Mental health of asylum seekers

Research suggests that asylum seekers and displaced persons worldwide report high rates of pre-migration trauma, ^{Reference Sinnerbrink, Silove, Field, Steel and Manicavasagar4} and therefore of trauma-related mental health problems. In a meta-analysis of worldwide studies investigating the mental health of refugees (including asylum seekers and displaced persons), Porter & Haslam found high rates of psychopathological disorder among refugees worldwide compared with non-refugee control groups. ^{Reference Porter and Haslam5} There is therefore consistent evidence to suggest that asylum seekers and refugees have higher rates of mental health difficulties than are usually found within the general population. The process of seeking asylum in Western countries places additional demands on this group. These include stressful legal processes. In an Australian study comparing post-migratory stress in refugees, asylum seekers and immigrants, Silove et al showed that the ‘refugee determination process’ (including interviews by immigration officials) was regarded as stressful by asylum seekers. ^{Reference Silove, Steel, McGorry and Mohan6} In addition, a Dutch study showed that longer asylum processes result in increased risk of psychiatric disorder. ^{Reference Laban, Gernaat, Komproe, Schreuders and De Jong7} The authors also reported increased anxiety, depression and somatoform disorders in individuals who had lived in The Netherlands for more than 2 years compared with refugees who had arrived within the preceding 6 months. Consistent with these findings was the observation of high rates of post-traumatic stress disorder (PTSD) symptoms in both groups. Post-migratory stressors seem, therefore, to be negatively affecting this population, who are already vulnerable to mental health difficulties as a result of their previous exposure to traumatic events.

Asylum seekers in detention

Asylum seekers who are detained in the host country experience a further and more specific set of stressors, reflecting the detention process itself and the detention centre environment, which may adversely affect their mental health status. These include loss of liberty, uncertainty regarding return to country of origin, social isolation, abuse from staff, riots, forceful removal, hunger strikes and self-harm. ^{Reference Fazel and Silove8–Reference Keller, Rosenfeld, Trinh-Shvrin, Meserve, Sachs and Leviss10} Given the well-documented vulnerability of asylum seekers as a result of experience of trauma prior to arrival, a number of clinicians have expressed concern that detention increases mental health difficulties in adult and child asylum seekers, and have called for an end to such practices. ^{Reference Salinsky11–Reference Fazel and Stein13} This conflicts with current government policy aimed at reducing numbers of asylum seekers. ^{Reference Silove, Steel and Watters2} The practice of detaining asylum seekers is therefore currently a prominent and contentious issue in terms of policy, health and social care. These issues are outlined in more depth by Steel & Silove. ^{Reference Steel and Silove14}

The aim of this review was to identify studies that have investigated the impact of immigration detention on the mental health of detainees held in Australia, the UK and the USA. Studies that investigated the impact of detention on children and adolescents were included in the review. The results of these studies were reviewed in order to consider whether there is evidence for an association between increased prevalence and severity of mental health problems and immigration detention.

Search strategy

The search terms ‘asylum’, ‘detention’, ‘depression’, ‘anxiety’ and ‘PTSD’ were used to identify relevant studies in the databases PsycINFO, PubMed, PubCentral and PILOTS, with a cut-off date of April 2007. In addition, manual searches of the reference lists of relevant studies were used to identify further relevant studies.

Case series

Bracken & Gorst-Unsworth and Arnold et al have both described opportunistic case series of asylum detainees in the UK. ^{Reference Bracken and Gorst-Unsworth15,Reference Arnold, Beeks, Fluxman, Katona and de Zulueta16} They described therapeutic and assessment work undertaken by clinicians within organisations working with refugees and asylum seekers: the Medical Foundation for the Care of Victims of Torture (www.torturecare.org.uk) and Medical Justice (www.medicaljustice.org.uk). Although the studies took place 15 years apart, both describe manifest psychological or psychiatric difficulties experienced by detainees with whom the authors had worked. Bracken & Gorst-Unsworth described the evidence leading to their clinical conclusion that there was a high level of psychological disturbance and a consistent pattern of symptoms in each of ten detainees, all of whom had previously experienced torture. ^{Reference Bracken and Gorst-Unsworth15} This included intense fear and anxiety, sleep disturbance and nightmares, irritability and frustration, as well as profound hopelessness and concerns about their own mental health. The authors did not report the use of any standardised ratings, which compromises the validity of their results. They described six of the individuals as expressing anxiety regarding their mental health, and all ten reporting depressed mood, appetite loss and other physical problems. Four detainees had suicidal ideation, two of whom had made actual suicide attempts. Although these qualitative impressions are useful in indicating emotional problems experienced by detained persons, a detailed case study is provided for only one individual and the extent to which these experiences are generalisable to other detainees is not addressed.

More recently, Arnold et al reported the results of the medical assessment of 56 consecutive patients who were either in detention or had been recently released: 59% fulfilled ICD–10 criteria for PTSD or depression. ^{Reference Arnold, Beeks, Fluxman, Katona and de Zulueta16} Adherence to a specified diagnostic tool increases the validity of these findings but it is unclear how representative their sample is of the asylum detention population as a whole. These reports of high levels of mental health problems among detainees are consistent with a participant–observer account reported by Sultan & O'Sullivan, an Iraqi doctor who had been detained in an Australian immigration centre and a clinical psychologist working at the same facility. ^{Reference Sultan and O'Sullivan17} These authors suggested that psychological difficulties observed among detainees increase through successive stages, triggered by negative outcomes on asylum decisions. Their descriptive reports are supported by a survey of 33 detainees (predominantly men), using a semi-structured interview. ^{Reference Sultan and O'Sullivan17} Individuals who had been detained for over 9 months were invited to participate; the findings represent 89% of this target population and are therefore highly representative of this subsample of detainees. Among those interviewed, 85% reported chronic depressive symptoms, 65% reported suicidal ideation, 39% were experiencing paranoid delusions and 21% showed signs of psychosis. In addition, 57% of participants required psychotropic medication. Information regarding the nature of the measures used to rate symptoms was not provided.

The findings from these case series provide clear documentary evidence of serious mental health problems among detainees. However, as collections of case studies they are subject to selection bias and may not be representative of detainees, many of whom do not come into contact with these specialist organisations. In addition, the association between detention experience and poor mental health is confounded by high rates of previous trauma exposure in the detainees studied. The extent to which detention can be described as having an independent negative impact on mental health is therefore not fully clarified by these reports. However, the high rate of disorder reported compared with the general refugee literature and the description of increasing symptom severity with increasing time in detention at least suggest an independent or confounding role for the detention process. ^{Reference Sultan and O'Sullivan17}

Systematic studies

More systematic methods were used in a small number of studies, mainly carried out in Australia. Thompson et al compared a group of 25 detained Tamil asylum seekers from Sri Lanka with a community-based group of Tamil asylum seekers. ^{Reference Thompson, McGorry, Silove and Steel18} They also compared findings with those in non-detained refugees and migrants from the same ethnic background. The comparison group data came from a previous study in which levels of trauma and symptoms had been investigated. ^{Reference Steel, Silove, McGorry, Mohan, Silove and Steel19} The authors used self-report measures commonly applied in research with asylum seekers and refugees: the Harvard Trauma Questionnaire, ^{Reference Mollica, Caspi-Yavin, Bollini, Truong, Tor and Lavelle20} the Depression, Anxiety and Somatisation scales of the Hopkins Symptom Checklist–25, ^{Reference Mollica, Wyshak, de Marneffe, Khoun and Lavelle21} as well as the Four Measures of Mental Health Panic Scale, ^{Reference Beiser and Fleming22} and a measure of exposure to post-migration stressors designed for the study. Respondents in the detainees group were more depressed and suicidal, and experienced more extreme post-traumatic panic and anxiety as well as increased somatic distress. An additional important finding was that detained participants had been exposed to a greater number of trauma experiences. However, levels of pre-migration trauma did not account entirely for the differences. The authors concluded that this suggests that the conditions of detention also contributed to the mental health difficulties experienced by detainees. Although the descriptive data are useful in demonstrating that detention was associated with increased incidence of mental health difficulties on all measures, it is difficult to draw conclusions from this in the absence of any details regarding the statistical techniques used or the results of multivariate analyses.

In a study based on a population of Mandaean refugees in Australia, Steel et al found further evidence for a relationship between detention and mental health problems. ^{Reference Steel, Silove, Brooks, Momartin, Alzuhairi and Susljik23} Using snowball sampling techniques, the authors recruited 241 participants (constituting an estimated 60% of the Mandaean refugee population living in Sydney) and assessed the prevalence of PTSD and of major depressive disorder as well as stress factors which were related to past trauma, detention and temporary protection. Clinical measures included the Harvard Trauma Questionnaire and the Hopkins Symptom Checklist–25. The mental and physical health component summaries of the Medical Outcomes Study – Short Form were used to measure mental and physical health status and disability. ^{Reference Gandek, Ware, Aaronson, Apolone, Bjorner and Brazier24} The authors used the Detention Symptom Checklist to assess adverse effects of particular experiences associated with detention. ^{Reference Steel, Momartin, Bateman, Hafshejani and Silove25} The authors also compared the effects of being granted a temporary protection visa v. a permanent protection visa. Temporary protection is a less secure status, requiring the holder to demonstrate a need for continued protection every 3–5 years. Further constraints applied to temporary protection visa holders include restricted access to employment and educational opportunities. ^{Reference Momartin, Steel, Coello, Aroche, Silove and Brooks26} Within this sample, 58% held temporary visas, whereas 42% held permanent visas; 62% of the sample had a history of asylum detention and had been released, on average, 3 years previously.

Using multivariate analyses, the authors investigated whether there was an independent effect of detention and of temporary v. permanent protection visa status. Time in immigration detention and temporary protection contributed independently to all three measures of mental health difficulties (P<0.05), even when other significant risk factors such as trauma history, family composition, age, gender and length of residency were taken into account.

Additional univariate analyses within the group who had experienced detention showed greater levels of traumatic distress related specifically to past detention among those who had been detained for longer than 6 months compared with those who had been detained for shorter periods. A higher proportion of those who had been detained in excess of 6 months met diagnostic cut-offs for PTSD, depression and moderate to severe mental health-related disability than those who had been detained for shorter periods or who had not been detained. A parallel study by the same group investigated the same factors in Persian-speaking refugees. ^{Reference Momartin, Steel, Coello, Aroche, Silove and Brooks26} All former detainees rated detention experiences as causing serious or very serious stress. These included fears of being sent home, being told they should return home by officers in detention facilities and language difficulties (reported as causing serious or very serious stress by 95%). In this study, 42% of participants had temporary protection visa (TPV) status and 58% held permanent protection visas (PPVs). Temporary protection visa status included previous experience of detention for an average of 1 year, whereas holders of a PPV had not experienced detention. Univariate analyses using two-sample t-tests showed significantly higher scores in TPV holders compared with PPV holders on anxiety (P<0.001), depression (P<0.001), PTSD (P<0.001) and general distress (P=0.003) (as measured by the General Health Questionnaire ^{Reference Goldberg and Williams27} ). These differences were not attributable to experience of pre-migration trauma.

Again, multivariate analyses were used to test whether post-migration factors had an independent adverse effect. Holding a TPV was the most significant predictor of PTSD. In order to explore further the negative impact of post-migration stress, the number of negative detention experiences, the number of current living difficulties and current separation from spouse replaced TPV status in a second multivariate model, as the combination of these stressors were shown to have ‘almost perfect collinearity’ with TPV status. This second model was used to analyse the independent effects of these stressors on measures of mental health and disability, while adjusting for the effects of age, gender and pre-migration trauma experiences. Of particular interest in this review is the independent impact of detention. Past detention stressors predicted PTSD (β=0.47, P<0.001) and mental health-related disability (β=0.30, P<0.001), but did not predict anxiety, depression, general distress and physical health-related disability.

These studies are the first to investigate the mental health implications of detention followed by temporary compared with permanent protection. They suggest that the combination of the detention experience and continued immigration status uncertainty following release is particularly harmful to mental health. The study by Steel et al shows that the damaging effects of detention persist following release. ^{Reference Steel, Silove, Brooks, Momartin, Alzuhairi and Susljik23} The use of measures that have been used widely in refugee research in addition to measures specifically adapted for detention experiences increases the validity of findings. Administration of interviews through one native Persian speaker maximised transcultural validity. ^{Reference Momartin, Steel, Coello, Aroche, Silove and Brooks26} The use of snowball sampling techniques resulted in recruitment of a large proportion of the target populations, ^{Reference Steel, Silove, Brooks, Momartin, Alzuhairi and Susljik23} although as the authors of the study acknowledge, it is possible that more distressed individuals were more likely to participate. Studies that have targeted a specific group are useful in allowing within-group comparisons and in increasing the number of populations in which these issues have been addressed, although the findings from any one study may not generalise to other refugee groups from different backgrounds.

In the only longitudinal study we identified, Keller et al looked at rates of symptoms of depression, anxiety and PTSD in detained asylum seekers and compared them with rates of mental disorder at follow-up. ^{Reference Keller, Rosenfeld, Trinh-Shvrin, Meserve, Sachs and Leviss10} Initial baseline interviews were conducted with 70 detainees (73% of those eligible) with a median duration of detention of 5 months. Follow-up data were collected at a median of 101 days later in 61 of the original participants. Some participants had been released (n=26) but the rest were still in detention. The authors used the PTSD sub-scale of the Harvard Trauma Questionnaire and the Hopkins Symptom Checklist–25 to measure anxiety and depression. Spearman correlation coefficients were used to examine the relationship between time in detention and anxiety (r=0.34, P=0.004), depression (r=0.28, P=0.017) and PTSD symptoms (r=0.28, P=0.019). Consistent with findings in the Australian studies, ^{Reference Steel, Silove, Brooks, Momartin, Alzuhairi and Susljik23} time in detention was directly related to severity of symptoms. This was also consistent with subjective reports; 70% of detainees reported deterioration in their mental health while in detention. Overall, clinically significant levels of depression, anxiety and PTSD were found in a high proportion of detainees in the sample (86%, 77% and 50% respectively). Independent sample t-tests were used to compare symptom scores from baseline at follow-up in those who had been released and those who were still detained. At baseline there was no significant difference between group scores in those who were eventually released and those still detained. The difference at follow-up was significant; however, reductions in symptoms were observed at follow-up in those who had been released (P<0.0001). The repeated-measures design allows for measurement of the effect of release from detention in comparison with continued detention in two groups of detainees who did not differ at baseline. These studies consistently suggest a possible independent effect of detention on mental health.

In the first study to investigate the impact of detention within the UK, Pourgourides et al provide qualitative explanations of how detention affects people psychologically. ^{Reference Pourgourides, Sashidharan and Bracken9} In-depth interviews with 15 male former detainees as well as focus group discussions (comprising former detainees, members of detainee support and campaigning groups, professional advisors and health professionals) informed a grounded theory account of how and why mental health problems arise in this group. The results of psychiatric diagnostic interviews with the detainees were reported: 27% were diagnosed with PTSD, 60% with depression, 7% with panic disorder and 7% with psychosis. Detailed qualitative findings from the study are outlined at length by the authors but are not described here. The authors highlighted a number of specific difficulties unique to the detention experience, which adversely affect mental health. The detention experience incapacitates detainees, in that it does not allow utilisation of usual coping skills, and constitutes a meaningless environment. Detainees are therefore preoccupied by time and experience extreme boredom and frustration as well as a sense of having no future. The potential for the detention environment to reactivate and exacerbate previous traumas was also raised as a theme. The authors conclude that the high incidence of hopelessness, depression and despair among detainees can be regarded as normal reactions to an abnormal situation, and detention itself as an ongoing trauma.

Impact of detention on children, adolescents and their families

Three studies that met our inclusion criteria looked at the impact of detention on children and their families. ^{Reference Sultan and O'Sullivan17,Reference Steel, Momartin, Bateman, Hafshejani and Silove25,Reference Mares and Jureidini28} In addition to the observations of mental health problems in adults, the participant–observer study by Sultan & O'Sullivan also reported on the effects on children. ^{Reference Sultan and O'Sullivan17} These effects were considered to be mediated through negative characteristics of the detention environment itself as well as through the impact of detention on the parents' mental well-being. Observed disturbances included separation anxiety, disruptive conduct, nocturnal enuresis, sleep disturbances (including nightmares, night terrors and sleepwalking) and impaired cognitive development. Severe symptoms of distress (including mutism, refusal to eat and drink, and stereotypical behaviours) were also reported in some cases. These reflect observations made by the authors, and unfortunately no information regarding the prevalence of these observations is reported, severely limiting the reliability of these data.

Mares & Jureidini reported on ten consecutive referrals made to a child and adolescent mental health service in Australia from a detention centre. ^{Reference Mares and Jureidini28} The study involved 16 adults and 20 children who were detained at the time of the study and had been in detention for 1–2 years at the time of the first interview. The study also included a follow-up interview for five families at 12 months. All participants were interviewed by child psychiatrists or allied health clinicians and consensus diagnoses were made. The authors reported that all children had at least one parent with mental health problems, but only two adults reported mental health difficulties prior to arriving in Australia. Among the 16 adults within the sample, 87% had major depression, 56% showed clinical symptoms of PTSD and 25% had a psychotic illness. Self-harm was also common – 31% had made significant repeated attempts. Among children aged less than 5 years, developmental delays were common. Out of ten such children, half had delays in language and social development. Emotional and behavioural dysregulation as well as attachment problems were observed.

Among the children in the older age range (6–17 years) mental health difficulties were extensive. All ten of these children met clinical criteria for PTSD. In addition, all ten had major depression and expressed suicidal ideation. Eight children had actually engaged in self-harm and the authors noted that a culture of self-harm existed within the detention centre. Seven had symptoms of anxiety disorder and half had persistent physical health symptoms. Children regularly reported boredom, a sense of injustice, sleep difficulties, anxiety regarding delays in educational progress and a sense of shame.

At the 12-month follow-up the well-being of the five families who were still detained had deteriorated. The authors reported that an initial improved sense of well-being among families who had been released from detention had generally not persisted. However, neither clinical diagnostic information nor outcomes from self-report measures accompanied these clinical observations in the report. In addition, the representativeness of these cases for other families held within this centre is unclear. The initial severity of the psychopathological symptoms in these children might have precipitated the initial referral to the service and their subsequent inclusion in the study.

Steel et al investigated mental health difficulties among detained families from a single ethnic group in one centre in Australia, ^{Reference Steel, Momartin, Bateman, Hafshejani and Silove25} by administering psychiatric interviews by telephone. The ethnic origin of the detainees interviewed was not specified in order to protect their anonymity. The study sample consisted of nearly all the members of the particular ethnic group within a single detention centre, comprising 10 families (14 adults and 20 children) out of 11 families who were eligible. Participants had been detained for a minimum of 2 years. All adult participants stated that they had experienced traumatic events prior to leaving their country of origin as well as en route to Australia. The semi-structured interviews included the Structured Clinical Interview for DSM–IV Axis I Disorders (SCID–IV) for adults and the Schedule for Affective Disorders and Schizophrenia for School-age Children – Present and Lifetime version (K–SADS–PL) for children. ^{Reference First, Spitzer, Gibbon and Williams29,Reference Kaufman, Birmaher, Brent, Rao, Flynn and Moreci30} These measures allowed assessment of prevalence of psychiatric disorder during detention as well as allowing diagnoses to be made retrospectively for the incidence of mental health difficulties prior to detention. The interviews also consisted of measures of common experiences in detention (Detention Experiences Checklist), as well as a list of nine stress symptoms taken from standard PTSD measures but focused on the detention experience itself (designed for this study): the Detention Symptom Checklist. A parenting questionnaire, again designed for the purpose of the study, was also included. All adults reported pre-migration trauma as well as traumatic experiences occurring in detention.

Psychiatric assessment indicated the prevalence of psychiatric disorders in adults and children to have increased markedly since the participants were detained. All adult participants were diagnosed with major depressive disorder, whereas only 21% reported symptoms (retrospectively) that would indicate a diagnosis of depression prior to detention. Two (14%) were diagnosed in detention with severe depressive disorder with psychotic features. A large proportion (86%) of adults were also diagnosed with PTSD while in detention – 50% of these cases were also retrospectively diagnosed with this. There was also an increase in suicidal ideation, with 93% of adults experiencing persistent suicidal ideation. Prior to detention, none of the participants had experienced persistent suicidal ideation nor had self-harmed; 36% of adults had self-harmed when assessed during detention. Overall, the authors reported a threefold increase in psychiatric problems. Parents also reported a marked decrease in their parenting capabilities since being detained. Only one adult responded positively to being ‘able to care for and support children’ and ‘able to control the behaviour of children’ while detained, whereas all participants stated that they were able to do this prior to detention.

Similarly to the results for parents, a substantial increase in psychiatric disorders was reported at the time of assessment among children. Comparison of the diagnoses made at the time of assessment while children were detained, with retrospective diagnoses for incidence of psychiatric disturbance prior to detention, revealed a tenfold increase in psychiatric difficulties among these children. All children were diagnosed with at least one psychiatric disorder at the time of assessment, the majority (80%) exhibiting multiple disorders. While in detention, 50% of the children were diagnosed with PTSD (with some re-experiencing symptoms being directly related to events in detention), all but one child received a diagnosis of major depressive disorder and 50% were diagnosed with separation anxiety disorder. Oppositional defiant disorder was also common, occurring in 45% of cases. Enuresis was present in 20% of the total sample but in four out of seven children aged 6–10 years, the age range in which this disorder usually occurs. Suicidal ideation was reported by 55% of the sample and 25% had self-harmed by cutting their wrists or head banging. In contrast, prior to detention, only one child met criteria for multiple disorders (depression, PTSD and separation anxiety). Two other children were retrospectively diagnosed with either depression or separation anxiety prior to being detained.

In addition to these findings of a deterioration in mental health during detention, the use of the Detention Experiences Checklist and the Detention Symptom Checklist allowed analysis of the extent to which current difficulties were directly linked to detention experiences. All participants reported experiences of traumatic events occurring during detention. ‘Sudden and upsetting memories of the time in detention’ and ‘images of threatening or humiliating events in detention’ were reported as causing distress in all adults and in 90% of the children. Increased anger and ‘feeling extremely sad and hopeless’ were also reported by all adults. All of the other nine detention-related symptoms were reported as causing distress in 86–100% of adults and in 53–90% of children. Although this study is limited by the methodological problems associated with administering interviews by telephone and the possibility of recall bias, the findings suggest an overall deterioration occurring as a response both to the detention process itself and to specific detention experiences.