Comparison of oxidative potential of fine particles at urban and rural sites

Abstract

Exposure to ambient fine particulate matter (PM2.5) has been associated with various adverse health effects (WHO., 2013). PM2.5 mass concentration is the most used exposure metric. However, PM2.5 mass concentration cannot fully explain the toxicity of PM2.5 because their multiple physicochemical factors could lead to adverse health effects. (Dergham et al., 2015). Several studies reported that PM2.5 concentration could lead to a misunderstanding of PM2.5 health risks by poor prediction of exposure level (Abrams et al., 2017, Bates et al.,2019). Shi et al. (2016) reported increased health risk at mass concentration below WHO air quality guideline (<15 μg/m3). The oxidative potential (OP), the ability of particle to induce oxidative stress, has been proposed as a more relevant health predictor for respiratory and cardiovascular diseases than PM2.5 mass concentration (Delfino et al., 2011, Bates et al.,2019). Therefore, if OP of PM2.5 can be predicted by their chemical components, predicted OP could complement PM2.5 mass concentration, providing a better understanding of health risks. In this study, multiple linear regression (MLR) model for OP was developed by using chemical components of PM2.5 in various environments. The ambient PM2.5 samples were collected at three different sites (urban Beijing (China) and Gwangju (Korea) in the winter of 2018, 2019, 2020 and summer of 2019 and rural Gimje (Korea) in the summer of 2020 and winter of 2021). The OP was measured by the cell-free dithiothreitol (DTT) assay. The average PM2.5 mass concentrations were 39.7, 27.1, and 25.8 µg/m3, at Beijing, Gwangju and Gimje sites, respectively. And the corresponding volume-normalized OP (OP-DTTv) were 2.204, 1.293, and 0.571 nmol/min/m3. Although PM2.5 concentrations were similar, OP-DTTv was significantly higher at Gwangju site than at Gimje site. This result derived from high mass normalized OP (OP-DTTm) at urban site. Seasonality in OP-DTTm was observed at Beijing (Summer>Winter) and Gimje sites (Winter>Summer). Principal components analysis (PCA) indicated that seasonal variation in OP-DTTm was mainly associated with water-soluble organic carbon (WSOC) at Beijing site. Meanwhile Cr, Zn, Pb, and Mn affected the seasonal variation in OP-DTTm at Gimje site. Despite low PM2.5 concentration (<15 µg/m3), relatively high OP-DTTv was observed during the days with high OP-DTTm due to increased organics and metals. Depending on the chemical composition, OP-DTTv could be high even at low PM2.5 concentration. To develop a prediction model for OP-DTTv, MLR was conducted using PM2.5 chemical components at Beijing, Gwangju and Gimje sites. Developed model suggested that WSOC, water-insoluble organic carbon (WISOC), Mn, and Zn can predict 50% of the variability of OP-DTTv at three sites. This model could provide a better explanation for regional differences in OP than PM2.5 mass concentration itself.