Association between housing and health of refugees and asylum seekers in Germany: explorative cluster and mixed model analysis

BACKGROUND: Accommodation for asylum seekers and refugees (ASR) in Germany differs in many ways depending on a range of political, structural, social, and environmental factors. These contextual differences present a challenge for assessing health impacts of refugee accommodation. We aimed to devise a broad typology of refugee accommodation that allows to assess associations between housing and health of ASR. METHODS: We performed a cluster analysis of population-based, cross-sectional secondary data in Germany to identify clusters of refugee accommodation. We then assessed health disparities across clusters by performing bivariate analysis and linear mixed model regression analysis. RESULTS: We identified four clusters, three of them reflected different types of private accommodation and one pointed to collective accommodation. The collective accommodation cluster clearly differed from the private accommodation clusters in terms of space, area, level of restrictions, social connections and respondent satisfaction. Across private accommodation clusters we also found differences in space, area, and level of restrictions. In regression analysis, belonging to one of the private accommodation cluster was significantly associated with better mental health compared to belonging to the collective accommodation cluster. Physical health was significantly lower in one private accommodation cluster characterized by poor access to public transport and a higher level of restrictions compared to a private accommodation cluster showing better connections and a lower level of restrictions. CONCLUSION: We demonstrate that unfavourable conditions cluster in collective accommodation with negative outcomes for mental health but not for physical health. We also found health disparities across types of private accommodation. We conclude that housing plays a role in the production of health inequalities in ASR but needs to be assessed in a differentiated, multidimensional way. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12889-021-12458-1.