广州市多污染物联合暴露的健康效应评估

本研究应用贝叶斯核机器回归(BKMR)分析了广州市2015~2018年大气主要污染物(SO₂、NO₂、PM₁₀、PM_2.5、CO和O₃)与非意外死亡之间的联合健康效应。结果表明,6种污染物对健康结局均存在较大影响。随着其它污染物固定百分位数浓度的增加,SO₂、O₃或NO₂的浓度从第25百分位数变化到第75百分位数导致的效应值的绝对值逐渐升高;相反,PM_2.5、CO或PM₁₀的浓度从第25百分位数变化到第75百分位数导致的效应值的绝对值逐渐减少。在累积滞后0~1d时,多种空气污染物的混合暴露对非意外死亡人数的影响效应值为正,与最低浓度相较,当所有污染物的浓度增至第90百分位数时,人数非意外死亡人数将增加21.98%(95%CI:3.04%,44.41%)。通过应用BKMR模型,本研究证实多种污染物的暴露可对广州市人群公共健康造成联合的综合影响。在污染物的防控治理上,不仅需要针对当地主要空气污染物,还需加强对多污染物暴露的防控。

Health risk assessment of exposure to multiple pollutants in Guangzhou

This study attempted to use the Bayesian kernel machine regression (BKMR) to analyze the health effects of multiple air pollutants (i.e, SO₂、NO₂、PM₁₀、PM_2.5、CO and O₃) on non-accidental mortality in Guangzhou during 2015~2018. We found that all the pollutants presented significant effects on non-accidental mortality. The absolute values of effect estimates associated with the change in SO₂, O₃ or NO₂ concentration from the 25th percentile to the 75th percentile gradually increased with the increment of the fixed percentile of other pollutants; On the contrary, the absolute values of effect estimates decreased gradually when the concentration of PM_2.5, CO or PM₁₀ changes from the 25th percentile to the 75th percentile. The cumulative effect estimates associated with exposure to multiple pollutants on non-accidental mortality at lag 0~1 day was positive. Compared with the lowest concentration, when the concentration of all pollutants increased to the 90th percentile, the number of non-accidental deaths will increase by 21.98% (95% CI: 3.04%, 44.41%). With the application of BKMR model, this study found that exposure to multiple pollutants could produce a combined impact on public health in Guangzhou. In terms of air pollution prevention and control, it is necessary to not only target the main local air pollutants, but also strengthen the policies on protecting the public from exposure to multi-pollutants.