RESUMEN
PURPOSE: To test whether upward socioeconomic neighborhood change has an effect on probability of distant metastasis at diagnosis of breast cancer among women who live there. METHODS: Census tract data (N = 1,137) from Cook County. IL, from 1990 and 2000 and cancer registry data for female breast cancer cases for these census tracts from 1994-2000 (N = 21,516) were used. A multilevel model of 1990 baseline socioeconomic status (SES) of neighborhoods and degree of neighborhood change 1990-2000 (compositional characteristics) and patient's age and race/Hispanic status (individual characteristics) was constructed to predict distant metastasis (vs. local and regional stage) at diagnosis. RESULTS: While residence in a census tract with lower baseline SES in 1990 (higher concentrated disadvantage and immigration and lower concentrated affluence) and being African American were associated with increased odds of distant metastasis at diagnosis, residence in an improving census tract was also associated with increased odds of distant metastasis at diagnosis. CONCLUSIONS: Paradoxically, both measures of initial neighborhood disadvantage and upward neighborhood socioeconomic change were independently associated with greater odds of distant metastasis at diagnosis of breast cancer. Neighborhood social and economic change can affect health.