Context, health and migration: a systematic review of natural experiments.

Background: Migration health research pays little attention to the places into which people migrate. Studies on health effects of contextual factors are often limited because of the ability of individuals to self-select their environment, but natural experiments may allow for the causal effect of contexts to be examined. The objective was to synthesise the evidence on contextual health effects from natural experiments among migrant groups. Methods: We performed a systematic review of natural experiments among migrant populations in PubMed/MEDLINE, The Cochrane Library, Web of Science, CINAHL and Google Scholar for literature published until 13 October 2022. 5870 articles were screened in duplicate using the following inclusion criteria: quantitative natural experiment design, migrant population, context factor as treatment variable and health or healthcare outcome variable. Synthesis without meta-analysis was performed following quality appraisal using the EPHPP tool for quantitative studies and data extraction (PROSPERO: CRD42020169236). Findings: The 46 included articles provide evidence for negative effects of neighbourhood disadvantage on physical health and mortality, while finding mixed effects on mental health. Articles comparing migrants with those that stayed behind demonstrate detrimental effects of migration and adverse post-migratory contexts on physical health and mortality, while demonstrating favourable effects for mental and child health. Natural experiments of policy environments indicate the negative impacts of restrictive migration and social policies on healthcare utilization, mental health and mortality, as well as the positive health effects when restrictions are lifted. Interpretation: Natural experiments complement observational studies and provide robust evidence to advocate for more inclusive migration, health and social policies as well as neighbourhood improvement programmes. In order to strengthen the methodological approach, future research utilising natural experiments should be more explicit in the mechanisms underlying the experiment and provide details on potential causal mechanisms for the observed effects. Funding: German Science Foundation (FOR: 2928/GZ: BO5233/1-1).