Background

In the United States (U.S.) Latinos experience disproportionate health disparities. Of the 34.2 million with diabetes, Latinos remain among the most affected groups [1]. Compared to non-Hispanic whites, they have a greater prevalence of diabetes (17 vs. 8%) and experience more complications and deaths [2, 3]. Latino health disparities also are evident during the coronavirus pandemic: while representing only 18% of the U.S. population, 38% of Latinos have coronavirus disease 2019 (COVID-19) [4], and in California, where 38.9% of the population is Latino, they account for 58.5% of COVID-19 cases and 46.8% of COVID-19-related deaths [5]. We now have evidence that this disparity is related to structural factors, such as social determinants of health (e.g., disproportionate occupational exposure to COVID-19, household density), and to higher rates of conditions associated with COVID-19 complications, such as obesity [6,7,8]. Indeed, Latino women (Latinas) have one of the highest rates of obesity in the U.S. (78.8% are overweight or obese), regardless of age [9,10,11]. However, even when offered access to weight management programs, they are less likely to participate in lifestyle-related interventions and, when they participate, are less likely to complete the programs and be successful. Overcoming barriers to successful engagement and completion of lifestyle interventions is especially critical and consequential for the already disenfranchised and often underinsured Latinos [12].

For undocumented Latinos, these disparities are amplified by poor mental health, increased substance use, more isolation, lack of health insurance, and less use of healthcare and social services even when they have access to care [13,14,15,16]. These behaviors and conditions, several of which appear to be associated with living with the constant fear of deportation, could potentially be aggravated by news of immigration-related raids and arrests [17].

Worldwide, policies and increased rhetoric promoting intolerance toward immigrants are known to negatively impact their health [18, 19]. In the U.S. few national events have produced such steep spikes in the incidence of mental and physical conditions of Latino immigrants as immigration raids [20]. In recent years the U.S. administration’s aggressive Immigration and Customs Enforcement (ICE) activities, deportations separating parents and children [21], and even detention of Latino-looking U.S. citizens have been documented. Widespread broadcasting of these ICE activities has fostered increasing fear and distrust of law-enforcement entities and distress among many [22,23,24,25].

As of 2020, the U.S. has the world’s largest infrastructure for detaining undocumented immigrants, with immigration detention centers (IDCs) among the nation’s fastest growing prison industries [26]. Detention and correctional facilities tend to be placed in rural locations and are often located in neighborhoods where racial and ethnic minorities reside [27,28,29,30,31]. For example, detention facilities in Elizabeth, NJ, Houston, TX, Del Rio, TX, and Albuquerque, NM are in zip codes where Latinos comprise more than 50% of the population. Thus, it is not unusual to have Latino immigrants—both documented and undocumented—living near IDCs and constantly exposed to visual reminders of ICE activities (ICE buildings, buses and law enforcement agents driving by with arrested immigrants in view), as is the case in the rural town where this study was conducted. The facility in this study is located in a zip code reporting 100% of residents being Latinos. Residents are exposed regularly to white ICE buses displaying the words “Homeland Security” with the associated blue logo, which are clearly distinguishable from other buses [32, 33]. Aside from the stress of living in the U.S. unauthorized, it is conceivable that the sight of these buses, and the close proximity to the IDCs, could trigger even more distress [34].

While some municipalities welcome correctional facilities as sources of revenue [35, 36], individuals residing near those facilities report fear of losing property value [37], and most consider their proximity undesirable, sometimes even feeling stigmatized for living nearby [38]. Not surprisingly, employees and healthcare providers living and working near IDCs have expressed frustration and a heavy emotional burden as they have witnessed the impact on the Latino communities they serve [39,40,41]. Yet little is known about the impact on the health of Latinos living near IDCs. With plans to expand some of the IDCs in the country, including the one in Adelanto [42], CA, where this study took place, our findings and recommendations may be particularly timely.

Theoretical/Conceptual Framework

Immigration arrests or “raids” occur unexpectedly where individuals live or work, and are usually conducted by heavily militarized agents, often leading to post-traumatic stress symptoms among those experiencing or witnessing these events [43, 44]. Moreover, individuals can be susceptible to traumatic stress when continuously exposed to reminders of potential threats, even without ever experiencing them. Symptoms similar to those manifested in post-traumatic stress disorder are likely and are reported to be worse among women than men [45]. The resulting condition, coined “continuous traumatic stress disorder” (CTSD) by Frank Chikane in 1986 [46], is frequently experienced by systematically oppressed and marginalized communities that are unable to benefit from an “external protection” system [47,48,49]. Continuous traumatic stress has been defined as a series of “emotional or behavioral responses to actual or current danger that the narrator describes as resulting from living under conditions of ongoing danger” [50]. In this case, the sight of reminders of possible threats could lead to a state of hypervigilance in anticipation of a traumatic event. This chronic exposure to threats and risks, perceived vulnerability, and lack of control can eventually bring about depression, learned helplessness, and their associated effects [51]. Based on the previous information, it is conceivable that living with the constant threat of deportation could lead undocumented Latinos, and even more so Latinas, to experience behavior changes associated with CTSD.

The purpose of this study was to assess differences in completion and outcomes of a health promotion program for Latinas based on their intervention group’s relative geographic proximity to an IDC. By doing so, we hope to fill an important gap in the literature and provide critical information to help healthcare providers and policy-makers better address the needs of Latinas. We hypothesized that news of immigration arrests would affect attendance and outcomes of a health promotion program for Latinas living near IDCs, in particular those undocumented, more than those of participants living farther away, and that the impact may be associated with symptoms of continuous traumatic stress.

Methods

Participants

This mixed-methods study was conducted as part of the evaluation of Full Plate Living, a lifestyle-based weight management and diabetes prevention program for Latinas residing in the Inland Empire (Southern California), an area with a high proportion of Latinos [52, 53]. As part of a larger study, the eight weekly program sessions were delivered in Spanish to 53 women self-identifying as “Latina”. Of these, 21 resided within 15 miles of an IDC (Near-IDC group) and 32 lived much farther away (approximately 60 miles away—at least 1 h away by car during hours of normal traffic) from an IDC (Far-IDC group). While the program was free, participants completed informed consent forms in Spanish or English, and received a $10 gift certificate to a local grocery store as incentive for completing each survey and participating in the focus group discussion. Ethical approval for this study was obtained from the Loma Linda University Institutional Review Board (IRB #5180068).

Setting: The Detention Center

The Adelanto ICE detention center is the largest IDC in the U.S. and is located in a rural, predominantly Latino area, four miles from the Latino church building where the intervention was delivered. The associated ICE processing center is only 2.5 miles away. As of June 2020, this IDC has housed nearly 2,000 immigrants [54], and ICE buses can be seen going to and from this facility carrying detained immigrants.

Data Collection

Quantitative data were collected as part of our ongoing evaluation of the Full Plate Living program, an intervention described in a separate document (Appendix A). Prior to and immediately after program completion, participants completed a standardized set of biometric measures and self-report surveys in Spanish, which included demographic information and assessed relevant behaviors, mental health, and perceived social support. The program was facilitated by trained bilingual community health workers, and data were collected and entered by research assistants.

Qualitative Procedures

As part of the process evaluation we observed an unexpected attendance drop in our Near-IDC group, formerly one of the most engaged groups. We then conducted a follow-up qualitative exploration using a grounded theory approach by Charmaz [55] guided by the continuous traumatic stress theory. The resulting semi-structured guide was used to explore the reasons for the significant drop in attendance during a focus group discussion (FGD) with a purposive sample of Latinas living near an IDC (N = 10), and a key informant interview (KII) with a community health worker (CHW) who facilitated the program. Probes were used to expand the exploration and allow new issues to emerge (see Appendix B for the interview and focus group guides). The FGD and KII explored two main topics: (1) the effects of ICE-related activity on program participants and the local Latino community, and (2) proposed solutions to counteract or mitigate the deleterious impact of ICE activity, if any. Both the FGD and KII were conducted in Spanish on August 7, 2019, six weeks after program completion and one day after the nationally publicized (and locally much discussed) Mississippi ICE raid where 600 immigrants were arrested, resulting in an increase in fear of deportation among Latinos throughout the nation [56]. Of note, around the same time period residents spoke of a local church raid, which further deepened the discussions about everyone’s fear of raids and deportation. The FGD and KII were conducted by a bilingual researcher, audio-recorded with participants’ permission, and transcribed by a bilingual translator. A team of two researchers then independently coded the transcripts using emergent thematic coding [57]. The team met to compare the codes, and a third independent researcher facilitated code discrepancy discussions. Once the team agreed to a final codebook, it was applied to code the text.

Quantitative Measures and Analyses

Biometric Measures

Height and weight were measured using a Seca portable stadiometer (Seca North America, Chino, CA) and InBody 270 scale (InBody USA, Cerritos, CA), with a person standing without shoes or heavy clothing. Results were reported in inches and pounds (lbs), respectively.

Body mass index (BMI) was calculated as weight × 703/(height)2 and reported in kg/m2. Overweight was defined as BMI between 25 and 29.9 kg/m2, and obesity as BMI > 30 kg/m2.

Self-Reported Measures

Demographics: we assessed age (years), relationship status (4 categories), household size (1 = living alone; 5 = at least 5 persons), number of children in the home (0 = none; 5 = 5 or more children), participants’ education level (1 = no formal education; 2 to7 = increasing levels of education), employment status (no/yes), and annual family income (5 categories).

Immigration status: due to the highly sensitive nature of the issue, and out of concern for participants’ safety, individuals’ immigration status was not collected as part of the quantitative data. After much discussion, and with the advice of our CHWs, the decision was made to ask participants to state only their nativity (1 = United States, 2 = Mexico, 3 = Central/South America, 4 = Caribbean) and the length of time living in the U.S. We then analyzed the data comparing individuals born in the U.S. and those not born in the U.S. The variable was coded immigration/nativity to indicate that the latter group immigrated (regardless of legality).

Food insecurity was assessed based on the response (0 = never true, 1 = sometimes true, 2 = often true) to the following two statements: “Within the past 12 months we worried whether our food would run out before we got money to get more” and “Within the past 12 months we didn’t have money to get more food”.

Dietary patterns and physical activity (short international physical activity questionnaire [58]) were assessed using a flipchart with pictures of locally available foods (fruits, vegetables, legumes, grains, etc.) to assess food frequency (1 = once a day; 4 = 4 or more times) and pictures of exercise types; lower numbers meant less consumption/activity.

Social support for healthy behaviors, such as engagement in physical activity and healthy food consumption (fruits, vegetables, low-fat, low-sugar, low-salt food), was assessed using a 6-point Likert scale.

Depression and anxiety were assessed with the PHQ-9 scale [59] and the GAD-7 scale [60], respectively.

We used SPSS, v.25 for quantitative analyses, including attrition analyses and group comparisons (BMI, demographics, behavior, mental health, and familial support). Food insecurity and employment status comparisons were made using chi-square tests, significance was assessed with independent t tests, and intervention effects were analyzed using paired t tests.

Transcripts were analyzed for emergent themes supported by critical quotes using the 2017 computer software program MAXQDA® (v.12, Berlin, GDR) [61], and then quotes were translated from Spanish to English.

Results

Quantitative Analyses Results

Table 1 compares baseline characteristics of participants from the Near-IDC group with those from the Far-IDC group. Average participant age was 47 (± 14) years, and most participants were married and immigrants born in Mexico. Participants in the Near-IDC and Far-IDC groups were comparable in age, marital status, BMI, birth country, years of residence in the U.S., educational level, family income, food insecurity status, number of meals per day, depression, and anxiety. The Near-IDC group differed from the Far-IDC group in that, at baseline, they reported more walking (p = 0.03) and vegetable consumption (p = 0.01) and perceived more social support for healthy eating (p = 0.007) and for physical activity (p = 0.000). Near-IDC group participants were also more likely to be unemployed (p = 0.01).

Table 1 Baseline characteristics of Near-IDC and Far-IDC comparison groups—all enrolled and completers
Full size table

Attrition, defined as missing more than two sessions including the last session, was significantly higher in the Near-IDC group (51% of enrollees) than in the Far-IDC group (25% of enrollees; p = 0.01). Completion was defined as attending at least six sessions including the last session. When comparing “completers” with those who dropped out (“dropout group”) within each location, attrition analyses revealed no baseline differences between the dropouts and completers within the Far-IDC group. However, several differences existed between Near-IDC dropouts and completers: completers were more likely to receive social support for physical activity (p = 0.001), tended to eat more frequently during the day (p = 0.053), and tended to be born in the U.S. In fact, all Near-IDC dropouts were foreign-born (p = 0.055).

Figure 1 depicts the timing of increased immigration activities and attendance changes among Near-IDC participants. Approximately a week into the program, President Trump’s rhetoric against California and his stance toward immigrants grew more intense, with threats of raids highly publicized [62] as he ordered that buses drop off undocumented immigrants in Southern California [63]. On the very day that Latino immigrants were dropped off, attendance decreased by a third, and by the time the program ended, half of the participants had stopped attending.

Fig. 1
figure 1

Timing of ICE activity and changes among Near-IDC participants

Full size image

Table 2 shows within- and between-group differences at program completion. When comparing completers in both groups, Near-IDC completers continued to report more vegetable consumption and perceived more emotional support for healthy behaviors than Far-IDC completers. Yet, when comparing program results between groups, those who remained in the Near-IDC group seemingly benefited less from the intervention than participants who completed the program in the Far-IDC group. In contrast with the Far-IDC group, Near-IDC overweight and obese participants had a non-statistically significant trend toward weight gain, with 60% gaining weight and 40% losing weight. Far-IDC overweight and obese participants had a mean weight loss of 2 ± 4.2 lbs (p = 0.04), and 76% of them lost weight. At program completion the Far-IDC group reported less consumption of unhealthy food and sugar-sweetened beverages (p = 0.009 and p = 0.002, respectively), more fruit consumption (p = 0.007), more walking (p = 0.02), and more vigorous physical activity (p = 0.02) compared to baseline. Meanwhile, Near-IDC group completers reported improvement in only two behaviors: vegetable consumption (p = 0.047) and moderate activity (p = 0.049). Vegetable consumption group differences remained statistically significant (p = 0.02). The Far-IDC group reported more social support by program completion, with support for physical activity (p = 0.056) approximating statistical significance. Although the Near-IDC group continued to report more social support for healthy behaviors than the Far-IDC group at the end of the program, they experienced a non-statistically significant downward trend in social support. Last, while the Far-IDC group had a statistically significant decrease in both depression and anxiety levels (p = 0.004), neither the reduction in anxiety nor in depression was statistically significant in the Near-IDC group.

Table 2 Intervention effects (pre-post changes) in Near-IDC (NI) and Far-IDC (FI) groups
Full size table

Qualitative Data Results

Table 3 describes the five qualitative themes that emerged when we contextually explored the impact of increased ICE activity and the disproportional dropout in the Near-IDC group: (1) high program attrition; (2) sadness, fear, anxiety, and stress; (3) reduced mobility; (4) reduced local retail and commercial activity; (5) request for word about their plight to be spread and for tools/resources for self-protection. Briefly, participants shared that the heightened ICE activity in their community resulted in sadness and high anxiety/depression levels, fear of being separated from their children and other family members, and a feeling of disappointment. This in turn resulted in social withdrawal (regardless of immigration status), including limiting all non-essential activities to avoid putting their loved ones at risk of deportation. Some expressed a renewed determination to stand up against the intimidation, while others expressed a sense of helplessness. Still, they requested and hoped to receive tangible help to address the unanticipated, chilling effects of the ICE activities, as they—including U.S. citizens—did not know how to deal with ICE confrontations.

Table 3 Themes derived from focus group discussion and key informant interview
Full size table

Discussion

Although our research was not originally designed to investigate the impact of living near an IDC on Latinas, drastic changes were observed in group dynamics and program results when delivering our health promotion program, which coincided with heightened nationwide ICE activity. The authors decided to compare attendance and program results with those of another group that attended the same program more than 1 h away by car from an IDC facility. In addition, we qualitatively explored whether and how differences related to people’s relative proximity to an IDC and legal status affected outcomes. The comparison group was similar in education level, income, food insecurity, and household size, all factors that have been shown to affect program results [64]. Also, the groups did not differ in nativity or length of residence in the U.S., factors that presumably could have played a role in the response to immigration activities, nor in baseline depression and anxiety levels. However, the group living farther from the IDC was more likely to be employed and less likely to engage in healthy behaviors (eating greens, walking) and to perceive social support for healthy eating and physical activity.

Our quantitative analyses indicate that the Far-IDC group benefited more from the health promotion program, but these results need to be interpreted with caution since the near-IDC group already walked more, consumed more vegetables, and perceived more social support for healthy eating and physical activity at baseline. The lack of statistically significant improvement in the latter group may very well be due to an initial higher baseline (plateau effect). Still, this would not explain the opposite trends in perceived support for healthy behaviors between these groups. Indeed, despite the Near-IDC group initially reporting more family support and engagement in healthy behaviors than the Far-IDC group, the social support they initially had for healthy behaviors dwindled by program completion. It is possible that collective stress may have caused a priority shift, which would be in keeping with Maslow’s hierarchy of needs model [65].

Dropouts from the Near-IDC group were all foreign-born, perceived less family encouragement for healthy behaviors, and reported eating fewer meals per day at baseline. Although not direct indicators of immigration and legal status, the last two characteristics have been associated with acculturation: studies have shown that foreign-born Latinas tend to be healthier, while those living in the U.S. longer tend to be more obese, engage in less physical activity, and have more meals/snacks per day [66,67,68]. Based on eating pattern, we suspect then that the dropout group included more undocumented and/or fewer acculturated Latinas, which may explain why the only participant who reported living less than 6 years in the U.S. was among the dropouts. This may be an indication of the degree of fear or distress that undocumented participants experienced. Unfortunately, CHWs were unable to reach the dropouts who, though initially highly engaged, did not answer phone calls or even respond to personal visits by our CHWs. However, baseline surveys indicated that most were foreign-born, and our qualitative data suggest that this occurred due to extreme fear of being identified and deported.

Moreover, according to our qualitative findings, news of ICE arrests of fellow Latinos seems to have destabilized the morale of both documented and undocumented Latinas living near the immigration detention center, and had negative physical, emotional, and financial impacts [22]. These fears took a distinct toll on participants who initially had high attendance and overall readiness to engage with the program. Even those who remained in the program reported sadness, anxiety, and stress related to the immigration raids and felt that there was a demoralizing reduction in local social activities. The authors learned that ICE raids caused the entire local Latino community to “hunker down” following the news. Interestingly, participants described a phenomenon that the entire country would experience months later during the COVID-19 pandemic: eerily quiet public spaces, drastic reduction in economic activity, and “sheltering-in-place” of entire extended families. Furthermore, this change in circumstances also heightened participants’ desire for resources to protect/defend themselves from potential threats to their families.

While anxiety and depression were measured and results suggest a pattern in line with the interpretation presented in this paper, the authors were unable to assess stress via a validated scale and to gauge the extent to which Near-IDC participants experienced continuous traumatic stress. However, feelings and behaviors associated with CTSD, such as hypervigilance, stress, worry, perceived vulnerability, and lack of control, were evident in the qualitative data. “Immigration stress” is one of the policy-related social determinants of health—conditions “in which people are born, grow, live, work and age”—that have been associated with negative health consequences [69, 70]. In this study it seems plausible that all participants would have experienced some degree of traumatic stress regardless of legal status: some undocumented participants dropping out altogether, others choosing to complete the program but remaining hypervigilant and experiencing increased stress, and still others with authorized “legal” status being negatively emotionally affected and benefiting less from the program.

One of our study limitations is the lack of information about, and thus inability to control for, our participants’ legal status. However, no statistically significant baseline differences in the variables we used as proxies were seen between the Near-IDC and Far-IDC groups. We also did not control for social support or employment differences between groups, and no focus group discussion was conducted in the Far-IDC group. However, FGD comments were specific to the area where Near-IDC participants live (the sight of the bus driving by, the drastic reduction in number of cars parked in the shopping areas). Furthermore, the Far-IDC group did not have the same degree of attrition. Having both self-reported and biometric measures as well as an affected and a comparison group were strengths of our study, as was the mixed methods triangulation used to further explore this phenomenon.

Healthcare professionals and policy-makers should be aware of Latina patients’ complex lives being significantly affected by the proximity to a detention center regardless of their legal status. Moreover, providing education and resources to Latinas living in these communities and equipping them to handle the emotional trauma and fear of possible legal confrontations would be a welcome service. Future research should expand on the potential impact of stress related to ICE activities and the continuous traumatic stress associated with it to determine the health ramifications on this high-need population. While expansion of IDCs may be economically beneficial to some, the health and prevention needs of the local population should remain an important consideration. The results of such studies may provide a more accurate estimate of the true cost of hosting a detention center in the midst of a Latino community.

New Contribution to the Literature

This publication adds to the literature by shedding light on the unanticipated negative impact of immigration policies and location of immigration detention centers on Latinas, a group already struggling with high rates of health disparities. Latinas living in these settings, regardless of their legal status, are susceptible to added stress and anxiety and seem to benefit less from much needed health interventions. The realities that prevail in communities where IDCs are often placed should be understood and further assessed. Not only should addressing the mental health stressors be part of a regular healthcare protocol, but also prevention services should be promoted more aggressively. Steps should also be taken to provide these vulnerable populations with resources and legal support. Lastly, our results seem to indicate that health and politics are not separate, and that the reality of Latinos’ health disparities cannot be addressed by offering health or prevention services that ignore immigration policies.