Globalization in the 21st century has translated into an increasingly interconnected world, with more people than ever before in human history moving across borders in search of better employment opportunities and lifestyles. Accordingly, a steep upward trend in international migrant patterns has been observed in recent years, with an estimated 258 million migrants worldwide as of 2017, which is up from 220 million in 2010 and 173 million in 2000 [1]. For many migrants, uprooting to a new country is often accompanied by drastic changes in multiple aspects of their lives. The process of assimilating to new surroundings and cultural practices can lead to considerable levels of acculturative stress [2,3], which in turn has been linked with the manifestation of psychiatric disorders (e.g., depression) [4,5]. Globally, depression has been identified as the leading cause of illness and disability, with an aggregate point, one-year and lifetime prevalence of 12.9%, 7.2%, and 10.8%, respectively [6]. In its most severe form, it can lead to suicide. Approximately 800,000 people die from suicide annually, and it remains the second leading cause of death in 15 to 29 year-olds [7].

With that being said however, evidence of a definite cause-and-effect relationship between migration, acculturative stress and subsequent development of depression appears arbitrary at best. The association between migration and depression remains contentious and numerous studies have found the link to be inconsistent [8,9]. Given the conflicting evidence available with current empirical research, the relationship between migration and depression presents worthwhile questions to examine. The aim of this systematic review and meta-analysis is therefore to explore the aggregate prevalence of depression among international migrants, and to investigate for any possible variations in prevalence with demographic and educational factors.

Given the heterogeneity among international migrants, multiple varying definitions of the term “migrant” exist in literature, but none has so far been universally accepted. For the purposes of this paper, we adopted the definition as provided by the United Nations Convention of the Rights of Migrants according to UNESCO (The United Nations Educational, Scientific and Cultural Organization) [10], which defines “migrant” as a term “covering all cases where the decision to migrate is taken freely by the individual concerned, for reasons of ‘personal convenience’ and without intervention of an external compelling factor”. Children, refugees, asylum seekers as well as second generation and later migrants were therefore excluded from this definition.

Of the 2831 results that were obtained from the initial online electronic search, a total of 25 studies were finally included in this review. The process of study selection is summarised with the PRISMA flow diagram, as depicted in Figure 1. All the final included studies were cross-sectional in design and had utilized validated standardised instruments as methods to assess depression prevalence. The various methods that were used in the final included studies were as follows: the Beck Depression Inventory (BDI) [18], the Beck Depression Inventory 2nd edition (BDI-II) [19], the Centre for Epidemiologic Studies Depression Scale (CES-D) [20] and its Korean variation (CES-D-K) [21], the World Health Organization Composite International Diagnostic Interview (CIDI) [22] and its Turkish computer-assisted version (CIDI DIA-X) [23], the Disaster-Related Psychological Screening Test (DRPST) [24], the Hopkins Symptom Checklist-25 (HSCL-25) [25], the Mini International Neuropsychiatric Interview (MINI) [26], the Patient Health Questionnaire-9 (PHQ-9) [27] and its two-item version (PHQ-2) [28], and the Primary Care Evaluation of Mental Disorders (PRIME-MD) [29]. There were a total number of 31,391 participants included in this review, with 16,121 migrants and 15,270 native respondents covered altogether. The prevalence of depression and characteristics of each included article are presented in Appendix A.

Following analysis of data obtained from 16,121 migrant participants across 20 different countries, this systematic review and meta-analysis reports an aggregate prevalence of depression among international migrants of 15.6%.

Although such a figure suggests that depression is a common and substantial mental health problem affecting migrants worldwide, we did not note any significant difference in depression prevalence between migrants and native participants. This appears to correspond with the findings of a previous systematic review that found no conclusive evidence of an increase in risk of depression or other mood disorders in general among members of the migrant community [30]. It is difficult to account for such a finding, as one would expect the acculturative stress of migration to be an important risk factor for depression. The same paper however attributed this apparent disparity to a selection hypothesis, which postulates that those genetically predisposed to mood disorders develop stronger attachments to people in their home countries and are therefore less likely to migrate [31]. An alternative explanation ties in the concept of resilience, a construct that represents positive adaptation, social, and psychological competence in times of significant adversity or trauma [32]. It is possible that migrants who voluntarily uproot themselves may possess resiliency factors not shared in common with natives, which explains why no differential risk for adverse mental outcomes were noted between the two groups. Migration may also at times be associated with upward social mobility and improved quality of life for individuals who move away from environmental pathogens such as high crime and unemployment rates. The improved psychosocial outcomes from this change in environment may negate the increased risk of depression from acculturative stress. Therefore, to categorically conclude that the process of migration has minimal association with depression without first accounting for factors specific to migrant individuals and their country of migration would undoubtedly be overt reductionism.

The wide range of populations that were examined in this systematic review and meta-analysis contributed to the significant heterogeneity in depression prevalence across studies. Meta-regression analysis identified moderators, including level of educational attainment, employment status, and length of residency spent in country of migration as significant contributors to the heterogeneity in aggregate prevalence. Studies which included a larger percentage of migrants who had graduated from high school and college tended to report greater depression prevalence. Given that education has long been regarded as a vehicle for social mobility, such a finding would appear to be contrary to what should be expected. However, various possible explanations for this observed trend have been proposed. An earlier study found that the prevalence of mood dysfunction was greater in Asian immigrants with higher education who had immigrated at a younger age when compared to those with lower educational attainment but who had immigrated at older ages [33]. This implies that the association between the level of educational attainment and depression may in fact be confounded by specific factors related to the migration process itself such as an individual’s age at the time of migration. Unfortunately, we could not investigate these factors in this paper since only a limited number of studies had provided the relevant data. Further research is therefore necessary and warranted for any conclusions to be drawn.

Apart from the confounding effect of other moderators, a separate study also found that educational level does not always directly translate to equivalent economic opportunity in ethnic minorities [34], of which migrants often find themselves a part of upon uprooting to a foreign land. Highly educated migrants may experience employment difficulties due to language difficulties or potential employers’ devaluation of foreign education credentials [35]. It is also possible that a higher level of educational attainment may increase awareness of social injustices and racial discrimination, with such feelings of perceived unfairness in turn resulting in feelings of distress that eventually manifest as mood dysfunction [36].

The somewhat paradoxical relationship between level of educational attainment and depression may also be accounted for by increased levels of distress caused by a discrepancy between one’s lifestyle and economic status [37]. Many a times, increased level of educational attainment may heighten expectations for a certain standard of living. Failure to realise such expectations may result in stress and ultimately depression.

Employment status was found to be a significant moderator for depression prevalence among migrants, with current employment associated with lower depression prevalence. This comes as no surprise as the association between unemployment and poor mental health has long been established [38,39]. A previous meta-analysis found evidence to support a causal relationship between the two, as substantiated by an individual’s mental well-being typically declining following unemployment, but improving after subsequent reemployment [40]. A number of theories have been proposed to explain such an association.

One such theory postulated that mental distress arising from unemployment stems primarily from the lack of five positive latent functions of employment [41]. It posits that employment in fact imposes a time structure on the day, increases social contact with others, offers one a sense of collective purpose, allows for increased status, and encourages activity. Unemployment inevitably causes the deprivation of these important latent functions, in turn resulting in mental distress and poor quality of life [42]. Such an effect is all the more pertinent among migrants, who are often blighted with difficulties involving social and economic integration in their newly adopted country. This is corroborated by previous research that reported that the negative effects of unemployment are likely increased among minority groups [43]. It is possible that the social and financial instability unemployment confers upon individuals aggravates the acculturative stress socioeconomically disadvantaged migrants already face and increase their susceptibility to mood dysfunction.

An inverse relationship between depression prevalence and length of residency was also observed. Studies that had included migrants with a shorter duration of stay in their country of migration were inclined to report greater depression prevalence. This is largely consistent with reports in the literature [44] and may be attributed to the fact that levels of acculturative stress typically peak in the early post-migration phase. During this period of assimilation, migrants commonly encounter problems with integration as they adjust to a host of changes in their lives, including chaotic disruptions to their social networks and challenges to their cultural values. The high levels of acculturative stress that accompany this process of adaptation may in turn result in increased mental distress and subsequent mood dysfunction.

The strengths of this review include an extensive literature search in identifying a large number of relevant articles, established methodology using the PRISMA guidelines, and a comprehensive meta-analysis that included regression and subgroup analysis. Studies that were included were also conducted in a broad range of diverse countries, which meant that any differential risk as a result of cultural or ethnic differences was minimized. Other strengths include a lack of significant publication bias and the application of a random-effects model in establishing robust aggregate depression prevalence among migrants.

Nonetheless, several limitations have been identified. First, this review has a high level of heterogeneity typical of meta-analyses of large number of studies. Second, meta-regression analyses were only performed on selected demographic moderators as there was limited data available on other moderators including age of migration. Data pertaining to pre-migration characteristics such as socio-economic background and psychological health status was also inadequate. Third, meta-regression merely denotes an observational association and is limited by ecological fallacy [45]. Fourth, all of the included studies were cross-sectional by design, which meant that a temporal causality between migration and depression could not be identified. Further reviews in the future should include more longitudinal studies to address this. Lastly, our inclusion and exclusion criteria meant that studies that were published in languages apart from English were excluded from our analysis. Studies were also retrieved primarily from two online databases, and thus this review may not be entirely representative.

In conclusion, this systematic review calculated an aggregate prevalence of depression among migrants of 15.6%. Meta-regression analyses showed that level of educational attainment, employment status, and length of residency were significant contributors to the high level of heterogeneity in prevalence of depression.

Our findings highlight the fact that migrants, particularly those newly arrived and unemployed, remain susceptible to developing mood dysfunction. With the number of migrants worldwide set to increase for the foreseeable future, such a trend is likely to persist unabated. These findings have underlined the gap in existing social and cultural supports for recent migrants, and it is imperative that more community resources be made available to help ease the process of social and economic integration. Screening for mood dysfunction should also be commenced early among the members of this vulnerable population.

Future research should attempt to elucidate the impact pre-migration mental health and host country factors has on depression prevalence among migrants. Qualitative studies that assess the role that specific cultural factors has on depression predisposition should also be pursued.

We would like to thank National University of Singapore and its related personnel for their support in the completion of this review.

Conceptualization, S.Q.F. and R.C.H.; methodology, S.Q.F., R.C.H. and W.W.T.; software, R.C.H. and W.W.T.; formal analysis, S.Q.F., R.C.H. and W.W.T.; investigation, S.Q.F.; resources, R.C.H. and W.W.T.; data curation, S.Q.F.; writing—original draft preparation, S.Q.F.; writing—review and editing, S.Q.F., C.S.H., B.X.T., L.H.N., R.S.M., R.C.H.; visualization, S.Q.F.; supervision, R.C.H.; project administration, R.C.H.

This research received no external funding.

The authors declare no conflict of interest.