Long-term exposure to air pollution at residential and workplace addresses and breast cancer risk: A case-control study nested in the French E3N-Générations cohort from 1990 to 2011

Background: An increasing evidence links air pollution to breast cancer (BC) risk. Yet, pollutant exposure estimates at the workplace location in pollution exposure assessment have not been considered. Objectives: This study investigates the association between particulate matters (PM_2·5, PM₁₀) and nitrogen dioxide (NO₂) atmospheric concentrations (1990–2011), at the women's residential and workplace locations, and BC risk. Methods: This case-control study of 2419 BC cases and 2984 controls, was nested in the French prospective E3N cohort. The annual mean PM_2·5, PM₁₀ and NO₂ concentrations were estimated using a Land Use Regression model (50 m x 50 m resolution) and assigned to the women's geocoded residential and workplace locations, from cohort recruitment to their index date (date of case diagnosis). Odds ratios (OR) and 95 % confidence intervals (CI) were estimated using multivariate logistic regression models. Results: An increased BC risk was observed for a 10 µg/m³ increase of the 1990–2011 average PM_2·5 concentration estimates (OR=1·28; CI 1·00, 1·63). An increased risk was suggested for a 10 µg/m³ increase for PM₁₀ (OR=1·09; CI 0·92, 1·30) and NO₂ (OR=1·05; CI 0·97, 1·13). No effect modification by menopausal status, nor difference by hormone receptor status were observed. Discussion: This study is the first to estimate BC risk and long-term air pollutant exposure from both, residential and workplace location histories. Results suggest that residential PM_2·5, PM₁₀ and NO₂ concentrations are strongly correlated with workplace ones, indicating that residential data may serve as proxy for overall exposure. Future studies should consider exposure during commuting.