Mental health of displaced and refugee children resettled in high income countries: Risk and protective factors

Abstract

We undertook a systematic search and review of individual, family, community, and societal risk and protective factors for mental health in children and adolescents who are forcibly displaced to high-income countries. Exposure to violence has been shown to be a key risk factor, whereas stable settlement and social support in the host country have a positive effect on the child's psychological functioning. Further research is needed to identify the relevant processes, contexts, and interplay between the many predictor variables hitherto identified as affecting mental health vulnerability and resilience. Research designs are needed that enable longitudinal investigation of individual, community, and societal contexts, rather than designs restricted to investigation of the associations between adverse exposures and psychological symptoms. We emphasise the need to develop comprehensive policies to ensure a rapid resolution of asylum claims and the effective integration of internally displaced and refugee children.

Full text

Review
266
www.thelancet.com Vol 379 January 21, 2012
Introduction
Children and adolescents who fl ee persecution and
resettle in high-income countries often endure great
physical and mental challenges during displacement,
and suff er continuing hardships after arrival. Most of
these refugees come from geographically distant, low-
income settings.1 The adverse events that necessitated
their fl ight are often only the beginning of a long period
of turbulence and uncertainty. Young people might travel
for weeks or months in dangerous circumstances to seek
asylum in a high-income country, and are sometimes
temporarily or permanently separated from family and
need to use professional traffi ckers to reach their
destination.2 The challenges typically encountered after
arrival include, fi rst, the complex legal immigration
processes that asylum seekers must negotiate to gain
refugee status or be repatriated,3 and second, the huge
social, cultural, and linguistic diff erences between the
place of origin and the new setting.
The process of sociocultural adaptation can be quite
gradual, and refugees integrate to diff erent extents with
the host community.4 Children with disrupted or
minimal school education are suddenly immersed in a
new edu cation system. Racial discrimination and
bullying, exacer bated by policies to accommodate
asylum seekers in already impoverished and dis-
advantaged areas, are widespread.5 Immigration policies
for dispersal and deten tion can negatively aff ect
refugees’ attempts to settle in their host community.6
However, rapid resolution of asylum decisions eases
access to social, health, education, and employment
opportunities and infra structures. Refugee children in
high-income countries do not usually lack basic material
necessities, yet certain factors nonetheless place their
healthy development at risk. In this Review, we draw
attention to the specifi c risk and protective factors that
aff ect the psychological wellbeing of refugee children.
Table 1 summarises all the studies included in this
Review. Table 2 summarises the main fi ndings
according to individual factors, and table 3 according to
family, community, and societal factors.
Mental health of displaced and refugee children resettled in
high-income countries: risk and protective factors
Mina Fazel, Ruth V Reed, Catherine Panter-Brick, Alan Stein
We undertook a systematic search and review of individual, family, community, and societal risk and protective factors
for mental health in children and adolescents who are forcibly displaced to high-income countries. Exposure to violence
has been shown to be a key risk factor, whereas stable settlement and social support in the host country have a positive
eff ect on the child’s psychological functioning. Further research is needed to identify the relevant processes, contexts,
and interplay between the many predictor variables hitherto identifi ed as aff ecting mental health vulnerability and
resilience. Research designs are needed that enable longitudinal investigation of individual, community, and societal
contexts, rather than designs restricted to investigation of the associations between adverse exposures and psychological
symptoms. We emphasise the need to develop comprehensive policies to ensure a rapid resolution of asylum claims
and the eff ective integration of internally displaced and refugee children.
Lancet 2012; 379: 266–82
Published Online
August 10, 2011
DOI:10.1016/S0140-
6736(11)60051-2
Oxford University,
Oxford, UK (M Fazel DM,
Prof A Stein FRCPsych); Oxford
Health NHS Foundation
Trust, Oxford, UK
Search strategy and selection criteria
The Medline, Scopus, PsycINFO, Embase, Web of Science citation, and Cochrane
databases were systematically searched for studies about risk and protective factors
that were reported from January, 1980, to July, 2010. Searches of similar terms were
combined such as “asylum seeker”, “refugee”, “displaced person”, “migrant” with
“child”, “adolescent”, “young”, “minor”, “youth” or “teenage”, and terms including
“psychiatr*”, “psycholog*”, “psychosocial”, “mental”, “resilience”, “outcome”,
“development”, “protective factor”, “adaptation”, “modifying factor”, “vulnerability
factor”, “risk factor”, “recovery”, “wellbeing”, “emotion”, “behaviour”, “behavior”,
“trauma”, “traumatic”, and “adjustment”. We also searched for specifi c countries of
origin. Adaptations to the terms and MeSH searching were implemented, depending
on the search style of each database. Additionally, reference and citation lists in
published works, grey literature, and the authors’ databases were reviewed. Inclusion
criteria included study population, publication date, data about risk or protective
factors, and sample size. There were no language restrictions.
We included studies of risk and protective factors for psychological, emotional, or
behavioural disorders with a minimum sample size of 50 participants, and studies with
25 participants or more if a predictor variable was assessed for which there was minimal
evidence from larger studies. Studies with participants aged up to and including the age
of 18 years were eligible for inclusion; those with wider age categories were only included
if all participants were younger than 25 years and mean age was 18 years or younger. We
contacted investigators who had undertaken more than one study to clarify whether
samples overlapped. Countries were defi ned by income in accordance with the World
Bank classifi cation.
5296 potentially relevant reports were identifi ed through database searches, of which
1581 were duplicates. 737 summaries were reviewed and 257 full-text papers were
obtained. Our fi nal sample consisted of 44 studies from high-income countries, with
5776 displaced children and adolescents (nine studies had overlapping samples). They
included forcibly displaced children from Bosnia, Cambodia, Central America, Chile,
Croatia, Cuba, Iraq, Middle East, Somalia, Sudan, Vietnam, and the former Yugoslavia, who
were either internally displaced or resettled in Australia, Belgium, Canada, Croatia,
Denmark, Finland, the Netherlands, Sweden, the UK, and the USA. Mental health
outcomes measured in these studies were generally grouped as internalising or emotional
problems, including depression, anxiety, and post-traumatic stress disorder; and
externalising or behavioural problems. We adhered to the terms used in each study
describing the mental health outcomes and groups of displaced or refugee children.
A meta-analysis was not done because of clinical and methodological heterogeneity.

Review
www.thelancet.com Vol 379 January 21, 2012
267
(R V Reed MRCPsych); and Yale
University, New Haven, CT,
USA (Prof C Panter-Brick, DPhil)
Correspondence to:
Dr Mina Fazel, Department of
Psychiatry, Oxford University,
Oxford OX3 7JX, UK
mina.fazel@psych.ox.ac.uk
Eff ects of displacement
Although there are a reasonable number of reports
about children exposed to confl ict,51 the importance of
displacement, as an additional variable to exposure to
organised violence, has only been assessed in four
Croatian studies.23,29,31,50 In a study31 with a 30-month
follow-up, post-traumatic stress disorder, depression,
and somatic complaints decreased with time in
internally displaced and non-displaced children, but
psychosocial adaptation remained worse in displaced
children and did not improve with time. In another
study,29 compari son of Bosnian refugee children and
displaced and non-displaced Croatian children showed
that the refugee children had higher anxiety and had
Study site Study population Number Age* (years) Domain assessed Measurements
Ajdukovic et al,7 1993 Croatia Internally displaced
children and their mothers
319 Up to 18 Family Semistructured interviews, authors’
own exposure and stress (checklist of
emotional, behavioural, and
psychosomatic symptoms) scales
Almqvist et al,8 1997 Sweden Iranian refugee children 50 4–8 Individual and family Semistructured interview and
observation of child’s play
Almqvist et al,9 1999 Sweden Iranian refugee children 39 6–10 (at follow-up) Individual, family, and
community
Semistructured interview, Social
Adjustment Index, Global Self-Worth
Angel et al,10 2001 Sweden Bosnian refugee children 99 6–16 Individual and family Clinical interviews, observation of
child, short Cederblad questionnaire
Bean et al,11 2007 Netherlands
and Belgium
Unaccompanied refugee
children in the
Netherlands; Dutch
normative sample;
immigrant and refugee
group in Belgium
920 UASC (and
1059 Dutch
children, and
1294 migrants and
refugees in
Belgium)
12–18 Individual and family HSCL-37A, SLE, RATS
Bean et al,12 2007 Netherlands Unaccompanied refugee
children from
48 countries
582 12–18 Individual, family, community,
and society
HSCL-37A, SLE, RATS, CBCL for age
4–18 years (guardian report), TRF for
4–18 years
Berthold et al,13 1999 USA Khmer refugee
adolescents (born in
Cambodia or in refugee
camps or in Vietnam)
76 11–19 (mean 16) Individual SCECV, LA PTSD Index, CIS
Berthold et al,14 2000 USA Khmer refugee
adolescents
144 14–20 (mean 16) Individual, family, and
community
HTQ part 1, modifi ed SCECV, LA
PTSD index, CES-DC, Personal Risk
Behaviour scale, Perceived Social
Support from Family and Friends,
Orthogonal Cultural Identifi cation
Scale
Cohn et al,15 1985 Denmark Chilean children whose
parents had been
tortured
85 (58 born
in Chile)
Not stated Family Clinical interview
Daud et al,16 2008 Sweden Second-generation Iraqi
children whose parents
had been tortured versus
North African children
whose parents had not
been tortured
80 7–16 Family WISC-III, DICA-R, PTSS, I Think I Am
scale, SDQ
Derluyn et al,17 2007 Belgium Unaccompanied refugee
children
166 9–18 Individual HSCL-37A, SDQ-self, RATS, SLE, social
workers completed CBCL for age
6-18 years and SDQ-parent
Derluyn et al,18 2009 Belgium Newly arrived adolescents;
10% were UASC
124 UASC 11–18 Family HSCL-37A, SLE, RATS
Ekblad et al,19 1993 Sweden Refugee children and their
mothers from former
Yugoslavia residing in a
refugee camp
66 5–15 Individual, family, community,
and society
Structured interview
Ellis et al,20 2008 USA Somali adolescent
refugees
135 11–20 (mean 15) Individual, family, community,
and society
UCLA-PTSD RI, WTSS, DSRS, Everyday
Discrimination Scale, Adolescent Post
War Adversities Scale, Acculturative
Hassles Inventory
Fox et al,21 1999 USA Vietnamese and
Cambodian adolescents
47 9–15 (mean 11) Individual and society Structured interview, authors’ own
assessment of emotional eff ect of
violence; CDI
(Continues on next page)

Review
268
www.thelancet.com Vol 379 January 21, 2012
Study site Study population Number Age* (years) Domain assessed Measurements
(Continued from previous page)
Geltman et al,22 2005 USA Sudanese unaccompanied
adolescents
304 Mean 18 (range not
stated)
Individual, family, community,
and society
HTQ-CHQ
Grgic et al,23 2005 Croatia Croatian camps for
internally displaced
people
112 12–15 (mean 14) Individual and family CDI, WTQ, Index of Family Relations
Grgic et al,24 2005 Croatia Returned previously
displaced adolescents
versus never displaced
57 16–18 (mean 17) Displacement CDI, HSC, C-PTSDI
Hjern et al,25 1991 Sweden Chilean refugees 50 2–15 (mean 6) Individual Adapted Cederblad questionnaire
Hjern et al,26 1998 Sweden Chilean and Middle
Eastern refugee children
18 months after arrival
63 2–15 (mean 6) Individual, family, community,
and society
Authors’ own questionnaires for
organised violence, family stress, social
situation in exile, school and nursery
teacher questionnaire, and Cederblad
questionnaire
Hodes et al,27 2008 UK 78 unaccompanied and
35 accompanied refugee
adolescents from various
countries
113 13–18 Individual, family, and society HTQ, IES, BDSR
Kia-Keating et al,28 2007 USA Somali adolescent
refugees
76 12–19 (mean 16) Individual and community WTSS, PSSM, UCLA-PTSD RI, DSRS,
Multidimensional Scales of Perceived
Self-Effi cacy
Kocijan-Hercigonja et al,29
1998
Croatia Non-displaced and
displaced Croatian
children, and refugee
children from Bosnia
35 displaced
and 35 refugees
(and
35 non-displaced)
6–14 Displacement SCSI, questionnaires on psychosomatic
symptoms, psychosocial adjustment,
anxiety and depression
Kovacev et al,30 2004 Australia Adolescents from former
Yugoslavia
83 12–19 (mean 15) Family and community Social Support Scale for Children,
Global Self Worth and Peer Social
Acceptance scales, Acculturation
Attitudes Scale
Kuterovac JagodiĆ et al,31
2000
Croatia Displaced and
non-displaced children
resident in the same area
93 displaced (and
161 non-displaced)
Mean 13 at follow-
up
Displacement Locally developed questionnaires for
war experiences, psychosocial
adaptation, PTSD, depression and
somatic symptoms
Liebkind et al,32 1993 Finland Vietnamese refugee
adolescents and their
parents or carers
159 14–24 (mean 18) Individual, community, and
society
HSCL-25, Vietnamese Depression Scale
Liebkind et al,33 1996 Finland Vietnamese refugee
adolescents
159 14–24 (mean 18) Individual and community Vietnamese Depression Scale, HSCL-25,
RCRG
Montgomery et al,34 2006 Denmark Middle Eastern
refugee children
311 3–15 (mean 8) Individual and family Structured parental interview for
exposures to violence and current
mental state
Montgomery et al,35 2008 Denmark Middle Eastern refugee
adolescents
131 11–23 (mean 15 at
follow-up)
Individual, family, community,
and society
Structured parental interview,
semistructured interview, YSR, YASR
Montgomery et al,36 2010 Denmark Middle Eastern refugee
adolescents
131 11–23 Individual and family Structured interviews with young
person and parents: YSR or YASR
depending on age at follow-up
Nielsen et al,37 2008 Denmark Accompanied refugee
children in Danish Red
Cross asylum centres—
mixed country of origin
246 4–16 Community and society Teacher SDQ, and self-report SDQ for
children aged 11–16 years
Porte et al,38 1987 USA Indochinese refugee
adolescents;
58 unaccompanied, in
various types of alternative
care; 24 accompanied
82 12–19 (mean 16) Family CES-DC, authors’ own method of
assessment of acculturation and
support systems
Reijneveld et al,39 2005 Netherlands 6 9 UASC in a restrictive
reception centre;
53 UASC in a routine
reception centre
122 14–18 Individual and society HSCL-25, RATS
Rothe et al,40 2002 USA Cuban refugee children in
a refugee camp
87 6–17 (mean 15) Individual PTSD RI, TRF
(Continues on next page)

Review
www.thelancet.com Vol 379 January 21, 2012
269
fewer eff ective coping strategies to manage stressful
situations than did the displaced and non-displaced
children. In a further study,50 the prevalence of
depression in internally displaced Croatian children
was not diff erent from that in non-displaced children.
Prevalence of hopelessness, post-traumatic stress
disorder, or depression was not diff erent in a comparison
of internally displaced Croatian teenagers (aged
16–18 years) who had returned after 7 years in exile with
those who were displaced for 6 months or less.23 Because
the results of these studies have not been conclusive,
further larger studies are needed to provide useful
insights into whether displacement places children at
additional risk of poor mental health, and whether they
have diff erent needs from local children exposed
to confl ict.
Individual factors
Exposure to violence
Direct experience of adverse events is associated with an
increased likelihood of psychological disturbance in
refugee children. The degree of post-traumatic stress
disorder was associated with personal experiences of
traumatic events, especially those occurring when away
from home.22 Internalising diffi culties in the initial phase
after displace ment were associated with adverse events
before migration,35 whereas the rates of sleep disturbances
and anxiety were increased in children with direct
Study site Study population Number Age* (years) Domain assessed Measurements
(Continued from previous page)
Rousseau et al,41 1998 Canada 100 southeast Asian, and
56 Central American
refugee children
156 Mean 10 (range not
stated)
Family, community, and
society
CBCL, a trauma scale, a separation
index, FES, SRDS for parental
depression, a social network score
Rousseau et al,42 1999 Canada Cambodian adolescent
refugees
67 Mean 14 (range not
stated)
Family YSR, AFI, measurements of school
performance, authors’ index of trauma
exposure
Rousseau et al,43 2000 Canada 76 Cambodian and
82 Central American
refugee adolescents
158 14–15 Individual and family CBCL, YSR
Rousseau et al,44 2003 Canada Cambodian refugee
adolescents
57 Mean 14 (range not
stated)
Family YSR, a measurement of risk behaviour,
AFI, SES, CSES, Racism Experience Scale
Rousseau et al,45 2004 Canada Cambodian refugee
adolescents
67 Mean 14, 16 at
follow-up
Individual, family, and
community
YSR, CBCL, FES, BAS
Slodnjak et al,46 2002 Slovenia Bosnian refugee
adolescents and local
Slovenian adolescents
265 refugees (and
195 local
adolescents)
14–15 Individual CDI, IES, adapted WTQ, local Teacher’s
Report
Sourander et al,47 1998 Finland UASC in an asylum
reception centre awaiting
placement
46 6–17 Individual CBCL
Sujoldzic et al,48 2006 Bosnia,
Croatia, and
Austria
Adolescents from Bosnia
resettled in three
contexts: internally
displaced in Bosnia or
living as refugees in
Austria or Croatia
499 refugee
Bosnians (and
comparator groups
of 359 internally
displaced and
424 non-displaced
Bosnians)
15–18 (mean 17) Individual, family, community,
and society
Perceived health problems, objective
health problems, index of
psychological distress, SES, FAS, scale
of risk and protective environmental
factors, religious commitment scale
Tousignant et al,49 1999 Quebec Refugees from various
countries
203 13–19 Individual, family, community,
and society
Diagnostic Interview Scale for
Children (version 2.25) and Children’s
Global Assessment Scale
Zivcic et al,50 1993 Croatia Displaced and non-
displaced Croatian
children in Rijeka during
wartime
160 refugees (and
320 non-displaced)
8–15 (mean 11) Displacement CDI, authors’ own emotion scale
Nine studies were either follow-up or used the same children in more than one study, and so were not recounted.9,12,25,33,35,36,42,44,45 UASC=unaccompanied asylum-seeking children. HSCL-37A=Hopkins Symptom
Checklist-37 for Adolescents. SLE=Stressful Life Events checklist. RATS=Reactions of Adolescents to Traumatic Stress. CBCL=Child Behaviour Checklist. TRF=Teacher’s Report Form (CBCL). FAS=Family Affl uence
Scale. SCECV=Survey of Children’s Exposure to Community Violence. LA PTSD=Los Angeles post-traumatic stress disorder. CIS=Columbia Impairment Scale. Cederblad questionnaire=Cederblad questionnaire for
children’s mental health. HTQ=Harvard Trauma Questionnaire. IES=Impact of Event Scale. CES-DC=Center for Epidemiologic Studies Depression Scale for Children. WISC=Wechsler Intelligence Scale for Children.
DICA-R=Diagnostic Interview for Children and Adolescents. PTSS=Post-Traumatic Symptom Scale. SDQ=Strengths and Diffi culties Questionnaire. UCLA-PTSD RI=University of California at Los Angeles
Post-traumatic Stress Disorder Reaction Index. WTSS=War Trauma Screening Scale. DSRS=Depression Self-Rating Scale. CDI=Children’s Depression Inventory. HTQ=Harvard Trauma Questionnaire.
CHQ=Child Health Questionnaire. WTQ=War Trauma Questionnaire. HSC=Hopelessness Scale for Children. C-PTSDI=Children’s Post-Traumatic Stress Disorder Inventory. BDSR=Birleson Depression Self-Rating
Scale. PSSM=Psychological Sense of School Membership. PTSD=post-traumatic stress disorder. SCSI=Schoolagers’ Coping Strategy Inventory. HSCL-25=Hopkins Symptom Checklist-25. RCRG=Role Construct
Repertory Grid (interview and questionnaire on acculturation). YSR=Youth Self Report (CBCL). YASR=Young Adult Self Report. PTSD RI=Post-Traumatic Stress Disorder Reaction Index. FES=Family Environment
Scale. SRDS=Self-Report of Depressive Symptoms. AFI=Adolescent Friendship Inventory. SES=Rosenberg’s Self-Esteem Scale. CSES=Collective Self-Esteem Scale. BAS=Behavioural Acculturation Scale. *Mean ages
have been rounded to the nearest whole number.
Table 1: Summary of studies

Review
270
www.thelancet.com Vol 379 January 21, 2012
exposure to adverse events before migration.25 The degree
of perceived personal threat during traumatic exposures
was a determinant of generalised anxiety10 and
post-traumatic stress disorder.40
Additionally, migration journeys and postmigration
experiences might be highly distressing. Thus, Cuban
children who witnessed violence while they were detained
in a refugee camp en route to the USA showed more
withdrawn behaviour than did children without exposure
to violence in the camp.40 Direct and indirect exposure to
violence or other potentially traumatic events, after entry
into the host country, was associated with a range of
negative psychological outcomes in most13,35,36,48 but not
all21 studies.
Cumulative exposure to traumatic events is associated
with a broad range of psychological problems in refugee
groups exposed to violence during war.8,17,20,24,46 However, in
some studies, the number of traumatic events before
migration was not a predictor of post-traumatic stress
disorder.13,14 The results of two studies have indicated that
the number of lifetime traumatic events could be more
consequential than are predisplacement events,13,36 empha-
sising the importance of considering the refugee’s whole
experience so far rather than just the premigration events.
However, in one longitudinal study, a high number of
adverse events before displacement continued to aff ect the
mental health of refugees even 9 years after arrival,
but those who subsequently recovered from initial
symptoms were likely to have suff ered fewer additional
adverse events after displacement than had those who
remained symptomatic.36
Physical, psychological, or developmental disorders
In a longitudinal study in Sweden,9 pre-existing vulner-
ability (consisting of delayed development, long-term
physical illness, or psychological problems) was a
predictor of mental ill-health, poor social adjustment,
and low self-worth 3·5 years after arrival, whereas the
absence of evidence of such vulnerability before exposure
to adverse events was a strong predictor of emotional
wellbeing. Personal injury that was sustained during
potentially traumatic pre migration events was associated
with an increased risk of post-traumatic stress disorder.
Head injury, in particular, was associated with a doubling
of risk.22 These potentially important factors were not
investigated in other studies.
Age and sex
The relation between age and psychological symptoms
is not clear from existing evidence because of the
diffi culty in diff erentiation of potential confounding
factors, including age at the onset of adverse events,
age at migration, and age-related policies for education,
accommodation, and the decision-making processes
for asylum in host countries. These variables intersect
with the nature and duration of adverse exposure,
aff ecting age-specifi c responses. For example, children
in their late teens confronted with a short period of
exposure to violent confl ict are likely to have benefi ted
from a long period of stable psychosocial development,
whereas children growing up in situations of long-
standing confl ict are likely to have had greater
cumulative adversity. This greater adversity might
increase the likelihood of psychological diffi culties in
these children, or conversely, strengthen their capacities
for resilience.
In a UK cross-sectional survey of unaccompanied
asylum-seeking children (UASC), increase in age was
associated with an increase in symptoms of post-
traumatic stress disorder, whereas accompanied children
had fewer problems with increasing age.27 This diff erence
might indicate a diff erence in immigration status—
ie, unaccompanied adolescents feared possible depor-
tation after the age of 18 years, whereas most accompanied
adolescents had been granted the right to remain in the
UK as refugees. Similar fi ndings have been reported
in other studies of UASC in Belgium and the
Netherlands.11,12,17 The living arrangements of UASC
might have a negative eff ect particularly on children
younger than 15 years, as reported in one study of asylum
centres.47 No independent relation was noted between
age and psychological disorders in other studies.10,13,20,26,49
Relations between sex and psychological functioning
also show much variation. In about half the studies of
accompanied and unaccompanied children, the
prevalence of mental health disorders, notably depres-
sion and internalising diffi culties, was higher in girls
than in boys.13,17,27,32,39 These disorders sometimes occur-
red with other diagnoses,48,49 including post-traumatic
stress disorder.12,46 No sex-related diff erences were noted
in the remainder of the studies.10,13,20,26,28,34,40,43 Changes
in sex-related eff ects with time were inconsis-
tent.13,14,20,26,28,34,40,43,45 A protective eff ect of male sex for
internalising disorders is consistent with fi ndings from
non-refugee populations52 and Reed and colleagues’
Review53 about low-income and middle-income
countries, but biological and social causal pathways
need to be assessed further.
Education
The period of formal education before displacement
was unrelated to psychological distress or behavioural
problems.12 Bosnian adolescent refugees to Slovenia
with high educational achievement were more likely to
have post-traumatic stress disorder than were those with
low achievement.46 Although the reasons for this
diff erence are not clear, evidence suggests that good
overall functioning in refugee children can coexist with
mental health symptoms.54 In a longitudinal study,
refugees whose mental health improved at follow-up
after 8–9 years were more likely to be in education or
employment than were those who remained sympto-
matic, but whether education or employment were
contributing to recovery or whether children with

Review
www.thelancet.com Vol 379 January 21, 2012
271
persistent symptoms tended to withdraw from such
situations is not clear.36
Family factors
Exposure to violence
Familial experiences of adverse events aff ect children’s
psychological functioning. Some types of parental
exposures are more strongly associated with children’s
mental health problems than are children’s own
exposures,34,44 particularly if parents have been tor-
tured9,15,16,34,55 or are missing.34 Familial adverse events
before the child’s birth were a major determinant of
children’s later psychological outcomes in Central
American, but not in southeast Asian56 or Middle Eastern
refugees.34 Family communi cation might be relevant—
awareness of a parent’s detention was an independent
predictor of post-traumatic stress disorder in the child34
and a lack of discussion by the family about adverse events
was protective with respect to the child’s mental health.10
These fi ndings need to be replicated because of the small
sample sizes, to compare consistency with other reports
about family communication, and especially because of
the substantial evidence from other situations that key
family processes play an important part in helping family
members to recover in times of crisis.36,57 Further research
is needed to identify which kinds of communication are
helpful or not in the diff erent contexts.
Family composition and bereavement
Being unaccompanied on entry to the host country puts a
child at risk of psychological disorders,11,27,58 although the
experiences of UASC and accompanied children are
heterogeneous. UASC often experienced higher numbers
of adverse events than did accompanied children.18,27
Separation from the immediate family was associated
with post-traumatic stress disorder in one study,22 though
Summary
Exposure to violence
Almqvist et al,8 1997 Post-traumatic stress disorder frequency was linked to degree of adverse event exposure (38%, if severe exposure; 11%, if some exposure)
Angel et al,10 2001 Degree of perceived personal threat during adverse event was predictive of anxiety symptoms subsequently
Berthold et al,13 1999 Lifetime and postmigration violence correlated with symptoms of post-traumatic stress disorder, but premigration violence did not
Berthold et al,14 2000 Although degree of violence exposure was not predictive of diagnoses of depression and post-traumatic stress disorder, high rates of symptoms were associated
with increased exposure
Derluyn et al,17 2007 Number of traumatic experiences were predictive of symptoms of anxiety, depression, post-traumatic stress, and emotional problems
Ekblad et al,19 1993 Experience of direct violence was associated with poor mental health of child
Ellis et al,20 2008 Exposure to traumatic events was most strongly associated with post-traumatic stress disorder, and was also associated with depression
Fox et al,21 1999 Violence before migration was associated with depression subsequently. Frequency of violence while in a refugee camp was not associated with depression
subsequently
Geltman et al,22 2005 Direct personal trauma was associated with post-traumatic stress disorder, but witnessing assaults on other people was not
Post-traumatic stress disorder was doubled in children who suff ered head trauma. Children in their own village at the time of an adverse event had a lower risk of
post-traumatic stress disorder than did those who experienced adverse events when away from their village
Grgic et al,24 2005 A relation was noted between the number of war traumas and Children’s Depression Inventory score
Hjern et al,25 1991 Sleep disturbances and separation anxiety were signifi cantly associated with direct experiences of persecution
Hjern et al,26 1998 Witnessing violence was a signifi cant predictor of symptom scores at follow-up after 17–19 months
Montgomery et al,34 2006 Witnessing violent acts and direct exposure to organised violence were predictive of various psychological symptoms, but not the full symptom complex of
post-traumatic stress disorder
Montgomery et al,35 2008 High numbers of diff erent premigration traumatic experiences were predictive of high internalising scores
Witnessing attacks on other people after arrival in Denmark was associated with an increase in externalising behaviour
Montgomery et al,36 2010 Traumatic experiences before arrival were signifi cantly more common in those who were symptomatic at arrival and at follow-up than in those who never attained
symptom thresholds. Numbers of types of stressful events after arrival in the host country were much lower in children who recovered from symptoms during
follow-up than in those who remained symptomatic
Rothe et al,40 2002 Degree of perceived personal threat during an event was predictive of withdrawn behaviour later
Witnessing violence in a refugee camp was predictive of withdrawn behaviour, but time spent in a camp and separation from family in the camp and witnessing
suicide attempts were not signifi cant
Fear of dying during migration while at sea was associated with withdrawn behaviour but not with post-traumatic stress disorder
Slodnjak et al,46 2002 Small correlation between Children’s Depression Inventory and trauma exposure
Sujoldzic et al,48 2006 Violence from peers and adults was associated with poor psychological functioning in Bosnian children who were resettled in Croatia
Physical, psychological, or developmental disorders
Almqvist et al,8 1999 Absence of reported signs of vulnerability (poor physical, emotional wellbeing, or delayed development) before exposure to violence strongly determined
emotional wellbeing at follow-up
Geltman et al,22 2005 Personal injury during premigration was associated with an increased risk of post-traumatic stress disorder; head injury, particularly, was associated with a doubled
risk of post-traumatic stress disorder
Time since displacement
Geltman et al,22 2005 Residence in resettlement country for less than 6 months was not associated with post-traumatic stress disorder
(Continues on next page)

Review
272
www.thelancet.com Vol 379 January 21, 2012
arriving with a family member was not protective in
another study.20 UASC who had at least one family
member already resident in the host country had lower
scores for internalising diffi culties and post-traumatic
stress symptoms.12
Accompanied children subsequently separated from
their relatives were also at risk of poor mental health.26
Children whose relatives were in diffi cult circumstances
(eg, imprisoned), and those who had diffi culty contacting
their relatives had worse psychological functioning.7 Boys
living with both parents had rates of psycho logical
symptoms fi ve times lower than those living in other
family arrangements,49 and fewer changes of family
structure were protective for boys. From interviews, many
single mothers had diffi culty asserting their authority over
adolescent boys, and their authority could be undermined
by the boys’ peers. Confl ict with a mother’s new partner
was also postulated as a contributory factor. Adolescents
Summary
(Continued from previous page)
Age
Angel et al,10 2001 No association was noted between age and scores on symptom scales
Bean et al,11 2007 Psychological distress was higher with increasing age in unaccompanied asylum-seeking children
Bean et al,12 2007 Older age was associated with higher internalising, externalising, and Reactions of Adolescents to Traumatic Stress scores at follow-up
Berthold et al,13 1999 No diff erence was noted in rates of post-traumatic stress disorder or functioning
Berthold et al,14 2000 No association of age with post-traumatic stress disorder or depression was noted
Derluyn et al,17 2007 Adolescents aged 17–18 years had higher depression scores; conduct problems peaked at age 16 years
Ellis et al,20 2008 Age was not associated with post-traumatic stress disorder or depression
Hjern et al,26 1998 Age was not an independent predictor of symptom scores in regression analyses
Hodes et al,27 2008 For unaccompanied asylum-seeking children, Impact of Event Scale scores increased with age, and decreased with age for accompanied children
Montgomery et al,35 2008 Fewer externalising behavioural problems were noted with increasing age
Rothe et al,40 2002 Children had higher scores of post-traumatic stress disorder; Child Behaviour Checklist scores were not aff ected by age
Sourander et al,47 1998 Children younger than 15 years had more psychological distress and behavioural problems than did older children
Tousignant et al,49 1999 No association of age with the prevalence of psychopathology was noted
Sex
Angel et al,10 2001 No association between sex and scores on symptom scales was noted
Bean et al,12 2007 Girls had higher Reactions of Adolescents to Traumatic Stress and internalising scores, but not externalising scores
Berthold et al,13 1999 No diff erence was noted in rates of post-traumatic stress disorder or functioning
Berthold et al,14 2000 Girls were more likely to be depressed than were boys
Derluyn et al,17 2007 Girls had higher scores for anxiety, depression, withdrawal, intrusion on the Reactions of Adolescents to Traumatic Stress, and total problems
Ellis et al,20 2008 Sex was not associated with post-traumatic stress disorder or depression
Hjern et al,26 1998 Girls had an increase in symptoms soon after arrival in the host country, but no diff erence was noted between boys and girls at 18 months after arrival
Hodes et al,27 2008 Girls had higher scores on the Impact of Event Scale and Birleson Depression Self-Rating Scale than did boys
Kia-Keating et al,28 2007 Sex was not a predictor of depression or post-traumatic stress disorder
Liebkind et al,32 1993 Girls had higher scores than did boys
Liebkind et al,33 1996 Premigration traumatic experiences were predictive of boys’ Hopkins Symptom Checklist-25 scores
Montgomery et al,34 2006 Sex was not a predictor of post-traumatic stress disorder
Montgomery et al,35 2008 Male sex was predictive of fewer internalising behaviours at follow-up
Reijneveld et al,39 2005 Girls had higher internalising symptoms than did boys
Girls in a restrictive reception facility had higher anxiety, depression, and emotional problems than did girls in a less restrictive reception setting, whereas boys in a
restrictive reception setting had similar rates to those in a less restrictive reception facility
Rothe et al,40 2002 Sex was not associated with scores for Child Behaviour Checklist or post-traumatic stress
Rousseau et al,43 2000 Sex was not associated with internalising or externalising symptoms
Rousseau et al,45 2004 No sex diff erence was noted in early adolescence (mean age 14 years); but boys had more internalising symptoms at follow-up in mid-adolescence (16 years)
Slodnjak et al,46 2002 Girls had higher frequencies of post-traumatic stress disorder and depression than did boys
Sujoldzic et al,48 2006 Girls had worse functioning than did boys
Tousignant et al,49 1999 All disorders except conduct disorder were more prevalent in girls than in boys
Education
Bean et al,12 2007 Period of child’s education before migration had no eff ect on scores for behavioural problems or psychological distress
Montgomery et al,36 2010 Refugees whose mental health improved by 8–9-year follow-up were more likely to be in education or employment than were those who remained symptomatic
Slodnjak et al,46 2002 Adolescents with better school achievement had more symptoms of post-traumatic stress disorder than did those with lower school achievement
Table 2: Summary of principal fi ndings in relation to individual factors assessed in each study

Review
www.thelancet.com Vol 379 January 21, 2012
273
living with both parents had lower internalising scores in
mid-adolescence, whereas those in single-parent house-
holds reported greater feelings of competence.45
Family functioning and parental health
Family cohesion and perception of high parental support
were associated with fewer psychological diffi culties in
children than were poor family support or cohesion.13,23,30,45,48
The results of one study showed that children whose
parents divorced after displacement showed more
psychological symptoms than did those whose parents
did not divorce,26 but this fi nding was not supported in
another study.9 Displaced Croatian children whose
mothers reported adaptation diffi culties or confl icts had
high levels of psychological distress.7 Evidence about the
eff ect of postmigration changes in parental behaviour is
insuffi cient and equivocal.7,34 In particular, the role of
communication within families is potentially important,
but there has been little investigation thereof; a lack of
information about experiences of migration or exile was
associated with an increased likelihood of psychological
disorders19 or had no eff ect.34 Situations in which children
spoke often with mothers about diffi culties were
associated with an absence of substantial psychological
diffi culties at arrival in Denmark and at follow-up
8–9 years later.36 Good parental mental health, particularly
in mothers, is an important protective factor,19,21,26 in
keeping with the evidence from non-refugee populations.59
Refugee parents with poor mental health were noted in
one study to direct feelings of anger towards their
children.60 Among Bosnian refugee families in Sweden,
shared stressful experiences accounted for the correlation
between the parent and child’s mental health.10
Household socioeconomic circumstances
Socioeconomic status might provide some protection, in
that access to material and social resources could allow an
early fl ight from confl ict and reduction in cumulative
exposure to adversity.5,61 However, fi ndings of studies are
contradictory, with socioeconomic circumstances before
displacement not being predictive of post-traumatic stress
disorder;34,41 or parental employment status, particularly
employment in a private enterprise rather than admin-
istrative or manual work, being associated with post-
traumatic stress disorder in the child,34 whereas being like
others in the middle stratum of society might be protective
(Montgomery E, Rehabilitation and Research Centre for
Torture Victims, Denmark, personal communication).
Attention to mental health in the context of everyday
stressors, including socioeconomic disadvantage, needs
greater attention.62 Economic circumstances after dis-
place ment can aff ect the child’s psychological function-
ing. Parental worries about fi nancial problems have a
particular adverse eff ect on the mental health of refugee
children.7,63 Thus, low socioeconomic status of Bosnian
refugee adolescents was linked to more depressive
symptoms and poor self-esteem.48 By contrast, clear
correlations were not noted between psychological
disorders and markers of socioeconomic status in two
reports.26,49 Paternal unemployment for longer than
6 months during the fi rst year of resettlement was
predictive in one study,49 but not in another study43 of
psychological disorders in children. Conclusions cannot
yet be drawn about the eff ect of socioeconomic status,
and causal pathways are diffi cult to interpret because
most of the studies have heterogeneous designs in terms
of the inclusion of predictor variables.
Parental education
Education of the parents has a variable eff ect—having
educated parents might be protective35,36,41 or have no
eff ect.34,49 In a longitudinal study,36 refugees whose
symptoms resolved during follow-up had fathers with
long periods of education in the home country, perhaps
indicating resources within the family that fostered
resilience. Educated, intellectual families might be
targeted in some political confl icts,64 which could outweigh
any protective eff ect of parental education.
Community factors
Social support and community integration
Perceptions of acceptance or discrimination within host
countries are highly relevant. In a study of displaced
Bosnian adolescents,48 those internally displaced or
displaced to Croatia reported more perceived discrimination
than did those who had resettled in Austria. Low peer
violence and discrimination were positively linked to self-
esteem. Boys were more likely to report discrimination
than were girls,35,48 and this diff erence was predictive of
poorer psychological functioning. Perceived discrimination
was the best predictor of outcomes such as depression and
post-traumatic stress disorder in Somali adolescents in the
USA,20 but had no eff ect in another study.33 41% of refugee
children in a Swedish study9 reported bullying and those
with few peers to play with were likely to show poor general
adaptation.9,35 High perceived peer support was associated
with improved psychological functioning.14,30 Subjective
childhood experiences, including the strength of peer
relationships, are integral to healthy psychological
development; however, longitudinal data in the refugee
context are lacking, and actual eff ect sizes are often left
unspecifi ed in reports.
A perceived sense of safety at school has been associated
with low risk of post-traumatic stress disorder,22 and an
increased sense of school belonging was shown to protect
against depression28,45,48 and anxiety.48 This sense of
belonging is important because of the potential for
modifi cation of school learning and social environments.
In one study, a change of school was not associated with
deterioration in psychological functioning of UASC,11
whereas in an 8-year follow-up in Denmark attendance at
several diff erent schools was predictive of high
externalising behaviour scores in young refugees.35
Strong school connectedness was positively linked to

Review
274
www.thelancet.com Vol 379 January 21, 2012
Summary
Family exposure to violence
Almqvist et al,8 1997 Most children whose fathers had been imprisoned showed abnormal play behaviour
Almqvist et al,9 1999 Severe traumatic exposure in parents correlated with severe traumatic exposure in children
Angel et al,10 2001 Lack of family discussions about adverse events during confl ict was associated with fewer psychological problems
Cohn et al,15 1985 Children whose parents had been tortured showed high rates of emotional and somatic disorders
Daud et al,16 2008 Values for I Think I Am and Strengths and Diffi culties questionnaires were similar for children with traumatised and non-traumatised parents
Montgomery et al,34 2006 Two violent exposures—mother tortured and father disappeared—and a child being informed about parents’ detention were independently predictive of
post-traumatic stress disorder (many other exposures were predictive of individual symptoms but not the post-traumatic stress disorder complex); parental
torture before the child’s birth did not increase risk
Rousseau et al,41 1998 Family trauma before the child’s birth was a predictor of internalising and externalising symptoms in Central American but not southeast Asian children
Rousseau et al,42 1999 Family trauma before the child’s birth was protective in terms of externalising symptoms, risk behaviour, and school failure in boys; in girls, it was associated
positively with social adjustment
Rousseau et al,44 2003 Family trauma after a child’s birth showed no relation with any assessment measurements. No relation was noted between internalising and externalising
symptoms in adolescence with prebirth familial trauma. In boys, the severity of prebirth exposure to familial trauma correlated with increased self-esteem in later
adolescence, and decreased perception of racism, but no eff ect was noted in girls
Family composition and bereavement
Bean et al,11 2007 Unaccompanied adolescents had higher internalising problems, traumatic stress reactions, and stressful life events than did other groups
Bean et al,12 2007 Change of guardian did not aff ect measurements. Reactions of Adolescents to Traumatic Stress and internalising scores were much higher if no family member
was in the host country, or if living in a large residential setting rather than foster care or a small group setting
Derluyn et al,18 2009 Unaccompanied children were more likely to have experienced all types of adverse events than were accompanied children
Ellis et al,20 2008 Arriving with or without parents was not associated with post-traumatic stress disorder or depression
Geltman et al,22 2005 Separation from family was associated with post-traumatic stress disorder, whereas presence of any biological family member or friend during migration or in
camp was not associated with post-traumatic stress disorder. Sudanese children living in a group home or foster care without other Sudanese people were more
likely to have post-traumatic stress disorder than were those in foster care with other Sudanese people (other fostered children or foster family)
Hodes et al,27 2008 Unaccompanied asylum-seeking children had higher numbers of traumatic events and scored higher on the Impact of Event Scale than did accompanied refugee
children; Birleson Depression Self-Rating Scale scores were not diff erent
Unaccompanied asylum-seeking children in low-support living arrangements showed signifi cantly higher scores for post-traumatic stress symptoms than did
those living with more support
For accompanied children, there was no diff erence in scores between those living in two-parent or single-parent households
Porte et al,38 1987 Refugee children in homes with people of the same ethnic origin (own families or foster care) were much less depressed than were those living in either foster or
group homes with local families
Rousseau et al,41 1998 In children of southeast Asian origin, but not Central American children, being from a single parent household was a predictor of internalising symptoms; number
of people in household was not a predictor
Rousseau et al,43 2000 Increasing household size and living with a single parent were associated with internalising symptoms in Cambodian but not Central American refugees
Rousseau et al,45 2004 Living in a one-parent or two-parent household had no eff ect in early adolescence (mean age 14 years); at follow-up in mid-adolescence (16 years), single-parent
households were associated with higher feelings of competence, but those in two-parent households had fewer internalising symptoms; no association was noted
between one-parent or two-parent household and externalising symptoms
Tousignant et al,49 1999 Prevalence of psychological problems was fi ve times lower in boys living with both parents than in boys living in other family arrangements. Number of residences
since birth was higher in boys with a diagnosis than in boys without a diagnosis. Children in host families with no kin contacts had more psychological problems
Family functioning and parental health
Ajdukovic et al,7 1993 Children had high stress scores if their mothers reported diffi culties in adapting to their new environment or felt burdened with confl icts, if they had diffi cult
mother–child relationships, if their separated family members were in diffi cult circumstances (such as being prisoners of war), or if they had diffi culty contacting
separated family members
Children living in host families rather than in shelters had lower stress scores, but had high stress scores if their mothers had negative views of the people they
were living with
Almqvist et al,9 1999 Maternal emotional wellbeing was a determinant of the child’s emotional wellbeing at follow-up
Marital discord predicted poor general adaptation
Angel et al,10 2001 Children whose parents were in need of psychiatric treatment had high total problem and anxiety scores, but association was attributable to shared stressors
in multiple regression analysis; parents’ talking about the war seemed to exacerbate negative eff ects of stress in worst-aff ected children, but not in those
least aff ected
Berthold et al,14 2000 Perceived family support negatively correlated with post-traumatic stress disorder and depression
Ekblad et al,19 1993 Having a mother who was described as apathetic or unstable was linked to poor mental health of the child; an optimistic mother was linked to good mental health
Lack of information about the fl ight was associated with poor mental health in children
Grgic et al,23 2005 Number of traumatic events was associated with Hudson’s Index of Family Relations score; most children perceived family relationships as positive, but
21% reported substantial problems with family relationships
Hjern et al,26 1998 Separation from family members and parental divorce after migration were associated with high symptom scores initially and at follow-up. Having a parent in
psychiatric care was associated with high symptom scores at follow-up. Recent family stressors were a predictor of poor mental health
Kovacev et al,30 2004 Greater parental support was associated with high Global Self Worth scores
(Continues on next page)

Review
www.thelancet.com Vol 379 January 21, 2012
275
Summary
(Continued from previous page)
Montgomery et al,34 2006 Current family structure, a change in parental behaviour towards the child, and the child being informed about family exposures to violence and the reason for
escape were not predictors of post-traumatic stress disorder
Montgomery et al,36 2010 Refugees without many symptoms were likely to speak to their mothers frequently about problems, both at arrival and 8–9-year follow-up
Rousseau et al,43 2000 No association of emotional and behavioural problems was noted with parental unemployment
Rousseau et al,45 2004 Protective eff ect of family cohesion was associated with feelings of competence in early adolescence (mean age 14 years), and reduced externalising symptoms in
mid-adolescence (16 years)
Family confl ict did not increase risk in early adolescence, but was associated with increased parental report of externalising symptoms in late adolescence
Sujoldzic et al,48 2006 Poor family connectedness was associated with depression
Household socioeconomic circumstances
Ajdukovic et al,7 1993 Children whose mothers reported having poorer fi nancial and material support had high stress scores
Angel et al,10 2001 No independent eff ect of prefl ight socioeconomic status was noted on the child’s psychological wellbeing at follow-up
Hjern et al,26 1998 Socioeconomic conditions in the host country did not correlate with symptom scores
Hodes et al,27 2008 Premigration socioeconomic circumstances of unaccompanied asylum-seeking children and accompanied refugee children were not diff erent
Montgomery et al,34 2006 Parents who had premigration occupations and fathers who had worked in private enterprises were predictive of post-traumatic stress disorder, but the parents’
economic situation and social class were not
Sujoldzic et al,48 2006 Low affl uence was associated with depression
Tousignant et al,49 1999 Downward professional mobility or current parental employment status was not associated with psychological problems, but a period of prolonged paternal
unemployment in the fi rst year of arrival was associated with psychological problems
Parental education
Ekblad et al,19 1993 Father’s high educational achievement was linked to poor mental health of child
Montgomery et al,34 2006 Parents’ education was not a predictor of post-traumatic stress disorder
Montgomery et al,35 2008 Less maternal education was associated with more externalising behaviours
Montgomery et al,36 2010 Refugees whose symptoms resolved during follow-up had fathers with long periods of education in the home country
Rousseau et al,41 1998 In children of southeast Asian origin, parental education was associated with externalising behaviour
Rousseau et al,43 2000 No association with emotional or behavioural problems was noted with parental education
Tousignant et al,49 1999 Parental education was not a predictor of psychological problems
Social support and community integration
Almqvist et al,9 1999 Having peers to play with was predictive of higher scores of general adaptation and social adjustment
Bean et al,12 2007 Change of school during 1-year follow-up did not aff ect levels of psychological distress or behavioural problems
Berthold et al,14 2000 Perceived support from friends negatively correlated with post-traumatic stress disorder and depression
Ekblad et al,19 1993 Social support associated with good mental health in child
Ellis et al,20 2008 Carer fl uency in the host language (English) was not associated with post-traumatic stress disorder or depression; acculturative stressors were related to
post-traumatic stress disorder, but not depression; perceived discrimination was associated with post-traumatic stress disorder and depression
Hjern et al,26 1998 Extent of the social network did not aff ect symptom scores
Geltman et al,22 2005 In Sudanese children, living in a group home or foster care with an American family without other Sudanese people was associated with post-traumatic
stress disorder, whereas living with a Sudanese family or with an American family alongside other Sudanese children was not, and feeling safe at home was
associated with a reduced risk
Feeling safe at school was associated with a reduction in risk of post-traumatic stress disorder
This disorder was associated with children feeling less comfort with host society and culture, feeling lonely or isolated where they were living, and reduced
participation and satisfaction in group activities
Kia-Keating et al,28 2007 Sense of school belonging was negatively predictive of depressive symptoms, irrespective of past exposure to adversities
Kovacev et al,30 2004 Increased support from classmates and close friends was associated with high Global Self Worth scores
The acculturation styles of assimilation and separation were not predictive of Global Self Worth scores; integration was positively predictive of Global Self Worth
and marginalisation was negatively predictive
Liebkind et al,32 1993 Children identifi ed increasingly with host country with time
Liebkind et al,33 1996 Presence of people of the same ethnic origin in the community was protective against anxiety for girls but not boys
Discrimination experiences were not predictors of disorders
Dissociation from traditional family values and a positive attitude towards acculturation was predictive of anxiety symptoms in girls but not boys
For boys, adherence to traditional family values had a protective eff ect
Montgomery et al,35 2008 The number of types of postmigration discrimination experiences was associated with increased internalising behaviours
Not attending school or work and high number of school changes were associated with externalising behaviour
Higher numbers of Danish friends was associated with lower internalising behaviours
Nielsen et al,37 2008 Children who had at least four relocations had poor mental health
Rousseau et al,41 1998 In children of southeast Asian origin, having a large potential social network of the same ethnic origin was a predictor of internalising and externalising symptoms
Central American children having an active social network from the host country was a predictor of internalising behaviour, possibly related to greater demands
and obligations associated with an extended family network in some cultures
Number of people in the household was not a predictor
(Continues on next page)

Review
276
www.thelancet.com Vol 379 January 21, 2012
self-esteem, whereas low social support at school was
correlated with increased depression.48
Little connectedness to the neighbourhood was associated
with depression.48 The presence of wide kin contacts and
the mother often receiving visitors at home were
protective.49 Living and socialising alongside other people
of the same ethnic origin seems to provide protection from
psychological morbidity, particularly while in foster care.22,38
The presence of people of the same ethnic origin had a
protective eff ect against anxiety in Vietnamese girls, but no
eff ect in boys.33 However, the extent of social networks per
se was not associated with psychological functioning in
one report,26 and large social networks with people of the
same ethnic origin were associated with poor functioning
in another report.41
Acculturation is usually a slow, subtle, and continuous
process, and is especially diffi cult to measure quantita-
tively.65 It is usually assessed as linguistic competency and
time since migration, which are only some of its
components. Notably, the usefulness of the term
acculturation in health research has been extensively
debated by social scientists. Some degree of alignment
with the host culture is probably protective. In an
Australian study30 of acculturation in adolescent refugees,
integration into the host society (maintaining the
individual’s original culture while participating in the host
society, as assessed with a range of scales to measure self-
worth, peer acceptance, and attitudes to acculturation) was
linked to improved psychosocial adjustment. Separation
(mainly maintaining the individual’s own culture) or
assimilation (adaptation to the values of the host society)
were not predictors of psychosocial adjustment, whereas
marginalisation had negative eff ects.30
Many adolescents perceive themselves to be more
acculturated than are their parents, and an increasing
gap in acculturation during adolescence can generate
Summary
(Continued from previous page)
Rousseau et al,45 2004 Acculturation alone had no eff ect on externalising or internalising symptoms, or on competence, but it interacted with sex such that less acculturated boys had
more symptoms
The degree of acculturation did not change with the duration of follow-up
Parental fl uency in the host language had no eff ect
Sujoldzic et al,48 2006 Poor school connectedness was associated with depression, anxiety, and somatic stress. Poor attachment to the neighbourhood was associated with depression;
perceived discrimination was associated with poor psychological functioning in Bosnians in Croatia but not Austria
Tousignant et al,49 1999 Parental fl uency in the host language, social integration, and frequency of conversations outside the home were not associated with psychological eff ects, but
mothers with few visitors and few wider kin contacts were both associated with psychological problems
Ideological and religious contexts
Montgomery et al,35 2008 Being Christian or Muslim negatively predicted clinically signifi cant internalising behaviour, but not overall internalising or externalising symptoms
Sujoldzic et al,48 2006 Religious commitment was protective against depression and anxiety
Ethnic origin
Bean et al,12 2007 Unaccompanied asylum-seeking children from Eritrea, Ethiopia, and Guinea had higher externalising and internalising scores than did those from other African
countries and China
Hjern et al,26 1998 Region of origin had no eff ect on symptom scores
Hodes et al,27 2008 Region of origin had no eff ect on trauma scores, but refugees from the Middle East had higher Birleson Depression Self-Rating Scale scores than did those from
Europe or Africa
Rousseau et al,41 1998 Ethnic origin was the main determinant of children’s emotional profi le
Tousignant et al,49 1999 Rates of psychological problems were similar for all regions except the Middle East, which showed lower rates
Resettlement location
Bean et al,12 2007 Children relocated to another regional immigration offi ce during 1 year follow-up had higher internalising scores than did those who were not relocated
Ellis et al,20 2008 Housing adequacy was not associated with post-traumatic stress disorder or depression
Fox et al,21 1999 No association was noted between adverse events in refugee camps and subsequent psychological problems
Geltman et al,22 2005 Living in an urban neighbourhood in the host country was not associated with post-traumatic stress disorder
Liebkind et al,32 1993 Area of residence in host country had no eff ect on mental health
Nielsen et al,37 2008 In children who had four or more relocations within the host country, the odds ratio of having mental health diffi culties was 3
Tousignant et al,49 1999 No association was noted between residence in a refugee camp en route to the host country and psychological eff ects later. Boys with psychiatric diagnoses had a
greater number of changes in residence
Immigration process
Bean et al,12 2007 Lack of a temporary residence permit was linked to high internalising scores
Ekblad et al,19 1993 Long period in a refugee camp in the host country was linked to poor mental health
Ellis et al,20 2008 Number of years in the USA was negatively associated with depression; no association with post-traumatic stress disorder was noted
Resettlement stress was associated with post-traumatic stress disorder but not depression
Nielsen et al,37 2008 Children who had been seeking asylum for more than 1 year had poor mental health
Reijneveld et al,39 2005 Unaccompanied girls seeking asylum in a restrictive reception centre showed more internalising symptoms than did girls in a less restrictive setting
Table 3: Summary of principal fi ndings in relation to family, community, and societal factors assessed in each study

Review
www.thelancet.com Vol 379 January 21, 2012
277
discord.66,67 Achievement of competence in the host
country’s language can be associated with a reduced
likelihood of depressive symptoms68 and internalising
behaviour scores35 in young refugees, but the parents’
language profi ciency seems unrelated to children’s
psychological outcomes.20,45,49 Adherence to traditional
values of family hierarchy according to age and sex
seemed to be protective, whereas dissociation from these
values and a positive attitude towards adoption of the
host country’s culture were predictive of poor
psychological functioning.33 In Somali adolescents
resettled in the USA, closer alignment with the Somali
culture was associated with better mental health for girls,
whereas closer alignment with the American culture was
associated with better mental health for boys.69 A high
rate of post-traumatic stress disorder was predicted by
acculturative stress in the same population,20 and among
Sudanese refugees who felt lonely, isolated, or less
comfortable in US society than in their own.22 Bosnian
refugees who felt connected to their neighbourhoods had
low rates of depression,48 and refugees with support from
friendships had improved psychosocial adjust ment.30 In
a longitudinal study of Cambodian refugee adolescents
in Canada,45 changes in acculturation were not noted over
2 years, although the refugees were assessed 10–12 years
after arrival. The degree of acculturation alone (as
measured by the adoption of customs, habits, and
language of the host country) was not associated with
psychological functioning, but when combined with sex
the least acculturated boys seemed most vulnerable. 45
This evidence shows complex associations between the
experiences of the adolescent, family, and society, as
expected from the ecological model of concentric spheres
of eff ect.53,70
Societal factors
Ideological and religious contexts
The evidence for religious beliefs is mixed. Among
Bosnian adolescents resettled in Austria and Croatia,
religious commitment (assessed as a composite of
frequency of participation in religious activities and
degree of subjective personal belief) was associated with
low anxiety and depressive symptoms.48 Among Middle-
Eastern groups in Denmark, Muslim and Christian
refugee adolescents had lower scores for internalising
behaviours than did those who belonged to a persecuted
minority religion or had changed or abandoned their
faith.35 Spiritual attributions about the meaning of
adverse events might also be important, but no studies
met our inclusion criteria; nonetheless, in a small
qualitative study of unaccompanied Sudanese boys
(aged 16–18 years), attribution of adverse events to God’s
will contributed to fairly good functioning.71 The links
between religion, faith, hope, agency, and sense of
coherence and responsibility implicated in risk and
resilience pathways are complex, as shown for confl ict-
aff ected populations,72 and reports often do not do justice
to these complexities.73
Ethnic origin
The eff ect of ethnic origin on child mental health has
been assessed in several studies,26,27,41,49 with mixed
fi ndings indicating that refugees from diff erent
countries of origin have diff erent types and duration of
exposure to potentially traumatic events and
premigration circum stances, as well as cultural
diff erences in the response to distressing events. In
Bean and colleagues’ study12 of UASC, coming from
48 diff erent countries, adolescents from diff erent
Domain
assessed
Number of studies* Total number
of children†
Risk or protective factor
Exposure to premigration violence Individual 138,10,14,17,19,20,22,24,26,35,40,46,48 3099 Risk
Female sex Individual 1112,14,17,26,27,32,35,39,46,48,49 3425 Risk (mainly for internalising or emotional
problems)
High parental support and family cohesion Family 414,30,45,48 1576 Protective
Self-reported support from friends Community 48,14,30,35 397 Protective
Unaccompanied Family 311,22,27 3690 Risk
Perceived discrimination Community 320,35,48 1548 Risk
Exposure to postmigration violence Individual 313,35,48 1489 Risk
Self-reported positive school experience Family 328,30,48 1441 Protective
Several changes of residence in host country Community 312,37,49 1031 Risk
Parental exposure to violence Family 38,34,41 517 Risk
Poor fi nancial support Family 27,48 1601 Risk
Same ethnic-origin foster care Family 222,38 386 Protective
Single parent Family 241,49 359 Risk
Parental psychiatric problems Family 210,26 162 Risk
Only factors that were validated in at least two studies resulting in the same direction of eff ect, were included in the Review. *Reported as one study if the same sample was
used in more than one reported study. †Includes forcibly displaced children and comparator groups, hence high numbers of participants.
Table 4: Summary of risk and protective factors for mental health outcomes in forcibly displaced children

Review
278
www.thelancet.com Vol 379 January 21, 2012
countries of origin had diff erent patterns of
psychological distress. Therefore, the heterogeneity of a
refugee population needs careful attention in terms of
policy and clinical need.
Resettlement location
No associations were noted between mental health
outcome and whether children had lived in a refugee
camp before arriving in the host country,49 or specifi c
adverse events in a refugee camp.21
UASC who were transferred to a diff erent regional
asylum offi ce had higher scores for internalising
diffi culties and traumatic stress symptoms than did those
who were not transferred.12 Four or more relocations
within the asylum system were predictive of poor mental
health in children and adolescents in Denmark.37 The
fewer the lifetime residence changes for boys the better
their mental health outcomes.49 High-support living
arrangements reduced psychological symptoms for
UASC,12,17,27 although a change of guardian during a 1-year
follow-up did not adversely aff ect psychological distress
or behaviour in UASC.11
Displaced Croatian children whose families were
accommodated with host families rather than in shelters
had few symptoms of stress.7 Accommodation in centres,
rather than living alone or in foster care, was associated
with poorer functioning in UASC in Belgium.17 Feeling
safe in an individual’s own home was associated with low
occurrences of post-traumatic stress disorder,22 but
housing adequacy was not a predictor of depression or
post-traumatic stress disorder in Somali adolescents in
the USA.20 Area of residence within the host country,
whether urban or rural, does not seem to be associated
with psychological functioning.22,32,49 Thus a sense of
safety and privacy rather than housing quality or location
might be most important to wellbeing.
Time since displacement
Residence duration in the host country has been negatively
associated with depression20 but not post-traumatic stress
disorder.20,22 Results of long-term studies indicate a trend
towards reduction of symptoms with time.36
Immigration process
Postmigration detention seems to be especially
detrimental to children’s mental health. Cuban refugee
children, detained for many months in Guantanamo
Bay before entry to the USA, showed high levels of
psychological symptoms.40 The high prevalence of
psychiatric illness during and after children’s detention
have been shown in small-scale studies,39,74,75 and girls
might be especially vulnerable to the adverse eff ects of
restrictive reception settings.39 After detention, intrusive
memories are common,74 since children might be
exposed to fi res, rioting, violence, and self-harm
attempts by parents or others while detained.76 Children
are more likely to suff er adverse mental health
consequences when detained in restrictive rather than
routine reception facilities.39
Rapid but careful resolution of asylum claims reduces
the duration of uncertainty, insecurity, and associated
distress for children. Insecure asylum status is associated
with a range of psychological problems.12,37,77 Experiences
during immigration interviews22 and detention after migra-
tion can be especially distressing for children, compounding
premigration negative experiences of authority and placing
them in situations that can be perceived as being worse
than adversity before migration.76
Long-term outcomes
The importance of longitudinal studies to help understand
prospectively which risk and protective factors are causally
associated with psychological outcomes should not be
underestimated. Few such studies, however, have been
reported.35,78–80 The results of Hjern and colleagues’78
6–7-year follow-up of refugee children in Sweden showed
Panel 1: Suggestions for further research
Information about specifi c groups of children
• Children with pre-existing physical or psychological disorders, or learning diffi culties.
• Ex-combatants.
• Traffi cked children.
• Children with alternative carers.
• Children living in refugee camps, or those forced to live and work on the streets.
• Children with uncertain immigration status.
• Refugees or asylum seekers returning (involuntarily or voluntarily) to their home
country from high-income settings.
Longitudinal studies
• Prospective predictors of mental health, and pathways to risk and resilience.
• Long-term eff ect of forced migration on individual psychological outcomes; and
on structure and functioning of families and their interactions with displaced and
host communities.
• Individual, family, community, and societal contexts aff ecting experiences and mental
health outcomes, rather than designs restricted to quantifi cation of associations
between adverse exposures and poor mental health outcomes.
Eff ects of various risk and protective factors on child development
• Specifi c types of violence exposures, and links between domestic, structural, and
collective violence.
• Diff erent parenting styles.
• Past periods of stable family life and education.
• Fostering within the child’s own ethnic or language group.
• The role of social networks in promoting resilience.
• Support to integrate into local communities.
• Foreign language acquisition for parents and children.
Best interventions
• Evidence for which interventions work best is insuffi cient. Although there is evidence
for individual-level interventions such as pharmacotherapy for depression and
individual-focused trauma treatments for post-traumatic stress disorder, the eff ects of
community or societal interventions, including group psychotherapeutic treatments
or school treatments, parental and teacher interventions, family-based treatments,
and housing initiatives, are still unknown.

Review
www.thelancet.com Vol 379 January 21, 2012
279
improvements in mental health outcomes with time,
although past exposure to violence in the home country
and recent family stressors were predictive of psychological
disturbance. In Sack and colleagues’ 12-year follow-up
study79,81 of Cambodian adolescents, starting when they
were aged 14–20 years (mean 17 years) in the USA,
depression was more closely related to postmigration
stressors than to past confl ict-related events, whereas
diagnoses of post-traumatic stress disorder were linked to
adverse confl ict-related experiences. Mental health
trajectories were variable with time, and although post-
traumatic stress disorder tended to persist, depression
initially decreased substantially during 3–6 years of follow-
up, only to rise again between 6–12 years of follow-up. The
refugees in this study seemed to be fairly resilient overall,
having few comorbid problems such as behavioural
problems or substance misuse, and most were in
education or employment. In Montgomery’s follow-up
study35 of 131 refugees in Denmark, the long-term eff ects
of premigration adversity were mediated by a variety of
diff erent risk and protective factors.35 Aspects of social life
in Denmark and stresses experienced in exile were more
predictive of psychological problems 8–9 years after arrival
than were adverse experiences before arrival, emphasising
the importance of the postmigration environment in
easing recovery from distressing experiences.
Conclusions and recommendations
Many diff erent factors aff ect the mental health of
forcibly displaced children in the presence of substantial
life challenges. Table 4 summarises the key protective
and risk factors. In accord with Reed and colleagues’
Review,53 premigration exposure to violence was strongly
predictive of psychological disturbance. Family factors
and living arrangements have received much more
attention in high-income settings than in the low-
income and middle-income settings. Overall, the ability
to integrate into the host society while maintaining a
sense of one’s cultural identity is protective, but its eff ect
has not been quantifi ed.
The evidence lends support to the idea of spirals of loss,82
drawing attention to the way many challenges aff ect
refugees at all stages of their journeys. The after-eff ects of
migration on the wellbeing of refugee children are wide-
ranging and powerful, and many are modifi able. Increased
prevalence of mental health disorders among displaced
children is likely to be a result of the increased exposure to
risk factors. Postmigration factors provide opportunities
for high-income countries to intervene directly to achieve
improved outcomes for vulnerable children, yet the
possibility of intervention by governments and non-
governmental organisations in high-income countries to
keep negative exposures to a minimum in countries of
origin and countries of transit should not be neglected.
Cumulative adversities usually worsen health outcomes,
exerting more powerful eff ects than any factor alone.83
The most harmful pathways are those that involve
exposure to violence—whether individually experienced,
witnessed, or feared—and the loss of family support by
death or violence,84 for both behavioural and emotional
mental health outcomes. As emphasised by the WHO
frame work,85 risks cannot be simply added up, but the
inter-related pathways that lead to the outcomes need to
be assessed.86,87 Thus, although distal or premigration
factors contribute to childhood adversities, repeated
exposure to violence and lack of safety soon after migration
or displacement are of pivotal importance. In the model
proposed by Pynoos and colleagues88 to understand the
complex processes involved in trauma-related
psychological changes, one important issue is that new
traumatic experiences can reawaken previous traumatic
memories, erode previous adaptation, and create
secondary adversities; however, the possibility of post-
traumatic growth—in which individuals might be better
able to achieve various goals after their experiences—
might be an alternative trajectory, but remains to be
investigated in depth in relation to refugee children.89
Health professionals need to assess the multiplicity of
ongoing challenges to the wellbeing of refugees, if they
are to advocate on refugees’ behalf when the imple-
mentation of immigration, social, or health policies is to
the disadvantage of a highly vulnerable community. The
ecological model70 provides a helpful conceptual
framework to shape humanitarian responses to children
in crisis.90 This model emphasises that children develop
in a social milieu in which family, community, and
society contribute to the quality of daily life.90 Thus
prioritisation of certain policies such as the reunifi cation
of children with their families or other carers, the
reinstitution of school education, and community-
building activities are key.91
Panel 2: Summary of policy implications for forcibly displaced children in
low-income, middle-income, and high-income settings
Individual
• Reduction in postmigration exposure to diff erent types of violence and threat,
including interventions that address intrafamilial violence, bullying, and racism
• Access to physical and psychological health services
Family
• Harness local community resources to assist with integration of children and families
• Support safe and appropriate cultural beliefs and social practices
• Prioritise reunion of children with families or other carers
• Provide support for families to remain intact and to reduce confl ict
Community and societal
• Stable settlement in host country, including rapid resolution of asylum claims;
minimisation of relocation; supported educational placements and employment
opportunities for children older than 16 years and parents; specifi c support for
unaccompanied children; improved provision of services, enabling cultural continuity
for religion and language
• Concerted action in health, social, economic, and political sectors to reduce
inequalities in access to resources

Review
280
www.thelancet.com Vol 379 January 21, 2012
Successful intervention with distressed refugee children
requires not only psychotherapeutic skills, but also these
in combination with structural interventions such as
those targeting adequate housing and psychosocial
interventions like access to skills training.92 Ideally such
resources should be available to ease integration for
refugee children and their families from the time of
arrival,93 with the aim of preventing adverse mental health
outcomes. Additionally, the ability to advise families of
ways they can optimise and integrate all the important
factors identifi ed in this Review, as they forge a new path
in the host country, is essential.
The elucidation of protective factors provides the
building blocks in the identifi cation of pathways to
resilience in children. There is general recognition that
an understanding of resilience is important for the
development of interventions: focus on the identifi cation
and mobilisation of adaptive systems within the
individual, family, and cultural systems is key.91 Hodes
and colleagues27 have suggested further investigation of
the role of past periods of stable family life and education,
the young person’s own appraisal of adversity, the role of
fostering within the individual’s own ethnic or language
group, and the value of social networks in promoting
resilience.27 Little is understood about the nature and
eff ects of parenting styles in refugee families94 and
whether interventions could promote resilience and
modify outcomes for parents and children. Help in terms
of support to integrate into local communities, and
language acquisition for both parents and children are
interventions that warrant formal assessment. One way
forward is to provide a comprehensive and sophisticated
approach to understanding the inter-relationships
between individual, family, community, or societal risk
and protective factors in the assessment of the causal
pathways that link psychological problems to mental
health outcomes.53 More research is needed (panel 1).
First, we need to improve our understanding of children
in particular groups who have, thus far, received little
attention. Second, longitudinal study designs are needed
to understand the processes and pathways involved in
mental health outcomes, including elucidation of
mediating and moderating variables.95 Third, we need to
have an improved understanding of the family,
community, and societal contexts in which refugee
children live. Fourth, we need to develop interventions
that are then assessed according to internationally
agreed guidelines.
Consideration should also be given to the long-term
outcomes. Evidence suggests that complex comorbidities
of post-traumatic stress disorder and other disorders are
not uncommon in adult survivors of childhood forced
displacement.96,97 Furthermore, evidence suggests that
the next generation is also aff ected;98 even children born
to refugee parents after migration are at increased risk of
psychotic disorders compared with the native popu-
lation,99 whereas second-generation labour immigrants
are not. Such complexities of the intergenerational
aspects of coping with adverse experiences and social
disadvantage are poorly understood in the context of
refugee families. Limitations of the work so far include
the assessment of a narrow set of predictor variables,
particularly those that focus on individual exposures,
heterogeneous research designs in studies, and, in many
cases, the lack of eff ect sizes, which restrict our ability to
draw defi nitive conclusions. Panel 2 summarises some
policy recommendations.
Prompt, but fair and thorough, investigation and
resolution of refugee status is essential to enable
individuals with a genuine claim to settle rapidly in the
host country. Frequent moves, delays, and prolonged
bureaucratic processes have negative eff ects on children’s
mental health. Equitable and prompt access to services
for physical and psychological health, and access to good
housing and schooling are central to adaptation and
positive mental health. A means of livelihood for families
is not only important for adequate nutrition and
wellbeing, but allows families to integrate into the new
society. Unaccompanied children are especially vul-
nerable, and need specifi c support to ensure they can
benefi t from long-term stability of residence and social
environment. Prolonged uncertainty about asylum status
endangers their mental health. Since mental health
problems originating in childhood and adolescence are
often longlasting, high-income countries must implement
immigration, health-care, and social policies that support
family units and keep deleterious consequences for child
health and development to a minimum.
Contributors
All authors were involved in the conceptualisation and the design of the
Review. RVR undertook the literature searches. RVR and MF selected
the studies. RVR gathered data from the studies. RVR and MF compiled
the tables. RVR, MF, CP-B, and AS wrote the Review. All authors have
read and approved the fi nal version of the Review.
Confl icts of interest
We declare that we have no confl icts of interests.
Acknowledgments
We thank M Berthold, I Derluyn, A Hjern, E Montgomery, and
C Rousseau, the authors of papers we included in our Review for their
helpful correspondence; K Welch for assistance in undertaking the
literature search, particularly of the grey literature; and Lynne Jones and
the anonymous reviewers for their constructive comments that led to an
improved Review.
References
1 UNHCR. 2008 global trends: refugees, asylum seekers, returnees,
internally displaced and stateless persons. Geneva: UN High
Commissioner for Refugees, 2009.
2 Feijen L. The challenges of ensuring protection to unaccompanied
and separated children in composite fl ows in Europe.
Refugee Surv Q 2008; 27: 63–73.
3 Ryan D, Dooley B, Benson C. Theoretical perspectives on
post-migration adaptation and psychological well-being among
refugees: towards a resource-based model. J Refugee Stud 2008;
21: 1–18.
4 Berry JW. Comparative studies of acculturative stress. Int Migr Rev
1987; 21: 491–511.
5 Goldin S, Levin L, Persson LA, Hagglof B. Stories of pre-war, war
and exile: Bosnian refugee children in Sweden. Med Confl Surviv
2001; 17: 25–47.

Review
www.thelancet.com Vol 379 January 21, 2012
281
6 Silove D, Steel Z, Watters C. Policies of deterrence and the mental
health of asylum seekers. JAMA 2000; 284: 604–11.
7 Ajdukovic M, Ajdukovic D. Psychological well-being of refugee
children. Child Abuse Neglect 1993; 17: 843–54.
8 Almqvist K, Brandell-Forsberg M. Refugee children in Sweden:
post-traumatic stress disorder in Iranian preschool children
exposed to organized violence. Child Abuse Neglect 1997; 21: 351–66.
9 Almqvist K, Broberg AG. Mental health and social adjustment in
young refugee children 3 1/2 years after their arrival in Sweden.
J Am Acad Child Psychiatry 1999; 38: 723–30.
10 Angel B, Hjern A, Ingleby D. Eff ects of war and organized violence
on children: a study of Bosnian refugees in Sweden.
Am J Orthopsychiatry 2001; 71: 4–15.
11 Bean T, Derluyn I, Eurelings-Bontekoe E, Broekaert E, Spinhoven P.
Comparing psychological distress, traumatic stress reactions, and
experiences of unaccompanied refugee minors with experiences of
adolescents accompanied by parents. J Nerv Ment Dis 2007;
195: 288–97.
12 Bean TM, Eurelings-Bontekoe E, Spinhoven P. Course and
predictors of mental health of unaccompanied refugee minors in
the Netherlands: one year follow-up. Soc Sci Med 2007; 64: 1204–15.
13 Berthold SM. The eff ects of exposure to community violence
on Khmer refugee adolescents. J Trauma Stress 1999; 12: 455–71.
14 Berthold SM. War traumas and community violence: psychological,
behavioral, and academic outcomes among Khmer refugee
adolescents. J Multicult Soc Work 2000; 8: 15–46.
15 Cohn J, Danielsen L, Holzer KIM, et al. A study of Chilean refugee
children in Denmark. Lancet 1985; 2: 437–38.
16 Daud A, af Klinteberg B, Rydelius PA. Resilience and vulnerability
among refugee children of traumatized and non-traumatized
parents. Child Adolesc Psychiatry Ment Health 2008; 2: 1–11.
17 Derluyn I, Broekaert E. Diff erent perspectives on emotional
and behavioural problems in unaccompanied refugee children
and adolescents. Ethn Health 2007; 12: 141–62.
18 Derluyn I, Mels C, Broekaert E. Mental health problems in
separated refugee adolescents. J Adolesc Health 2009; 44: 291–97.
19 Ekblad S. Psychosocial adaptation of children while housed in a
Swedish refugee camp: aftermath of the collapse of Yugoslavia.
Stress Med 1993; 9: 159–66.
20 Ellis BH, MacDonald HZ, Lincoln AK, Cabral HJ. Mental health
of Somali adolescent refugees: the role of trauma, stress, and
perceived discrimination. J Consult Clin Psychol 2008; 76: 184–93.
21 Fox PG, Cowell JM, Montgomery AC. Southeast Asian refugee
children: violence experience and depression.
Int J Psychiatr Nurs Res 1999; 5: 589–600.
22 Geltman PL, Grant-Knight W, Mehta SD, et al. The “ lost boys of
Sudan”: functional and behavioral health of unaccompanied
refugee minors re-settled in the United States.
Arch Pediatr Adolesc Med 2005; 159: 585–91.
23 Grgic M, Vidovic V, Butkovic-Soldo S, Vuksic-Mihaljevic Z,
Degmecic D, Laufer D. The mental health of children upon their
return home after a long displacement period. Coll Antropol 2005;
29: 537–42.
24 Grgic M, Vidovic V, Soldo-Butkovic S, Koic O. Depression and
perceived family functioning in Croatian displaced children.
Drustvena Istrazivanja 2005; 14: 597–608.
25 Hjern A, Angel B, Hojer B. Persecution and behavior: a report of
refugee children from Chile. Child Abuse Neglect 1991; 15: 239–48.
26 Hjern A, Angel B, Jeppson O. Political violence, family stress and
mental health of refugee children in exile. Scand J Soc Med 1998;
26: 18–25.
27 Hodes M, Jagdev D, Chandra N, Cunniff A. Risk and resilience
for psychological distress amongst unaccompanied asylum seeking
adolescents. J Child Psych Psychiatry 2008; 49: 723–32.
28 Kia-Keating M, Ellis BH. Belonging and connection to school in
resettlement: young refugees, school belonging, and psychosocial
adjustment. Clin Child Psychol Psychiatry 2007; 12: 29–43.
29 Kocijan-Hercigonja D, Rijavec M, Marusic A, Hercigonja V. Coping
strategies of refugee, displaced, and non-displaced children in a war
area. Nord J Psychiat 1998; 52: 45–50.
30 Kovacev L. Acculturation and social support in relation to
psychosocial adjustment of adolescent refugees resettled in
Australia. Int J Behav Dev 2004; 28: 259–67.
31 PSIH. Sarajevo 2000. The psychosocial consequences of war.
Results of empirical research from the territory of former
Yugoslavia. http://www.psih.org./2000e.pdf (accessed May 11, 2011).
32 Liebkind K. Self-reported ethnic identity, depression and anxiety
among young Vietnamese refugees and their parents. J Refugee Stud
1993; 6: 25–39.
33 Liebkind K. Acculturation and stress: Vietnamese refugees
in Finland. J Cross Cult Psychol 1996; 27: 161–80.
34 Montgomery E, Foldspang A. Validity of PTSD in a sample of
refugee children: can a separate diagnostic entity be justifi ed?
Int J Methods Psychiatr Res 2006; 15: 64–74.
35 Montgomery E. Long-term eff ects of organized violence on young
Middle Eastern refugees’ mental health. Soc Sci Med 2008;
67: 1596–603.
36 Montgomery E. Trauma and resilience in young refugees: a 9-year
follow-up study. Dev Psychopathol 2010; 22: 477–89.
37 Nielsen SS, Norredam M, Christiansen KL, Obel C, Hilden J,
Krasnik A. Mental health among children seeking asylum in
Denmark—the eff ect of length of stay and number of relocations:
a cross-sectional study. BMC Public Health 2008; 8: 293.
38 Porte Z, Torney-Purta J. Depression and academic achievement
among Indochinese refugee unaccompanied minors in ethnic and
nonethnic placements. Am J Orthopsychiatry 1987; 57: 536–47.
39 Reijneveld SA, de Boer JB, Bean T, Korfker DG. Unaccompanied
adolescents seeking asylum: poorer mental health under a
restrictive reception. J Nerv Ment Dis 2005 Nov; 193: 759–61.
40 Rothe EM, Lewis J, Castillo-Matos H, Martinez O, Busquets R,
Martinez I. Posttraumatic stress disorder among Cuban children
and adolescents after release from a refugee camp. Psychiatr Serv
2002; 53: 970–76.
41 Rousseau C, Drapeau A, Corin E. Risk and protective factors
in Central American and Southeast Asian refugee children.
J Refugee Stud 1998; 11: 20–37.
42 Rousseau C, Drapeau A, Platt R. Family trauma and its association
with emotional and behavioral problems and social adjustment in
adolescent Cambodian refugees. Child Abuse Neglect 1999; 23: 1263–73.
43 Rousseau C, Drapeau A, Platt R. Living conditions and emotional
profi les of Cambodian, Central American, and Quebecois youth.
Can J Psychiatry 2000; 45: 905–11.
44 Rousseau C, Drapeau A, Rahimi S. The complexity of trauma
response: a 4-year follow-up of adolescent Cambodian refugees.
Child Abuse Neglect 2003; 27: 1277–90.
45 Rousseau C, Drapeau A, Platt R. Family environment and
emotional and behavioural symptoms in adolescent Cambodian
Refugees: infl uence of time, gender, and acculturation.
Med Confl Surviv 2004; 20: 151–65.
46 Slodnjak V, Kos A, Yule W. Depression and parasuicide in refugee
and Slovenian adolescents. Crisis 2002; 23: 127–32.
47 Sourander A. Behavior problems and traumatic events of
unaccompanied refugee minors. Child Abuse Neglect 1998;
22: 719–27.
48 Sujoldzic A, Peternel L, Kulenovic T, Terzic R. Social determinants
of health—a comparative study of Bosnian adolescents in diff erent
cultural contexts. Coll Antropol 2006; 30: 703–11.
49 Tousignant M, Habimana E, Biron C, Malo C, Sidoli-LeBlanc E,
Bendris N. The Quebec Adolescent Refugee Project:
psychopathology and family variables in a sample from 35 nations.
J Am Acad Child Psychiatry 1999; 38: 1426–32.
50 Zivcic I. Emotional reactions of children to war stress in Croatia.
J Am Acad Child Psychiatry 1993; 32: 709–13.
51 Betancourt TS, Khan KT. The mental health of children aff ected by
armed confl ict: protective processes and pathways to resilience.
Int Rev Psychiatry 2008; 20: 317–28.
52 Green H, McGinnity A, Meltzer H, Ford T, Goodman R. Mental
health of children and young people in Great Britain, 2004.
Basingstoke: Palgrave MacMillan, 2005.
53 Reed RV, Fazel M, Jones L, Panter-Brick C, Stein A. Mental health of
displaced and refugee children resettled in low-income and middle-
income countries: risk and protective factors. Lancet 2011; published
online Aug 10. DOI:10.1016/S0140-6736(11)60050-0.
54 Sack WH, Clarke GN, Kinney R, Belestos G, Him C, Seeley J.
The Khmer Adolescent Project. II: functional capacities in two
generations of Cambodian refugees. J Nerv Ment Dis 1995; 183: 177–81.

Review
282
www.thelancet.com Vol 379 January 21, 2012
55 Almqvist K, Brandell-Forsberg M. Iranian refugee children in
Sweden: eff ects of organized violence and forced migration on
preschool children. Am J Orthopsychiatry 1995; 65: 225–37.
56 Rousseau C, Said TM, Gagne M-J, Bibeau G. Resilience in
unaccompanied minors from the north of Somalia. Psychoanal Rev
1998; 85: 615–37.
57 Luthar SS. Resilience in development: a synthesis of research
across fi ve decades. In: Cicchetti D, Cohen DJ, eds. Developmental
Psychopathology. 2nd edn. Hoboken, NJ: Wiley, 2006: 739–95.
58 Derluyn I, Broekaert E, Schuyten G. Emotional and behavioural
problems in migrant adolescents in Belgium.
Eur Child Adolesc Psychiatry 2008; 17: 54–62.
59 Stein A, Ramchandani P, Murray L. Impact of parental psychiatric
disorder and physical illness. In: Rutter M, Bishop D, Pine D, et al,
eds. Rutter’s child and adolescent psychiatry. 5th edn. Oxford:
Blackwell, 2008: 407–20.
60 Hinton DE, Rasmussen A, Leakhena N, Pollack MH, Good M-J.
Anger, PTSD, and the nuclear family: a study of Cambodian
refugees. Soc Sci Med 2009; 69: 1387–94.
61 Mghir R, Raskin A. The psychological eff ects of the war in
Afghanistan on young Afghan refugees from diff erent ethnic
backgrounds. Int J Soc Psychiatry 1999; 45: 29–36.
62 Miller KE, Rasmussen A. War exposure, daily stressors, and mental
health in confl ict and post-confl ict settings: Bridging the divide
between trauma-focused and psychosocial frameworks.
Soc Sci Med 2010; 70: 7–16.
63 Heptinstall E, Sethna V, Taylor E. PTSD and depression in
refugee children: associations with pre-migration trauma and
post-migration stress. Eur Child Adolesc Psychiatry 2004;
13: 373–80.
64 Rousseau C, Drapeau A, Corin E. The infl uence of culture and
context on the pre- and post-migration experience of school-aged
refugees from Central America and southeast Asia in Canada.
Soc Sci Med 1997; 44: 1115–27.
65 Phinney JS. Ethnic identity in adolescents and adults: review
of research. Psychol Bull 1990; 108: 499–514.
66 Nguyen NA, Williams HL. Transition from east to west: Vietnamese
adolescents and their parents. J Am Acad Child Psychiatry 1989;
28: 505–15.
67 Rick K, Forward J. Acculturation and perceived intergenerational
diff erences among Hmong youth. J Cross Cult Psychol 1992; 23: 85–94.
68 Sack WH. Multiple forms of stress in refugee and immigrant
children. Child Adolesc Psychiatr Clin N Am 1998; 7: 153–67.
69 Ellis BH, Lincoln A, MacDonald HZ, Klunk-Gillis J, Strunin L,
Cabral HJ. Discrimination and mental health among Somali
refugee adolescents: the role of acculturation and gender.
Am J Orthopsychiatry 2010; 80: 564–75.
70 Bronfenbrenner U. The ecology of human development:
experiments by nature and design. Cambridge, MA: Harvard
University Press, 1979.
71 Goodman JH. Coping with trauma and hardship among
unaccompanied refugee youths from Sudan. Qual Health Res 2004;
14: 1177–96.
72 Eggerman M, Panter-Brick C. Suff ering, hope, and entrapment:
Resilience and cultural values in Afghanistan. Soc Sci Med 2010;
71: 71–83.
73 Barber B. Adolescents and war: how youth deal with political
violence. Oxford: Oxford University Press, 2009.
74 Steel Z, Momartin S, Bateman C, et al. Psychiatric status of asylum
seeker families held for a protracted period in a remote detention
centre in Australia. Aust N Z J Public Health 2004; 28: 527–36.
75 Lorek A, Etntholt K, Nesbitt A, et al. The mental and physical health
diffi culties of children held within a British immigration detention
center: a pilot study. Child Abuse Neglect 2009; 33: 573–85.
76 Mares S, Newman L, Dudley M, Gale F. Seeking refuge, losing
hope: Parents and children in immigration detention.
Australasian Psychiatry 2002; 10: 91–96.
77 Bodegard G. Pervasive loss of function in asylum-seeking children
in Sweden. Acta Paediatr 2005; 94: 1706–07.
78 Hjern A, Angel B. Organized violence and mental health of
refugee children in exile: a six-year follow-up. Acta Paediatr 2000;
89: 722–27.
79 Sack WH, Clarke G, Him C, et al. A 6-year follow-up study
of Cambodian refugee adolescents traumatized as children.
J Am Acad Child Psychiatry 1993; 32: 431–37.
80 Sack WH, Him C, Dickason D. Twelve-year follow-up study of
Khmer youths who suff ered massive war trauma as children.
J Am Acad Child Psychiatry 1999; 38: 1173–79.
81 Sack WH, McSharry S, Clarke GN, Kinney R, Seeley J, Lewinsohn P.
The Khmer Adolescent Project. I. Epidemiologic fi ndings in two
generations of Cambodian refugees. J Nerv Ment Dis 1994;
182: 387–95.
82 Hobfoll SE. The infl uence of culture, community, and the
nested-self in the stress process: Advancing conservation of
resources theory. Appl Psychol 2001; 50: 337–421.
83 Rutter M, Maughan B, Mortimore P, Ouston J. Fifteen thousand
hours. Secondary schools and their eff ects on children. London:
Open Books, 1979.
84 Catani C, Gewirtz A, Wieling E, Schauer E, Elbert T, Neuner F.
Tsunami, war, and cumulative risk in the lives of Sri Lankan
schoolchildren. Child Dev 2010; 81: 1176–91.
85 CSDH. Closing the gap in a generation: Health equity through
action on the social determinants of health. Geneva: World Health
Organization, 2008.
86 Kessler RC, McLaughlin M, Green J, et al. Childhood adversities
and adult psychopathology in the WHO World Mental Health
Surveys. Br J Psychiatry 2010; 197: 378–85.
87 Panter-Brick C, Goodman A, Tol W, Eggerman, M. Mental health
and childhood adversities: A longitudinal study in Kabul,
Afghanistan. J Am Acad Child Adol Psychiatry 2011; 50: 349–63.
88 Pynoos RS, Steinberg AM, Piacentin JC. A developmental
psychopathology model of childhood traumatic stress and
intersection with anxiety disorders. Biol Psychiatry 1999;
46: 1542–54.
89 Powell S, Rosner R, Butollo W, Tedeschi RG, Calhoun LG.
Posttraumatic growth after war: A study with former refugees
and displaced people in Sarajevo. J Clin Psychol 2003; 59: 71–83.
90 Kostelny K. A culture-based, integrative approach. In: Boothby N,
Strang A, Wessells M, eds. A world turned upside down: Social
ecological approaches to children in war zones. Bloomfi eld, CT:
Kumarian, 2006: 19–38.
91 Ager A, Stark L, Akesson B, Boothby N. Defi ning best practice in
care and protection of children in crisis-aff ected settings: a Delphi
study. Child Dev 2010; 81: 1271–86.
92 Betancourt TS, Williams T. Building an evidence base on mental
health interventions for children aff ected by armed confl ict.
Intervention (Amstelveen) 2008; 6: 39–56.
93 Correa-Velez I, Giff ord SM, Barnett AG. Longing to belong: social
inclusion and wellbeing among youth with refugee backgrounds in
the fi rst three years in Melbourne, Australia. Soc Sci Med 2010;
71: 1399–408.
94 Williams N. Establishing the boundaries and building bridges.
A literature review on ecological theory: implications for research
into the refugee parenting experience. J Child Health Care 2010,
14: 35–51 .
95 Tol W, Komproe I, Susanty D, Jordans MJ, Macy R, de Jong J.
School-based mental health intervention for children aff ected by
political violence in Indonesia. A cluster randomized trial.
JAMA 2008; 300: 655–62.
96 Hubbard J, Realmuto GM, Northwood AK, Masten AS. Comorbidity
of psychiatric diagnoses with posttraumatic stress disorder in
survivors of childhood trauma. J Am Acad Child Psychiatry 1995;
34: 1167–73.
97 Kuwert P, Brahler E, Glaesmer H, Freyberger HJ, Decker O.
Impact of forced displacement during World War II on the
present-day mental health of the elderly: a population-based study.
Int Psychogeriatr 2009; 21: 748–53.
98 Sagi-Schwartz A, MH van IJzendoorn, Grossman KE, et al.
Attachment and traumatic stress in female Holocaust child
survivors and their daughters. Am J Psychiatry 2003; 160: 1086–92.
99 Leao TS, Sundquist J, Johansson LM, Johansson SE, Sundquist K.
Incidence of mental disorders in second-generation immigrants in
Sweden: a four-year cohort study. Ethnic Health 2005; 10: 243–56.