were used to test whether pregnancy outcome related to the categorical risk factors and t-tests were used to test whether pregnancy outcome related to the continuous measures. Women terminating their first pregnancy were younger at the time of the first pregnancy, p < .0005, had higher education levels, p < .0005, and were less likely to be married at the time of the interview than women who delivered, p < .0005. Women aborting the first pregnancy were more likely to have experienced pre-pregnancy depression, p < .001, pre-pregnancy alcohol dependence, p < .0005, and pre-pregnancy sexual violence, p = .001, and marginally more likely to have experienced pre-pregnancy suicidal ideation, p < .07, and pre-pregnancy drug dependence, p < .06, compared to women who delivered the first pregnancy.

Table 1.

Descriptive statistics of risk factors by first pregnancy outcome (% within column)

Abortion (n= 218) Delivery (n= 1547) p-value
Age at first pregnancy 20.0 (.31) 22.0 (.19) <.0005
Marital statusa <.0005
 Married/cohabitating 64.9 76.8
 Separated/widowed/divorced 16.2 17.0
 Never married 18.9 6.2
Race ns
 White 75.3 73.1
 Hispanic 10.5 9.0
 Black 12.5 14.8
 Other 1.7 3.1
Incomea ns
 $0 - $19,999 23.6 28.0
 $20,000 - $34,999 20.0 26.1
 $35,000 - $69,999 40.0 32.7
 $70,000 + 16.3 13.3
Educationa < .0001
 0-11 Years 4.3 16.5
 12 Years 34.7 42.3
 13-15 Years 29.3 24.1
 16 + Years* 31.7 17.1
Sexual violence before first pregnancy .001
 Yes 26.3 14.0
 No 73.7 86.0
Other violence before first pregnancy ns
 Yes 12.5 9.8
 No 87.5 90.2
Alcohol dependence before first pregnancy <.0005
 Yes 7.1 2.8
 No 92.9 97.2
Drug dependence before first pregnancy <.06
 Yes 4.8 2.6
 No 95.2 97.4
Depression before first pregnancy .001
 Yes 14.6 6.9
 No 85.4 93.1
Suicidal ideation before first pregnancy .07
 Yes 12.5 7.9
 No 87.5 92.1
Open in a new tab

Note. Self-esteem and background and economic factors were assessed at the time of the interview. The values for depression and suicidal ideation are odds ratios and for self-esteem are unstandardized regression coefficients. Bolded values are significantly different at p < .05.

Tables 2 presents bivariate relationships between mental health risk factors and post-pregnancy depression, suicidal ideation, and self-esteem. Again, chi-squares were used to test whether depression or suicidal ideation related to the categorical variables and t-tests were used to test whether depression or suicidal ideation related to the continuous measures. Compared to women who were not depressed or did not have suicidal ideation women who were depressed or had suicidal ideation after their first pregnancy were younger at their first pregnancy, less likely to be married and more likely to be separated/widowed/divorced at the time of the interview, ps ≤ .001. Additionally, compared to women who did not have suicidal ideation after their first pregnancy, women who had suicidal ideation had lower incomes at the time of the interview, p ≤ .001. Finally, compared to women without post-pregnancy depression or suicidal ideation, those with post-pregnancy depression or suicidal ideation were more likely to have experienced pre-pregnancy sexual violence, ps < .0005, or other violence, ps < .005, and were more likely to have had pre-pregnancy depression, ps < .0005, suicidal ideation, ps < .0005, and alcohol, ps < .05, or drug dependence, ps < .0005. Women had lower self-esteem if they were younger at time of the first pregnancy (data not shown) B = .012, p ≤ .01, married compared to never married, p < .02, identified as White or Hispanic compared to Black, p < .05, had lower household incomes, p < .02, or had lower educational attainment, p < .05. Additionally, women who had experienced pre-pregnancy sexual or other violence, ps < .0005 or pre-pregnancy depression, suicidal ideation, or alcohol or drug dependence, ps ≤ .005 had lower self-esteem.

Table 2.

Descriptive statistics of risk factors by presence or absence of depression or suicidal ideation and mean self-esteem at each level of each risk factor

Depression Suicidal ideation Self-esteem
Yes (n = 372) No (n = 1393) Yes (n = 203) No (n = 1562) (n = 1749)
Age at first pregnancy 21.1 (.21) 21.9 (.19) 19.6 (.30) 22.0 (.17) Not applicable
Marital Status
 Married/cohabitating 67.3 77.4 62.0 77.0 3.45a
 Separated/wid/div 25.1 14.7 28.9 15.4 3.46ab
 Never married 7.6 7.9 9.1 7.6 3.31b
Race
 Hispanic 8.9 9.3 11.2 8.9 3.29a
 Black 10.5 15.6 18.8 13.9 3.58b
 Other 3.5 2.8 3.2 2.9 3.46ab
 White 77.2 72.4 66.8 74.3 3.44a
Income
 $0-$19,999 30.0 26.8 39.0 26.0 3.28a
 $20,000-$34,999 26.6 25.0 30.2 24.7 3.38a
 $35,000-$69,999 33.6 33.6 24.9 34.7 3.54b
 $70,000+ 9.9 14.7 5.9 14.7 3.66c
Education
 0-11 years 14.0 15.3 16.0 14.9 3.33a
 12 years 45.7 40.2 49.6 40.3 3.37a
 13-15 years 23.2 25.2 20.6 25.3 3.51b
 16+ years 17.2 19.4 13.9 19.6 3.59b
Sexual violence before first pregnancy
 Yes 27.4 12.3 32.9 13.3 3.20a
 No 72.6 87.7 67.3 86.7 3.49b
Other violence before first pregnancy
 Yes 16.7 8.4 25.9 8.1 3.19 a
 No 83.3 91.6 74.1 91.9 3.47 b
Alcohol dependence before first pregnancy
 Yes 6.8 2.3 6.0 2.9 3.10a
 No 93.2 97.7 94.0 97.1 3.45b
Drug dependence before first pregnancy
 Yes 5.5 2.2 7.4 2.3 3.11 a
 No 94.5 97.8 92.6 97.7 3.45 b
Depression before first pregnancy
 Yes 30.6 1.7 15.6 6.8 2.98 a
 No 69.4 98.3 84.4 93.2 3.48 b
Suicidal ideation before first pregnancy
 Yes 15.5 6.6 31.7 5.4 2.99a
 No 84.5 93.4 68.3 94.6 3.48b
Open in a new tab

Note. Self-esteem, marital status, income, and education are at time of interview. For depression and suicidal ideation, bolded values indicate significant chi-square or t-test differences, p < .05. For self-esteem superscripts with different letters indicate significantly different means, p < .05. Because the age of first pregnancy and self-esteem are both continuous, the first cell under self-esteem is left blank.

Many factors common among women who have abortions are also common among women who have post-pregnancy major depression, suicidal ideation, or lower self-esteem, such as being younger at the time of the first pregnancy, experiencing sexual violence, and having pre-pre pregnancy mental health problems.

Table 3 presents the effects of model adjustments on the odds ratio or regression coefficient comparing women who aborted with those who delivered a first pregnancy. Unadjusted analyses, reported in the first column, show that women who had abortions on their first pregnancy were more likely to have subsequent depression, OR = 1.53, p < .05, and suicidal ideation, OR = 2.02, p < .001, than women who delivered their first pregnancy. There were no differences in self-esteem for the two groups. The second through fourth columns show the odds ratios or regression coefficients when controlling for only individual sets of risk factors. There were no differences between women who abort versus deliver in post- pregnancy depression when controlling for background and economic characteristics, pre-pregnancy violence experience, or pre-pregnancy mental health. For suicidal ideation, however, only the background and socioeconomic status adjustments fully explained the bivariate relationship found with pregnancy outcome. For self-esteem, women who had abortions and those who delivered had similar levels of self-esteem, regardless of adjustment for other factors. The final column of Table 3 shows the odds ratio or regression coefficient of pregnancy outcome when controlling for all risk factors.

Table 3.

Adjusted odds ratio or regression coefficients (and confidence interval) of abortion versus delivery for subsequent depression or suicidal ideation

No factors Background and economic factors Violence Experience Pre-pregnancy mental health All factors
Depression 1.53* (1.05-2.22) 1.31 (.85-2.02) 1.36 (.89-2.06) 1.18 (.81-1.71) .87 (.54-1.37)
Suicidal ideation 2.02*** (1.40-2.92) 1.57 (.97-2.56) 1.78** (1.19-2.66) 1.86*** (1.29-2.70) 1.19 (.70-2.02)
Self-esteem -.02 (-.16-.12) -.04 (-.18-.10) .02 (-.11-.15) .04 (-.09-.17) .03 (-.09-.15)
Open in a new tab

Note. Self-esteem and background and economic factors were assessed at the time of the interview. The values for depression and suicidal ideation are odds ratios and for self-esteem are unstandardized regression coefficients.

*

p ≤ .05,

**

p ≤ .01,

***

p ≤ .001

Table 4 presents the relationship of all the risk factors to major depression, suicidal ideation, and self-esteem in the fully adjusted models. A strong predictor of depression after the first pregnancy event was depression before the first pregnancy event (OR = 32.1, p < .0005). Similarly, strong predictors of suicidal ideation after the first pregnancy event were suicidal ideation (OR = 8.0, p < .0005) and drug dependence (OR = 2.72, p < .05) before the first pregnancy outcome. Being younger at the first pregnancy (OR = .90, p < .0005) was also significantly related to having post-pregnancy depression and suicidal ideation. For self-esteem, both pre-pregnancy depression (B = -.38, p < .0005) and suicidal ideation (B = -.28, p ≤ .005) predicted self-esteem. The only significant predictor of all three mental health outcomes was experiencing pre-pregnancy sexual abuse. The OR was 1.8 to 2.3, for depression and suicidal ideation, respectively, ps ≤ .01, and B = -.19, p ≤ .002 for self-esteem.

Table 4.

Adjusted odds ratios or regression weights (and their standard errors) of all factors in the final model

Depression p-value Suicidal ideation p-value Self-esteem p-value
Abortion versus delivery on first pregnancy .87 (.19) .53 1.19 (.31) .52 .03 (.06) .69
Background and economic factors
 Age at first pregnancy .90 (.02) < . 0005 .81 (.03) < .0005 .78 (.44) ns
 Marital status
  Married 1.12 (.33) ns 1.35 (.37) ns .003 (.004) ns
  Div/sep/wid 2.53 (.69) < .005 2.72 (.69) < .0005 .16 (.09) 10
  Never married 1.00 1.00
 Race
  Black .60 (.17) <.10 1.24 (.38) ns .17(.05) .001
  Hispanic 1.12 (.24) ns 1.00 (.33) ns -.06 (.07) ns
  Other 1.38 (.61) ns 1.20 (.52) ns .04 (.14) ns
  White 1.00 1.00
 Income
  0-19,999 .87 (.22) ns 1.57 (.63) ns -.32 (.08) < .0005
  20,000-34,999 1.06 (.31) ns 1.85 (.67) ns -.24 (.06) .001
  35,000-69,999 1.26 (.32) ns 1.60 (.60) ns -.09 (.05) ns
  70,000 + 1.00 1.00
 Education
  0-11 years .91 (.36) ns .48 (.23) ns -.10 (.07) ns
  Completed high 1.28 (.39) ns .91 (.35) < .10 -.13 (.06) .02
  Some college 1.10 (.32) ns .61 (.24) ns -.02 (.06) ns
  College + 1.00 1.00
 Pre-pregnancy violence experience
  Sexual violence 2.27 (.38) <.0005 1.81 (.37) .006 -.19 (.06) .002
  Other violence 1.34 (.35) ns 2.38 (.75) .008 -.14 (.06) .02
 Pre-pregnancy mental
  Depression 32.12 (8.21) < .0005 1.22 (.37) ns -.38 (.07) < .0005
  Suicidal ideation .90 (.27) ns 8.01 (2.503) < .0005 -.28 (.09) .005
  Alcohol dependence 1.48 (.57) ns .66 (.35) ns -.06 (.10) ns
  Drug dependence .92(.37) ns 2.72 (1.33) <.05 -.04 (.12) ns
Open in a new tab

Note. Self-esteem and background and economic factors were assessed at the time of the interview. The values for depression and suicidal ideation are odds ratios and for self-esteem are unstandardized regression coefficients. For the dichotomous outcomes of depression and suicidal ideation, the reference category is the one with a 1.00. For the continuous measure of self-esteem, the reference category is the one with a dash. Reference categories for pre-pregnancy violence and mental health are not shown (since each of these variables take on only two values). Bolded coefficients are significant, p < .05

Discussion

Our study provides evidence that the relationship of abortion and mental health is explained by other factors related to both unintended pregnancy and mental health. Consequently, it is not surprising that studies that fail to control for important risk factors find poor psychological outcomes following an abortion. Indeed, in our unadjusted models, women who had an abortion were more likely to have subsequent mental health problems. However, the bivariate association between abortion and mental health was explained by other risk factors that were highly correlated with both abortion and subsequent mental health. Specifically, in the final model pre-pregnancy mental health, sexual violence, and being younger at the first pregnancy were significant predictors of depression, suicidal ideation, and self-esteem. These findings are consistent with other research on abortion and mental health (American Psychological Association, 2008; Major et al., 2000; Steinberg & Russo, 2008; Taft & Watson, 2008). Furthermore, the belief expressed by Justice Kennedy, that women who have abortions must suffer a loss of esteem was not found, even when no risk factors were taken into account, supporting other research (Bradshaw & Slade, 2003; Pope et al., 1999; Russo & Dabul, 1997; Russo & Zierk, 1992).

The finding that pre-pregnancy sexual violence was strongly and consistently related to poor mental health and having an abortion, supports other research showing the experience of early sexual violence in particular, as opposed to physical violence, puts women at high risk of unintended pregnancy and poor mental health (Beitchman et al., 1992; Neumann et al., 1996; Roosa et al., 1997). Even if sexual violence is not the direct cause of a first pregnancy, that sexual violence prior to the first pregnancy was a strong predictor of post-pregnancy mental health in the fully adjusted model speaks to the impact of such an experience on women’s lives. Moreover, other research finds that women with histories of sexual violence are more likely to be victims of subsequent sexual violence than women without such histories (Classen, Palesh, & Aggarwal, 2005; Garcia-Moreno & Stockl, 2009; Macy, 2008), putting them at repeated risk for both unintended pregnancy and poor mental health.

Limitations

Underreporting of abortion is a common problem among studies comparing women who abort to other women (Jagannathan, 2001; Jones & Forrest, 1992; Jones & Kost, 2007; Udry, Gaughan, Schwingl, & van den Berg, 1996; Smith, Adler, & Tschann, 1999). In the current study’s sample, of all women’s reported pregnancies (abortions and deliveries), 9.0 % ended in abortion between the years of 1954 and 1991. Based on data collected by periodic surveys of all identifiable abortion providers in the US from 1973 to 1991, the percentage of pregnancies (abortion and deliveries) ending in abortion ranged from 19.3% to 30.4% among women 15-44 (Jones et al., 2009). Records of the proportion of pregnancies ending in abortion before 1973 are non-existent; however, estimates are approximately 16 percent in the late 60s (Henshaw, 2009). If we assume a steady percentage of 16% from 1954 to 1972, and we use the published statistics from 1973 to 1992 reported in Jones et al. (2009), 22.5 % of all pregnancies (abortions and deliveries) between 1954 and 1991 ended in abortion among women 15-44 years of age. This suggests approximately 40% (9.0/22.5) of abortions were reported in the NCS data. In the National Survey of Family Growth (NSFG), Jones and Kost (2007) estimated that 35% to 59% of abortions were reported between 1973 and 2001. Compared to the NSFG, abortion underreporting in our sample is within the estimated range of Jones and Kost (2007). In the current study, it would be problematic if underreporting of abortion were associated with mental health, such that those with poor mental health did not report past abortions. However, previous research suggests that women less likely to report abortion have better mental health than those reporting (Jagannathan, 2001; Schmiege & Russo, 2005). In the present study, if women with better mental health were less likely to report a previous abortion, the associations found between pregnancy outcome and mental health would be attenuated.

Another limitation of the present study is that we were unable to control for important contextual factors such as pregnancy intention because it was not measured. Thus, we were unable to ascertain the extent to which pregnancy intention could contribute to the relationship of first pregnancy outcome and subsequent mental health. Indeed, Steinberg and Russo (2008) show and Charles, Polis, Sridhara, and Blum (2008) discuss the importance of pregnancy intention for subsequent mental health. If a significant relationship between abortion and poor mental were found in our study, then our results would still not be strong evidence in support of the claim that abortion causes poor mental health. This is because pregnancy intention and other contextual factors could not be ruled out as alternative plausible explanations accounting for the relationship of abortion and mental health. However, because no significant relationship existed when we controlled for pre-pregnancy risk factors of later mental health, we can be confident that the claim that abortion causes poor mental health is not supported by our data.

Conclusions

Rather than focusing on abortion as the cause of mental health problems, policies and clinical practices should be tailored to helping women most at risk of poor mental health, regardless of pregnancy outcomes. We found that that those at risk of serious post-pregnancy mental health problems were those with histories of mental health problems or an experience of sexual violence before their pregnancy. Sadly, the experience of early sexual abuse has long-lasting implications for women’s lives, and this finding should be the focus of policy and practices.

Acknowledgments

We would like to thank the Ellertson Fellowship for supporting our effort. Dr. Henderson’s effort was also supported by an NIH/NICHD Mentored Research Scientist Development Award in Population Research (K01HD054495) and resources from NIH/NCRR/OD UCSF-CTSI (KL2 RR024130). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.

Footnotes

Publisher's Disclaimer: The following manuscript is the final accepted manuscript. It has not been subjected to the final copyediting, fact-checking, and proofreading required for formal publication. It is not the definitive, publisher-authenticated version. The American Psychological Association and its Council of Editors disclaim any responsibility or liabilities for errors or omissions of this manuscript version, any version derived from this manuscript by NIH, or other third parties. The published version is available at www.apa.org/pubs/journals/ort.

References

  1. Adler NE. Abortion and the null hypothesis. Archives of General Psychiatry. 2000;57:785–786. doi: 10.1001/archpsyc.57.8.785. [DOI] [PubMed] [Google Scholar]
  2. Beitchman JH, Zucker KJ, Hood JE, DaCosta GA, Akman D, Cassavia E. A review of the long-term effects of child sexual abuse. Child Abuse & Neglect. 1992;16:101–118. doi: 10.1016/0145-2134(92)90011-f. [DOI] [PubMed] [Google Scholar]
  3. Bradshaw Z, Slade P. The effects of induced abortion on emotional experiences and relationships: A critical review of the literature. Clinical Psychology Review. 2003;23:929–958. doi: 10.1016/j.cpr.2003.09.001. [DOI] [PubMed] [Google Scholar]
  4. Cairney J, Krause N. The social distribution of psychological distress and depression in older adults. Journal of Aging and Health. 2005;17:807–835. doi: 10.1177/0898264305280985. [DOI] [PubMed] [Google Scholar]
  5. Charles VE, Polis CB, Sridhara SK, Blum RW. Abortion and long-term mental health outcomes: A systematic review of the evidence. Contraception. 2008;78:436–450. doi: 10.1016/j.contraception.2008.07.005. [DOI] [PubMed] [Google Scholar]
  6. Cougle JR, Reardon DC, Coleman PK. Depression associated with abortion and childbirth: a long-term analysis of the NLSY cohort. Medical Science Monitor. 2003;9(4):CR105–CR112. [PubMed] [Google Scholar]
  7. Cougle JR, Reardon DC, Coleman PK. Generalized anxiety following unintended pregnancies resolved through childbirth and abortion: a cohort study of the 1995 National Survey of Family Growth. Journal of Anxiety Disorders. 2005;19:137–142. doi: 10.1016/j.janxdis.2003.12.003. [DOI] [PubMed] [Google Scholar]
  8. Coleman PK. Resolution of unwanted pregnancy during adolescence through abortion versus childbirth: Individual and family predictors and psychological consequences. Journal of Youth and Adolescence. 2006;35:903–911. [Google Scholar]
  9. Coleman PK, Coyle CT, Shuping M, Rue VM. Induced abortion and anxiety, mood, and substance abuse disorders: isolating the effects of abortion in the national comorbidity survey. Journal of Psychiatric Research. 2009;43:770–776. doi: 10.1016/j.jpsychires.2008.10.009. [DOI] [PubMed] [Google Scholar]
  10. Coleman PK, Reardon DC, Cougle JR. Substance use among pregnant women in the context of previous reproductive loss and desire for current pregnancy. British Journal of Health Psychology. 2005;10:255–268. doi: 10.1348/135910705X25499. [DOI] [PubMed] [Google Scholar]
  11. Coleman PK, Reardon DC, Rue VM, Cougle J. A history of induced abortion in relation to substance use during subsequent pregnancies carried to term. American Journal of Obstetrics & Gynecology. 2002a;187:1673–1678. doi: 10.1067/mob.2002.127602. [DOI] [PubMed] [Google Scholar]
  12. Coleman PK, Reardon DC, Rue VM, Cougle J. State-funded abortion versus deliveries: a comparison of outpatient mental health claims over 4 years. American Journal of Orthopsychiatry. 2002b;72:141–152. doi: 10.1037/0002-9432.72.1.1410155. [DOI] [PubMed] [Google Scholar]
  13. Coyne JC, Downey G. Social factors and psychopathology: Stress, social support, and coping processes. Annual Review of Psychology. 1991;42:401–425. doi: 10.1146/annurev.ps.42.020191.002153. [DOI] [PubMed] [Google Scholar]
  14. Dietz PM, Spitz AM, Anda RF, Williamson DF, McMahon PM, Santelli JS, Kendrick JS. Unintended pregnancy among adult women exposed to abuse or household dysfunction during their childhood. Journal of the American Medical Association. 1999;282:1259–1264. doi: 10.1001/jama.282.14.1359. [DOI] [PubMed] [Google Scholar]
  15. Fergusson DM, Horwood LJ, Boden JM. Abortion and mental health disorders: evidence from a 30-year longitudinal study. British Journal of Psychiatry. 2008;193:444–451. doi: 10.1192/bjp.bp.108.056499. [DOI] [PubMed] [Google Scholar]
  16. Fergusson DM, Horwood LJ, Ridder EM. Abortion in young women and subsequent mental health. Journal of Child Psychology and Psychiatry. 2006;47:16–24. doi: 10.1111/j.1469-7610.2005.01538.x. [DOI] [PubMed] [Google Scholar]
  17. Finer LB, Henshaw SK. Disparities in rates of unintended pregnancy in the United States, 1994 and 2001. Perspective of Sexual and Reproductive Health. 2006;38:90–96. doi: 10.1363/psrh.38.090.06. [DOI] [PubMed] [Google Scholar]
  18. Gazmararian JA, Adams MA, Saltzman LE, Johnson CH, Bruce C, Marks JS. The Prams Working Group. The relationship between pregnancy intendedness and physical violence in mothers of newborns. Obstetrics & Gynecology. 1995;85:1031–1038. doi: 10.1016/0029-7844(95)00057-x. [DOI] [PubMed] [Google Scholar]
  19. Gazmararian JA, Petersen R, Spitz AM, Goodwin MM, Saltzman LE, Marks JS. Violence and reproductive health: current knowledge and future directions. Maternal Child Health Journal. 2000;4:79–84. doi: 10.1023/a:1009514119423. [DOI] [PubMed] [Google Scholar]
  20. Gissler M, Berg C, Bouvier-Colle MH, Buekens P. Injury deaths, suicides and homicides associated with pregnancy. Finland, 1987-2000. European Journal of Public Health. 2005;15:459–463. doi: 10.1093/eurpub/cki042. [DOI] [PubMed] [Google Scholar]
  21. Gissler M, Berg C, Bouvier-Colle MH, Buekens P. Methods for identifying pregnancy-associated deaths: Population-based data from Finland 1987-2000. Paediatric and Perinatal Epidemiology. 2004a;18:448–455. doi: 10.1111/j.1365-3016.2004.00591.x. [DOI] [PubMed] [Google Scholar]
  22. Gissler M, Berg C, Bouvier-Colle MH, Buekens P. Pregnancy-associated mortality after birth, spontaneous abortion, or induced abortion in Finland, 1980-2000. American Journal of Obstetrics and Gynecology. 2004b;190:422–427. doi: 10.1016/j.ajog.2003.08.044. [DOI] [PubMed] [Google Scholar]
  23. Gissler M, Hemminki E, Lonnqvist J. Suicides after pregnancy in Finland, 1987-94: Register linkage study. British Medical Journal. 1996;313:1431–1434. doi: 10.1136/bmj.313.7070.1431. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Gissler M, Kauppila R, Merilainen J, Toukomaa H, Hemminki E. Pregnancy-associated deaths in Finland 1987-994—Definition problems and benefits of record linkage. Acta Obstetrica et Gynecologica Scandinavica. 1997;76:651–657. doi: 10.3109/00016349709024605. [DOI] [PubMed] [Google Scholar]
  25. Gold RB, Nash N. State abortion counseling policies and the fundamental principles of informed consent. Guttmacher Policy Review. 2007;10(4):6–13. [Google Scholar]
  26. Carhart Gonzales v. 550 U S. 2007;380 [Google Scholar]
  27. Goodwin MM, Gazmararian JA, Johnson CH, Gilbert BC, Saltzman LE. Pregnancy intendedness and physical abuse around the time of pregnancy: findings from the pregnancy risk assessment monitoring system, 1996-1997. PRAMS Working Group. Pregnancy Risk Assessment Monitoring System. Maternal Child Health Journal. 2000;4:85–92. doi: 10.1023/a:1009566103493. [DOI] [PubMed] [Google Scholar]
  28. Henshaw SK. Personal Communication 2009 [Google Scholar]
  29. Jagannathan R. Relying on surveys to understand abortion behavior: Some cautionary Evidence. American Journal of Public Health. 2001;91:1825–1831. doi: 10.2105/ajph.91.11.1825. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Jones EF, Forrest JD. Underreporting of abortion in surveys of US. Women: 1976 to 1988. Demography. 1992;29:113–126. [PubMed] [Google Scholar]
  31. Jones RK, Darroch JE, Henshaw SK. Patterns of socioeconomic characteristics of women obtaining abortions in 2000-2001. Perspectives on Sexual and Reproductive Health. 2002;34(5):226–235. [PubMed] [Google Scholar]
  32. Jones RK, Kost K. Underreporting of induced and spontaneous abortion in the United States: An analysis of the 2002 National Survey of Family Growth. Studies in Family Planning. 2007;38:187–197. doi: 10.1111/j.1728-4465.2007.00130.x. [DOI] [PubMed] [Google Scholar]
  33. Kessler RC, McGonagle DA, Zhao S, Nelson CB, Hughes M, Eshelman S, Kendler K. Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Archives of General Psychiatry. 1994;51:8–19. doi: 10.1001/archpsyc.1994.03950010008002. [DOI] [PubMed] [Google Scholar]
  34. Link BG, Phelan J. Social conditions as fundamental causes of disease. Journal of Health and Social Behavior, Spec. 1995:80–94. [PubMed] [Google Scholar]
  35. Neumann DA, Houskamp BM, Pollock VE, Briere J. The long-term sequelae of childhood sexual abuse in women: A meta-analytic review. Child Maltreatment. 1996;1:6–16. [Google Scholar]
  36. Pedersen W. Abortion and depression: A population-based longitudinal study of young women. Scandinavian Journal of Public Health. 2008;36:424–428. doi: 10.1177/1403494807088449. [DOI] [PubMed] [Google Scholar]
  37. Pope LM, Adler NE, Tschann JM. Postabortion psychological adjustment: Are minors at increased risk? Journal of Adolescent Health. 2001;29:2–11. doi: 10.1016/s1054-139x(01)00212-9. [DOI] [PubMed] [Google Scholar]
  38. Reardon DC, Coleman PK. Relative treatment rates for sleep disorders and sleep disturbances following abortion and childbirth: a prospective record-based study. Sleep. 2006;29:105–106. doi: 10.1093/sleep/29.1.105. [DOI] [PubMed] [Google Scholar]
  39. Reardon DC, Coleman PK, Cougle JR. Substance use associated with unintended pregnancy outcomes in the National Longitudinal Survey of Youth. American Journal of Drug and Alcohol Abuse. 2004;30:369–383. doi: 10.1081/ada-120037383. [DOI] [PubMed] [Google Scholar]
  40. Reardon DC, Cougle JR. Depression and unintended pregnancy in the National Longitudinal Survey of Youth: a cohort study. British Medical Journal. 2002;19:151–152. doi: 10.1136/bmj.324.7330.151. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Reardon DC, Cougle JR, Rue VM, Shuping MW, Coleman PK, Ney PG. Psychiatric admissions of low-income women following abortion and childbirth. Canadian Medical Association Journal. 2003;168:1253–1256. [PMC free article] [PubMed] [Google Scholar]
  42. Reardon DC, Ney PG, Scheuren F, Cougle J, Coleman PK, Strahan TW. Deaths associated with pregnancy outcome: a record linkage study of low income. Southern Medical Journal. 2002;95:834–41. [PubMed] [Google Scholar]
  43. Rosenberg M. Society and the Adolescent Self-Image. Princeton, NJ: Princeton University Press; 1965. [Google Scholar]
  44. Russo N, Dabul AJ. The relationship of abortion to well-being: Do race and religion make a difference? Professional Psychology: Research and Practice. 1997;28:23–31. doi: 10.1037/0735-7028.28.1.23. [DOI] [PubMed] [Google Scholar]
  45. Russo NF, Denious JE. Understanding the relationship of violence against women to unwanted pregnancy and its resolution. In: Beckman LJ, Harvey MS, editors. The New Civil War: The Psychology, Culture, and Politics of Abortion Psych of Women book series. Washington, DC: American Psychological Association; 1998. pp. 25–59. [Google Scholar]
  46. Russo NF, Denious JE. Violence in the lives of women having abortions: Implications for practice and public policy. Professional Psychology: Research and Practice Special Issue: Innovative interventions in the practice of health psychology. 2001;32:142–50. [Google Scholar]
  47. Russo NF, Zierk KL. Abortion, childbearing, and women’s well-being. Professional Psychology: Research and Practice. 1991;23:269–280. doi: 10.1037/0735-7028.28.1.23. [DOI] [PubMed] [Google Scholar]
  48. Schmiege S, Russo NF. Depression and unwanted first pregnancy: longitudinal cohort study. British Medical Journal. 2005;331:1303–1307. doi: 10.1136/bmj.38623.532384.55. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Siegel RB. Dignity and the politics of protection: Abortion restrictions under. Casey/Carhart The Yale Law Journal. 2008;117:1692–1800. [Google Scholar]
  50. Siegel RB. The right’s reasons: Constitutional conflict and the spread of woman-protective antiabortion argument. Duke Law Journal. in press. [PubMed] [Google Scholar]
  51. Smith LB, Adler NE, Tschann JM. Underreporting sensitive behaviors: The case of young women’s willingness to report abortion. Health Psychology. 1999;18:37–43. doi: 10.1037//0278-6133.18.1.37. [DOI] [PubMed] [Google Scholar]
  52. South Dakota. Codified Laws, § 34-23A-10.1. 2006 Retrieved June 1, 2009 from, http://legis.state.sd.us/statutes/DisplayStatute.aspx?Type=Statute&Statute=34-23A-10.1.
  53. Steinberg JR, Russo NF. Abortion and Anxiety: What’s the relationship? Social Science & Medicine. 2008;67:238–52. doi: 10.1016/j.socscimed.2008.03.033. [DOI] [PubMed] [Google Scholar]
  54. Steinberg JR, Russo NF. Evaluating research on abortion and mental health. Contraception. doi: 10.1016/j.contraception.2009.06.003. in press. [DOI] [PubMed] [Google Scholar]
  55. Taft AJ, Watson LF. Depression and termination of pregnancy (induced abortion) in a national cohort of young Australian women: the confounding effect of women’s experience of violence. BMC Public Health. 2008;8:75–82. doi: 10.1186/1471-2458-8-75. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Taylor SE, Stanton AL. Coping resources, coping processes, and mental health. Annual Review of Clinical Psychology. 2007;3:377–401. doi: 10.1146/annurev.clinpsy.3.022806.091520. [DOI] [PubMed] [Google Scholar]
  57. Turner RJ, Lloyd DA, Roszell P. Personal Resources and the social distribution of depression. American Journal of Community Psychology. 1999;27:643–672. doi: 10.1023/A:1022189904602. [DOI] [PubMed] [Google Scholar]
  58. Udry R, Guaghan M, Schwingl PJ, van den Berg BJ. A medical record linkage analysis of abortion underreporting. Family Planning Perspectives. 1996;28:228–231. [PubMed] [Google Scholar]
  59. Wilmoth GH. Abortion, public health policy, and informed consent legislation. Journal of Social Issues. 1991;48:1–17. doi: 10.1111/j.1540-4560.1992.tb00895.x. [DOI] [PubMed] [Google Scholar]
  60. Zhao S, Kessler RC, Wittchen HU. Diagnostic Algorithms for NCS/DSM-III-R Di NCS Working Paper # 7. 1994. Unpublished manuscript. [Google Scholar]

RESOURCES