Health status and air pollution related socioeconomic concerns in urban China

Environmental pollution has attracted great attention in China due to its rapid development of industrialization and urbanization in recent years. The report of city smog has widely occurred in the mass media in China [1]. The social implication of air pollution has indicated an increasing number of severe problems of air pollution in China¹ and the health effects associated with air pollution. Air pollution may cause acute or chronic health problems including mild irritation to the upper respiratory tract, chronic respiratory system disease, heart disease, lung cancer, children’s acute respiratory infection and adults’ chronic bronchitis. In addition, air pollution exacerbates asthma, heart disease or lung disease, and increases the risk of death [2, 3]. Meanwhile, health effects caused by air pollution are significantly different for people with different socioeconomic statuses. Even though Beck argued that poverty was hierarchical and chemical smog was democratic [4], some researches have pointed out that individuals and groups with different socioeconomic statuses were exposed to air pollution at different levels and suffered from different health effects. These differences refer to environmental inequality and health inequality.

Some studies have examined the health inequalities associated with air pollution at regional level. For example, a study of six districts in Sao Paulo, Brazil, demonstrated that PM10 had less effect on respiratory mortality among older adults in areas with a higher proportion of college education populations and high-income families, and it had greater effect in areas where the proportion of the poor population was high [5]. Another study based on the city of Hamilton in Canada showed that air pollution exposure had a relatively large effect on acute mortality in the zones with a higher proportion of low educational attainment and high manufacturing employment [6]. In addition, a study of Hong Kong, China, found that air pollution had a greater effect on the mortality of people living in public rental than those living in private homes, and a greater effect on blue-collar workers than those who had never worked and white-collar workers [7]. Other studies focusing on developed areas, such as Rome and Norway, revealed similar conclusion that neighborhood socioeconomic deprivation could exacerbate mortality caused by air pollution [8, 9].

On the other hand, some studies have examined the health inequalities associated with air pollution at individual level. For example, a study of 20 cities in the United States showed that individual education significantly modified the relationship between PM10 and mortality; specifically, the higher education level of the individual the lower the effect of PM10 on mortality [10]. In addition, several studies about China found similar results that people with lower socioeconomic status normally experienced a higher health risk from air pollution, while people with middle or high socioeconomic status barely had a health risk since they had more ways to avoid air pollution [11]. Moreover, as the severity of air pollution increased, health disparities among people with different socioeconomic status would intensify [12]. On the contrary, a good natural environment could reduce health inequality [13].

There are two possible reasons that explain the stronger effect of air pollution among people in low socioeconomic class. First, the level of exposure to air pollution is higher among those living in low socioeconomic communities [14]. Pearce and Kingham claimed that in some countries or regions the socially deprived, the low income and the ethnic minority were exposed to a higher level of air pollution [15]. Furthermore, Schoolman and Ma found that townships with higher proportion of rural migrants were exposed to higher level of air pollution in Jiangsu province, China [16]. Second, compared to people in high socioeconomic status, populations with low socioeconomic status are more susceptible to air pollution. The susceptibility of people with low socioeconomic status is caused by health-related social, behavioral and psychological factors including poor health status (such as diabetes and obesity), addictions (such as smoking), other pollutant exposures (such as passive smoking), psychological stress, low intake of nutrition, and even genetic make-up [17].

Forastiere pointed out that the second reason (i.e. different susceptibility of individuals) was more convincible than the first one (i.e. differential exposure to air pollution) for explaining the stronger effect of air pollution among low socioeconomic groups [8]. This is because several researchers have found no significant difference in levels of exposure in air pollution among people with different socioeconomic status, and some even have reached the opposite conclusion that high socioeconomic groups experience high levels of air pollution. For example, Goodman et al. examined the traffic-related air pollution in London and found that even though the level of average air pollution concentrations was higher in low socioeconomic positions, reversed direction of association was expected to occur in central London area [18]. Similar results were found in studies by Crouse and Cesaroni et al. [19, 20]. People with a college education or living in high socioeconomic communities were more likely to be exposed to traffic-related air pollution. Havard et al. asserted that the relationship between air pollution exposure and deprivation was nonlinear and the medium level deprivation areas were the most exposed to air pollution [21]. Thus, the association between socioeconomic status and air pollution exposure was still under uncertainty and required further investigation [22]. However, it should be noted that several studies did not find a modification effect of socioeconomic status on air pollution [23], or they found just a partial impact that education modified the effect of air pollution but household income did not [6]. Some researchers even found opposite results, i.e. the effect of air pollution on mortality was higher among people with high socioeconomic status [24]. Therefore, Laurent et al. argued that the modification effect of socioeconomic status depended on the regional level at which socioeconomic characteristics were measured [17]. If socioeconomic characteristics were measured at city level, no modification effect was found; if at community level, the result was mixed; if individually measured socioeconomic characteristics were used, the result indicated disadvantaged subjects were affected more by pollution.

To sum up, although existing studies have found that people with different socioeconomic status experienced varying levels of exposure to air pollution and health effects, unanimous conclusion has not been reached. In addition, most of the previous studies have focused on developed countries or regions. Therefore, studies need to be further extended to developing countries like China, which is now at a stage of accelerated industrialization and urbanization. Study of air pollution and health inequality in China can provide new empirical evidence for the debate. Therefore, this study tried to extend the literature with the aim to explore whether there was a significant difference in air pollution exposure and health effects among different socioeconomic status individuals or communities in China. To be more specific, there are two sub-questions: (1) Is the level of exposure to air pollution different among communities with different socioeconomic status in China, and what happens to the level of air pollution exposure as the community’s socioeconomic status continues to improve? (2) Is the effect of air pollution on health associated with socioeconomic status? With the increase of socioeconomic status, either at community level or at individual level, does the effect of air pollution on health decrease?

First, we examined the relationship between community socioeconomic status, community air pollution and community health. As shown in Table 3, the community socioeconomic status was positively related to community air pollution. As the community socioeconomic status increased, the level of community air pollution rose simultaneously. However, the coefficient of squared community socioeconomic status was significantly negative (P < 0.01), which indicated a curved relationship between community socioeconomic status and community air pollution. In other words, it was possible that the relationship between community socioeconomic status and community air pollution was mediated by community socioeconomic status.

Figure 1 presents the predicted values of community air pollution (percentage of serious air pollution assessment in community samples) at different levels of community socioeconomic status based on the models in Table 3. It can be seen that as community socioeconomic status improved, air pollution in community continued to increase. However, when community socioeconomic status increased to a certain extent, air pollution in community declined as community socioeconomic status increased.

Additionally, Table 3 demonstrates a negative relationship between community air pollution and community health, i.e. when the level of air pollution in community increases, the health at community level will decline. Table 3 also shows the coefficient of the interaction term is 0.141 (P < 0.1), which means that health effects associated with community air pollution were mediated by community socioeconomic status. As community socioeconomic status increased, the effect of community air pollution on community health decreased. As Fig. 2 illustrates, among the communities with low socioeconomic status, the more serious the air pollution, the worse the health conditions in the community. However, as the socioeconomic status increased, the difference of health conditions among communities with different levels of air pollution shrunk. That is to say, air pollution had relatively small impact on health in communities with higher socioeconomic status. It was important to point out that in communities with very high socioeconomic status, the health status of communities with higher air pollution was better than that of communities with relatively low air pollution.

In addition, we also examined the relationship between self-rated air pollution and self-rated health, and the mediating role of individual socioeconomic status. Table 4 shows the results of the mixed effects model. Model 1 in Table 4 contained only control variables and model 2 added self-rated air pollution. According to the likelihood ratio test results of the two models, model 2 (χ²(1)=24.11, P = 0.000) was better than model 1, which indicated self-rated air pollution significantly affected self-rated health. The more severe the self-rated air pollution of the community where the individual is living, the worse self-rated health will be. In model 3, individual socioeconomic status variables were added. Likelihood ratio test results showed model 3 (χ²(1)=89.47, P = 0.000) was better than model 2, which indicated individual socioeconomic status was significantly related to self-rated health. Individuals with higher socioeconomic status were significantly less likely to have bad self-rated health status than those with lower socioeconomic status. In model 4 of Table 4, we have added an interaction term of individual socioeconomic status and self-rated air pollution. Likelihood ratio test results showed model 4 (χ²(1)=4.76, P = 0.029) was better than model 3, which indicated the health effects associated with self-rated air pollution were mediated by individual socioeconomic status. The coefficient of the interaction term was −0.247, which was statistically significant (P < 0.05). Thus, with the improvement of individual socioeconomic status, the impact of self-rated air pollution on self-rated health would be significantly reduced.

As Fig. 3 illustrates, among individuals with low socioeconomic status, those who rated air pollution as ‘serious’ were less likely to rate health as “very good” or “good” than those who rated air pollution as ‘not serious’. However, as the level of individual socioeconomic status increased, the difference of health effects caused by self-rated air pollution decreased. From Fig. 3, among individuals with high socioeconomic status, the predicted probability of self-rated health as ‘very good’ and ‘good’ among those who rated air pollution as ‘serious’ was very close to that among those who rated air pollution as ‘not serious’. Figure 3 also implies that individuals rating air pollution as ‘serious’ were more likely to report ‘moderate’, ‘bad’ and ‘very bad’ for health status than those rating air pollution as ‘not serious’, especially among those with lower socioeconomic status. Similarly, among individuals with higher socioeconomic status, the predicted probability of self-rated health as ‘moderate’, ‘bad’ and ‘very bad’ among those who rated air pollution as ‘serious’ was very close to that among those who rated air pollution as ‘not serious’. In short, the level of individual socioeconomic status was negatively correlated to the health effects associated with self-rated air pollution.

This study found a non-linear relationship between community air pollution and community socioeconomic status in urban China. On the one hand, as community socioeconomic status improves, air pollution in the community is also on the rise, which, to some extent, indicates air pollution is a ‘byproduct’ of economic development of the society. This finding is consistent with some previous research. For example, in cities such as London and Rome, communities with high socioeconomic status had a higher level of exposure to air pollution [18, 20]. However, when the community socioeconomic status was raised to a certain level, the relationship between it and community air pollution began to reverse. In other words, as community socioeconomic status further increased, community air pollution would decrease. It implied that air pollution exposure was the highest among the communities with moderate socioeconomic status [21], while communities with lowest or highest socioeconomic status experienced relatively less air pollution.

In general, we can classify the relationship between community socioeconomic status and air pollution into five categories: low community socioeconomic status - low community air pollution (Class 1), moderate community socioeconomic status - moderate community air pollution (Class 2), high community socioeconomic status - high community air pollution (Class 3), high community socioeconomic status - moderate air pollution (Class 4) and high community socioeconomic status - low community air pollution (Class 5). At present, most urban areas in China are at the stage of rapid socioeconomic development and increasing air pollution (Class 1). Whether these urban areas transform from Class 2 to Class 3 or from Class 2 to Class 4 depends on changes in the economic development mode and investment in air pollution control. Additionally, some cities are at the stage of high socioeconomic status and high air pollution (Class 3), such as Beijing and Tianjin. With the change of economic development mode and the increase of government’s investment in air pollution control, air pollution in these cities like Beijing will continue to decline in the future (from Class 3 to Class 4). Finally, some cities in China have higher socioeconomic status but lower air pollution (Class 4 or Class 5), such as Shenzhen and Xiamen.

The factors to be considered in explaining the relationship between community socioeconomic status and community air pollution include economic structure, industrialization patterns, urbanization patterns, the natural environment and individual choices. While higher socioeconomic groups tend to prefer communities with better air quality, this is not the case in most of China’s cities. Now when urban residents choose their living location, they consider work opportunity, educational resources, medical resources and transportation more than air quality. Because of the homogeneity of density of air pollution in a city, such as PM2.5 [55, 56], individuals have to move to another city if they want to live in a community with better air quality. Considering the communities with serious air pollution are usually featured by highly developed economics, abundant educational and medial resources, and convenient transportation in China, people with high socioeconomic status do not demonstrate strong selection preference for high-level air quality. Therefore, high socioeconomic groups experience higher level of air pollution exposure in China. However, if air pollution becomes worse or its health risks are perceived to become higher, higher socioeconomic people are likely to make air quality an important consideration for relocation, such as moving from air polluted areas to areas with good air quality. If this becomes the case, the result will be that higher socioeconomic people will live in communities with good air quality, while the lower socioeconomic people will live mostly in poor air quality communities.

In addition, this study also found that there were significant inequalities in health effects associated with air pollution. With the improvement of socioeconomic status, the health effects associated with air pollution had been declining. As a result, the differences in self-rated health among people living in areas with severe air pollution and those living in areas with less polluted air had been gradually reduced. To be more specific, the health effects associated with air pollution on lower socioeconomic people was significantly greater than that of higher socioeconomic people. The result may be explained by vulnerability and avoidance of people with different socioeconomic status [14]. First, compared with higher socioeconomic people, the health status of lower socioeconomic groups is relatively poor [57], making them more sensitive to severe air pollution and more vulnerable to health damage. Second, people with lower socioeconomic status have significantly less access to good quality medical services than those with higher socioeconomic status [58], such that they cannot receive medical services in a timely manner when air pollution aggravates and affects their health. Third, people with low socioeconomic status usually have relatively low capability of prevention. Even under the same level of air pollution, people with low socioeconomic status are exposed into air pollution more due to work environment (for example, outdoor) and indoor environment (for example, the low capability of protection against air pollution). Lastly, because of the low educational background, people with low socioeconomic status usually lack the knowledge of health and environmental pollution as well as the health effects caused by air pollution. Therefore, they are more likely to have lower level of awareness of self-protection.

Because people with low socioeconomic status are less likely to be well protected from air pollution, they may experience more health effects than those with high socioeconomic status. Given the lack of capacity of lower socioeconomic groups to avoid the risk of air pollution, the public service of air pollution prevention provided by the government is very important. At present, when the air is heavily polluted, the main task of the government is to reduce pollutant emissions. Policies related to air pollution risk management are seriously lagging behind, resulting in lower socioeconomic people (such as children from poor families, the elderly, patients, etc.) failing to receive the related public services. Therefore, it is imperative for the regions with more serious air pollution to urgently formulate public policy to mitigate the negative impact of air pollution exposure. The main goals of public policy makers should be to increase the government’s financial investment to provide masks, air purification equipment and other protective means and information consulting services for lower socioeconomic people. In addition, the government should continue to improve existing health care policies, enhance the accessibility of high-quality medical services to the lower socioeconomic people, and minimize the health damage caused by air pollution.

This study has some limitations as follows. First, due to the limited number of air quality monitoring stations in China, only the city-level data was available rather than community-level data. Therefore, air pollution was measured in a subjective way in this study. Although some studies suggested that subjectively perceived community characteristics (such as perception of community air quality) played an important role in health research [59], and some studies also found that perceived air pollution was significantly correlated with official monitoring data [60], more effort should be on the relative role of objective air pollution and subjectively perceived air pollution in influencing health conditions. Additionally, whether the differences in the impact of air pollution on the health status of different socioeconomic people are caused by the differences in perceptions of air pollution also requires further study. Second, since this study used cross-sectional data, there is no in-depth study on causal mechanisms between socioeconomic status, air pollution and health status. Future studies may consider using panel data to examine how changes in socioeconomic status affect exposure to air pollution and how changes in air pollution exposure affect health. Third, although the self-rated health used in this study contained the evaluation of one’s own overall health status as well as WHO’s definition of health, the impact of air pollution on some objective health indicators needs further investigation. The association between air pollution and different health indicators, and whether this relationship is modified by socioeconomic status also requires further investigation. Finally, since the sample of this study only included urban residents over the age of 45, the conclusions of this study cannot be generalized to the large population. We need to extend the study of the health effects associated with air pollution in the general population and the mediating role of socioeconomic status. Moreover, we still need to elaborate the difference in different subgroups such as age group differences.

This study used nationwide multilevel data to investigate how socioeconomic factors matter in the relationship between air pollution and health status in China. It verified the different levels of exposure to air pollution and inequality in health effects among different socioeconomic groups in China. The findings indicated a nonlinear relationship between community socioeconomic status and community air pollution, and the highest level of the relationship was found in communities with moderate socioeconomic status. It was also found that air pollution had the greatest impact on the health of the lower socioeconomic groups. With the increase of socioeconomic status, the effect of air pollution on health was decreased. Therefore, it is imperative for the government to urgently formulate public policies to enhance the ability of the lower socioeconomic groups to circumvent air pollution and reduce the health damage caused by air pollution. On account of the dynamic evolution of socioeconomic development and level of air pollution, the modification effect on the relationship between environment and health from the socioeconomic status is changing. Therefore, it is important to further investigate and analyze the inequality of exposure to environmental pollution and health effects associated with air pollution in China.