1547

Background: There is growing interest in the relationship between exposure to greenness and health. However, very few studies have evaluated the potential modification of this relationship by the degree of urbanity. We aimed to examine the association between exposure to greenness and mortality in a French cohort, according to urban and rural areas. Methods: We used data from the population-based Gazel cohort with a follow up of 26 years, which enrolled 20,625 participants in 1989. We assigned the satellite-derived Normalised Difference Vegetation Index (NDVI) in different buffers (100, 300, 500 and 1000m) around participants’ residential addresses during follow up. We used the extended Cox proportional hazard model to estimate the association between long-term exposure to greenness and natural mortality with a 10-year lag, after adjusting for relevant covariates including sex, age, alcohol use, smoking status, BMI⋯ Stratified analyses were conducted by urban-rural areas and sex. Results: Among the 19,851 participants eligible in this study [mean age at enrollment 43.7 years (SD 3.5), 73% males], we registered 1847 deaths. Each 0.1 increase in NDVI at 100m was associated with 6% higher mortality rate regardless the type of area [Hazard Ratio (HR)=1.06, 95%CI: 1.03-1.10]. Within population subsets, we found 23% higher mortality rate in rural areas (HR=1.23, 95%CI: 1.15-1.31), but 2% lower mortality rate in urban areas (HR=0.98, 95%CI: 0.93-1.04), with a significant difference of the estimates (P<0.001). These urban-rural differences were slightly amplified by sex, with higher HRs for males in rural areas and lower in urban areas. Results were similar for the other buffers.Conclusions: Greenness exposure was associated with higher mortality risk in rural areas with some indications of a potentially protective association in urban areas. Our study suggests a potential effect modification by urban-rural areas, which may provide relevant information for future research and policy developments.

2201

Background: Fine particulate matter (PM_2.5) is associated with mortality, and recent evidence identified black carbon (BC) as one component of PM_2.5 that may at least partly explain the health effects. Yet, this remains poorly documented. This study aimed to examine the association between long-term exposure to BC and mortality in a population-based French cohort. Methods: We used data from the Gazel cohort collected between 1989-2015, including geocoded residential address history. We used land use regression models with temporal extrapolation to estimate the exposure to BC and PM_2.5 for 19,850 participants. We used extended Cox models with attained age as time-scale and time-varying average exposure to BC, adjusted for relevant covariates including sex, smoking status and cumulative pack-years, and including a 10-year lag, to estimate the association between long-term exposure to BC and all-cause and cardiovascular mortality. To handle confounding by PM_2.5, we regressed BC against PM2.5 and used the residuals as the exposure variable in a sensitivity analysis. Separate analyses by sex and smoking status were also done to examine effect modification for all-cause mortality. Results: The median long-term BC exposure was 2.34 10^-5/m (inter-quartile range (IQR): 1.09). We found a significant association between BC and all-cause mortality (n=1794) using long-term average BC and residuals, with respective hazard ratios (HR) of 1.10 (95%CI: 1.04-1.17) and 1.19 (1.11-1.26) per IQR increase. We found a similar association between BC and cardiovascular mortality (n=271) with a HR of 1.15 (0.99-1.35). The association with all-cause mortality slightly varied across population subsets with HRs of 1.10 (1.03-1.18) and 1.09 (0.96-1.24) for men and women, and 1.10 (0.97-1.25) and 1.14 (1.07-1.23) for never- and ever-smokers, respectively.Conclusions: We found positive associations between long-term exposure to BC and increased mortality, reinforcing the emerging evidence that BC is a harmful component of PM.