Abstract
Introduction
Abortion restrictions on pregnant people, such as those that limit gestational age or ban abortion altogether, have been associated with infant mortality, but little is known about whether laws restricting abortion providers and clinics providing abortions have similar associations. This study examines the association between laws that restrict abortion providers, known as TRAP (targeted regulations of abortion providers) laws, and infant mortality on a state level.
Methods
This cross‐sectional analysis was performed utilizing 2016–2021 Phase 8 data from the Pregnancy Risk Assessment Monitoring System (PRAMS). PRAMS data were linked with Guttmacher Institute data about state TRAP laws from 2016 to 2021. Infant death rates were calculated using PRAMS data of self‐reported infant deaths. Multivariable Poisson regression testing the association between the existence of TRAP laws in a state and infant mortality was performed, adjusting for individual‐level sociodemographic factors available in PRAMS. The exposure was defined as the existence of at least one TRAP law in a state during the study period.
Results
Among 178,210 participants, representing a weighted population of 9,099,586 birthing people, 47.5% people lived in states (n = 19) that had at least one TRAP law during the study period. Overall, 0.4% of participants reported an infant death. In adjusted analysis, there was a 36% greater likelihood of infant mortality (adjusted incident rate ratio [aIRR], 1.36; 95% confidence interval [CI], 1.13–1.63) in states with TRAP laws in comparison to those without TRAP laws.
Conclusions
TRAP laws are associated with a greater likelihood of infant mortality, even after accounting for potential individual‐level confounders. Restricted access to abortion may result in people with high‐risk maternal or fetal health conditions continuing pregnancy rather than receiving desired abortion services.
Keywords: abortion, health policy, infant mortality, neonatal mortality, reproductive rights, state policy, targeted regulation of abortion providers
1. INTRODUCTION
Infant mortality has generally been decreasing in the United States, yet there was an increase in infant mortality rates from 2021 to 2022 [1]. Infant mortality rates related to maternal complications and bacterial sepsis increased during this time, but no singular cause for this increase has been identified [1]. With ongoing restriction of abortion in the United States, more stringent restrictions on individuals’ access to abortion have also been associated with an increase in infant mortality [2, 3, 4]. Abortion restrictions may act to increase neonatal mortality through multiple pathways, including continuation of pregnancies affected by congenital anomalies, maternal high‐risk health conditions, or limited access to resources [2, 4].
In contrast to restrictions placed on pregnant people with regard to abortion access, targeted regulation of abortion providers (TRAP) laws are medically unindicated restrictions on clinicians or healthcare facilities that limit the provision of abortion, even in states where abortion is legal [5]. TRAP laws have resulted in the closure of clinics providing abortions, more limited access to abortion services, and adverse neonatal and maternal outcomes [6, 7]. TRAP laws, in addition to other abortion restrictions, exacerbate perinatal disparities [7, 8]. Given that the effect of TRAP laws on a smaller scale, is similar to post‐Dobbs policies—including closure of abortion clinics and increased travel time to receive care—studying TRAP laws may inform an understanding of abortion bans in the post‐Dobbs era [7]. Thus, we aimed to evaluate the association of the existence of state‐level TRAP laws and state rates of infant mortality.
2. METHODS
This cross‐sectional analysis was performed utilizing 2016–2021 Phase 8 data from the Pregnancy Risk Assessment Monitoring System (PRAMS), a surveillance system by the Centers for Disease Control and Prevention. PRAMS data were linked with archived Guttmacher Institute data about state TRAP laws from 2016 to 2021 (Supporting Information Section I). PRAMS is a mixed‐mode, representative surveillance system exploring maternal and child health indicators and linked to birth certificate data [9]. Forty states and the District of Columbia were included in PRAMS 2016–2021 data (Supporting Information Section II).
The exposure, existence of a TRAP law in each state, was defined as a state having at least one TRAP law in place during the 2016–2021 study period (Supporting Information Section III). TRAP laws include regulations about where abortion can occur, requirements (e.g., procedure room size, corridor width, distance to hospital) for facilities where abortions are provided, and clinician requirements for people providing abortions [5]. This study occurred prior to the Supreme Court decision in Dobbs v Jackson Women's Health Organization. Thus, although some states at this time enforced restrictions on abortion (e.g., gestational age limitations), no states could legally ban abortion. TRAP laws, therefore, were a strategy to further restrict abortion in the absence of the ability to enact abortion bans [10].
Incidence of infant mortality was calculated using PRAMS data, which uses respondent self‐report. The PRAMS questionnaire is administered only to people who recently gave birth to a live‐born infant. In the PRAMS questionnaire, participants are asked “Is your baby alive now?” If participants answer “no,” they are coded as having an infant death. The terminology “infant death” was chosen because the PRAMS survey does not collect information about the timing of infant death and thus cannot differentiate between a neonatal death (death at 0 to 27 days of life) and a post‐neonatal death during infancy (death at 28 days to 1 year of life). This measure only captures those who experienced an infant death prior to completing the PRAMS survey, which occurs between two and six months postpartum. Covariates available in PRAMS and selected for this analysis included infant birth year, maternal age, race and ethnicity, household income, educational attainment, marital status, preconception uninsurance, Medicaid coverage at time of delivery, maternal health conditions (prepregnancy body mass index [BMI], pregestational diabetes, gestational diabetes, chronic hypertension, hypertensive disorders of pregnancy), and neonatal outcomes (low birth weight, preterm birth). Respondent household income was the only covariate with more than 5% missing (9.0%) and was regression imputed (Supporting Information Section I).
Multivariable Poisson regression testing the association between the existence of TRAP laws in a state and incidence of infant mortality was performed, adjusting for all the above individual‐level sociodemographic factors (Supporting Information Section I). Two sensitivity analyses were performed (Supporting Information Section I). We first examined the possibility that other state‐level policies may confound our results and included other state‐level abortion restrictions, Medicaid expansion statutes, and state expenditures in our regression (Supporting Information Section IV). Subsequently, we evaluated the association between TRAP laws and markers of neonatal morbidity including low birth weight, preterm birth, and infant transfer. Statistical analysis was performed utilizing population‐weighted data using the complex survey module of Stata 18. This study was deemed exempt from our Institutional Review Board as it utilized publicly available, deidentified data.
3. RESULTS
Among 178,210 participants, representing a weighted population of 9,099,586, 47.5% people lived in states (n = 19) that had at least one TRAP law during the study period (Supporting Information Section V). States with at least one TRAP law in place during the study period had more respondents identifying as non‐Hispanic Black or non‐Hispanic White and fewer respondents who identified as Hispanic. States with at least one TRAP law also had fewer respondents with an annual income greater than $85,000, lower rates of more than 12 years of education among birthing people, higher rates of smoking during pregnancy, and higher rates of respondents with a BMI of 30 or greater (Table 1).
TABLE 1.
Sociodemographic characteristics for respondents living in states with and without TRAP laws.
|
No TRAP laws (%) n = 93,469, weighted n = 4,779,711 |
At least 1 TRAP law (%) n = 77,377 weighted n = 4,319,875 |
|
|---|---|---|
| Sociodemographic characteristics by state TRAP a law policy | ||
| Infant death | 0.3 | 0.5 |
| Infant birth year | ||
| 2016 | 14.2 | 14.2 |
| 2017 | 16.6 | 15.2 |
| 2018 | 18.4 | 18.5 |
| 2019 | 17.9 | 17.9 |
| 2020 | 15.8 | 18.8 |
| 2021 | 17.2 | 15.5 |
| Age (years) | ||
| Less than 20 | 3.2 | 4.3 |
| 20–29 | 43.2 | 50.4 |
| 30–39 | 49.7 | 42.7 |
| 40–45 | 33.9 | 2.6 |
| Race and ethnicity b | ||
| American Indian | 0.7 | 0.5 |
| Asian | 6.9 | 3.3 |
| Black | 12.7 | 16.9 |
| Hispanic | 17.1 | 9.1 |
| Multiple races or “other” race | 3.3 | 3.0 |
| Native Hawaiian or Other Pacific Islander | 0.2 | 0.1 |
| White | 59.1 | 67.1 |
| Annual household income | ||
| ≤$20,000 | 19.3 | 23.6 |
| $20,001–$40,000 | 18.0 | 19.0 |
| $40,001–$60,000 | 11.9 | 12.9 |
| $60,001–$85,000 | 10.3 | 11.6 |
| >$85,000 | 40.4 | 32.9 |
| Educational attainment | ||
| Less than 12 years | 9.4 | 9.6 |
| 12 years | 22.7 | 25.5 |
| More than 12 years | 68.0 | 64.8 |
| Married b | 65.2 | 61.7 |
| Uninsured preconception | 10.2 | 11.2 |
| Medicaid coverage at the time of delivery | 39.0 | 40.7 |
| Maternal health conditions | ||
| Prepregnancy BMI (kg/m2) | ||
| <18.5 | 3.3 | 3.2 |
| 18.5–24 | 46.3 | 43.4 |
| 25–29 | 26.2 | 25.3 |
| >29 | 24.2 | 28.2 |
| Smoking during pregnancy | 5.5 | 8.7 |
| Pregestational diabetes | 3.3 | 3.0 |
| Gestational diabetes | 9.4 | 9.6 |
| Chronic hypertension | 5.3 | 6.5 |
| Hypertensive disorders of pregnancy | 12.9 | 14.0 |
| Neonatal outcomes | ||
| Low birth weight c | 6.0 | 6.5 |
| Preterm birth d | 7.5 | 8.2 |
Note: PRAMS data are Phase 8 from 2016 to 2021; states excluded from PRAMS were California, Connecticut, Florida, Idaho, Nevada, North Carolina, Ohio, Oklahoma, South Carolina, and Texas. Vermont and Alaska were included in PRAMS but excluded from analysis because of non‐reporting of certain sociodemographic data.
Abbreviations: BMI, body mass index; PRAMS, Pregnancy Risk Assessment Monitoring System; TRAP, targeted regulation of abortion providers.
TRAP laws represent author's analysis of Guttmacher data about TRAP laws.
Race and ethnicity data and categorization and maternal marital status in PRAMS are derived from linked National Vital Statistics System birth certificate data.
Birth weight less than 2500 g.
Delivery at less than 37 weeks’ gestational age.
During the 2016–2021 study period, 0.4% of participants reported an infant death. The respondent‐reported infant death rate was 0.3% in states with no TRAP laws and 0.5% in states with at least one TRAP law. In adjusted analysis, there was a 36% higher likelihood of infant mortality (adjusted incident rate ratio [aIRR], 1.36; 95% confidence interval [CI], 1.13–1.63) in states with TRAP laws in comparison to those without TRAP laws (Table 1). Results from the fully specified model can be found in Supporting Information Section VI. The sensitivity analysis to assess the association between other state‐level policies and infant death produced statistically significant results consistent with the primary analysis (Supporting Information Section VII). The sensitivity analysis to assess the association between TRAP laws and markers of neonatal morbidity was not statistically significant (Supporting Information Section VIII).
4. DISCUSSION
In this cross‐sectional analysis of representative US data from PRAMS, we demonstrated that TRAP laws are associated with a higher likelihood of infant mortality, even after accounting for individual‐level social and demographic risk factors that may act as potential confounders. Results from sensitivity analyses support our findings.
In literature on patient‐level restrictions to comprehensive reproductive health care, restricted abortion access has been associated with an increase in infant mortality [2, 3, 4, 11]. Our study adds to this literature by evaluating TRAP laws, which function to restrict abortion by placing medically unnecessary burdens on abortion clinics and providers, resulting in the closure of facilities that provide abortions. Other studies have hypothesized that restrictions in abortions may result in people with pregnancies affected by congenital anomalies or with high‐risk maternal health conditions continuing pregnancies, which may be a mechanism underlying this finding as well [2, 12]. While we did not find an association between markers of neonatal morbidity and TRAP laws in our sensitivity analysis, further examination of this potential consequence of abortion restriction is needed. We did not examine the association between TRAP laws and congenital anomalies because of the poor sensitivity of PRAMS and birth certificate data for congenital anomalies [13, 14].
This study has limitations to note. First, this study does not employ quasi‐experimental methods, therefore causal inference cannot be implied. Because causal inference cannot be assessed through this study, there is also a possibility of reverse temporality, in which states with worse perinatal outcomes are more likely to pass TRAP laws. Research has demonstrated that states with the most restrictive abortion policies have less access to maternal and family social services [15]. However, the addition of TRAP laws could further exacerbate this relationship. While we controlled for well‐known individual‐level factors including measures of socio‐economic status and maternal and infant health characteristics, there may exist other state‐level characteristics that influence infant mortality. PRAMS is a retrospective survey given at two to six months postpartum and thus may be subject to selection bias if people who experienced an infant death are less likely to complete the questionnaire; however, such bias would bias the findings toward the null. PRAMS does not include data describing the cause of infant deaths. While PRAMS is open to all states, there were 10 states that were excluded from PRAMS during our study period.
Abortion restrictions unequally affect people with lower socioeconomic status, a group with already higher rates of infant mortality and increased risk of adverse perinatal outcomes in the setting of reproductive rights restrictions [15, 16]. Thus, access to abortion and other reproductive health services is a health equity and racial justice issue. In the United States, it is well established that infants who are racially minoritized have an increased risk of mortality within the first year of life in comparison to infants who are white; this disparity may be even further exacerbated by the disproportionate burden that abortion restrictions place on Black birthing people [8].
In conclusion, we identified an association between the existence of state TRAP laws and a higher likelihood of infant mortality. This pre‐Dobbs study demonstrates one possible mechanism by which limitations on reproductive justice are associated with harm. Further studying this association will be important to understand how abortion restrictions, even in the absence of outright bans, may be associated with adverse outcomes.
AUTHOR CONTRIBUTIONS
Madeline F. Perry: Conceptualization; investigation; writing—original draft; methodology; formal analysis. Joe Feinglass: Conceptualization; writing—review and editing; methodology; supervision. Lynn M. Yee: Conceptualization; methodology; writing—review and editing; supervision.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflicts of interest.
FUNDING INFORMATION
The authors received no specific funding for this work.
ETHICS STATEMENT
This study was deemed exempt from our Institutional Review Board as it utilized publicly available, deidentified data.
Supporting information
Supporting File 1: pmf270320‐sup‐0001‐SuppMat.pdf
ACKNOWLEDGMENTS
We thank the PRAMS Working Group, which includes the PRAMS Team, Division of Reproductive Health, CDC, and the following PRAMS sites, for their role in conducting PRAMS surveillance and allowing the use of their data.
This abstract was presented as a poster presentation to the 45th Annual Meeting for the Society of Maternal‐Fetal Medicine, January 27–February 1, 2025, Aurora, CO.
This paper is being exclusively submitted to Pregnancy and has not been previously published. The work behind this paper meets applicable ethical standards. All authors are willing to take public responsibility for the work and satisfy the requirements of authorship.
DATA AVAILABILITY STATEMENT
The Pregnancy Risk Assessment Monitoring System is publicly available data.
REFERENCES
- 1. Ely, D. M. , and Driscoll A. K.. Infant Mortality in the United States: Provisional Data From the 2022 Period Linked Birth/Infant Death File . Hyattsville, MD: National Center for Health Statistics. https://www.cdc.gov/nchs/data/vsrr/vsrr033.pdf. [PubMed]
- 2. Gemmill, A. , Margerison C. E., Stuart E. A., and Bell S. O.. 2024.“Infant Deaths After Texas' 2021 Ban on Abortion in Early Pregnancy.” JAMA Pediatrics 178(8): 784–91. 10.1001/jamapediatrics.2024.0885. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Karletsos, D. , Stoecker C., Vilda D., Theall K. P., and Wallace M. E.. 2021. “Association of State Gestational Age Limit Abortion Laws With Infant Mortality.” American Journal of Preventive Medicine 61(6): 787–94. 10.1016/j.amepre.2021.05.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Burdick, K. J. , Coughlin C. G., D'Ambrosi G. R., Monuteaux M. C., Economy K. E., Mannix R. C., and Lee L. K.. 2024. “Abortion Restrictiveness and Infant Mortality: An Ecologic Study, 2014–2018.” American Journal of Preventive Medicine 66(3): 418–26. 10.1016/j.amepre.2023.10.010. [DOI] [PubMed] [Google Scholar]
- 5. Targeted Regulation of Abortion Providers. https://www.guttmacher.org/state‐policy/explore/targeted‐regulation‐abortion‐providers.
- 6. Cunningham, S. D. , Lindo J. M., Myers C., and Schlosser A.. “How Far Is Too Far? New Evidence on Abortion Clinic Closures, Access and Abortions.” Working Paper No. 23366, NBER, 2017.
- 7. Gardner, G. 2024. “The Maternal and Infant Health Consequences of Restricted Access to Abortion in the United States.” Journal of Health Economics 98: 102938. 10.1016/j.jhealeco.2024.102938. [DOI] [PubMed] [Google Scholar]
- 8. Kozhimannil, K. B. , Hassan A., and Hardeman R. R.. 2022. “Abortion Access as a Racial Justice Issue.” New England Journal of Medicine 387(17): 1537–9. 10.1056/NEJMp2209737. [DOI] [PubMed] [Google Scholar]
- 9. Shulman, H. B. , D'Angelo D. V., Harrison L., Smith R. A., and Warner L.. 2018. “The Pregnancy Risk Assessment Monitoring System (PRAMS): Overview of Design and Methodology.” American Journal of Public Health 108(10): 1305–13. 10.2105/ajph.2018.304563. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Mercier, R. J. , Buchbinder M., and Bryant A.. 2016. “TRAP Laws and the Invisible Labor of US Abortion Providers.” Critical Public Health 26(1): 77–87. 10.1080/09581596.2015.1077205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Gemmill, A. , Franks A. M., Anjur‐Dietrich S., Ozinsky A., Arbour D., Stuart E. A., Ben‐Michael E., Feller A., and Bell S. O.. 2025. “US Abortion Bans and Infant Mortality.” JAMA 333(15): 1315–23. 10.1001/jama.2024.28517. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Singh, P. and Gallo M. F.. 2024. “National Trends in Infant Mortality in the US After Dobbs.” JAMA Pediatrics 178(12): 1364–6. 10.1001/jamapediatrics.2024.4276. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Salemi, J. L. , Tanner J. P., Sampat D. P., Rutkowski R. E., Anjohrin S. B., Marshall J., and Kirby R. S.. 2017. “Evaluation of the Sensitivity and Accuracy of Birth Defects Indicators on the 2003 Revision of the U.S. Birth Certificate: Has Data Quality Improved?.” Paediatric and Perinatal Epidemiology 31(1): 67–75. 10.1111/ppe.12326. [DOI] [PubMed] [Google Scholar]
- 14. Snell, L. M. , Little B. B., Knoll K. A., Johnston W. L. Jr., Rosenfeld C. R., and Gant N. F.. 1992. “Reliability of Birth Certificate Reporting of Congenital Anomalies.” American Journal of Perinatology 9(3): 219–22. 10.1055/s-2007-999325. [DOI] [PubMed] [Google Scholar]
- 15. Madden, N. , Trawick E., Watson K., and Yee L. M.. 2024. “Post‐Dobbs Abortion Restrictions and the Families They Leave Behind.” American Journal of Public Health 114(10): 1043–50. 10.2105/ajph.2024.307792. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Wallace, M. E. , Evans M. G., and Theall K.. 2017. “The Status of Women's Reproductive Rights and Adverse Birth Outcomes.” Women's Health Issues 27(2): 121–8. 10.1016/j.whi.2016.12.013. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting File 1: pmf270320‐sup‐0001‐SuppMat.pdf
Data Availability Statement
The Pregnancy Risk Assessment Monitoring System is publicly available data.