Abstract
BACKGROUND: Within the Multiethnic Cohort (MEC), we examined the association between air pollution and mortality among African American, European American, Japanese American, and Latina American women diagnosed with breast cancer. METHODS: We used a land use regression (LUR) model and kriging interpolation to estimate nitrogen oxides (NO(x ,) NO(2)) and particulate matter (PM(2.5), PM(10)) exposures for 3,089 breast cancer cases in the MEC, who were diagnosed from 1993 through 2013 and resided largely in Los Angeles County, California. Cox proportional hazards models were used to examine the association of time-varying air pollutants with all-cause, breast cancer, cardiovascular disease (CVD), and non-breast cancer/non-CVD mortality, accounting for key covariates. RESULTS: We identified 1,125 deaths from all causes (474 breast cancer, 272 CVD, 379 non-breast cancer/non-CVD deaths) among the 3,089 breast cancer cases with 8.1 years of average follow-up. LUR and kriged NO(X) (per 50 ppb) and NO(2) (per 20 ppb), PM(2.5) (per 10 µg/m(3)), and PM(10) (per 10 µg/m(3)) were positively associated with risks of all-cause (Hazard Ratio (HR) range = 1.13-1.25), breast cancer (HR range = 1.19-1.45), and CVD mortality (HR range = 1.37-1.60). Associations were statistically significant for LUR NO(X) and CVD mortality (HR = 1.60; 95% CI: 1.08-2.37) and kriged NO(2) and breast cancer mortality (HR = 1.45; 95% CI 1.02-2.07). Gaseous and PM pollutants were positively associated with breast cancer mortality across racial/ethnic group. CONCLUSION: In this study, air pollutants have a harmful impact on breast cancer survival. Additional studies should evaluate potential confounding by socioeconomic factors. These data support maintaining clean air laws to improve survival for women with breast cancer.