PM2.5 pollution in China has decreased dramatically, but how its health effects change is not clear. There are 120 old industrial cities in China, where the sources, composition, and health effects of PM2.5 may be significantly different with other cities. Huangshi, an old industrial city in central China, underwent intense green transformations from 2015 to 2018. In this study, we collected ambient PM2.5 samples in 2015 and 2018 at an urban site in Huangshi. The average PM2.5 concentration decreased from 83.44 ± 48.04 μg/m3 in 2015 to 68.03 ± 39.41 μg/m3 in 2018. However, the average volume-normalized dithiothreitol (DTTv) of PM2.5 increased from 1.38 ± 0.45 nmol/min/m3 to 2.14 ± 1.31 nmol/min/m3 and the DTT normalized by particulate mass (DTTm) increased from 20.6 ± 10.1 pmol/min/μg to 40.07 ± 21.9 pmol/min/μg, indicating increased exposure risk and inherent toxicity. The increased toxicity of PM2.5 might be related to the increased trace elements (TEs) concentrations. The positive matrix factorization and multiple linear regression methods were employed to quantify the contributions of emission sources to PM2.5 and DTTv. The results showed that the contribution of coal combustion, industry, and dust to PM2.5 decreased significantly from 2015 to 2018, while that of vehicle emission and secondary sources increased. Despite the decreased fraction of coal combustion and industry sources, their contribution to DTTv increased slightly, which was caused by the increased intrinsic toxicity. The increased intrinsic toxicity was possibly caused by increased TEs, such as Pb, Cu, and V. Besides, the contribution of vehicle emission to DTTv also increased. Overall, these results provide valuable insights into the effectiveness of controlling strategies in reducing particulate health impacts in old industrial cities, and stress the necessity of formulating toxicity-oriented controlling strategies, with special attention to TEs from coal combustion and industry sources as well as vehicle emissions.
