When evaluating the safety of any medical intervention, it is important not only to consider the immediate potential complications but also to evaluate the potential for associated long-term health effects. Research on abortion’s potential long-term health consequences has focused on reproductive and mental health outcomes, as well as other outcomes including breast cancer risk and premature death. This chapter reviews research on the long-term physical and mental health effects of having an abortion.1 The focus is on four putative areas of potential harm:
- future childbearing and pregnancy outcomes (e.g., secondary infertility; ectopic pregnancy; spontaneous abortion and stillbirth; complications of pregnancy; and preterm birth, small for gestational age, and low birthweight);
- risk of breast cancer;
- mental health disorders; and
- premature death.
1 This chapter reviews the epidemiological research on abortion’s long-term physical and mental health effects. In epidemiology, an odds ratio is the statistic used by researchers to measure the association between an “exposure” (e.g., a prior abortion) and an outcome of interest. Odds ratios compare the relative odds of a particular health outcome, given the exposure, and can indicate whether the exposure is a risk factor, as well as the magnitude of the risk. The confidence interval (CI) indicates the precision of the estimate.
As noted in Chapter 1, some states require that abortion patients be offered or provided information indicating that abortion negatively affects future fertility (Arizona, Kansas, Nebraska, North Carolina, South Dakota, and Texas); risk of breast cancer (Arkansas, Kansas, Mississippi, Oklahoma, and Texas); and/or mental health disorders (Idaho, Kansas, Louisiana, Michigan, Nebraska, North Carolina, North Dakota, Oklahoma, South Dakota, Texas, Utah, and West Virginia).2 Some observers have also questioned whether abortion may lead to premature death (Reardon et al., 2002).
This chapter first describes the limitations of the abortion literature and the committee’s criteria for identifying scientifically valid research on abortion’s long-term health consequences. The remainder of the chapter presents the committee’s review of what is known about each of the above putative harms of abortion.
While randomized controlled trials are the gold standard research design for assessing the health effects of a medical intervention, they are not appropriate for studies assessing the long-term risks of abortion. Women seeking an abortion cannot be randomized to an experimental group that has the abortion or a control group that does not have the abortion. Researchers must use observational study designs (e.g., cohort, case control, and cross-sectional studies) to examine abortion’s long-term potential for harm. However, the risk of bias is greater for observational studies than for randomized studies, and it is imperative that published studies be assessed for potential sources of bias that might affect their findings (IOM, 2011). In research, bias refers to systematic error in a study’s design or execution that leads to an incorrect result (Cochrane Collaboration, 2018).
Sources of Bias
Observational studies of abortion’s long-term health effects have two important sources of information bias: selective recall bias and selection bias. Each is described below.
Selective Recall Bias
Several studies have demonstrated underreporting of past abortions in surveys of American women (Beral et al., 2004; Jones and Kost, 2007; Steinberg et al., 2011). Women who have had an abortion have a tendency
not to recall—or not to report—having had an abortion when asked (Anderson et al., 1994; Bouyer et al., 2003; Hogue, 1975; Jones and Kost, 2007; Lindefors-Harris et al., 1991). Jones and Kost (2007) pooled data from the 1997 to 2001 annual National Survey of Family Growth to assess the accuracy of women’s reports of the number of their past pregnancies and the timing and outcome of each pregnancy. Data were collected from face-to-face interviews and computer-assisted, self-administered questionnaires. Tallies of the survey responses were compared with national estimates of the number of abortions performed during the time period. Overall, the number of self-reported abortions was only 47 percent of the total estimated number of abortions performed in the United States during the study period. Inconsistencies also were seen between women’s responses during the interviews and on the self-administered questionnaire.
The committee found that selective recall bias—which occurs when there are systematic differences in study subjects’ reporting of a past exposure (e.g., a prior abortion)—affects much of the published research on abortion’s long-term health effects.3 In case control studies, for example, subjects who had a particular negative health outcome may be more likely to report having had the past exposure compared with subjects who also had the exposure but did not have the outcome. As a consequence, healthy subjects with an exposure are more likely than unhealthy subjects to be mistakenly assigned to a nonexposure study group. When this occurs, a study is likely to conclude erroneously that the negative health outcome is associated with the exposure. Many published abortion studies are flawed in this way because women who have had an abortion are less likely to report their prior abortion if they have not experienced a long-term adverse health outcome. Such selective recall bias was first documented in a 1975 study of low-birthweight births among women who had had an abortion (Hogue, 1975) and has been demonstrated in other studies of abortion’s long-term effects (Anderson et al., 1994; Bouyer et al., 2003; Lindefors-Harris et al., 1991).
Recall bias is best addressed by using registry or medical record data to document prior abortions and link the abortion histories with reliable records of subsequent patient outcomes.
3 Selective (or differential) recall bias has a different effect on study results relative to general (or nondifferential) recall bias. Nondifferential recall bias occurs when all study subjects are equally likely to misrecall an exposure. When this happens, study results are likely to under-report the association between the exposure and the health outcome (Alexander et al., 2015). In studies of abortion recall and health outcomes, only differential bias has been documented. This suggests that studies that rely on women’s reports of abortion and that find no association with the condition under study might be used to conclude that there is evidence of no true association. However, the committee did not utilize this assumption, but rather relied exclusively on studies that document the subjects’ abortion history.
Selection Bias (Comparability of Study Populations)
Selection bias occurs when one of the baseline characteristics of the study population is associated with the outcome of interest (e.g., future pregnancy and birth complications after abortion). In such circumstances, the statistical analysis should take into account (or “control for”) the differences in the prevalence of the confounding factors in the study and control groups. If the objective of a study is to determine whether having an abortion raises the risk of future mental health problems, for example, the study should control for women’s mental health status at baseline (i.e., before they have the abortion). This is particularly important in mental health research because women who have an abortion report higher rates of mental health disorders before undergoing the procedure compared with women who give birth (Steinberg et al., 2014). Other confounding variables that may affect future health and pregnancy outcomes include socioeconomic status, race and ethnicity, smoking, and substance use. Much of the research on abortion’s long-term effects has been conducted outside the United States, and a substantial volume of literature is based on abortion care in countries where such factors as socioeconomic conditions, culture, population health, health care resources, and/or the health care system are markedly different from those in the United States. In addition to the other selection criteria listed below, the committee determined the applicability of published research based on the likelihood that the abortion interventions examined reflected contemporary abortion care in the United States (e.g., in European countries). Although the committee identified studies of long-term health effects in Africa, China, India, Taiwan, and Vietnam, these studies were excluded.
The Committee’s Selection Criteria
The committee’s literature search strategy for this chapter is provided in Appendix D. The bibliographies of retrieved articles were reviewed to find additional relevant research. Each identified article was reviewed to determine whether the study met the criteria listed below. The findings reported in this chapter draw solely on studies that met these criteria:
- for the study population, there was objective medical record or patient registry documentation of a prior induced abortion (excluding spontaneous abortion or miscarriage);
- the study population (women with a documented abortion) was compared with a control group of women with no documented abortion history;
- the analysis controlled for mental health status prior to the abortion (if assessing the mental health effects of abortion);
- the study was published in 2000 or later and included abortions performed in 1980 or later (to help ensure that reported outcomes reflected contemporary abortion methods); and
- the clinical settings and care delivery were similar to those in the United States.
Many women are likely to desire and experience a future pregnancy after having had an abortion. Abortion has been investigated for its potential effect on secondary infertility; ectopic pregnancy; spontaneous abortion and stillbirth; pregnancy complications that can lead to adverse maternal or fetal health; and preterm birth, low birthweight, and/or weight that is small for gestational age. This section reviews this research.
Does Abortion Increase the Risk of Secondary Infertility?
Secondary infertility is defined as difficulty conceiving a pregnancy or carrying a pregnancy to term after a previous pregnancy. The committee found one study on abortion and risk of secondary infertility that met its selection criteria. Holmlund and colleagues (2016) examined the pregnancy-related outcomes of 57,406 first-time mothers in Finland who gave birth between 2008 and 2010; 5,167 of the women had had a prior abortion. Using national registry data, the researchers linked the mothers’ birth records with records documenting a prior abortion and/or subsequent treatment for infertility. First-time mothers with a prior abortion were significantly less likely to be treated for infertility compared with women in their first pregnancy (1.95 versus 5.14 percent, p <.0001), thus suggesting that there is no association between abortion and secondary infertility.
Does Abortion Increase the Risk of Ectopic Pregnancy?
An ectopic pregnancy occurs when a fertilized egg grows outside of the uterus, most commonly in a fallopian tube. As the pregnancy progresses, the fallopian tube may rupture, causing major internal bleeding (ACOG, 2017). In 2013, an estimated 0.68 percent of commercially insured pregnant women and 0.57 percent of Medicaid-insured women in the United States were diagnosed and treated for an ectopic pregnancy (Tao et al., 2017). Women with a history of upper genital tract infection (e.g., in the
uterus or fallopian tubes) are at increased risk of an ectopic pregnancy (Sivalingam et al., 2011). As noted in Chapter 2, serious infection after an abortion is rare and has become even rarer since antibiotic prophylaxis became standard practice. If untreated, an abortion-related infection may increase the risk of subsequent ectopic pregnancy.
While several literature reviews have concluded that abortion is not associated with increased risk of ectopic pregnancy (Lowit et al., 2010; RCOG, 2011b; Thorp et al., 2003), all the published reviews are methodologically flawed because they include studies based on maternal recall and/or rely heavily on studies of abortions performed before the introduction of contemporary abortion methods. The committee could identify no primary literature without these limitations.
Spontaneous Abortion (Miscarriage) and Stillbirth
Does Abortion Increase the Risk of Spontaneous Abortion and Stillbirth?
Spontaneous abortion, also referred to as miscarriage, is the spontaneous death of a fetus prior to 20 weeks’ gestation. Stillbirth refers to spontaneous fetal death after 20 weeks’ gestation (ODPHP, 2017). Several literature reviews have concluded that abortion is not associated with increased risk of either spontaneous abortion or stillbirth (Lowit et al., 2010; RCOG, 2011b; Thorp et al., 2003). However, as in the published reviews of abortion and subsequent ectopic pregnancy, these literature reviews include studies of abortions performed with outdated methods and are methodologically flawed by reliance on studies based on maternal recall (rather than objective documentation of an abortion). The committee could identify no relevant primary literature without these limitations, and thus was unable to draw a conclusion regarding the association between abortion and risk of spontaneous abortion and stillbirth.
The committee identified three primary research studies4 that used documented records of receipt of an abortion to assess the effect of abortion
4 The committee identified a fourth study that used Danish registry data to assess the effects of abortions performed from 1980 to 1982 on subsequent pregnancy complications. This study, by Zhou and colleagues (2001), was excluded from the committee’s review because it is unlikely to reflect the outcomes of contemporary abortion methods. The authors report that almost all (99.7 percent) of the abortion procedures included in the study were followed by curettage. As noted in Chapter 2, sharp-metal curettage is no longer recommended because it is associated with risk of injury (NAF, 2017; RCOG, 2011b, 2015; Roblin, 2014; SFP, 2013; WHO, 2012).
on the risk of complications in a future pregnancy. The study findings are summarized below. Study details, including adjusted odds ratios (aORs), sample sizes, and the years in which the abortions occurred, are provided in Table 4-1.
In a retrospective cohort study, Jackson and colleagues (2007) linked medical records and obstetrics databases in two Chicago-area hospitals to compare the pregnancy outcomes of women who had and had not previously received a dilation and evacuation (D&E) abortion. Holmlund and colleagues (2016) linked Finnish birth and abortion registry data to compare the first full-term pregnancies of women with and without a prior abortion. Finally, Woolner and colleagues (2014) used Scottish registry data to compare the risk of preterm delivery and other birth outcomes among women with and without a prior abortion. This latter analysis has a number of strengths not characteristic of most of the available research on abortion and subsequent birth outcomes. The researchers had a large enough sample and sufficient data to control for maternal age, socioeconomic variables, weeks’ gestation (≤13 weeks versus >13 weeks), and smoking, as well as to stratify the sample by type of abortion (e.g., medication versus aspiration). The study population groups included 3,186 women with a documented termination of their first pregnancy and 42,446 primigravid women.5
Does Abortion Increase the Risk of Hypertension of Pregnancy?
Hypertension of pregnancy includes preeclampsia and chronic and gestational hypertension. It is associated with increased risk of both maternal complications, such as placental abruption and gestational diabetes, and poor birth outcomes, such as preterm birth, having a baby that is small for gestational age, and infant death (CDC, 2016). Hypertension complicates about 5 to 10 percent of pregnancies (Garovic and August, 2013). The Woolner et al. (2014) and Holmlund et al. (2014) studies described above compared the subsequent pregnancies of women who had had prior abortions and women in their first pregnancy (without prior abortions) and found no increased risk of hypertension in pregnancy or preeclampsia among the women with prior abortions. In the Woolner et al. (2014) study, women who had had an abortion had a lower risk of hypertensive disease relative to women in their first pregnancy. This finding persisted when the outcomes were analyzed by type of abortion (i.e., medication and aspiration). The authors also found that the timing of abortion had no impact on hypertensive disease. Women who had undergone both early and late abortions had a lower risk of hypertension in pregnancy compared with women
5 “Primigravid” refers to women who are pregnant for the first time.
|Author (year)||Location (time period)||Abortion Group||Comparison Group|
|Holmlund et al. (2016)||Finland (1983–2007)||All methods n = 5,167||Primigravid n = 52,239|
|Woolner et al. (2014)||Scotland (1986–2010)||All methods n = 3,186||Primigravid n = 42,446|
|Aspiration n = 1,800||Primigravid n = 42,446|
|Medication n = 1,385||Primigravid n = 42,446|
|All methods *13 weeks n = 431||Primigravid n = 42,446|
|All methods <13 weeks n = 2,315||Primigravid n = 42,446|
|Jackson et al. (2007)||United States (1995–2003)||D&E 12–24 weeks n = 85||Women with no history of midtrimester abortion, matched by age n = 170|
NOTES: The term “primigravid” refers to a first pregnancy.
a All methods n = 2,497; primigravid n = 33,520.
b All methods n = 689; primigravid n = 8,916.
c Aspiration n = 1,421; primigravid n = 33,520.
d Aspiration n = 367; primigravid n = 8,916.
e Medication n = 1,064; primigravid n = 33,520.
|Adjusted Odds Ratios (Confidence Intervals)|
|Hypertension||Antepartum Hemorrhage||Postpartum Hemorrhage||Notes|
|1.07 (0.92–1.24)||n/a||n/a||aOR (95% CI); control variables not reported|
|0.69 (0.61–0.78)||1.26 (1.10–1.45)||Vaginal deliverya 1.14 (0.97–1.33) Cesarean sectionb 1.01 (0.74–1.36)||aOR (99% CI) adjusted for maternal age at delivery, smoking, and social class|
|0.73 (0.62–0.85)||1.33 (1.11–1.59)||Vaginal deliveryc 0.88 (0.71–1.11) Cesarean sectiond 0.96 (0.64–1.45)|
|0.63 (0.52–0.76)||1.18 (0.95–1.46)||Vaginal deliverye 1.49 (1.21–1.85) Cesarean sectionf 1.06 (0.69–1.62)|
|0.83 (0.71–0.98)||1.02 (0.84–1.25)||Vaginal deliveryg 1.05 (0.85–1.29) Cesarean sectionh 0.94 (0.60–1.45)|
|0.64 (0.55–0.75)||1.29 (1.10–1.51)||Vaginal deliveryi 1.09 (0.90–1.31) Cesarean sectionj 0.94 (0.66–1.35)|
|n/a||n/a||n/a||Odds ratios not reported; outcomes reported as percentages (abortion versus control)|
|Abnormal placentation: 4.8% versus 2.4% (p = .310)|
|Hemorrhage: 2.3% versus 2.3% (p = 1.0)|
f Medication n = 321; primigravid n = 8,916.
g All methods ≥13 weeks n = 344; primigravid n = 33,520.
h All methods ≥13 weeks n = 87; primigravid n = 8,916.
i All methods <13 weeks n = 1,818; primigravid n = 33,520.
j All methods <13 weeks n = 497; primigravid n = 8,916.
in their first pregnancy (aOR = 0.64; 99% CI = 0.55–0.75 and aOR = 0.83; 99% CI = 0.71–0.98, respectively).
Does Abortion Increase the Risk of Complications of the Placenta?
Abnormal placentation, including placenta previa and accreta, is associated with maternal hemorrhage requiring transfusion (Saleh, 2008). The committee identified one study that assessed the risk of placenta complications in a full-term pregnancy following an abortion. As noted earlier, Jackson and colleagues (2007) compared the pregnancy outcomes of women with and without a prior D&E in two Chicago-area hospitals (Jackson et al., 2007). The authors found no association between D&E and abnormal placentation.
Does Abortion Increase the Risk of Hemorrhage in Subsequent Pregnancy?
Two types of hemorrhaging may occur during pregnancy and childbirth: antepartum and postpartum. Antepartum hemorrhage—bleeding from or into the genital tract—affects from 3 to 5 percent of pregnancies globally (Lange and Toledo, 2017). Its causes vary with weeks’ gestation. Before 20 weeks, bleeding may be due to abnormal embryo implantation, miscarriage, ectopic pregnancy, gestational trophoblastic disease, and benign and malignant tumors of the reproductive tract. After 20 weeks’ gestation (and before birth), the most common causes of antepartum bleeding are cervical change due to preterm labor and disorders of the placenta. Postpartum hemorrhage is usually defined as the loss of 500 ml or more of blood from the genital tract within 24 hours of childbirth (RCOG, 2017); in 2006, it affected an estimated 2.9 percent of pregnancies in the United States (Callaghan et al., 2010). The most common cause (79 percent) is uterine atony, the failure of the uterus to contract following delivery (Bateman et al., 2010).
The Woolner et al. (2014) registry study contains the only analysis the committee could identify on abortion’s association with hemorrhage in subsequent pregnancy (Woolner et al., 2014). The study found that prior aspiration abortion was associated with a higher risk of antepartum hemorrhage (aOR = 1.33; 99% CI = 1.11–1.59). However, unlike the other outcomes the researchers investigated (e.g., see the above review of hypertensive disorders and preterm birth), antepartum hemorrhage was not clearly defined, and the finding of higher risk does not appear to be clinically significant as there was no association with preterm birth or hospitalization. In contrast, Woolner and colleagues (2014) clearly defined postpartum hemorrhage as >500 ml blood loss for vaginal delivery or >1,000 ml blood loss for cesarean section. The risk of postpartum hemorrhage during a vaginal
delivery was higher among women who had had a medication abortion (but not an aspiration abortion) compared with women in their first pregnancy (aOR = 1.49; 99% CI = 1.21–1.85). It is unclear how a medication abortion might lead to postpartum hemorrhage in a later pregnancy. There were no differences for cesarean section.
Preterm Birth, Small for Gestational Age, and Low Birthweight
Preterm birth (birth at <37 weeks’ gestation) and low birthweight are related. Infants born prematurely are more likely than full-term infants to weigh <2,500 g (the definition of low birthweight)—although full-term infants that are small for gestational age may also weigh <2,500 g, and low-birthweight and preterm infants can be the appropriate weight for their gestation. Each of these adverse outcomes may have different underlying causes. The risks to neonatal survival and development vary at different birthweights and weeks’ gestation. Thus, investigators often examine both gestation at delivery and fetal growth to determine whether a prior abortion is a risk factor for poor pregnancy outcomes in the future.
First births are at greater risk of preterm delivery than are subsequent births (Ananth et al., 2001). Thus, if the risk of preterm delivery is compared for women who have had an abortion and those who have not, the findings are likely to be biased if they do not take into account the increased risk of preterm delivery for first births (Hogue et al., 1982). To address this source of bias and ensure the comparability of study populations, the committee limited its review to studies of preterm birth that compare the outcomes of a first birth or control for the number of previous births. In the following discussion, the extent to which studies adjust for other confounding variables (e.g., smoking, maternal age, the provision of prophylactic antibiotics at the time of abortion, the type of abortion method, weeks’ gestation, and the number of prior abortions) is noted.
The committee identified five studies that met its criteria for assessing the association of abortion with birth outcomes (see Table 4-2). These include the Woolner et al. (2014) and Jackson et al. (2007) studies described above and three studies that used linked Finnish medical records during different but overlapping time periods (KC et al., 2017b; Klemetti et al., 2012; Mannisto et al., 2017). The findings from these studies are presented below. See Table 4-2 for further details on study designs and results.
Do Early-Gestation Aspiration or Medication Abortions Increase the Risk of Preterm Birth?
The Woolner et al. (2014) study is the only available reliable analysis (meeting the committee’s criteria) of the association between abortion
|Authors (year)||Location (abortion period)||Abortion Group||Comparator||Adjusted Odds Ratios (Confidence Intervals)|
|KC et al. (2017b)||Finland (1983–2013)||1 medication abortion n = 12,183||No previous abortion n = 365,356||Very PTB 0.92 (0.70–1.21) Extremely PTB 0.96 (0.68–1.35)|
|1 aspiration abortion n = 33,840||No previous abortion n = 365,356||Very PTB 1.01 (0.87–1.18) Extremely PTB 1.11 (0.92–1.33)|
|>1 medication abortion n = 1,267||No previous abortion n = 365,356||Very PTB 0.50 (0.16–1.55) Extremely PTB 0.82 (0.26–2.58)|
|>1 aspiration abortion n = 4,819||No previous abortion n = 365,356||Very PTB 1.14 (0.81–1.59) Extremely PTB 1.51 (1.03–2.23)|
|Mannisto et al. (2017)||Finland (2000–2009)||0 to <6 months interpregnancy interval n = 2,956||18 to <24 months interpregnancy interval n = 2,076||n/a|
|6 to <12 months interpregnancy interval n = 3,203||18 to <24 months interpregnancy interval n = 2,076||n/a|
|12 to <18 months interpregnancy interval n = 2,623||18 to <24 months interpregnancy interval n = 2,076||n/a|
|≥24 months interpregnancy interval n = 9,036||18 to <24 months interpregnancy interval n = 2,076||n/a|
|Adjusted Odds Ratios (Confidence Intervals)|
|All PTB <37 Weeks||Spontaneous PTB <37 Weeks||LBW or SGA||Notes|
|0.85 (0.77–0.93)||n/a||SGA 1.05 (0.97–1.15)||aOR (95% CI) adjusted for maternal age, marital status, maternal smoking, maternal residence of municipality, and birth year of child.|
|0.99 (0.94–1.05)||n/a||SGA 1.07 (1.02–1.13)|
|0.70 (0.51–0.96)||n/a||SGA 1.05 (0.82–1.34)|
|1.00 (0.88–1.14)||n/a||SGA 1.07 (0.95–1.21)|
|1.35 (1.02–1.77)||n/a||LBW 1.22 (0.90–1.66)
SGA 1.15 (0.80–1.66)
|aOR (95% CI) adjusted for parity, prepregnancy body mass index (BMI), cohabitation, type of residence, socioeconomic status, maternal age, smoking, type of termination of pregnancy, and gestational age at termination of pregnancy.|
|1.14 (0.86–1.50)||n/a||LBW 1.10 (0.82–1.49)
SGA 1.21 (0.85–1.71)
|0.94 (0.70–1.26)||n/a||LBW 0.87 (0.62–1.20)
SGA 0.86 (0.58–1.26)
|1.19 (0.93–1.53)||n/a||LBW 1.06 (0.81–1.39)
SGA 0.89 (0.65–1.22)
|Authors (year)||Location (abortion period)||Abortion Group||Comparator||Adjusted Odds Ratios (Confidence Intervals)|
|Woolner et al. (2014)||Scotland (1986–2010)||All methods n = 3,186||No previous pregnancy n = 42,446||0.81 (0.54–1.21)|
|Aspiration n = 1,800||No previous pregnancy n = 42,446||0.73 (0.42–1.26)|
|Medication n = 1,385||No previous pregnancy n = 42,446||0.91 (0.52–1.60)|
|All methods <13 weeks n = 2,315||No previous pregnancy n = 42,446||0.84 (0.53–1.33)|
|All methods ≥13 weeks n = 431||No previous pregnancy n = 42,446||1.02 (0.64–1.62)|
|Klemetti et al. (2012)||Finland (1983–2008)||1 previous abortion n = 31,083||No previous abortion n = 264,190||1.19 (0.98–1.44)|
|2 previous abortions n = 4,417||No previous abortion n = 264,190||1.69 (1.14–2.51)|
|3 or more previous abortions n = 942||No previous abortion n = 264,190||2.78 (1.48–5.24)|
|1 or more previous abortions n = 36,442||No previous abortion n = 264,190||1.27 (1.06–1.52)|
|Adjusted Odds Ratios (Confidence Intervals)|
|All PTB <37 Weeks||Spontaneous PTB <37 Weeks||LBW or SGA||Notes|
|1.05 (0.88–1.26)||1.05 (0.83–1.32)b||LBW 1.14 (0.94–1.39)||aOR (99% CI) adjusted for maternal age at delivery, smoking, and social class.|
|1.10 (0.88–1.39)||1.06 (0.78–1.44)c||LBW 1.08 (0.84–1.38)|
|0.98 (0.75–1.29)||1.03 (0.72–1.46)d||LBW 1.23 (0.92–1.68)|
|1.03 (0.83–1.27)||0.97 (0.73–1.28)e||LBW 1.13 (0.90–1.41)|
|1.13 (0.91–1.40)||1.25 (0.97–1.60)f||LBW 1.01 (0.79–1.29)|
|0.98 (0.93–1.03)||n/a||LBW 0.96 (0.90–1.01) Very LBW 1.03 (0.91–1.18)||aOR (95% CI) adjusted for age, marital status, socioeconomic position, urbanity, smoking, miscarriage, and ectopic pregnancy.|
|1.01 (0.89–1.15)||n/a||LBW 1.02 (0.89–1.18) Very LBW 1.13 (0.84–1.54)|
|1.35 (1.07–1.71)||n/a||LBW 1.43 (1.12–1.84) Very LBW 2.25 (1.43–3.52)|
|0.99 (0.94–1.04)||n/a||LBW 0.98 (0.93–1.03) Very LBW 1.06 (0.94–1.19)|
|Authors (year)||Location (abortion period)||Abortion Group||Comparator||Adjusted Odds Ratios (Confidence Intervals)|
|Jackson et al. (2007)||United States (1995–2003)||D&E 12–24 weeks n = 85||Women with no history of midtrimester abortion, matched by age n = 170||n/a|
NOTES: aOR = adjusted odds radio; CI = confidence interval; LBW = <2,500 g; n/a = not applicable; very LBW = <1,500 g.
a Very PTB is defined as the following in these studies: KC et al. (2017b) defines very PTB as a birth before 32 weeks’ gestations and extremely PTB as a birth before 28 weeks’ gestation; Woolner et al. (2014) defines very PTB as a birth before 33 weeks’ gestation; Klemetti et al. (2012) defines very PTB as a birth occurring before 28 weeks’ gestation.
b All methods n = 2,093; primigravid n = 28,012.
c Aspiration n = 1,187; primigravid n = 28,012.
d Medication n = 906; primigravid n = 28,012.
e All methods <13 weeks n = 1,504; primigravid n = 28,012.
f All methods ≥13 weeks n = 304; primigravid n = 28,012.
g The authors of this study note that this finding was not likely to be clinically significant and may have been confounded by multiple factors (e.g., an unusually low proportion of PTB in the control group), and that controls were lacking for important variables, including race, socioeconomic status, smoking, and prior uterine surgery (including dilation and curettage) (Jackson et al., 2007, Table 1).
|Adjusted Odds Ratios (Confidence Intervals)|
|All PTB <37 Weeks||Spontaneous PTB <37 Weeks||LBW or SGA||Notes|
|n/a||n/a||n/a||Odds ratios not reported; outcomes reported as percentages (abortion versus control):|
|PTB <37 weeks: 9.5% versus 2.9% (p = .025)g|
|PTB <34 weeks: 0.0% versus 0.6% (p = 1.0)|
|Spontaneous PTB <37 weeks: 6.0% versus 2.4% (p = .144)|
method at ≤13 weeks’ gestation and preterm birth in the next pregnancy. As noted above, the authors controlled for smoking history and other potential confounding variables and also stratified the sample to assess differences in the outcomes of women with a prior medication abortion (n = 1,385), a prior aspiration abortion (n = 1,800), or no prior abortion (n = 42,446) both before and after 13 weeks’ gestation. The authors found no statistically significant association between an abortion at <13 weeks’ gestation in the first pregnancy and preterm (aOR = 1.03; 99% CI = 0.83–1.27), spontaneous preterm (aOR = 0.97; 99% CI = 0.73–1.28), or very preterm births (aOR = 0.84; 99% CI = 0.53–1.33) in the next pregnancy.
Do Later-Gestation Abortions Increase the Risk of Preterm Birth?
Woolner and colleagues (2014) found no significant association between a medication or aspiration abortion after 13 weeks’ gestation and a later preterm birth (aOR = 1.13; 99% CI = 0.91–1.40). The small hospital-based study by Jackson and colleagues (2007) compared the birth outcomes of women undergoing D&Es at 12 to 24 weeks’ gestation (n = 85) and women without a history of abortion (n = 170). Controlling for maternal age, multiparity, prior preterm birth, first-trimester dilation and curettage, first-trimester spontaneous delivery, and prior cervical surgery, the researchers found no significant difference in the risk of a later spontaneous preterm birth between the groups of women with a prior D&E abortion and no prior abortion.
Does a Short Interval Between Abortion and Subsequent Pregnancy Increase the Risk of Preterm Birth?
A number of studies indicate that a short interpregnancy interval between live births (conception less than 6 months after the previous pregnancy) may be a risk factor for preterm birth (Smith et al., 2003; Wong et al., 2016). Pregnancies occurring 18 to 23 months after a previous birth have been found to have the lowest risks of preterm births and other adverse events (Ball et al., 2004). The committee identified one study that examined whether a short interpregnancy interval after abortion increases the risk of preterm birth in a subsequent birth. In a Finnish-based study of linked medical records, Mannisto and colleagues (2017) addressed this question by comparing postabortion pregnancies occurring less than 6 months after an abortion with those occurring 18 to 23 months postabortion with respect to risk of preterm birth. They found a slight but significant increase in the estimated risk of preterm birth (aOR = 1.35; 95% CI = 1.02–1.77). This finding is consistent with those from studies focused on other pregnancy outcomes in the index pregnancy (Conde-Agudelo et al., 2012; Shachar et
al., 2016; Wendt et al., 2012), although some have challenged whether the association is causal or related to maternal factors rather than the interval itself (Ball et al., 2014; Hanley et al., 2017; Klebanoff, 2017). If the association between short interpregnancy interval and preterm birth is causal, extending the interval between pregnancies beyond 6 months should reduce the risk of preterm birth associated with shorter intervals. In the case of abortion, effective postabortion contraceptive counseling and use could reduce this concern.
Do Multiple Abortions Increase the Risk of Preterm Birth?
The committee identified two studies (both based in Finland) that examined whether having multiple abortions is associated with a greater risk of preterm birth (KC et al., 2017b; Klemetti et al., 2012). Using 1983–2008 national registry data, Klemetti and colleagues (2012) compared the outcomes of first-time births among women with two prior abortions of any type or gestation (n = 4,417), three or more prior abortions (n = 942), and no prior abortion (n = 264,190). Their analysis adjusted for maternal age, marital status, socioeconomic status, urban residence, smoking, miscarriage, and ectopic pregnancy. The researchers found a dose-response relationship between the number of prior abortions before a first birth and an increased risk of very preterm birth (<28 weeks’ gestation)6 after two abortions (aOR = 1.69; 95% CI = 1.14, 2.51) and after three or more abortions (aOR = 2.78; 95% CI = 1.48, 5.24) compared with first births among women with no abortion history. In addition, three or more abortions were associated with preterm birth at <37 weeks’ gestation (aOR = 1.35; 95% CI = 1.07–1.71).
KC and colleagues (2017b)7 used Finnish registry data that extended into a more recent time period (1983–2013) to examine first-birth outcomes following more than one medication or aspiration abortion. The analysis controlled for maternal age, marital status, smoking, maternal residence (by municipality), and birth year. The researchers found that first births were at an increased risk of very preterm birth8 after more than one aspiration abortion (n = 4,819) compared with women with no abortions (n = 365,356) (aOR = 1.51; 95% CI = 1.03–2.23). No association was found between multiple medication abortions (n = 1,267) and preterm birth.
6 Incidence of very preterm births (number per 1,000 births): no previous abortions, 3/1,000; one previous abortion, 4/1,000; two previous abortions 6/1,000; three or more previous abortions 11/1,000 (Klemetti et al., 2012).
8KC and colleagues (2017b) use the term “extremely preterm” to refer to births at <28 weeks. Other than the aORs described, the authors did not report the number of very preterm births in the study groups.
Does Abortion Increase the Risk of Breast Cancer?
The association of breast cancer and abortion has been examined in the literature over several decades. Pregnancy has been shown to have a protective effect against breast cancer (NCI, 2016). The original hypotheses suggesting a possible association of abortion with breast cancer drew on animal studies indicating that an interrupted pregnancy might reduce the protective effect of full-term pregnancy on future risk of breast cancer (Russo and Russo, 1980). Epidemiological studies have explored this possible association in analyses of women who have had an abortion. However, much of this literature, including systematic reviews, meta-analyses, and primary research, is flawed by recall bias and lack of controls for such clinically important confounding factors as age at first live birth. The risk of hormone receptor-positive breast cancer increases with a woman’s age at first full-term pregnancy (NCI, 2016).
The committee identified three case control studies of insured women with abortion coverage that used documented records of a prior abortion (Brewster et al., 2005; Goldacre et al., 2001; Newcomb and Mandelson, 2000). The studies controlled for a variety of confounding variables, such as parity, age at delivery of first child, age at breast cancer diagnosis, family history of breast cancer, race, and socioeconomic status.
In a case control study of Scottish women, Brewster and colleagues (2005) linked National Health Service (NHS) hospital discharge and maternity records with national cancer registry and death records dating from 1981 to 1998. The analysis included 2,833 cases (women with a first-time breast cancer diagnosis before age 55) and 9,888 matched controls (women without cancer who had been admitted to an acute care hospital for a non-obstetric, nongynecological condition). Controls were matched with cases by birth year, year of breast cancer diagnosis, residence, and socioeconomic status. The sample was stratified by the same variables, as well as age at breast cancer diagnosis, parity, and age at delivery of first child. Women who had had a prior abortion were no more likely than other women to develop breast cancer (aOR = 0.80; 95% CI = 0.72–0.89). Age at abortion, number of abortions, weeks of gestation, time since abortion, and temporal sequence of live births and abortions also were not found to increase the risk of breast cancer.
In another case control study using linked NHS records, Goldacre and colleagues (2001) analyzed 28,616 breast cancer cases in the Oxford health region of England from 1968 to 1998. The matched control group included 325,456 women who had been hospitalized for reasons other than cancer. The sample was stratified by age, year of the case or control event, residence, and social class. Women with a prior abortion were found not
to be at higher risk of breast cancer than women with no abortion history (OR = 0.83; 95% CI = 0.74–0.93).
Newcomb and Mandelson (2000) analyzed the risk of breast cancer among members of the Group Health Cooperative of Puget Sound (Washington State) by linking health plan and local cancer registry data. The analysis included 138 cancer cases and 252 matched controls and adjusted for race, age at first birth, menopause status, family breast cancer history, and body mass index. The control group was matched by age and period of enrollment in the health plan. The analysis found no association between a history of abortion and breast cancer; compared with women with no prior abortion, the adjusted relative risk of breast cancer in women with an abortion was 0.9 (95% CI = 0.5–1.6).
Does Abortion Increase the Risk of Long-Term Mental Health Problems?
The committee identified a wide array of research on mental health outcomes, including systematic reviews (Bellieni and Buonocore, 2013; Charles et al., 2008; Coleman, 2011; Fergusson et al., 2013; Major et al., 2008, 2009; NCCMH, 2011), prospective cohort studies (Biggs et al., 2015, 2016, 2017; Foster et al., 2015; Munk-Olsen et al., 2011), cohort studies (Fergusson et al., 2006; Gomez, 2018; Herd et al., 2016; Pedersen, 2007, 2008; Steinberg and Russo, 2008; Steinberg et al., 2011; Warren et al., 2010), and analyses linking medical record or registry data (Coleman et al., 2002; Gissler et al., 2015; Leppalahti et al., 2016; Munk-Olsen et al., 2011; Reardon et al., 2003). Most of the studies focused on whether abortion increases women’s risk of depression, anxiety, and/or posttraumatic stress disorder (PTSD).
The utility of most of the published research on mental health outcomes is limited by selective recall bias, inadequate controls for confounding factors, and inappropriate comparators (Major et al., 2008; NCCMH, 2011). Moreover, systematic reviews and meta-analyses are not reliable if they do not assess the quality of the primary research they include (IOM, 2011). As noted earlier, objective documentation of a prior abortion is essential to assessing whether abortion is associated with any outcomes, including subsequent mental health problems. Yet while self-reported data are not reliable sources of abortion history, self-reports are the basis of much of the available primary research on the association between abortion and mental health (Fergusson et al., 2006; Gomez, 2018; Herd et al., 2016; Nilsen et al., 2012; Pedersen, 2007, 2008; Steinberg and Finer, 2011; Steinberg and Russo, 2008; Steinberg et al., 2011; Sullins, 2016; Warren et al., 2010). In addition, as noted earlier, if a study’s objective is to determine whether
having an abortion raises the risk of future mental health problems, it is important that the study control for women’s mental health status at baseline (i.e., before they had the abortion). For example, Steinberg and colleagues (2014) found that women who have abortions report higher rates of mood disorders (depression, bipolar disorder, and dysthymia) (21.0 percent) before undergoing the procedure compared with women with no abortion history who give birth (10.6 percent). Studies by Coleman and colleagues (2002) and Reardon and colleagues (2003) failed to control adequately for preexisting mental disorders. Munk-Olsen and colleagues (2011, 2012) report that their analyses are limited because they were unable to control for a woman’s reason for having an abortion and whether the pregnancy was unwanted. Terminations of pregnancies due to fetal abnormalities, for example, may have very different psychological consequences than abortions for unwanted pregnancies.9
The committee identified seven systematic reviews on the association between abortion and long-term mental health problems (Bellieni and Buonocore, 2013; Charles et al., 2008; Coleman, 2011; Fergusson et al., 2013; Major et al., 2008, 2009; NCCMH, 2011). The 2011 review conducted by the UK National Collaborating Center for Mental Health (NCCMH)10 is particularly informative (NCCMH, 2011). Building on the previously published reviews, the NCCMH (2011) used GRADE11 to analyze the quality of individual studies on several research questions, including the focus of this review, that is, whether women who have an abortion experience more mental health problems than women who deliver an unwanted pregnancy. The two reviews published after the NCCMH report (Bellieni and Buonocore, 2013; Fergusson et al., 2013) identified no additional studies that met the committee’s selection criteria. After extensive quality checks of the primary literature, including controlling for previous mental health problems, NCCMH (2011) found that “the rates of mental health problems for women with an unwanted pregnancy were the same whether they had an abortion or gave birth” (p. 8).
The committee identified several more recent studies that met its selection criteria but were published after the NCCMH and other systematic reviews (Biggs et al., 2015, 2016, 2017; Foster et al., 2015; Leppalahti et
9 The committee did not examine the literature on the mental health consequences of terminations of pregnancies due to fetal abnormalities.
10 The NCCMH was established by the Royal College of Psychiatrists, in partnership with the British Psychological Society, to develop evidence-based mental health reviews and clinical guidelines (NCCMH, 2011).
11 GRADE refers to the Grading of Recommendations Assessment, Development, and Evaluation. It is a tool, used by the Cochrane Collaboration and many other health care research organizations, for assessing the quality of evidence in health care and the strength of clinical recommendations (GRADE Working Group, 2004, 2018).
al., 2016). Four recent articles draw on the Turnaway study, a prospective longitudinal cohort study designed to address many of the limitations of other studies (Biggs et al., 2015, 2016, 2017; Foster et al., 2015).
The Turnaway study contributes unique insight into the consequences of receiving a desired abortion versus being denied the procedure and carrying the pregnancy to term. The study sample included 956 English- and Spanish-speaking women aged 15 and over who sought abortions between 2008 and 2010 from 30 abortion facilities in the United States. The sample design was unique because it drew from groups of women who presented up to 3 weeks beyond a facility’s gestational age limit and were denied an abortion, women presenting within 2 weeks of the limit who received an abortion, and women who received a first-trimester abortion. The women were followed via semiannual phone interviews for 5 years (Dobkin et al., 2014). The investigators collected baseline data on mental health (history of depression, anxiety, suicidal ideation), as well as data on factors known to be important predictors of mental health problems (e.g., history of trauma and abuse). The study groups were specifically designed to enable comparisons of women who had had abortions and those who had been turned away (wanted an abortion but were denied one).
Results from the Turnaway study suggest that there are few psychiatric consequences of abortion, including risk of depression, anxiety, or PTSD. At 2 years, women who had received an abortion had similar or lower levels of depression and anxiety than women denied an abortion (Foster et al., 2015). The study also examined new self-reports of professional diagnoses of either depression or anxiety at 3 years postabortion. Women who had obtained abortions near facility gestational limits were at no greater mental health risk than women who had sought an abortion and carried an unwanted pregnancy to term (Biggs et al., 2015). At 4 years follow-up, the participants completed a measure of PTSD risk (Biggs et al., 2016). Women who had received an abortion were at no higher risk of PTSD than women who had been denied an abortion. At 5 years follow-up, women completed measures of mental health (depression and anxiety) and well-being (self-esteem and life satisfaction) (Biggs et al., 2017). Compared with having had an abortion, having being denied an abortion may be associated with greater risk of initially experiencing more anxiety symptoms; levels of depression were similar among both groups of women.
Two recent studies used Finnish registry data to analyze mental health outcomes after abortion. Leppalahti and colleagues (2016) conducted a longitudinal retrospective cohort study of girls born in Finland in 1987 to examine the effect of abortion on adolescent girls. The comparison groups were girls who had had an abortion (n = 1,041) or given birth (n = 394) before age 18 and a group with no pregnancies up to age 20 (n = 25,312). The girls were followed until age 25. The researchers found no significant
differences between the underage abortion group and childbirth group with respect to risk of any psychiatric disorder (including psychoactive substance use disorder, mood disorder, or neurotic or stress-related disorders) after the index pregnancy (aOR = 0.96; 95% CI = 0.67–1.40). Other recent Finnish research provides some evidence that monitoring for mental health status in a follow-up visit after abortion may help reduce the consequences of serious mental health disorders (Gissler et al., 2015).
Does Abortion Increase the Risk of Premature Death?
Mortality following abortion is an important long-term outcome to consider. As noted in Chapter 2 (see Table 2-4), when mortality rates from abortion and childbirth are compared, abortion is associated with fewer maternal deaths than carrying a pregnancy to term (Grimes, 2006; Raymond and Grimes, 2012). However, the follow-up period in these short-term studies may not have been of sufficient length to account for late complications leading to death. The committee identified several studies that examined long-term mortality and abortion. These studies—one U.S. study (Reardon et al., 2002) and four studies using Finnish registries (Gissler et al., 2004, 2005, 2015; Jalanko et al., 2017)—are based on linked records. In comparing groups on mortality, however, it is important to adjust for both individual characteristics and social risk factors, as they are likely to differ between women who give birth and those who have an abortion. Minority women and those who are young, unmarried, or poor are more likely than more advantaged women to have unwanted pregnancies and subsequent abortions (Boonstra et al., 2006). Without robust risk adjustments for these social differences, attributing outcomes to such factors as having an abortion or not, especially when the outcomes are rare, is inappropriate. As a result of the inability to control for the many ways in which women who have unwanted pregnancies differ from those who do not, no clear conclusions regarding the association between abortion and long-term mortality can be drawn from these studies.
This chapter has reviewed the epidemiological evidence on abortion’s long-term effects on future childbearing and pregnancy outcomes, risk of breast cancer, mental health disorders, and premature death. The committee found that much of the published literature on these topics fails to meet scientific standards for rigorous, unbiased research. Reliable research on these outcomes uses documented records of a prior abortion, analyzes
comparable study and control groups, and controls for confounding variables shown to affect the outcome of interest. Thus, this chapter has focused on the findings of research that meets these basic standards. The committee did not find well-designed research on abortion’s association with future ectopic pregnancy, miscarriage or stillbirth, or long-term mortality. Findings on hemorrhage during a subsequent pregnancy are inconclusive.
The committee identified high-quality research on numerous outcomes of interest and concludes that having an abortion does not increase a woman’s risk of secondary infertility, pregnancy-related hypertensive disorders, abnormal placentation (after a D&E abortion), preterm birth, breast cancer, or mental health disorders (depression, anxiety, and PTSD). An increased risk of very preterm birth (<28 weeks’ gestation) in a woman’s first birth was found to be associated with having two or more prior aspiration abortions compared with first births among women with no abortion history; the risk appears to be associated with the number of prior abortions. Preterm birth is associated with pregnancy spacing after an abortion: it is more likely if the interval between abortion and conception is less than 6 months (the same is also true of pregnancy spacing in general).
ACOG (American College of Obstetricians and Gynecologists). 2017. Ectopic pregnancy. https://www.acog.org/Patients/FAQs/Ectopic-Pregnancy (accessed November 13, 2017).
Alexander, L. K., B. Lopes, K. Ricchetti-Masterson, and K. B. Yeatts. 2015. ERIC notebook. sources of systematic error or bias: Information bias, 2nd ed. https://sph.unc.edu/files/2015/07/nciph_ERIC14.pdf (accessed December 4, 2017).
Ananth, C. V., D. P. Misra, K. Demissie, and J. C. Smulian. 2001. Rates of preterm delivery among black women and white women in the United States over two decades: An age-period-cohort analysis. American Journal of Epidemiology 154(7):657–665.
Anderson, B. A., K. Katus, A. Puur, and B. D. Silver. 1994. The validity of survey responses on abortion: Evidence from Estonia. Demography 31(1):115–132.
Ball, S. J., G. Pereira, P. Jacoby, N. de Klerk, and F. J. Stanley. 2014. Re-evaluation of link between interpregnancy interval and adverse birth outcomes: Retrospective cohort study matching two intervals per mother. British Medical Journal 349:g4333. doi:10.1136/bmj.g4333.
Bateman B. T., M. F. Berman, L. E. Riley, and L. R. Leffert. 2010. The epidemiology of postpartum hemorrhage in a large, nationwide sample of deliveries. Anesthesia-Analgesia 110(5):1368–1373.
Bellieni, C. V., and G. Buonocore. 2013. Abortion and subsequent mental health: Review of the literature. Psychiatry and Clinical Neurosciences 67(5):301–310.
Beral, V., D. Bull, R. Doll, R. Peto, and G. Reeves. 2004. Breast cancer and abortion: Collaborative reanalysis of data from 53 epidemiological studies, including 83,000 women with breast cancer from 16 countries. Lancet 363(9414):1007–1016.
Biggs, M. A., J. M. Neuhaus, and D. G. Foster. 2015. Mental health diagnoses 3 years after receiving or being denied an abortion in the United States. American Journal of Public Health 105(12):2557–2563.
Biggs, M. A., B. Rowland, C. E. McCulloch, and D. G. Foster. 2016. Does abortion increase women’s risk for post-traumatic stress? Findings from a prospective longitudinal cohort study. British Medical Journal Open 6(2):e009698. doi:10.1136/bmjopen-2015-009698.
Biggs, M. A., U. D. Upadhyay, C. E. McCulloch, and D. G. Foster. 2017. Women’s mental health and well-being 5 years after receiving or being denied an abortion: A prospective, longitudinal cohort study. JAMA Psychiatry 74(2):169–178.
Boonstra, H. D., R. Gold, C. Richards, and L. Finer. 2006. Abortion in women’s lives. New York: Guttmacher Institute.
Bouyer, J., J. Coste, T. Shojaei, J. L. Pouly, H. Fernandez, L. Gerbaud, and N. Job-Spira. 2003. Risk factors for ectopic pregnancy: A comprehensive analysis based on a large case-control, population-based study in France. American Journal of Epidemiology 157(3):185–194.
Brewster, D., D. Stockton, R. Dobbie, D. Bull, and V. Beral. 2005. Risk of breast cancer after miscarriage or induced abortion: A Scottish record linkage case-control study. Journal of Epidemiology and Community Health 59(4):283–287.
Callaghan, W. M., E. V. Kuklina, and C. J. Berg. 2010. Trends in postpartum hemorrhage: United States, 1994–2006. American Journal of Obstetrics and Gynecology 202(4):353, e351–e356.
CDC (Centers for Disease Control and Prevention). 2016. Pregnancy complications. https://www.cdc.gov/reproductivehealth/maternalinfanthealth/pregcomplications.htm (accessed February 2, 2018).
Charles, V. E., C. B. Polis, S. K. Sridhara, and R. W. Blum. 2008. Abortion and long-term mental health outcomes: A systematic review of the evidence. Contraception 78(6):436–450.
Cochrane Collaboration. 2018. Assessing risk of bias in included studies. http://methods.cochrane.org/bias/assessing-risk-bias-included-studies (accessed January 5, 2018).
Coleman, P. K. 2011. Abortion and mental health: Quantitative synthesis and analysis of research published 1995–2009. The British Journal of Psychiatry 199(3):180–186.
Coleman, P. K., D. C. Reardon, V. M. Rue, and J. Cougle. 2002. State-funded abortions versus deliveries: A comparison of outpatient mental health claims over 4 years. American Journal of Orthopsychiatry 72(1):141–152.
Conde-Agudelo, A., A. Rosas-Bermudez, F. Castano, and M. H. Norton. 2012. Effects of birth spacing on maternal, perinatal, infant, and child health: A systematic review of causal mechanisms. Studies in Family Planning 43(2):93–114.
Dobkin, L. M., H. Gould, R. E. Barar, M. Ferrari, E. I. Weiss, and D. G. Foster. 2014. Implementing a prospective study of women seeking abortion in the United States: Understanding and overcoming barriers to recruitment. Women’s Health Issues 24(1):e115–e123.
Fergusson, D. M., L. J. Horwood, and E. M. Ridder. 2006. Abortion in young women and subsequent mental health. Journal of Child Psychology and Psychiatry 47(1):16–24.
Fergusson, D. M., L. J. Horwood, and J. M. Boden. 2013. Does abortion reduce the mental health risks of unwanted or unintended pregnancy? A re-appraisal of the evidence. Australian & New Zealand Journal of Psychiatry 47(9):819–827.
Foster, D. G., J. R. Steinberg, S. C. Roberts, J. Neuhaus, and M. A. Biggs. 2015. A comparison of depression and anxiety symptom trajectories between women who had an abortion and women denied one. Psychological Medicine 45(10):2073–2082.
Garovic, V. D., and P. August. 2013. Preeclampsia and the future risk of hypertension: The pregnant evidence. Current Hypertension Reports 15(2). doi:10.1007/s11906-013-0329-4.
Gissler, M., C. Berg, M.-H. Bouvier-Colle, and P. Buekens. 2004. Pregnancy-associated mortality after birth, spontaneous abortion, or induced abortion in Finland, 1987–2000. American Journal of Obstetrics and Gynecology 190(2):422–427.
Gissler, M., C. Berg, M.-H. Bouvier-Colle, and P. Buekens. 2005. Injury deaths, suicides and homicides associated with pregnancy, Finland 1987–2000. The European Journal of Public Health 15(5):459–463.
Gissler, M., E. Karalis, and V. M. Ulander. 2015. Decreased suicide rate after induced abortion, after the current care guidelines in Finland 1987–2012. Scandinavian Journal of Public Health 43(1):99–101.
Goldacre, M., L. Kurina, V. Seagroatt, and D. Yeates. 2001. Abortion and breast cancer: A case-control record linkage study. Journal of Epidemiology and Community Health 55(5):336–337.
Gomez, A. M. 2018. Abortion and subsequent depressive symptoms: An analysis of the National Longitudinal Study of Adolescent Health. Psychology Medicine 48(2):294–304. doi:10.1017/S0033291717001684.
GRADE (Grading of Recommendations Assessment, Development and Evaluation) Working Group. 2004. Grading quality of evidence and strength of recommendations. British Medical Journal 328(7454):1490.
GRADE Working Group. 2018. GRADE. http://www.gradeworkinggroup.org (accessed January 19, 2018).
Grimes, D. A. 2006. Estimation of pregnancy-related mortality risk by pregnancy outcome, United States, 1991 to 1999. American Journal of Obstetrics and Gynecology 194(1):92–94.
Hanley, G. E., J. A. Hutcheon, B. A. Kinniburgh, and L. Lee. 2017. Interpregnancy interval and adverse pregnancy outcomes: An analysis of successive pregnancies. Obstetrics & Gynecology 129(3):408–415.
Herd, P., J. Higgins, K. Sicinski, and I. Merkurieva. 2016. The implications of unintended pregnancies for mental health in later life. American Journal of Public Health 106(3):421–429.
Hogue, C. J. 1975. Low birth weight subsequent to induced abortion. A historical prospective study of 948 women in Skopje, Yugoslavia. American Journal of Obstetrics and Gynecology 123(7):675–681.
Hogue, C. J., W. Cates, Jr., and C. Tietze. 1982. The effects of induced abortion on subsequent reproduction. Epidemiologic Reviews 4(1):66–94.
Holmlund, S., T. Kauko, J. Matomaki, M. Tuominen, J. Makinen, and P. Rautava. 2016. Induced abortion—impact on a subsequent pregnancy in first-time mothers: A registry-based study. BMC Pregnancy and Childbirth 16(1):325.
IOM (Institute of Medicine). 2011. Finding what works in health care: Standards for systematic reviews. Washington, DC: The National Academies Press.
Jackson, J. E., W. A. Grobman, E. Haney, and H. Casele. 2007. Mid-trimester dilation and evacuation with laminaria does not increase the risk for severe subsequent pregnancy complications. International Journal of Gynaecology & Obstetrics 96(1):12–15.
Jalanko, E., S. Leppalahti, O. Heikinheimo, and M. Gissler. 2017. Increased risk of premature death following teenage abortion and childbirth: A longitudinal cohort study. European Journal of Public Health 5(1):845–849.
Jones, R. K., and K. Kost. 2007. Underreporting of induced and spontaneous abortion in the United States: An analysis of the 2002 National Survey of Family Growth. Studies in Family Planning 38(3):187–197.
KC, S., E. Hemminki, M. Gissler, S. M. Virtanen, and R. Klemetti. 2017a. Perinatal outcomes after induced termination of pregnancy by methods: A nationwide register-based study of first births in Finland 1996–2013. PLoS One 12(9):e0184078. doi: 10.1371/journal. pone.0184078.
KC, S., M. Gissler, S. M. Virtanen, and R. Klemetti. 2017b. Risks of adverse perinatal outcomes after repeat terminations of pregnancy by their methods: A nationwide register-based cohort study in Finland 1996–2013. Paediatrics and Perinatal Epidemiology 31(6):485–492.
Klebanoff, M. A. 2017. Interpregnancy interval and pregnancy outcomes: Causal or not? Obstetrics & Gynecology 129(3):405–407.
Klemetti, R., M. Gissler, M. Niinimaki, and E. Hemminki. 2012. Birth outcomes after induced abortion: A nationwide register-based study of first births in Finland. Human Reproduction 27(11):3315–3320.
Lange, E. M. S., and P. Toledo. 2017. Chapter 33. Antepartum hemorrhage. In Complications in anesthesia, 3rd ed., edited by L. A. Fleisher and S. H. Rosenbaum. Philadelphia, PA: W. B. Elsevier.
Leppalahti, S., O. Heikinheimo, I. Kalliala, P. Santalahti, and M. Gissler. 2016. Is underage abortion associated with adverse outcomes in early adulthood? A longitudinal birth cohort study up to 25 years of age. Human Reproduction 31(9):2142–2149.
Lindefors-Harris, B. M., G. Eklund, H. O. Adami, and O. Meirik. 1991. Response bias in a case-control study: Analysis utilizing comparative data concerning legal abortions from two independent Swedish studies. American Journal of Epidemiology 134(9):1003–1008.
Lowit, A., S. Bhattacharya, and S. Bhattacharya. 2010. Obstetric performance following an induced abortion. Best Practice & Research in Clinical Obstetrics & Gynaecology 24(5):667–682.
Major, B., M. Appelbaum, L. Beckman, M. A. Dutton, N. F. Russo, and C. West. 2008. Report of the APA Task Force on Mental Health and Abortion. Washington, DC: American Psychological Association.
Major, B., M. Appelbaum, L. Beckman, M. A. Dutton, N. F. Russo, and C. West. 2009. Abortion and mental health: Evaluating the evidence. American Psychologist 64(9):863–890.
Mannisto, J., A. Bloigu, M. Mentula, M. Gissler, O. Heikinheimo, and M. Niinimaki. 2017. Interpregnancy interval after termination of pregnancy and the risks of adverse outcomes in subsequent birth. Obstetrics & Gynecology 129(2):347–354.
Munk-Olsen, T., T. M. Laursen, C. B. Pedersen, Ø. Lidegaard, and P. B. Mortensen. 2011. Induced first-trimester abortion and risk of mental disorder. New England Journal of Medicine 364(4):332–339.
Munk-Olsen, T., T. M. Laursen, C. B. Pedersen, Ø. Lidegaard, and P. B. Mortensen. 2012. First-time first-trimester induced abortion and risk of readmission to a psychiatric hospital in women with a history of treated mental disorder. Archives of General Psychiatry 69(2):159–165.
NAF (National Abortion Federation). 2017. 2017 Clinical policy guidelines for abortion care. Washington, DC: NAF.
NCCMH (National Collaborating Centre for Mental Health). 2011. Induced abortion and mental health: A systematic review of the mental health outcomes of induced abortion, including their prevalence and associated factors. London, UK: Academy of Medical Royal Colleges.
NCI (National Cancer Institute). 2016. Reproductive history and cancer risk. https://www.cancer.gov/about-cancer/causes-prevention/risk/hormones/reproductive-history-fact-sheet (accessed January 31, 2018).
Newcomb, P. A., and M. T. Mandelson. 2000. A record-based evaluation of induced abortion and breast cancer risk (United States). Cancer Causes Control 11(9):777–781.
Nilsen, W., C. A. Olsson, E. Karevold, C. O’Loughlin, M. McKenzie, and G. C. Patton. 2012. Adolescent depressive symptoms and subsequent pregnancy, pregnancy completion and pregnancy termination in young adulthood: Findings from the Victorian Adolescent Health Cohort Study. Journal of Pediatric and Adolescent Gynecology 25(1):6–11.
ODPHP (Office of Disease Prevention and Health Promotion). 2017. National Vital Statistics System—fetal death. https://www.healthypeople.gov/2020/data-source/national-vital-statistics-system-fetal-death (accessed November 10, 2017).
Pedersen, W. 2007. Childbirth, abortion and subsequent substance use in young women: A population-based longitudinal study. Addiction 102(12):1971–1978.
Pedersen, W. 2008. Abortion and depression: A population-based longitudinal study of young women. Scandinavian Journal of Social Medicine 36(4):424–428.
Raymond, E. G., and D. A. Grimes. 2012. The comparative safety of legal induced abortion and childbirth in the United States. Obstetrics & Gynecology 119(2 Pt. 1):215–219.
RCOG (Royal College of Obstetricians and Gynaecologists). 2011. The care of women requesting induced abortion (Evidence-based clinical guideline number 7). https://www.rcog.org.uk/globalassets/documents/guidelines/abortion-guideline_web_1.pdf (accessed July 27, 2017).
RCOG. 2015. Best practice in comprehensive abortion care (Best practice paper no. 2). London, UK: RCOG.
RCOG. 2017. Prevention and management of postpartum haemorrhage. British Journal of Obstetrics & Gynaecology 124(5):e106–e149.
Reardon, D., J. Cougle, P. Ney, F. Scheuren, P. Coleman, and T. Strahan. 2002. Deaths associated with delivery and abortion among California Medicaid patients: A record linkage study. The Southern Medical Journal 95(8):834–841.
Reardon, D. C., J. R. Cougle, V. M. Rue, M. W. Shuping, P. K. Coleman, and P. G. Ney. 2003. Psychiatric admissions of low-income women following abortion and childbirth. Canadian Medical Association Journal 168(10):1253–1256.
Roblin, P. 2014. Vacuum aspiration. In Abortion care, edited by S. Rowlands. Cambridge, UK: Cambridge University Press.
Russo, J., and I. H. Russo. 1980. Susceptibility of the mammary gland to carcinogenesis. II. Pregnancy interruption as a risk factor in tumor incidence. American Journal of Pathology 100(2):497–512.
Saleh, H. 2008. Placenta previa and accreta. https://www.glowm.com/section_view/heading/Placenta%20Previa%20and%20Accreta/item/121 (accessed November 13, 2017).
SFP (Society of Family Planning). 2013. Surgical abortion prior to 7 weeks of gestation. Contraception 88(1):7–17.
Shachar, B. Z., J. A. Mayo, D. J. Lyell, R. J. Baer, L. L. Jeliffe-Pawlowski, D. K. Stevenson, and G. M. Shaw. 2016. Interpregnancy interval after live birth or pregnancy termination and estimated risk of preterm birth: A retrospective cohort study. British Journal of Obstetrics & Gynaecology 123(12):2009–2017.
Sivalingam, V. N., W. C. Duncan, E. Kirk, L. A. Shephard, and A. W. Horne. 2011. Diagnosis and management of ectopic pregnancy. The Journal of Family Planning and Reproductive Health Care 37(4):231–240.
Smith, G. C. S., J. P. Pell, and R. Dobbie. 2003. Interpregnancy interval and risk of preterm birth and neonatal death: Retrospective cohort study. British Medical Journal 327(7410):313.
Steinberg, J. R., and L. B. Finer. 2011. Examining the association of abortion history and current mental health: A reanalysis of the national comorbidity survey using a common-risk-factors model. Social Science and Medicine 72(1):72–82.
Steinberg, J. R., and N. F. Russo. 2008. Abortion and anxiety: What’s the relationship? Social Science & Medicine 67(2):238–252.
Steinberg, J. R., D. Becker, and J. T. Henderson. 2011. Does the outcome of a first pregnancy predict depression, suicidal ideation, or lower self-esteem? Data from the National Comorbidity Survey. American Journal of Orthopsychiatry 81(2):193–201.
Steinberg, J. R., C. E. McCulloch, and N. E. Adler. 2014. Abortion and mental health: Findings from the National Comorbidity Survey-Replication. Obstetrics & Gynecology 123(2 Pt. 1):263–270.
Sullins, D. P. 2016. Abortion, substance abuse and mental health in early adulthood: Thirteen-year longitudinal evidence from the United States. SAGE Open Medicine 4:2050312116665997.
Tao, G., C. Patel, and K. W. Hoover. 2017. Updated estimates of ectopic pregnancy among commercially and Medicaid-insured women in the United States, 2002–2013. Southern Medical Journal 110(1):18–24. doi:10.14423/SMJ.0000000000000594.
Thorp, Jr., J. M., K. E. Hartmann, and E. Shadigian. 2003. Long-term physical and psychological health consequences of induced abortion: Review of the evidence. Obstetrical & Gynecological Survey 58(1):67–79.
Warren, J. T., S. M. Harvey, and J. T. Henderson. 2010. Do depression and low self-esteem follow abortion among adolescents? Evidence from a national study. Perspectives on Sexual and Reproductive Health 42(4):230–235.
Wendt, A., C. M. Gibbs, S. Peters, and C. J. Hogue. 2012. Impact of increasing inter-pregnancy interval on maternal and infant health. Paediatric and Perinatal Epidemiology 26(Suppl. 1):239–258.
WHO (World Health Organization). 2012. Safe abortion: Technical and policy guidance for health systems, 2nd ed. http://apps.who.int/iris/bitstream/10665/70914/1/9789241548434_eng.pdf (accessed September 12, 2017).
Wong, L. F., J. Wilkes, K. Korgenski, M. W. Varner, and T. A. Manuck. 2016. Risk factors associated with preterm birth after a prior term delivery. British Journal of Obstetrics & Gynaecology 123(11):1772–1778.
Woolner, A., S. Bhattacharya, and S. Bhattacharya. 2014. The effect of method and gestational age at termination of pregnancy on future obstetric and perinatal outcomes: A register-based cohort study in Aberdeen, Scotland. British Journal of Obstetrics & Gynaecology 121(3):309–318.
Zhou, W., G. L. Nielsen, H. Larsen, and J. Olsen. 2001. Induced abortion and placenta complications in the subsequent pregnancy. Acta Obstetricia et Gynecologica Scandinavica 80(12):1115–1120.