We found adverse associations in health outcomes for individuals who were incarcerated either directly before or upon hospital discharge from giving birth. Further, babies born to this population had higher rates of negative health outcomes that make them higher risk for both immediate and long-term complications.

In contrast to studies from a single facility, we included hospitals from across California, enhancing generalizability to the broad pregnant population affected by incarceration. This is an advantage over prior studies [12, 15, 22] that draw conclusions about birth outcomes from a single carceral facility making it difficult to generalize to the incarcerated population as a whole, in part because perinatal care varies between individual facilities and states. Prior ecological studies have examined regional incarceration rates and pregnancy outcomes or examined parental incarceration and outcomes, but have been limited by not accounting specifically for individual maternal incarceration [16, 23, 24].

The proportion of non-Hispanic Black people in the incarcerated population was 3.4 times greater than the non-incarcerated control population (17.0% vs. 5.0%), similar to the disparity in incarceration rates among non-Hispanic black people versus other groups seen in the general population [25]. The AI/AN racial group accounted for 1.8% of births in the incarcerated sample, which was 6 times greater than their representation among non-incarcerated controls and 4.5 times greater than the California AI/AN proportion of births (1.8% vs. 0.3% vs. 0.4%) [25]. The overrepresentation of births affected by incarceration to individuals who are Black and AI/AN reinforces other aspects of structural racism embedded in our legal system. Incarcerated individuals had higher rates of prenatal care paid for by the government and lower rates of higher education compared to non-incarcerated controls, which is unsurprising given the lives of many are afflicted with poverty and limited access to healthcare [5, 26, 27].

The roots of unequal incarceration of Black, indigenous, and economically disadvantaged people have been well described. The racial disparity in mass incarceration has been compared to slavery (1619–1865), the Jim Crow segregation in the South (1865–1965), and the urban ghetto in the North (1915–1968) [28]. The collateral damage of mass incarceration to Black / African American communities includes damage to social networks, distortion of social norms, and destruction of social citizenship [29]. As Dorothy Roberts noted, “we need to reconsider the meaning of reproductive liberty to take into account its relationship to racial oppression” [30].

Previous research shows parental incarceration is associated with child health problems such as asthma, migraines, high cholesterol, early grade retention, depression, anxiety, post-traumatic stress disorder, antisocial behavior, human immunodeficiency virus / acquired immunodeficiency deficiency syndrome, and poor health [31,32,33,34]. Health problems such as increased likelihood of worse mental and physical health extend into adulthood [35, 36]. We found that 81% of pregnant incarcerated individuals had prior children, which is higher than a prior Bureau of Justice Statistics report that found that 58% of females in prisons across the United States have minor children; this may indicate that some pregnant individuals affected by incarceration have older children who are no longer minors [37]. Prior to the COVID-19 pandemic, there were approximately 5 million U.S. children with at least one parent in prison at one time or another [38], and 77% of mothers in state prison who had lived with their children just prior to incarceration provided most of the children’s daily care [4]. The high frequency of incarcerated parents with children reminds us of the multigenerational ramifications of incarceration during pregnancy.

Individuals in prison or jail are more likely to have ever had a chronic condition, an infectious disease, a cognitive disability, or any disability [39, 40] consistent with our finding that compared to non-incarcerated controls, the rate of severe maternal morbidity was increased 1.4 times, the rate of placental abruption nearly double, and the rate of infections more than three times increased. While our findings suggest that incarcerated individuals have an increased likelihood of complex medical needs, our study troublingly found they were more likely to have later onset of prenatal care compared to controls, with one-third of incarcerated pregnant people initiating care in either the 2nd or 3rd trimesters. Such delay in care speaks to the instability of healthcare of individuals affected by incarceration, some of whom may enter into the carceral system, especially jail where the average length of stay is 26 days [2] and leave before childbirth but may still have deleterious effects from the experience [26].

The incarcerated population is much more socially at risk for poverty and other adverse life circumstances, with significant pediatric ramifications. Maternal perinatal depression and anxiety alone can lead to adverse childhood experience, and is associated with preterm birth, low birthweight, intrauterine growth restriction, as well as impaired social interaction, and delays in language, cognitive, and social-emotional development [41]. The developmental outcomes extend beyond infancy into childhood and adolescence [41, 42]. Complications from these conditions can have enduring ramifications for not only the birthing parent but also the offspring and entire family unit [43, 44]. Compared to non-incarcerated pregnant controls, the rate of mental health disorder was 6.9 times higher and substance use was 15.6 times higher in the incarcerated pregnant group. The ramifications for the immediate perinatal health of the birthing parent can be severe. A report from nine maternal mortality review committees found that mental health conditions and substance use disorders (SUDs) were linked to 12.9% and 8.2% of pregnancy-related deaths, respectively [45]. As discussed by Haffajee et al. [44], the number of states that have adapted punitive policies (including fines, years of jail time, and loss of custody) directed at pregnant individuals with SUDs has increased from 12 states in 2000 to 26 states in 2017. In California, mental health disorders and substance use are screened at intake, with follow-up evaluations and referrals made available to those in need for treatment. Policies and programs that can facilitate rehabilitation and treatment as an alternative to punitive measures for those with SUDs may help to optimize the health for both the pregnant individual and their child.

We found significantly higher rates of preterm births of babies born to incarcerated mothers than controls, where 87% of babies born to incarcerated mothers were term compared to 91% of babies in the control population. The rates for both populations exceeded California’s preterm birth rate of 8.5% in 2015 and the nation’s preterm birth rate of 9.85% in 2016 [25], which may reflect the nature of hospitals that tend to care for incarcerated individuals. We also observed higher rates of small for gestational age and low birth weight in babies born to incarcerated individuals than controls. In addition, we found significantly higher rates of NICU admission, reaching 13% for babies born to incarcerated mothers compared to 8% for controls. The discrepant rates of preterm births, small for gestational age, and NICU admissions persisted even when correcting for maternal age, multiple gestation, parity, smoking, and body mass index. We did not observe a significant difference in very low birthweight.

Findings from our study indicate incarceration has negative ramifications for birth outcomes and stands in contrast to other studies [7,8,9,10,11,12]. In fact, some prior studies have shown that incarceration has a protective effect on birth outcomes, especially when incarcerated mothers are compared to at-risk controls, such as non-jailed methadone-maintained women [10], women living in “high-risk” residential areas [12], or people with history of criminal conviction or drug use [9, 11], leading to speculation that incarceration may provide protective effects on pregnancy by providing shelter, food, healthcare, and reduction of substance misuse [7, 8]. Those prior studies indicate that directed efforts to optimize pregnancy healthcare during incarceration can have a positive impact. The population affected by incarceration had significantly higher rates of medical morbidities, mental health disorders and drug dependence (Table 2), indicating that public health investment in optimizing healthcare in these areas outside of the carceral system, may mitigate risk for these individuals.

The increased rates of maternal morbidities among women who are incarcerated have been previously described. Using the National Inpatient Sample from 2015 to 1018, Logue et al. [46] found that pregnancy complications associated with incarceration included non-transfusion severe maternal morbidity, hypertensive disorders of pregnancy, preterm delivery, placental abruption and postpartum hemorrhage. Similar to our findings, they found increased abruption among incarcerated pregnant women, and because the associations remained significant after adjustment for confounders, it is suggested that some contribution to maternal risk is conferred by incarceration. This study would not have had linkage to the vital records as in our study.

We excluded 61 California hospitals in order to minimize the possible bias encountered by hospitals that do not routinely admit incarcerated pregnant individuals. It is possible that this could lead to clusters of negative birth outcomes in selection bias. It is also possible that individual differences in risks of adverse birth outcomes, such as timing and duration of incarceration during pregnancy, play a role, but are not captured in administrative datasets. Finally, our findings may be related to the control population, a statewide population of non-incarcerated women in California. The 9.1% preterm birth rate in California ranks lower than most states and may have provided a control group healthier than other studies [47], some of which use control groups of populations similarly disadvantaged to incarcerated individuals, such as women who were incarcerated during the same time but not pregnant [48].

Scholars such as Angela Davis, Mariame Kaba, and Ruth Wilson Gilmore have stated that the incarceration system is designed to punish and dehumanize people [49,50,51]. Comprehensive progress may come from shifting away from investing in reforming our current carceral system and instead reimagining the justice system through centering the voices of people most impacted by incarceration, through rehabilitation and restoration, and through investing in underserved communities. In 2021, the Minnesota government enacted the Healthy Start Act that allows individuals who are serving short sentences in state prison to be released conditionally for the duration of their pregnancy up to the first year after birth [52].

Our study had several limitations. The carceral system in California may differ from other states. While California does not have prisons specifically designated for pregnant people, as many as 17 states have one specific facility that houses incarcerated pregnant individuals [5]. Female incarceration rates vary by state, with Texas having 178 women incarcerated per 100,000 residents, compared to 88 per 100,000 in California [53]. The dataset we used does not distinguish between facilities (i.e. jail, state, and federal prisons) and the exact duration and time of incarceration in relation to pregnancy is unknown. The disposition variable to determine admission to or discharge from the hospital in relation to the carceral system may not be well coded in healthcare facilities. We worked within the constraints of using birth certificate data and hospital discharge data, recognizing that both have limitations and are sometimes discordant; our prior experiences informed the study methods [54, 55]. Fetal birth certificate data were not included so pregnancies that ended before 20 weeks of gestation were not captured in this study. We could not definitively determine whether some maternal comorbidities, such as obesity, were present during the current pregnancy or if they were from an individual’s past medical history. The size of the dataset makes correcting for some confounding variables infeasible.