REVIEW

Systematic Review and Meta-analysis: The Prevalence

of Mental Illness in Child and Adolescent Refugees and
Asylum Seekers
Rebecca Blackmore, BaSocSci (Hons), Kylie M. Gray, PhD, Jacqueline A. Boyle, PhD,
Mina Fazel, MRCPsych, Sanjeeva Ranasinha, MSc (Applied Statistics), Grace Fitzgerald, MBBS,
Marie Misso, PhD, Melanie Gibson-Helm, PhD

Objective: Over half of the world’s refugee population are under the age of 18 years. This systematic review aims to summarize the current body of
evidence for the prevalence of mental illness in child and adolescent refugee populations.
Method: Eight electronic databases, gray literature, and Google Scholar were searched for articles from 1 January 2003 to 5 February 2018. Strict
inclusion criteria regarding the diagnosis of mental illness were imposed. Study quality was assessed using a template according to study design, and
study heterogeneity using the I2 statistic. Random effects meta-analyses results were presented given heterogeneity among studies. The protocol for this
systematic review was registered with PROSPERO (CRD42016046349).
Results: Eight studies were eligible, involving 779 child and adolescent refugees and asylum seekers, with studies conducted in 5 countries. The overall
prevalence of posttraumatic stress disorder (PTSD) was 22.71% (95% CI 12.79
32.64), depression 13.81% (95% CI 5.96
21.67), and anxiety
disorders 15.77% (95% CI 8.04
23.50). Attention-deficit/hyperactivity disorder (ADHD) was 8.6% (1.08
16.12) and oppositional defiant disorder
(ODD) was 1.69% (95% CI
0.78 to 4.16). Because of the high heterogeneity, further subgroup analyses were conducted.
Conclusion: Refugee and asylum seeker children have high rates of PTSD, depression, and anxiety. Without the serious commitment by health and
resettlement services to provide early support to promote mental health, these findings suggest that a high proportion of refugee children are at risk for
educational disadvantage and poor social integration in host communities, potentially affecting their life course.
Key words: refugee, systematic review, mental illness
J Am Acad Child Adolesc Psychiatry 2020;59(6):705–714.

hildren and adolescents comprise 51% of the Despite comprising half the world’s refugee and asylum
C world’s refugees and asylum seekers,1 many of
whom have experienced significant social and
emotional disruption. Many are exposed to potentially
seeker population, there is a lack of high-quality prevalence
estimates of mental illness in refugee children and adoles-
cents. There is therefore a lack of clarity on the extent of the
traumatic experiences, harsh living conditions, deprivation public health problem when new, forcibly displaced pop-
of basic health care, separation from or loss of family ulations of children and adolescents arrive in a country of
members, and an interrupted education.2,3 Unaccompanied resettlement, potentially hampering efforts to improve
refugee minors face even greater risks, including increased awareness among the populations and institutions of host
vulnerability to traffickers and exploitation without the countries, to advocate for services, and to mobilize resources
possibly protective buffer of primary caregivers.4 Experi- within health systems and resettlement programs.
encing violence and instability at a young age and at critical Previous systematic and narrative reviews examining the
points in a child’s cognitive and emotional development can prevalence of mental illness in child and adolescent refugee
have a lasting psychological impact.5,6 Although many populations have three limitations: (1) a reliance on self-
young people demonstrate resilience and an ability to report questionnaires with cutoff scores to determine di-
overcome the chaos and challenges of forced displacement, agnoses that might not be validated for the populations
others are at risk for developing serious mental illness.7-9 being studied10-12; (2) a focus on specific geographical

Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 705
Volume 59 / Number 6 / June 2020
BLACKMORE et al.

locations of refugee populations13; and (3) a focus on METHOD

posttraumatic stress disorder (PTSD) rather than the Search Strategy and Selection Criteria
breadth of mental illness.6 The protocol for this systematic review and meta-analysis
Previous research has largely relied on the use of symp- was registered with PROSPERO CRD42016046349
tom rating measures and self-report questionnaires. This can (https://www.crd.york.ac.uk/prospero/display_record.php?
explain some of the variability in the reported prevalence of RecordID¼46349). We followed the Preferred Reporting
mental illness in this population, as these types of measures Items for Systematic Reviews and Meta-Analyses
tend to overestimate symptomatology.14-16 Self-report mea- (PRISMA).22 A PRISMA checklist is provided in
sures, although practical for their ease of administration, Table S1 (available online). The protocol was based on that
particularly in low-resource settings, are prone to response used in the earlier systematic review by Fazel et al.21 (which
inaccuracy and issues with item interpretation, which can included adult, child, and adolescent refugee populations)
result in false-positive results.17 Individual clinical interviews but expanded the range of databases searched and the
using validated assessment measures are viewed as the gold number of search terms, imposed stricter criteria for mental
standard in assessing mental illness in children and adoles- illness diagnosis, and placed no restrictions on resettlement
cents.18 A recent review on the epidemiology of PTSD and countries or publication language. Studies pertaining to
depression in refugee minors highlighted the need for more child and adolescent populations were extracted for this
rigorous research involving professionals trained in clinical specific systematic review and meta-analysis.
mental health assessments, to increase the accuracy of the We undertook a systematic search, developed using the
reported prevalence of these conditions.19 OVID platform, combining Medical Subject Headings
The global conflicts forcing increasing numbers of (MeSH terms) with related text-based search words, and
populations to become displaced are changing in nature, translated to other databases. An example of the search
with the involvement of more civilian populations and strategy used in MEDLINE is provided in Table S2,
increasing targeting of children.20 Therefore, establishing available online. Studies about refugees or asylum seekers in
more current prevalence estimates to inform public policy, combination with terms related to mental illness, diagnosis,
in terms of resettlement support and mental health care, for and trauma were searched across eight databases: MED-
refugee populations is needed. Furthermore, although the LINE; MEDLINE In-Process; EBM Reviews; Embase;
vast majority of refugees live in low- and middle-income PsycINFO; CINAHL; PILOTS, and Web of Science. Any
countries, these populations have often been omitted from mental illness that met our criteria for diagnostic assessment
systematic reviews and studies; hence their inclusion here. was accepted. Searching the gray literature using the data-
Previous systematic reviews have focused on PTSD. A base Open Grey (http://www.opengrey.eu/) as well as
comprehensive 2005 systematic review and meta-analysis of Google Scholar allowed for the possible inclusion of reports
adult, child, and adolescent refugees resettled in western from refugee, government, and other nongovernment
countries identified only 5 studies reporting the prevalence agencies. The date limits were January 1, 2003, to February
of PTSD in children and none reporting diagnoses of other 5, 2018. This start date reflects the end date of the search
mental illnesses (from 1996 to 2002).21 These studies conducted by Fazel et al.21 Therefore, none of the included
provided data on a total of 260 refugee children, of whom studies from the earlier systematic review were included in
11% were diagnosed with PTSD. The samples of children this analysis, to provide a contemporary estimate on mental
were drawn from Bosnia, Central America, Iran, Kurdistan, illness within this population. The reference lists of 67
and Rwanda. It is therefore timely to update and expand systematic reviews, 13 of which pertained to child and
that review. adolescent samples, were identified through the search and
We have therefore conducted a systematic review that screened for any further relevant articles. This resulted in
aims to overcome some of the methodological limitations of the manual checking of 17 additional articles, 8 of which
the current body of evidence and to establish a current es- were relevant to child and adolescent populations.
timate of mental illness in child and adolescent refugee Studies were included in this review if the sample
populations. To our knowledge, this is the first systematic included refugee and/or asylum seeker children and/or ad-
review to use rigorous inclusion criteria regarding the olescents; if the study had a sample size larger than 50, to
method of diagnosis determination, the first to place no overcome potential sampling bias associated with smaller
limits on country of origin or resettlement, to place no sample sizes21; and if the study reported quantitative prev-
limits on publication language, and to include a review of alence estimates of a mental illness as classified by the
gray literature. DSM23 or the International Classification of Diseases

706 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry
Volume 59 / Number 6 / June 2020
 MENTAL ILLNESS REVIEW OF CHILD REFUGEES

(ICD).24 The diagnosis must have been made as a result of a of mental illness calculated with 95% CIs in the pooled
clinical interview using a validated diagnostic assessment data. Random effects were presented given heterogeneity
measure. Studies that based diagnoses solely on self-report among studies. This statistical model is based on the
questionnaires or symptomatology rating scales were assumption that the samples of the included studies are
excluded. The interview needed to have been conducted drawn from different populations.33 Heterogeneity was
either by a mental health professional (psychiatrist, psy- assessed using the I2 statistic.34 The I2 value provides a
chologist, psychiatric nurse) or other trained para- measure of the variation explained by the differences be-
professional (psychology research assistant, trained tween the included studies rather than chance. In the case of
researcher). In studies that administered the World Health 5 or more studies being available, publication bias was
Organization, World Mental Health Composite Interna- assessed by visual inspection of funnel plots and applying
tional Diagnostic Interview (WHO WMH-CIDI),25 non- the Egger test35 set at a threshold of a p < .05 to indicate
clinicians who had completed official WHO-recommended funnel plot asymmetry. Prevalence rates were for current
training requirements were accepted. The WHO WMH- diagnoses and were combined by direct summation of nu-
CIDI is a fully structured interview for the assessment of merators and denominators across studies, thereby
mental disorders intended for use by trained lay in- providing a pooled estimate.
terviewers. Studies were selected if they had recruited Possible sources of heterogeneity between studies were
representative samples of refugee children; hence those investigated, where reported data allowed, by subgroup
recruiting participants solely from medical clinics were analyses. These included the following: sex, duration of
excluded to reduce selection bias. Studies stating that the displacement (timeframes to be determined by the reported
sample included asylum seekers whose applications had data), visa status, use of native interviewer (whereby the
been rejected were excluded if the results were not dis- diagnostic interview was conducted in the preferred lan-
aggregated, or if the mental health assessment was not guage of the child or adolescent), and current residence
conducted prior to rejection (when the individuals met the status (residing in the local community versus refugee fa-
definition of asylum seekers). Qualitative or case report cility/reception center).
studies were excluded. When multiple articles used data
from the same study, the article providing data that best met Risk-of-Bias Appraisal
the search criteria was included. Methodological quality of the included studies was assessed
Two reviewers (R.B. and M.G.H./G.F.) independently by two independent reviewers (R.B. and K.M.G.) using a
assessed all the titles, abstracts, and key words of every risk-of-bias assessment template (Table S3, available online)
article retrieved against the selection criteria. Full-text arti- according to study design.36 This template incorporates the
cles were then assessed if the title and abstract suggested that Newcastle
Ottawa Scale (NOS) for assessing the quality of
the study met the selection criteria or if there was any doubt nonrandomized studies in meta-analyses and includes
regarding eligibility of the article. Disagreements were additional risk-of-bias components.37 It has been used in
resolved by discussion and, where appropriate, we contacted international evidence-based guidelines and other systematic
the study authors for further information. The reviewers reviews.38-40 Individual items related to study quality such
contacted eight study authors to obtain further information as internal and external validity, reporting bias, confound-
regarding methodology and data, of whom seven respon- ing, and conflict of interest were assessed. Studies were
ded.26-32 Studies in languages other than English were assigned a rating of low, moderate, or high risk of bias. Any
assessed first by a native speaker where possible or via disagreement was to be reviewed by a third author, but this
Google translate, and then officially translated by a profes- was not required.
sional translation service if they potentially met inclusion
criteria. RESULTS
The entire search, including electronic databases and other
Data Extraction and Analysis sources, yielded a total of 17,423 documents. A flowchart
Two review authors (R.B. and M.G.H.) independently outlining the search results and selection of studies is pro-
extracted statistical data from the included studies into Stata vided in Figure 1. After removing duplicates, 10,419 doc-
software version 14.1 (StataCorp LP, College Station, TX) uments were excluded based on title and abstract, and a
for the meta-analysis. Study characteristics such as sample further 965 documents were screened by full-text retrieval.
size, sampling framework, diagnostic instrument, diagnostic Eight studies pertaining to child and adolescent populations
criteria, and use of native interviewer were also extracted. met the inclusion criteria,29-32,41-44 after a final exclusion of
Meta-analysis results were expressed as prevalence estimates three papers that reported duplicate datasets.45-47 One study
Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 707
Volume 59 / Number 6 / June 2020
BLACKMORE et al.

FIGURE 1 Search Results and Selection of Studies sample originating from up to 15 different countries
(12.5%).30 In 2 of the studies, a proportion of the partic-
ipants had been born in the host nation (4.8%, n ¼ 5,30
and 31.4%, n ¼ 3231). Disaggregated data were not re-
ported in these studies.
Seven diagnostic measures were used to assess mental
illness (Table S4, available online). One of the measures, the
Posttraumatic Stress Symptoms in Children (PTSS-C) was
specifically developed as a cross-cultural, semi-structured
interview to diagnose PTSD.31 Six studies made mention of
the psychometric properties of the instruments used and/or
previous use with child refugee populations.29-31,41-43 Four
studies conducted the diagnostic interview in the native
language of the child or adolescent,31,41,43,44 three with
assistance from interpreters29,30,42 and one with a combi-
nation of native interviewers and interpreters.32
PTSD was investigated in 7 studies, with data for a
total of 681 children and adolescents.29,31,32,41,42,44
Overall, 22.71% (95% CI 12.79
32.64) were diagnosed
with PTSD (Figure 2). Participants had a weighted mean
age of 12.3 years, and 40% were girls. There was sub-
stantial heterogeneity between the studies (I2 ¼ 91.1%,
p ¼ .000); therefore, subgroup analyses were conducted for
duration of displacement, visa status, use of native inter-
viewer, and current residence (Figure 3). The subgroup
analysis for sex could not be conducted because of a lack of
was published in German and professionally translated for reported data. The PTSD prevalence was higher for those
inclusion.30 The 8 eligible studies provided data on 779 displaced less than 2 years and for those with an insecure
child and adolescent refugees and asylum seekers. Six visa status. Conducting the diagnostic interview in the
studies included both children and adolescents in their native language of the child or adolescent and current
samples30-32,41,43,44; however 2 studies recruited solely community residence resulted in lower reported prevalence
adolescents, 10 to 19 years42 and 15 to 18 years.29 The of PTSD.
included studies provided data on the following range of Five studies of depression were identified, providing
mental illnesses: PTSD, depression, anxiety disorders, data for a total of 492 children and adolescents.29,30,42-44
attention-deficit/hyperactivity disorder (ADHD), and Overall, 13.81% (95% CI 5.96
21.67) were diagnosed
oppositional defiant disorder (ODD). with depression (Figure 4). Participants had a weighted
Characteristics of the included studies are provided in mean age of 12.9 years, and 36% of the sample were girls.
Table S3 (available online). The age range for each sample One study reported sex data for the total sample (N ¼ 104);
was consistent with the World Health Organization defi- however, only 98 of these individuals completed the full
nition of child or adolescent (19 years or younger)48 except diagnostic assessment.30 Two studies reported prevalence of
for one sample, from a study specifically recruiting unac- major depression,30,42 two studies reported prevalence of
companied asylum-seeking children,29 that contained some any depressive disorder,43,44 and one study reported prev-
participants 20 years of age. Considering the focus of the alence of major depression and dysthymia, which was
study and that the mean age of participants (16.23  0.83) combined for the analysis.29 There was substantial hetero-
was within the adolescent range, the authors decided to geneity between the studies (I2 ¼ 86.5%, p ¼ .000), so
include it. Studies were undertaken in 5 countries: Ger- subgroup analyses were conducted for duration of
many (197 refugees),30,32 Malaysia (90),42 Norway (160),29 displacement, visa status, use of native interviewer, and
Sweden (191),31,41 and Turkey (144).43,44 Refugee samples current residence (Figure 5). The subgroup analysis for sex
were drawn from the Middle East (45%), a combination of could not be conducted because of a lack of reported data.
Middle Eastern and African countries (31%), and Southern Depression prevalence was higher for those displaced less
Asia and the Middle East (11.5%). One study reported a than 2 years, those with refugee visa status, use of native
708 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry
Volume 59 / Number 6 / June 2020
 MENTAL ILLNESS REVIEW OF CHILD REFUGEES

FIGURE 2 Prevalence of Posttraumatic Stress Disorder in Child and Adolescent Refugees and Asylum Seekers

Note: Horizontal lines indicate 95% CIs; open diamond denotes subtotals.

interviewer for diagnostic assessment, and current com- breakdown of diagnosis across individual anxiety disor-
munity residence. ders,29,30 which were combined for analysis. These disorders
Four studies reporting the prevalence of anxiety disor- included: generalized anxiety disorder, separation anxiety
ders were identified, consisting of data for a total of 402 disorder, obsessive compulsive disorder, social anxiety dis-
children and adolescents.29,30,43,44 Overall, 15.77% (95% order, agoraphobia, and specific phobia. Participants had a
CI 8.04
23.50) were diagnosed with an anxiety disorder weighted mean age of 12.7 years, and 32% of the sample
(Figure S1, available online). Two studies reported diagnosis were girls. There was substantial heterogeneity between the
of any anxiety disorder,43,44 and two studies provided a studies (I2 ¼ 76.0%, p ¼ .006), so subgroup analyses were

FIGURE 3 Prevalence of Posttraumatic Stress Disorder by Various Study Characteristics

Note: p values are derived from random-effects models. Horizontal lines indicate 95% CIs. One study29 was not included in this analysis, as both native interviewers and
interpreters were used for the assessment, and disaggregated data were not available.

Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 709
Volume 59 / Number 6 / June 2020
BLACKMORE et al.

FIGURE 4 Prevalence of Depression in Child and Adolescent Refugees and Asylum Seekers

Note: Horizontal lines indicate 95% CIs; open diamond denotes subtotals.

conducted for duration of displacement, visa status, use of of the sample were girls. There was substantial heterogeneity
native interviewer, and current residence status (Figure S2, among the studies (I2 ¼ 84.9%, p ¼ .000), so subgroup
available online). Anxiety prevalence was higher for those analyses were conducted for duration of displacement, visa
displaced less than 2 years, with refugee visa status, use of status, use of native interviewer, and current residence sta-
native interviewer for diagnostic assessment, and current tus. The ADHD prevalence was higher for those displaced
community residence. longer than 2 years and for those with an insecure visa status
Four studies reporting the prevalence of ADHD were (Figure S4, available online).
identified, consisting of data for a total of 322 children and Two studies reported prevalence of ODD, consisting of
adolescents.30,41,43,44 Overall, 8.6% (95% CI 1.08–16.12) data for a total 178 children and adolescents.30,41 Overall,
were diagnosed with ADHD (Figure S3, available online). 1.69% (95% CI
0.78 to 4.16) were diagnosed with ODD
Participants had a weighted mean age of 11 years, and 52% (Figure S5, available online). There was a low level of

FIGURE 5 Prevalence of Depression by Various Study Characteristics

Note: P values are derived from random-effects models. Horizontal lines indicate 95% CIs.

710 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry
Volume 59 / Number 6 / June 2020
 MENTAL ILLNESS REVIEW OF CHILD REFUGEES

heterogeneity between the studies (I2 ¼ 9.6%, p ¼ .293). PTSD, depression, and anxiety disorders were all higher
Subgroup analyses were not conducted, as there were only for those displaced less than 2 years, compared to those
two studies. displaced more than 2 years. However, the prevalence of
ADHD was higher among those displaced more than 2
Publication Bias years. This may be a result of the small number of included
There is some evidence of publication bias for PTSD and studies. Alternatively, this might be partly explained by the
depression, based on the results of the Egger test where the phenomena of spontaneous or natural recovery, which can
p values were less than .05 (Egger test plots provided in occur in some cases of PTSD and depression,49,50 whereas
Figures S6 and S7, available online). The funnel plots ADHD persists in childhood with some change in presen-
showed some asymmetry in the scatter of the studies, which tation as individuals become older.51 PTSD was higher for
can be an indication of publication bias; hence a search of those with insecure visa status and temporary residence.
gray literature was conducted. These results, however, This was not the case for depression and anxiety disorders,
should be interpreted with caution, as they may instead be a which were higher for those with refugee visa status and
result of the small number of studies published in the field. community residence. Rigorous longitudinal research is
required to truly understand the relationship between
Risk of Bias refugee experiences, different mental illnesses, visa status
Five of eight studies were assigned a low risk of bias and and resettlement experiences, and trajectories of recovery.
determined to be of high quality.29-32,42 Three studies The results from our systematic review show higher
demonstrated moderate risk of bias.41,43,44 Moderate ratings prevalence for PTSD, depression, anxiety, and ADHD in
were assigned because there were potential issues with the refugee and asylum seeker populations compared to data in
representativeness of the samples. In one study, there was a the literature for nonrefugee populations. A recent meta-
high rate of nonparticipation.43 Interviews were conducted analysis of 3,563 trauma-exposed children and adolescents
with a small number of participants compared to the total reports a PTSD prevalence of 15.9%.52 Our systematic
number screened. In the second study, immigration lists review found an overall PTSD prevalence of 22.71% and
were unable to be used for recruitment because of the high even higher for those displaced less than 2 years (35%).
mobility of the population.44 Therefore, only families who However, direct comparisons of PTSD rates are challenging,
had registered with the obstetrics and gynecology de- as prevalence varies according to trauma type and sex.52 In
partments of the local hospitals were contacted for recruit- regard to the other mental illnesses, the reported worldwide
ment. The third study investigated the comorbidity of pooled prevalence of any depressive disorder in general
PTSD and ADHD and the impact of a parental history of populations of children and adolescents is 2.6%, any anxiety
torture.41 This study recruited a group of children who had disorder 6.5%, ADHD 3.4%, and ODD 3.6%.53 Although
at least one parent with a history of torture, and a com- our results were higher for depressive disorders (13.81%),
parison group of children whose parents had no history of anxiety disorders (15.77%), and ADHD (8.6%), the prev-
torture. It could therefore be argued that a proportion of the alence of ODD (1.69%) was comparable to the general
sample had a greater risk of mental illness due to parental population estimate, and in fact slightly lower in refugee
experience of torture. No study was assigned a high risk and asylum seeker children. Overall, these comparison data
of bias. confirm that refugee children and adolescents have sub-
stantial need for mental health services.
DISCUSSION These findings build on the previous systematic review
This systematic review aimed to overcome some of the by Fazel et al.,21 by providing an updated prevalence esti-
methodological limitations of the current body of evi- mate for PTSD as well as new estimates for depression and
dence, and to establish new estimates of mental illness in other mental illnesses, based on rigorous diagnostic
child and adolescent refugee populations. The review methods. The prevalence of PTSD reported in this sys-
identified a limited number of high-quality studies tematic review is higher. The fact that this current review
measuring prevalence estimates of mental illness, despite was able to contribute data for the prevalence of depression,
the substantial number of children and adolescents dis- anxiety, ADHD, and ODD highlights some growth within
placed globally. Overall, the findings have shown that the field. More than half of the world’s current refugee and
refugee and asylum seeker children have high rates of asylum seeker population are aged 18 years and less1; yet
PTSD (22.71%), anxiety (15.77%), and depres- this review was able to pool data for only 779 children and
sion (13.81%). adolescents. This not only limited the precision of our

Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 711
Volume 59 / Number 6 / June 2020
BLACKMORE et al.

estimates, but demonstrates that this is an underrepresented limited recognition of cultural perspectives.59 The diag-
population in the research literature. nostic framework for PTSD was largely established using
Our subgroup analyses and quality assessments high- adult military samples and single-incident trauma survivors
light the effect that study design can have on prevalence from high-income nations, which may not adequately
rates and the importance of considering these factors when capture the posttraumatic symptomatology of refugee
interpreting current evidence or designing new studies. youths.60
Previous research has suggested that the use of native in- Our quality assessment identified a few limitations of
terviewers during mental health assessments results in lower individual studies that are worth noting here. First, one
reported prevalence of mental illness.21 This was the case for study compared children with and without parents who had
PTSD, whereby lower prevalence was reported when the a history of torture.41 Therefore it was not a truly repre-
interviews were conducted in the child’s or adolescent’s sentative community sample, which may explain the high
native language. However, for depression and anxiety, the rates of PTSD and ADHD reported in this study. Another
prevalence was higher when the interview was conducted in study, which reported a high prevalence of depression in
the native language, and for ADHD the results were not comparison to that in the other included studies, described
significant. issues with a low participation rate from the initial screened
Although our intention was to determine the current study sample.43 It is possible that families with children or
prevalence of mental illnesses, the generalizability of the adolescents experiencing more severe mental distress had a
results was limited by the lack of studies. Few restrictions greater motivation to participate. The reported PTSD
were placed on characteristics of the refugee experience, in prevalence from this study was also high in comparison to
the hope of including all possible studies, and as a result the other studies.
meta-analysis yielded expectedly high statistical heteroge-
neity. Taking this heterogeneity into account, random ef- Implications for Clinical Practice and Policy
fects models were used in order to calculate conservative Despite the limitations, the findings have implications for
confidence intervals. Meta-regression is often used to assess clinical practice and resettlement policy. There are imme-
sources of heterogeneity, but was not possible because of diate and detrimental effects of pre-migration trauma,
limited covariates reported in the included studies. forced displacement, and the postmigration environment on
Although subgroup analyses were conducted to investigate the mental health of child and adolescent refugee and
the possible sources of heterogeneity, some subgroup ana- asylum seekers. With the exception of ADHD, all of the
lyses were also not possible due to a lack of reported data, other mental illnesses showed higher prevalence for in-
such as sex. As PTSD prevalence rates have been shown to dividuals recently displaced (2 years or less), emphasizing
vary by sex, future research should include this analysis.52 the need for early support following arrival into a country of
Subgroup analysis for country of origin could not be un- refuge. This support may involve adequately resourced
dertaken, as many of the study samples were highly diverse. refugee centers designed to protect children from further
Relevant information pertaining to some aspects of the traumatization and to address the many facets of resettle-
refugee experience were lacking, such as number of re- ment stressors that can increase risk of poor social integra-
locations, which has been shown to increase risk of devel- tion and educational disadvantage.55,61 The high prevalence
oping mental illness.54 There was limited information of mental illness within this population leads also to the
reported on the prevalence of comorbid illnesses, with only need to provide youth- and refugee-appropriate, cross-
two studies providing such data.41,43 Only one study culturally valid screening in refugee centers to streamline
explored the relationship between parent mental health and allocation to clinical assessment and treatment services.
child mental health outcomes.41 The role of family,
particularly parental psychopathology, is crucial in its ability Implications for Future Research
to mediate or exacerbate child mental health symptoms.6,55-58 To determine more accurate prevalence estimates, there is
The decision to include the two studies,30,31 in which a an urgent need for large studies that incorporate random
proportion of the sample had been born in the host nations, sampling of populations, use rigorous diagnostic methods,
may affect the generalizability of the results. clearly characterize the study sample, report relevant de-
The cross-cultural application of a western psychiatric tails of the refugee experience, and provide diagnosis data
framework must also be acknowledged as a limitation. Ef- based on sex and age. The limited body of high-quality
forts to enhance cultural validity have been made in the research in this field is a barrier to the provision of
DSM-5; however the included studies applied DSM-IV and appropriate and informed mental health care for this
DSM-III diagnostic criteria, which have been criticized for population.8 The results have shown that, whereas the
712 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry
Volume 59 / Number 6 / June 2020
 MENTAL ILLNESS REVIEW OF CHILD REFUGEES

limited literature available focuses on PTSD, depression Accepted November 19, 2019.
and anxiety disorders also appear to be highly prevalent Drs. Boyle, Misso, Gibson-Helm, Mss. Blackmore, Fitzgerald, and Mr. Ranasi-
within this population. This highlights the need for nha are with Monash Centre for Health, Research & Implementation, School of
Public Health and Preventive Medicine, Monash University, Kanooka Grove,
future research to investigate the full range of mental Clayton, Victoria, Australia. Dr. Gray is with the Centre for Developmental
illnesses. The majority of studies in this field are often Psychiatry & Psychology, School of Clinical Sciences at Monash Health, Mon-
ash University, Monash Medical Centre, Clayton, Victoria, Australia, and the
undertaken in high-income host countries, which may Centre for Educational Development, Appraisal and Research (CEDAR), Uni-
not be countries of first asylum and often differ in terms versity of Warwick, Coventry, UK. Dr. Fazel is with the University of Oxford, UK.

of postsettlement support. Although the studies of this The authors have reported no funding for this work.

review included key refugee host nations such as Turkey Mr. Ranasinha served as the statistical expert for this research.
and Malaysia, future research needs to be conducted in The authors thank the following authors for providing additional information
regarding their studies: Prof. Hubertus Adam, PhD, of Martin Gropius Kran-
less-resourced host nations such as Pakistan, Lebanon, kenhaus, Associate Professor Abdulbaghi Ahmad, PhD, of the University of
Libya, and Ethiopia. Further research is also needed to Uppsala, and Dr. Joseph Ssenyonga, PhD, of the Mbarara University of Science
and Technology. The authors would also like to thank Ms. Anne Young, from
identify appropriate and effective real-world examples of the Monash University library, for her assistance with conducting the database
mental health support services for refugee children and search. The authors sincerely thank the following individuals who assisted with
the screening of articles across a number of languages: Dr. Chau Tay, MBBS, of
adolescents, and their families. Monash University, Ms. Daisy Coles, Ms. Natalie Pekin, BSc, of Monash Uni-
The results of this review contribute not just an updated versity, Dr. Karin Hammarberg, PhD, of Monash University, Ms. Karin Stanzel, M
PublicHlth, of Monash University, Dr. Rashad Hasanov, PhD, of Deakin Uni-
prevalence estimate for PTSD but the largest analysis of versity, and Dr. Craig Pickett, PhD, of Edith Cowan University and Victoria
PTSD prevalence for this population, based on rigorous University. Also, a thank you to Ms. Emily Gilbert, from Monash University, who
assisted with the screening of the gray literature and Google search results.
diagnostic methods. It has also expanded the current evi-
Disclosure: Drs. Boyle and Gibson-Helm are National Health and Medical
dence base by contributing prevalence estimates for Research Council Fellows. Ms. Blackmore has received funding by scholarships
depression, anxiety disorders, ADHD, and ODD. Overall, from Australian Rotary Health (Ian Scott, PhD Scholarship), the Windermere
Foundation, and the Monash Centre for Health Research and Implementation.
the findings confirm that refugee and asylum seeker chil- Drs. Gray, Fazel, Fitzgerald, and Misso and Mr. Ranasinha have reported no
dren have high rates of PTSD, depression, and anxiety. biomedical financial interests or potential conflicts of interest.

Without a serious commitment by health and resettlement Correspondence to Melanie Gibson-Helm, PhD, Monash Centre for Health,
Research & Implementation, School of Public Health and Preventive Medicine,
services to provide early support to promote mental health, Monash University, Level 1, 43-51 Kanooka Grove, Clayton Victoria, Australia;
these findings suggest that a high proportion of refugee e-mail: [email protected]

children are at risk for educational disadvantage and poor 0890-8567/$36.00/ª2019 American Academy of Child and Adolescent
Psychiatry
social integration in host communities, potentially further
https://doi.org/10.1016/j.jaac.2019.11.011
affecting their life course.

REFERENCES
1. United Nations High Commissioner for Refugees. Mid-year trends 2018. Available at: 11. Giacco D, Laxhman N, Priebe S. Prevalence of and risk factors for mental disorders in
https://www.unhcr.org/statistics/unhcrstats/5c52ea084/mid-year-trends-2018.html. refugees. Semin Cell Dev Biol. 2018;77:144-152.
Accessed February 7, 2019. 12. Turrini G, Purgato M, Ballette F, Nos
e M, Ostuzzi G, Barbui C. Common mental
2. Bean T, Derluyn I, Eurelings-Bontekoe E, Broekaert E, Spinhoven P. Comparing psy- disorders in asylum seekers and refugees: umbrella review of prevalence and intervention
chological distress, traumatic stress reactions, and experiences of unaccompanied refugee studies. Int J Ment Health Syst. 2017;11:51.
minors with experiences of adolescents accompanied by parents. J Nerv Ment Dis. 2007; 13. Kien C, Sommer I, Faustmann A, et al. Prevalence of mental disorders in young refugees
195:288-297. and asylum seekers in European countries: a systematic review. Eur Child Adolesc
3. Measham T, Guzder J, Rousseau C, Pacione L, Blais-McPherson M, Nadeau L. Refugee Psychiatry. 2019;28:1295-1310.
children and their families: supporting psychological well-being and positive adaptation 14. de Jong JTVM, Komproe IH, Van Ommeren M. Common mental disorders in post-
following migration. Curr Probl Pediatr Adolesc Health Care. 2014;44:208-215. conflict settings. Lancet. 2003;361:2128-2130.
4. Mitra R, Hodes M. Prevention of psychological distress and promotion of resilience 15. Lim GY, Tam WW, Lu Y, Ho CS, Zhang MW, Ho RC. Prevalence of depression
amongst unaccompanied refugee minors in resettlement countries. Child Care Health in the community from 30 countries between 1994 and 2014. Sci Rep. 2018;
Dev. 2019;45:198-215. 8:2861.
5. Eruyar S, Huemer J, Vostanis P. Review: how should child mental health services 16. Richardson LK, Frueh BC, Acierno R. Prevalence estimates of combat-related post-
respond to the refugee crisis? Child Adolesc Ment Health. 2018;23:303-312. traumatic stress disorder: critical review. Aust N Z J Psychiatry. 2010;44:4-19.
6. Reed RV, Fazel M, Jones L, Panter-Brick C, Stein A. Mental health of displaced and 17. Fan X, Miller BC, Park K-E, et al. An exploratory study about inaccuracy and invalidity
refugee children resettled in low-income and middle-income countries: risk and pro- in adolescent self-report surveys. Field Methods. 2006;18:223-244.
tective factors. Lancet. 2012;379:250-265. 18. Zuckerbrot RA, Jensen PS. Improving recognition of adolescent depression in primary
7. Betancourt TS, Newnham EA, Layne CM, et al. Trauma history and psychopathology in care. JAMA Pediatr. 2006;160:694-704.
war-affected refugee children referred for trauma-related mental health services in the 19. Reavell J, Fazil Q. The epidemiology of PTSD and depression in refugee minors who
United States. J Trauma Stress. 2012;25:682-690. have resettled in developed countries. J Ment Health. 2017;26:74-83.
8. Fazel M, Betancourt TS. Preventive mental health interventions for refugee children and 20. Guha-Sapir D, Schl€ uter B, Rodriguez-Llanes JM, Lillywhite L, Hicks MH-R. Patterns of
adolescents in high-income settings. Lancet Child Adolesc Health. 2018;2:121-132. civilian and child deaths due to war-related violence in Syria: a comparative analysis from
9. Karam EG, Mneimneh ZN, Dimassi H, et al. Lifetime prevalence of mental disorders in the Violation Documentation Center dataset, 2011-2016. Lancet Glob Health. 2018;6:
Lebanon: first onset, treatment, and exposure to War. PLoS Med. 2008;5:e61. e103-e110.
10. Bronstein I, Montgomery P. Psychological distress in refugee children: a systematic re- 21. Fazel M, Wheeler J, Danesh J. Prevalence of serious mental disorder in 7000 refugees
view. Clin Child Fam Psychol Rev. 2011;14:44-56. resettled in western countries: a systematic review. Lancet. 2005;365:1309-1314.

Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 713
Volume 59 / Number 6 / June 2020
BLACKMORE et al.

22. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred Reporting Items for Systematic 41. Daud A, Rydelius PA. Comorbidity/overlapping between ADHD and PTSD in relation
Reviews and Meta-Analyses: the PRISMA statement. BMJ. 2009;339:b2535. to IQ among children of traumatized/non-traumatized parents. J Atten Disord. 2009;13:
23. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 188-196.
5th ed. Arlington, VA: American Psychiatric Association; 2013. 42. Gosnell NM. Mental health and emotion regulation among refugee students in Malaysia
24. World Health Organization. The ICD-10 classification of mental and behavioural dis- [thesis] 2017. Retrieved from Digital Repository at the University of Maryland. Available
orders: clinical descriptions and diagnostic guidelines. Geneva: World Health Organi- at: https://doi.org/10.13016/M2X862. Accessed February 10, 2018.
zation; 1992. 43. Nasiroglu S, Ceri V. Posttraumatic stress and depression in Yazidi refugees. Neuro-
25. Kessler RC, Ustun TB. The World Mental Health (WMH) survey initiative version of psychiatr Dis Treat. 2016;12:2941-2948.
the World Health Organization (WHO) Composite International Diagnostic Interview 44. Sapmaz SY, Tanriverdi BU, Ozturk M, Gozacanlar O, Ulker GY, Ozkan Y. Immigra-
(CIDI). Int J Methods Psychiatr Res. 2004;13:93-121. tion-related mental health disorders in refugees 5-18 years old living in Turkey. Neu-
26. Muhtz C, von Alm C, Godemann K, et al. Long-term consequences of flight and ropsychiatr Dis Treat. 2017;13:2813-2821.
expulsion in former refugee children. Psychother Psychosom Med Psychol. 2011;61: 45. Daud A, Rydelius PA. Resilience and vulnerability among refugee children of traumatized
233-238. and non-traumatized parents. Child Adolesc Psychiatry Ment Health. 2008;2:7.
27. Ssenyonga J, Owens V, Olema DK. Posttraumatic cognitions, avoidance coping, suicide, 46. Jakobsen M, DeMott MAM, Wentzel-Larsen T, Heir T. The impact of the asylum
and posttraumatic stress disorder among adolescent refugees. In: Maree K, ed. World process on mental health: a longitudinal study of unaccompanied refugee minors in
Conference on Psychology and Sociology. 2012;82:261-265. Available at: https://www. Norway. BMJ Open. 2017;7:8.
sciencedirect.com/science/article/pii/S1877042813013232. Accessed April 20, 2020. 47. Jakobsen M, Meyer DeMott MA, Heir T. Validity of screening for psychiatric disorders
28. Adam H, Klasen F. Trauma und Vers€ohnung: Vers€ohnungsbereitschaft bei traumati- in unaccompanied minor asylum seekers: use of computer-based assessment. Transcult
uchtlingskindern ¼ Trauma and reconciliation: child refugees in Hamburg.
sierten Fl€ Psychiatry. 2017;54:611-625.
Trauma Gewalt. 2011;5:356-369. 48. World Health Organization. Adolescent development. 2017. Available at: https://www.
29. Jakobsen M, Demott MA, Heir T. Prevalence of psychiatric disorders among unac- who.int/maternal_child_adolescent/guidelines/en. Accessed December 15, 2018.
companied asylum-seeking adolescents in Norway. Clin Pract Epidemol Ment Health. 49. Tedeschi FK, Billick SB. Pediatric PTSD: clinical, forensic, and diagnostic understand-
2014;10:53-58. ing. J Am Acad Psychiatr Law. 2017;45:161-169.
30. Ruf M, Schauer M, Elbert T. Prevalence of traumatic stress and mental health 50. Whiteford HA, Harris MG, McKeon G, et al. Estimating remission from untreated
problems in children of asylum-seekers in Germany. Z Klin Psychol Psychother. major depression: a systematic review and meta-analysis. Psychol Med. 2013;43:
2010;39:151-160. 1569-1585.
31. Ahmad A, Von Knorring A-L, Sundelin-Wahlsten V. Traumatic experiences and post- 51. Harpin VA. The effect of ADHD on the life of an individual, their family, and com-
traumatic stress symptoms in Kurdish children in their native country and in exile. munity from preschool to adult life. Arch Dis Child. 2005;90(Suppl 1):i2-i7.
Child Adolesc Ment Health. 2008;13:193-197. 52. Alisic E, Zalta AK, van Wesel F, et al. Rates of post-traumatic stress disorder in trauma-
32. Soykoek S, Mall V, Nehring I, Henningsen P, Aberl S. Post-traumatic stress disorder in exposed children and adolescents: meta-analysis. B J Psychiatr. 2014;204:335-340.
Syrian children of a German refugee camp. Lancet. 2017;389:903-904. 53. Polanczyk GV, Salum GA, Sugaya LS, Caye A, Rohde LA. Annual research review: a
33. Hedges LV, Vevea JL. Fixed- and random-effects models in meta-analysis. Psychol meta-analysis of the worldwide prevalence of mental disorders in children and adoles-
Methods. 1998;3:486-504. cents. J Child Psychol Psychiatr. 2015;56:345-365.
34. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta- 54. Nielsen SS, Norredam M, Christiansen KL, Obel C, Hilden J, Krasnik A. Mental health
analyses. BMJ (Clin Res Ed). 2003;327:557-560. among children seeking asylum in Denmarkethe effect of length of stay and number of
35. Egger M, Smith GD, Schneider M, Minder C. Bias in meta-analysis detected by a simple, relocations: a cross-sectional study. BMC Public Health. 2008;8:293.
graphical test. BMJ. 1997;315:629-634. 55. Fazel M. Focusing a lens on refugee families to address layers of avoidance. Lancet Public
36. Monash Centre for Health Research and Implementation. Evidence synthesis program Health. 2019;4:e318-e319.
template for critical appraisal of a cohort study. Melbourne: Monash Centre for Health 56. Diab M, Peltonen K, Qouta SR, Palosaari E, Punam€aki R-L. Effectiveness of psycho-
Research and Implementation; 2014. social intervention enhancing resilience among war-affected children and the moderating
37. Wells G, Shea B, O’Connell D, et al. The Newcastle-Ottawa Scale (NOS) for assessing role of family factors. Child Abuse Negl. 2015;40:24-35.
the quality of nonrandomised studies in meta-analyses. 2018. Available at: http://www. 57. Thabet AA, Ibraheem AN, Shivram R, Winter EA, Vostanis P. Parenting support and
ohri.ca/programs/clinical_epidemiology/oxford.asp. Accessed December 11, 2018. PTSD in children of a war zone. Int J Soc Psychiatry. 2009;55:226-237.
38. Balen AH, Morley LC, Misso M, et al. The management of anovulatory infertility in 58. Yaylaci FT. Trauma and resilient functioning among Syrian refugee children. Dev Psy-
women with polycystic ovary syndrome: an analysis of the evidence to support the chopathol. 2018;1-14.
development of global WHO guidance. Hum Reprod Update. 2016;22:687-708. 59. Alarcon RD. Culture, cultural factors and psychiatric diagnosis: review and projections.
39. Teede HJ, Misso ML, Costello MF, et al. Recommendations from the international World Psychiatry. 2009;8:131-139.
evidence-based guideline for the assessment and management of polycystic ovary syn- 60. Specker P, Liddell BJ, Byrow Y, Bryant RA, Nickerson A. A factor analytic investigation
drome. Clin Endocrinol. 2018;89:251-268. of DSM-5 PTSD symptoms in a culturally diverse sample of refugees resettled in
40. Goldstein RF, Abell SK, Ranasinha S, et al. Association of gestational weight gain with Australia. Confl Health. 2018;12:26.
maternal and infant outcomes: a systematic review and meta-analysis. JAMA. 2017;317: 61. Graham HR, Minhas RS, Paxton G. Learning problems in children of refugee back-
2207-2225. ground: a systematic review. Pediatrics. 2016;137.

714 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry
Volume 59 / Number 6 / June 2020