All-Cause Mortality in Reproductive-Aged Females by State: An Analysis of the Effects of Abortion Legislation : Obstetrics & Gynecology

Secondary Logo

Journal Logo

Original Research

All-Cause Mortality in Reproductive-Aged Females by State

An Analysis of the Effects of Abortion Legislation

Harper, Lorie M. MD, MSCI; Leach, Justin M. PhD; Robbins, Lindsay MD, MPH; Blanchard, Christina MS; Metz, Torri D. MD, MS; Mazzoni, Sara MD, MPH; Nash, Elizabeth MPP; Szychowski, Jeff PhD

Author Information
Obstetrics & Gynecology: November 1, 2022 - Volume - Issue - 10.1097/AOG.0000000000005035
doi: 10.1097/AOG.0000000000005035
  • Free
  • PAP

Given the Supreme Court decision in Dobbs v Jackson Women's Health Organization, it is anticipated that approximately 33 million U.S. women will live in states without available abortion services.1 Even before this decision, abortion access varied widely across the United States; in 2019, 43% of reproductive-aged females lived in states that had policies that were overall restrictive to abortion2 due to the 26 abortion bans that had been enacted, with many more having been introduced.2 Other states have passed laws supportive of and protective of abortion.

Many of the laws restricting abortion were purported to protect women's health by regulating abortion facilities as ambulatory surgery centers, dictating the types of health care professionals who can perform abortion, and requiring admitting privileges to hospitals for clinicians who provide abortion care; opponents of these regulations assert that they are designed to restrict access without evidence that they decrease the risks associated with abortion. Prior studies have demonstrated that safe access to abortions may prevent up to 13% of maternal deaths3; thus, it is possible that these laws purporting to protect women with safer abortions have actually resulted in increased pregnancy-related deaths due to restricted abortion access.

Abortion regulations and restrictions also focus on protecting the fetus (and subsequent neonate).4 However, prior studies have demonstrated an association between abortion access restriction and increased infant mortality.5

We assessed the association between state-level abortion legislation and all-cause mortality among females of reproductive age and maternal, fetal, and infant mortality. To do so, we performed a retrospective cohort study that compared mortality rates in states categorized as supportive, moderate, and restrictive with regard to their abortion laws.

METHODS

We conducted a retrospective cohort study using the Centers for Disease Control and Prevention’s (CDC) WONDER (Wide-ranging ONline Data for Epidemiologic Research) database, a publicly available database with information on mortality by state.6–10 This study was exempt from IRB review because only publicly available, de-identified data were used. The STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines for reporting cohort studies were followed.11

Data were available for all-cause, maternal, and infant mortality from 2000 to 2019; fetal mortality data were available from 2005 to 2019. We selected this 20-year period to evaluate because data were available for both maternal and infant mortality for all years selected and state-level abortion legislation shifted dramatically over this period. A large time period was required to account for changes in state legislation (resulting in changing exposure groups) and to provide adequate power for detecting differences in rare outcomes. The database was queried for state- and year-specific deaths in reproductive-aged females, infant deaths, and fetal deaths. The primary outcome was all-cause mortality in reproductive-aged females (age 15–49 years). Secondary outcomes included maternal mortality, fetal mortality, and infant mortality.

All-cause mortality in reproductive-aged females was defined as death in a female with a reported age of 15–49 years. All-cause mortality in reproductive-aged females was selected as the primary outcome because maternal mortality is unreliable in national data sets due to errors in recording and linking pregnancy events with death certificates.12–14 As a secondary analysis of maternal mortality specifically, we used a previously published technique15 to identify maternal mortality in the CDC WONDER database; any causes of death coded with the International Classification of Diseases, Tenth Revision as A34, O-O95 and O98-99 were considered related to maternal mortality. Infant death was defined as death within 1 year of life.8–10Fetal death was defined as fetal death at 20 weeks of gestation or more7; the database makes no distinction between spontaneous and induced fetal deaths. Data were aggregated at the state level, and states served as the units of analysis.

The Guttmacher Institute has analyzed the abortion policy landscape in each state (Fig. 1).16 The laws of each state were reviewed and scored on a scale of −6 to +6 based on the number of supportive policies (+1 point) and the number of restrictive policies (−1 point). Based on this score from −6 to +6, states were classified as either restrictive (−6 through −2), moderate (−1, 0, or +1), or supportive (+2 through +6) for every year from 2000 to 2019. Policies were not counted in effect if they were blocked by a court order or if the effective date had not been reached.

F1
Fig. 1.:
Abortion policy landscape: 2000 (A), 2010 (B), and 2019 (C). Data from Guttmacher Institute. State abortion policy landscape: from hostile to supportive. Accessed October 17, 2022. https://www.guttmacher.org/article/2019/08/state-abortion-policy-landscape-hostile-supportive. Created with mapchart.net.

We estimated the influence of specific, common legal restrictions placed on abortion, including in-person counseling, waiting periods, restrictions on insurance coverage, laws requiring inaccurate or misleading counseling (such as information on medication abortion reversal or fetal personhood), prohibition of telemedicine for medication abortion, parental consent for minors, and clinic regulations (TRAP [Targeted Regulation of Abortion Providers] laws). We also considered the potential effect of trigger laws, or laws that would ban all abortions in the event that Roe v Wade was overturned.

Statistical analyses were conducted using R 4.1.3. To account for repeated outcome measures over time, we modeled all outcomes using generalized estimating equations with year as a covariate17; generalized estimating equation models were fit using the R package geepack 1.3.4.18 Key predictors (eg, the number of laws, presence of specific laws, or percentage of the population in poor health) also varied over time. We used a cross-sectional approach in our analyses, for example, for a specific year, the predictor was whether the law was in effect for that year; an individual state may change exposure categories (restrictive, moderate, supportive) in any given year. To avoid biased estimates in the presence of time-varying predictors, we used independence working correlation with robust standard errors in our generalized estimating equation models.19 All models controlled for the percentage of the state population in poor health. The percentage of the population in poor health was obtained from the CDC's Behavioral Risk Factor Surveillance System, an ongoing, state-based, random-digit–dialed telephone survey of noninstitutionalized civilian adults aged 18 years and older.20 All hypothesis tests were conducted at the α=0.05 level using Wald statistics, and we report effect estimates, 95% CIs, and P-values. As a sensitivity analysis to address multiple testing, we also applied a Benjamini-Hochberg correction to control the false discovery rate at 5%.21

RESULTS

We first assessed mortality rates by year and by abortion legislation classification (Appendices 1–4, available online at https://links.lww.com/AOG/C957). The mean all-cause mortality rate in reproductive-aged females per year ranged from 108.6 to 117.7 per 100,000, without a clearly increasing or decreasing trend over the dates of interest. Mean maternal mortality demonstrated an increase over the study period, ranging from 8.24 to 32.21 per 100,000. Mean infant mortality ranged from 5.7 to 7.1 per 1,000 live births, with a clearly decreasing trend from the start of the study period to the final year data were available. Fetal mortality was steady at 5.7–6.0 per 1,000 births over the years that data were available. When all-cause mortality in reproductive-aged females was compared among restrictive (reference), moderate, and supportive states, moderate and supportive states were not associated with a significant decrease in all-cause mortality compared with restrictive states (Table 1). However, maternal mortality was significantly lower in moderate states (5.8 fewer maternal deaths/100,000 live births, 95% CI −9.9 to −1.7) compared with restrictive states. The difference in maternal mortality between supportive states and restrictive states was not statistically significantly lower (2.5 fewer maternal deaths/100,000 live births, 95% CI −6.7 to 1.7). Infant mortality was significantly lower in both moderate (0.6 fewer infant deaths/1,000 live births, 95% CI −1.1 to −0.04) and supportive (1.1 fewer infant deaths/1,000 live births, 95% CI −1.6 to −0.6) states. Fetal mortality was also significantly lower in moderate states (0.7 fewer fetal deaths/1,000 births, 95% CI −1.2 to −0.2), with a nonsignificantly lower fetal mortality rate in supportive states (0.6 fewer fetal deaths/1,000 births, 95% CI −1.1 to 0.1).

Table 1. - Estimates and Hypothesis Tests for Level of Restriction by Type of Mortality Rate, 2000–2019*
Estimate 95% CI P
Primary outcome
 All-cause mortality rate in reproductive-aged females (per 100,000)
  Supportive states −6.70 −15.65 to 2.24 .256
  Moderate states 0.59 −6.31 to 7.50
Secondary outcomes
 Maternal mortality rate (per 100,000 live births)
  Supportive states −2.51 −6.75 to 1.72 .021
  Moderate states −5.79 −9.88 to −1.70
 Infant mortality rate (per 1,000 live births)
  Supportive states −1.10 −1.56 to −0.64 <.001
  Moderate states −0.56 −1.09 to −0.04
 Fetal mortality rate (per 1,000 births)
  Supportive states −0.64 −1.41 to 0.13 .019
  Moderate states −0.69 −1.18 to −0.20
*Restrictive states is the reference group.
The P-value presented is a type 3 P-value, which tests the null hypothesis that all levels of a categorical predictor have the same effect on the outcome as the reference category, conditional on the other covariates in the model, that is, year and percentage of the population in poor health.
Remained significant after Benjamini-Hochberg correction to control false discovery rate at q=0.05.

We then evaluated the association between the number and types of abortion-related laws and mortality. In the analysis of all-cause mortality in reproductive-aged females, the direction of effect of the number of laws and each individual type of law was to increase all-cause mortality, except for insurance restriction (Table 2). Types of laws that were significantly associated with an increase in all-cause mortality in reproductive-aged females were trigger laws (laws that would eliminate access to abortion if Roe v Wade were overturned), laws that limit access to medication abortion, and laws that require parental consent for minors to access abortion, although only trigger laws remained significant after adjustment for false discovery rate.

Table 2. - All-Cause Mortality in Reproductive-Aged Females, Per 100,000
Estimate 95% CI P
No. of laws 1.25 −0.16 to 2.66 .082
Trigger law (bans abortion on overturn of Roe v Wade) 12.82 7.93–17.71 <.001*
Prohibition of abortion after a certain gestational age 2.36 −3.76 to 8.48 .45
Laws that require certain types of counseling 5.78 −1.09 to 12.65 .10
Laws requiring in-person counseling or consent or both 5.36 −1.75 to 12.46 .14
Laws that require an ultrasonogram before abortion procedure 2.27 −4.92 to 9.45 .54
Laws that prohibit insurance coverage of abortion −0.02 −5.78 to 5.74 .99
Laws that prohibit Medicaid coverage of abortion 6.94 −0.12 to 14.00 .05
Laws that limit access to medication abortion 8.00 1.90–14.10 .010
Laws that require parental consent for abortion in those younger than 18 y 8.21 1.50–14.92 .017
TRAP laws 3.28 −2.80 to 9.36 .29
TRAP, Targeted Regulation of Abortion Providers.
*Remained significant after Benjamini-Hochberg correction to control false discovery rate at q=0.05.

For maternal mortality (Table 3), each additional abortion regulation was associated with an increase of 1.09 maternal deaths per 100,000 live births (95% CI 0.36–1.82, P=.003). Types of laws that were significantly associated with an increase in maternal mortality were trigger laws, laws that prohibit abortion after a certain gestational age (not significant after adjustment for false discovery rate), laws that require certain types of counseling, laws that require in-person appointments, laws limiting access to medication abortion, and TRAP laws.

Table 3. - Maternal Mortality Rate, Per 100,000 Live Births
Estimate 95% CI P
No. of laws 1.09 0.36–1.82 .003*
Trigger law (bans abortion on overturn of Roe v Wade) 5.69 2.38–9.01 .001*
Prohibition of abortion after a certain gestational age 3.13 0.22–6.04 .04
Laws that require certain types of counseling 4.52 0.55–8.50 .026*
Laws requiring in-person counseling or consent or both 3.88 0.47–7.29 .026*
Laws that require an ultrasonogram before abortion procedure 1.80 −3.13 to 6.74 .47
Laws that prohibit insurance coverage of abortion 2.50 −0.72 to 5.73 .128
Laws that prohibit Medicaid coverage of abortion 1.13 −2.55 to 4.81 .546
Laws that limit access to medication abortion 5.29 1.01–9.57 .015*
Laws that require parental consent for abortions in those younger than 18 y −0.12 −3.73 to 3.50 .950
TRAP laws 5.09 1.76–8.43 .003*
TRAP, Targeted Regulation of Abortion Providers.
*Remained significant after Benjamini-Hochberg correction to control false discovery rate at q=0.05.

Each additional abortion regulation was associated with an increase of 0.20 infant deaths per 1,000 live births (95% CI 0.13–0.26, P<.001, Table 4). Infant mortality also increased with the passage of trigger laws, laws requiring in-person counseling, laws requiring an ultrasonogram before abortion, laws limiting insurance and Medicaid coverage of abortion, laws limiting access to medication abortion, laws that require parental consent for minors to access abortion, and TRAP laws. The only types of abortion legislation not associated with increases in infant mortality were laws limiting abortion after 20 weeks of gestation and laws requiring inaccurate counseling.

Table 4. - Infant Mortality, Per 1,000 Live Births
Estimate 95% CI P
No. of laws 0.20 0.13–0.26 <.001*
Trigger law (bans abortion on overturn of Roe v Wade) 1.12 0.81–1.41 <.001*
Prohibition of abortion after a certain gestational age 0.27 −0.15 to 0.69 .203
Laws that require certain types of counseling 0.43 −0.02 to 0.88 .059
Laws requiring in-person counseling or consent or both 0.67 0.10–1.24 .021*
Laws that require an ultrasonogram before abortion procedure 0.75 0.36–1.14 .001*
Laws that prohibit insurance coverage of abortion 0.53 0.16–0.91 .005*
Laws that prohibit Medicaid coverage of abortion 0.81 0.40–1.22 .001*
Laws that limit access to medication abortion 0.74 0.40–1.07 <.001*
Laws that require parental consent for abortions in those younger than 18 y 0.90 0.55–1.26 <.001*
TRAP laws 0.59 0.13–1.04 .011*
TRAP, Targeted Regulation of Abortion Providers.
*Remained significant after Benjamini-Hochberg correction to control false discovery rate at q=0.05.

Although fetal mortality was increased in all secondary analyses, only restrictions on insurance coverage were significantly associated with the increase (Table 5), and no associations were significantly different after correction for false discovery rate.

Table 5. - Fetal Mortality Rate, Per 1,000 Births (Live Births and Fetal Deaths)
Estimate 95% CI P *
No. of laws 0.11 −0.01 to 0.23 .077
Trigger law (bans abortion on overturn of Roe v Wade) 0.58 −0.69 to 1.85 .37
Prohibition of abortion after a certain gestational age 0.45 −0.03 to 0.93 .065
Laws that require certain types of counseling 0.12 −0.51 to 0.76 .70
Laws requiring in-person counseling or consent or both 0.20 −0.51 to 0.92 .577
Laws that require an ultrasonogram before abortion procedure 0.67 −0.10 to 1.44 .090
Laws that prohibit insurance coverage of abortion 0.47 0.01–0.93 .047
Laws that prohibit Medicaid coverage of abortion 0.38 −0.18 to 0.93 .180
Laws that limit access to medication abortion 0.28 −0.14 to 0.71 .190
Laws that require parental consent for abortions in those younger than 18 y 0.46 −0.10 to 1.02 .109
TRAP laws 0.47 −0.12 to 1.05 .116
TRAP, Targeted Regulation of Abortion Providers.
*No tests remained significant after Benjamini-Hochberg correction to control false discovery rate at q=0.05.

DISCUSSION

In this national retrospective cohort using CDC WONDER data, increasingly restrictive abortion legislation was not associated with an increase in all-cause mortality in reproductive-aged females. However, increasing number of laws was associated with increasing maternal and infant mortality. Notably, the presence of trigger laws was significantly associated with all-cause mortality in reproductive-aged females, maternal mortality, and infant mortality. Additionally, laws enacted with the purported intent of protecting women's health and fetal health by regulating abortion were not associated with decreases in all-cause mortality in reproductive-aged women or maternal mortality.

The strong observed association between trigger laws and all-cause mortality in reproductive-aged females, maternal mortality, and infant mortality is notable because none of these laws were actually in effect at the time of this study. In other words, although the trigger laws prevented no abortions from actually occurring during this study period, they were still associated with an increase in mortality for women and infants in those states. This may be related to the number of other abortion restrictions already in effect and, therefore, may serve as a marker of abortion access, because all states with trigger laws are classified as restrictive states.

Prior studies have demonstrated increases in maternal and infant mortality in association with abortion restrictions.5,22 In a recent article assessing the effect of the Texas state abortion ban after detection of embryonic cardiac activity, Nambiar et al23 reported on increasing numbers of obstetric complications when institutions and physicians interpret state legislation as prohibiting definitive management of previable obstetric complications. In that report, 16 of 28 (57%) patients experienced maternal morbidity and only one neonate survived beyond 1 day of life.

This study has several strengths. It is unique in that we simultaneously examined all-cause mortality in reproductive-aged females, maternal mortality, infant mortality, and fetal mortality. Additionally, we were able to examine not just the cumulative number of laws to create a supportive or restrictive abortion regulation environment, but also the effect of individual types of laws on each type of mortality. In our primary analysis, we examined laws as either present or absent in the year that they went into effect. Abortion restrictions may have a delayed effect, particularly on infant mortality; thus, this may have reduced the estimated effect size of the law.

Our study is limited by factors inherent to large database studies, and our findings should be considered in light of those limitations. Maternal mortality can be difficult to ascertain accurately through large databases, but we used validated methodology to identify maternal deaths using this data source. Our study period ranged from 2000 to 2019; therefore, we could not examine the association between the Supreme Court decision in June 2022 and mortality. However, it is important to note that the mere presence of trigger laws that were not yet enforced was associated with a substantial increase in all-cause mortality in reproductive-aged females, maternal mortality, and infant mortality. Although data will not be available for several more years, it will be important to study whether mortality rates will increase further with implementation of these laws. Deaths were assigned to the state in which they occurred rather than examining the state of residence for decedents. Fetal mortality was included as an outcome and restricted to deaths beyond 20 weeks of gestation; this methodology does not account for fetal deaths related to termination of pregnancy before 20 weeks of gestation. Because we could not differentiate between spontaneous fetal death and termination of pregnancy after 20 weeks of gestation, this may account for the lack of difference in fetal death rates between supportive and restrictive states. However, we thought it significant that allowing abortion at any gestational age did not substantially increase fetal death rates compared with states that restricted termination after a certain gestational age. Finally, although we controlled for the general health of the population by state, residual confounding may exist, influencing our point estimates.

In conclusion, although no difference was detected in all-cause mortality in reproductive-aged females by restrictive legislation, certain types of legislation (trigger laws, limiting access to medication abortion, and parental consent laws) were associated with higher rates of all-cause mortality in reproductive-aged females. States restrictive of abortion access have higher rates maternal mortality and infant death, even after adjusting for the general health of the population. Although the relationship between abortion legislation and mortality rates is undoubtedly complex, repealing certain types of laws may decrease all-cause mortality in reproductive-aged females, maternal mortality, and infant mortality. Additionally, states with restrictive abortion laws should consider proven countermeasures to offset increased mortality rates, including expansion of Medicaid.24,25

REFERENCES

1. Cai W, Johnston T, McCann A, Schoenfeld Walker A. Half of US women risk losing abortion access without Roe. Accessed May 7, 2022. https://www.nytimes.com/interactive/2022/05/07/us/abortion-access-roe-v-wade.html
2. Nash E. Abortion rights in peril—what clinicians need to know. N Engl J Med 2019;381:497–9. doi: 10.1056/nejmp1906972
3. Vricella LK, Gawron LM, Louis JM. Executive summary: reproductive services for women at high risk for maternal mortality workshop, February 11–12, 2019, Las Vegas, Nevada. Am J Obstet Gynecol 2019;220:B2–12. doi: 10.1016/j.ajog.2019.02.011
4. Berer M. Abortion law and policy around the world: in search of decriminalization. Health Hum Rights 2017;19:13–27.
5. Pabayo R, Ehntholt A, Cook DM, Reynolds M, Muennig P, Liu SY. Laws restricting access to abortion services and infant mortality risk in the United States. Int J Environ Res Public Health 2020;17:3773. doi: 10.3390/ijerph17113773
6. Centers for Disease Control and Prevention, National Center for Health Statistics. CDC WONDER: multiple cause of death, 1999-2000 request, deaths occurring through 2020. Accessed July 7, 2022. http://wonder.cdc.gov/mcd-icd10.html
7. Centers for Disease Control and Prevention, National Center for Health Statistics. CDC WONDER: fetal deaths, 2005-2020 request. Accessed July 10, 2022. http://wonder.cdc.gov/fetal-deaths-current.html
8. Centers for Disease Control and Prevention, National Center for Health Statistics. CDC WONDER: linked birth/infant death records, 1999-2002 request. Accessed July 10, 2022. http://wonder.cdc.gov/lbd-v2002.html
9. Centers for Disease Control and Prevention, National Center for Health Statistics. CDC WONDER: Linked birth/infant death records, 2003-2006 request. Accessed July 10, 2022. http://wonder.cdc.gov/lbd-v2006.html
10. Centers for Disease Control and Prevention, National Center for Health Statistics. CDC WONDER: linked birth/infant death records, 2007-2019 request. Accessed July 10, 2022. http://wonder.cdc.gov/lbd-current.html
11. von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. PLoS Med 2007;4:e296. doi: 10.1371/journal.pmed.0040296
12. Horon IL. Underreporting of maternal deaths on death certificates and the magnitude of the problem of maternal mortality. Am J Public Health 2005;95:478–82. doi: 10.2105/ajph.2004.040063
13. Hoyert DL. Maternal mortality rates in the United States, 2020. National Center for Health Statistics; 2022. https://dx.doi.org/10.15620/cdc:113967
14. Hoyert DL, Uddin SFG, Miniño AM. Evaluation of the pregnancy status checkbox on the identification of maternal deaths. Natl Vital Stat Rep 2020;69:1–25.
15. Moaddab A, Dildy GA, Brown HL, Bateni ZH, Belfort MA, Sangi-Haghpeykar H, et al. Health care disparity and pregnancy-related mortality in the United States, 2005-2014. Obstet Gynecol 2018;131:707–12. doi: 10.1097/aog.0000000000002534
16. Nash E. State abortion policy landscape: from hostile to supportive. 2018. Accessed August 26, 2019. https://www.guttmacher.org/article/2019/08/state-abortion-policy-landscape-hostile-supportive
17. Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes. Biometrics 1986;42:121–30. doi: 10.2307/2531248
18. Højsgaard S, Halekoh U, Yan J. The R package geepack for generalized estimating equations. J Stat Softw 2006;15:1–11. doi: 10.18637/jss.v015.i02
19. Diggle PJ, Heagerty P, Liang K-Y, Zeger SL. Analysis of longitudinal data. 2nd ed. Oxford University Press; 2002.
20. Centers for Disease Control and Prevention. Behavioral Risk Factor Surveillance System survey data. Accessed July 10, 2022. https://www.cdc.gov/brfss/annual_data/annual_data.htm
21. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Ser B (Methodological) 1995;57:289–300. doi: 10.1111/j.2517-6161.1995.tb02031.x
22. Addante AN, Eisenberg DL, Valentine MC, Leonard J, Maddox KEJ, Hoofnagle MH. The association between state-level abortion restrictions and maternal mortality in the United States, 1995-2017. Contraception 2021;104:496–501. doi: 10.1016/j.contraception.2021.03.018
23. Nambiar A, Patel S, Santiago-Munoz P, Spong CY, Nelson DB. Maternal morbidity and fetal outcomes among pregnant women at 22 weeks' gestation or less with complications in 2 Texas hospitals after legislation on abortion. Am J Obstet Gynecol 2022;227:648–50.e1. doi: 10.1016/j.ajog.2022.06.060
24. Bhatt CB, Beck-Sagué CM. Medicaid expansion and infant mortality in the United States. Am J Public Health 2018;108:565–7. doi: 10.2105/ajph.2017.304218
25. Eliason EL. Adoption of Medicaid expansion is associated with lower maternal mortality. Women's Health Issues 2020;30:147–52. doi: 10.1016/j.whi.2020.01.005

Supplemental Digital Content

© 2022 by the American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.