Emotional and behavioural problems amongst Afghan unaccompanied asylum-seeking children: results from a large-scale cross-sectional study
- First Online:
- Cite this article as:
- Bronstein, I., Montgomery, P. & Ott, E. Eur Child Adolesc Psychiatry (2013) 22: 285. doi:10.1007/s00787-012-0344-z
- 917 Views
Unaccompanied asylum-seeking children (UASC) are considered at high risk for mental health problems, yet few studies focus on single ethnic populations. This study presents results from the largest Afghan UASC mental health survey in the UK. Specifically, the study aims to estimate the prevalence of emotional and behavioural problems and to investigate the associations of these problems with demographic variables, cumulative traumatic events, and care and migration variables. A census sample of 222 Afghan UASC was interviewed using validated self-report screening measures. Emotional and behavioural problems were screened using the Hopkins Symptoms Checklist 37A (HSCL-37A). Pre-migration stressful life events were screened using the Stressful Life Events Questionnaire. Administrative data on care and asylum were provided by the local authority social services and the UK Border Agency. Approximately one-third (31.4 %) scored above cut-offs for emotional and behavioural problems, 34.6 % for anxiety and 23.4 % for depression. Ordinary least squares regression indicated a significant dose–response relationship between total pre-migration traumatic events and distress as well as between increased time in the country and greater behavioural problems. Compound traumatic events in the pre-migration stages of forced migration have a deleterious association with UASC well-being. Increased time in country suggests a possible peer effect for these children. Consistent with other studies on refugee children, it should be stressed that the majority of UASC scored below suggested cut-offs, thus displaying a marked resilience despite the experience of adverse events.
KeywordsRefugee childrenMental healthAsylumAfghan
The United Nations High Commissioner for Refugees (UNHCR) reported in 2010 that there were approximately 43 million forcibly displaced persons . A small but significant number arrives in Western countries including the United States, Australia, and members of the European Union [2, 3]. One particularly vulnerable subset is unaccompanied asylum-seeking children (UASC) who have been separated from their parent or primary care giver and claim asylum alone.
UASC may experience war, the death of family members, persecution, violence, rape, escape from forced recruitment into military or paramilitary organisations, and forced domestic labour . It is these experiences, coupled with other developmental stressors from childhood and adolescence, which increase the likelihood of mental health problems for UASC [5–7]. Indeed, the literature concerning UASC indicates high levels of psychological distress in the form of emotional and behavioural problems and PTSD . There is variability in scores across studies on refugee children; however, that is likely due to contextual differences such as countries of origin, country of asylum, as well as the measurements used, thus limiting generalisation [9, 10].
Afghans have been the largest group of all forced migrants for the last three decades . During this time, the Afghan population experienced chronic conflict [17–19]. Several studies reporting on children in Afghanistan indicate a high likelihood that the children experience violent conflict related to war, massacres, torture, the loss of loved ones, and destruction of homes [20–22]. In addition, the same studies report violence irrespective of the military conflict. This includes being slapped, shouted at, hit with an object, punched, kicked, hair pulled, burned, threatened verbally, family member being hit, belittled, and/or ridiculed .
There is emerging evidence concerning the migration experiences of Afghan UASC to Europe. A recent study which interviewed 150 male Afghan UASC in the UK, reported journeys that lasted approximately 6 months, with almost all of the children travelling overland . Experiences included sleeping rough, living in cramped and unsanitary conditions, physical abuse by smugglers and government officials, witnessing the death of friends, and living in near-constant privation. A smaller report of 20 male Afghan UASC in the UK and Norway reported experiences of living in containers without food or water and seeing friends suffocated to death .
Approximately 30,000 UASC arrived in the UK in the decade 2000–2010 . Consistent with worldwide trends, the largest group comes from Afghanistan. The number of Afghan UASC has increased in total percentage from 10.5 % in 2004 to 51.0 % by the end of 2009 .
Two aspects of the UASC UK context are the care framework and the immigration and asylum framework. UASC in the UK are cared for by local authority social services, under mandate from the Children Act 1989 [26, 27]. This framework provides a package of care including the provision of living arrangements, education, access to health professionals, and social work support. Studies indicate that UASC who enter the UK at a younger age are more likely to be placed in foster care [27–29]. UASC over the age of 16 are more likely to be placed in semi-independent accommodation [27, 30].
The child’s claim for asylum is reviewed by the immigration and asylum system . Most UASC receive only a temporary status to remain in the country (discretionary leave, DL) and face formal status proceedings at age 17.5. For example, of the 2,359 initial decisions on UASC asylum applications in 2010, 68 % were granted DL . Many UASC face the end of their proceedings around age 18—often choosing between destitution and returning to their country . Both the care framework and legal asylum system, therefore, have an important role in the lives of these young people and may affect mental health well-being. There is evidence to suggest that the level of support provided within living arrangements may be associated with UASC distress [29, 34, 35]. While not UK specific, the temporary nature of the asylum system has also been indicated as leading to greater distress among asylum-seekers .
This study is the first to investigate the prevalence of emotional and behavioural problems amongst Afghan unaccompanied asylum-seeking adolescents in the UK. The objectives are (1) to provide an estimate concerning the prevalence of emotional and behavioural problems for Afghan UASC in the UK and (2) to investigate the relationship of these problems with demographic variables, the accumulation of pre-migration traumatic events, and post-migration contextual factors concerning care and migration.
Study design and settings
Ethical approval for the study was provided by the Oxford University Central University Research Ethics Committee (CUREC) and the local authority. This paper is compliant with the reporting guidelines from the STROBE statement .
Demographic and care data were collected from the local authority databases. Asylum data were collected from the Home Office databases. Specific care measures applied in this study include the type of UASC living arrangements and the total number of placements. Foster care is defined as placement within a foster family. Semi-independent care is defined as shared living accommodation with other UASC and support provided by care workers. Two other forms of living arrangements include bed and breakfast placements and family placements. The former is used as an emergency type of housing immediately after UASC arrive. The latter refers to private fostering arrangements. The immigration and asylum variable reflects the amount of days remaining from the start of the data collection until the expiration of discretionary leave (asylum status expiration time—ASET).
The Hopkins Symptoms Check List 37A questionnaire was used to screen for emotional and behavioural problems [HSCL; 39]. This is a validated, 37-item, self-report screening questionnaire developed for UASC and used to screen for symptoms of anxiety, depression, and behavioural problems. The HSCL-37A is a modified version of the original Hopkins Symptom’s Check List 25 . The HSCL-37A was designed and validated as a reliable measure specifically for UASC. Items on the HSCL-37A are scored on a Likert scale of 1–4 from least (1) to greatest (4). Higher scores indicate greater difficulties. Suggested cut-off scores for caseness are: total score (69.00), anxiety (20.00), and depression (33.21). No cut-off has been suggested for externalising scores. The psychometric quality of the HSCL-37A has been previously indicated as good , and the Cronbach’s alpha values obtained in this study were also good (anxiety 0.81; depression 0.82; externalising problems 0.84).
Traumatic events for Afghan UASC were obtained using the Stressful Life Events Questionnaire [SLE; 42]. The SLE contains 12 ‘yes’ or ‘no’ questions concerning potential traumatic events to do with loss, conflict, and violence that the UASC may have experienced in the pre-migration stages of their forced migration. The SLE was specifically designed for the reading level of adolescents and developed and piloted with UASC from a range of different cultural backgrounds . The SLE presented to all UASC participating in the study with a high item response, with most questions exceeding a 99 % response rate (total response rate 98.1 %).
Questionnaire completion took approximately 60–90 min, depending on the reading ability of the young person. Young people were provided with refreshments throughout the data collection. All questionnaires underwent a double-blind back translation into Dari and Pashto.
Non-parametric statistical tests were performed for all analyses conforming to the suggested procedures for non-normal data . Dependent variables were the HSCL total score and different subscores. Descriptive analyses were performed to check for demographic and care differences between participant and non-participant groups using χ2 and Mann–Whitney (U) tests. Bivariate analyses for linear independent variables (age, time in country, total number of traumatic events, number of placements, and ASET were performed using non-parametric Spearman’s rho (ρ) correlation coefficient and for categorical variables (language, type of living arrangement) using ANOVA . An ordinary least squares (OLS) regressions were performed to investigate the strongest predictors for HSCL scores including the total score, the internalising score, anxiety and depression sub-scores, and the externalising scores. The significant variables from the bivariate analyses were entered into the model in one of three steps (demographic, pre-migration, post-migration) where time in country was considered a demographic variable. A value of p < 0.05 was considered statistically significant for regression analyses. All analyses were conducted using SPSS version 18 for Mac.
Demographic and social characteristics
Demographic, asylum, and care characteristics of the sample
Total Group (N = 326)
Participants (N = 222) (68 %)
(N = 104) (32 %)
Age, mean (SD)
U = 11,183.0, ns
231 (70.9 %)
155 (69.8 %)
76 (73.1 %)
χ2 = 0.237, ns
94 (28.5 %)
66 (29.7 %)
28 (26.9 %)
1 (0.6 %)
1 (0.5 %)
Time in country
Mean in days (SD)
U = 10,543.40, ns
Asylum status expiration time (ASET)
N = 255
N = 173
N = 82
Mean in days (SD)
U = 6,419.00, ns
0 to −1,948
0 to −1,218
0 to −1,948
χ2 = 0.393, ns
200 (61.3 %)
139 (62.6 %)
61 (58.7 %)
115 (35.3 %)
76 (34.2 %)
39 (37.5 %)
Bed and breakfast
9 (2.8 %)
7 (3.2 %)
2 (1.9 %)
2 (2.2 %)
0 (0.0 %)
2 (1.9 %)
Mean number of placements
U = 11,064.0, ns
The mean number pre-migration stressful life events were 6.6 (SD 2.7). An item-by-item breakdown is reported elsewhere . There were no correlations or group differences between any of the independent variables and the total number of stressful life events.
Emotional and behavioural problems
Bivariate associations for HSCL 37A total score and sub-scores with independent variables
Total number of traumatic events
Time in country
Type of living arrangement
Number of total placements
HSCL-37A total Score
F(1,216) = 0.556
F(1,218) = 5.445*
F(2,217) = 0.286
F(1,218) = 6.329*
F(2,214) = 0.076
F(1,218) = 3.850****
F(2,211) = 0.586
F(1,212) = 5.682**
F(2,219) = 0.784
F(1,220) = 0.001
HSCL total score
The regression model accounts for 17.2 % of the total variance (r2 = 0.172, adjusted r2 = 0.161). Age, results from the stressful life events questionnaire (SLE), and the type of living arrangement for the UASC appear to explain nearly one-fifth of the variance of the emotional and behavioural problems. The only predictor of the HSCL total score is SLE which accounts for 86 % of the variance in the model, or 15 % of the total variance for the total score.
The regression model accounts for 20.4 % of the total variance (r2 = 0.204, adjusted r2 = 0.193). Age, SLE, and the type of living arrangement appear to explain one-fifth of the variance of the emotional problems as determined by the HSCL-37A internalising score. SLE is the only variable to remain significant when all other variables are held constant, accounting for 85 % of the variance in the model, or 17 % of the total variance within the HSCL internalising score.
The regression model accounts for 16.1 % of the total variance (r2 = 0.161, adjusted r2 = 0.149). Age, SLE, and the type of living arrangement appear to explain 16 % of the variance of the anxiety scores reported by UASC. SLE was the only variable to remain significant when all other variables are held constant, accounting for 88 % of the variance in the model, or 14 % of the total variance within the HSCL anxiety score.
The regression model accounts for 18.8 % of the total variance (r2 = 0.188, adjusted r2 = 0.177). Age, SLE, and the type of living arrangement explain nearly one-fifth (19 %) of the variance of the depression scores reported by the UASC. SLE was the only variable to remain significant when all other variables are held constant. SLE accounts for 85 % of the variance in the model, or 16 % of the total variance within the HSCL depression score. Statistical analyses for interactions between age and living arrangements were considered in regressions for HSCL total, internalising, depression, anxiety scores. The improvement in the model fit was not statistically significant; the models without the interaction fit the data as well as the model with the interactions.
Ordinary least square regressions for HSCL 37A total score and sub-scores with independent variables
Total Adj. R2
Standardised B (for final step)
HSCL total score
HSCL internalising score
Time in country
Approximately one-third of the sample reported scores above the suggested cut-off for the HSCL-37A. Broken down into the sub-scales of anxiety and depression, the prevalence estimates are 35 and 23 %, respectively. These results reflect UASC trends for emotional and behavioural problems from the wider literature [for a review see: 8], but the generalisability of the results of this study must be cautiously discussed given the specific ecological model contexts of Afghan UASC in the UK.
Two variables had a significant association with emotional and behavioural problems: pre-migration stressful events and the time in country. The former was associated positively with emotional and behavioural problems including all sub-scales of anxiety, depression, and behavioural problems. This is unsurprising and consistent with previous evidence concerning refugee and asylum-seeking populations [47–49].
An increase in the time in the UK was associated with a greater level of behavioural problems. Another study has previously reported a similar result . In both studies, the length of the time in country was the more robust predictor of behavioural problems compared to pre-migration stressful life events. What time in the UK means for UASC is a question that demands greater exploration. One starting point of the ‘temporal’ trend may be the arrival of the UASC in the country. In the initial stages after arrival, the child is likely to be alone and in need of basic safety and care. As time passes, however, they may settle and acculturate; they will acquire language, norms, and build strong social networks through the care or educational networks [51–53]. These may be expected to be associated with positive outcomes; however, the passage of time in this study indicated poorer behavioural outcomes. Conduct problems may be due to several mechanisms. One may be increased closeness with peers (i.e. peer effect), where there is the possibility that the UASC learn patterns of deviant behaviour from other children in the community [6, 54, 55].
Another mechanism for acting out may emanate from the liminal situation with regard to staying in the UK. Specific to the UK context, the passage of time marks an approach towards the date when the young person will be faced with the threat of return to his or her country of origin, an event that may be out of their control. This places the young person within a liminal position and presents confusing messages. One message, that of the care framework, provides physical needs such as shelter, a sense of protection, and, arguably, belonging. The contrary message from the asylum and immigration framework presents an uncertain position in the country and perhaps feelings of being unwanted, uncertainty towards the future, and fear of another forced migration—this time back to their country of origin [56, 57]. For the latter, UASC lack power in the asylum decision processes, which may bring about behavioural problems in what has been described as planful behaviour as a coping mechanism, increased conduct problems associated with the desire to gain control over life . Similarly, levels of confidence, and mental well-being may become more salient as the immediate feelings of relief recede. Interestingly, the effect of time until status determination was non-significant on the HSCL-37A externalizing score, weakening the theory of liminality increasing stress and problem behaviours. A sense of discomfort, uncertainty, and liminality may not be directly correlated with the legal status determination itself and may be more strongly correlated to a heightened sense of awareness from time in the country.
Comparison of HSCL-37A across studies
Oxford study (2010) (N = 222)
HSCL-37A total score (SD)
HSCL-37A internalising score (SD)
HSCL-37A anxiety score (SD)
HSCL-37A depression score (SD)
HSCL-37A externalising score (SD)
The cross-sectional nature of this research limits any causal inferences from the results . Cross-sectional studies do provide information concerning problems within a community; however, and the large sample size here provides a strong indication of the levels of problems within this community. Similar to other studies concerning UASC, it is not possible to confirm the historical accuracy of the stressful life events of these children . It should also be noted that those who did not participate in the research may be more or less vulnerable in comparison to those who participated although no significant differences on key demographic variables were found. A further limitation with this population concerns the cultural mediation and translation of mental health concepts. Concepts and ideas may not be transferable across cultures, and translated measures may thus result in decreased validity of mental health measures [10, 61]. This study made use of HSCL 37A, which is arguably the most appropriate and valid instrument for UASC mental health. It was developed specifically and validated for this population.
Social desirability bias reflects an over- or under-reporting on a questionnaire based upon a desire to maximise gains . For example, UASC may perceive that their answers could result in a positive financial, social, or immigration status gain. Steps taken to minimise social desirability bias include the use of detailed information sheets describing the study and how answers would remain confidential, alongside self-report methods. The latter allowed the UASC to answer without feeling pressured.
This study focused solely on self-report methods in the collection of information concerning mental health and behavioural problems, thus limiting the results to the subjective observations of the young people. Arguably, multi-informant methods and triangulation of sources are considered the best practice for collecting data . Cultural differences and frequency of interaction between adults such as social workers and foster carers may give rise to doubts on the consistency and accuracy of their interpretation of forms of distress. In this context, self-reports may be the most valid source of information .
The sample set used for this research came from one London borough, and therefore the results of the research should be generalised with a degree of caution. Although the research was able to access the largest sample of Afghan UASC in the UK (and Europe), this population may differ both in internal composition (e.g. ethnic groupings) and in external experiences (e.g. caseworkers and structural support) from other Afghan UASC in Europe and from other refugee populations. Indeed, it should be highlighted that the term Afghan includes several different ethnic, religious, linguistic, and social groupings . It was not possible within the framework of this research to fully explore how these other variables may interact with resilience and distress, e.g. the Conservation of Resources Theory .
This research presents the first strong estimate concerning emotional and behavioural problems for male Afghan UASC in the UK. Approximately one-third of Afghan UASC scored above suggested cut-off for emotional and behavioural problems on the HSCL-37A. The compounding of traumatic events in the pre-migration stages of forced migration has a deleterious association on UASC well-being. Importantly, this only accounts for one-fifth of the variance in emotional and behavioural problems, suggesting that there are other items that need to be identified. Some of these items may include displacement and non-displacement factors, which act both as risk or protective variables [9, 67]. The findings concerning time in country suggest that UASC externalise the longer they are in the country. This may be due to peer effect, or perhaps seeking to control what may be a situation beyond their control. Finally, it must be stressed that while over 30 % of the Afghan UASC in this study scored above suggested cut-offs for emotional and behavioural problems, the majority of children did not. In other words, Afghan UASC display a marked resilience considering the amount of adverse events they have experienced. Further research should explore UASC well-being via more resilience-based paradigms.
This study was funded by the John Fell Oxford University Fund Unaccompanied Asylum-Seeking Children Project. A special debt of gratitude goes to Lindsay Shepard for all her care and assistance with the data collection, to the dedicated team of social workers and interpreters, and to all the young people.
Conflict of interest
The authors declare that they have no conflict of interest.