Content uploaded by Patricia L. East
Author content
All content in this area was uploaded by Patricia L. East on Oct 16, 2017
Content may be subject to copyright.
Vol.:(0123456789)
1 3
J Immigrant Minority Health
DOI 10.1007/s10903-017-0624-2
ORIGINAL PAPER
The Impact ofRefugee Mothers’ Trauma, Posttraumatic Stress,
andDepression onTheir Children’s Adjustment
PatriciaL.East1· SheilaGahagan1· WaelK.Al‑Delaimy2
© Springer Science+Business Media, LLC 2017
Keywords Somali refugees· Refugee trauma·
Intergenerational traumatization· Transmission of trauma·
Children of refugees
Introduction
As of 2015, approximately 150,000 Somalis were living in
the U.S., with over 43,000 Somali refugees entering since
2000 [1]. The first large wave of Somalis seeking refuge
began in 1991 at the start of the Somali civil war. To escape
violence from military unrest as well as the famine of
1991–1992, many Somalis fled to Kenya and were placed
in the Dadaab refugee camp. The Dadaab camp has been
described as overcrowded, unsanitary, and with limited
access to food, clean water, and medical care [2]. Resettle-
ment out of the camp is a long and arduous process, with
final relocation often taking several years. Like many refu-
gees, Somalis enter countries of resettlement with high lev-
els of trauma exposure and significant mental health issues
[3]. Torture and civil war before migration, and displace-
ment, loss, and hopelessness during long periods in refugee
camps contribute to the development of psychopathology
for many Somali refugees [4].
Within the public health literature, the trauma and
comorbid depressive symptoms of adult refugees have
been well documented [5]. However, there is limited
understanding of how the trauma experienced by adult
refugees impacts their children. Intergenerational trau‑
matization [6], or parent-to-child transmission of trauma,
is the process whereby those close to a trauma survivor
develop psychological distress symptoms similar to those
of the survivor. Intergenerational transmission of trauma
is believed to occur through the effect of trauma on par-
ents’ ability to function as a caregiver and parent [7, 8].
Abstract The mechanisms linking refugee parents’
trauma onto their children’s functioning are not well under-
stood. The current study sought to identify how Somali ref-
ugee mothers’ past trauma and current mental health impact
their children’s psychosocial adjustment. One hundred and
ninety-eight Somali mothers (M age = 39 years) and their
children (M age = 10 years; 56% male) were studied. On
average, mothers spent 7 years in refugee camps, expe-
rienced significant trauma, and some had been tortured.
Measures of mothers’ posttraumatic stress and depression
were analyzed as three symptom clusters: volatility/panic,
withdrawn/detached, and depressed mood. Most children
were born in the U.S. and their indirect exposure to trauma
was statistically controlled. Results from structural equa-
tion modeling indicated that there was no direct association
between trauma of the mother and their children’s well-
being, however, mothers’ posttraumatic stress and depres-
sive symptoms significantly mediated the effects of moth-
ers’ past torture on their children’s adjustment—a pattern
indicative of intergenerational traumatization. Findings
enhance our understanding of how refugees’ traumatization
lingers and possibly affects their and their children’s health
and well-being.
* Patricia L. East
peast@ucsd.edu
1 Department ofPediatrics, University ofCalifornia, San
Diego School ofMedicine, 9500 Gilman Drive, Mail Code
0927, LaJolla, CA92093-0927, USA
2 Department ofFamily Medicine andPublic Health,
University ofCalifornia, San Diego School ofMedicine,
LaJolla, CA92093, USA
J Immigrant Minority Health
1 3
A socialization mode of transmission is emphasized, as is
role modeling or the child observing and imitating a trau-
matized parent’s maladaptive behaviors [8, 9].
Evidence of intergenerational traumatization has
been confirmed largely by documenting the adjust-
ment difficulties of children of traumatized parents [10].
Studied initially among the children of Holocaust sur-
vivors, reports indicated that such children suffered
from problems with separation, depression, and aggres-
sion [11]. The children of war veterans with PTSD also
show behavior problems of aggression, delinquency,
and hyperactivity [9]. More recently, among the chil-
dren of refugees, parents’ torture experiences [7, 10]
and, separately, parents’ PTSD symptoms [12, 13] have
been shown to relate to children’s psychological distress,
internalizing problems, and hyperactivity. However, the
complete intergenerational traumatization process—from
parents’ past trauma to parents’ current mental health
to child adjustment difficulties—is not well understood.
The current study sought to better understand the spe-
cific mental health symptoms that act as traumatization
transmission mechanisms, or how the past trauma expe-
rienced by Somali refugee mothers affects their current
posttraumatic stress and depressive symptoms which, in
turn, impact their child’s adjustment. We examined three
symptom clusters common among traumatized refugees:
depressed mood (melancholic thoughts, feelings and
anhedonia), and the withdrawn/detached and volatility/
panic components of posttraumatic stress disorder [14].
The three distinctive symptom clusters noted above may
have unique disruptive effects on children’s adjustment. For
example, depressive symptoms interfere with parents’ emo-
tional engagement, playful interactions, and verbal respon-
sivity toward their child, with the children of depressed
mothers having greater social difficulties and internalizing
symptoms than the children of non-depressed mothers [15].
The withdrawn/detached cluster of posttraumatic stress
entails avoiding activities with others and alienating one-
self from social interaction. Studying the children of tor-
ture victims, Daud and colleagues speculated that mothers’
withdrawal and detachment could signal indifference of the
child and thereby adversely affect the child’s self-esteem
[7]. A traumatized parents’ withdrawal and detachment can
also negatively affect parents’ capacity to provide adequate
care and socialization to their child, which can lead to child
depression, acting-out behaviors, and youth violence [16,
17]. The posttraumatic stress symptoms of hyperarousal
and volatility (angry outbursts, panic attacks, irritability)
might likely take the form of rash and punitive parenting
among traumatized adults [18, 19]. Among Cambodian ref-
ugee parents, anger directed toward a child was four times
more common than anger directed toward a spouse [20].
Parental anger and PTSD symptoms of hyperarousal have
been found to be associated with child depression, anxiety,
and child displays of aggression [19, 20].
The Current Study
We sought to identify the associations among Somali
refugee mothers’ past trauma, their current posttraumatic
stress and depressive symptoms, and their child’s adjust-
ment. Specifically, we investigated the extent to which the
three symptom clusters of depressed mood, withdrawal/
detachment, and volatility/panic mediate the association
between mothers’ past trauma and torture and their chil-
dren’s functioning. The conceptual model that guided this
study is shown in Fig.1. We examined the unique effects
associated with mothers’ past torture and, separately, the
extent of trauma experienced, given evidence of distinctive
effects associated with each [3, 5]. The effects of parents’
trauma on their children are also related to several addi-
tional factors, such as the child’s age and gender [10, 21].
These and other factors were controlled in analyses to iso-
late the effects of mothers’ past trauma and current mental
health on their child’s functioning. Most mothers included
in this study fled Somali in the late 1990s and gave birth to
their children in the U.S. Thus, most children in this study
are second-generation U.S. and did not experience direct
trauma from Somali’s civil war nor refugee camp life.
Methods
Participants
Participants were 198 Somali refugee women and their
children who were interviewed and completed question-
naires as part of a research study in San Diego, CA. Somali
mother–child pairs were recruited in 2013 through door-to-
door soliciting in neighborhoods and apartment complexes
in City Heights, San Diego, where there are high concentra-
tions of Somali residents. Three female Somali community
workers were assigned the recruitment given their bilingual
and cultural understanding of this population. The study
was explained to potential eligible participants, or Somali
women who had spent time in a refugee camp and who had
at least one child 7–14years of age. Mothers and the oldest
eligible-aged child were invited to participate.
Procedure
The study interview took place in the participant’s home
or in a neighboring community room available by private
reservation per the preference of the participating mother.
Questionnaires were offered in English and Somali per the
J Immigrant Minority Health
1 3
preference of the mother and child. All questionnaires were
translated to Somali and then back translated to English
to verify linguistic and cultural comparability. All meas-
ures were also pilot tested prior to conducting the study
to ensure their acceptability for the study population. The
questionnaires were read to mothers and children by a
bilingual, bicultural Somali female community worker who
was trained in the administration of the study’s instruments
and verified the comprehension of the survey questions.
Completion of the questionnaires took approximately 1h.
All participants’ responses were entered onto a secure digi-
tal platform and exported as excel files using a password-
protected server. This study was approved by the Univer-
sity of California Human Subjects Protection Committee.
Mothers provided informed, written consent for their and
their child’s participation, and children provided signed
assent. All consent/assent/permission forms were verbally
reviewed with mothers and children in native Somali.
Mothers were given a $30 incentive to compensate them
for their time participating.
Measures
Indices of mothers’ trauma included an extensive assess-
ment of past traumatic and torture experiences prior to
arriving in the U.S. As indicators of children’s functioning,
children completed measures of their depressive symptoms,
and experiences of racism and peer bullying victimiza-
tion. Bullying and racism directed toward refugee children
have been considered key factors in post-migration social
inclusion and social integration [22], and are studied here
as indicators of children’s social adaptation. For example,
experiences of peer bullying and discrimination among
resettled youth are integrally related to youths’ self-esteem,
mental health and general well-being [23, 24].
Measures Completed byMothers
Harvard Trauma Questionnaire (HTQ)
Mothers completed the HTQ, which was developed by the
Harvard Program in Refugee Trauma to inquire about trau-
matic events and emotional symptoms uniquely associated
with refugee trauma [25]. The HTQ has been modified and
adapted to the characteristics of several cultural groups, as
the actual traumatic events and meanings attributed to them
vary per the historical, political, and social context in which
the trauma occurred [26]. The present study used a version
of the HTQ in which the items were translated into Somali
and adapted to the Somali context and then back translated
to English by bilingual staff. The questionnaire included
questions about experiences of 56 traumatic events deter-
mined to affect refugees, including: material deprivation,
warlike conditions, bodily injury, forced confinement, and
coercion to inflict harm, harm to a loved one, witnessing
violence, and head injuries. Respondents answered “Yes,”
or “No” to whether they experienced each event. Affirma-
tive responses were summed to indicate the extent of trau-
matic events experienced (range: 0–56). An additional 33
items listed specific torture events, to which the respondent
Fig. 1 Conceptual model of
mothers’ mental health symp-
toms mediating the association
between mothers’ past trauma
and children’s current adjust-
ment. Paths from mothers’ tor-
ture to child’s racism and from
mothers’ traumatic events to
child depressive symptoms are
also hypothesized but are not
shown for ease of presentation
Mothers’ Past Trauma Mothers’ Mental Health Children’s Adjustment
Perceived
Racism
Bullying
Victimization
Depressive
Symptoms
Depressed
mood
Withdrawn/
Detached
Torture
Experiences
Traumatic Events
Volatility/
Panic
J Immigrant Minority Health
1 3
responded whether she had experienced that event (“yes,”
“no”). Affirmative responses to these items were summed
to indicate the extent of torture the participant had experi-
enced (0–33).
The HTQ also includes 42 items that describe post-
traumatic stress symptoms, categorized into four areas of
functioning [26]: re-experiencing, arousal, avoidance, and
symptoms specific to refugees or those associated with per-
secution and displacement (“survival guilt,” “ashamed,”
“betrayed”; 12 items). Mothers responded regarding the
extent to which each symptom bothered them within the
last week, with response options of: “not at all,” “a little,”
“quite a bit,” and “extremely,” scored as 0–3, respectively.
Hopkins Symptom Checklist (HSCL)
The HSCL measures 15 symptoms of depression and 10
symptoms of anxiety, experienced during the past week
[27]. Response options consisted of how much each symp-
tom bothered the respondent within the past week: “not at
all,” “a little,” “quite a bit,” and “extremely,” coded 0–3,
respectively.
Measures Completed byChildren
Children’s Depressive Symptoms
Children’s depressive symptoms were assessed by child-
ratings on the Children’s Depression Inventory-2 (CDI-2),
which consists of 28 items describing symptoms of depres-
sion [28]. Children responded to each item indicating the
extent to which the item applied to them during the last 2
weeks. Response options ranged from 0 to 2 [e.g., “I am
sad once in a while” (0), “I am sad all the time” (2)]. The
CDI-2 can be analyzed as five subscales: negative mood,
interpersonal problems, ineffectiveness, anhedonia, and
negative self-esteem [28]. To determine if that factor struc-
ture held within the current sample, we computed princi-
pal axis factoring using direct oblimin rotation. Results
suggested that a 4-factor solution best explained the data.
The variance explained by the solution was 56.30%, and
the four factors individually accounted for 25.6, 11.9,
10.6, and 8.3% of the variance, respectively. Using the pat-
tern matrix for interpretation, eight items loaded onto the
first factor (factor loadings >0.48), six items loaded onto
the second factor (>0.40), five items loaded onto the third
factor (>0.55), and six items loaded onto the fourth factor
(>0.38). The factors were identified as negative self-esteem
(“I do everything wrong”), negative mood (“I am sad all
the time”), ineffective (“I do badly in school”), and inter-
personal problems (“I have no friends”). The intercorrela-
tions among the factors were >0.52.
Perceived Racism Scale‑Children (PRS‑C) [29]
This 10-item scale assesses: (a) the frequency with which
the child has personally experienced each of six racist
events (“Teachers think I am not as smart because I am
Somali”), by indicating a score of 0 (not occurred) to 5
(occurred several times a day); and (b) the level of insti-
tutional racist beliefs (“Many Somali schools are in bad
condition because of racism”) by indicating disagreement
(1 = strongly disagree) or agreement (5 = strongly agree).
Peer Victimization Scale
The 16-item Multicultural Peer Victimization scale [30]
was administered to children to assess peer bullying
within the most recent school year, or the following types
of victimization, assessed by four items each: physical
victimization (“kicked me”), verbal victimization (“swore
at me”), social manipulation (“refused to talk to me”),
and attacks on property (“broke something of mine”).
Response options were: 0 (“almost never”), 1 (“once”),
and 2 (“more than once”).
Covariates
Mothers’ Background Characteristics
The study questionnaire included several questions about
mothers’ background (years of education, marital status,
language most often spoken, etc.) and relocation expe-
riences (Table 1). In addition, several questions asked
about mothers’ social status in Somali, her perceived
social status in the U.S., and her perceived change in sta-
tus from Somali to the U.S.
Children’s Exposure toTraumatic Events
Children completed the Harvard-Uppsala Trauma Ques-
tionnaire for Children (HUTQ-C) [31] in a semi-struc-
tured interview format with a Somali female researcher
to provide clarification and verification as needed. The
HUTQ-C assesses children’s exposure to traumatic
events. Children were asked to indicate if they had expe-
rienced, witnessed, or heard about a list of 30 events
(e.g., a kidnapping, murder of a stranger). Affirmative
responses were summed individually to indicate the num-
ber of events experienced, witnessed, and heard about.
Prior to use in this study, the HUTQ-C was translated
into Somali, and back-translated by our native Somali
research staff to verify accuracy.
J Immigrant Minority Health
1 3
Analytic Strategy
In efforts to identify cohesive and reliable indices of
maternal symptoms, we first identified questionnaire items
corresponding to the three symptom clusters of mothers’
withdrawal/detachment, volatility/panic and depressed
mood. Items that reflected withdrawal from social activi-
ties and emotional detachment as outlined in criteria C of
posttraumatic stress disorder in the DSM‑IV [14] involved
the seven items on the HTQ avoidance scale and one item
from the HSCL (“no interest in things”). Principal compo-
nent analysis (PCA) identified one factor (all factor load-
ings ≥0.67) which involved five items (shown in Table2).
Items reflecting volatility/panic as outlined in criteria D of
PTSD in the DSM‑IV [14] involved the ten items from the
HTQ anxiety scale, the four items on the HTQ re-experi-
encing scale, and the five items on the HTQ arousal scale.
PCA identified three factors (all factor loadings ≥0.50)
of six, five and three items each, which we termed vola-
tility, panic, and fear, respectively (Table 2). Depressed
mood items were drawn from the 15 items on the HSCL
depression scale, one item from the HTQ avoidance scale
(“don’t have a future”), and one item added by the inves-
tigators (“want to die”). PCA identified two factors (load-
ings ≥0.53), which we termed melancholic thoughts and
feelings (eight items), and low energy (four items). All
scales had good internal consistency (Table3) and signifi-
cant factor loadings in our analytic model (≥0.57; Fig.2;
described below in “The Measurement Model”). Previous
studies have examined similar symptom clusters associated
with posttraumatic stress and depression and find unique
associations between these clusters and individual adjust-
ment indicators [32].
Once reliable indicators were identified, structural
equation modeling using Mplus 6.0 [33] was used to
evaluate the model shown in Fig.1. Because we were
primarily interested in mediation effects from mothers’
past trauma and torture to children’s adjustment as medi-
ated by mothers’ mental health symptoms, we first tested
the full model that included all direct effects, including
direct effects from mothers’ past trauma and torture to
children’s adjustment (Fig. 1). We then tested a nested
model, excluding the direct effects from mothers’ past
trauma and torture to child’s adjustment. A Chi square
difference test was conducted to determine the better
fitting model. For all models, model fit was determined
by reviewing indices of good model fit [34], includ-
ing: a nonsignificant Chi square, comparative fit index
(CFI) >0.90, root mean square error of approximation
(RMSEA) <0.08, and standardized root mean square
residual (SRMR) <0.08 Mediation was tested using the
INDIRECT command within Mplus, which estimates
indirect effects with the delta method standard errors
Table 1 Description of sample (198 mothers, 198 children)
a Includes: Syria, Saudia Arabia, Egypt, United Arab Emirates, South
Africa, and Switzerland
b Assessed using a ladder diagram of ten rungs, where the top rung
represents the highest social status within the community and the bot-
tom rung represents the lowest status
c Response options were: 1 = doing worse financially in the U.S. than
in Somalia; 2 = about the same; 3 = doing better financially in the U.S.
than in Somalia
d Response options were: 1 = poorly, 2 = fairly well, 3 = very well
M or % Min Max SD
Mothers’ age (years) 39.4 26 57 7.2
Marital status
Married 91%
Divorced 4%
Separated 3%
Widowed 2%
Language speak most often
Somali 95%
English 1.5%
Other 3.5%
Religion
Muslim 99.5%
Christian 0.5%
Educational level
No schooling 35%
Elementary school 7%
Middle school 17%
High school 35%
Trade or business school 2%
Bachelor’s degree or higher 4%
Length of resettlement (years) 7.3 0 23 6.0
Area of resettlement
Kenya 79%
Ethiopia 12%
Yemen 3%
Othera6%
Number of relocations
1 (directly from refugee camp) 90%
2+ 10%
Years in the U.S. 13.7 0.20 26 5.9
Social status in Somaliab6.2 1 10 1.6
Social status in U.S.b2.7 1 10 1.0
Financial status changec2.0 1 3 2.0
Ability to meet financial needsd1.6 1 3 0.6
Child age (years) 10.4 7 14 2.1
Child grade 5.4 2 9 2.1
Child gender (% male) 56%
Child born in the U.S. 76%
J Immigrant Minority Health
1 3
[33]. Analyses were conducted using maximum likeli-
hood estimators, which are robust to non-normality.
There was no missing data for any of the variables.
Given the expected associations among the within-
domain variables, all within-domain variables were cor-
related a priori.
Results
Descriptive Results
Mothers’ average age was 39years, all but one were Mus-
lim, and 91% were currently married (Table1). The average
level of mothers’ formal education was elementary school,
with 35% of mothers having no formal schooling. Mothers
experienced an average of 7.3years in refugee camps (range:
0–23years), predominantly in Kenya, before arriving in the
U.S. in the 1990s. Ninety percent of mothers arrived in the
U.S. directly from a refugee camp; 10% had two or more
relocations before arriving in the U.S. Mothers had spent an
average of 13.7years in the U.S. (range 0.20–26.0years) at
the time of the study. Participating children were an aver-
age age of 10 years old, most were in the 5th grade, and
56% were boys. Most children were born in the U.S. (76%).
Those who were not born in the U.S., had spent an average
of 4.9years in the U.S. (range 0.20–10.0years).
Table3 provides descriptive results of the study’s meas-
ures. Mothers reported experiencing an average of 24 trau-
matic events (range: 0–56), including: the murder or violent
death of a family member (77%), witnessing someone being
physically harmed (beaten, knifed, 42%), witnessing mass
execution of civilians (41%), home confiscated or destroyed
(49%), and imprisonment (39%). Mothers had, on average,
experienced one torture event prior to arriving in the U.S.
(range: 0–33), including: forced to stand for long periods of
time (5.6%), deprived of food and water for long periods of
time (6.1%), subjected to mock executions (5.1%), starva-
tion (6.1%), and punched, slapped, kicked, or stricken with
objects (4.5%). Children reported experiencing or witness-
ing very few traumatic events themselves. However, chil-
dren reported hearing about approximately three traumatic
events, determined subsequently by research staff to be chil-
dren overhearing parents talking about a traumatic event or
a parent talking on the phone about a traumatic event.
Correlations andInclusion ofCovariates
Intercorrelations were computed among the model vari-
ables and correlations were computed between the model
variables and all participant background variables shown
in Table 1. Based on these results (Table 4), all partici-
pant background variables that were significantly related
to a model variable were included as a covariate on that
variable in the modeling analyses. For example, mothers’
educational level was included as a covariate on mothers’
torture experiences, children’s bullying victimization, and
children’s depressive symptoms.
Table 2 Items used as indicators of latent variables
a Assessed by the Harvard Trauma Questionnaire
b Assessed by the Hopkins Symptom Checklist
Withdrawn/detached
Withdrawn-detacheda
Can’t feel emotionsa
Less interest in daily routinea
Avoid activitiesa
No interest in thingsb
Volatility/panic symptoms
Volatility
Angry outbursta
Sudden emotional, physical reactiona
Jumpy, easily startleda
On guarda
Restlessb
Hard to concentratea
Panic
Terror, panic attacksb
Tremblingb
Heart pounding, racingb
Nervous, shakyb
Feel trappedb
Fear
Scaredb
Fearfulb
Tense, keyed upb
Depressed mood
Melancholic thoughts, feelings
Hopelessa, b
Want to dieb
Lonelyb
Worthlessb
Feel blueb
Suicidal thoughtsb
Don’t have a futurea
Cry easilyb
Low energy
Poor appetiteb
Low energyb
Everything an effortb
Lost interest in sexb
J Immigrant Minority Health
1 3
The Measurement Model
The latent variables were constructed within Mplus for the
mediating and child-rated endogenous variables (Fig. 2).
All standardized loadings were generally large and statis-
tically significant (values ranged from 0.56 to 0.91). The
standard coefficient of the correlation between number of
torture experiences and number of traumatic events was
0.38 (p < .000, not shown in Fig. 2). However, the corre-
lations between mothers’ volatility/panic and mothers’
depressed mood and withdrawal/detachment exceeded
one. We thus correlated the observed variables of each
mother-rated latent variable and fixed the parameter of
volatility/panic to 1. This resulted in a measurement model
with adequate fit (χ2 [123] = 271.95, p < .001, CFI = 0.896,
RMSEA = 0.078, SRMR = .106), sufficient to proceed to
the structural model.
Modeling Results
The model that included the direct paths from mothers’
torture and trauma to children’s outcomes was compared
to a model that excluded these direct effects. The fit statis-
tics for the full model were: χ2 [112] = 213.93 (p < .001),
Table 3 Scores of mothers’
past trauma, current mental
health symptoms and children’s
functioning
a Scores reflect sum of events
b Scores range from 0 = not bothered at all by symptoms to 3 = bothered extremely by symptoms
c Scores range from 0 = not occurred to 5 = occurred several times a day
d 1 = strongly disagree to 5 = strongly agree
e Scores range from 0 = occurred almost never to 2 = occurred more than once
f Scores indicate extent of depressive symptoms (range: 0–2)
Mean SD Possible range Number of
items
Cronbach alpha
Mother‑reported
Traumatic eventsa24.35 9.65 0–56 56 –
Torture experiencesa1.01 3.88 0–33 33 –
Posttraumatic stress symptomsb
Withdrawn-detached 0.74 0.34 0–3 5 0.77
Volatility-panic
Volatility 0.51 0.69 0–3 6 0.86
Panic 0.27 0.56 0–3 5 0.88
Fear 0.49 0.75 0–3 3 0.81
Depressed moodb
Melancholic thoughts 0.33 0.63 0–3 8 0.92
Low energy 0.46 0.66 0–3 4 0.82
Child‑reported
Racism
Experiencesc0.71 0.80 0–5 6 0.90
Institutional racist beliefsd1.73 1.10 0–5 4 0.87
Bullying/victimizatione
Physical 0.20 1.08 0–2 4 0.80
Manipulation 0.28 1.23 0–2 4 0.72
Verbal 0.43 1.40 0–2 4 0.75
Property 0.24 1.11 0–2 4 0.71
Depressive symptomsf
Negative self-esteem 0.26 0.29 0–2 8 0.73
Negative mood 0.13 0.27 0–2 6 0.76
Ineffectiveness 0.28 0.38 0–2 5 0.74
Interpersonal problems 0.25 0.31 0–2 6 0.71
Controls
Child experienced traumatic eventa0.14 0.70 0–30 30 –
Child witnessed traumatic eventa0.09 0.54 0–30 30 –
Child heard about traumatic eventa2.86 7.73 0–30 30 –
J Immigrant Minority Health
1 3
CFI = 0.923, RMSEA = 0.068, SRMR = .046, and the fit
statistics for the nested, more parsimonious model were:
χ2 [118] = 220.78 (p < .001), CFI = 0.922, RMSEA = 0.066,
SRMR = .049. The Chi square difference test between the
two models was nonsignificant (χ2 [6] = 6.85, p = .33),
however, the Akaike Information Criterion and Bayesian
Information Criterion were smaller for the nested model,
and the paths from mothers’ torture and trauma to each
of the child adjustment variables were nonsignificant (β’s
ranged from −0.06 to 0.08, p > .37). These results support
the more parsimonious model (without the direct effects) as
the better fitting model and this model was examined for its
path coefficients.
(Fig.3).
Fig. 2 Measurement model.
Model fit indices: Chi square
[123] = 271.95, p < .001,
CFI = 0.896, RMSEA = 0.078,
SRMR = .106. N = 198. The
items and corresponding load-
ings for the withdrawn/detached
factor are: withdrawn-detached
0.80, can’t feel emotions 0.70,
less interest in daily routine
0.62, avoid activities 0.60, no
interest in things 0.57
Mothers’ Mental Health Children’s Adjustment
Perceived
Racism
Bullying
Victimization
Depressive
Symptoms
Depressed
Mood
Withdrawn/
Detached
Experiences
Institutional
Neg mood
Ineffectiveness
Neg self-esteem
Interpersonal
.81
.73
.70
.74
.56
Physical
Manipulation
Verbal
Property
.88
.75
.79
Items 1 -5
Panic
Fear
Volatile
.66
.70
Melancholia
Low energy
.76
.76
.80
.81
.64.69
.65
.63
Volatility/
Panic
.57-.80
.91
.89
.75
Table 4 Intercorrelations among model variables and correlations between model variables and significant controls
*p<0.05;**p<0.01;***p<0.001
a Sum of traumatic events experienced
b Sum of torture experiences
c Latent variable; correlations computed within Mplus
d 1 = Somali, 0 = other
e Coded as: 1 = doing worse financially in the U.S. than in Somalia; 2 = about the same; 3 = doing better financially in the U.S. than in Somalia
f 1 = male, 0 = female
g Number of traumatic events heard about
12345678
1. Mothers’ past traumatic eventsa–
2. Mothers’ past tortureb0.41*** –
3. Mothers’ withdrawal/detachmentc0.29*** 0.24** –
4. Mothers’ volatility/panicc0.19** 0.25** 0.77*** –
5. Mothers’ depressed moodc0.20** 0.24** 0.69*** 0.89*** –
6. Children’s experienced racismc0.10 0.07 0.32*** 0.41*** 0.49*** –
7. Children’s peer bullying victimizationc0.16* 0.20* 0.29*** 0.31*** 0.42*** 0.65*** –
8. Children’s depressive symptomsc0.03 0.02 0.14 0.26** 0.32*** 0.67*** 0.67*** –
Controls
9. Mothers’ languaged−0.16* −0.05 −0.02 0.12 −0.04 −0.04 −0.02 0.12*
10. Mothers’ education −0.12 −0.13* −0.06 −0.03 0.00 0.00 −0.18* −0.18*
11. Mothers’ change in financial statuse−0.03 0.01 0.04 −0.15* −0.22* −0.03 −0.07 −0.20*
12. Child sexf0.02 0.03 0.03 0.04 0.09 0.09 −0.14 −0.19*
13. Children’s traumatic experiencesg−0.10 0.03 0.11 −0.01 −0.03 0.22* 0.15* 0.19*
J Immigrant Minority Health
1 3
Results of the path coefficients indicated that the extent
of mothers’ past traumatic events was significantly related
to mothers’ current withdrawal/detachment symptoms;
however, mothers’ withdrawal/detachment symptoms were
not related to any of the child adjustment variables. The
extent of mothers’ past torture experiences was signifi-
cantly related to mothers’ volatility/panic symptoms which,
in turn, were related to children’s more frequent bullying
victimizations. Additionally, mothers’ past torture expe-
riences were significantly related to mothers’ depressive
symptoms which, in turn, were significantly related to all
three indices of child functioning, that is, to children’s more
frequent depressive symptoms, bullying victimizations, and
to higher levels of perceived racism.
Tests ofIndirect Effects
The results of the indirect effect tests indicated that all sig-
nificant mediating effects involved mothers’ torture experi-
ences and all but one involved mothers’ depressed mood
(Table 5). Specifically, mothers’ torture experiences were
related to children’s perceived racism, peer victimization,
and marginally, children’s depressive symptoms by way of
mothers’ depressed mood. Additionally, mothers’ volatility/
panic symptoms mediated the association between moth-
ers’ past torture and children’s bullying victimization.
Discussion
The current findings indicate that there was no direct
association between trauma of the mother and their chil-
dren’s well-being; however, mothers’ posttraumatic stress
and depressive symptoms significantly mediated the
effect of mothers’ past torture on their children’s adjust-
ment—a pattern indicative of intergenerational traumati-
zation. Thus, refugee mothers’ depressive and volatility-
panic symptoms serve as transmission mechanisms that
pass traumatization effects from mother to child. The
most widespread associations with children’s function-
ing was derived from mothers’ depressed mood which,
in turn, derived from mothers’ past torture experiences.
Certainly, experiences of torture can lead to a fundamen-
tal distrust in others and a breakdown in how the survivor
relates to the world around them [35]. Such basic distor-
tions of self, as well as intense preoccupations of hope-
lessness and worthlessness,likely impair a mother’s abil-
ity to foster a loving bond with her child [8]. Melancholic
thoughts and feelings, as well as mothers’ anhedonia can
have a devastating impact on a mother’s ability to social-
ize, respond, and interact effectively with her child, leav-
ing her child vulnerable to a wide range of adjustment
difficulties [7]. From an attachment perspective, trauma-
tized mothers suffering from depression might be unable
to invest emotionally in their child, signaling a sense
Fig. 3 Structural model
results. Standardized coef-
ficients are shown. Covariates
not shown. All within-domain
variables were correlated a
priori (not shown, p < .000).
Model had good fit: Chi square
[118] = 220.78, p < .001,
CFI = 0.922, RMSEA = 0.066,
SRMR = .049. N = 198.
*p < .05. **p < .01. ***p < .001
Mothers’ Past Trauma Mothers’ Mental Health Children’s Adjustment
Perceived
Racism
Bullying
Victimization
Depressive
Symptoms
Depressed
Mood
Withdrawn/
Detached
Torture
Experiences
Traumatic Events
.24**
.21**
.20* .40*
.76***
R2 = .10
R2 = .12
R2 = .31
R2 = .27
R2 = .17
.51**
Volatility/
Panic
.65**
R2 = .10
Table 5 Summary of indirect
effects stemming from mothers’
past torture
+p < .09. *p < .05
B (SE)
Torture → mothers’ depressed mood → children’s perceived racism 0.113* (0.054)
Torture → mothers’ depressed mood → children’s bullying victimization 0.015* (0.006)
Torture → mothers’ depressed mood → children’s depressive symptoms 0.037+(0.022)
Torture → mothers’ volatility/panic → children’s bullying victimization 0.011* (0.005)
J Immigrant Minority Health
1 3
of worthlessness and ineffectiveness in their child [36].
Given that mothers’ depressed mood was related to their
torture experiences which occurred prior to their arrival
in the U.S. (on average, approximately 14years prior),
it is possible that mothers’ depressive symptoms were
ongoing and present throughout their child’s upbring-
ing, thereby having a persistent negative impact on their
child. In all, the current findings confirm that mothers’
depressed mood is a critical mechanism mediating the
relation between mothers’ past torture and their children’s
adjustment.
It is interesting to note that mothers’ depressed mood
was more strongly related to children’s perceived racism
and bullying victimization than to children’s depressive
symptoms. We believe this pattern might be rooted in the
relatively smaller variance surrounding children’s depres-
sion scores relative to the variance surrounding scores of
children’s bullying victimization and racism (Table 3).
These variances reveal that children’s experiences of bully-
ing and racismwere more variable within the current sam-
ple than their experiences of depression. This is in line with
studies documenting the relatively low incidence of depres-
sion in children of this age [37]. It would be informative to
see this study’s conceptual model applied to older children
of refugees, when rates of depression are higher [38].
Other study findings indicate that mothers’ past torture
experiences were related to mothers’ volatile/panic behav-
iors, which in turn, were related to children’s bullying vic-
timization. It may be that mothers’ angry outbursts and
volatile, erratic behavior initiates the child’s expectation of
aggression and, over time, the child might be more likely
to interpret benign social gestures as acts of aggression and
respond by aggressing, which in turn, might elicit aggres-
sion from others [39]. Alternately, mothers’ volatility/
panic symptoms might model explosive behaviors for their
children, predisposing their child to act out aggressively
[9, 18], thereby provoking bullying from peers and class-
mates. Several studies show that bullying victims are often
themselves aggressive and provocative [40, 41]. There is
also evidence that the posttraumatic stress symptoms of
volatility/panic (which includes angry outbursts and sud-
den emotional or physical reactions) manifest themselves
as psychologically punitive parenting [18, 19]. Moreover,
refugee children from war-torn countries are known to be
exposed to violence inside the family and experience sig-
nificant parent-directed violence and anger pre- as well
as post-settlement [42]. Thus, whether modeling directly
mothers’ volatile behavior or whether mothers’ volatile
behavior alters children’s social expectations of aggres-
sion, the current findings suggest that mothers’ volatile,
unpredictable behaviors expose their child to bullying vic-
timization. Because mothers’ volatile/panic behaviors and
child victimization were measured concurrently, we cannot
establish the direction of effects. Nevertheless, the current
findings lend insight into possible maternal precursors to
peer victimization in refugee children [22].
It was somewhat surprising that mothers’ withdrawal/
detachment symptoms did not relate to any of this study’s
indicators of children’s functioning given that these symp-
toms specifically involved avoiding social activities and
being emotionally detached. In the literature, parental
withdrawal and detachment have been found to relate to
a broad array of children’s adjustment difficulties, includ-
ing depression, poor interpersonal functioning, negative
self-esteem, and issues with anger and hostility [43]. Other
research suggests that child anger and aggression result-
ing from parental withdrawal and detachment can eventu-
ally lead to peer bullying and victimization [44]. It appears,
though, that when considered jointly and simultaneously
with mothers’ volatility/panic and depressive symptoms,
these latter symptoms were more important for children’s
functioning than mothers’ withdrawal/detachment. Further
study of the varying effects of mental health symptom clus-
ters experienced by trauma victims on their significant oth-
ers is needed.
Finally, the results of the within-domain relations indi-
cate that mothers’ mental health symptoms were highly
mutually associated, as were children’s indices of func-
tioning. The relations between children’s depression and
their experiences of victimization and racism corroborates
findings of other studies that show the damaging effects of
bullying and discrimination on young refugees [24]. These
concurrent associations, though, might also reflect chil-
dren’s depression and apathy giving rise to being targeted
as a victim. In either case, it is interesting to point out that
Somali children’s experiences of racism and bullying vic-
timization were unrelated to their age, gender, family social
status, length of time in the U.S., or whether they were born
in the U.S. Only mothers’ lower education and children’s
indirect exposure to traumatic events were related to more
victimization. Future research that uncovers possible ame-
liorating factors that protect against the harmful effects of
racism and bullying among the children of refugees would
be important.
Limitations andStrengths
Certain study limitations are important for interpreting
the findings. First, as stated above, the cross-sectional
data preclude causal interpretation of our findings. Addi-
tionally, a certain degree of reciprocal causality might
exist among the variables studied and underlie the rela-
tions observed. For example, mothers’ depressed mood
might heighten children’s depressive symptoms, which in
turn might exacerbate mothers’ depression further. This
bidirectionality among the study variables is possible and
J Immigrant Minority Health
1 3
limits the conclusions one can draw about the direction of
effects. Second, mothers’ trauma and torture history were
recalled retrospectively and possibly influenced by both
memory and current mental health issues. The study also
did not involve assessments of mothers’ parenting or the
mother–child relationship, an important potential trans-
mission mechanism between mothers’ poor mental health
and children’s functioning [13, 19]. We also did not recruit
fathers into the study because we wanted to have consist-
ency focusing on mothers as the caregiver of the child and
because fathers are less likely to report trauma or mental
health symptoms due to stigma, according to the local
Somali community leaders. Finally, so as not to burden
the child participants, this study did not include measures
of children’s adjustment in several key areas of functioning
(e.g., anxiety, aggression, hyperactivity), which would have
been useful to corroborate the results of existing studies on
intergenerational traumatization effects.
Important strengths of the study are that children rated
their own adjustment and, thus, reports of children’s func-
tioning were not biased by parent perceptions. In several
studies, children’s functioning has been assessed by par-
ent report, which might likely be biased by parents’ own
traumatic experiences or mental health issues. The effect of
refugees’ trauma on their children’s mental health has been
difficult to study because in many instances, children were
themselves in refugee camps and exposed to a myriad of
trauma [31]. Within the current sample, most children were
born in the U.S. and, thus, not exposed to refugee camp liv-
ing themselves. Moreover, most children did not witness
any direct trauma and children’s experiences of hearing
about traumatic events were controlled in the current study.
Thus, the relations that emerged did so while controlling
for children’s indirect exposure to others’ traumatic expe-
riences. Additionally, although clinicians have identified
intergenerational traumatization in the children of trauma-
tized parents, much of this work has been among the chil-
dren of Holocaust survivors or the children of war veterans
[9, 11]. The current study’s focus on the intergenerational
traumatization of the children of refugees is an impor-
tant and timely addition. Finally, using community-based
workers who were fluent in the local language and culture
allowed better reporting of these sensitive topics of trauma
and poor mental health.
Conclusions
Research on refugees is relatively scarce, but tremen-
dously needed. Not only has there been a recent surge
in refugees globally, but they are a population with high
rates of trauma and mental health problems [1, 5]. This
study’s limitations notwithstanding, the current results
suggest that refugee traumatization lingers and is related
not only to the refugee’s future mental health but to that
of their children as well. As such, the psychological dis-
tress and psychosocial difficulties of the children of ref-
ugees should be considered, even in cases in which the
children themselves did not experience trauma. Mental
health services focused on refugees and their families are
needed, as well as better training and understanding of
the persistent and diffusive effects of refugee trauma.
Funding Funding associated with the conduct of this research was
provided by pilot funding from the Department of Pediatrics at the
University of California San Diego for data collection and participant
compensation purposes.
Compliance with Ethical Standards
Conflict of interest The authors declare that they have no conflict
of interest.
Ethical Approval There were human subjects involved in this
research. All procedures performed in this study were in accordance
with the ethical standards of the authors’ institutional research com-
mittee and with the 1964 Helsinki declaration and its later amend-
ments or comparable ethical standards. The University of California
San Diego Human Subjects Committee approved the study.
Informed Consent Informed consent was obtained from all indi-
vidual participants included in the study.
Human and Animal Rights This article does not contain any stud-
ies with animals performed by any of the authors.
References
1. United Nations High Commissioner for Refugees. Global
trends: forced displacements in 2015. Geneva, Switzerland;
2016.
2. Rawlence B. City of thorns: nine lives in the world’s largest
refugee camp. New York: Picador Press; 2016.
3. Jaranson JM, Butcher J, Halcon L, Johnson DR, Robertson C,
Savik K, Spring M, Westermeyer J. Somali and Oromo refu-
gees: correlates of torture and trauma history. Am J Public
Health. 2004;94:591–8.
4. Bhui K, Abdi A, Abdi M, Pereira S, Dualeh M, Robertson D,
Sathyamoorthy G, Ismail H. Traumatic events, migration char-
acteristics and psychiatric symptoms among Somali refugees.
Soc Psychiatry Psychiatr Epidemiol 2003;38:35–43.
5. Porter M, Haslam N. Predisplacement and postdisplacement
factors associated with mental health of refugees and internally
displaced persons: A meta-analysis. JAMA. 2005;294:602–12.
6. Prager J. Lost childhoods, lost generations: the intergenera-
tional transmission of trauma. J Hum Rights. 2003;2:173–81.
7. Daud A, Skoglund E, Rydelius PA. Children in families of tor-
ture victims: transgenerational transmission of parents’ traumatic
experiences to their children. Int J Soc Welf. 2005;14:23–32.
8. Scheeringa M, Zeanah CH. A relational perspective on PTSD
in early childhood. J Trauma Stress. 2001;14:799–815.
J Immigrant Minority Health
1 3
9. Dekel R, Goldblatt H. Is there intergenerational transmission of
trauma? The case of combat veterans’ children. Am J Orthopsy-
chiatry. 2008;78:281–9.
10. Sangalang CC, Vang C. Intergenerational trauma in refugee
families: a systematic review. J Immigr Minor Health, 2016.
(Online).
11. van IJzendoorn MH, Bakermans-Kranenburg JJ, Sagi-Schwartz
A. Are children of Holocaust survivors less well-adapted?
A meta-analytic investigation of secondary traumatization. J
Trauma Stress. 2013;16:459–69.
12. Daud A, af Klinteberg B, Rydelius PA. Resilience and vulner-
ability among refugee children of traumatized and non-trauma-
tized parents. Child Adolesc Psychiatr Ment. Health (London).
2008;2:1–12.
13. Van Ee, E., Kleber, R. J., & Mooren, T. (2012). War trauma
lingers on: Associations between maternal posttraumatic stress
disorder, parent–child interaction, and child development. Infant
Ment Health J 2012;3:459–468.
14. American Psychiatric Association. (2000). Diagnostic and sta-
tistical manual of mental disorders. 4th ed. Washington, DC:
Author; 2000.
15. Feng X, Shaw DS, Skuban EM, Lane T. Emotional exchange in
mother-child dyads: Stability, mutual influence, and associations
with maternal depression and child problem behavior. J Fam
Psychol. 2007;21:714–25.
16. Ruscio AM, Weathers FW, King LA, King DW. Male war-zone
veterans’ perceived relationships with children: The impact of
emotional numbing. J Trauma Stress. 2002;15:351–57.
17. Spencer JH, Le TN. Parent refugee status, immigration stressor,
and Southeast Asian youth violence. J Immigr Minor Health
(London). 2006;8:359–68.
18. Palosaari E, Punamaki RL, Qouta S, Diab M. Intergenera-
tional effects of war trauma among Palestinian families medi-
ated via psychological maltreatment. Child Abuse Neglect.
2013;37:955–68.
19. Van Ee E, Kleber RJ, Jongmans MJ, Mooren TTM, Out D.
Parental PTSD, adverse parenting and child attachment in a refu-
gee sample. Attach Hum Dev. 2016;18:273–91.
20. Hinton DE, Rasmussen A, Nou L, Pollack MH, Good M. Anger,
PTSD, and the nuclear family: a study of Cambodian refugees.
Soc Sci Med. 2009;69:1387–94.
21. Valentino K, Berkowitz S, Stover CS. Parenting behaviors
and posttraumatic symptoms in relation to children’s symp-
tomatology following a traumatic event. J Trauma Stress.
2010;23:403–7.
22. Montgomery E, Foldspang A. Discrimination, mental problems,
and social adaptation in young refugees. Eur J Public Health.
2008;18:156–61.
23. Correa-Velez I, Gifford SM, Barnett AG. Longing to belong:
Social inclusion and wellbeing among youth with refugee back-
ground in the first three years in Melbourne, Australia. Soc Sci
Med. 2010;71:1399–408.
24. Ellis BH, MacDonald HZ, Lincoln AK, Cabral HJ. Men-
tal health of Somali adolescent refugees: The role of trauma,
stress, and perceived discrimination. J Consult Clinical Psychol.
2008;76:184–93.
25. Mollica R, Caspi-Yavin Y, Bollini P, Truong T, Tor S, Lavell
J. The Harvard Trauma Questionnaire: validating a cross-cul-
tural instrument for measuring torture, trauma, and posttrau-
matic stress disorder in Indochinese refugees. J Nerv Ment Dis.
1992;180:111–6.
26. Mollica R, McDonald L, Massagli M, Silove D. Measuring
Trauma, Measuring Torture: Instructions and Guidance on the
Utilization of the Harvard Program in Refugee Trauma’s Ver-
sions of the Hopkins Symptom Checklist-25 (HSCL-25) and
the Harvard Trauma Questionnaire (HTQ). Harvard Program in
Refugee Trauma. Cambridge, MA; 2004.
27. Derogatis L, Lipman R, Rickels K, Uhlenhuth E, Covi L. The
Hopkins symptom checklist (HSCL): A self-report symptom
inventory. Behav Sci. 1974;19:1–15.
28. Kovacs M. The Children’s depression inventory-2: technical
manual. North Tonawanda: Multi-Health Systems Inc; 2011.
29. Nyborg VM, Curry JF. The impact of perceived racism: Psycho-
logical symptoms among African American boys. J Clin Child
Adolesc Psychol. 2003;32:258–66.
30. Mynard H, Joseph S. Development of the multidimensional peer-
victimization scale. Aggress Behav. 2000;26:169–78.
31. Wahlsten VS, Ahmad A, von Knorring A. Traumatic experiences
and post-traumatic stress reactions in children from Kurdistan
and Sweden. Acta Paediatr. 2001;90:563–8.
32. Solomon Z, Dekel R, Zerach G. The relationships between post-
traumatic stress symptom clusters and marital intimacy among
war veterans. J Fam Psychol. 2008;22:659–66.
33. Muthén LK, Muthén BO. Mplus user’s guide. 6th ed. Los Ange-
les: Authors; 2010.
34. Kline RB. Principles and practice of structural equation mod-
eling. 3rd ed. New York: Guilford Press; 2011.
35. Kira IA, Templin T, Lewandowski L, Clifford D, Wiencek P,
Hammad A, Mohanesh J, Al-haidar AM. The effects of torture:
two community studies. Peace Confl. 2006;12:205–28.
36. Dalgaard NT, Todd BK, Daniel SIF, Montgomery E. The trans-
mission of trauma in refugee families: Associations between
intra-family communication style, children’s attachment security
and psychosocial adjustment. Attach Hum Dev. 2016;18:69–89.
37. Hankin BL, Abramson LY, Moffitt TE, Silva PA, McGee R,
Angell KE. Development of depression from preadolescence to
young adulthood: Emerging gender differences in a 10-year lon-
gitudinal study. J Abnorm Psychol. 1998;107:128–40.
38. 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. Lan-
cet. 2012;379:250–65.
39. Crick NR, Dodge KA. A review and reformulation of social
information-processing mechanisms in children’s social adjust-
ment. Psychol Bull. 1994;115:74–101.
40. Pellegrini AD, Bartini M, Brooks F. School bullies, vic-
tims, and aggressive victims: Factors relating to group affili-
ation and victimization in early adolescence. J Educ Psychol.
1999;91:216–24.
41. Schwartz D, Dodge KA, Petit GS, Bates JE. The early socializa-
tion and adjustment of aggressive victims of bullying. Child Dev.
1997;68:665–75.
42. Bjorgo EMS, Jensen TK. Unaccompanied refugee minors’ early
life narratives of physical abuse from caregivers and teachers in
their home countries. Soc Sci Med. 2015;48:148–59.
43. Khaleque A. Perceived parental neglect, and children’s psycho-
logical maladjustment, and negative personality dispositions:
A meta-analysis of multi-cultural studies. Child Fam Stud.
2015;24:1419–28.
44. Kawabata Y, Alink LR, Tseng WL, van Ijzendoorn MH, Crick
NR. Maternal and paternal parenting styles associated with rela-
tional aggression in children and adolescents: A conceptual anal-
ysis and meta-analytic review. Dev Rev. 2011;31:240–78.
A preview of this full-text is provided by Springer Nature.
Content available from Journal of Immigrant and Minority Health
This content is subject to copyright. Terms and conditions apply.