Correction
17 Feb 2015: The PLOS ONE Staff (2015) Correction: Promoted Relationship of Cardiovascular Morbidity with Air Pollutants in a Typical Chinese Urban Area. PLOS ONE 10(2): e0118427. https://doi.org/10.1371/journal.pone.0118427 View correction
Figures
Abstract
Background
A large number of studies about effects of air pollutants on cardiovascular mortality have been conducted; however, those investigating association between air pollutants and cardiovascular morbidity are limited, especially in developing countries.
Methods
A time-series analysis on the short-term association between outdoor air pollutants including particulate matter (PM) with diameters of 10 µm or less (PM10), sulfur dioxide (SO2) and nitrogen dioxide (NO2) and cardiovascular morbidity was conducted in Tianjin, China based on 4 years of daily data (2008–2011). The morbidity data were stratified by sex and age. The effects of air pollutants during the warm season and the cool season were also analyzed separately.
Results
Each increase in PM10, SO2, and NO2 by increments of 10 µg/m3 in a 2-day average concentration was associated with increases in the cardiovascular morbidity of 0.19% with 95 percent confidence interval (95% CI) of 0.08–0.31, 0.43% with 95% CI of 0.03–0.84, and 0.52% with 95% CI of −0.09–1.13, respectively. The effects of air pollutants were more evident in the cool season than those in the warm season, females and the elderly were more vulnerable to outdoor air pollution.
Conclusions
All estimated coefficients of PM10, SO2 and NO2 are positive but only the effect of SO2 implied statistical significance at the 5% level. Moreover, season, sex and age might modify health effects of outdoor air pollutants. This work may bring inspirations for formulating local air pollutant standards and social policy regarding cardiovascular health of residents.
Citation: Tong L, Li K, Zhou Q (2014) Promoted Relationship of Cardiovascular Morbidity with Air Pollutants in a Typical Chinese Urban Area. PLoS ONE 9(9): e108076. https://doi.org/10.1371/journal.pone.0108076
Editor: Qinghua Sun, The Ohio State University, United States of America
Received: March 17, 2014; Accepted: August 25, 2014; Published: September 23, 2014
Copyright: © 2014 Tong et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This research was financially supported by the Ministry of Science and Technology, People's Republic of China as a 973 project (grant No. 2011CB503802), and by the National Natural Science Foundation as a key project (grant No. 21037002). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
As the largest harbor in northern China, Tianjin is a fast-growing and economically developed city. It has an area of approximately 11,919 km2 and a population of 10 million. A continental monsoonal climate featuring hot and humid summer, and dry and cold winter is typical here. The mean annual temperature and precipitation is 13.1°C and 389.4 mm, respectively. It is wetter in summer than in winter. The domestic heating season is generally between November and March. The average wind speeds range from 16 to 24 km/h. A comprehensive industrial system including integrated machinery, electronics, petroleum and chemicals, metallurgy, textiles and vehicles has been built in this city. The industrial prosperity brings about severe air pollution problems. From 2008 to 2011, both daily concentrations of PM10 and SO2 (80.2 µg/m3 and 148.1 µg/m3, respectively) exceeded the Class I levels of the Chinese national standards (daily average: 50 µg/m3 for both PM10 and SO2) [1] and the standards of the World Health Organization (WHO) (daily average: 50 µg/m3 and 24 µg/m3 for PM10 and SO2, respectively) [2]. Air pollution exerts tremendous burdens to public health, ranking as the 13th leading cause of mortality [3]–[5].
There are numerous studies of similar questions that have been conducted in North America and Western Europe. Several years ago, the Health Effects Institute sponsored some studies under the acronym PAPA (http://pubs.healtheffects.org/types.php?type=1 for published reports), including "Public Health and Air Pollution in Asia (PAPA): Coordinated Studies of Short-Term Exposure to Air Pollution and Daily Mortality in Two Indian Cities" and "Public Health and Air Pollution in Asia (PAPA): Coordinated Studies of Short-Term Exposure to Air Pollution and Daily Mortality in Four Cities". Pathophysiologic mechanisms of air pollutant-induced cardiovascular morbidity and mortality are also being studied widely [6]–[8], besides, some studies on the associations between air pollutants and cardiovascular diseases have been conducted in terms of epidemiology [9]–[13]. However, investigations on the relationship between air pollutants and cardiovascular morbidity are scarce at present [14], especially in Asian countries where, arguably, both living conditions and health indicators may be quite different.
This study presents an investigation of this issue in a typical Chinese city to increase our understanding of cardiovascular health risks associated with polluted air and provide some scientific basis for establishment of public health and environmental protection policies [15].
Materials and Methods
2.1. Subject Data
Daily air pollution data on PM10, SO2 and NO2 were obtained from the website of the Tianjin Environmental Monitoring Centre. Daily mean temperature and relative humidity were obtained from the China Meteorological Data Sharing Service System (Data S1).
Daily cardiovascular morbidity data from 1 January 2008 to 31 November 2011 was collected from the Centers for Disease Control and Prevention of Urban Districts in Tianjin, China (Nankai, Heping, Hexi, Hedong and Hongqiao districts), covering around 77000 local residents (SI). Information of patients including those under the age of 18 was anonymized and were validated each year by China CDC. They were coded according to the ICD-10 (the 10th revision of International Classification of Diseases) and classified into cardiovascular causes including cerebral infarction, primary diagnosed hypertension, cerebral hemorrhage, acute myocardial infarction and subarachnoid hemorrhage [16]. They were also stratified by sex and age (0–18, 18–44, 45–64, and ≥65 years) [17]. The ethical committee of the coordinating center of five urban CDCs in Tianjin approved the study (Full name is "CDC biomedical ethics council").
Air pollutant concentrations and meteorological measures including temperature and humidity are shown in Table 1 and 2.
2.2. Statistical methods
All analyses were conducted with statistical software package SAS version 9.1. Time-series analysis was utilized to explore modification effects of season, age and sex on the association between air pollutants and morbidity in Tianjin from 2008 to 2011. In detail, the generalized linear model with natural splines (ns) functions of time, weather conditions accommodate nonlinear and non-monotonic relationships of morbidity with time, temperature and relative humidity were utilized for the analysis [17]. In the basic model, morbidity outcomes were included without air pollution variables. Residuals of the basic models were examined to determine whether there were discernible patterns and autocorrelation by means of residual plots and partial autocorrelation function (PACF) plots. Day of the week (DOW) was considered as a dummy variable in the basic modes. After the establishment of the basic model, PM10 and covariates (temperature, humidity, and SO2 and NO2 concentrations) were introduced into it and their effects were analyzed. The selection of df (degrees of freedom) for time trends was done with the PACF [17], [18]. 4 df per year for time trends were used in our basic models for cardiovascular morbidity. In addition, we used 3 df (during whole period of study) for temperature and humidity. This modeling procedure was carried out for each series studied and the core models were assessed with plots of model residuals and fitted values as well as plots of the estimated partial autocorrelation functions.
The estimated pollution log-relative rate β is obtained through fitting of the following log-linear generalized linear model:
logE(Yt) = βZt + DOW + ns(time, df) + ns(temperature/humidity, 3) + intercept (1)
Where E(Yt) and β represent the expected morbidity numbered at day t and the log-relative rate of morbidity corresponding to a unit increase of air pollutants, respectively; Zt indicates pollutant concentrations at day t; DOW is dummy variable for day of the week; ns (time, df) is the ns function of calendar time with 4 df to adjust for seasonality and other time-varying influences on admissions (e.g. influenza epidemics and longer-term trends); and ns (temperature/humidity, 3) is the ns function for temperature and humidity with 3 df.
The data were stratified by sex and age. We analyzed the associations for the warm season (April-September), the cool season (October-March) and the entire year, respectively (Kan et al., 2008b). The basic models of seasonal analyses were different from those of the whole-period in terms of df for time trends. The effects were quantified on the basis of the percentage change in risk per 10 mg/m3 increase in the concentration of each pollutant. The statistical significance was defined as p<0.05.
Results
3.1. Time series plot of the morbidity, exposure-response relationships, ACF and PACF of morbidity
The time series plot of cardiovascular morbidity from 2008 to 2011 in Tianjin is shown in Fig. 1. Fig. 2 demonstrates the exposure–response relationships for PM10, SO2 and NO2 (2-day moving average of air pollutant concentration (lag 01)) with cardiovascular mortality during 2008–2011 in Tianjin. In Fig. 3, ACF and PACF of cardiovascular morbidity are depicted as original data (log). ACF and PACF of cardiovascular morbidity as original data (log+Difference) are shown in Fig. 4. Fig. 5 gives ACF and PACF of residual after modeling (basic model).
The x-axis is the pollutant concentrations; the y-axis is the estimated percent change in cardiovascular mortality; the solid blue lines indicate the estimated mean percent change in daily mortality outcomes with the dashed lines representing the 95% CI.
The x-axis is the lag number.
The x-axis is the lag number.
3.2. Effects by sexes and ages
As shown in Table 3, the percent increases associated with air pollutant concentrations varied by sex or age groups. The effect estimates of pollutants among females were greater than those among males, especially the estimate for NO2 was thrice as much as that among males, although their between-sex differences were insignificant statistically. The effects among those ≥65 were greater than those in the other two groups.
3.3. Effects by seasons
The daily average morbidity in the warm and cool season was 51.8±0.5 and 50.9±0.6, respectively. For the entire period, the average value was 51.4±0.4 and an increase of 10 µg/m3 of 2-day average concentration of PM10, SO2 and NO2 correspond to 0.19% (95% CI, 0.08–0.31), 0.43% (95% CI, 0.03–0.84) and 0.52% (95% CI, −0.09–1.13) increases, respectively (Table 4).
Sensitivity analyses results for PM10 as shown in Table 5 and Table 6. They showed that the results were fairly robust for various concentrations, specifications for temperature, methods of aggregating daily data, df used in the smoothers, and alternative spline models.
Discussion
PM10 in developed countries are mainly discharged from various automobiles including large amount of secondary organic aerosols. In Tianjin, ground dust, vehicle, cement dust and incineration are the primary PM10 sources [19]. Coal and crude oil are responsible for 66% and 30% of energy supply, respectively [20]. Crude oil dominated by PM with fine fractions carries more toxic substance and is easier to penetrate into the circulation system of humans than coal. The main components of other PM sources such as raise dust and sand dust are less toxic inorganic minerals.
Generally speaking, the smoking habit exerted much more oxidative and inflammatory influences on males than air pollution [21]. In China, smoking is much more prevalent among men than women [22], this may have affected male health to a greater degree than environmental factors, therefore we suspect females were more sensitive to smoking than males. As for age, older people ≥65 were more vulnerable to air pollution (especially SO2) than people in the other two age groups. In terms of their occupation, percent increase of myocardial infarction exposed to welding and soldering fumes in the Copenhagen male study population was 1.1 (95%CI, 0.6–2.2) [23], much higher than the effect estimates on males in this study (Table 3). This is mainly attributed to the higher concentration PM exposure with more toxicity in the workplace than public environment.
Though not measured in this study, other factors such as underlying diet could also have affected our observed results. Because nutrients with natural chelating properties, including antioxidants, herbs, minerals, essential amino acids and fiber, can detoxify human bodies, they can protect humans from oxidative stress of free radicals derived from air pollutants to some extent. Vitamins C, E and A in majority of plant and fish foods favored by residents in Tianjin as an important coastal city can interfere with or scavenge reactive oxygen species within cells [24]–[26]. To better understand the modification effects of living habits, socioeconomic and demographic factors on the associations, more individual features (smoking and eating habit, occupation, education attainment levels, physical activity, socioeconomic status, etc.) should be taken into consideration in the future.
As for seasonal effects, some studies indicated that high temperatures in the warm season and low temperatures in the cold season are associated with increased cardiovascular mortality [17], [27]–[29]. It is likely that they can also affect cardiovascular morbidity. Therefore, morbidity data were stratified by the warm and cool season defined as April-September and the cool season defined as October-March, respectively. Therefore, this research also demonstrated that the effect estimates of air pollutants in Tianjin might be modified by seasons. Exposure patterns may contribute to this season-specific observation. Personal exposure was reduced by decreased time spent outdoors caused by heavy rain. In contrast, people are more likely to go outdoors and open windows in the cool season with the drier and less variable in Tianjin. The effect estimates of SO2 and NO2 in the cool season were much higher than those in the warm season. Significant associations were observed for PM10 and SO2 in the cool season (0.24 (0.20, 0.28) and 0.66 (0.35, 0.97), respectively), while the effect of SO2 in the warm season was insignificant. It may be caused by coal combustion accounting for the largest proportion of heat supply in the cool season. Besides, the significant Pearson correlation coefficient between PM10 and SO2 (0.197, p<0.01) might partially explain this phenomenon. NO2 was mainly released from vehicle emission and partially from coal combustion, which may account for the insignificant relationship between NO2 and morbidity. The constituents of the complex mix of PM10 may vary by seasons, different from the gaseous pollutants (SO2 and NO2).
There are some limitations in our analyses. Ozone was not included in our studies due to a lack of monitoring. Our ability to separate the independent effect of each pollutant was limited by high correlations between PM and gaseous pollutants including SO2 and NO2 in Tianjin. Moreover, the role of each specific component of air pollution should be examined to determine the combination of particles responsible for the increases in environment-induced health concerns. These investigations are paramount for policy makers to carry out improved interventions that will impact health hazards related with air pollutants, especially on the increased risks of cardiovascular morbidity.
Conclusions
In summary, we found that PM10 and SO2 were significantly associated with cardiovascular morbidity in the cool season in Tianjin in this time-series analysis. The mean percent increases (95% CI) of cardiovascular morbidity outcomes per 10 µg/m3 increase in PM10 and SO2 concentrations by season were 0.24 (0.20, 0.28) and 0.66 (0.35, 0.97), respectively. Besides modification effect of a season, individual features (sexes, ages) might also interfere with the effect estimates of air pollutants. Females were more sensitive than males to air pollutants. The effects of air pollutants on elder people (≥65) were greater than those <65. These data can enable policy makers to enforce or improve existing legislation that control air pollution after weighing the disadvantages of potentially slowing rapid economic development.
Supporting Information
Data S1.
Cardiovascular, air pollutants and meteorological measurements data.
https://doi.org/10.1371/journal.pone.0108076.s001
(XLS)
Acknowledgments
We cordially thank the Centers for Disease Control and Prevention of Urban Districts in Tianjin for assisting us to collect medical data.
Author Contributions
Performed the experiments: LT. Analyzed the data: KL LT QZ. Wrote the paper: LT QZ KL.
References
- 1.
GB/T 3095-2012 (2012) Ambient air quality standards. Chinese Standard Press, Beijing.
- 2.
WHO (2005) Air quality guidelines global update for particulate matter, ozone, nitrogen dioxide and sulfur dioxide. Geneva: World Health Organization.
- 3. Roberts S (2011) Can mortality displacement mask thresholds in the concentration-response relation between particulate matter air pollution and mortality? Atmo Environ 45: 4728–4734.
- 4. Kan HD, Chen RJ, Tong SL (2012) Ambient air pollution, climate change, and population health in China. Environ Int 42: 10–19.
- 5. Cheng Z, Jiang JK, Fajardo O, Wang SX, Hao JM (2013) Characteristics and health impacts of particulate matter pollution in China (2001–2011). Atmos Environ 65: 186–194.
- 6. Brook DR (2008) Cardiovascular effects of air pollution. Clin Sci 115: 175–187.
- 7. Koulova A, Frishman HW (2014) Air Pollution Exposure as a Risk Factor for Cardiovascular. Cardiology 22: 30–36.
- 8. Shrey K, Suchit A, Deepika D, Shruti K, Vibha R (2011) Air pollutants: The key stages in the pathway towards the development of cardiovascular disorders. Environ Toxicol Phar 31: 1–9.
- 9. Wong CM, Atkinson WR, Anderson RH, Hedley AJ, Ma S, et al. (2002) A tale of two cities: effects of Air Pollution on Hospital Admissions in Hong Kong and London Compared. Environ Health Persp 110: 67–77.
- 10. Moolgavkar HS (2000) Air pollution and hospital admissions for diseases of the circulatory system in three U.S. metropolitan areas. Air & Waste Manage Assoc 50: 1199–1206.
- 11. Roberts S (2003) Combining data from multiple monitors in air pollution mortality time series studies. Atmo Environ 37: 3317–3322.
- 12. Guo YM, Barnett AG, Zhang YS, Tong SL, Yu WW, et al. (2010) The short-term effect of air pollution on cardiovascular mortality in Tianjin, China: Comparison of time series and case-crossover analyses. Sci Total Environ 409: 300–306.
- 13. Tsai DH, Wang JL, Chuang KJ, Chan CC (2010) Traffic-related air pollution and cardiovascular mortality in central Taiwan. Sci Total Environ 408: 1818–1823.
- 14. Brook DR, Rajagopalan S, Pope AC III, Brook RJ, Bhatnagar A, et al. (2010) Particulate Matter Air Pollution and Cardiovascular Disease: An Update to the Scientific Statement from the American Heart Association. Circulation 1: 2331–2378.
- 15. Hu XG, Zhou QX (2013) Health and ecosystem risks of graphene. Chem Rev 113: 3815–3835.
- 16.
WHO (1993) International Classification of Diseases, Tenth Revision. Geneva: World Health Organization.
- 17. Kan HD, London JS, Chen GH, Zhang YH, Song GX, et al. (2008) Season, sex, age, and education as modifiers of the effects of outdoor air pollution on daily mortality in Shanghai, China: The Public Health and Air Pollution in Asia (PAPA) study. Environ Health Persp 116: 1183–1188.
- 18. Klea K, Giota T, Evangelia S, Alexandros G, Alain LT, et al. (2001) Confounding and effect modification in the short-term effects of ambient particles on total mortality: results from 29 European cities within the APHEA2 Project. Epidemiology 12: 521–531.
- 19. Xiao ZM, Wu JH, Han SQ, Zhang YF, Xu H, et al. (2012) Vertical characteristics and source identification of PM10 in Tianjin. J Environ Sci-China 24: 112–115.
- 20. Li WF, Bai ZP, Liu AX, Chen J, Li C (2009) Characteristics of major PM2.5 components during winter in Tianjin, China. Aerosol Air Qual Res 9: 105–119.
- 21. Grundtvig M, Hagen PT, Amrud SE, Reikvam Å (2013) Reduced life expectancy after an incident hospital diagnosis of acute myocardial infarction- Effects of smoking in women and men. Int J Cardiol 167: 2792–2797.
- 22. Johnson AC, Palmer HP, Chou CP, Pang ZC, Zhou Dunjin, et al. (2006) Tobacco use among youth and adults in Mainland China: The China Seven Cities Study. Public Health 120: 1156–1169.
- 23. Fang CS, Cassidy A, Christiani CD (2010) A Systematic Review of Occupational Exposure to Particulate Matter and Cardiovascular Disease. Inter J Environmental Research Pub Heal 7: 1773–1806.
- 24. Kampa M, Castanas E (2008) Human health effects of air pollution. Environ Pollu 151(2): 362–367.
- 25. Kelly JF (2004) Dietary antioxidants and environmental stress. P Nutr Soc 63: 579–585.
- 26. Chalamaiah M, Dinesh kumar B, Hemalatha R, Jyothirmayi T (2012) Fish protein hydrolysates: Proximate composition, amino acid composition, antioxidant activities and applications: A review. Food Chem 135: 3020–3038.
- 27. Curriero FC, Heiner KS, Samet JM, Zeger LS, Strug L, et al. (2002) Temperature and mortality in 11 cities of the eastern United States. Am J Epidemiol 155: 80–87.
- 28. Kan HD, Heiss G, Rose MK, Whitsel AE, Lurmann F, et al. (2008) Prospective analysis of traffic exposure as a risk factor for incident coronary heart disease: The Atherosclerosis Risk in Communities (ARIC) Study. Environ Health Persp 116: 1463–1468.
- 29. Zanobetti A, Schwartz J (2008) Temperature and mortality in nine US cities. Epidemiology 19: 563–570.