The relationship between ethnic composition of the residential environment and self-reported health among Turks and Moroccans in Amsterdam

BACKGROUND: Previous studies from the US and UK suggest that neighbourhood ethnic composition is associated with health, positive or negative, depending on the health outcome and ethnic group. We examined the association between neighbourhood ethnic composition and self-reported health in these groups in Amsterdam, and we aimed to explore whether there is spatial variation in this association. METHODS: We used micro-scale data to describe the ethnic composition in buffers around the home location of 2701 Turks and 2661 Moroccans. Multilevel regression analysis was used to assess the association between three measures of ethnic composition (% co-ethnics, % other ethnic group, Herfindahl index) and three measures of self-reported health: self-rated health, Physical and Mental Component Score (PCS, MCS). We adjusted for socioeconomic position at individual and area level. We used geographically weighted regression and spatially stratified regression analyses to explore whether associations differed within Amsterdam. RESULTS: Ethnic heterogeneity and own ethnic density were not related to self-rated health for both ethnic groups. Higher density of Turks was associated with better self-rated health among Moroccans at all buffer sizes, with the most significant relations for small buffers. Higher heterogeneity was associated with lower scores on PCS and MCS among Turks (suggesting worse health). We found spatial variation in the association of the density of the other ethnic group with self-rated health of Moroccans and Turks. We found a positive association for both groups, spatially concentrated in the sub-district Geuzenveld. CONCLUSIONS: Our study showed that the association of ethnic composition with self-reported health among Turks and Moroccans in Amsterdam differed between the groups and reveals mainly at small spatial scales. Among both groups, an association of higher density of the other group with better self-rated health was found in a particular part of Amsterdam, which might be explained by the presence of a relatively strong sense of community between the two groups in that area. The study suggests that it is important to pay attention to other-group density, to use area measurements at small spatial scales and to examine the spatial variation in these associations. This may help to identify neighbourhood characteristics contributing to these type of area effects on urban minority health.