Abstract Background and Aims According to the Global Burden of Disease, mortality attributed to chronic kidney disease (CKD) is steadily increasing worldwide, with an estimated increase of more than 30% since 2005. Among the candidates mentioned to explain this inflation, air pollution is attracting more and more attention. Although the link between air pollution and the incidence of CKD and kidney failure seems increasingly clear, there is currently very little data regarding the potential effect of exposure to air pollutants on mortality in dialysis patients. Beyond the lack of data in the literature, their interpretation is hampered by the conventional analysis of air pollutants on an individual basis, whereas ambient pollution actually corresponds to a complex combination of multiple pollutant exposure. Composite spatial indices that account for multi-exposition have recently been developed and are now recognized as relevant indicators for global air exposure assessment. The main objectives of this project are to study the association of the level of multi-exposure to air pollutants with all-cause and cause-specific mortality in dialysis patients. Method We included all incident dialysis patients in France between January 1, 2012 and December 31, 2020, identified in the REIN registry. Annual mean levels of exposure to fine particulate matter (PM10, PM2.5) and nitrogen dioxide (NO2) were extracted over the entire study period based on the data from the French ‘Institut National de l’Environnement Industriel et des Risques (INERIS)’. A composite environmental score was created from the first component of a standardized, annualized principal component analysis of exposure levels to the three pollutants and estimated at the municipal level of each patient residency. This score was used as a continuous and categorical indicator (highest levels corresponding to the most exposed areas). The association between this score and all-cause mortality in dialysis patients was analyzed by multivariate Cox models considering comorbidities, modalities of dialysis start, socio-environmental factors (population density, social deprivation and distance to the nearest dialysis center), and the center effect. These analyses were declined according to the cause of death (all-cause, cardiovascular, infectious or cancer) and to different pre-dialysis exposure lags (1, 2 or 3 years). Interaction studies and subgroup analyses were performed according to age, gender and level of autonomy. Results A total of 90,373 patients were included in this study (64% men, median age 72 years) for a median follow-up of 46 months. Over the study period, 44,242 deaths were observed (20% of cardiovascular origin). The mean annual levels of each air pollutant exposure were 11.5 (± 2.5), 18.0 (± 2.5) and 16.8 (± 7.0) μg/m3 respectively for PM2.5, PM10 and NO2. For each point increment in the environmental score, we observed a linear 1.6% increase in the risk of all-cause mortality (HR = 1.016 [1.007-1.025] p < 0.001). Compared with the lowest exposure areas (Q1), patients residing in quartile 3 and 4 municipalities had an increased risk of all-cause death (HRQ2 = 1.00 [0.96-1.04] p = 0.99, HRQ3 = 1.06 [1.02-1.10] p = 0.005, and HRQ4 = 1.09 [1.05-1.13] p < 0.001 for quartiles 2, 3, and 4 of the environmental score respectively – see Figure). A significant interaction was observed with the level of autonomy of the patients, the risk of death associated with environmental exposure being higher in partially or totally dependent patients compared to fully autonomous ones (p for interaction = 0.006). These results were similar for deaths from cardiovascular or infectious causes and consistent regardless of exposure lag and each air pollutant taken individually. Conclusion In total, this is to our knowledge the first study describing an association between environmental multi-exposure and the risk of all-cause or cause-specific death in dialysis patients on a national scale. These results argue for intensified efforts to limit air pollution, particularly for the highly vulnerable dialysis patient population.
Read full abstract