in patients aged 25-74 years (a total of 73 159 patients). Multilevel logistic regression analyses were performed with individual-level characteristics at the first level and level of neighborhood deprivation at the second level.
RESULTS:The age-standardized prostate cancer mortality rate was 1.5 times higher in men living in high-deprivation neighborhoods than in those living in the most affluent neighborhoods. Mortality rates were also associated with certain individual-level characteristics, that is, age, marital status, family income, educational attainment, immigration status, urban/rural status, mobility and comorbidity. For example, there was a strong relationship between prostate cancer mortality and being unmarried, having a low income or educational attainment, and hospitalization for chronic obstructive pulmonary disease. In the full model, the risk of prostate cancer mortality was 25% higher in men living in the most deprived neighborhoods than in those living in the most affluent neighborhoods.CONCLUSIONS: High level of neighborhood deprivation independently predicts prostate cancer mortality. This raises important clinical and public health concerns. Both individual-and neighborhood-level approaches are important in healthcare policies.