A bird's eye view of the air pollution-cancer link in China.

Air pollution in China comes from multiple sources, including coal consumption, construction and industrial dust, and vehicle exhaust. Coal consumption in particular directly determines the emissions of three major air pollutants: dust, sulfur dioxide (SO(2)), and nitrogen oxide (NOx). The rapidly increasing number of civilian vehicles is expected to bring NOx emission to a very high level. Contrary to expectations, however, existing data show that the concentrations of major pollutants [particulate matter-10 (PM10), SO(2), and nitrogen dioxide (NO(2))] in several large Chinese cities have declined during the past decades, though they still exceed the national standards of ambient air quality. Archived data from China does not fully support that the concentrations of pollutants directly depend on local emissions, but this is likely due to inaccurate measurement of pollutants. Analyses on the cancer registry data show that cancer burden related to air pollution is on the rise in China and will likely increase further, but there is a lack of data to accurately predict the cancer burden. Past experience from other countries has sounded alarm of the link between air pollution and cancer. The quantitative association requires dedicated research as well as establishment of needed monitoring infrastructures and cancer registries. The air pollution-cancer link is a serious public health issue that needs urgent investigation.

Serious air pollution, especially the frequent haze (also called Wu Mai in Chinese) affecting northern China in recent years, has attracted global attention. The widespread fear of this modern day threat of public health is not without precedent. The London Smog Episode in 1952, the biggest air pollution event in the history of the United Kingdom, was reported to be associated with approximately 12,000 premature deaths[1]. According to the latest assessment of global disease burden, 3.2 million people died from air pollution in 2010, of which 2.1 million were from Asia[2]. Accumulating evidence suggests that air pollution is associated with increased risk of cancers, including lung cancer[3]–[9], nasopharyngeal cancer[10]–[12], breast cancer[13]–[16], lymphohematopoietic cancer[17], [18], and bladder cancer[19],[20]. Outdoor air pollution and particulate matter (PM) from outdoor air pollution were established as class I carcinogens by the International Agency of Research on Cancer (IARC)[21].

Nevertheless, because cancer has a long latency (averaging 30 years) and individual exposure has often been inaccurately measured[22], the air pollution–cancer (APC) link has not gained the same widespread attention as other relationships, such as the link between tobacco use and lung cancer. Frequently reported haze episodes around the world have recently brought the APC link back into the spotlight, especially in China. In January 2013, a hazardous dense haze covered 1.4 million square kilometers of China and affected more than 800 million people[23]. A heavy haze that winter shrouded northern and eastern China, reducing visibility to less than 50 meters in some regions[24].

Because of fast economic growth over the past three decades, China has become the world's second largest economy in terms of gross domestic product (since 2010) and the world's largest energy consumer (since 2009)[25]. Although this rapid economic development has improved living conditions, it also introduced a health threat caused by air pollution. In this paper we systematically analyze archived data on air pollution in China and discuss the potential association with cancer burden.

Trends of Major Air Pollution Sources and Pollutant Emission in China

Modern-day air pollution is much more complex than that in the past, such as that in the London Smog in the 1950s. This is because it is caused by multiple sources, including coal consumption, construction and industrial dust, and vehicle exhaust[25],[26]. Coal is the main source of energy in China. In 2012, coal accounted for 67% of the country's total energy consumption, and the coal consumption in China accounted for 50% of the world's total coal consumption in that year[27]. In the same year, China's production capacity of cement and steel has reached 2.20 billion tons and 0.95 billion tons, respectively ()[27]. The number of civilian vehicles has increased rapidly in China, from 1.36 million in 1978 to 109.33 million in 2012 ()[27],[28].

Figure 1.

Trends of annual coal consumption, cement production, steel production, and number of civilian vehicles in China from 1978 to 2012.

Data were incomplete for coal consumption due to the inconsistent data reported from different yearbook. Only data in the latest yearbook was selected. The left vertical axis represents coal consumption, cement production, and steel production. The right vertical axis represents civilian vehicles. Sources of data: China Statistical Yearbook 1983-2013[27],[28],[31],[52] and China Energy Statistical Yearbook 2012[32].

Unprecedented industrial development, overly coal-dependent energy consumption, and rapidly increasing number of vehicles have inevitably caused air pollutant emission to rise in China. Dust emission has decreased since 1997, the year in which the first emission standard of air pollutants in China was released[29], and sulfur dioxide (SO2) emission has decreased since 2006, the year in which national management of industrial desulfurization was piloted[30]. Nevertheless, dust and SO2 emission is still high, with level of 12.34 million tons and 21.18 million tons in 2012, respectively (). Nitrogen oxide (NOx) emission has steadily increased since monitoring was initiated in 2006, reaching a level of 23.38 million tons in 2012 ().

Figure 2.

Trends of annual emission of sulfur dioxide (SO2), nitrogen oxide (NOX), and dust in China from 1989 to 2012.

Dust emission included both the industrial dust and dust from the living environments, but only the industrial dust was reported in 1995 and 1996. NOx emission was just monitoring since 2006. Sources of data: Chinese Environment Statistical Yearbook 1998-2012[53], [54] and Chinese National Environment Statistical Report 1995-2012[55],[56].

Relationship Between Major Air Pollution Sources and Pollutant Emission in China

Although the annual emissions of dust, SO2, and NOx do not appear to parallel with the three potential major pollution sources at the national level, the relationship is clear at the province level. In Hebei Province, both dust emission (1.32 million tons) and NOx emission (1.80 million tons) were higher than those in any other provinces in 2011[31]. At the same time, Hebei Province also produced the largest amount of steel (0.19 million tons)[31]. In 2011, Shandong Province had the highest SO2 emission (1.83 million tons), consumed the largest amount of coal (0.39 billion tons), and produced the largest amount of cement (0.15 billion tons)[31],[32]. By contrast, Tibet had the lowest emissions of the three air pollutants because it has the fewest industries and civilian vehicles ().

Figure 3.

Scatter diagram of four air pollution sources and emissions for three major pollutants in 31 provinces/municipalities/autonomous regions in China in 2011.

Red dot in each sub-graph represents the level of air pollution sources for one province/municipality/autonomous region, and the blue dot represents the corresponding level of air pollutant emission. Sources of data: China Statistical Yearbook 2012[31] and China Energy Statistical Yearbook 2012[32].

The scatter diagrams in and show the relationship between four major pollution sources and the emissions of the three major pollutants in 31 provinces/municipalities/autonomous regions in China in 2011. The data show that the emissions of all three major pollutants were significantly associated with the four pollution sources, especially coal consumption[31],[32]. Indeed, coal consumption and NOx emission had the strongest positive correlation (r = 0.944), and coal consumption was also positively correlated with SO2 emission (r = 0.917) and dust emission (r = 0.839). Furthermore, the number of civilian vehicles had a strong positive correlation with NOx emission (r = 0.746) but not with SO2 emission (r = 0.574) or dust emission (r = 0.432). These results suggest that coal consumption may still be the most important pollution source in current China. As the number of civilian vehicles increase, NOx emission will increase more significantly compared with SO2 emission and dust emission.

Figure 4.

Relationship between four air pollution sources and emissions of three major pollutants in 31 provinces/municipalities/autonomous regions in China in 2011.

Each blue dot in each sub-graph represents one province/municipality/ autonomous region with corresponding level of air pollution source and air pollutant emission. The linear relationship between air pollution source and air pollutant emission is presented with a formula, and the coefficient of determination (R2) for the relationship is also calculated. Sources of data: China Statistical Yearbook 2012[31] and China Energy Statistical Yearbook 2012[32].

Trends of Annual Mean Concentrations of Major Air Pollutants in China

Although there is no comprehensive national data to reflect the overall level of air pollution in China, annual monitoring data from selected major cities can nonetheless give us a general estimate of the concentrations of particulate matter-10 (PM10), SO2, and NO2 in the country. In Beijing, Shanghai, and Tianjin, the annual mean concentrations of all three pollutants decreased from 2003 to 2012. However, the annual mean concentrations of PM10 in the three cities still exceeded the class II standard (70 µg/m3) in China in 2012 ()[27]. The trend for NO2 was similar to PM10. Although the annual mean concentration of SO2 in all the three cities was in compliance with the class II standard (60 µg/m3), the concentrations still exceeded the class I standard (20 µg/m3) ()[27].

Figure 5.

Trends for the annual mean concentrations of PM10, SO2, and NO2 in Beijing, Shanghai, and Tianjin from 2003 to 2012.

Sources of data: China Statistical Yearbook 2004-2013[27], [52].

By contrast to Beijing, Shanghai, and Tianjin, other cities in China did not experience a similar decrease in pollutant concentration. Between 2003 and 2012, the annual mean concentrations of pollutants actually increased in several cities. As shown in , PM10 concentration increased in three cities (Chengdu, Shenyang, and Urumqi), SO2 concentration increased in five cities (Zhengzhou, Nanjing, Shenyang, Lhasa, and Changchun), and NO2 concentration increased in eight cities (Nanjing, Wuhan, Kunming, Chengdu, Haikou, Zhengzhou, Urumqi, and Changchun). In 2012, the annual mean concentrations of NO2 and PM10 exceeded the national class II standards in six and three cities, respectively ()[27]. These results suggest that NO2 concentration may be more difficult to control than PM10 or SO2. Government regulations must better emphasize the control of NO2 concentration in the future.

Figure 6.

Changes in the annual mean concentrations of PM10, SO2, and NO2 in 14 selected cities in China in 2003 and 2012.

The blue and red bars represent the annual mean concentrations of air pollutants in 2003 and 2012, respectively. And the green bar represent the change of annual mean concentrations of air pollutant between the two years. A green bar with negative value indicates a decrease in the level of air pollutant, and a positive value indicates an increase. Sources of data: China Statistical Yearbook 2004-2013[27], [52].

In 2013, China began to monitor the concentration of PM2.5 in 74 selected cities[33]. As shown in , the monthly mean concentration of PM2.5 dropped to its lowest level in July (40 µg/m3) but increased to high levels by December (130 µg/m3). The concentrations of PM10, SO2, and NO2 followed similar trends (). This apparent seasonal distribution of pollutants is probably due to coal burning for heat during winter months. Although we were unable to find detailed information on the distribution of pollutants in northern and southern China, we expect a difference would certainly exist between these locations, as was reported in a recent study that examined the impact of sustained exposure to air pollution on life expectancy around the Huai River in China[34].

Figure 7.

Trends of the monthly mean concentrations (µg/m3) of PM2.5, PM10, SO2, and NO2 from January to December 2013 in 74 cities with monitored air quality in China.

Sources of data: Monthly reports of air quality monitoring in 74 cities in China (January to December in 2013)[33], [57].

Relationship Between Concentrations of Major Pollutants and Pollutant Emission in China

Available data from 31 major cities in China in 2012 showed that the concentrations of PM10, SO2, and NO2 were not linearly associated with the emissions of the corresponding air pollutants ()[27]. However, there was a weak exponential relationship between the concentrations and the emissions of these air pollutants (). The relationship between annual mean concentration of SO2 and SO2 emission was stronger than that between concentration of PM10 and dust emission and that between the concentration of NO2 and NOX emission ().

Figure 8.

Scatter diagram (A) and relationship (B) between annual mean concentrations and emissions of PM10, SO2, and NO2 in 31 major cities in China in 2012.

Red dot in each sub-graph of part A represents the level of air pollutant concentration for one city, and the blue dot represents the corresponding level of air pollutant emission. Each blue dot in each sub-graph of part B represents one city with corresponding level of air pollutant concentration and air pollutant emission. The exponential relationship between air pollutant concentration and air pollutant emission is presented with a formula, and the coefficient of determination (R2) for the relationship is also calculated. Sources of data: China Statistical Yearbook 2013[27].

On one hand, if this exponential relationship was real, further studies would be needed to confirm whether the concentrations of the pollutants would reach a plateau as the emissions increased and where the plateau would be. On the other hand, the unexpected weak relationship between pollutant concentrations and pollutant emissions in these cities could be a result of where emissions occur, how monitoring locations are selected, what measures are taken to reduce pollutant levels in different regions, and how accurately pollutants are measured. Inaccurate measurement of pollutant emission and concentration is probably the most important determinant of this weak relationship. To be clear about the true relationship, all provincial regions must improve at measuring pollutants. This is also needed to accurately assess the relationship between air pollution and cancer burden.

Trends of Cancer Incidence and Mortality in China

From 1989 to 2008, the crude overall cancer incidence in cancer registration areas increased from 184.81/100,000 to 286.69/100,000 in China, with an annual change of 2.4%. After age standardization, no obvious change was found in overall incidence[35]. In addition, although the crude overall cancer mortality increased from 156.93/100,000 to 184.67/100,000 (annual percent change = 1.0%), the age-standardized overall cancer mortality significantly decreased, with an annual percent change of –1.2% ()[36].

Figure 9.

Trends of overall cancer incidence and mortality in China, 1989–2008.

Sources of data: Chinese Cancer Registry Report[35], [36].

In 2009, the crude overall cancer incidence and mortality were 285.91/100,000 and 180.54/100,000, respectively, in China[37]. Lung cancer, gastric cancer, colorectal cancer, liver cancer, and esophageal cancer were the five most common cancers, accounting for 51.8% of all cancer cases[37]. Among the 72 cancer registration areas, Dalian, Shanghai, and Yangzhong reported the highest overall cancer incidence (410.95/100,000, 410.936/100,000, and 383.39/100,000); Dalian, Haimen, and An'shan reported the highest lung cancer incidence (85.69/100,000, 79.90/100,000, and 77.46/100,000); and Qidong, Fusui, and Haimen reported the highest liver cancer incidence (75.61/100,000, 69.99/100,000, and 53.53/100,000) ()[37].

Figure 10.

Cancer (overall, lung, and liver) incidence and mortality in 72 cancer registration areas in China in 2009.

Sources of data: Chinese Cancer Registry Report[37],[58],[59].

Relationship Between Cancer Burden and Concentrations of Major Pollutants in China

Several factors may account for the increasing cancer incidence and mortality in China, including the aging population, high rate of smoking among males, limited time for physical exercise, changing diet patterns, and increased numbers of obese and overweight individuals[38]. Notably, these factors were observed not only in urban China but also in rural China[38]. Assessing other specific risk factors may provide more information on the cause of certain cancers in China. As shown in , an obviously opposite urban-rural distribution was observed between the incidences of lung cancer and liver cancer in China. For lung cancer, the incidence in urban China was consistently higher than that in rural China over the past 20 years, but the opposite pattern was observed for liver cancer. This urban-rural distribution of liver cancer probably implies that diet factors are the major determinant. Differences in the distribution of lung cancer between urban and rural areas are probably due to air pollution.

Figure 11.

Trends of the incidence of lung cancer and liver cancer in urban and rural China from 1989 to 2008.

Sources of data: Chinese Cancer Registry Report[35].

To better clarify the APC link, the annual mean concentrations of PM10, SO2, and NO2 were linked with cancer incidence in selected cities (10 for which there are matched data). As shown in , neither the overall cancer incidence nor lung/liver cancer incidence in 2009 in these areas was strongly associated with mean concentration of PM10 or SO2 from 2003 to 2009. However, the annual mean concentration of NO2 was associated with overall cancer incidence and lung cancer incidence (). The relationship between pollutant concentration and cancer mortality followed a similar trend as that between pollutant concentration and cancer incidence.

Figure 12.

Scatter diagram of cancer incidence (overall, lung, and liver) in 2009 and annual mean concentrations of major pollutants (PM10, SO2, and NO2) from 2003 to 2009 in select Chinese cities.

Blue dot in each sub-graph represents the cancer incidence (overall, lung, and liver) for one city in 2009, and other color dots represent the corresponding level of air pollutant concentration for one city from 2003 to 2009. Sources of data: Chinese Cancer Registry Report[37],[58],[59] and China Statistical Yearbook 2004–2010[27],[52].

After excluding the data from Shenyang (as shown in the scatter diagram, data from Shenyang seems to be an outlier), the correlation became stronger. Moreover, the incidence and mortality of overall cancer and lung cancer were correlated with both the current and past annual mean concentrations of NO2. For example, the overall cancer incidence in 2009 was most strongly correlated with the concentration of NO2 in 2005 (r = 0.784), and lung cancer incidence in 2009 with the concentration of NO2 in 2008 (r = 0.836). Stronger correlation was found between lung cancer mortality in 2009 and the concentration of NO2 in 2004 (r = 0.923) (). These results support the assumption of latency between exposure to air pollution and development of cancer. However, it should be noted that the latency may be not as long as the reported 30 years[39]. Lung cancer may manifest more quickly because of exposure to a lasting, unprecedented, high level of complex pollutants resulting from coal consumption, cement and steel production, and civilian vehicle use. Even if a 30-year latency exists, the jump in cancer burden related to air pollution will likely occur in China in the near future, given that the remarkable economic transformation started 20–30 years ago.

Table 1.

Correlation between cancer (overall and lung) incidence and mortality in 2009 and annual mean concentrations of nitrogen dioxide (NO2) from 2003 to 2009 in selected cities in China

Cancer incidence/mortality	Pearson correlation (r)
	2003	2004	2005	2006	2007	2008	2009
Data with Shenyang
Overall cancer incidence	0.527	0.375	0.650*	0.656*	0.517	0.637*	0.626
Lung cancer incidence	0.337	0.281	0.381	0.375	0.250	0.421	0.357
Liver cancer incidence	0.413	0.169	0.375	0.311	0.434	0.515	0.479
Overall cancer mortality	0.392	0.435	0.570	0.574	0.324	0.506	0.414
Lung cancer mortality	0.303	0.440	0.417	0.522	0.180	0.276	0.153
Liver cancer mortality	0.453	0.502	0.563	0.561	0.407	0.566	0.395
Data without Shenyang
Overall cancer incidence	0.648	0.493	0.784*	0.716*	0.681*	0.780*	0.759*
Lung cancer incidence	0.711*	0.678*	0.766*	0.602	0.684*	0.836**	0.738*
Liver cancer incidence	0.372	0.089	0.329	0.276	0.398	0.489	0.447
Overall cancer mortality	0.531	0.600	0.730*	0.652	0.495	0.668*	0.556
Lung cancer mortality	0.698*	0.923**	0.844**	0.801**	0.622	0.680*	0.502
Liver cancer mortality	0.508	0.574	0.629	0.583	0.481	0.639	0.443

*, Correlation is significant at the 0.05 level (two-tailed). **, correlation is significant at the 0.01 level (two-tailed). Sources of data: Chinese Cancer Registry Report[37],[58],[59] and China Statistical Yearbook 2004-2010[27],[52].

Challenges in Pollution Control for China

Controlling air pollution and understanding the real APC link in China are very challenging because the sources of pollution and types of pollutants are very different from European countries and the US, and they are also diverse and vary across different regions in China[40]. The government has demonstrated strong intent to tackle this problem; nevertheless, there are deficiencies in pollution monitoring systems, and there are only a few well established cancer registries in selected cities in China. At the other hand, reduction of air pollution will possibly impact the economy and the modern lifestyle to which Chinese people have become accustomed, but past experience from other parts of the world points to a grim picture of cancer burden and personal suffering. Some hard choices have to be made. We hereby suggest a few measures that should be taken.

First, establish and enforce the rules of the Clear Air Acts in China. No doubt, this will be challenging; but the Clean Air Acts would be the most direct and effective strategy for controlling air pollution. Such enforcement has already been undertaken in other countries[41],[42], with the Clean Air Act of 1956 after the London Smog Episode and the Pollution-Related Health Damage Special Measures Law after the Yokkaichi Asthma Episode in Japan. In recent years, the Chinese government has increased the emission standards[43], launched a National Plan on Air Pollution Control in Key Regions during the 12th Five-Year Plan (2011–2015) in 2012, and released the first National Action Plan on Air Pollution Prevention and Control (2013–2017) in September 2013[25]. The emphasis has to be enforcement of the laws and implementation of the standards. More detailed regulations are also needed in the areas of the deamination, dedusting, and desulfurization processes in coal consumption. Clear regulations to reduce NOx emission in vehicle exhaust should be established and enforced.

Second, improve the monitoring network for air pollutants. As discussed earlier, distribution of air pollutants in China exhibits obvious regional and time differences. A careful study is needed to determine the optimal number of monitoring stations in different regions of China to gain accurate information on current pollution levels and any changes as a measure of the effectiveness of new laws and regulations. The data from the monitoring network need to be maintained in a centralized database and reported to the public in a timely fashion.

Third, establish an effective national cancer registry that accurately reports cancer burden on a yearly basis. A difficulty in establishing a clear APC link in China is a lack of comprehensive cancer incidence data from many provinces and cities. As a result, only two cohort studies[40],[44] and very limited case-control studies[45]–[48] have been reported to support the association between air pollution and cancer burden among the Chinese population. And most studies investigated indoor air pollution; only Cao's cohort study found that every 10 µg/m3 increase in outdoor PM2.5, SO2, and NOx was associated with a 3.4%, 4.2%, and 2.7% increased risk of lung cancer mortality in China[44]. However, this APC link was likely to be underestimated according to the studies from other parts of the world[3],[49]–[51]. For example, the latest European Study of Cohorts for Air Pollution Effects (ESCAPE), including 312,944 subjects from 17 European cohort studies, showed a 22% increased risk of lung cancer associated with every 10 µg/m3 increase of PM10[3]. The same increments of PM10 and PM2.5 were associated with hazard ratios for lung adenocarcinomas of 1.51 (1.10–2.08) and 1.55 (1.05–2.29), respectively[3]. The Harvard Six Cities study showed that each increase in PM2.5 (10 µg/m3) was associated with a 37% (95% CI: 7%–75%) adjusted increased risk of lung cancer mortality[49].

Future Directions on Air Pollution and Cancer Research

Although the APC link is largely unclear and very limited evidence has been provided to support the link in China, the high level of multicomponent air pollution in large cities provides a unique opportunity and natural laboratory for studying this relationship. To study the pathologic role of air pollution among the Chinese, large longitudinal studies with extensive and accurate measurement of outdoor air pollution, as well as controls for potential confounding factors, are urgently needed in the future. Improving the measurement of exposure to air pollution at the individual level will be a common challenge in all studies focused on this link. Biomarkers for exposure, such as DNA adducts and molecular alterations (e.g., loss of heterozygosity, gene mutations, and aberrant gene promoter methylation), will be very useful for understanding the mechanisms of the APC link and for facilitating early detection and treatment of high-risk groups.

Conclusions

As another demonstration of the Chinese Yin-and-Yang philosophy, China's rapid economic growth has also brought about a rapid increase in air pollutant emissions, which is expected to augment the cancer burden on Chinese people. Because of the latency for cancer development, this great threat has not gained the deserved attention. However, it is clear that a surge in cancer cases is emergent, if not happening already. All are feeling the urgency. Although the population looks to the government to solve the upcoming crisis, it is essential for the academic community to work with the government to establish a system with which the severity of the problem can be objectively measured and monitored. Only when Chinese people clearly see the danger of the cancer burden can they make appropriate choices. Without the public's buy in, Clean Air Act laws in China will be difficult to enforce.