|Home | About | Journals | Submit | Contact Us | Français|
Refugee research to date has predominantly focused on factors that make refugees more vulnerable for developing posttraumatic stress disorder (PTSD) and /or psychological distress. Few papers have studied potential protective factors such as resilience. A targeted non-random sample of Iraqi refugees (n=75) and a control group of non-Iraqi Arab immigrants (n=53) were recruited from a number of Iraqi/Arab community institutions in Michigan to complete a questionnaire that included measures for psychological distress, PTSD symptoms, exposure to trauma, and resilience. Refugees reported significantly more PTSD symptoms (T-test, p<.01) and psychological distress (p<.05) compared to immigrants. There was no difference in resilience between the two groups. In linear regression, pre-migration exposure to violence was a significant predictor of psychological distress (p<.01) and PTSD symptoms (p<.01). After controlling for migrant status and violence exposure, resilience was a significant inverse predictor of psychological distress (p<.001) but not of PTSD. Resilience is associated with less trauma-related psychological distress and should be considered in assessing risk and protective factors among victims of war-related violence.
Refugees are at a substantially increased risk for pre-migration exposure to violence rendering them more vulnerable to war- and trauma – related mental health disorders, including psychological distress, posttraumatic stress disorder (PTSD) and other symptoms (Rousseau & Depardeau, 2004; Momartin et al., 2004; Jamil et al., 2009). Studying adverse outcomes of trauma was the predominant aim of related research for a long time (Mohaupt, 2008). However, the wide range of mental health outcomes following trauma (Isakson, 2008), and the observation that overwhelmingly traumatic experiences might only result in minimal negative impact on overall daily functioning (Bonanno et al., 2011) have generated more interest in the adaptive processes that help the victims retain or resume adequate personal and social interactions (Bonanno, 2004; Agaibi et al., 2005). Thus, the concept of resilience as a protective factor against the development of trauma-related psychological disorders has attracted increasing attention recently (Yehuda, 2004; Almedom, 2008; McLaughlin, Doane, Costiuc & Feeny, 2009).
Within the framework of war and trauma, resilience is defined as personality traits that help protect against the psychological disorders resulting from exposure to terrifying incidents, such as mass violence or deportation under life-threatening circumstances; it encompasses bouncing back and positive adaptation in the face of safety-challenging experiences (Edward and Warelow, 2005; Hoge et al., 2007). In their review of resilience research among certain minority groups,McLaughlin et al. (2009) identified two characteristics that they considered fundamental to resilience: heightened vulnerability and adaptation to risk. Resilience can determine the capacity to face safety-threatening events and still perform adequately (Charney, 2004) and it explains how a victim of violence can deal positively with past traumatic experiences (Lee et al., 2008; Sossou et al., 2008).
Previous studies concerned with resilience and refugees have focused on the associations between resilience as a discrete concept and psychological symptoms. However most of those studies did not employ a specific scale for resilience. Instead, they used surrogate measures and related, but not identical, constructs including social support (Hooberman et al., 2010), sense of coherence (Ghazinour, 2003), sense of control over one’s life (Sundquist et al., 2000), and absence of mental symptoms (Turner et al., 2003; Aroian et al., 2000) to test the effect of resilience on mental health. None of these studies included a comparison group. Other studies that examined the association between traumatic exposure, resilience and psychological distress did not target refugees but community victims of acts of terrorism (Connor et al., 2003). We could not find any study that examined resilience and its association to psychological distress or PTSD among refugees from a war-afflicted country compared to a control group of immigrants from the same geographical regions that share similar cultural and social traditions, beliefs and language. Such studies are important to better understand and define the role and mechanism of resilience among refugees.
The aim of this study was to examine resilience as a potential protective factor against psychological distress and PTSD among Iraqi refugees exposed to traumatic events. In addition, Iraqi refugees were compared to a control group of non-Iraqi Arab immigrants with similar cultural, language and social traditions. We hypothesized that resilience is a protective factor which moderates the effects of trauma and is protective of psychological distress.
This study was part of a larger project examining mental health in Iraqi refugees in the United States compared to immigrants from non-war exposed Arabic countries. This study focused specifically on resilience in refugees versus immigrants and the possible attenuating effect of resilience on psychological distress and PTSD in the entire study population. Ethical approval was obtained from the Human Investigation Committee at Wayne State University.
Data were collected from a non-random targeted sample of Iraqi refugees (N=75) and non-war exposed Middle Eastern immigrants (N=53) residing in the state of Michigan in the U.S. The participants were recruited through several local community organizations in the metro Detroit tri-county area including 3 offices of the Arab and Chaldean Council (ACC), 5 offices of the Department of Human Services (DHS), the Egyptian Coptic church, the Islamic Center of Detroit and a number of other churches, mosques and cultural centers. Refugees and immigrants participated from all those sites in comparable numbers. A total of 134 individuals were asked to participate; of these 128 agreed, resulting in a response rate of 96%. A brief description of the study and inclusion criteria was provided to all prospective participants. Study descriptions were provided in both written and oral Arabic because some of the potential participants were functionally illiterate. Inclusion criteria included refugees from Iraq and non-Iraqi immigrants from four Arabic countries neighboring Iraq (Lebanon, Jordan, Egypt, Yemen) who entered the USA after 1991 at an age of 18 years or older. All participants were fluent in Arabic. Those agreeing to participate were scheduled for a meeting with representatives of the research team during which they received a detailed written and oral description of the study, were given the opportunity to pose questions, and gave their written consent. After translation of the original questionnaire by a bilingual psychiatrist, back translation was performed by a second bilingual language expert to check for any discrepancies with the original scales. In case of any discrepancies between the two, a joint meeting was arranged to agree on the proper translation. After the scales were finalized, the Arabic copies of the questionnaire were completed by the participants. Those who did not have formal education (5 refugees and 3 immigrants) received support from two bilingual specially-trained graduate research assistants. The participants with no formal education did not differ significantly from the educated participants on any of the main variables of interest including exposure to violence, psychological distress, PTSD or resilience.
Socio-demographic data included age, gender, marital status, employment, migrant status (refugee or immigrant) and years of residing in the U.S. Pre-migration exposure to violence was measured by a scale based on the Survey of Children’s Exposure to Violence (Richters & Saltzman, 1990) and modified for adult use (Berthold SM, 1999). Each participant was asked to respond on a 4–point scale (1=not at all to 4=several times) if he had experienced certain events, such as “being beaten up” before coming to the U.S. The scale yielded good internal reliability in our study (Cronbach’s alpha=.81).
Resilience was measured using an 8-item version of the resilence scale (Wagnild & Young, 1993). A sample item from this scale is, “My belief in myself helps me get through hard times.” Respondents rated on a 7-point scale the extent to which they strongly disagree (1) to strongly agree (7) with each positively stated self-description. Internal reliability for the scale was good (Cronbach's alpha=.81).
Psychological distress was measured using a 9-item scale which is a modified version of the General Health Questionnaire (GHQ, Goldberg and Williams, 1988). The GHQ’s application in research settings as a screening tool is well documented and reliability estimates for the scale in our study were good (Cronbach's alpha=.88). Respondents were asked to rate on a 4–point scale whether items such as “How often during the last month have you felt constantly under strain?” have happened not at all (1) up to much more than usual (4).
PTSD symptoms were assessed using the PTSD Checklist (PCL, Blanchard et al., 1996; Ruggiero et al., 2003). The PCL contains 17 items which measure the PTSD symptoms listed in the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders, or DSM- IV (1994). A sample item from this scale is “How much have you been bothered in the last month by repeated disturbing dreams of a stressful experience from the past?” Respondents rated on a 5–point scale how frequently this occurred (0=not at all to 4=extremely often). The total PCL score in our study demonstrated good internal consistency (Cronbach’s alpha = 0.89).
Chi-square was used to compare the two groups regarding gender, marital status and employment. All scales were examined for normality and were found to be normally distributed with the exception of pre-migration exposure to violence. Student’s t-test was used to compare the two groups with respect to age, psychological distress, PTSD symptoms, and resilience. The Mann-Whitney U-test was used to compare the groups’ median values for pre-migration exposure to violence. A 3-model linear regression was performed to identify predictors of psychological distress and PTSD symptoms, respectively. In model 1 we examined the association between the two outcome variables, respectively, and five socio-demographic variables: age, gender, migrant status, employment status, and years of living in the U.S. Pre-migration exposure to violence was entered in model 2, and resilience was added in the final model. Regression analyses were also conducted utilizing the natural log of pre-migration violence exposure and compared to regressions utilizing the continuous variable. Significance was set to a two-tailed p-value of <.05. All statistical analyses were conducted using IBM SPSS version 19.0.
Characteristics of the respondents are summarized in Table 1. The mean age of the study participants was 40.6 years (range 20–67, SD± 12.3), with no significant difference between refugees and immigrants. Refugees had been living in the U.S. a significantly shorter time than immigrants (p<.001). Fifty-nine percent of the total sample were women, with a significantly greater proportion of women in the refugee group (p<.01). There was no difference in marital status between the two groups. Unemployment was significantly higher among refugees (p< .001). Reported exposure to violence was significantly higher among refugees compared to immigrants (p<.001) and refugees also had significantly higher scores than immigrants for PTSD symptoms (p<.01) and psychological distress (p<.05). No significant difference was demonstrated between the 2 groups for resilience.
The results of the linear regression predicting psychological distress are summarized in Table 2. Migrant status was significant in model 1, i.e., being a refugee was associated with higher psychological distress than being an immigrant (p<.05). When history of exposure to violence was added to the analysis in model 2, migrant status was no longer significant but violence exposure was a significant predictor of psychological distress (p<.05). In the final model (model 3), low resilience (p<.001) and history of exposure to violence (p<.01) were the only variables that remained as significant predictors of psychological distress. This model explained 22% of the overall variance in psychological distress.
Results of the linear regression predicting PTSD symptoms are presented in Table 3. In model 1 migrant status was a predictor of higher PTSD symptoms but, as with psychological distress, that relationship became non-significant when previous exposure to violence was added in model 2. Trauma exposure was a significant predictor of PTSD symptoms in model 2, along with female gender. When resilience was added to the regression in model 3 it did not show a significant association to PTSD symptoms, although female gender and previous violence exposure remained significant predictors. This final model predicted 20% of the overall variance in PTSD symptoms. Regression analyses using the log transformed variable for pre-migration violence exposure did not alter results for predictors of psychological distress or PTSD symptoms.
This study examined a cross-sectional sample of Iraqi refugees and Arab immigrants in the U.S. to determine if resilience is a protective factor for psychological distress, a well-known mental health outcome among individuals exposed to violence (Schweitzer et al., 2007). We found that refugees, as compared to immigrants from a similar culture, reported more psychological distress and PTSD symptoms. Regardless of migrant status, pre-migration exposure to violence was a significant predictor of both psychological distress and PTSD symptoms. Resilience was a significant inverse predictor of psychological distress, but not of PTSD symptoms. Thus, our hypothesis that resilience would be protective factor against both psychological distress and PTSD symptoms was only partially confirmed.
Few previous studies have compared mental health outcomes among refugees and immigrants with similar cultural backgrounds. In line with our findings, Silove et al (1998) reported a higher level of anxiety and depression among Sri Lankan asylum seekers compared to Sri Lankan immigrants resettling in Australia. Although refugees and immigrants in the current study differed significantly with regard to psychological distress and PTSD symptoms, migrant status became non-significant when previous exposure to violence was included in the model. This supports previous studies that reported trauma exposure as a risk factor for psychological distress (Pine and Cohen, 2002). Notably, time since entry into the U.S. was not a significant predictor of either mental health outcome, even after adjusting for pre-migration trauma. This suggests that the effect of pre-migration trauma was not mitigated by the time elapsed since the event. These results are in line with findings by Marshall et al (2005), who reported high levels of psychopathology among Cambodian refugees two decades after migration. However, they contradict other previous research (Ehlers and Clark, 2003) suggesting that symptoms of PTSD would diminish with time.
In the final regression models, pre-migration trauma was a significant positive predictor of both psychological distress and PTSD symptoms. However, resilience was only a significant inverse predictor of psychological distress, not of PTSD. The items of the resilience scale reflect the positive and non-passive responses to traumatic events and participants’ high resilience was associated with lower psychological distress. The fact that we did not find a significant difference in the resilience score between the two study groups may be explained by the fact that resilience is shaped through factors and events that are fostered in the individuals’ minds during upbringing and through social norms and beliefs, many of which are mutually shared by the two study groups. Holtz (1998) compared Tibetan refugees exposed to torture with a control group of Tibetans without a history of torture and found significantly higher anxiety scores among the torture-exposed. Holtz (1998) concluded that a number of factors such as commitment, spirituality and preparedness foster resilience against psychological distress. Unlike our study, Holtz did not use a standardized resilience scale but inferred resilience from responses to open-ended questions.
Among the socio-demographic variables, unemployment - which has been reported to be significantly associated with psychological distress (Jackson et al., 1983; Kokko and Pulkkinen, 1998) - was more prevalent among refugees in our study but was not a significant predictor of either psychological distress or PTSD symptoms. Gender was a significant predictor of PTSD, with females at increased risk in models 2 and 3. Several previous studies have reported a higher prevalence of PTSD in females. Breslau et al (1997) studied a random sample of 1007 young adults drawn from a 400 000-member health maintenance organization (HMO) in Southeast Michigan and reported a higher prevalence of PTSD among women than men despite the fact that exposure to traumatic events did not vary between the 2 sexes. In a convenience sample of 325 refugees from all over the world enrolled in a program for survivors of torture, women also reported significantly higher levels of PTSD (Keller et al., 2006). However, neither of these studies controlled for resilience as was the case in our study. The gender difference in our study might be explained by the submissive social role assumed by many females in Iraq and other Arab countries leading to a higher exposure to trauma in the form of physical and psychological abuse (Douki et al., 2003; Barkho et al., 2011). Intimate partner abuse has been associated with higher rates of PTSD among females (Golding 1999; Dutton et al., 2006).
One of the limitations of this study was that the recruitment process targeted specific social service institutions, such as the Department of Human Services (DHS). This might result in a selection bias between the study groups. However, the large number of different sites for recruitment would tend to limit any systematic bias. Every refugee is instructed by the resettlement agencies to visit the DHS offices to apply for the services provided by those agencies, while only a small percentage of the immigrants would be eligible for the DHS services. Immigrants are expected to be supported by relative-sponsors and those immigrants attending the DHS offices, especially in the early years of coming to the U.S., would only be those with low income due to unemployment. However, regression analyses controlled for socio-demographic variables, including employment status.
Another possible selection bias would result from recruiting participants through churches and mosques, thus increasing the chance of recruiting the more religious individuals in both study groups. However, participants were also recruited from non-religious organizations, and refugees and immigrants were recruited in equal numbers from all institutions.
One of the limitations in attempting to associate PTSD symptoms with the history of exposure to violence is that the pre-migration violence scale does not inquire about the period that had elapsed between exposure and the interview. Symptoms of PTSD might improve over time with a steep decline in the first year after the exposure (Ehlers and Clark, 2003). In our study, 64% of the refugees had been in the U.S. two years or longer. The retrospective recall of traumas and symptoms might be subject to event reconstruction. However, time in the U.S. was not a significant predictor of either PTSD or psychological distress, while pre-migration exposure to violence was a predictor of both.
For unavailability of an alternative, our study employed a Western resilience scale that might not be culturally-sensitive to Iraqis and Arabs as it had not been tested previously on Arabic-speaking subjects. This might emphasize the need for standardizing the scales used to study non-English speaking ethnic groups instead of depending solely on expert translation. However, this cannot explain findings of no differences between refugees and immigrants and the overall finding that resilience was protective.
Another important consideration is that certain questions in the trauma scale (specifically the questions about being raped or forced to engage in sexual activities to survive) are considered taboo; even if participants were exposed to such traumas, they might not be willing to admit them. Only one refugee and one immigrant admitted being raped before coming to the United States. This would bias the score of the trauma scale towards lower exposure.
Finally, this study compared groups that differ on two variables, i.e., migrant status (refugee vs. immigrant) as well as nationality (Iraqi vs. non-Iraqi), which might confound results. However, migrant status was not a significant predictor of either mental health outcome once exposure to trauma was entered into the respective equations. Moreover, many cultural and social traditions and health beliefs are similar between these nationalities. This indicates that violence and trauma exposure played a more significant role in mental health outcomes than either migrant status or nationality.
To the best of our knowledge, this is the first study that compared the protective effect of resilience against psychological distress between refugees and culturally–similar immigrants in the same post-settlement country. Our findings supported the hypothesis that resilience is a protective factor against psychological distress, but not preferentially so in refugees as compared to immigrants. More research is needed to study factors related to resilience and different variables that augment an individual’s potential to rebound after trauma. A resilience-oriented rather than a symptom-oriented approach, putting more emphasis on studying the protective and recovery-fostering individual assets rather than focusing on illness-expectancy is what is needed with the large numbers of re-settlers arriving from conflict and unstable zones. We recommend studying resilience in prospective research of refugees and immigrant populations who are also likely to face large number of post-migration traumas. Such studies would minimize recall bias that is encountered in retrospective research.
All authors were partly supported by award number R01MH085793 (Principal Investigator B. Arnetz) from the National Institute of Mental Health. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Mental Health or the National Institutes of Health.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflict of Interest
The authors declare no conflicts of interest.