Asian Dust and Daily Mortality in East Asian Cities: Analysis Using the European Union Reference Method for Decomposing PM₁₀

Chris Fook Sheng Ng,Aurelio Tobías,Jorge Pey,Sara Basart,Massimo Stafoggia,Satbyul Estella Kim,Xerxes Seposo,Lina Madaniyazi,Ho Kim,Haidong Kan,Shih Chun Pan,Yue Guo,Xavier Querol,Masahiro Hashizume
DOI: https://doi.org/10.1289/isee.2023.op-172
2023-01-01
Abstract:BACKGROUND AND AIM: The association between desert dust and daily mortality has been investigated previously in East Asia, but with a different analytical approach and methods for quantifying dust exposure, rendering it difficult to compare results with other regions. We conducted a study to examine the association using the European Union (EU) Reference Method (Directive 2008/50/EC) to identify desert dust events and quantify exposure as a continuous measurement of particulate matter with diameters of 10 micrometers or less (PM₁₀). METHOD: We collected data on daily mortality (total respiratory and cardiovascular), PM₁₀, and average temperature from four cities – Beijing, Seoul, Fukuoka, and Taipei – with a period of 4-6 years between 2011-2017. Desert dust days were identified using aerosol maps (NAAPS-NRL), back trajectories of air masses (HYSPLIT) and reanalysis product (MERRA-2). We estimated the background level of PM₁₀ and quantified PM₁₀ concentrations by desert and non-desert sources. City-specific Poisson regression models with distributed lags were used to estimate the associations between mortality and source-specific PM₁₀ (by 3 sources: desert and non-desert PM₁₀ on days with or without dust events). RESULTS: Cities located closer to the desert areas had more dust events, ranging from 45% of the days in Beijing to 10% in Taipei. The exposure response curves tend to be more curvilinear for cities located closer to desert areas. We observed associations between mortality and PM₁₀ originating from desert dust notably in Beijing. The largest relative risks estimated for total, respiratory and cardiovascular mortality were 1.05 (95% confidence interval: 1.01, 1.10), 1.20 (1.06, 1.36), and 1.08 (1.02, 1.14), respectively, in comparison to the risk at the 1st percentile (3 microgram/m). Local PM₁₀ were also associated daily mortality but the association varied by cities. CONCLUSIONS: PM₁₀ from desert dusts is an independent risk factor for mortality in populations living near arid areas affected by desert dust.
What problem does this paper attempt to address?