Risk of Psychiatric Disorders Among Refugee Children and Adolescents Living in Disadvantaged Neighborhoods

Else Foverskov; Justin S White; Trine Frøslev; Henrik T Sørensen; Rita Hamad

doi:10.1001/jamapediatrics.2022.3235

. 2022 Sep 12;176(11):1107–1114. doi: 10.1001/jamapediatrics.2022.3235

Risk of Psychiatric Disorders Among Refugee Children and Adolescents Living in Disadvantaged Neighborhoods

Else Foverskov ^1,^2,^3,^✉, Justin S White ^2,⁴, Trine Frøslev ¹, Henrik T Sørensen ^1,⁵, Rita Hamad ^2,⁶

¹Department of Clinical Epidemiology, Aarhus University and Aarhus University Hospital, Aarhus, Denmark

²Philip R. Lee Institute for Health Policy Studies, University of California, San Francisco

³Department of Public Health, University of Copenhagen, Copenhagen, Denmark

⁴Department of Epidemiology & Biostatistics, University of California, San Francisco

⁵Clinical Excellence Research Center, Stanford University, Stanford, California

⁶Department of Family & Community Medicine, University of California, San Francisco

Accepted for Publication: July 10, 2022.

Published Online: September 12, 2022. doi:10.1001/jamapediatrics.2022.3235

^✉

Corresponding Author: Else Foverskov, PhD, Department of Public Health, University of Copenhagen, Oester Farimagsgade 5, PO Box 2099, 1014 Copenhagen, Denmark (elfo@sund.ku.dk).

Author Contributions: Ms Foverskov had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Foverskov, White, Sørensen, Hamad.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Foverskov.

Critical revision of the manuscript for important intellectual content: White, Frøslev, Sørensen, Hamad.

Statistical analysis: Foverskov, White, Sørensen.

Obtained funding: Hamad.

Supervision: Sørensen.

Conflict of Interest Disclosures: Dr Foverskov reported grants from the National Institute on Aging during the conduct of the study. Dr Hamad reported grants from the National Institutes of Health during the conduct of the study and from Robert Wood Johnson Foundation outside the submitted work. No other disclosures were reported.

Funding/Support: This work was supported by grant R01 AG063385 from the National Institute on Aging (to Dr Hamad).

Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

^✉

Corresponding author.

PMCID: PMC9468942 PMID: 36094528

Key Points

Question

Is quasi-randomly assigned neighborhood socioeconomic disadvantage associated with risk of developing psychiatric disorders among refugee children and adolescents?

Findings

In this population-based cohort study of 18 709 refugees aged 0 to 16 years at the time of resettlement in Denmark, higher neighborhood disadvantage was associated with an increase in risk of psychiatric disorders before age 30 years. The association did not differ by sex, age at arrival, or family structure.

Meaning

Targeted placement of refugee families in less disadvantaged areas and improvements of disadvantaged neighborhoods may have positive impacts on the long-term mental health of refugee children and adolescents.

This cohort study evaluates the association between neighborhood socioeconomic disadvantage and risk of psychiatric disorders among refugee children and adolescents in Denmark.

Abstract

Importance

Refugee children and adolescents are at increased risk of mental health difficulties, but little is known about how the characteristics of the neighborhood in which they resettle may affect vulnerability and resilience.

Objective

To test whether neighborhood socioeconomic disadvantage is associated with risk of psychiatric disorders among refugee children and adolescents and examine whether the association differs by sex, age at arrival, and family structure.

Design, Setting, and Participants

This quasi-experimental register-based cohort study included refugees in Denmark aged 0 to 16 years at the time of resettlement from 1986 to 1998. A refugee dispersal policy implemented during those years assigned housing to refugee families in neighborhoods with varying degrees of socioeconomic disadvantage in a quasi-random (ie, arbitrary) manner conditional on refugee characteristics observed by placement officers. Cox proportional hazard models were used to examine the association between neighborhood disadvantage and risk of psychiatric disorders, adjusting for relevant baseline covariates.

Exposures

A neighborhood disadvantage index combining information on levels of income, education, unemployment, and welfare assistance in the refugees’ initial quasi-randomly assigned neighborhood.

Main Outcomes and Measures

First-time inpatient or outpatient diagnosis of a psychiatric disorder before age 30 years.

Results

Median (IQR) baseline age in the sample of 18 709 refugee children and adolescents was 7.9 (4.7-11.7) years; 8781 participants (46.9%) were female and 9928 (53.1%) were male. During a median (IQR) follow-up period of 16.1 (10.2-20.8) years, 1448 refugees (7.7%) were diagnosed with a psychiatric disorder (incidence rate, 51.2 per 10 000 person-years). An increase of 1 SD in neighborhood disadvantage was associated with an 11% increase in the hazard of a psychiatric disorders (hazard ratio [HR], 1.11; 95% CI, 1.03-1.21). This association did not differ between male and female individuals, refugees who arrived at different ages, or those from single- vs dual-parent households. In secondary analyses using prescribed psychiatric medication as the outcome, a similar association with neighborhood disadvantage was found (HR, 1.08; 95% CI, 1.03-1.14).

Conclusions and Relevance

In this cohort study, neighborhood disadvantage was associated with an increase in risk of psychiatric disorders. The results suggest that placement of refugee families in advantaged neighborhoods and efforts to enhance the neighborhood context in disadvantaged areas may improve mental health among refugee children and adolescents.

Introduction

Psychiatric disorders often originate in early life and can have lifelong negative effects on health and social outcomes.^1,2,3,4 Research increasingly suggests that growing up in a socioeconomically disadvantaged neighborhood can influence the likelihood of developing psychiatric problems and represents a structural determinant of health.^5,6,7,8,9,10 Refugee children may be particularly vulnerable to the neighborhood context, given their history of displacement and likely exposure to trauma and adverse childhood experiences throughout the migration process, already predisposing them to mental health problems.^11,12,13 Characteristics of the neighborhoods in which they resettle may contribute to their likelihood of experiencing mental health difficulties as differential access to neighborhood resources, such as social cohesion, school quality, and social and health services, can affect vulnerability and resilience.^14,15 Because of the unique circumstances of refugees, studies on the effects of neighborhoods on health in the general population may not generalize to refugees, and little is known about the extent to which the neighborhood context is associated with psychiatric morbidity among refugee children and adolescents.¹²

Moreover, the existing literature on associations between neighborhood characteristics and health is predominantly based on nonexperimental studies with potential bias due to confounding, that is, selection of unhealthy individuals into disadvantaged neighborhoods.^16,17 Experimental designs with random allocation of neighborhoods can ameliorate such threats to internal validity. However, experimental evidence on the effects of neighborhood disadvantage on youth mental health is rare and mainly based on results from the US Moving to Opportunity (MTO) study,¹⁸ in which low-income families were randomly assigned to receive a housing voucher allowing relocation to lower-poverty neighborhoods. Female children and adolescents in the MTO treatment group had a reduced risk of psychological distress after 4 to 7 years, while male children and adolescents experienced an increase in risk.^18,19,20,21 The final follow-up 10 to 15 years later showed similar results.^22,23 Results from the MTO study and other work have furthermore shown differential neighborhood effects depending on children’s age^8,24,25,26 and family characteristics.^8,24,27 This work highlights how associations between neighborhood characteristics and mental health may vary across subgroups. Evaluating potential differential associations has high policy relevance as it can help inform which subgroups may benefit most from interventions.

In the current study, we leveraged a natural experiment in which refugees arriving in Denmark from 1986 to 1998 were dispersed in neighborhoods across the country in a quasi-random (ie, arbitrary) fashion.²⁸ This dispersal policy led to assignment of refugee families to neighborhoods with different levels of socioeconomic disadvantage, which we used to provide association estimates not confounded by residential selection bias. We focused on refugees who arrived in Denmark as children or adolescents and tested the hypothesis that quasi-randomly assigned neighborhood disadvantage at resettlement was associated with an increase in risk of psychiatric disorders. Additionally, we tested whether associations differed by children’s sex, age at arrival, and family structure. This study contributes evidence on the association of neighborhood disadvantage with mental health among refugees that may be used to inform programs and policies at a time of increased global displacement.

Methods

Study Design and Participants

Data

Data were drawn from several longitudinal registers covering the entire Danish population (eTable 1 in the Supplement).²⁹ Data on neighborhood characteristics and sociodemographic structure were obtained from registers maintained by the Danish Census Bureau.³⁰ Data on psychiatric disorders were obtained from the Psychiatric Central Register,³¹ which has recorded information on admissions to psychiatric inpatient facilities since 1970 and information on contacts with psychiatric hospital outpatient clinics and emergency services since 1995. Disease diagnoses were based on the International Classification of Diseases, Eighth Revision (ICD-8), through 1994 and subsequently on the Tenth Revision (ICD-10). In secondary analyses, data on prescribed psychiatric medication were drawn from the Prescription Drug Register,³² containing information on anatomical therapeutic chemical codes for purchased drugs since 1995.

We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline. The study was approved by the Danish Data Protection Agency. Danish legislation does not require informed consent for registry-based studies.

Natural Experiment: Refugee Dispersal Policy

The Danish refugee dispersal policy was in place from 1986 to 1998 and aimed at distributing a growing number of incoming refugees more evenly across the country. Refugees who acquired a residence permit during this period were allocated first to counties and then to municipalities proportional to the number of inhabitants.²⁸ Although refugees were free to find a residence on their own, official statistics suggest that at least 90% of incoming refugees were provided with a place of residence under the terms of the dispersal policy.³³ In comparison, roughly half of the participants in the US MTO study moved in response to receiving a voucher.³⁴ Placement officers were in charge of finding a permanent residence for refugees in the allocated municipality. The placement officers did not meet refugees in person and only had access to information on age, marital status, family size, and nationality.^35,36 Thus, refugees were assigned to neighborhoods with different levels of disadvantage in a quasi-random fashion, conditional on information available to placement officers. For this reason, neighborhood placement was unlikely to be influenced by mental health status prior to arrival or other unobserved individual-level factors that may confound the association between neighborhood disadvantage and psychiatric conditions. The natural experiment created by this policy has been used in other studies on neighborhood effects, with historical documentation and analyses supporting the assumption of quasi-random neighborhood assignment.^{35,36,37,38,39} The study is akin to a randomized encouragement design, as the policy placed no restriction against families moving, and welfare support was not conditional on staying in the assigned residence.

Study Population

According to aggregated historical data, Denmark granted residence permits to 76 209 refugees during the period of the dispersal policy (eTable 2 in the Supplement). The 8 largest refugee-sending countries were former Yugoslavia, Iraq, Iran, Afghanistan, Sri Lanka, Vietnam, Somalia, and Lebanon. Of all refugees receiving residence permits during the period, 70 891 (93%) were granted to refugees from these countries (eTable 2 in the Supplement). Because no individual-level data on refugee status exist from this period, we defined the population of refugees similar to prior studies,^33,37,38 including individuals arriving from 1 of the 8 countries listed above from 1986 to 1998. We excluded those who had a spouse already living in Denmark at the time of resettlement, as people who were granted family reunification were not covered by the dispersal policy. Given the focus on childhood exposures, we limited the refugee sample to children and adolescents aged 0 to 16 years when their family obtained a Danish residence permit. We excluded children and adolescents unaccompanied by at least 1 parent and those who immigrated more than 1 year after their parent(s), as they may not have been included under the dispersal policy. After additionally excluding refugees with missing information on their initial assigned neighborhood, the final analytic sample consisted of 18 709 individuals (Figure 1).

Measures

Psychiatric Morbidity

The main outcome of psychiatric disorders was measured as first-time psychiatric hospital contact (inpatient, outpatient clinic, or emergency services) in childhood or early adulthood (ages 0 to 30 years) during which any psychiatric diagnosis was assigned (ICD-8 codes 291-315 or ICD-10 codes F10-F99). A secondary outcome was first-time redemption of a prescribed psychiatric medication (anatomical therapeutic chemical codes N05 and N06A).

Neighborhood Disadvantage

The exposure of interest was socioeconomic disadvantage in the neighborhood to which refugees were quasi-randomly assigned during resettlement. We used residential history data from population registers to identify assigned neighborhoods, as no official records of placement decisions exist. In accordance with prior research, we used parish borders, which represent small, historically meaningful geographic units, to define neighborhoods.^40,41,42 Our data consisted of 2097 parishes nested within 271 municipalities.

Principal component analysis was used to create an index of neighborhood disadvantage for each year from 1986 to 1998. The index combined aggregated information from 4 standardized measures: median family income, low educational attainment, unemployment rate, and welfare benefits (eMethods, eTables 3-6, and eFigure in the Supplement). The index was standardized within the analytical sample, so that a 1-unit change represented a 1-SD change in the index (higher values indicating greater disadvantage).

Statistical Analysis

Primary Analysis

First, we assessed baseline characteristics by tertiles of the neighborhood disadvantage index. Next, we estimated the association between neighborhood disadvantage and psychiatric disorders using Cox proportional hazards models, with time since resettlement as the underlying time scale. Study participants were monitored from the date of resettlement in their assigned neighborhood until the first date of diagnosis with a psychiatric disorder, their 30th birthday, end of the study (February 2019), emigration, or death, whichever came first. To model the association between the disadvantage index and psychiatric disorders, we applied restricted cubic splines with 3 knots placed at quantiles 10, 50, and 90, following established methods.⁴³ We estimated multivariable models adjusted for the characteristics available to placement officers when assigning a residence, including baseline age, sex, country of origin, number of children in the family, parental age, and parental marital status. Baseline age and parental age were specified as continuous variables and all other covariates as categorical. For parental age and marital status, we used information on the mother if available (97%) and otherwise on the father. We also adjusted for year of placement to account for secular trends. Finally, we included fixed effects for initial municipality to account for time-invariant municipality characteristics and strengthen unconfounded identification of associations. As a result, we identified estimates of the neighborhood association from differences in outcomes for refugee children and adolescents placed in neighborhoods with different degrees of disadvantage within the same municipality. Municipality fixed effects were incorporated using stratification, thus allowing different municipalities to have different baseline hazard functions.⁴⁴ Models included robust standard errors clustered by neighborhood to account for associated outcomes among those assigned to the same neighborhood, including nested levels, such as families. The proportional hazards assumption was not violated based on tests of the association between time and the Schoenfeld residuals for each variable.

Secondary Analyses

In secondary analyses, we analyzed effect measure modification (ie, heterogeneity in association estimates) by key characteristics. We estimated separate models for each subgroup of interest: child sex, child age (age groups: 0-4, 5-8, 9-12, and 13-16 years), and parental marital status as an indicator of single- vs dual-parent households. We used bootstrapping with 1000 replications to obtain test statistics for the difference between the estimated associations across subgroups. Additionally, we analyzed the first date of redeeming prescribed psychiatric medication as an alternative outcome using the same model specifications as in the main analysis. The prescription drug register was not available until 1995, resulting in left-censoring of this secondary analysis. Additional sensitivity analyses are described in the eMethods in the Supplement. Hypothesis tests were 2-sided with P < .05 as the threshold for statistical significance. Analyses were performed with Stata version 17 (StataCorp).

Results

Baseline characteristics of the 18 709 children and adolescents (median [IQR] age, 7.9 [4.7-11.7] years; 8781 (46.9%) female and 9928 (53.1%) male) did not vary substantially by neighborhood disadvantage level (Table; see eTable 7 in the Supplement for standardized differences). During a median (IQR) follow-up of 16.1 (10.2-20.8) years, 1448 individuals (7.7%) were diagnosed with a psychiatric disorder (incidence rate, 51.2 per 10 000 person years). The median (IQR) age at diagnosis was 21.3 (17.7-25.0) years. Sixty-nine participants were censored due to death. While 10 371 individuals were censored due to emigration (Table), we found no association between neighborhood disadvantage and the likelihood of emigration (eTable 8 in the Supplement).

Table. Characteristics of Refugee Children and Adolescents Resettling in Denmark From 1986-1998, Overall and by Neighborhood Disadvantage Level.

Characteristic	No. (%)	Tertile of neighborhood disadvantage at resettlement, No. (%)
Characteristic	Total population (N = 18 709)	Low (n = 6519)	Moderate (n = 5732)	High (n = 6458)
Baseline characteristics
Sex
Female	8781 (46.9)	3014 (46.2)	2696 (47.0)	3071 (47.6)
Male	9928 (53.1)	3505 (53.8)	3036 (53.0)	3387 (52.4)
Age at resettlement, y
0-4	3726 (19.9)	1379 (21.2)	1118 (19.5)	1229 (19.0)
5-8	5809 (31.1)	1990 (30.5)	1838 (32.1)	1981 (30.7)
9-12	4762 (25.5)	1650 (25.3)	1470 (25.6)	1642 (25.4)
13-16	4412 (23.6)	1500 (23.0)	1306 (22.8)	1606 (24.9)
Median (IQR)	7.9 (4.7-11.7)	7.8 (4.5-11.6)	7.8 (4.7-(11.6)	8.1 (4.7-12.0)
Marital status of primary parent at resettlement
Not married	2168 (11.6)	818 (12.5)	612 (10.7)	738 (11.4)
Married	16 541 (88.4)	5701 (87.5)	5120 (89.3)	5720 (88.6)
No. of children in the family at resettlement
1	2384 (12.7)	808 (12.4)	775 (13.5)	801 (12.4)
2	5940 (31.8)	2051 (31.5)	1832 (32.0)	2057 (31.9)
3	4118 (22.0)	1442 (22.1)	1231 (21.5)	1445 (22.4)
≥4	6267 (33.5)	2218 (34.0)	1894 (33.0)	2155 (33.4)
Age of primary parent at resettlement, y
17-23	756 (4.04)	293 (4.5)	212 (3.7)	251 (3.9)
24-30	5033 (26.9)	1774 (27.2)	1536 (26.8)	1723 (26.7)
31-37	7234 (38.7)	2466 (37.8)	2289 (39.9)	2479 (38.4)
≥38	5686 (30.4)	1986 (30.5)	1695 (29.6)	2005 (31.1)
Median (IQR)	33.4 (28.9-38.2)	33.4 (28.8-38.2)	33.4 (29.1-38.1)	33.5 (28.8-38.3)
Country of origin
Former Yugoslavia	5759 (30.8)	1698 (26.0)	1961 (34.2)	2100 (32.5)
Somalia	2730 (14.6)	856 (13.1)	710 (12.4)	1164 (18.0)
Afghanistan	665 (3.6)	251 (3.9)	196 (3.4)	218 (3.4)
Sri Lanka	1309 (7.0)	511 (7.8)	543 (9.5)	255 (3.9)
Iraq	2553 (13.7)	980 (15.0)	649 (11.3)	924 (14.3)
Iran	1614 (8.6)	707 (10.8)	446 (7.8)	461 (7.1)
Vietnam	977 (5.2)	244 (3.7)	308 (5.4)	425 (6.6)
Lebanon	3102 (16.6)	1272 (19.5)	919 (16.0)	911 (14.1)
Outcome and censoring
Psychiatric disorders	1448 (7.7)	500 (7.7)	440 (7.7)	508 (7.9)
Death	69 (0.4)	25 (0.4)	20 (0.4)	24 (0.4)
Emigration	10371 (55.4)	3644 (55.9)	3123 (54.5)	3604 (55.8)
Reached age 30 y	5750 (30.7)	2007 (30.8)	1822 (31.8)	1921 (29.8)
End of follow-up (Feb 2019)	1071 (5.7)	343 (5.3)	327 (5.7)	401 (6.2)

Open in a new tab

Neighborhood Disadvantage and Psychiatric Disorders

In graphical analyses using restricted cubic splines, increased neighborhood disadvantage was linearly associated with an increased hazard of psychiatric disorders (Figure 2). Only at higher index values was increased neighborhood disadvantage significantly associated with higher hazard of psychiatric disorders. An increase of 1 SD in neighborhood disadvantage was associated with an 11% increase in the hazard of psychiatric disorders (hazard ratio [HR], 1.11; 95% CI, 1.03-1.21) (Figure 3).

Figure 2. — Hazard ratio and 95% CI obtained from a Cox proportional hazard model with neighborhood disadvantage modeled as a restricted cubic spline with 3 knots and adjusted for sex, age, number of children in the family, parental marital status, parental age, year at time of resettlement, and initial municipality. Standard errors are clustered by neighborhood. N = 18 709.

Secondary Analyses

The association between neighborhood disadvantage and psychiatric disorders did not differ by sex or age (Figure 3). The association was stronger among those in single-parent households (HR, 1.40; 95% CI, 1.08-1.81) than those in dual-parent households (HR, 1.09; 95% CI, 1.00-1.19), but the confidence interval for the single-parent household estimate was wide and the difference not significantly different from zero (eTable 9 in the Supplement).

During follow-up, 2808 individuals (15.0%) redeemed prescriptions for psychiatric medication. An increase in hazard of psychiatric medication use was found in association with greater neighborhood disadvantage (Figure 4). Assuming linearity, an increase of 1 SD in neighborhood disadvantage was associated with an 8% increase in the hazard of psychiatric medication (HR, 1.08; 95% CI, 1.03-1.14). Results of sensitivity analyses are described in the eResults in the Supplement.

Figure 4. — Hazard ratio and 95% CI obtained from a Cox proportional hazard model with neighborhood disadvantage modeled as a restricted cubic spline with 3 knots and adjusted for sex, age, number of children in the family, parental marital status, parental age, year at time of resettlement, and initial municipality. Standard errors are clustered by neighborhood. N = 18 709.

Discussion

We leveraged a natural experiment to examine the association between neighborhood disadvantage and first-time diagnosis with a psychiatric disorder during childhood and early adulthood among refugees aged 0 to 16 years at time of resettlement in Denmark. We found that quasi-random assignment to a disadvantaged neighborhood was associated with increased risk of psychiatric disorders. At higher levels of neighborhood disadvantage, increases of 1 SD in the disadvantage index were associated with an 11% increase in risk of psychiatric disorders. This association was consistent when using psychiatric prescriptions as an alternative outcome that may capture psychiatric morbidity among children and adolescents who not had psychiatric hospital-based contacts but were treated in primary care or private practice. Although the effect size was modest at the individual level, the implications are important at a population level, especially when considering the potential cumulative effects of neighborhood disadvantage.^45,46 Results suggest that 292 psychiatric disorders (20% of all cases) could have been avoided if children and adolescents placed in the most disadvantaged neighborhoods (disadvantage index ≥2) had instead been placed in the least disadvantaged neighborhoods (disadvantage index ≤2). Our findings align with previous observational studies^5,6,9 showing positive associations between neighborhood disadvantage and mental health problems in nonrefugee children and adolescents and with a recent study⁴⁷ that found an association between neighborhood disadvantage and worsened mental health among adult refugees. Our study is among the first to examine this association among refugee children and adolescents. It is also unique in using the Danish dispersal policy as a natural experiment to reduce confounding.

We found no difference by sex in the association between neighborhood disadvantage and psychiatric disorders. This contrasts with experimental evidence from the US MTO experiment, which showed that female children and adolescents experienced reduced psychological distress while male children and adolescents experienced increased psychological distress when moving to low-poverty neighborhoods.^18,19,20,21 The adverse effect for male children and adolescents in the MTO treatment group has been hypothesized to stem from the disruption of social networks and loss of adult role models.⁴⁸ In our study sample, all male children and adolescents experienced the disruption of placement in a new neighborhood environment, unlike in the MTO study, where it was only the treatment group who had this experience. Earlier observational studies^8,26 have also suggested that neighborhood effects on mental health among children increase with age, possibly due to longer exposure and increasing involvement with the neighborhood environment. We found no effect modification by age, perhaps because the outcome was too uncommon to capture differences between age groups. Prior research is inconsistent on whether living in vulnerable families, including single-parent households, modifies the association of neighborhood disadvantage with mental health among children and adolescents.^8,24,27 Although confidence intervals were wide, we found larger estimates for refugees in single-parent households. Further research with more power to detect subgroup differences is needed to identify whether those from single-parent families may be more susceptible to the adverse effects associated with highly disadvantaged neighborhoods.

Neighborhood disadvantage is likely associated with multiple potential mediating risk factors underlying psychiatric disorders. The neighborhood context may be associated with socioeconomic opportunities, health behaviors, social support, and collective efficacy, which can affect psychiatric morbidity in susceptible children and adolescents either directly or through processes within the family, peer group, and local insitutions.^10,14,49 Future studies are needed to clarify the underlying processes. Larger studies with more power to assess specific psychiatric disorders are also warranted, as associations may differ for different illnesses with different etiologies.

Limitations

This study has several limitations. First, results may not generalize to other time periods or countries. Second, we could not identify refugee families who found a place to live on their own and thus were not covered by the dispersal policy. Including children from these families in the study sample may have biased the results, but the bias is likely small, as official statistics suggest that at least 90% of refugees obtained a residence under the policy.³³ Third, we relied on residential history data to identify the neighborhood of placement, as no data on the assigned addresses exist. This may have led to misclassification of the exposure, which could bias results. Fourth, we had no data on preexisting psychiatric disorders among refugees prior to arrival in Denmark. While this may have reduced precision, it is not likely to have biased results, as preexisting cases should be equally spread across neighborhoods due to placement officers’ lack of knowledge of family members’ health status.³⁶ A sensitivity analysis supported this conclusion, as excluding those who had psychiatric disorders within the first 2 years after resettlement did not change results. Fifth, research has shown that refugee children and adolescents use psychiatric health care less than their Danish-born peers^50,51; if undertreatment was higher among refugees in high-disadvantage neighborhoods, the association may be an underestimate. Importantly, all refugees are covered by free national health insurance with psychiatric services available nationwide,⁵² so results are unlikely to be biased by issues concerning insurance coverage and health access present in typical studies in the US. Sixth, we cannot claim with certainty that neighborhood socioeconomic disadvantage is affecting the mental health of the refugee children and adolescents as opposed to other associated unmeasured neighborhood factors. Seventh, we did not model the association for neighborhoods to which refugee families moved after their initial assignment. Since these moves occurred after the initial assignment, they represent mediators rather than confounders, and therefore will not bias results.

Conclusions

This quasi-experimental register-based cohort study is among the first to show that resettling in a disadvantaged neighborhood is associated with a higher risk of psychiatric disorders among refugee children and adolescents, highlighting the potential contribution of a key structural factor that can affect health at the population level. At a time when high numbers of refugee children are seeking asylum in high-income countries,⁵³ our findings suggest that targeted placement of families in less disadvantaged areas and investments to improve disadvantaged neighborhoods may have a positive long-term impact on mental health among refugee children and adolescents.

Supplement.

eMethods

eResults

eReferences

eTable 1. Data sources

eTable 2. Residence permits granted to refugees from 1986 to 1998

eTable 3. Description of the aggregated socioeconomic variables included in the neighborhood disadvantage index, 1986-1998

eTable 4. Results from the principal component analyses: Eigenvalues and proportion of variance explained

eTable 5. Results from principal component analyses: Variable loadings for the first component

eTable 6. Descriptive neighborhood-level socioeconomic characteristics by disadvantage level, 1986-1998

eTable 7. Standardized differences of baseline characteristics by neighborhood disadvantage level among refugee children and adolescents

eTable 8. Association between neighborhood disadvantage and emigration from Denmark among refugee children and adolescents

eTable 9. Tests for differences in associations between neighborhood disadvantage and first-time diagnosis with a psychiatric disorder by subgroups among refugee children and adolescents

eFigure. Neighborhood socioeconomic disadvantage index classified into quintiles in the first and last year of the dispersal policy

Click here for additional data file.^{(878.6KB, pdf)}

References

1.Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry. 2005;62(6):593-602. doi: 10.1001/archpsyc.62.6.593 [DOI] [PubMed] [Google Scholar]
2.Patel V, Flisher AJ, Hetrick S, McGorry P. Mental health of young people: a global public-health challenge. Lancet. 2007;369(9569):1302-1313. doi: 10.1016/S0140-6736(07)60368-7 [DOI] [PubMed] [Google Scholar]
3.Costello EJ, Maughan B. Annual research review: optimal outcomes of child and adolescent mental illness. J Child Psychol Psychiatry. 2015;56(3):324-341. doi: 10.1111/jcpp.12371 [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Currie J, Stabile M. Mental Health in Childhood and Human Capital. In: Jonathan G, ed. The Problems of Disadvantaged Youth. University of Chicago Press; 2009:115-148. doi: 10.7208/chicago/9780226309477.003.0005 [DOI] [Google Scholar]
5.Sundquist J, Li X, Ohlsson H, et al. Familial and neighborhood effects on psychiatric disorders in childhood and adolescence. J Psychiatr Res. 2015;66-67:7-15. doi: 10.1016/j.jpsychires.2015.03.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Jablonska B, Kosidou K, Ponce de Leon A, et al. Neighborhood socioeconomic characteristics and utilization of ADHD medication in schoolchildren: a population multilevel study in Stockholm County. J Atten Disord. 2020;24(2):265-276. doi: 10.1177/1087054716643257 [DOI] [PubMed] [Google Scholar]
7.Jonsson KR, Vartanova I, Södergren M. Ethnic variations in mental health among 10-15-year-olds living in England and Wales: the impact of neighbourhood characteristics and parental behaviour. Health Place. 2018;51:189-199. doi: 10.1016/j.healthplace.2018.03.010 [DOI] [PubMed] [Google Scholar]
8.Humphrey JL, Root ED. Spatio-temporal neighborhood impacts on internalizing and externalizing behaviors in U.S. elementary school children: effect modification by child and family socio-demographics. Soc Sci Med. 2017;180:52-61. doi: 10.1016/j.socscimed.2017.03.014 [DOI] [PubMed] [Google Scholar]
9.Schneiders J, Drukker M, van der Ende J, Verhulst FC, van Os J, Nicolson NA. Neighbourhood socioeconomic disadvantage and behavioural problems from late childhood into early adolescence. J Epidemiol Community Health. 2003;57(9):699-703. doi: 10.1136/jech.57.9.699 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Visser K, Bolt G, Finkenauer C, Jonker M, Weinberg D, Stevens GWJM. Neighbourhood deprivation effects on young people’s mental health and well-being: a systematic review of the literature. Soc Sci Med. 2021;270:113542. doi: 10.1016/j.socscimed.2020.113542 [DOI] [PubMed] [Google Scholar]
11.Fazel M, Reed RV, Panter-Brick C, Stein A. Mental health of displaced and refugee children resettled in high-income countries: risk and protective factors. Lancet. 2012;379(9812):266-282. doi: 10.1016/S0140-6736(11)60051-2 [DOI] [PubMed] [Google Scholar]
12.Scharpf F, Kaltenbach E, Nickerson A, Hecker T. A systematic review of socio-ecological factors contributing to risk and protection of the mental health of refugee children and adolescents. Clin Psychol Rev. 2021;83:101930. doi: 10.1016/j.cpr.2020.101930 [DOI] [PubMed] [Google Scholar]
13.Wood S, Ford K, Hardcastle K, Hopkins J, Hughes K, Bellis M. Adverse childhood experiences in child refugee and asylum seeking populations. Cardiff: Public Health Wales NHS Trust. Published 2020. Accessed January 17, 2022. https://www.researchgate.net/publication/343365371_Adverse_Childhood_Experiences_in_child_refugee_and_asylum_seeking_populations
14.Hill TD, Maimon D. Neighborhood Context and Mental Health. In: Aneshensel CS, Phelan JC, Bierman A, eds. Handbook of the Sociology of Mental Health. Springer Netherlands; 2013:479-501. doi: 10.1007/978-94-007-4276-5_23 [DOI] [Google Scholar]
15.Fazel M, Betancourt TS. Preventive mental health interventions for refugee children and adolescents in high-income settings. Lancet Child Adolesc Health. 2018;2(2):121-132. doi: 10.1016/S2352-4642(17)30147-5 [DOI] [PubMed] [Google Scholar]
16.Oakes JM. The (mis)estimation of neighborhood effects: causal inference for a practicable social epidemiology. Soc Sci Med. 2004;58(10):1929-1952. doi: 10.1016/j.socscimed.2003.08.004 [DOI] [PubMed] [Google Scholar]
17.Diez Roux AV. Estimating neighborhood health effects: the challenges of causal inference in a complex world. Soc Sci Med. 2004;58(10):1953-1960. doi: 10.1016/S0277-9536(03)00414-3 [DOI] [PubMed] [Google Scholar]
18.Kling JR, Liebman JB, Katz LF. Experimental analysis of neighborhood effects. Econometrica. 2007;75(1):83-119. doi: 10.1111/j.1468-0262.2007.00733.x [DOI] [Google Scholar]
19.Schmidt NM, Glymour MM, Osypuk TL. Does the temporal pattern of moving to a higher-quality neighborhood across a 5-year period predict psychological distress among adolescents? results from a federal housing experiment. Am J Epidemiol. 2021;190(6):998-1008. doi: 10.1093/aje/kwaa256 [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Osypuk TL, Schmidt NM, Bates LM, Tchetgen-Tchetgen EJ, Earls FJ, Glymour MM. Gender and crime victimization modify neighborhood effects on adolescent mental health. Pediatrics. 2012;130(3):472-481. doi: 10.1542/peds.2011-2535 [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Osypuk TL, Tchetgen EJT, Acevedo-Garcia D, et al. Differential mental health effects of neighborhood relocation among youth in vulnerable families: results from a randomized trial. Arch Gen Psychiatry. 2012;69(12):1284-1294. doi: 10.1001/archgenpsychiatry.2012.449 [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Sanbonmatsu L, Ludwig J, Katz L, et al. Moving to Opportunity for Fair Housing Demonstration Program—Final Impacts Evaluation. US Department of Housing & Urban Development; 2011. [Google Scholar]
23.Kessler RC, Duncan GJ, Gennetian LA, et al. Associations of housing mobility interventions for children in high-poverty neighborhoods with subsequent mental disorders during adolescence. JAMA. 2014;311(9):937-948. Retracted in: JAMA. 2016;316(2): 227-228. doi: 10.1001/jama.2016.6187 [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]
24.Nguyen QC, Rehkopf DH, Schmidt NM, Osypuk TL. Heterogeneous effects of housing vouchers on the mental health of US adolescents. Am J Public Health. 2016;106(4):755-762. doi: 10.2105/AJPH.2015.303006 [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Schmidt NM, Glymour MM, Osypuk TL. adolescence is a sensitive period for housing mobility to influence risky behaviors: an experimental design. J Adolesc Health. 2017;60(4):431-437. doi: 10.1016/j.jadohealth.2016.10.022 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Li M, Johnson SB, Musci RJ, Riley AW. Perceived neighborhood quality, family processes, and trajectories of child and adolescent externalizing behaviors in the United States. Soc Sci Med. 2017;192:152-161. doi: 10.1016/j.socscimed.2017.07.027 [DOI] [PubMed] [Google Scholar]
27.Nguyen QC, Schmidt NM, Glymour MM, Rehkopf DH, Osypuk TL. Were the mental health benefits of a housing mobility intervention larger for adolescents in higher socioeconomic status families? Health Place. 2013;23:79-88. doi: 10.1016/j.healthplace.2013.05.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Damm AP. The Danish dispersal policy on refugee immigrants 1986-1998: a natural experiment? Department of Economics, Aarhus School of Business, University of Aarhus working paper 05-03. Published 2005. Accessed May 21, 2022. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3782634
29.Schmidt M, Pedersen L, Sørensen HT. The Danish Civil Registration System as a tool in epidemiology. Eur J Epidemiol. 2014;29(8):541-549. doi: 10.1007/s10654-014-9930-3 [DOI] [PubMed] [Google Scholar]
30.Thygesen LC, Daasnes C, Thaulow I, Brønnum-Hansen H. Introduction to Danish (nationwide) registers on health and social issues: structure, access, legislation, and archiving. Scand J Public Health. 2011;39(7)(suppl):12-16. doi: 10.1177/1403494811399956 [DOI] [PubMed] [Google Scholar]
31.Mors O, Perto GP, Mortensen PB. The Danish Psychiatric Central Research Register. Scand J Public Health. 2011;39(7)(suppl):54-57. doi: 10.1177/1403494810395825 [DOI] [PubMed] [Google Scholar]
32.Pottegård A, Schmidt SAJ, Wallach-Kildemoes H, Sørensen HT, Hallas J, Schmidt M. Data resource profile: the Danish National Prescription Registry. Int J Epidemiol. 2017;46(3):798-798f. doi: 10.1093/ije/dyw213 [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Damm AP. Ethnic enclaves and immigrant labor market outcomes: quasi experimental evidence. J Labor Econ. 2009;27(2):281-314. doi: 10.1086/599336 [DOI] [Google Scholar]
34.Ludwig J, Duncan GJ, Gennetian LA, et al. Long-term neighborhood effects on low-income families: evidence from moving to opportunity. Am Econ Rev. 2013;103(3):226-231. doi: 10.1257/aer.103.3.226 [DOI] [Google Scholar]
35.Damm AP. Neighborhood quality and labor market outcomes: evidence from quasi-random neighborhood assignment of immigrants. J Urban Econ. 2014;79:139-166. doi: 10.1016/j.jue.2013.08.004 [DOI] [Google Scholar]
36.Hasager L, Jørgensen M. Sick of your poor neighborhood? CEBI working paper 02/21. Published 2021. Accessed May 21, 2022. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3782634
37.Damm AP, Dustmann C. Does growing up in a high crime neighborhood affect youth criminal behavior? Am Econ Rev. 2014;104(6):1806-1832. doi: 10.1257/aer.104.6.1806 [DOI] [Google Scholar]
38.Hamad R, Öztürk B, Foverskov E, et al. Association of neighborhood disadvantage with cardiovascular risk factors and events among refugees in Denmark. JAMA Netw Open. 2020;3(8):e2014196. doi: 10.1001/jamanetworkopen.2020.14196 [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Damm AP. Determinants of recent immigrants’ location choices: quasi-experimental evidence. J Popul Econ. 2009;22(1):145-174. doi: 10.1007/s00148-007-0148-5 [DOI] [Google Scholar]
40.Meijer M, Engholm G, Grittner U, Bloomfield K. A socioeconomic deprivation index for small areas in Denmark. Scand J Public Health. 2013;41(6):560-569. doi: 10.1177/1403494813483937 [DOI] [PubMed] [Google Scholar]
41.Schofield P, Thygesen M, Das-Munshi J, et al. Ethnic density, urbanicity and psychosis risk for migrant groups—a population cohort study. Schizophr Res. 2017;190:82-87. doi: 10.1016/j.schres.2017.03.032 [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Meijer M, Kejs AM, Stock C, Bloomfield K, Ejstrud B, Schlattmann P. Population density, socioeconomic environment and all-cause mortality: a multilevel survival analysis of 2.7 million individuals in Denmark. Health Place. 2012;18(2):391-399. doi: 10.1016/j.healthplace.2011.12.001 [DOI] [PubMed] [Google Scholar]
43.Harrell FE. General aspects of fitting regression models. In: Regression Modeling Strategies: With Applications to Linear Models, Logistic and Ordinal Regression, and Survival Analysis. Springer International Publishing; 2015:13-44. [Google Scholar]
44.Allison PD. Fixed effects models for events history data. In: Fixed Effects Regression Models. SAGE Publications; 2009. doi: 10.4135/9781412993869.d26 [DOI] [Google Scholar]
45.Funder DC, Ozer DJ. Evaluating effect size in psychological research: sense and nonsense. Adv Methods Pract Psychol Sci. 2019;2(2):156-168. doi: 10.1177/2515245919847202 [DOI] [Google Scholar]
46.Guyatt GH, Osoba D, Wu AW, Wyrwich KW, Norman GR; Clinical Significance Consensus Meeting Group . Methods to explain the clinical significance of health status measures. Mayo Clin Proc. 2002;77(4):371-383. doi: 10.4065/77.4.371 [DOI] [PubMed] [Google Scholar]
47.Foverskov E, White JS, Norredam M, et al. Neighbourhood socioeconomic disadvantage and psychiatric disorders among refugees: a population-based, quasi-experimental study in Denmark. Soc Psychiatry Psychiatr Epidemiol. Published online May 21, 2022. doi: 10.1007/s00127-022-02300-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Clampet-Lundquist S, Kling JR, Edin K, Duncan GJ. Moving teenagers out of high-risk neighborhoods: how girls fare better than boys. AJS. 2011;116(4):1154-1189. doi: 10.1086/657352 [DOI] [PubMed] [Google Scholar]
49.Leventhal T, Dupéré V. Neighborhood effects on children’s development in experimental and nonexperimental research. Annu Rev Dev Psychol. 2019;1(1):149-176. doi: 10.1146/annurev-devpsych-121318-085221 [DOI] [Google Scholar]
50.de Montgomery CJ, Petersen JH, Jervelund SS. Psychiatric healthcare utilisation among refugee adolescents and their peers in Denmark. Soc Psychiatry Psychiatr Epidemiol. 2020;55(11):1457-1468. doi: 10.1007/s00127-020-01878-w [DOI] [PubMed] [Google Scholar]
51.Barghadouch A, Kristiansen M, Jervelund SS, Hjern A, Montgomery E, Norredam M. Refugee children have fewer contacts to psychiatric healthcare services: an analysis of a subset of refugee children compared to Danish-born peers. Soc Psychiatry Psychiatr Epidemiol. 2016;51(8):1125-1136. doi: 10.1007/s00127-016-1260-1 [DOI] [PubMed] [Google Scholar]
52.Schmidt M, Schmidt SAJ, Adelborg K, et al. The Danish health care system and epidemiological research: from health care contacts to database records. Clin Epidemiol. 2019;11:563-591. doi: 10.2147/CLEP.S179083 [DOI] [PMC free article] [PubMed] [Google Scholar]
53.United Nations High Commissioner for Refugees . Global trends: forced displacement in 2020. Accessed January 25, 2022. https://www.unhcr.org/60b638e37/unhcr-global-trends-2020

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials