Multicontaminant air pollution in Chinese cities

Abstract Objective To investigate multicontaminant air pollution in Chinese cities, to quantify the urban population affected and to explore the relationship between air pollution and urban population size. Methods We obtained data for 155 cities with 276 million inhabitants for 2014 from China's air quality monitoring network on concentrations of fine particulate matter measuring under 2.5 μm (PM2.5), coarse particulate matter measuring 2.5 to 10 μm (PM10), nitrogen dioxide (NO2), sulfur dioxide (SO2) and ozone (O3). Concentrations were considered as high, if they exceeded World Health Organization (WHO) guideline limits. Findings Overall, 51% (142 million) of the study population was exposed to mean annual multicontaminant concentrations above WHO limits – east China and the megacities were worst affected. High daily levels of four-contaminant mixtures of PM2.5, PM10, SO2 and O3 and PM2.5, PM10, SO2 and NO2 occurred on up to 110 days in 2014 in many cities, mainly in Shandong and Hebei Provinces. High daily levels of PM2.5, PM10 and SO2 occurred on over  146 days in 110 cities, mainly in east and central China. High daily levels of mixtures of PM2.5 and PM10, PM2.5 and SO2, and PM10 and SO2 occurred on over  146 days in 145 cities, mainly in east China. Surprisingly, multicontaminant air pollution was less frequent in cities with populations over 10 million than in smaller cities. Conclusion Multicontaminant air pollution was common in Chinese cities. A shift from single-contaminant to multicontaminant evaluations of the health effects of air pollution is needed. China should implement protective measures during future urbanization.


Introduction
Air pollution in cities is a major concern worldwide, irrespective of a country's level of development. In high-income countries, air quality has improved substantially since the 1970s; however, the adverse health effects of exposure to relatively low-level pollution remains a public concern. 1 In contrast, air quality in some middle-and low-income countries, such as China and India, has seriously deteriorated. 2 Before the 1920s, the main cause of urban air pollution in high-income countries was the rapid spread of coal-fired industry during the second phase of the Industrial Revolution. The major contaminants produced by coal combustion are particulate matter and sulfur dioxide (SO 2 ). After the1920s, a new source of air pollution emerged with the widespread use of the automobile, which emits particulate matter, nitrogen dioxide (NO 2 ), lead and other contaminants. However in some middle-and low-income countries, e.g. China, the development of coal-fired industries and increased automobile use have overlapped, which has resulted in the emission of a complex mix of air contaminants. 3,4 Most studies of the health effects of air pollution have focused on individual contaminants, such as particulate matter, NO 2 , SO 2 , ozone (O 3 ) and carbon monoxide, with each considered to have an independent impact. [5][6][7] However, in reality the urban atmosphere is never confronted with a single contaminant but is actually exposed to a complex mix of different contaminants at varying times of the day and year. Consequently, people are more likely to be exposed to a mixture of contaminants than to a single substance, the resultant impact on human health can be highly varied. 8 For instance, some contaminants (e.g. NO 2 and O 3 ) affect the respiratory system, some (e.g. particulate matter) affect the circulatory system and cause heart disease and others (e.g. SO 2 ) affect the skin and mucous membranes. Although few epidemiological studies have looked at the combined effect of several air contaminants, it can be assumed that they will have an impact on different parts of human body. For example, the combination of NO 2 and particulate matter pollution will affect both respiratory and cardiovascular systems. 5,6 As it can lead to these complex conditions, exposure to multicontaminant air pollution is important and should be quantified, especially in rapidly urbanizing developing countries where mixtures of contaminants are common. 4,9 Previous research has paid particular attention to understanding how specific contaminants affect public health in developing countries. Although important, this approach may underestimate the actual impact of urban air pollution on public health. In fact, there have been calls for a shift from a single-contaminant to a multicontaminant approach to countering the health effects of air pollution. 5 The aims of this study were: (i) to document the mixture of air contaminants in Chinese cities both annually and diurnally; (ii) to determine the proportion of the urban population affected by multicontaminant air pollution; and (iii) to investigate the relationship between the size of the urban population and the frequency of occurrence of high levels of multicontaminant air pollution.

Methods
We obtained data on air quality for 155 cities (including all 31 provincial capitals and 124 major prefectural cities) from China's urban air quality monitoring network, which reports concentrations of air contaminants under the newly upgraded ambient air quality standard GB3095-2012. For this study, we used hourly concentrations of fine particulate matter less than or equal to 2.5 μm in diameter (PM 2.5 ), coarse particulate matter with a diameter between 2.5 and 10 μm (PM 10 ), NO 2 , SO 2

and O 3 for the whole of 2014. To
Objective To investigate multicontaminant air pollution in Chinese cities, to quantify the urban population affected and to explore the relationship between air pollution and urban population size. Methods We obtained data for 155 cities with 276 million inhabitants for 2014 from China's air quality monitoring network on concentrations of fine particulate matter measuring under 2.5 μm (PM 2.5 ), coarse particulate matter measuring 2.5 to 10 μm (PM 10 ), nitrogen dioxide (NO 2 ), sulfur dioxide (SO 2 ) and ozone (O 3 ). Concentrations were considered as high, if they exceeded World Health Organization (WHO) guideline limits. Findings Overall, 51% (142 million) of the study population was exposed to mean annual multicontaminant concentrations above WHO limits -east China and the megacities were worst affected. High daily levels of four-contaminant mixtures of PM 2.5 , PM 10 , SO 2 and O 3 and PM 2.5 , PM 10 , SO 2 and NO 2 occurred on up to 110 days in 2014 in many cities, mainly in Shandong and Hebei Provinces. High daily levels of PM 2.5 , PM 10 and SO 2 occurred on over 146 days in 110 cities, mainly in east and central China. High daily levels of mixtures of PM 2.5 and PM 10 , PM 2.5 and SO 2 , and PM 10 and SO 2 occurred on over 146 days in 145 cities, mainly in east China. Surprisingly, multicontaminant air pollution was less frequent in cities with populations over 10 million than in smaller cities. Conclusion Multicontaminant air pollution was common in Chinese cities. A shift from single-contaminant to multicontaminant evaluations of the health effects of air pollution is needed. China should implement protective measures during future urbanization.
assess pollution levels and their potential impact on public health, we used guideline values for annual and daily ambient air quality provided by the World Health Organization (WHO; Table 1). 10 We averaged hourly concentrations to obtain annual means for all contaminants, 24-hour means for PM 2.5 , PM 10 and SO 2 and 8-hour means for O 3 . For the NO 2 concentration, we retained the hourly values. Finally, we determined how frequently annual and daily multicontaminant air pollution due to various combinations of three, four and five contaminants (  Table 1 for individual substances. We obtained the size of the population in each of the 155 cities, as reported in the 2010 census, from the National Bureau of Statistics of China. 11 In total, these cities accounted for 41.2% of China's urban population in 2010. The main variable of interest in our study was exposure to a high level of multicontaminant air pollution, which was defined as occurring when the concentration of a contaminant exceeded the relevant WHO value in Table 1. Annual exposure to multicontaminant air pollution was assessed for combinations of two or three contaminants and daily exposure was assessed for combinations of two, three, four or five contaminants (Table 2). To investigate the impact of urbanization on air pollution, we determined whether there was a correlation between the size of the urban population and the proportion of days in 2014 during which the concentration of specific contaminants exceeded WHO guideline values. For this analysis, cities were divided into five groups by population size, according to China's new urban size standard: 12 (i) less than 0.5 million; (ii) 0.5 to less than 1 million; (iii) 1 to less than 5 million; (iv) 5 to less than 10 million; and (v) 10 million or more. The correlation between the population size and the percentage of days in 2014 with a high level of multicontaminant air pollution was determined using nonlinear regression analysis.

Results
In total, 56 of the 155 cities analysed (36%) were exposed to mean annual concentrations of the contaminants PM 2.5 , PM 10 and NO 2 above WHO guideline values (Fig. 1). These cities had a combined population of 142 million out of a total study population of 276 million (i.e. 51%). In addition, all 155 cities were exposed to high annual concentrations of two-contaminant mixtures of PM 2.5 and PM 10 and 56 cities, with a total population of 142 million, were exposed to high annual concentrations of PM 2.5 and NO 2 and of PM 10 and NO 2 . The cities with high annual multicontaminant exposure to either (i) PM 2.5 , PM 10 and NO 2 ; (ii) PM 2.5 and NO 2 ; or (iii) PM 10 and NO 2 were mainly located in east China, specifically in Hebei, Henan, Jiangsu, Shandong and Zhejiang Provinces and in the megacities of Beijing, Guangzhou, Shenzhen and Tianjin (Fig. 2).

Daily multicontaminant exposure
Only two cities, Dongying and Linyi in Shandong Province, had mean daily concentrations of all five contaminants (i.e. PM 2.5 , PM 10 , SO 2 , O 3 and NO 2 ) above WHO guideline values for 11-15 days (3-4%) in 2014 (Fig. 3). Weifang and Zibo in Shandong Province were exposed to high daily concentrations of the five contaminants for 8-11 days (2-3%) in the year. Jining in Shandong Province, Wuhan in Hubei Province and Jiayuguan and Jinchang in Gansu Province were exposed to high daily concentrations for 4-8 days (1-2%; Fig. 4). Other cities had less than 4 days (1%) with high concentrations of all five contaminants.
Exposure to high mean daily concentrations of four contaminants was more common. In some locations, daily concentrations exceeded WHO guideline values for 73-110 days (20-30%)   Exposure to high mean daily concentrations of the three contaminants PM 2.5 , PM 10 and SO 2 was even more common: 110 cities with a total population of 173 million were exposed to this level of air pollution for more than 146 days (40%) in 2014 (Fig. 9). Those cities were mainly located in east and central China, particularly in Hebei, Henan, Shandong and Shanxi Provinces ( Fig. 10). In addition, exposure to high daily concentrations of mixtures of the following three-contaminant combinations were observed on 18-146 days    (Table 3). However, high daily concentrations of the three-contaminant mixtures of (i) PM 2.5 , O 3 and NO 2 , (ii) PM 10 , SO 2 and NO 2 , and (iii) NO 2 , O 3 and SO 2 were observed on less than 18 days (5%) in 2014 in major Chinese cities. Exposure to high daily concentrations of two contaminants was extremely common: 145 cities with a total population of 269 million were exposed to mean daily concentrations of PM 2.5 and PM 10 above WHO guideline values for more than 146 days (40%) in 2014 ( Fig. 11). High concentrations of the two contaminants PM 2.5 and SO 2 were also observed on more than 146 days (40%) in 116 cities with a total population of 184 million ( Fig. 12; available at: http://www.who.int/bulletin/volumes/96/4/17-195560) and high concentrations of PM 10 and SO 2 were equally frequently observed in 111 cities with a total population of 175 million

Population size
In general, daily multicontaminant air pollution was less frequent in cities with a population greater than 10 million than in smaller cities in our study. For example, the proportion of days in 2014 during which the mean daily concentrations of all five contaminants exceeded WHO guideline values was not significantly affected by population size in cities with fewer than 10 million inhabitants but the portion was substantially lower in cities with a population greater than 10 million (Fig. 17). Similarly, the frequency of exposure to high mean daily concentrations of four contaminants was comparable among cities with populations ranging from 0.5 to 10 million but was lower in cities with a population less than 0.5 million or greater than 10 million (Fig. 18). This variation was also observed for high mean daily concentrations of three contaminants: the frequency was similar in cities with populations rang-ing from 0.5 to 10 million but lower in those with a population less than 0.5 million or greater than 10 million (Fig. 19). For exposure to high daily  concentrations of two contaminants, there was no substantial variation in frequency among cities with a population less than 10 million, whereas the frequency was markedly lower in cities with a population greater than 10 million (Fig. 20). There was a significant inverse U-shaped relationship between the size of the urban population and the observed frequency of high mean daily concentrations of four contaminants (Fig. 18). In addition, there were inverse U-shaped relationships between population size and the frequency of high mean daily concentrations of three and two contaminants but the relationships were weaker ( Fig. 19 and Fig. 20).

Discussion
Although our study was based on data for only one year, it provides a snapshot of air pollution in major Chinese cities and demonstrates that multicontaminant air pollution was very common in 2014. These findings underscore the need to assess multiple air contaminant concentrations at the same time to obtain a more realistic picture of urban air quality and its potential impact on public health. Consequently, a change in air quality guidelines is required, with the establishment of guidelines on multi-contaminant mixtures. The globally recognized, ambient air quality guidelines produced by WHO were designed to help reduce the health effects of air pollution in 1987. They were based on a review of the scientific evidence and its implications. The guidelines, which were updated in 1997 and 2005, now specify daily and annual limits for five major ambient air contaminants. In addition, some regions and countries have established their own air quality standards. For instance, the European Union, Japan and the United States of America were quick to update their air quality guidelines, whereas some   middle-and low-income countries, e.g. China, established their own standards in response to high levels of pollution. However, all these guidelines and standards treat each contaminant in isolation or choose a single major contaminant as an indicator of air quality. For example, China uses an air quality index based on the maximum value of each individual contaminant's concentration to indicate air quality. 13 Multicontaminant ambient air pollution is also important for public health research at both the urban and regional level. In the past, very little attention has been paid to multicontaminant exposure and research efforts have primarily focused on the health effects of individual contaminants. Initially, the reason for this focus was the difficulty of evaluating the medical effects of exposure to several contaminants. In addition, there was little understanding that multicontaminant ambient air pollution is common. 5 However, without detailed research into the medical consequences of multicontaminant exposure, the disease burden will be underestimated. The influential Global Burden of Disease Study 2013 considered both ambient and household air pollution. 14 Still, the only ambient air contaminants included were particulate matter and ozone, no consideration was given to other contaminants. We recommend that research into air pollution and its health effects should pay more attention to multicontaminant ambient air pollution, especially in middle-and low-income counties where current pollution levels are often higher than in high-income countries. In particular, by devoting attention to multicontaminant mixtures, researchers could raise public awareness of the complex nature of ambient air quality and stimulate greater interest in air pollution prevention.
As a result of rapid urbanization during the last century, more than half of the world's population now lives in cities. 15 This rise in the urban population and the associated intensification of social and economic activity have had a substantial impact on urban air quality. Thus, urbanization and its effect on air quality are among the most important issues for achieving sustainable urban and regional development.
Researchers have studied the relationship between urbanization and typical air contaminants in both developed and developing countries. 9,16 For example, the concentration of the traditional air contaminant NO 2 has been observed to increase exponentially with population size, though the value of the exponent varies between locations. 16 In contrast, for PM 2.5 , the relationship between its  concentration and urban population size is much more variable across continents and countries. 9 In our study, we found an inverse U-shaped relationship between urban population size and the frequency of high daily concentrations of three contaminants, whereas other researchers have demonstrated no clear relationship. Furthermore, we discovered that a high level of multicontaminant air pollution was less common in cities with a population of more than 10 million than in smaller cities, which is contrary to general expectations that larger cities would be more polluted. The likely explanation is that large cities have implemented extensive environmental protection measures and that many polluting industries have been relocated to smaller cities. 3

Contaminación del aire con múltiples contaminantes en las ciudades chinas
Objetivo Investigar la contaminación atmosférica con múltiples contaminantes en las ciudades chinas, cuantificar la población urbana afectada y explorar la relación entre la contaminación del aire y el tamaño de la población urbana.         PM 2.5 : fine particulate matter less than or equal to 2.5 μm in diameter; SO 2 : sulfur dioxide. Note: The mean daily air contaminant concentration was classed as high if it exceeded the World Health Organization guideline value (Table 1).