Clinician's Corner
August 3, 2005

Predisplacement and Postdisplacement Factors Associated With Mental Health of Refugees and Internally Displaced PersonsA Meta-analysis

Author Affiliations

Author Affiliations: Columbia Presbyterian Medical Center, New York, NY (Dr Porter); University of Melbourne, Melbourne, Australia (Dr Haslam).

JAMA. 2005;294(5):602-612. doi:10.1001/jama.294.5.602

Context The global refugee crisis requires that researchers, policymakers, and clinicians comprehend the magnitude of the psychological consequences of forced displacement and the factors that moderate them. To date, no empirical synthesis of research on these issues has been undertaken.

Objective To meta-analytically establish the extent of compromised mental health among refugees (including internally displaced persons, asylum seekers, and stateless persons) using a worldwide study sample. Potential moderators of mental health outcomes were examined, including enduring contextual variables (eg, postdisplacement accommodation and economic opportunity) and refugee characteristics.

Data Sources Published studies (1959-2002) were obtained using broad searches of computerized databases (PsycINFO and PILOTS), manual searches of reference lists, and interviews with prominent authors.

Study Selection Studies were selected if they investigated a refugee group and at least 1 nonrefugee comparison group and reported 1 or more quantitative group comparison on measures of psychopathology. Fifty-six reports met inclusion criteria (4.4% of identified reports), yielding 59 independent comparisons and including 67 294 participants (22 221 refugees and 45 073 nonrefugees).

Data Extraction Data on study and report characteristics, study participant characteristics, and statistical outcomes were extracted using a coding manual and subjected to blind recoding, which indicated high reliability. Methodological quality information was coded to assess potential sources of bias.

Data Synthesis Effect size estimates for the refugee-nonrefugee comparisons were averaged across psychopathology measures within studies and weighted by sample size. The weighted mean effect size was 0.41 (SD, 0.02; range, −1.36 to 2.91 [SE, 0.01]), indicating that refugees had moderately poorer outcomes. Postdisplacement conditions moderated mental health outcomes. Worse outcomes were observed for refugees living in institutional accommodation, experiencing restricted economic opportunity, displaced internally within their own country, repatriated to a country they had previously fled, or whose initiating conflict was unresolved. Refugees who were older, more educated, and female and who had higher predisplacement socioeconomic status and rural residence also had worse outcomes. Methodological differences between studies affected effect sizes.

Conclusions The sociopolitical context of the refugee experience is associated with refugee mental health. Humanitarian efforts that improve these conditions are likely to have positive impacts.