Starting from scratch: prevalence, methods, and functions of non-suicidal self-injury among refugee minors in Belgium

The Geneva convention defines a ‘refugee’ as “someone who is unable or unwilling to return to their country of origin owing to a well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group, or political opinion” [1]. Unaccompanied minors are less than 18 years of age, not accompanied by any person exercising parental authority or custody under the national law of the minor, and originating from countries other than those in the European Economic Area [2]. The above definition of ‘refugee’ implies that most of them have experienced insecurity and stressful or even traumatic life events in their country of origin. Adverse life events [e.g. 3, 4], psychological distress [e.g. 5, 6], identity confusion [e.g. 7], and ethnic status [e.g. 8] are all risk factors for non-suicidal self-injury (NSSI) that can be expected in the majority of refugee minors. However, research on NSSI in refugee adolescents is scarce. Consequently, the primary aim of this study is to situate NSSI within a refugee adolescent population.

Refugee minors who have been exposed to war and political violence report traumatic loss, bereavement, separation, forced displacement, community and domestic violence, physical abuse, emotional abuse, impairment in the caregiver’s caregiving performance, etcetera [9]. During their transit, numerous stressful and dangerous situations may have occurred. Once they have arrived at their destination, a long asylum procedure, a difficult integration and an uncertain future await [e.g. 10–13]. Due to different origins, ethnicities, cultures, family and personal histories, refugees constitute a diverse, heterogeneous group with increased levels of psychological distress as a common factor. More specifically, post-traumatic stress disorder, depression and anxiety disorders are frequently reported in refugee children [9, 11, 14]. Despite pre- and post-migration distress, young refugees, like other adolescents, begin to develop a personal identity. Rejection by peers of the same ethnicity is an obstacle to this development. The integration of racial and ethnic identities into new social and cultural contexts might complicate this already demanding process, causing acculturative distress [9, 15, 16].

In comparison to Belgian adolescents, peers with a migration background report significantly more traumatic events, symptoms of severe post-traumatic stress, as well as higher avoidance scores. They do, however, show less anxiety symptoms and comparable amounts of depressive and emotional symptoms. The amount of traumatic experiences influences the prevalence of emotional and behavioural problems [10]. Migrant adolescents report less externalising problems and lower hyperactivity scores than their Belgian counterparts [10, 17], as well as very low levels of high-risk behaviours (sexual risk taking, running away, etc.), crime involvement, and alcohol abuse, common in Western traumatized samples [9]. One possible explanation could be their striving for a better future [10, 17]. However, when parents or social workers are questioned about adolescent refugees’ emotional and behavioural problems, the refugee group scores significantly higher on both internalising and externalising problems than natives do [18]. These differing findings could suggest that the behaviour of refugee minors is either perceived as more problematic by others than by themselves or that refugee minors underestimate or underreport their own problems.

A literature review on self-harm in refugees and asylum seekers found that the hopelessness and loss of future aspirations, combined with a traumatic background, common in refugees, is a risk factor for self-harm [19].

A negative association was determined between engagement in NSSI at some point in life and levels of affirmation, belonging, and commitment to one’s ethnic group. Therefore, a sense of belonging could be seen as a protective factor against engagement in NSSI, but other factors such as socioeconomic status (SES) and gender, might influence aforementioned relationships [8, 20]. While ethnic/racial identity (being aware of and understanding social/historical/cultural aspects of one’s ethnic group) might be a protective factor, ethnic status is a significant predictor of non-suicidal self-injurious behaviour [8]. Religion, especially Baptism and Islam, serves as a protective factor [8, 21].

Unaccompanied refugee adolescents report more emotional problems, more symptoms of anxiety, more depression, and more post-traumatic stress reactions than those living with their parents in the country of asylum [10, 22, 23]. One study comparing the inpatient psychiatric care between accompanied and unaccompanied refugee minors found that the latter exhibited more self-harm and suicidal behaviour [23]. Jensen et al. [24] found that 11% of unaccompanied refugee minors displayed suicidal ideation. However, even though unaccompanied Afghan refugee minors are all likely to have lived through a range of traumatic experiences, only 34% reported clinical levels of PTSD in a study by Bronstein et al. [25] in comparison to the .4 to 10% measured in the general population. The mere fact that they succeeded in their flight could be an indication of their resilience and capacities [13, 18]. Bhui and colleagues [26] also hypothesised that people with certain mental disorders, such as psychosis, are less likely to succeed in their flight to a safer country. It might indeed be that only the strongest and most resilient of refugees make it to the Western world.

Wester and Trepal [20] found a sense of belonging to be negatively related to the number of methods reported. No significant differences are found between ethnic groups (Caucasian, African American, Hispanic, Asian American, and multiracial groups) concerning the number of methods used in NSSI [20].

In Western studies on adolescents, automatic reinforcement functions are reported more frequently than social reinforcement functions [27]. It is unclear whether this also applies on non-Western samples. A study on Hong Kong adolescents, for example, found the regulation of interpersonal issues to be the main function while NSSI did not serve to regulate negative emotions. Another study on university students in India found that the function of minor forms of NSSI is to regulate social environments by means of avoidance, while the function of moderate to severe NSSI serves to regulate emotions [8].

Research on non-suicidal self-injury has been predominantly conducted on White samples in Western countries [8]. (Western) adolescents engaging in NSSI show higher levels of psychological symptoms than not self-injurious youths [6]. Approximately one out of five young adults engaging in NSSI exhibit high clinical symptomatology [28]. Psychological distress measured at age 12 is considered a significant predictor of NSSI [5]. Many studies link adverse life events and trauma symptoms to self-injurious behaviour [e.g. 3, 4]. Literature on (non-clinical) Western adolescents reports a lifetime NSSI prevalence of 17.2–18% [29, 30], and a 12-month prevalence of 9.6% to 28.4% [30]. Research demonstrates equivalency across gender [30]. As one singular episode is sufficient for being included in the lifetime prevalence statistics, some studies differentiate between the more common occasional (e.g. one to four reported lifetime episodes) forms and repetitive forms of self-injury. The American Psychiatric Association proposes a minimum of five occasions in the last year as one of the DSM-5 criteria for ‘nonsuicidal self-injury’ [31]. Zetterqvistet al. found that 6.7% of adolescents in a community sample meet the DSM-5 criteria for an NSSI disorder diagnosis [32]. In two studies by Brunner, approximately one out of every four adolescents engaging in ‘deliberate self-injurious behaviour’ or D-SIB (without suicidal intent), did so on a repetitive basis (i.e. five or more instances), rather than occasionally. Repetitive self-injury is related to a greater psychological burden [33, 34].

Many adolescents (39.8–47.75%) restrict themselves to one singular method of NSSI (e.g. cutting, burning, etc.), although 11.26% to 22.8% apply four methods or more [34–36]. Females tend to prefer methods like scratching and cutting to punching objects with the intention of hurting themselves. The latter is more common in male subjects [37]. An average of 4.3 NSSI functions per adolescent was found by Zetterqvist et al. [4].

Automatic reinforcement functions (e.g. to feel something or to relieve tension) are reported more commonly than social reinforcement functions (e.g. to avoid activities or to get help) [4, 27, 38].

Although NSSI is non-suicidal in its primary intention, research does link NSSI to suicidal thoughts and behaviours [39–41]. There is a high co-morbidity in adolescence. NSSI is seen as a significant risk factor for suicidal ideation, with an almost threefold risk for suicidality after even one act of NSSI [40]. Suicidal ideation has been associated with automatic functions, as well as with the number of methods used and the urge of self-injury [41]. In their review, Hamza et al. [39] distil several studies on NSSI and suicide into three theories: (1) the ‘Gateway Theory’, which places NSSI and suicide as extremes on a continuum on which NSSI may build up towards committing suicide, (2) the ‘Third Variable Theory’, in which a third variable (e.g. psychiatric disorder or psychological distress) is responsible for both the engagement in NSSI and the suicidal behaviour, instead of NSSI increasing the suicidal risk, and (3) ‘Joiner’s Theory of Acquired Capability for Suicide’, in which NSSI can be seen as one of many means to practise suicide by learning to overcome the fear and pain associated with it. However, in Joiner’s Theory, other conditions have to be fulfilled as well for NSSI to result in suicide (i.e. social isolation and the feeling of being a burden to others). According to Whitlock and colleagues [40], focusing on enhancing the perceived meaning in life and positive relationships with others could reduce the risk of NSSI behaviour developing into suicidal thoughts or actions.

Research exploring non-suicidal self-injurious behaviour in adolescent refugees seems to be scarce, though some research points to cultural differences. The primary aim of this study is to explore the prevalence of NSSI behaviour within a refugee minor population, as well as the methods used and the functions ascribed to it. This study also aims to compare the results with existing literature on the Western adolescent population.

As many of the risk factors previously described (e.g. adverse life events, psychological distress, identity confusion, ethnic status, lower SES) can be applied to refugee minors, we hypothesize that the prevalence of NSSI for this population will exceed the prevalence known for their Western adolescent counterparts. We also hypothesize a higher prevalence of self-injurious behaviour within the group living without parents. We predict no gender differences in NSSI prevalence.

Studies on non-Western populations show no differences in the number of methods used for NSSI between samples of differing ethnicities. We therefore hypothesize similar numbers used by adolescent refugees. As research on methods and functions in non-Western populations is limited and indecisive, we will also compare methods and functions of NSSI common in refugee minors with the existing literature on Western adolescent samples.

This study is an attempt to fill a void in the existing literature by exploring NSSI prevalence, methods, and functions in refugee minors in Belgium.

Contrary to expectations, refugee minors do not differ greatly from their Western counterparts in prevalence, methods or functions of NSSI behaviour. A lifetime prevalence of 17.4% was measured and is comparable to the 17.2% and 18% found by Swannell et al. [30], and Muehlenkamp et al. [29]. The 12-month prevalence of 11.4% is in accordance with the 9.6% to 28.4% found by Swannell et al. [30]. The resilience approach, allowing different mental outcomes when experiencing similar risks, may serve as a plausible explanation [25], combined with the hypotheses that people with certain mental disorders are less likely to succeed in their search for safer grounds [26]. However, 68.4% of self-injurers reported five or more acts of NSSI, indicating a greater psychological burden, as this number fluctuates around 25% in studies on Western adolescents [33, 34]. The age of onset of 13.11 years falls within the 12 to 14 years found for a Belgian and Dutch adolescent population [7, 35]. As in most literature on Western samples [e.g. 30], no statistical differences were found between boys and girls relating to lifetime or 12-month prevalence, or age of onset. Girls did report significantly more emotional problems and boys addressed more conduct problems (cf. SDQ). Being accompanied or not, having both parents around, or living in an asylum centre, did not influence the prevalence of NSSI. These findings are in accordance with research by Bean et al. [12], where very little variance in mental health outcome for refugees was found when examining gender, presence of family, and living in a centre. Unaccompanied refugees did, however, estimate the self-perceived impact of personal problems in their different life domains (i.e. home life, friendships, classroom learning, and leisure activities) to be higher in comparison to their accompanied peers (cf. SDQ Impact Scale).

Currently, a common idea among Belgian caretakers working with refugees is that Afghan males engage more in NSSI than other refugees do. Within this study however, no statistical differences in NSSI prevalence were found between Syrian and Afghan refugee minors. Moreover, Syrians reported significantly more difficulties than Afghan refugees (cf. Total Difficulties Scale, SDQ). The popular belief may be partially explained by the fact that there are more Afghan than Syrian refugees in Belgium [51]. A higher presence could result in more visibility to those working with refugees and can create the misconception of more mental health issues. This might also account for the idea that unaccompanied minors as well as youths residing in asylum centres are thought to be more sensitive to self-injury, since in both cases more caretakers are involved. Unaccompanied youths are more closely monitored than children who are part of a family and it needs no explanation that people living in asylum centres have less privacy than those inhabiting houses.

Literature has shown that religiosity is a protective factor for NSSI [8, 21]. The results of this study show no significant differences in NSSI behaviour between religious and non-religious individuals. However, the proportion of non-religious people in this sample is small (5%) and the manner used to investigate the nature of religiosity (i.e. through the questions ‘Are you religious?’, and ‘If yes: Christian, Muslim, Hindu, or other…’) seems too limited to jump to conclusions. Furthermore, stating that one is religious does not indicate how this religion is experienced or put into practice.

The number of methods used per person is also comparable with Western adolescents: 55% restricts themselves to one method (vs. 39.8–47.75% in a Western population), 20% applies five or more methods (vs. the 11.26–22.8% of Western adolescents using four or more methods) [34–36]. There seems to be less difference concerning the choice of method between boys and girls compared to a Western population with both genders preferring scratching, and banging or punching objects.

The mean number of six functions per refugee engaging in NSSI is comparable to the 4.3 functions per Western adolescent reported by Zetterqvist et al. [4].

Similar to their Western peers [e.g. 27, 38], refugees mainly report automatic functions. The most reported function was that of practising suicide. Joiner’s Theory of Acquired Capability for Suicide hypothesises that NSSI can lead to suicide when there is social isolation and the belief of being a burden to others [39]. The data of the SDQ questionnaire (cf. Table 4) clearly indicate the high proportion of peer problems (e.g. being solitary, not having many friends, not being liked or being bullied). In combination with the high impact participants engaging in NSSI estimate that their problems have on different areas of their daily life (i.e. home life, friendships, classroom learning, leisure activities) (cf. SDQ Impact Scale), these findings might suggest their considerable vulnerability for committing suicide. However, this study did not focus on suicide. Suicide and practising suicide were two functions from a long list of functions to be ticked if applicable. Further research is needed to establish how suicidal ideation and NSSI are connected in relation to refugee minors.

Research by Klonsky and Olino [28] indicates that approximately one out of five young adults engaging in NSSI exhibit high clinical symptomatology. Even though the SDQ is only a screening instrument, its results suggest that refugees engaging in NSSI suffer from more emotional and behavioural problems than their Western peers; with more than 50% reporting abnormal levels (i.e. a considerable risk of clinically significant problems) of total difficulties and emotional problems, as well as increased levels of conduct and peer problems.

In order to discern whether Western adolescents and refugee minors can be treated for NSSI in a similar way, it is necessary to take a closer look at this behaviour within the refugee population. Several circumstances in the past, present and future could have predicted higher NSSI outcomes. However, this study did not find notable differences, possibly suggesting different coping mechanisms upon which treatment could be focused. Furthermore, it might be possible that certain cultures have higher or lower levels of acceptance towards behaviours like NSSI and suicide. More research is needed to understand these findings and to learn how people around the world look upon NSSI behaviour.

The findings above hardly show any difference in NSSI prevalence, methods and functions between refugee and Western adolescents. This possibly suggests strong protective factors or different coping styles or self-regulation techniques in refugee minors. The findings do, however, indicate more repetitive forms of NSSI, a substantial risk of clinically significant problems in self-injuring refugees (i.e. high rates of abnormal scores on SDQ), and an increased risk of suicidal behaviour. More research is needed to explore protective factors, as well as the different reactions of refugees to stress and adverse life events. Research comparing refugees in less safe camps en route to Europe might also shed a light on the effects of arriving in a safe country of refuge.