Increased social identification is linked with lower depressive and anxiety symptoms among ethnic minorities and migrants: A systematic review and meta-analysis (Kristine Brance, et al.)

Share This

Highlights

  • •We examined the relationship between social identity and common mental disorders in ethnic minorities and migrants.
  • •Increased social identification is linked with decreased symptoms of anxiety and depression with small effect sizes.
  • •Seven variables significantly moderated the relationship between social identity and depression.
  • •Four variables significantly moderated the relationship between social identity and anxiety.

Abstract

Evidence suggests that social identities, which provide purpose and a sense of belonging to the social world, promote resilience against psychological strain and protect well-being. This is especially important in ethnic minorities, who experience exclusion and discrimination from the majority group, and in migrant populations where adverse experiences, such as prejudice, disconnection from previous identities and issues of integration into the host country, negatively impact well-being. Drawing from the social identity theory, a meta-analysis was conducted examining the influence of group memberships and sense of belonging on ethnic minority and migrant mental health (depression and anxiety). The final search on three databases (i.e., PubMed, PsycINFO, Web of Science) was conducted on August 9th, 2022, identifying 3236 citations before removing any duplicates within and between databases. Across the 74 studies that met the inclusion criteria for the meta-analysis, increased social identification (ethnic, national and other types of identification) was associated with low psychological symptoms. We found that social identification is protective against common psychological disorders but with small effect sizes for depression (r = − 0.09, CI = [− 0.12; − 0.06]) and anxiety (r = − 0.08, CI [− 0.12; − 0.03]). Results are discussed with regard to the role that social context plays on ethnic minority and migrant mental health and the importance of facilitating migrant integration with the host society after displacement.

1. Introduction

People have migrated throughout history, creating ethnically diverse communities across the world, with recent projections showing a future increase in ethnic minority groups (U.S. Census Bureau, P. D, 2019). Despite this trend, these minorities still face precarious socio-economic conditions and discrimination, which are consistent predictors of mental health disorders (e.g., Harris et al., 2006Karlsen & Nazroo, 2002Karlsen, Nazroo, McKenzie, Bhui, & Weich, 2005Nazroo, 2003). Epidemiological research seeking to explore ethnic disparities in mental health disorders points to the complexity in this association. For instance, research suggests that ethnic minorities in England and in other European countries experience elevated rates of common mental disorders (Missinne & Bracke, 2012Smith, Bhui, & Cipriani, 2020Weich et al., 2004). Ethnic minority status has also been identified as a risk factor for psychotic disorders (Leaune et al., 2019Tortelli et al., 2018). However, most studies conducted in the United States (US) produce contradictory results. For example, a large body of evidence shows that ethnic minorities in the US have a lower prevalence of psychiatric disorders, such as anxiety and major depression (Barnes & Bates, 2017Barnes, Keyes, & Bates, 2013Breslau et al., 2006Breslau, Kendler, Su, Gaxiola-Aguilar, & Kessler, 2005Harris, Edlund, & Larson, 2005Himle, Baser, Taylor, Campbell, & Jackson, 2009Williams et al., 2007). In the context of social stressors and mental health, these findings appear to contradict the social stress paradigm, which predicts that disadvantages, such as social status and discrimination, lead to mental health issues. Nonetheless, studies in the field, including the US, consistently indicate that mental health disorders tend to persist for longer in ethnic minorities (Breslau et al., 2005Williams et al., 2007), which may be attributed to their lower use of mental health services (Harris et al., 2005Wang et al., 2005Wang et al., 2005).

1.1. Migration and mental health

The literature on ethnic minorities with immigration status is more consistent, with findings globally indicating that this population is particularly vulnerable and has a greater likelihood of developing post-traumatic stress disorder (PTSD), major depression, anxiety, and nonaffective psychosis (Bas-Sarmiento, Saucedo-Moreno, Fernández-Gutiérrez, & Poza-Méndez, 2017Brandt et al., 2019Close et al., 2016Fazel, Wheeler, & Danesh, 2005Porter & Haslam, 2005). These findings are particularly important as, in recent years, the number of people who have moved between distant geographical regions has reached its highest humanity has ever seen; in 2020, the number of people who lived in a country other than the one in which they were born reached over 280 million, and this number is expected to increase further in the future (United Nations Department of Economic and Social Affairs, P. D, 2020).

Because of the wide range of economic, social, political, cultural, and environmental factors that foster migration, any simple definition of a migrant risks being reductive. The International Organization for Migration (IOM) confirms that there is no universally accepted definition and describes that a migrant is someone who moved within or outside the state of birth regardless of legal status, the reason for migration, whether the movement is temporary or permanent or voluntary or involuntary (Sironi, Bauloz, & Emmanuel, 2019). In practice, there are numerous reasons why people leave their usual place of residence. Some migrate out of choice in search of work opportunities or education. However, others have been forced to flee their homes either internally or outside their state of residence for reasons such as natural or other environmental disasters or in response to armed conflict and violence. By the end of 2021, the number of forcibly displaced people reached 89.3 million worldwide, including 53.2 million people who have relocated within their own country. Of these, 27.1 million are refugees and 4.6 million are asylum seekers (The United Nations High Commissioner for Refugees, 2021), where, according to the 1951 Geneva Convention, a refugee is a person who is forced to flee a country due to a well-founded fear of persecution based on reasons, such as race, religion, political beliefs, nationality, or a membership to a particular social group and who is unable to seek protection from that country (Sironi et al., 2019); in contrast to a refugee, who has already received protection, an asylum seeker is someone who is only seeking this protection.

Because of this lack of consensus, scholars tend to use the term migrant inconsistently, and some authors have failed to provide a clear explanation of whom they consider to be migrants in their research. For example, Close et al. (2016), in a recent systematic review of the literature on the mental health of 1st generation migrants (those who have made the journey from one country to another, as opposed to their descendants in the second, third generation etc.), use the definition proposed by IOM. Yet, in a study conducted in Germany by Geschke, Mummendey, Kessler, and Funke (2010), a migrant was considered anyone with a culture other than German (in other words, migrant status was confounded with ethnic minority status), while, in a US study by Keller, Joscelyne, Granski, and Rosenfeld (2017), migrants were defined simply as individuals who had arrived at the US border from the Northern Triangle of Central America. In light of this lack of consensus, the current study draws from the IOM definition of a migrant as anyone who moves away from their usual place of residence regardless of legal status, the reason for migration and the length of stay.

Given the distressing events forcibly displaced people experience, research has established that forced migration is a strong risk factor for developing psychiatric disorders, with most reviews in this area exploring PTSD followed by depression and anxiety (Uphoff et al., 2020). For example, a meta-analysis of 56 studies conducted in five different regions, including Africa, Latin America, the Middle East, Asia, and Europe, showed that refugees and internally displaced people report worse mental health outcomes relative to non-refugee groups (Porter & Haslam, 2005). Furthermore, a systematic review indicated that refugees resettled in Western countries are more likely to be diagnosed with PTSD and major depression than the general population in those countries (Fazel et al., 2005). Similarly, a review exploring first-generation migrants, including refugees and asylum seekers who had relocated to high-income countries, such as the US, Canada, United Kingdom, Sweden, and Australia, reported significantly higher prevalence rates of PTSD, depression, and anxiety compared to the native population in the host country (Close et al., 2016). Therefore, a recent meta-analysis on refugees in Western host countries confirmed that the traumatic events migrants experience prior to migration have also been shown to be a risk factor for the development of nonaffective psychosis (Brandt et al., 2019). Nonetheless, while those who migrate under adverse circumstances such as refugees have an elevated risk of developing psychological disorders, migration itself poses a potential psychological threat. A systematic review by Bas-Sarmiento et al. (2017) demonstrated that migrant populations across the world, including those who migrate out of choice, experience an increased risk of psychopathology, such as depression, anxiety, and somatic disorders, compared to the native population.

Scholars have tried to identify which premigration and postmigration factors contribute to this effect. For example, migrants who have experienced traumatic events such as exposure to torture and violence, suffered injuries, forced to evacuate under dangerous conditions, witnessed fighting between armed forces and who have been separated from family or lost a family member, are at a great risk for developing mental health issues (Cantekin & Gençöz, 2017Duraković-Belko, Kulenović, & Dapić, 2003Kira, Shuwiekh, Rice, al Ibraheem, & Aljakoub, 2017Lindencrona, Ekblad, & Hauff, 2008Rasmussen et al., 2010). This extensive literature has been synthesized by several reviews which have demonstrated that, despite varying prevalence rates across studies, war-related traumatic experiences are consistently linked with elevated rates of PTSD, depression, and anxiety (Porter & Haslam, 2005Steel et al., 2009). Moreover, the existing literature emphasizes the importance of the process of displacement, such as long and unsafe journeys, and of post-displacement experiences that may compound or alleviate migrant mental health outcomes. These challenges include lack of employment opportunities and poverty (Beiser & Hou, 2017Bernardes et al., 2010Papadopoulos, Lees, Lay, & Gebrehiwot, 2004Porter & Haslam, 2005Priebe et al., 2012Rasmussen et al., 2010Silove, Sinnerbrink, Field, Manicavasagar, & Steel, 1997); perceived interpersonal discrimination, such as verbal abuse and physical assault; as well as perceived institutional discrimination (Bernardes et al., 2010Branscombe, Schmitt, & Harvey, 1999Ellis, MacDonald, Lincoln, & Cabral, 2008Karlsen et al., 2005Karlsen & Nazroo, 2002); poor housing and living conditions (Bernardes et al., 2010Papadopoulos et al., 2004Porter & Haslam, 2005Rasmussen et al., 2010Steel et al., 2009); feelings of loss of cultural roots including unfamiliar environments, different values, traditions and beliefs, as well as language (Ager & Strang, 2004Papadopoulos et al., 2004Phillimore, 2011Priebe et al., 2012); lack of safety and access to resources (Ager & Strang, 2004Phillimore, 2011Rasmussen et al., 2010); social isolation and lack of social support due to the loss of social networks (Norris, Aroian, & Nickerson, 2011Papadopoulos et al., 2004Priebe et al., 2012Silove et al., 1997). An additional stressor for asylum seekers is their pending status, with research showing that prolonged time in detention centers has an adverse effect on migrant mental health (Keller et al., 2003Steel et al., 2004).

1.2. Social identity and belonging

While research has identified numerous social, economic and cultural displacement factors that need to be addressed to improve psychological well-being in ethnic minorities and migrants, one important psychological factor has been overlooked – the need to belong. The sense of belonging to the social world is one of the fundamental psychological needs (Baumeister & Leary, 1995), which enhances psychological well-being (Cruwys et al., 2013Cruwys et al., 2014Haslam, Jetten, Postmes, & Haslam, 2009). Hence, people’s social connectedness predicts psychologically and physically healthier lives (Holt-Lunstad, Smith, & Layton, 2010). According to the social identity theory, a person’s social identity can relate to any group that a person identifies as a psychologically meaningful description of the self (not just ethnic, cultural and national identity as focused on in this review, but also, for example, sexual identity, identification with school or neighborhood) which has resulted in studies in this research field using a variety of measures to assess social identification. While there is limited evidence on whether the different instruments measure the same concept, there is a growing body of evidence supporting the hypothesis that identification with groups has health benefits and is protective against a range of mental health issues in vulnerable populations (Jetten, Haslam, & Alexander Haslam, 2012). Within this context, evidence shows that increased social identification is a predictor of better mental health outcomes and coping strategies after major life transitions for stroke patients (Haslam et al., 2008), for people who suffered traumatic injuries (Jones et al., 2012), for people facing financial stress (Elahi et al., 2018), as well as for those who live in homeless shelters (Jetten et al., 2015).

While ethnic minorities and migrants have an increased likelihood of developing mental health issues (Brandt et al., 2019Close et al., 2016Weich et al., 2004), empirical evidence on the benefit of multiple social identities to ethnic minorities and migrants is scarce, with most research focusing on a single dimension of social identity. For example, literature indicates that ethnic identification plays a crucial role on ethnic minority mental health, predicting lower likelihood of developing a lifetime-psychiatric disorder, including depression and anxiety (Burnett-Zeigler, Bohnert, & Ilgen, 2013), as well as enhancing overall psychological well-being (Branscombe et al., 1999). Furthermore, research indicates that ethnic identification has a positive effect on perceived discrimination, buffering against the development of depressive symptoms for ethnic minorities (Ikram et al., 2016) and ethnic minorities with immigrant status (Thibeault, Stein, & Nelson-Gray, 2018). Other studies explored ethnic minority identification with their close environment, showing that a sense of belonging to a community protects from the development of depressive symptoms (Gonyea, Curley, Melekis, & Lee, 2018Hill, 2009).

With regards to migrant social identities, a recent study explored group identification of Syrian refugees, demonstrating that increased Syrian identification derived from the sense of belonging to the Syrian community and the perseveration of this identity after migration was linked with lower levels of depression and anxiety (Çelebi, Verkuyten, & Bagci, 2017). Similarly, Smeekes, Verkuyten, Çelebi, Acartürk, and Onkun (2017) found that Syrian refugees belonging to multiple social groups before migration were more likely to maintain group memberships after migration, which in turn was linked with a decreased risk of depression and greater life satisfaction. Other scholars examined the role migrant identification with the host culture plays, suggesting that migrants’ greater sense of belonging to the US culture is linked with decreased depressive and anxiety symptoms (Meca, Gonzales-Backen, Davis, Hassell, & Rodil, 2019Tikhonov, Espinosa, Huynh, & Anglin, 2019).

Despite this growing support for the positive mental health benefits of social identity in minorities and migrants, the consistency of the findings and strength of this effect remain uncertain. We therefore conducted a meta-analysis of relevant studies, focusing on common mental disorders, hypothesizing that increased social identification would be linked with lower levels of common mental disorders. In addition, we sought to assess the influence of methodological and contextual factors that may account for variations across the studies.

2. Methodology

2.1. Data sources and search strategy

A protocol of the review was developed prior and published on the International Prospective Register of Systematic Reviews (PROSPERO), registration number CRD42019129184, available from https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42019129184.

This meta-analysis was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines (Moher, Liberati, Tetzlaff, Altman, & Altman, 2009; see Appendix A). To achieve the objectives of the current study, we systematically identified articles on three databases: PubMed, PsycINFO, and Web of Science. All available records from 1970 to 2021 were searched using the following keyword combinations: (immigrant OR asylum seeker OR migrant OR refugee OR displaced person OR displaced people OR ethnic minorit* AND identity OR group belonging OR group membership OR group identification OR social identification OR identification OR sense of belonging AND common mental disorders OR depress* OR posttraumatic stress OR anxiety OR panic disorder OR obsessive-compulsive disorder). The final search on all databases was conducted on August 9th, 2022.

2.2. Inclusion criteria

Studies were included if they: (i) were published in a peer-reviewed journal; (ii) used a quantitative design (e.g., cross-sectional, longitudinal); (iii) included participants 18 years of age or older; (iv) explored ethnic minorities and/or migrants (v) using any type of instrument to measure (whether culturally adapted or not, see supplementary materials Appendix C) at least one of the common mental disorders defined by the National Institute for Health and Care Excellence including depression, generalized anxiety, panic, obsessive-compulsive, post-traumatic stress and social anxiety disorders (National Institute for Health and Care Excellence, 2011); (vi) used any type of social identification measure, including culturally adapted and validated or self-developed, which assesses any dimension of a person’s social identity (e.g., ethnic identity, national identity; see supplementary materials Appendix C); (vii) reported a quantitative finding of a direct association between social identity and common mental disorders.

2.3. Exclusion criteria

Studies were excluded if they: (i) used mixed methodology; (ii) drew the sample from a general population and then compared different groups in terms of ethnic background or migration status; (iii) did not report data separately for migrants or ethnic minorities; (iv) examined clinical samples.

2.4. Study selection

Following the conduct of the searches, K.B. reviewed all of the titles and/or abstracts of the studies and eliminated those studies that unambiguously failed to meet the inclusion criteria outlined above. A random selection of 229 (10%) of both the included and excluded studies was screened by the second researcher, V·C, who disagreed about 3 of the included studies (98.7% agreement equating to a kappa of 0.960, reflecting a prior decision to include studies for further examination if in doubt). After the initial screening, full-text articles were assessed for eligibility against the inclusion/exclusion criteria by K.B. A random selection of 48 (10%) of both the included and excluded studies were also examined by V.C.; of these, 40 were agreed to be excluded, 6 were included by both raters, and 2 were rejected by the second rater; hence there was agreement in 95,83% of papers, equating to a kappa of 0.833. The third author, R.B., was consulted for final agreement on the disputed papers.

2.5. Data extraction

K.B. extracted data from each study using a standardized form. The form included information on the title, author, publication year, study location, study design, study population characteristics, sample size, measurement instruments, the social identity dimension explored, and the association between social identity and common mental disorders. A random selection of 10% of the standardized forms was verified by V.C. Any discrepancies were resolved through discussion between researchers or by the third reviewer R.B.

2.6. Data coding

A coding manual was developed prior to data extraction by all researchers. Coding was done by the first researcher K.B. Data raising any questions was directly discussed with the second and third researchers to make the coding decision. Note that short-form or revised versions for both social identity and common mental disorders were coded under the same category as the original scales (see Appendix B for the full list of coded variables).

2.7. Assessment of methodological quality

There is no consensus on the assessment of methodological quality for observational studies (Sanderson, Tatt, & Higgins, 2007Shamliyan, Kane, & Dickinson, 2010), especially for cross-sectional studies in migrant research due to sampling challenges and language barriers. Whilst there is no golden rule to quality assessment, research suggests that the included quality components should be specific to the research area (Shamliyan et al., 2010). For example, in research on refugee mental health, language is identified as an important criterion when assessing the methodological quality of studies (Fazel et al., 2005). The current study used a five-point quality appraisal scale from Bogic, Njoku, and Priebe (2015), which was developed according to key quality criteria identified in previous reviews in this research area. The first three components relate to the sample selection bias minimization, while the remaining two evaluate the assessment validity of the studies. A cumulative quality score was calculated for each study raging from 0 to 5. Lower quality studies received a score between 0 and 3, whereas high quality studies received 4 or 5. The following criteria were assessed:

  • 1.The sampling
    • a.The use of random or inclusive sampling (non-random = 0, random or inclusive = 1)
    • b.The sample size if non-random sampling (<200 = 0, ≥200 = 1);
  • 2.The sample representativeness, i.e., the sample frame was a true or close representation of the target population (not representative = 0, representative = 1);
  • 3.The response rate (<60% or not mentioned at all = 0, ≥60% = 1);
  • 4.The use of validated and reliable measurements (valid and reliable measure not used = 0, valid and reliable measure used = 1);
  • 5.The language in which the survey was conducted (second language or through interpreter = 0, native language or participants were proficient in the assessment language = 1).

2.8. Analyses

The metric of choice for the current meta-analysis was Pearson’s r because the majority of the included studies (86.7%) reported data in terms of bivariate correlations. Other statistical methods included regression and logistic regression analyses. Other statistical measures were converted to r based on the statistical information extracted from each study through the following procedures.

First, beta coefficients (β) ranging from −0.50 to 0.50 were transformed using the following formula (Peterson & Brown, 2005):r=β+0.5λwhere λ = 1 when β is nonnegative and λ = 0 when β is negative. Two studies reported results in terms of unstandardized β coefficients. Because studies did not provide sufficient information to convert data into Pearson’s r, they were excluded from the meta-analysis (i.e., Cislo, Spence, & Gayman, 2010Tummala-Narra et al., 2018).

Second, log odds ratios (Log Odds Ratio) were converted to the standardized mean difference d using the following formula (Cooper, Hedges, & Valentine, 2009):d=LogOdds Ratio×3πwhich was then transformed from the standardized mean difference d to r using the following formula (Cooper et al., 2009):r=dd2+a

For those studies that included multiple measures of social identity, thus reported multiple correlations, for example, between ethnic identity and depression as well as national identity and depression, the average of all relevant correlations was taken. In order to do so, first, all the relevant Pearson correlation coefficients r were transformed to Fisher’s z using the following formula (Cooper et al., 2009):rz=0.5×ln1+r1−r

Then the average of Fisher’s z values was taken and back-transformed to Pearson’s r using the following formula (Cooper et al., 2009):r=e2z−1e2z+1

In one case, an analysis described as nonsignificant without any additional information was set to r = 0.00 (i.e., Tikhonov et al., 2019).

The converted effect size values were included in all statistical analyses below. Due to the considerable heterogeneity among the included studies in terms of sample characteristics and social identity dimensions explored, a random-effects model was used to estimate the magnitude of the effect across studies. Follow-up moderator analyses were conducted to investigate which potential participant characteristics (e.g., migration status, ethnic group) and methodological variables (e.g., social identity measure, sample size) account for heterogeneity, also applying the random-effects model. In addition, publication bias was assessed using the “trim-and-fill” method to estimate the number of potentially missing studies due to publication bias and to impute their values in the analysis to show the adjusted average effect size (Duval & Tweedie, 2000). Results of publication bias analyses were illustrated using the funnel plot; hence the “trim-and-fill” method assumes that studies in the funnel plot should be symmetrically distributed around the mean effect. All analyses within the current meta-analysis were conducted using the Comprehensive Meta-Analysis software version 3.

Read More

Leave a Comment