Abstract
Background:
Research on the impact of post-migration experiences on the mental health of migrant populations has shown a predictive link between post-migration living difficulties and psychological distress. While many studies have focussed on refugees and asylum seekers, there is a considerable gap in the literature concerning undocumented migrants.
Aims:
The aim of this study was to assess the differences in mental health between documented and undocumented migrants in Belgium. It identified the post-migration difficulties encountered by these two groups and measured their impact on their levels of trauma, resilience, anxiety, depression and their assumptive world.
Method:
This study involved 69 participants, aged 18 to 68 years, who were either documented or undocumented resident in Belgium. Our data collection included the Post-Migration Living Difficulties Checklist for the measurement of post-migration difficulties, the Posttraumatic Stress Disorder Checklist – DSM-V version for the measurement of trauma, the World Assumptions Questionnaire for the measurement of the assumptive world, the Hopkins Symptom Checklist-25 for the measurement of depressive and anxiety disorders and the Adult Resilience Measure – Revised for the measurement of resilience.
Results:
Our analyses showed that the mental health of undocumented migrants was poorer than that of documented migrants, with higher levels of trauma, anxiety and depression, a more negative assumptive world, a lower sense of control and lower total and personal resilience. Our results also revealed that post-migration living difficulties were more severe and more numerous for undocumented migrants, and that they were associated to different themes in both groups.
Conclusions:
The fact that post-migration experience plays such an important role in the mental health of migrant populations raises significant clinical, political and societal considerations. Furthermore, it would appear that undocumented migrants represent a distinct migratory group with their own specificities in terms of migration journey and mental health.
Introduction
According to the United Nations High Commissioner for Refugees (2022), there are more than 89.3 million people displaced worldwide by conflict, human rights abuses, climate change, food insecurity, etc. In 2021, 43.5% of asylum applications submitted to the Belgian authorities resulted in asylum being granted to 10,093 people, with a refusal rate of 56.5% (Asylum Information Database, 2022). In 2017, the Belgian authorities found that only 20% of dismissed migrants left the country (Migrants-Refugees, 2020). People in an irregular situation, also known as undocumented migrants, refer to people who do not have or no longer have a valid residence permit authorising them to stay in Belgium legally. Undocumented migrants have few legal rights (Ligue des droits de l’Homme, 2016).
The scientific literature has long established that migrant populations are likely to develop mental health disorders (Cleary et al., 2018; Knefel et al., 2020). The most frequently diagnosed would be post-traumatic stress disorder (PTSD), depressive disorders and finally anxiety disorders (Aragona et al., 2011; Brunnet et al., 2018; Silove et al., 1997).
Research on the predictors of mental health in migrant populations has mainly focussed on pre-migration and perimigration factors (Sinnerbrink et al., 1997; Steel et al., 2017). Recent literature has shown a predictive link between post-migration living difficulties and psychological distress in migrant populations (Alemi et al., 2016; Aragona et al., 2020; Gleeson et al., 2020). In the face of these advances, it has been observed that the post-migration journey is still insufficiently investigated in current research, and in treatment protocols, psychotherapies (Knefel et al., 2020). However, some authors have argued that post-migration difficulties have a more severe impact on mental health than adversities experienced before migration (Carswell et al., 2011; Laban et al., 2008). Undocumented migrants appear to be more heavily impacted by uncertain employment, limited social and family networks, restricted access to information, lack of legal protection, exploitation and access to housing (Croix-Rouge de Belgique, 2014; Garcini et al., 2017; Momartin et al., 2006; Nickerson et al., 2019). They are also more likely to be exposed to physical, verbal, psychological and gender-based violence (Keygnaert et al., 2012; Steel et al., 2006).
In view of these findings, the aim of this study was to assess the differences between documented and undocumented migrants in terms of mental health. Secondly, we identified the post-migration living difficulties experienced by these two groups, and measured their impact on their levels of trauma, resilience, anxiety, depression and their assumptive world.
Methods
Participants
Participants consisted of 69 persons divided into two groups. The first group was composed of 31 participants (45%) in an irregular situation, referred here as ‘undocumented migrants’. While the second group was composed of 38 participants (55%) in a regular situation, referred here as ‘documented migrants’. Among this second group, different residence permits could be identified: 17 people were holders of international protection (25%), 4 people were holders of subsidiary protection (6%), 13 people were naturalised (18%) and 4 people through family reunification (6%). Inclusion criteria in this study were: (a) aged 18 years or older, (b) having arrived in Belgium at least 1 year prior to the testing, (c) documented or undocumented background in Belgium and (d) literate in at least one of the study languages (French, English, Spanish, Arabic, Albanian or Russian). Asylum seekers were not included. Recruitment methods included advertising at services supporting migrants and snowball sampling. The study was approved by the Ethics Committee of the Faculty of Psychology, Logopaedics and Educational Sciences of the University of Liege.
Procedure and measures
All participants were asked to complete five questionnaires: the Post-Migration Living Difficulties Checklist, the Posttraumatic Stress Disorder Checklist – DSM-V version (PCL-5), the World Assumptions Questionnaire, the Hopkins Symptom Checklist-25 and the Adult Resilience Measure – Revised. These questionnaires were selected because they are regularly used in cross-cultural research. All questionnaires were translated from English into French and Spanish. We also benefited from the help of accredited translators.
Post-Migration Living Difficulties Checklist (PMLD)
The PMLD is a self-report measure consisting of 23 items (Silove et al., 1998). It was designed to assess the negative life experiences in a post-migration context (Alemi et al., 2016; Tay et al., 2019). The PMLD has been consistently identified as a predictor of mental health amongst displaced populations (Nickerson et al., 2010). The inventory was adapted, and we retained 18 of the 23 items. It has good psychometric properties, with Cronbach’s alpha (α = .72–.88; Usama et al., 2021).
Post-traumatic Stress Disorder Checklist – DSM-V version (PCL-5)
The PTSD Checklist – DSM-V version (PCL-5) is a self-reported measure consisting of 20 items (Weathers et al., 2013). Its psychometric properties showed that it provides a valid and accurate measure. This scale has good internal consistency (α = .96), good test-retest reliability (r = .84), as well as good convergent (rs = .74–.85) and discriminant validity (rs = .31–.60; Bovin et al., 2016).
Hopkins Symptom Checklist-25 (HSCL-25)
The HSCL-25 by Mollica et al. (1987) is a 25-item self-report measure assessing the presence of anxiety and depressive symptoms according to DSM-V criteria. It consists of two distinct parts: a subscale measuring anxiety, composed of 10 items, and a 15 items subscale measuring depression. This scale has good internal consistency (α = .89–.91), good test-retest reliability (r = .89) and correct validity (88% sensitivity; 73% specificity) (Bris, 2017).
World Assumptions Questionnaire (WAQ)
Kaler’s (2009) WAQ is a 22-item self-report measure that assesses the assumptive world of individuals. The questionnaire consists of four subscales: controllability of events (CE), trustworthiness and goodness of people (TGP), understandability and predictability of people (UPP) and safety and vulnerability (SV). In this study, we chose to administer only two categories with a total of 11 items.
Adult Resilience Measure – Revised (ARM-R)
The ARM-R by Liebenberg and Moore (2018) is a self-reported measurement tool consisting of 17 items, assessing resilience. This scale has good internal consistency with overall resilience (α = .87), personal resilience (α = .82) and relational resilience (α = .82). It has good validity, measured by a Rasch model (Resilience Research Centre, 2022).
Data analysis
All analyses were conducted using the SAS statistical program Version 9.4. There was less than 5% missing data on any of the variables included in the analyses.
Results
Demographics
Participants in this study had a mean age of 37 years (SD = 10.53) years (range 18–68 years), with over one third of participants being male (n = 25; 36.23%). We were able to identify 26 different nationalities. In the documented migrants group participants were originally from t: Russia and Colombia: 13%; El Salvador: 11%; Kosovo and Morocco: 8%; Syria and Mauritania: 5%; Iran, Palestine, Libya, Burundi, Cameroon, Algeria, Afghanistan, Senegal, Mexico, Tunisia, Turkey and Venezuela: 3%. In the undocumented migrants group participants were originally from: Syria and Venezuela: 13%; Albania and Cameroon: 10%; Burkina Faso, El Salvador and Guinea: 7%; Iraq, Morocco, DRC and Rwanda: 6%; Senegal, Central Africa and Tunisia: 3%.
Half of participants had a partner, 33.82% were married (n = 23) and 17.65% were in a relationship (n = 12). Among the participants having children, 11 were separated from them, with at least one child living abroad (15.94%).
Furthermore, the study reported that only 26% of our sample were employed (n = 18), 36% were financially dependent on social welfare (n = 25), 4% lived mainly on donations and financial support from friends and family (n = 3), while 33% did not receive any income (n = 23). The income stability of documented migrants averaged around 3.97 (SD = 1.22) on a Likert scale of 1 (unstable) to 5 (stable), while it reached only 1.71 (SD = 1.16) for undocumented migrants.
Finally, the duration of the migration journey for documented migrants was less than 1 year, with an average of 0.34 years (SD = 0.97), while it was on average more than 1 year for undocumented migrants, reaching 1.32 years (SD = 2.39). As for the duration of the asylum procedure for documented migrants, it was an average of 23.63 months (SD = 27.13), whereas for undocumented migrants it was an average of 37.06 months (SD = 34.77).
Post-traumatic stress disorder
Levels of trauma were higher in the undocumented group (U = 1,441.00; p ⩽ .0001). The trauma scores were statistically higher for undocumented migrants (M = 48.97; SD = 14.25) than for documented migrants (M = 27.11; SD = 20.42). There was a significant difference between documented and undocumented migrants for the prevalence of trauma (χ²(1) = 20.50; p ⩽ .0001). Thus, in our sample, we were able to observe a 37% prevalence of PTSD in the documented migrant group, while it was 90% in the undocumented group.
Anxiety and depression
Levels of depression and anxiety were higher in the undocumented migrant group. We were able to observe significant differences between the two groups for the total scores of depression (U = 1,333.00; p = .0005), anxiety (U = 1,308.50; p = .0014) and both disorders combined (U = 1,331.50; p = .0005). Anxiety and depression scores were statistically higher for undocumented migrants (M = 2.56, SD = 0.61; M = 2.55, SD = 0.70; M = 2.57, SD = 0.61) than for documented migrants (M = 1.90, SD = 0.84; M = 1.92, SD = 0.86; M = 1.89, SD = 0.86). Test confirmed this result for the prevalence of depression (χ²(1) = 18.21; p ⩽ .0001), anxiety (χ²(1) = 15.55; p ⩽ .0001) and both disorders combined (χ²(1) = 16.61 ; p ⩽ .0001). We were able to observe a 44% prevalence of anxiety for documented migrants, compared to 90% for undocumented migrants. For depression, these rates were 44% for documented migrants and 94% for undocumented migrants.
World assumptions
The assumptive world of undocumented migrants was more negative than that of documented migrants. There were significant differences between the two groups in terms of total scores (t = -2.70; p = .0088) and the sense of control measured by the WAQ (t = -3.26; p = .0018). In contrast, it could not show any significant difference between the two groups for the sense of security (t = -1.18; p = .2434). The assumptive world was statistically more negative for undocumented migrants (M = 35.06; SD = 4.61) than for documented migrants (M = 31.89; SD = 4.96). More specifically regarding the feeling of control, which appears lower for undocumented migrants (M = 15.68; SD = 2.81). However, the feeling of security did not vary between the two groups.
Resilience
Levels of resilience were lower in the undocumented migrant group. We found significant differences for the total scores obtained in the ARM-R (t = 2.67; p = .0095) and for the personal resilience scores (t = 2.67; p = .0095). In contrast, the Mann-Whitney U test did not allow us to identify difference with regard to relational resilience (U = 900.00; p = .0531). The levels of total resilience were statistically higher for documented migrants (M = 65.33; SD = 11.16) than for undocumented migrants (M = 57.35; SD = 13.27). More specifically, this would relate to personal resilience, (M = 38.31; SD = 6.84), while relational resilience did not vary between the two groups. These results were not confirmed by the chi-square test of independence (χ²(3) = 4.08; p = .2531).
Post-migration living difficulties
Undocumented migrants experienced more post-migration difficulties (t = -4.54; p ⩽ .0001). It also showed a significant difference in the number of post-migration difficulties between the two groups (U = 1,355.50; p = .0011). The number and severity of post-migration difficulties were statistically higher for undocumented migrants (M = 12.58, SD = 2.46; M = 62.90, SD = 10.24) than for documented migrants (M = 8.66, SD = 4.91; M = 47.79, SD = 17.09). In our sample, documented migrants encountered about 8 post-migration difficulties, while this figure was 12 for undocumented migrants. It also indicated that there would be significant differences between the two groups for the variables of income stability (U = 636.50; p ⩽ .0001) and duration of the asylum procedure (U = 1,305.00; p = .0080).
The incomes of documented migrants (M = 3.97; SD = 1.22) were significantly more stable than those of undocumented migrants (M = 1.71; SD = 1.16). The duration of the asylum procedure was statistically longer for undocumented migrants (M = 37.06; SD = 34.77) than for documented migrants (M = 23.63; SD = 27.13). It was just under 2 years on average for documented migrants, while it exceeded 3 years for undocumented migrants. Finally, there is a significant differences between both groups for the source of income (χ²(3) = 30.19; p ⩽ 0.0001). We found that only 13% of the undocumented migrants had a job, compared to 37% of the documented migrants and 68% of undocumented migrants had no source of income, compared to 5% of documented migrants.
In addition, we were able to observe different themes in the post-migration difficulties. The difficulties most often mentioned by documented migrants concerned the separation from their families (71%), followed by difficulties in obtaining appropriate housing (68%), and finally the inability to return to their country of origin in case of an emergency (63%). While the difficulties most reported by undocumented migrants were difficulties with employment and the fear of being sent back to their country of origin (90%), followed by problems related to the asylum application and worries about the family left behind (87%).
Relationship between mental health and post-migration living difficulties
Post-migration difficulties had an impact on the four mental health scales used in this study. Linear regressions on the whole sample showed a significant influence of the severity of post-migration difficulties. Thus, post-migration difficulties had a significant impact on the level of trauma (F = 98.05; p ⩽ .0001), assumptive world (F = 15.23, p = .0002), sense of control (F = 11.34, p = .0013) and security (F = 8.38, p = .0052), levels of depression and anxiety (F = 65.13, p ⩽ .0001; F = 49.36, p ⩽ .0001; F = 65.51, p ⩽ .0001) and levels of total, personal and relational resilience (F = 7.06, p = .0099; F = 6.12, p = .0160; F = 4.14, p = .0459), for the whole sample. However, these results could be contrasted when these linear regressions were applied to each group. The severity of post-migration difficulties still significantly affected the level of trauma (F = 31.14; p ⩽ .0001; F = 38.53; p ⩽ .0001) and the levels of anxiety and depression (F = 19.82, p ⩽ .0001; F = 14.69, p = .0005; F = 21.68, p ⩽ .0001; F = 35.00, p ⩽ .0001; F = 28.55, p ⩽ .0001; F = 26.53, p ⩽ .0001) for both groups. With regard to total assumptive world (F = 5.49, p = .0251), sense of security (F = 4.27, p = .0465), total resilience (F = 7.39, p = .0103) and relational resilience (F = 7.47, p = .0099), we observed a significant influence of post-migration difficulties for documented migrants, but not for undocumented migrants. Finally, we no longer found any influence of post-migration difficulties on the feeling of control and on personal resilience, for both groups.
Discussion
This study investigated the impact of post-migration living difficulties on mental health in a sample of documented and undocumented people living in Belgium. Our analyses confirmed that the mental health of undocumented migrants was poorer than that of documented migrants. However, we were unable to find any difference between the two groups in feelings of safety and relational resilience. Our results also revealed differences in the post-migration difficulties experienced by both groups. Finally, we found that post-migration difficulties had a significant impact on the mental health of the migrant populations. The majority of the results obtained in this study confirmed our initial hypotheses and were in line with previous international research.
First, we observed higher levels of PTSD in our sample for undocumented migrants. This difference could be explained in different ways: either undocumented migrants are more exposed to traumatic events during their pre-migration journey, or during their peri-migration journey, or after their resettlement in the host country, in the post-migration phase. These three hypotheses remain widely discussed in current scientific research and are not mutually exclusive (Knefel et al., 2020; Steel et al., 2017). Furthermore, authors have recently highlighted that the difficulties encountered in the host country have a greater impact on trauma than previous experiences in the country of origin (Carswell et al., 2011; Li et al., 2016; Zimbrean et al., 2014) and the development of PTSD is thought to be largely dependent on the conditions of resettlement in the host society (Schweitzer et al., 2011; Silove et al., 2000). Thus, studies have shed light on a disturbing effect of ‘re-traumatisation’ in Western countries (Posselt et al., 2020).
Then, we found higher levels of anxiety and depression in our sample for undocumented migrants. These results are in line with previous research (Newnham et al., 2019). This difference could be explained by a combination of many obstacles encountered in the host country. Studies have identified a risk of major depressive disorder that is twice as high for unemployed migrants (Croix-Rouge de Belgique, 2014; Heeren et al., 2014; Hocking et al., 2015a).
It is possible that the high levels of PTSD, anxiety and depression measured in the undocumented migrant group are due to their vulnerability to the development of mental health disorders. Given their low representativeness in prevalence studies of migrant populations, we might therefore expect an underestimation of the general mental health statistics for this particular group. Additionally, the levels of mental health disorders experienced by the documented migrants in our sample were particularly high compared to the prevalence observed in the general population, as demonstrated in previous studies (Schweitzer et al., 2011). This allows us to deduce that the restorative effect of the positive asylum decision would not erase the trauma experienced despite the granting of a protection status.
Moreover, our findings showed a more negative assumptive world and a lower sense of control for the undocumented migrants. In contrast, we could not find any difference between the two groups for the feeling of security. These results are new. On the other hand, the absence of difference between documented and undocumented migrants about the feeling of security is more surprising. Different explanations may be suggested. These results could indicate that documented migrants have a lower sense of security than expected. Thus, we could assume that recognised refugees continue to live in a context of insecurity. This could also indicate that undocumented migrants have a higher sense of security than expected. Thirdly, these results could be the consequence of an external factor or methodological biases.
In addition, we found lower total resilience and personal resilience in our sample for undocumented migrants. In contrast, we could not find any difference between the two groups for relational resilience. These results are new. In our view, the differences between documented and undocumented migrants could be explained by the Drive to Thrive theory developed by Hou et al. (2018). This theory postulates that resilience is determined by the maintenance of regularity and structure in the everyday environment, following the experience of traumatic events. This theory posits that in precarious conditions a reduction in the regularity of daily routines would be associated with poorer mental health and resilience capacities (Hou et al., 2018).
Furthermore, this study showed that the number and severity of post-migration living difficulties differed. Those experienced by undocumented migrants are both more severe and more numerous. These results are consistent with the current literature comparing migrants with temporary and permanent status (Garcini et al., 2017; Momartin et al., 2006; Nickerson et al., 2019).
Our results also show that the post-migration difficulties experienced vary in their content. On the one hand, the most reported difficulties for documented migrants are: separation from their families, followed by difficulties in obtaining appropriate housing, and finally the inability to return to their country of origin in case of an emergency. On the other hand, the difficulties most often mentioned by undocumented migrants concern issues related to employment and fear of being repatriated to their country of origin, followed by problems related to the asylum application and concerns about the family left behind. These data lead us to make several observations. First of all, despite the protection status they hold, documented migrants still seem to regularly face major post-migration difficulties related to housing or employment. Secondly, it is surprising to note that employment-related difficulties are equivalent to the fear of expulsion for undocumented migrants and outweigh all other post-migration difficulties. This underlines the fact that employment integration is the starting point for resettlement in the host country, providing opportunities for socialisation, financial resources, access to basic needs and a way of finding a role in society. Finally, we can also identify in these results the importance of the social and family network for both groups which is recognised as a protective factor in the literature (Fossion et al., 2006; Schweitzer et al., 2007).
Additionally, findings showed a significant impact of post-migration difficulties on measures of trauma, depression, anxiety, the assumptive world, sense of control and security and total, personal and relational resilience for the whole sample. These findings are consistent with recent studies (Li et al., 2016).
Finally, our research has highlighted important clinical, social and political implications for mental health professionals, policy makers and the general public.
a) Psychological interventions should take greater account of resettlement concerns. As many authors have pointed out, post-migration difficulties could have a greater impact on mental health than the previous experience in the country of origin or on the migration route (Carswell et al., 2011; Zimbrean et al., 2014). However, the therapeutic interventions currently developed are mainly focussed on past trauma (Knefel et al., 2020).
b) Our findings may lead to the consideration of three types of policy measures: the revision of restrictive migration measures on the one hand (Momartin et al., 2006; Steel et al., 2006), and governments should provide adequate resources to promote the social integration of displaced people (Hocking et al., 2015b). Finally, developing policies that focus on access to work for undocumented migrants (Hocking et al., 2015a, Posselt et al., 2020).
d) Psychosocial interventions focussing on post-migration difficulties could contribute to improve the living conditions of migrant populations (Newnham et al., 2019; Nickerson et al., 2011).
e) We identify several elements to investigate for future research. First, the need to focus research efforts on the field of on the mental health of undocumented populations and the field of irregular migration. Despite the considerable barriers in accessing this population, we believe that these often socially excluded people need to be recognised and deserve the attention of researchers. Secondly, this paper has shown the need to further develop current research into post-migration difficulties, and to move away from considering previous traumatic events as central to the symptomatology of migrant populations. It may therefore be relevant to identify the post-migration difficulties that have the most significant impact on mental health, in order to make the necessary clinical, social and policy adjustments to address them. Finally, we argue that resilience factors should be further explored in epidemiological and interventional research. We advocate a holistic approach to mental health, targeting both the difficulties, but also the strengths and resources of migrant populations. Developing our knowledge of the resilience factors of migrants can help to mobilise the necessary means to improve the care provided to these vulnerable groups.
Limitations
The current study had several limitations.
First, our sample differed in terms of the residence status obtained, age, gender, nationality or length of stay in Belgium. For example, the group of documented migrants included people from different backgrounds, who had been granted international protection status, subsidiary protection status, naturalisation or family reunification. Likewise, migration conditions differed. It is therefore not possible to account for the role of the different migration conditions or the different status obtained. Several differences that could have limited generalisability of findings (Li et al., 2016). However, this diversity was intended to gather a representative sampling of the highly diverse migrant population in Belgium.
Second, a bias often identified within the literature concerns the exaggeration of symptoms of precarious populations in order to be viewed favourably and attract public sympathy (Hocking et al., 2015a, 2015b, Steel et al, 2006). Within our study, this could have led to an amplification of difficulties and needs. Conversely, some participants might have concealed certain aspects of their experience so that disclosure would not undermine their status or procedure (e.g. in the case of undeclared work or false claims). We tried to avoid these reporting biases by making it clear to each participant that all data collected was anonymous and confidential, so that what was shared could not in any way promote or affect their immigration status.
Finally, the data collection undertaken using self-report measures rather than clinician administered diagnostic interviews. Our analyses relied exclusively on self-reporting by participants, as it is frequently done in migration research. This way of collecting data could have biased our results, due to its lack of objectivity, compared to diagnostic interviews administered by a clinician. Indeed, self-reporting of clinical symptoms may have led to an underestimation or overestimation of these symptoms and limit the accuracy of the diagnosis
