Annotated BibliographyStart Assignment summarize the source using the following as guides as to what to include: What are the main arguments?What is the point of this article?What topics are covered

United States, 2014–2018 Emilie Bruzelius, MPH, and Aaron Baum, PhD Objectives.To examine the relationship between aggressive enforcement of anti- immigration policies and mental health among Hispanics/Latinos in the United States before and after major national immigration policy changes.

Methods.Data were drawn from Behavioral Risk Factor Surveillance System surveys administered from 2014 to 2018. The exposure was the rate of immigration arrests in the 2 months before the survey date within the respondent’s state of residence.

Outcomes included past-month reporting of (1) number of days of poor mental health, (2) at least 1 day of poor mental health, and (3) frequent mental distress.

Results.There was no relationship between arrest rates and mental health among Hispanic/Latino respondents across the overall period. After consideration of policy changes, however, a 1-percentage-point increase in a state’s immigration arrest rate in the postpolicy period was significantly associated with each mental health morbidity outcome.

Conclusions.We found evidence supporting an association between worsening mental health among Hispanics/Latinos and increased arrest rates following the an- nouncement of several restrictive immigration policies. The potential public health effects of aggressive immigration enforcement must be better acknowledged and addressed in immigration debates. (Am J Public Health.2019;109:1786–1788. doi:10.2105/ AJPH.2019.305337) I mmigration and Customs Enforcement (ICE) has arrested more than 2 million immigrants living without authorization in the United States since 2008. 1Evidence suggests that immigration-related anxiety could have a detrimental impact on mental health, particularly among racial/ethnic groups that have been disproportionately targeted. 2In January and February 2017, a series of anti-immigration executive orders were announced that authorized a border wall with Mexico, banned US entry for people from several predominantly Muslim coun- tries, and modified ICE policies, sparking a period of increased immigration arrests, heightened anti-immigration sociopolitical rhetoric, and greater public awareness re- garding deportation. 3 We theorized that, in this ensuing climate, immigration arrests could be associated with poor mental health by increasingdeportation fears among undocumented individuals and their social or familial networks directly 4and by increasing dis- criminatory or exclusionary experiences among Hispanics/Latinos more generally. 5 We therefore sought to determine whether aggressive anti-immigration enforcement tar- geting undocumented immigrants worsens the mental health of Hispanic/Latino Americans in the general population and whether the highly publicized anti-immigration policies announced in early 2017 affected that relationship. METHODS Outcome and covariate data were drawn from the Behavioral Risk Factor Surveillance System (BRFSS), an annual telephone survey of noninstitutionalized adults. In our primary analyses, we focused on participants who completed surveys between October 2014 and March 2018 (n = 1 055 088) and, in re- sponse to the question“Are you Hispanic, Latino/a, or Spanish origin?”self-identified as being of Hispanic/Latino ethnicity, regardless of race (n = 118 883 [17% after the survey design had been taken into account]). Mental health outcomes were based on the question “Now, thinking about your mental health, which includes stress, depression, and prob- lems with emotions, for how many days during the past 30 days was your mental health not good?”We operationalized mental health status as follows: (1) number of poor mental health days, (2) any indication of poor mental health (1 or more days), and (3) fre- quent mental distress (on 14 or more days). 6 Using interview dates, we matched survey responses to monthly state-level ICE arrest rates. Arrest counts were obtained from Transactional Records Access Clearing- house 1data made public following Freedom of Information Act requests. This clearing- house compiles data on“interior arrests,” which exclude arrests at US borders. We calculated arrest rates by dividing states’ monthly arrest counts by corresponding census denominators.

We used quasi-Poisson regression to test the association between arrest rates in a ABOUT THE AUTHORSEmilie Bruzelius is with the Department of Epidemiology, Mailman School of Public Health, Columbia University, New York, NY. Aaron Baum is with the Department of Health System Design and Global Health, Icahn School of Medicine, Mount Sinai, New York, NY.

Correspondence should be sent to Aaron Baum, PhD, Icahn School of Medicine at Mount Sinai, 1216 5th Ave, New York, NY 10029 (e-mail: [email protected]). Reprints can be ordered at http://www.ajph.org by clicking the“Reprints”link.

This article was accepted August 7, 2019.

doi: 10.2105/AJPH.2019.305337 1786ResearchPeer ReviewedBruzelius and BaumAJPHDecember 2019, Vol 109, No. 12 AJPHOPEN-THEMED RESEARCH respondent’s state of residence during the 2 months preceding the interview date and self-reported mental health outcomes. We then refit the models, incorporating an in- teraction between arrest rates and a post-2017 policy indicator. We adjusted for established mental health risk factors including season- ality, age, gender, marital status, education, income, and employment. Models further included year-fixed effects (in addition to the binary post-2017 indicator) as well as state- fixed effects. All analyses accounted for the complex survey design of the BRFSS. Re- spondents with missing outcome (n = 2344) or covariate (n = 41 732) data were excluded.

We conducted several robustness checks to examine the sensitivity of the results (see appendix tables andfigures, available as a supplement to the online version of this article at http://www.ajph.org). First, we included all BRFSS respondents and specified a 3-way interaction between Hispanic/Latino eth- nicity, arrest rate, and a postpolicy indicator.

Second, we repeated the analyses after restricting the sample to respondents who completed the interview in a language other than English. Third, we explored heteroge- neity by state sanctuary laws.

RESULTS More than a third of Hispanic/Latino respondents (34.08%) reported at least 1 poor mental health day in the preceding month, and 11.38% reported frequent mental distress.

The mean number of past-month poor mental health days was 3.62 (SD = 0.05; Table A, available as a supplement to the online version of this article at http:// www.ajph.org). During the observation pe- riod, ICE made 440 601 arrests. Arrests in- creased sharply after the 2017 policy announcements (Figure A, available as a supplement to the online version of this article at http://www.ajph.org), with changes in arrest rates varying across states (Figure B, available as a supplement to the online version of this article at http://www.ajph.org).

Mental health outcomes among Hispanic/ Latino respondents in states with higher arrest rates were similar to those among Hispanic/ Latino respondents in states with lower rates (Table 1, models 3, 7, and 11, and Table B, available as a supplement to the online versionof this article at http://www.ajph.org). How- ever, the association between arrest rates and each mental health outcome was signif- icant in a comparison of the postpolicy and prepolicy periods (Table 1, models 4, 8, and 12). Relative to the prepolicy period, a 1-percentage-point increase in the arrest rate in the postpolicy period was associated with worse mental health (poor mental health days, relative risk [RR] = 1.13; 95% confidence interval [CI] = 1.03, 1.22; any indication of poor mental health, RR = 1.11; 95% CI = 1.04, 1.17; frequent mental distress, RR = 1.14; 95% CI = 1.02, 1.26).

In sensitivity analyses in which a 3-way interaction model was used and the sample was restricted to respondents who completed the interview in a language other than English or lived in states with antisanctuary laws, results were consistent with or larger inmagnitude and direction than the primary results were but not consistently significant (Tables C–E, available as a supplement to the online version of this article at http:// www.ajph.org). Of note, the statistical power of these tests was lower than that of our primary analysis. DISCUSSION Comparing changes in mental health after versus before anti-immigration policies, we found evidence that Hispanic/Latino Americans’mental health worsened more in states with larger increases in arrest rates. In models that did not account for these policies, arrest rates did not correlate with Hispanic/ Latino respondents’mental health. Although the magnitudes of the associations were TABLE 1—Relative Risks for Associations Between State-Level Immigration Arrest Rates and Poor Mental Health Among Hispanic/Latino Respondents Before and After Enactment of Anti-Immigration Policies: Behavioral Risk Factor Surveillance System, United States, 2014–2018 ModelArrest Rate, RR (95% CI)Post-2017 Policies, RR (95% CI)Arrest Rate·Post-2017 Policies, RR (95% CI) No. of poor mental health days0.87 (0.69, 1.04) 1.13 (1.03, 1.22) Model 1 0.86 (0.73, 1.01) Model 2 1.13 (0.94, 1.38) Model 3 1.01 (0.92, 1.11) Model 4 0.91 (0.79, 1.04) Any indication of poor mental health0.93 (0.82, 1.05) 1.11 (1.04, 1.17) Model 5 0.90 (0.81, 1.01) Model 6 1.12 (0.98, 1.28) Model 7 1.03 (0.96, 1.09) Model 8 0.93 (0.85, 1.02) Frequent mental distress 0.86 (0.64, 1.09) 1.14 (1.02, 1.26) Model 9 0.83 (0.67, 1.03) Model 10 1.13 (0.89, 1.44) Model 11 1.01 (0.88, 1.13) Model 12 0.90 (0.73, 1.06) Note. CI = confidence interval; RR = relative risk. All models present exponentiated results of quasi- Poisson regression models with robust variance estimations, interpretable as relative risks. Models 1, 5, and 9 present results for the pre-2017 period; models 2, 6, and 10 present results for the post-2017 period; and models 3, 4, 7, 8, 11, and 12 present results for the entire observation period (October 2014 to March 2018). All models controlled for season, age group, gender, marital status, education, income, and employment and included separate survey-yearfixed effects and state-fixed effects with errors clustered at the state level. Additional analyses involving negative binomial regression to account for potential overdispersion (for each of the 3 outcomes) and logistic regression (for frequent mental distress) produced similar results. Placebo tests incorporating a 24-month lagged arrest rate variable were nonsignificant (results are available on request). As a result of the way the BRFSS is administered, some individuals in the 2017 wave were actually interviewed in 2018. AJPHOPEN-THEMED RESEARCH December 2019, Vol 109, No. 12AJPHBruzelius and BaumPeer ReviewedResearch1787 relatively small, it is important to note that our results captured aggregate associations. We probably underestimated the impact of arrest rates on mental health among those directly targeted by anti-immigration policies and those living in states where the salience of national anti-immigration rhetoric may have been especially pronounced.

Our overallfindings are consistent with emerging evidence documenting the adverse effects of restrictive immigration policies.

Studies examining Arizona Senate Bill 1070, a measure allowing law enforcement personnel to arrest individuals suspected to be un- documented, showed that self-reported health 7and use of health care services 8de- clined in response to the bill’s enactment, whereas adverse birth outcomes increased. 9 Analyses of single ICE raids have further revealed that the health effects of raids can extend beyond noncitizens. Similar to the results of our study, in which we did not restrict analyses according to citizenship or nativity, previous research has shown that living in a community that has been targeted by ICE can increase the risk of poor physical and mental health among US-born Hispanics/Latinos. 10,11 Several limitations should be considered when interpreting our results. For example, we relied on pooled cross-sectional data and cannot rule out unmeasured confounding, especially given that the confidence intervals were close to null. Within-state changes in immigration arrest rates may be correlated with changes in a state’s proportion of un- documented residents or other policies affecting mental health, 12but we were unable to control for these factors.

Another important issue we could not address is the mechanisms through which immigration arrests may influence mental health. For instance, we do not know the extent to which respondents experienced immigration-related discrimination; also, we do not have information on whether they were aware of immigration arrests and whether this awareness increased stress related to immigration status.

Further research is needed to examine these possibilities directly, preferably with more granular geographical measures of ex- posure (e.g., neighborhood or county) that can support quasi-experimental methods such as interrupted time-series designs. Researchthat either directly measures immigration- related anxiety and discrimination experiences or better captures regional anti-immigration contexts is needed to examine plausible mechanisms and inform potential interventions. PUBLIC HEALTH IMPLICATIONS Additional research and, more important, public health action are urgently needed to mitigate the damaging effects of intensified anti-immigration enforcement measures and restrictive immigration policies on health.

Given that immigration policy continues to be a deeply contested topic, ensuring that the health and social consequences of aggressive enforcement are identified and acknowl- edged within national debates is a key priority. CONTRIBUTORSBoth authors contributed equally to this article.

ACKNOWLEDGMENTSA. Baum is supported by grants from the National Institute on Minority Health and Health Disparities.

We thank Natalie Levy of the Mailman School of Public Health at Columbia University for providing methodological guidance regarding several of our sensitivity analyses.

CONFLICTS OF INTERESTThe authors declare no conflicts of interest.

HUMAN PARTICIPANT PROTECTIONNo protocol approval was needed for this study because de-identified public data sets were used.

REFERENCES1. Transactional Records Access Clearinghouse. Immi- gration and Customs Enforcement arrests. Available at: https://trac.syr.edu/phptools/immigration/arrest.

Accessed September 10, 2019.

2. Williams DR, Medlock MM. Health effects of dramatic societal events—ramifications of the recent presidential election.N Engl J Med. 2017;376(23):2295–2299.

3. Keller J, Pearce A, Andrews W. Tracking Trump’s agenda, step by step. Available at: https://www.nytimes.

com/interactive/2017/us/politics/trump-agenda- tracker.html. Accessed September 10, 2019.

4. Hacker K, Chu J, Arsenault L, Marlin RP. Provider’s perspectives on the impact of Immigration and Customs Enforcement (ICE) activity on immigrant health.J Health Care Poor Underserved. 2012;23(2):651–665.

5. Chavez LR, Campos B, Corona K, Sanchez D, Ruiz CB. Words hurt: political rhetoric, emotions/affect, and psychological well-being among Mexican-origin youth.

Soc Sci Med. 2019;228:240–251.

6. Bossarte RM, He H, Claassen CA, Knox K, Tu X.

Development and validation of a 6-day standard for the identification of frequent mental distress.Soc Psychiatry Psychiatr Epidemiol. 2011;46(5):403–411.7. Anderson KF, Finch JK. Racially charged legislation and Latino health disparities: the case of Arizona’s S.B.

1070.Sociol Spectr. 2014;34(6):526–548.

8. Toomey RB, Umaña-Taylor AJ, Williams DR, Harvey-Mendoza E, Jahromi LB, Updegraff KA. Impact of Arizona’s SB 1070 immigration law on utilization of health care and public assistance among Mexican-origin adolescent mothers and their motherfigures.Am J Public Health. 2014;104(suppl 1):S28–S34.

9. Torche F, Sirois C. Restrictive immigration law and birth outcomes of immigrant women.Am J Epidemiol.

2019;188(1):24–33.

10. Lopez WD, Kruger DJ, Delva J, et al. Health im- plications of an immigration raid:findings from a Latino community in the midwestern United States.J Immigr Minor Health. 2017;19(3):702–708.

11. Novak NL, Geronimus AT, Martinez-Cardoso AM.

Change in birth outcomes among infants born to Latina mothers after a major immigration raid.Int J Epidemiol.

2017;46(3):839–849.

12. Hatzenbuehler ML, Prins SJ, Flake M, et al. Immi- gration policies and mental health morbidity among Latinos: a state-level analysis.Soc Sci Med. 2017;174:

169–178. AJPHOPEN-THEMED RESEARCH 1788ResearchPeer ReviewedBruzelius and BaumAJPHDecember 2019, Vol 109, No. 12 Copyright ofAmerican JournalofPublic Health isthe property ofAmerican PublicHealth Association anditscontent maynotbecopied oremailed tomultiple sitesorposted toa listserv without thecopyright holder'sexpresswrittenpermission. However,usersmayprint, download, oremail articles forindividual use.