Abstract
Executive Summary
The United States Immigration and Customs Enforcement (ICE) detains hundreds of thousands of noncitizens each year in the world’s largest immigration detention system. A recent proliferation of federal policies and funding has accelerated detention and deportation to unprecedented levels, increasing exposure to immigration enforcement across the US immigrant population. Yet despite this mass expansion, we know very little about how detention experiences may expose immigrants to unique vulnerabilities that worsen their health even after deportation. Using representative, repeated cross-sectional survey data from Central American immigrants deported from the United States between 2017 and 2019 and surveyed in the Survey of Migration in the Southern Border of Mexico (EMIF Sur), this study examines the extent to which one dimension of immigration detention — experiences of abuse and mistreatment while in the custody of US immigration authorities — is associated with self-rated health immediately following deportation. We find that experiencing physical abuse, verbal abuse, and perceived poor treatment by immigration authorities is associated with worse health among recently deported Guatemalan, Honduran, and Salvadoran immigrants. These findings suggest that immigration detention operates as a transnational carceral exposure, with health harms that extend beyond confinement and across national borders, with the potential to shape post-deportation well-being in contexts of return marked by precarity and limited access to health care.
Our findings have important policy implications for protecting the health and well-being of current and formerly detained populations. As immigration detention rapidly expands, identifying strategies to mitigate its harms has become increasingly urgent. We offer the following alternatives to mitigate the health harms of detention in the United States, as well as recommendations for receiving countries to identify detention experiences as a significant health risk factor among the deported population:
In the United States
End the use of detention in civil immigration proceedings. Reform federal immigration law to better align with economic and human realities by expanding lawful migration pathways and reducing the grounds for deportation. Prioritize release from detention, including access to bond hearings for all detained migrants. Formally evaluate alternatives to detention that do not exacerbate harms. Adopt state and local policy reforms aimed to limit exposure to the deportation pipeline. Implement administrative oversight reforms to the detention system.
In countries of return
Strengthen procedures within government agencies and NGOs in countries of return to systematically document harms experienced during immigration detention. Integrate screening for immigration detention and deportation histories into public health systems to recognize these processes as a durable social determinant of health. Identify deported individuals at elevated health risk and facilitate timely connections to appropriate care and support services at the earliest possible stage of return.
Introduction
The United States operates the world’s largest immigration detention and deportation system, detaining more than eight million immigrants between 2001 and 2025 and deporting up to 1 percent of the entire foreign-born population annually between 1996 and 2022 (Patler and Jones 2025). Despite the already massive scale of immigration enforcement, recent federal policy changes have dramatically expanded detention and deportation. The Trump administration has issued dozens of Executive Orders aimed at broadening immigration enforcement authority, deputized more than 1,000 local law enforcement agencies to enforce federal immigration law, and stripped hundreds of thousands of individuals of lawful protections such as Temporary Protected Status (TPS), rendering them newly deportable (Immigration Policy Tracking Project 2025; Kocher 2025; Patler and Jones 2025). Congress has reinforced this expansion through budget reconciliation legislation that nearly triples Immigration and Customs Enforcement (ICE)’s operating budget, as well as through passage of the Laken Riley Act, which mandates indefinite detention for immigrants convicted — or merely accused — of low-level offenses such as shoplifting (Chishti and Lacarte 2025). In parallel, the government has imposed a sweeping rule requiring indefinite detention for all immigrants who entered the United States without inspection — regardless of criminal history or individual circumstances — who comprised 62 percent of new immigration court cases in 2025 alone (Cassler 2025). Predictably, by late-January 2026, the average daily population in ICE detention had exceeded 70,500, the highest in the system’s modern history (Transactional Records Access Clearinghouse [TRAC] 2026).
Despite the rapid expansion of the US immigration detention system, research on its collateral consequences — especially its health consequences — remains limited, largely because ICE does not release detailed or comprehensive data on the populations it detains, and facility oversight is limited (Diaz et al. 2023a; Whitelaw et al. 2025a). Still, a growing body of research has begun to document the potential health harms associated with detention conditions, including overcrowded, unsafe, and unsanitary facilities; poor nutrition; abuse by facility guards and staff; and inadequate access to medical and mental health care (Saadi et al. 2020; Saadi, Patler, and De Trinidad Young 2022; Diaz et al. 2023b). For example, one study found that inhumane treatment by guards and facility staff can produce enduring harm, with detained immigrants describing detention as leaving lasting “mental scars” that persist long after release (Diaz et al. 2023b).
Yet existing scholarship on immigration detention and health remains limited in several respects. First, due largely to barriers to data access, few studies draw on representative samples of detained or formerly detained immigrants. In addition, most research focuses on health during detention or shortly after release on bond within the United States. Far less is known about how detention experiences may continue to shape health after deportation, when individuals are returned to contexts marked by economic precarity, disrupted social ties, and limited access to health-supporting institutions (Waldman 2022; Bakely et al. 2023). As a result, the role of immigration detention in stratifying post-deportation health outcomes remains poorly understood. More broadly, existing research has yet to conceptualize how exposure to immigration detention may generate harms that extend beyond the period of confinement and across national borders following deportation.
In this article, we address these and other gaps in the literature using representative, repeated cross-sectional data from the 2017–2019 waves of the Survey of Migration in the Southern Border of Mexico (EMIF Sur). We examine one critical dimension of the US immigration detention experience: the association between abuse and poor treatment while in the custody of US immigration authorities and post-deportation health among recently deported Central American immigrants from El Salvador, Guatemala, and Honduras. Using population-representative data also allows us to overcome a central limitation of much prior work in this area. Further, by examining health outcomes immediately following deportation, our study extends existing research beyond the period of confinement or release in the United States. Moreover, our focus on Central American immigrants is both substantively and empirically important. Migrants from El Salvador, Guatemala, and Honduras accounted for more than one in four (26 percent) undocumented immigrants in the United States as of 2023 (Gelatt, Ruiz-Soto, and Bachmeier 2025) and have been disproportionately targeted by contemporary US immigration enforcement (Franco, Patler, and Reiter 2022; Pedroza 2022). These populations also experience distinct migration trajectories shaped by violence, economic displacement, and family separation, yet remain underrepresented in research on detention and health. To our knowledge, our study is among the first to examine how US detention experiences are associated with post-deportation health outcomes among Central American immigrants.
Our analyses show that punitive detention experiences — including physical abuse, verbal abuse, and perceived poor treatment while in the custody of US immigration authorities — are associated with poorer self-rated health immediately following deportation. Taken together, these findings suggest that US immigration detention should be conceptualized as a form of transnational carceral exposure. By this term, we refer to institutional contact with the immigration detention system that produces harms extending beyond the moment of confinement and beyond US territorial borders, shaping migrants’ health and well-being after deportation. For deported migrants, detention represents a coercive institutional encounter whose consequences may travel with individuals into the social, economic, and health contexts to which they are deported. For example, during the height of the COVID-19 pandemic, the United States continued deporting immigrants who had been exposed to the coronavirus while detained — often without a treatment plan — which endangered the health of deported migrants and their communities upon return (Kerwin 2020). In this sense, the effects of detention do not end at the point of release but instead accompany migrants into their post-return lives. Conceptualizing immigration detention in this way bridges emerging research on immigration enforcement with scholarship on the long-term health consequences of incarceration, highlighting how carceral state practices structure health well beyond formal custody. This framing positions immigration detention as a punitive institutional exposure and an upstream determinant of health that may stratify post-deportation outcomes across migrants’ countries of origin. It thus allows for closer examination of the distinct and cumulative harms of the detention and deportation system.
These findings reveal the grave implications of a rapidly expanding system and carry immediate relevance for policymakers, advocates, and public health practitioners in both the United States and migrants’ countries of origin. In the United States, the renewed expansion of large-scale and punitive enforcement has intensified efforts to identify strategies to mitigate the harms of immigration detention; our results provide empirical evidence for these debates. At the same time, as deportations continue at historically high levels, governments in countries of origin are increasingly seeking ways to support the reintegration, resettlement, and well-being of returning citizens (Golash-Boza 2015; Hagan and Wassink 2020; Mora and Ruiz Soto 2025). This study provides evidence to inform such efforts by highlighting immigration detention as a consequential upstream determinant of health among the deported population.
Background
In this section, we review a multidisciplinary body of literature that identifies exposure to carceral settings as a key determinant of durable individual-level health outcomes. We then bridge this literature with research on the US immigration system, examining how the conditions and practices of immigration detention mirror the punitive nature of criminal incarceration and may produce similarly deleterious health effects that can extend across borders. Finally, we provide background on Central American immigrants and situate this growing yet understudied population as an important case for understanding detention experiences and post-deportation health outcomes.
Incarceration and Health
The expansion of mass incarceration since the mid-1970s has become a powerful stratifying institution in American society and a key driver of racialized health inequities (Garland 2001; Alexander 2010; Wakefield and Uggen 2010; Wildeman and Muller 2012). A substantial body of research shows that exposure to carceral settings harms individual health, increasing the risk of morbidity and mortality among incarcerated populations, including immigrants (Massoglia and Pridemore 2015; Zajdel and Patterson 2024; Patler, forthcoming). Incarcerated individuals experience disproportionately higher rates of chronic physical conditions, infectious diseases, poorer mental health, and other stress-related illnesses (Massoglia 2008; Binswanger, Krueger, and Steiner 2009; Schnittker, Massoglia, and Uggen 2012; Turney, Wildeman, and Schnittker 2012; Schnittker 2014; Udo 2019). Research on incarceration and mortality further indicates that imprisonment is associated with premature adult mortality (Massoglia et al. 2014; Massoglia and Remster 2019). Highlighting the enduring toll of incarceration, an analysis of New York State parole data from 1989 to 2003 found that each additional year spent in prison increased the likelihood of death among parolees by 15.6 percent (Patterson 2013).
Carceral settings harm health through multiple pathways (see Daza, Palloni, and Jones 2020), including the conditions and exposures within the prison environment itself (Brinkley-Rubinstein 2013). Prisons deprive incarcerated persons of fundamental human rights and needs, producing physical, mental, and social harms (Viggiani 2007). Incarcerated individuals are typically confined in remote, overcrowded, and unsanitary facilities that easily facilitate the spread of infectious diseases. They are also exposed to chronic stressors — such as harsh disciplinary practices and violence or abuse — that collectively contribute to adverse health outcomes, even after release (Brinkley-Rubinstein 2013; Massoglia and Pridemore 2015).
Immigration Detention and Health
As mass incarceration has expanded and US criminal law has increasingly converged with the immigration legal system — a phenomenon commonly referred to as crimmigration (see Menjívar, Gómez Cervantes, and Alvord 2018) — the confinement of noncitizens in the US immigration detention system has come to closely resemble penal incarceration in ways that are likely to generate similar health harms. Although immigration detention is formally defined as a civil process intended to facilitate removal rather than punishment (Schriro 2015), in practice, the US immigration system often operates through punitive logics and practices that mirror those of mass incarceration (Diaz et al. 2023a). Despite this legal distinction, detained immigrants are frequently constructed as threats to public safety and confined in harsh, carceral environments — often described as immigration prisons (García Hernández 2014, 2017), despite immigration law’s civil definition. The social and environmental conditions in these facilities commonly include severe overcrowding; unsafe and unsanitary environments; restricted access to showers; food and water that are inedible or unsafe; exposure to physical, verbal, sexual, and psychological abuse; and persistent barriers to necessary physical and mental health care (Ahmed, Appelbaum, and Jordan 2017; Saadi, Patler, and De Trinidad Young 2022; Diaz et al. 2023b; Cuevas et al. 2024).
Moreover, facilities subcontracted by ICE are often US jails themselves, or they are built and operated to function like jails by private, for-profit correctional corporations. Many of these facilities are located in remote areas, further limiting visitation and reducing contact with loved ones (Patler and Branic 2017). Staffing structures and population management strategies likewise replicate penal models of command and control, including the use of solitary confinement as a punitive measure (Schriro 2009; García Hernández 2014; Patler, Sacha, and Branic 2018; Franco, Patler, and Reiter 2022). Consistent with these practices, numerous reports document abuse and mistreatment in the custody of US immigration authorities. A recent study of detention facilities in Florida found that 72 percent of complaints submitted between December 2017 and May 2023 involved reports of physical or mental abuse, often perpetrated by facility officers (Cuevas et al. 2024). These complaints described a wide range of detention-related abuses, including physical violence (e.g., pepper spraying, beatings, and use of tasers), incidents of sexual harassment, and verbal and emotional abuse (e.g., threats of deportation or solitary confinement and racial slurs). Alarmingly, these issues occur within a broader context of limited oversight and accountability. Persistent deficiencies in ICE oversight, along with the operation of for-profit detention contractors, have contributed to preventable health emergencies and deaths in custody (Kelly 2018; Grassini et al. 2021; Dreisbach 2023; Whitelaw et al. 2025b).
In addition to conditions of confinement, the duration and structure of immigration detention may further intensify health risks. Although some immigrants are rapidly deported, average detention lengths are often comparable to those in state and local criminal facilities, even though most detained immigrants have no criminal record. At Adelanto — one of the largest detention facilities in the country — average detention lengths ranged from 100 to 200 days between 2013 and 2023 (TRAC 2024). Similarly, 15 percent of Mexican nationals deported from the United States in 2015 were detained for more than one year, with approximately 7 percent detained for two years or longer (Bermudez n.d.). By comparison, 42 percent of individuals released from state prison in 2018 served 1 year or less, with a median sentence length of 1.3 years (Kaeble 2018). The mandatory and often indefinite nature of immigration detention, therefore, likely compounds its cumulative physical and psychological toll (Martinez-Aranda 2020; Saadi, Patler, and Langer 2025).
Although research on the health of detained immigrants remains limited, in part because ICE does not release data on detainee health, a growing body of evidence links conditions of confinement to adverse physical and mental health outcomes during detention and after release (Patler, Sacha, and Branic 2018; Saadi, Patler, and De Trinidad Young 2022; Saadi, Patler, and Langer 2025). Studies of immigrants detained by ICE and later released in the United States documented high rates of poor or fair self-rated health, mental illness, and PTSD, particularly among those detained for longer periods (Saadi, Patler, and Langer 2025) and among those with existing mental health vulnerabilities (Patler, Saadi, and Langer 2025). Additional research shows that health-harming conditions such as sleep deprivation, difficulty accessing family visitation, difficulty accessing medical and mental health care, and harassment and abuse by facility staff were independently and cumulatively associated with poorer overall health among formerly detained immigrants released in the United States (Saadi, Patler, and De Trinidad Young 2022). It is not surprising, then, that detained immigrants consistently describe detention as a stressful and harmful experience that they perceive as unjust by design (Diaz et al. 2023b).
Despite the expanding literature on the health harms of detention, most research focuses on health outcomes during detention or following release on bond within the United States. Far fewer studies examine how detention experiences may shape health after deportation — and those that do generally analyze the experiences of Mexican returnees. For example, research using the EMIF Norte, a survey of migration flows to and from Mexico, has shown that deported Mexican immigrants report poorer health than voluntary returnees, suggesting that deportation itself is associated with health harms (Hamilton, Orraca-Romano, and Vargas Valle 2023). Other studies focused on deported Mexican populations find that experiences of physical or verbal abuse during detention are associated with poorer post-deportation health, including higher likelihoods of poor self-rated health and recent (past 15 days) symptoms of short-term fever, sadness, and feelings of desperation (Waldman 2022; Bakely et al. 2023).
Taken together, existing research documents the punitive conditions of immigration detention and their health consequences during confinement and shortly thereafter, but it leaves unresolved whether and how these harms may persist after deportation, particularly when individuals return to contexts marked by economic precarity, social disruption, and limited access to care (Brouwer et al. 2009; Dingeman 2018; Pinedo et al. 2018). This gap is especially consequential for Central American immigrants, many of whom migrate to the United States to flee violence and persecution rather than the primarily economic push factors that historically characterized Mexican migration (Menjívar 2000; Massey, Durand, and Pren 2014). Addressing this gap requires attention to detention as an experience with consequences that may unfold not only beyond the site of confinement but also beyond US borders.
Central American Immigrants: A Growing but Understudied Population
Central Americans represent a particularly important population for understanding the consequences of immigration detention and deportation. This is due to their rapid population growth in the United States, their increasing presence among undocumented and deported populations, and their unique profile as predominantly asylum-seekers.
As of 2023, approximately 4.3 million Central Americans were living in the United States, of whom an estimated 3.8 million were undocumented immigrants (Dasema and Batalova 2025). Together, they represent more than one in four (26 percent) of all undocumented immigrants in the country (Gelatt, Ruiz-Soto, and Bachmeier 2025). Over the past decade, as increasing numbers of Central Americans have fled their home countries with limited pathways to legal residency in the United States, they have become disproportionately overrepresented in the detention and deportation system. 1 For example, Central Americans accounted for nearly 47 percent of the detained population in 2015 and 25 percent of the deported population in 2017, despite representing only 15 percent–17 percent of the nation’s undocumented population (Franco, Patler, and Reiter 2022; Pedroza 2022). These trends are also reflected in more recent data from ICE arrests in New York City (NYC) during the second Trump administration: Guatemalans and Hondurans accounted for 8 percent and 6 percent of ICE arrests, respectively, even though each group represents less than 1.5 percent of the undocumented population in NYC (New York Immigration Coalition 2025). The persistent overrepresentation of Central Americans in the United States detention and deportation systems warrants careful attention, as it could signify disproportionate exposure to detention-related health harms and adverse health consequences following confinement and deportation.
Central Americans’ profile as mostly asylum-seekers further highlights their potential vulnerability within these expanding systems. Since 2014, many Central Americans apprehended at the US-Mexico border have been fleeing generalized violence in their home countries and seeking asylum (O’Connor, Batalova, and Bolter 2019; Córdova et al. 2025). Once in US custody, the restrictive and punitive conditions of immigration detention may further retraumatize this already highly vulnerable population (Menjívar, Gómez Cervantes, and Staples 2021). A systematic review of research on detained asylum seekers in the UK, Japan, Canada, and Australia found that detention was uniquely associated with a higher prevalence of mental health harm (e.g., clinically meaningful PTSD, depression, and anxiety) and self-harm (Filges et al. 2024).
Understanding the potential consequences of detention is also especially important in the current political moment, as Central Americans also face heightened vulnerability to deportation under the second Trump administration. For example, in September 2025, nearly 72,000 Hondurans lost protection from deportation after their TPS was terminated, and similar policy shifts could place more than 170,000 Salvadorans with TPS at risk (National Immigration Forum 2025). These populations, in becoming suddenly undocumented overnight, also become instantly vulnerable to immigration law enforcement, including detention and deportation (Menjívar 2006). Together, these trends underscore the urgent need to closely examine and better understand the experiences of Central Americans within the United States’ expansive detention and deportation system.
Data and Methodology
We use three waves of representative, repeated cross-sectional survey data from the 2017 to 2019 Encuesta sobre Migración en la Frontera Sur de México (Survey of Migration on the Southern Border of Mexico, or EMIF Sur), the most comprehensive data source on Guatemalan, Honduran, and Salvadoran immigrants deported from the United States to their countries of birth. The survey employs a probabilistic sampling design for mobile populations. It first identifies strategic locations where the study population can be identified and counted in their country of origin, and then randomly selects specific days and time periods (shifts) to form the sampling unit. Trained staff from El Colegio de la Frontera Norte (EL COLEF), in collaboration with academic institutions and other non-governmental organizations in each country, screen and interview eligible deported migrants at airports or reception centers upon arrival. This careful selection of time and place helps ensure that the survey reflects the overall population of deported migrants returning to Central America during the study period (see Secretaría de Gobernación 2023, 14). 2
The EMIF Sur provides a unique data source for examining detention experiences and self-rated health among recently deported migrants. First, the survey collects detailed information on respondents’ demographic characteristics and detention experiences, including verbal or physical abuse during apprehension and detention, perceived treatment by US immigration authorities, and self-rated health — information that ICE and other immigration agencies do not release and, in some cases, do not collect at all. We analyze data from the 2017–2019 survey waves, which include measures of abuse and mistreatment while in the custody of immigration authorities, as well as other variables of interest. 3 Although these data were collected during the first Trump administration, the findings remain highly relevant today and may be even more pertinent under its second administration, as ICE arrests in 2025 reached their highest levels in over a decade (East, Patler, and Cox 2026).
Second, the EMIF Sur is particularly well-suited for analyzing the association between detention and health outcomes because deported migrants are surveyed immediately upon return to their countries of birth. This timing reduces the likelihood of recall bias and limits the extent to which health assessments reflect post-return resettlement or reintegration (Diaz, Koning, and Martinez-Donate 2016). Although the cross-sectional data do not allow us to examine health outcomes over time, surveying respondents shortly after deportation provides valuable observational evidence on the health harms of immigration detention that extend beyond the moment of release across borders and at the point of return. Furthermore, the EMIF Sur is one of the only representative surveys capturing migration flows from the northern region of Central America to the United States, offering rare insight into a population that has received relatively little attention in research on the consequences of detention and deportation (Amuedo-Dorantes, Pozo, and Puttitanun 2015; Escamilla García and Cerón 2025).
The 2017–2019 EMIF Sur surveys included data on 11,921 deported migrants. We excluded two respondents under 18 and two others not born in El Salvador, Guatemala, or Honduras, yielding a sample of 11,917. When weighted, this represents a deported population of 275,495 individuals.
Measures
Outcome Variable
To assess the health of recently deported migrants, we use respondents’ self-rated health (SRH) status as a global measure of physical and mental health. SRH, the only health measure collected during the survey years, is measured on a 5-point Likert scale ranging from “very good” to “very bad” (in Spanish: muy bueno, bueno, regular, malo, and muy malo). A substantial body of research has shown that this measure of self-rated health is a valid and reliable predictor of mortality and overall health, supporting its use in our study (Idler and Benyamini 1997). Following prior research on the health of deported Mexicans surveyed in the EMIF Norte (Waldman 2022; Hamilton, Orraca-Romano, and Vargas Valle 2023), and to maintain comparability with these studies, we dichotomize self-rated health by coding regular/bad/very bad health as 1 (hereafter “poor” health) and very good/good health as 0 (hereafter “good” health). Because research shows that translating “fair” health as “regular” can lead Spanish-speaking respondents to report poorer health than they would in English (Viruell-Fuentes et al. 2011), we also conduct a sensitivity analysis using a continuous measure of health in an ordered logit regression, which we discuss further below in the Sensitivity Analyses and Supplemental Models section.
Explanatory Variables
We draw on three explanatory variables to measure experiences of abuse and mistreatment in immigration detention. The first two variables measure whether respondents experienced physical abuse and verbal abuse during their apprehension and detention by US immigration authorities or other law enforcement officers (hereafter “US immigration authorities”). Physical abuse is measured by asking respondents whether they were subjected to physical aggression (e.g., shoving, hitting, etc.). The variable is coded as 1 = experienced physical abuse and 0 = did not experience physical abuse. Similarly, verbal abuse is measured by asking whether respondents were subjected to mockery, contempt, insults, or shouting, with 1 = experienced verbal abuse and 0 = did not experience verbal abuse.
Our third explanatory variable measures respondents’ subjective evaluation of their treatment by US immigration authorities through the question, “How would you describe the treatment you received from the US immigration authorities who detained you?” Respondents rated their treatment on a 5-point Likert scale from “very good” to “very bad” (in Spanish: muy bueno, bueno, regular, malo, and muy malo). We recoded responses into a binary variable, with 1 indicating poor treatment (regular/bad/very bad) and 0 indicating good treatment (very good/good). 4
Covariates
We account for several background characteristics that may influence respondents’ health status: gender (0 = female, 1 = male); age (in years); educational attainment (0 = less than high school, 1 = high school or more); speaks English (0 = no, 1 = yes); total number of deportations from the United States (0 = 1 deportation, 1 = 2 or more deportations); total number of deportations from Mexico (0 = no deportation from Mexico, 1 = at least one deportation from Mexico); arresting agency (1 = Customs and Border Protection [CBP], 2 = ICE, 3 = local police 5 ); length of detention, measured continuously in months, with values ranging from 0 to 12 6 ; years spent in the United States during the last migration spell (1 = <1 year, 2 = 1–<5 years, 3 = 5–<10 years, and 4 = 10 years); and country of birth (1 = El Salvador, 2 = Guatemala, and 3 = Honduras). Finally, to account for potential time trends and seasonality, we include controls for the survey year and month, as prior research has found patterns of poorer self-rated physical and mental health during winter months (Jia and Lubetkin 2009).
Analytic Strategy
We begin by presenting weighted descriptive statistics for the analytic sample. We then estimate multiple logistic regression models to examine the relationships between the three explanatory variables — physical abuse, verbal abuse, and poor treatment — and self-reported health among deported migrants. We present baseline models examining the bivariate associations between these variables and poor health (controlling only for country of origin, month, and year of survey), as well as fully adjusted models that control for covariates. To aid interpretation, we report regression results as odds ratios (ORs) and calculate predicted probabilities of poor health related to the three detention experiences from the fully adjusted models. All analyses were conducted in STATA 18.5, using EMIF Sur sampling weights to account for survey design and produce estimates representative of the deported population between 2017 and 2019.
Results
As shown in Table 1, most respondents were from Guatemala (49.33 percent), followed by Honduras (33.05 percent) and El Salvador (17.62 percent), which is comparable to ICE removal rates from FY 2017 to FY 2019. 7 Most of the sample is male (83.50 percent) with an average age of 27.89 years (SD = 8.31). About one-quarter have completed high school or higher, and 18.91 percent speak English. Additionally, 16.44 percent report having been deported from the United States two or more times, and 5.72 percent experienced at least one deportation from Mexico to their country of birth. The average detention length was 1.81 months (SD = 2.39). Regarding the arresting agency, 41.48 percent were apprehended by ICE, 32.90 percent by CBP, and 25.62 percent by local police. Nearly 78.83 percent spent less than one year in the United States during their most recent migration spell.
Summary Statistics (Weighted N = 275,495).
Source: Authors’ compilation based on EMIF Sur 2017–2019.
Note: Unweighted N = 11,917.
Turning to experiences during immigration detention, 4.78 percent of respondents reported experiencing physical abuse, 11.18 percent reported experiencing verbal abuse, and 37.07 percent reported receiving poor treatment by immigration authorities. Finally, 14.50 percent of respondents reported poor overall health.
Table 2 presents results from baseline and fully adjusted logistic regression models examining the association between our three explanatory variables and poor health. All models control for country of origin and time-fixed effects (year and month of survey). According to Model 1a, deported migrants who reported physical abuse by immigration authorities had 2.05 times higher odds of reporting poor health (p < .001). This pattern persists in Model 1b, even after adjusting for covariates (OR = 1.93, p < .001). Additionally, each additional year of age, apprehension by ICE relative to CBP, and being from Guatemala or Honduras (relative to El Salvador) are each significantly associated with higher odds of poor health, while being male is associated with lower odds.
Odds Ratios From Logistic Regressions Predicting Poor Health Among Recently Deported Central American immigrants.
Source: Authors’ compilation based on EMIF Sur 2017–2019.
Note: Standard errors are in parentheses. Each model controls for month/year of survey.
p < .001. **p < .01. *p < .05.
Model 2a examines the relationship between verbal abuse and poor health. We find that respondents who reported experiencing verbal abuse had 2.09 times higher odds of reporting poor health than those who did not (p < .001), and this pattern also persists in Model 2b after adjusting for covariates (OR = 1.83, p < .001). The significance and direction of the control variables are comparable to those in Model 1b. Finally, Model 3a replaces the abuse variables with a broader measure of perceived poor treatment by immigration authorities. In this model, respondents who rated their treatment as poor had even higher odds (OR = 2.90) of reporting poor health (p < .001) than those who reported good treatment by US immigration authorities. This pattern persists in Model 3b: deported migrants who rated their treatment as poor had 2.91 times higher odds of reporting poor health (p < .001), net of covariates. Turning to the covariates, age and being from Guatemala or Honduras increase the odds of poor health. Similarly to the other fully adjusted models, being male reduces the odds of reporting poor health, a finding consistent with research showing that women around the world report worse self-rated health than men (Hosseinpoor et al. 2012).
To illustrate our findings, Figure 1 shows the predicted probabilities of poor health across experiences of abuse and mistreatment by immigration authorities during detention, based on the fully adjusted Models presented in Table 2. Among our sample of recently deported Central Americans, the likelihood of poor health was nearly 70 percent higher among those who reported experiencing physical abuse by immigration authorities (predicted probability = .22; 95 percent CI = 0.18–0.26), compared to those who did not report physical abuse (predicted probability = .13; 95 percent CI = 0.12–0.14). Additionally, the predicted probability of poor health was about 62 percent higher among deported migrants who reported experiencing verbal abuse (predicted probability = .21; 95 percent CI = 0.18–0.24), compared with those who did not (predicted probability = .13; 95 percent CI = 0.12–0.14). Furthermore, the models predict that respondents who rated their treatment by immigration authorities as poor had a predicted probability of poor health of .21 (95 percent CI = 0.20–0.23), compared with .09 (95 percent CI = 0.08–0.10) among those reporting good treatment, a 133 percent difference that highlights the health harms associated with punitive immigration detention experiences.
Predicted Likelihood of Poor Health by Experience With Immigration Authorities During Detention.
Sensitivity Analyses and Supplemental Models
We conducted two sensitivity analyses to assess the robustness of our results. First, we included a model that uses a continuous health measure, estimated with an ordered logit regression. The results are substantively consistent with our main findings. As shown in Supplemental Table A1 in the Supplemental Appendix, both the baseline and fully adjusted models indicate that three explanatory variables — physical abuse, verbal abuse, and poor treatment — are associated with higher odds of reporting worse self-rated health. In ordered logit models, this corresponds to an increased probability of moving to the next higher (worse) category on the self-rated health scale. For instance, in Model 1b, reporting experiences of physical abuse is associated with an increased probability of moving to a worse self-rated health category (OR = 1.72, p = .001), net of covariates. Similar associations are found in Model 2b for verbal abuse (OR = 1.83, p = .001) and in Model 3b for poor treatment by immigration authorities (OR = 1.69, p = .001). Second, we assessed respondents’ subjective evaluations of their treatment by immigration authorities as a continuous variable (Supplemental Table A2). Higher values on this scale indicate worse perceived treatment. We find that an increase in the treatment scale is associated with increased odds of reporting poor health in both the baseline model (Model 1a: OR = 1.75, p = .001) and the fully adjusted model (Model 1b: OR = 1.72, p = .001). These findings reinforce the robustness of the main results across alternative coding specifications.
We also conducted three supplementary analyses to further substantiate our results. First, existing studies show that prolonged detention is associated with poorer health (Bakely et al. 2023; Saadi, Patler, and Langer 2025), yet few examine the mechanisms underlying this relationship. One possible mechanism is that longer detention increases the likelihood of exposure to physical abuse, verbal abuse, and other forms of poor treatment (including repeated exposure to such harms). We tested this assumption by using time in detention to predict our key independent variables and found some support for this hypothesis: an increase in months in detention increased the predicted probability of reporting physical abuse, verbal abuse, and poor treatment by immigration authorities, net of covariates (see Supplemental Figure A1 in the Appendix).
Second, given evidence that detention and deportation can abruptly separate families and negatively affect the emotional well-being of deported persons upon their return (Andrews 2024; Escamilla García and Cerón 2025), we conducted an additional supplemental analysis that incorporated family characteristics into the baseline and fully adjusted models. Specifically, we controlled for whether respondents had family members in the United States (0/1) and whether they were parents (0/1). The inclusion of these variables does not alter the results (Supplemental Table A3). Finally, we estimated our base models predicting poor health among key subgroups defined by gender (male, female), country of origin (El Salvador, Guatemala, and Honduras), and education (less than high school degree, high school degree or greater; Supplemental Table A4). Overall, the associations are generally consistent across these subgroups.
Discussion
This article advances research on the health consequences of US immigration detention by making three interrelated contributions. First, we provide rare, representative evidence on the association between detention-related abuse and post-deportation health. Using repeated cross-sectional data from the 2017–2019 waves of the Survey of Migration in the Southern Border of Mexico (EMIF Sur), we move beyond the predominantly qualitative or non-representative samples that, due to data constraints, have characterized much of the existing literature. This allows us to assess whether detention experiences — particularly physical abuse, verbal abuse, and poor treatment by US immigration authorities — are systematically associated with poorer self-rated health after deportation. We find that Central American immigrants who experienced physical abuse and verbal abuse by immigration authorities had higher odds of poor health, as did those who rated their treatment by immigration authorities as poor. These results are robust to the inclusion of a range of demographic covariates, including time spent in the United States, detention length, and month- and year-fixed effects.
Second, our results support a broader theorization of US immigration detention as a form of transnational carceral exposure. Our findings suggest that the health consequences of detention are not confined to the period of custody but can persist as migrants exit detention and then be exported to their countries of origin via deportation. In this sense, immigration detention operates less as a discrete episode of confinement than as a punitive institutional exposure whose effects may extend beyond detention and across national contexts. While prior research has largely examined health during detention or shortly after release within the United States, our findings show that detention experiences — particularly those involving abuse and poor treatment — are associated with worse health even after individuals have been deported. The harms of detention, therefore, appear to travel with migrants across national borders, shaping well-being at the moment of return. By demonstrating that detention experiences are associated with health immediately following deportation, our study extends the emerging literature on immigration enforcement and health and connects it to broader scholarship on the durable consequences of carceral contact. Our findings suggest that the reach of the immigration detention system may extend beyond formal custody and beyond US territorial borders, shaping health outcomes among deported populations.
Third, and critically, we center Central American immigrants from El Salvador, Guatemala, and Honduras — populations central to contemporary US immigration enforcement yet underrepresented in quantitative research on detention and health. Migrants from these three countries accounted for more than one in four (26 percent) undocumented immigrants in the United States as of 2023 (Gelatt, Ruiz-Soto, and Bachmeier 2025) and experience distinct migration trajectories shaped by violence, economic displacement, and family separation. To our knowledge, this study is among the first to examine how US detention experiences are associated with post-deportation health among Central American immigrants, despite their growing prevalence and disproportionate exposure to detention and deportation.
Together, our findings demonstrate that punitive detention experiences — including physical and verbal abuse and perceived poor treatment — are associated with poorer self-rated health following deportation. These results carry immediate implications for policymakers, advocates, and public health practitioners in both the United States and migrants’ countries of origin. In the United States, the mass expansion of punitive immigration law enforcement under the second Trump administration has also intensified efforts to identify strategies to mitigate the harms of immigration detention; our findings provide empirical evidence to inform these debates. Beyond US borders, deported migrants often return to contexts characterized by economic precarity, severed family ties, and limited access to health-supporting institutions (Robertson et al. 2012; Menjívar, Morris, and Rodríguez 2018; Escamilla García and Cerón 2025; Masferrer, Hamilton, and Denier 2025). As deportations continue at historically high levels, governments in countries of origin are increasingly seeking ways to support reintegration (Golash-Boza 2015; Hagan and Wassink 2020; Mora and Ruiz Soto 2025). In our conclusion, we propose policy interventions to reduce the health harms of immigration detention in the United States and urge receiving countries to recognize prior US detention as a critical social determinant of health among deported populations.
Building on these contributions, future research can address remaining limitations to deepen understanding of how the US detention and deportation system shapes health outcomes. First, future research should incorporate pre-migratory factors (e.g., exposure to violence and abuse in countries of origin or during the migration journey) and post-migratory factors (e.g., worksite raids and fears of deportation). For example, many immigrants enter detention already in physically and psychologically vulnerable conditions — often following violent arrests, lengthy criminal sentences, abrupt separations from loved ones, or the perilous journey to the US border (Tovino 2016). These preexisting vulnerabilities could make the punitive nature of carceral settings especially harmful, as such environments can exacerbate prior trauma and expose individuals to new forms of victimization while in custody (Anderson, Pitner, and Wooten 2020). Incorporating these dimensions would allow scholars to assess how preexisting trauma interacts with detention experiences to produce lasting health harms (Saadi et al. 2020). Future work should also incorporate more direct and longitudinal measures of health and well-being across the pre-, during-, and post-detention periods, including indicators of chronic conditions, activity limitations, and mental health. At the same time, research would benefit from more ethical and systematic procedures for measuring exposure to violence. Without strong protections for confidentiality and careful attention to detainees’ fear of retaliation for reporting abuse (American Civil Liberties Union 2024), studies may underestimate the prevalence of victimization — particularly among women and transgender people (see Messing et al. 2015; Vogler and Rosales 2023) — thus obscuring the full magnitude of detention-related health harms. These measures are especially urgent considering recent reports documenting excessive use of force and abuse during arrest, confinement, and transfer by US authorities (Cheng 2018; Bustillo 2025).
Our findings also highlight the need for research examining how the health harms of immigration detention vary across intersecting axes of stratification, including, for example, gender, race and ethnicity, indigeneity, legal vulnerability, caregiving status, disability, and sexual orientation. For example, while many analyses predominantly include Latino men — reflecting the gendered and racialized nature of deportation, which disproportionately targets men from Latin America and the Caribbean (Golash-Boza and Hondagneu-Sotelo 2013) — other groups may experience distinct or compounded harms that merit more detailed inquiry (Menjívar, Gómez Cervantes, and Staples 2021; Franco, Patler, and Reiter 2022; Pepe, Saadi, and Molina 2023; Vogler and Rosales 2023). Taken together, advancing this research agenda is important for understanding immigration detention as a durable and stratifying determinant of health, with consequences that may extend well beyond the period of confinement and across national borders.
Conclusion and Policy Recommendations
The rapid expansion of mass detention during the second term of the Trump administration has reignited concerns about the human consequences of the US immigration enforcement system. As immigration detention has expanded, enforcement tactics have become increasingly aggressive. Community members, journalists, and advocacy organizations have documented hundreds of large-scale immigration raids across the country, many characterized by excessive force (Human Rights Watch 2025). At the same time, alarm has grown over unsafe, punitive, and often inhumane conditions of confinement in immigration detention, which may produce additional, lasting harms to immigrants’ physical and mental health in the absence of meaningful oversight (Gibson 2025). Evidence of these failures is stark: more than 78 percent of deaths in ICE custody between 2011 and 2018 involved violations of Performance-Based National Detention Standards (Grassini et al. 2021), and nearly all of the 32 deaths recorded in 2025 — more than in any year since 2004, including during the COVID-19 pandemic when the virus spread quickly through detention facilities (Tosh, Berg, and León 2021 — have been attributed to “dire and inhumane” detention conditions (Gibson 2025; Singh, Murphy Marcos, and Simmons 2026).
Our study’s focus on abuse by US immigration authorities provides timely new evidence on the potential health consequences of detention, underscoring the urgent need for policies that protect the well-being and rights of individuals who come into contact with the detention and deportation system. In the following section, we outline harm-mitigation strategies and policy recommendations to reduce health risks during detention and after deportation in receiving countries. Given the documented health harms of immigration detention around the world, these policy recommendations may also be relevant to other countries where similar detention conditions are observed (World Health Organization [WHO] 2026).
US Immigration Detention
Policy strategies to mitigate the harms of immigration detention include structural reforms, changes to legal and administrative systems, and interventions to reduce individuals’ exposure to harmful conditions. The most effective policy solution — widely advanced by legal and medical experts (e.g., National Immigrant Justice Center 2021) — is to end the use of immigration detention altogether, thereby eliminating a major source of physical, psychological, and social harm.
Reforms to immigration law could also substantially reduce exposure to detention. At the federal level, Congress can revise immigration law and visa policies to better reflect economic and human realities by expanding lawful migration pathways and narrowing the statutory grounds for deportation. Complementary policy approaches can focus on reducing or eliminating individuals’ direct exposure to detention, for example, by facilitating release from custody through guaranteed access to bond hearings for all detained immigrants. Indeed, evidence from a study of immigrants released on bond indicates that release is associated with reductions in psychological and physical stress, although stress levels remain elevated relative to the general population (Patler et al. 2021).
Policymakers should also formally evaluate and, where appropriate, recommend alternatives to detention (ATDs), while carefully assessing whether and how such programs may generate harm (MRS/USCCB and CMS 2015). For example, electronic monitoring devices have been shown to produce stress, anxiety, and physical harm (Martinez-Aranda 2022). In contexts where federal reforms remain stalled, state and local governments can adopt policies that reduce immigrants’ vulnerability to detention and deportation, including sanctuary policies and misdemeanor sentencing reforms designed to narrow the pipeline to deportation (Immigrant Legal Resource Center 2020).
Administrative reforms, though potentially less effective, could prioritize stronger oversight and monitoring of facility staff behavior to prevent avoidable harms and deaths in immigration detention. In addition, reforms must establish a standardized system for tracking morbidity and mortality in detention facilities. Since 2000, the Bureau of Justice Statistics has monitored health and deaths in jails and prisons. However, no comparable federal program exists for immigration detention, despite the carceral conditions that closely resemble penal incarceration. The absence of mortality tracking is particularly concerning given recent Department of Homeland Security reports on detainee deaths that have conflicted with medical examiner findings, including cases likely classified as homicides (MacMillan 2026). Furthermore, strengthening oversight of medical care within detention facilities could mitigate some harms generated within the system. However, given the punitive logic and structural conditions that define immigration detention, such measures are unlikely to meaningfully address the breadth or durability of harms documented in the literature (Diaz et al. 2023b). In the near term, stakeholders should also invest in public education on the harms of detention and expand funding for community-based initiatives that seek to mitigate those harms.
Post-Deportation in Countries of Return
Countries of origin also play an important role in addressing the health consequences of immigration detention among deported populations. Governmental agencies responsible for receiving deported individuals can integrate systematic documentation of detention-related harms into standard reception procedures, ideally in collaboration with health care providers and non-governmental organizations that serve deported communities. This documentation should capture reports of medical neglect, poor conditions of confinement, abuse by authorities, and the use of solitary confinement and other harmful practices. Collecting such information at the point of arrival is critical, as it becomes much more difficult to document these harms once individuals leave reception facilities and airports (Isacson and Flórez 2025). Given ICE’s limited data transparency, documenting detention-related harms during the reception process also plays an important role in understanding the full scope of health and human rights violations linked to US immigration detention.
Along with strengthening accountability efforts, systematically documenting detention-related harms can help identify and address health risks as soon as individuals arrive in the country of return. Reception facilities should be staffed with medical and mental health professionals who can conduct initial physical and mental health assessments, identify individuals at heightened risk, and provide immediate care or connect them with appropriate services. Referrals at the point of arrival are especially important because many deported individuals remain unaware of available assistance programs even years after their return, thus navigating the resettlement process with limited support (Flores Morán 2022; Mora and Ruiz Soto 2025).
Public health institutions in receiving countries should also recognize the experiences of US immigration detention as significant health risk factors for deported populations. Consistent with prior calls (see Saadi et al. 2020), healthcare providers should routinely screen for histories of abuse or mistreatment by immigration authorities, as these experiences can compound existing physical and mental health vulnerabilities. By formally recognizing immigration detention as a social determinant of health, public health systems can improve care for deported individuals and begin to address the longer-term health consequences of detention in countries of return. At every stage, screenings and assessments should be conducted using trauma-informed approaches that uphold dignity, minimize the risk of retraumatization, and avoid further marginalizing deported persons.
Supplemental Material
sj-docx-1-mhs-10.1177_23315024261435092 – Supplemental material for Detention-Related Health Harms Among Recently Deported Central American Immigrants
Supplemental material, sj-docx-1-mhs-10.1177_23315024261435092 for Detention-Related Health Harms Among Recently Deported Central American Immigrants by Adriana Cerón and Caitlin Patler in Journal on Migration and Human Security
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported (in part) by the National Science Foundation Graduate Research Fellowship Program (DGE-2139841), the American Sociological Association Doctoral Dissertation Research Improvement Grant, and the Russell Sage Foundation (Grant #2501-51469). Any opinions expressed are those of the authors alone and should not be construed as representing the opinions of the funder(s).
Supplemental Material
Supplemental material for this article is available online.
1
Apprehensions of Central Americans at the Southwestern border increased significantly in the past decade, surpassing the number of Mexican immigrants for the first time in 2014 and again in 2016. This demographic shift was also reflected in deportations: while ICE removals of Mexican nationals declined from 241,493 to 176,968 between FY2013 and FY2014, removals of Central American nationals increased from 106,400 to 122,298 during the same period (Escamilla García and Cerón 2025; 
).
2
Because the EMIF Sur does not collect data on voluntarily returnees or on non-migrants in Central America, we are unable to compare the health of deported migrants with those of other groups.
3
We do not analyze data from 2020 and onwards because the EMIF Sur data collection was paused following the COVID-19 pandemic.
4
We also analyze respondents’ subjective evaluation of their treatment in its continuous form and find consistent results, which we describe further in the Sensitivity Analyses and Supplemental Models section.
5
Due to small sample sizes, we combine local police (n = 2,219), highway patrol (n = 188), and sheriffs (n = 148) into a single category labeled “local police.” We exclude responses for FBI/DEA (n = 15) and other agencies (n = 12) from our measure of arresting agency.
6
Respondents who reported >12 months (n = 157) were recoded as 12 months due to the distribution of the data.
7
Among the three Central American groups removed during this period, 48.7 percent were from Guatemala, 32.6 percent from Honduras, and 18.7 percent from El Salvador (ICE 2017; ).
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
