Abstract
BACKGROUND: Studies on the association between health and neighborhood ethnic composition yielded inconsistent results, possibly due to methodological limitations. We assessed these associations at different spatial scales and for different measures of ethnic composition. METHODS: We obtained health survey data of 4673 respondents of Dutch, Surinamese, Moroccan, Turkish other non-Western and other Western origin. Neighborhood ethnic composition was measured for buffers varying from 50-1000 m. Associations with self-rated health were measured using logistic multilevel regression analysis, with control for socioeconomic position at the individual and area level. RESULTS: Overall ethnic heterogeneity was not related to health for any ethnic group. The presence of other Surinamese was associated with poor self-rated health among Surinamese respondents. The presence of Moroccans or Turks was associated with poor health among some groups. The presence of Dutch was associated with better self-rated health among Surinamese and Turks. In most cases, these associations were stronger at lower spatial scales. We found no other associations. CONCLUSIONS: In Amsterdam, self-rated health was not associated with ethnic heterogeneity in general, but may be related to the presence of specific ethnic groups. Policies regarding social and ethnic mixing should pay special attention to the co-residence of groups with problematic interrelations.