Assessing the Recent Estimates of the Global Burden of Disease for Ambient Air Pollution: Methodological Changes and Implications for Low- and Middle-Income Countries

摘要

The Global Burden of Disease (GBD) is a comparative assessment of the health impact of the major and well-established risk factors, including ambient air pollution (AAP), assessed by concentrations of PM2.5 (particles less than 2.5 microns) and ozone. Over the last two decades, major improvements have emerged for two important inputs in the methodology for estimating the impacts of PM2.5: The assessment of global exposure to PM2.5 and the development of integrated exposure risk models (lERs) that relate the entire range of exposures to PM2.5 to cause-specific mortality risks. As a result, the estimated annual mortality attributed to AAP increased from less than 1 million in 2000 to roughly 3 million for GBD in years 2010 and 2013, and to 4.2 million for GBD 2015. However, the magnitude of the recent change and uncertainty regarding its rationale have resulted, in some cases, in skepticism and reduced confidence in the overall estimates. To understand the underlying reasons for the change in mortality, we examined the estimates for the years 2013 and 2015 to determine the quantitative implications of alternative model input assumptions. We calculated that the year 2013 estimates increased by 8% after applying the updated exposure methods used in GBD 2015, and increased by 23% with the application of the updated lERs from GBD 2015. The application of both upgraded methodologies together increased the GBD 2013 estimates by 35% or about one million deaths. We also determined the impact of changes in demographics and assumptions about the threshold concentration. Since the global estimates of air pollution-related deaths will continue to change over time, a clear documentation of the modifications in the methodology and their impacts is necessary. In addition, there is need for additional monitoring and epidemiological studies to reduce uncertainties in the estimates for low-and medium-income countries.