Impact factor (WEB OF SCIENCE - Clarivate)

2 year: 7.664 | 5 year: 7.127

Articles

Ambient fine particulate matter chemical composition associated with in-hospital case fatality, hospital expenses, and length of hospital stay among patients with heart failure in China

Zhihan Lai, Jingyi Zhang, Shanshan Ran, Dashan Zheng, Jin Feng, Gan Wu, Miao Cai*, Hualiang Lin*

Department of Epidemiology, School of Public Health, Sun Yat-sen University, Guangzhou, China
*Joint senior authorship.

DOI: 10.7189/jogh.14.04032

Share:

Facebook
Twitter
LinkedIn
Abstract

Background

Previous studies have observed the adverse effects of ambient fine particulate matter pollution (PM2.5) on heart failure (HF). However, evidence regarding the impacts of specific PM2.5 components remains scarce.

Methods

We included 58 129 patients hospitalised for HF between 2013 and 2017 in 11 cities of Shanxi, China from inpatient discharge database. We evaluated exposure to PM2.5 and its components ((sulphate (SO42−), nitrate (NO3), ammonium (NH4+), organic matter (OM) and black carbon (BC)), along with meteorological factors using bilinear interpolation at each patients’ residential address. We used multivariable logistic and linear regression models to assess the associations of these components with in-hospital case fatality, hospital expenses, and length of hospital stay.

Results

Increase equivalents to the interquartile range (IQR) in OM (odds ratio (OR) = 1.13; 95% confidence interval (CI) = 1.02, 1.26) and BC (OR = 1.14; 95% CI = 1.02, 1.26) were linked to in-hospital case fatality. Per IQR increments in PM2.5, SO42−, NO3, OM, and BC were associated with cost increases of 420.62 (95% CI = 285.75, 555.49), 221.83 (95% CI = 96.95, 346.71), 214.93 (95% CI = 68.66, 361.21), 300.06 (95% CI = 176.96, 423.16), and 303.09 (95% CI = 180.76, 425.42) CNY. Increases of 1 IQR in PM2.5, SO42−, OM, and BC were associated with increases in length of hospital stay of 0.10 (95% CI = 0.02, 0.19), 0.09 (95% CI = 0.02, 0.17), 0.10 (95% CI = 0.03, 0.17), and 0.16 (95% CI = 0.08, 0.23) days.

Conclusions

Our findings suggest that ambient SO42−, OM, and BC might be significant risk factors for HF, emphasising the importance of formulating customised guidelines for the chemical constituents of PM and controlling the emissions of the most dangerous components.

Print Friendly, PDF & Email

Heart failure (HF), a clinical cardiac syndrome characterised by symptoms such as shortness of breath, fatigue, and fluid retention, has become as a prominent global public health concern and imposed substantial health burdens, including economic losses, disabilities, and fatalities [1,2]. It is the leading cause of mortality and morbidity in low- and middle-income countries [3]. As an ageing developing country, China is profoundly impacted by HF [4]; estimates suggest that approximately one-fifth of cardiovascular inpatients in China were affected by HF in 2003, and the related costs accounted for over 5% of the country’s total health care expenditure [5,6]. This identifying potential risk factors for HF and developing appropriate and effective preventive measures crucial for addressing this critical public health concern.

As a serious public health threat worldwide, air pollution has consistently been associated with elevated risks of cardiovascular diseases and related mortality [710]. Ambient particulate matter (PM) plays a vital role in this context; it constitutes a composite of solid and liquid particles, possessing aerodynamic diameters smaller than 2.5 µm (PM2.5) and ranging between 2.5 and 10 µm (PM10–2.5) [11]. According to the updated 2021 PM2.5 thresholds of the World Health Organization (WHO), approximately 1.41 billion people in China (99% of the total population) were exposed to unsafe levels of PM2.5 (>5 μg/m3), while 765 million (53% of the population) resided in areas with hazardous air quality (>35 μg/m3) [12]. Numerous studies found favourable associations between ambient PM and increased incidence, hospitalisation, readmission, and mortality of HF [1316]. However, they primarily used incidence or mortality rates of HF to assess the disease burden attributed to air pollution, without emphasising metrics such as hospitalisation expenses and length of hospital stay (LOS). These indicators better describe the influence of air pollution on individuals and give additional practical significance. Additionally, most studies used the mass concentration of PM as the indicator, overlooking the toxicity associated with its different chemical components ((sulphate (SO42−), nitrate (NO3), ammonium (NH4+), organic matter (OM), and black carbon (BC)). Their health effects are unlikely to be equally significant, considering the distinct chemical properties of each PM chemical component. Though WHO released the latest air quality guidelines in 2021, specific guidelines for these chemical components are still unavailable. Previous studies have shown that they can impair human health through different mechanisms such as systemic inflammation, oxidative stress, adipose tissue inflammation, impaired vascular function, and insulin resistance [17,18]. Therefore, studying the relationships between PM chemical components and HF could help with advancing the current understanding of PM toxicity and facilitating the development of more focused regulations and policies aimed at reducing the health burden caused by PM air pollution.

We hypothesised that PM2.5 and its chemical constituents are associated with increased risks of in-hospital mortality, hospitalisation costs, and LOS in patients hospitalised due to HF. We explored this by using data from the inpatient hospitalisation database of Shanxi, a province known for its highly developed coal mining industry and poor air quality in China. We assessed the impact of individual-level exposure to PM2.5 mass and its dust-free chemical constituents at HF patients’ residential addresses prior to hospitalisation.

METHODS

Study area

Shanxi Province, located on the eastern boundary of the Loess Plateau and west of Taihang Mountains in the northern region of China, stretches over 156 700 km2. It administratively encompasses 11 prefecture-level cities (Changzhi, Datong, Jincheng, Jinzhong, Linfen, Lvliang, Shuozhou, Taiyuan, Xinzhou, Yangquan and Yuncheng) with a population of 34.9 million in 2020. Due to the temperate continental climate, Shanxi is relatively dry and undergoes large temperature differences all year round. As a province rich in coal resources, Shanxi Province has a highly developed coal-related industry, which has resulted in severe air pollution problems, significantly impacting the health of residents.

Data collection

We retrieved data on individual patient’s demographics, medical diagnoses and procedures, hospital costs, days of hospital stay, residential addresses, and in-hospital health outcomes from the electronic medical records system of secondary and tertiary hospitals of Shanxi Province [1922]. The data spanned a period from 13 February 2013 to 31 December 2017. Before accessing the data, personal identifiers such as names and distinct identification codes of patients were removed. This study received ethical approval from the institutional review board of the School of Public Health, Sun Yat-sen University.

Identification of HF

We used the I50 primary diagnosis code at admission (per the International Classification of Diseases, 10th Edition (ICD-10)) to identify HF cases. We excluded individuals <18 years of age, those with unknown age, sex, ethnicity, or residential addresses, and those lacking data on exposure (Figure 1).

Figure 1.  Flowchart of sample selection, inclusion and exclusion criteria.

Air pollution exposure measurement

We obtained daily measurements of PM2.5 mass and its dust-free chemical constituents (SO42−, NO3, NH4+, OM, and BC) with a spatial resolution of 10 × 10 km from the Tracking Air Pollution database in China [2325]. The database integrates ground-based measurements, satellite-derived aerosol optical depth estimations, and emissions data from inventories to calculate the daily concentrations of PM chemical components. These estimates are generated using the Weather Research and Forecasting (version 3.9.1)-Community Multiscale Air Quality (version 5.2) chemical transport models in China since 2000 until the current date [23]. Providing information of daily concentrations of PM chemical components, the Tracking Air Pollution data set shows a mean out-of-bag cross-validation correlation coefficient (R) of 0.83. The cross-validation R values for the model were 0.70 for SO42−, 0.75 for NO3, 0.75 for NH4+, 0.72 for OM and 0.64 for BC, respectively, indicating strong agreement with ground measurements of PM chemical constituents [23,24].

We employed a two-step assessment strategy to estimate the average individual-level exposure to PM2.5 mass and its constituent chemicals, along with temperature and relative humidity, over a period of seven days. The first step involves geocoding the latitude and longitude coordinates of the study patients through the application programming interface (API) provided by ‘amap’ (Gaode map) based on their residential addresses prior to hospitalisation. In the second step, bilinear interpolation is used for the purpose of assessing the magnitudes of environmental variables, including PM2.5 mass and its chemical components, temperature, and relative humidity, at the residential addresses of the patients prior to their hospitalisation. The detailed calculation process can be found elsewhere [26,27]. Briefly, this algorithm improved the spatial precision of environmental variables at particular locations through the calculation of a weighted average derived from the closest four grid points, and the weights were determined by the distance between the case address and the grids of meteorologic variables, with closer distances having higher weights [28,29].

Short-term exposure was characterised by the seven-day average concentrations of PM2.5 and its constituents, along with the temperature and relative humidity leading up to the day of hospitalisation.

Outcomes

We obtained our outcomes of interest (in-hospital case fatality, hospitalisation expenses (CNY) and length of hospital stay (LOS in days)) from the collected electronic medical records. The hospital expenses encompass the subsequent elements: charges for medications, clinical examinations, nursing expenditures, and clinical operations. LOS refers to the cumulative number of days a patient remains hospitalised from admission to discharge.

Covariates

We selected the following covariates based on their availability in the database and their potential to confound the associations between PM2.5 mass, its chemical constituents, and HF: demographic variables, socioeconomic status, comorbidities, hospital level, payment methods and meteorological status. The demographic variables comprised age, sex, and ethnicity (Han vs non-Han). Socioeconomic status encompassed occupation (public sector, private sector, agriculture, unemployed, retired, or other) and marital status (married, single, widowed, divorced, or other). We determined existing comorbidities (Charlson comorbidity index) by extracting ICD-10 codes from the patient diagnosis fields, which we refined through text matching using regular expressions applied to clinical diagnosis descriptions [30]. We categorised hospital level into either ‘tertiary’ or ‘non-tertiary’ [19], and tertiary hospital represents a higher level of medical care. Payment methods included the New Rural Cooperative Medical Scheme, Urban Employee Basic Medical Insurance, Urban Resident Basic Medical Insurance, self-payment, and others. We collected daily meteorological information on temperature and relative humidity with a spatial resolution of 9 × 9 km through the fifth generation of European ReAnalysis– Land reanalysis data set [31,32]. We employed natural cubic splines with five degrees of freedom to incorporate the seven-day average temperature and relative humidity leading up to the day of hospitalisation. In this way, we aimed to accommodate potential nonlinear associations, considering that earlier studies have suggested the presence of relationships displaying J-shaped or U-shaped patterns. [33].

Statistical analyses

We summarised the sample chracteristics, stratified by fatality status, using means and standard deviations (SDs) for continuous and frequency with percentages for categorical variables. We used Spearman’s rank correlation to estimate the correlations between each pair of PM2.5 and its components.

We employed logistic regression models to estimate the odds ratios (ORs) of in-hospital case mortality for each interquartile range (IQR) increase in PM2.5 and its chemical components, along with 95% confidence intervals (CIs). Hospitalisation expenses and LOS showed a positively skewed distribution with pronounced tails. Nevertheless, considering the significant size of our study sample and considering the distribution patterns (Figure S1–2 in the Online Supplementary Document), we opted to use linear regression to investigate the association between exposure and these variables to more clearly describe influence of exposure on hospitalisation outcomes in patients with HF. We also used nonlinear categorical variable (quartile modes) of PM2.5 and its chemical components to repeat the above analyses and conducted trend tests. Additionally, to demonstrate potential nonlinear associations and estimate exposure-response curves, we incorporated PM2.5 and its chemical components as natural cubic splines with knots specified at quartiles. We did not use multi-pollutant models, since most air pollutants had moderate to high correlations with one another (Table S1 in the Online Supplementary Document).

We performed sensitivity analyses to assess the robustness of the findings. First, we evaluated various time lags (lag 0 to lag 7) of short-term exposure to PM2.5 and its chemical components using bilinear interpolation. Second, we only included participants with a main diagnosis of HF (n = 44 565, 76.7% of the full sample) in this study to repeat our analyses. Third, we excluded the uppermost 1% of data points for hospitalisation expenses (>47 652.27 CNY, n = 582) and LOS (>29 days, n = 647) to mitigate the influence of extreme values from these two right-skewed distributed variables. Lastly, we employed the interval between admission date and discharge date as the follow-up time and constructed Cox regression models to examine the associations between these pollutants and in-hospital mortality among HF patients.

All data processing, including data cleaning, statistical modelling, and data visualisation, were executed in R, version 4.2.1 (R Core Team, Vienna, Austria). We considered a two-way P-value <0.05 as statistically significant.

RESULTS

Descriptive results

The 58 129 patients with a primary diagnosis of HF had a mean age of 69.21 years, and 48.0% were female (Table 1). Among them, 630 (1.08%) experienced hospital fatality; they were more likely to be retired, divorced or widowed, receive treatment at tertiary hospitals, utilise Urban Resident Basic Medical Insurance, experience comorbidities, and have higher hospitalisation costs and shorter LOS. The seven-day average concentrations prior to hospitalisation were 62.14 (SD = 27.76) μg/m3 for PM2.5, 10.40 (SD = 4.05) μg/m3 for SO42−, 12.52 (SD = 7.39) μg/m3 for NO3, 8.24 (SD = 4.23) μg/m3 for NH4+, 14.64 (SD = 7.59) μg/m3 for OM and 2.87 (SD = 1.31) μg/m3 for BC. The mean concentrations of PM2.5, SO42−, OM, and BC were higher in those who experienced hospital fatality, though the difference for SO42− was not statistically significant.

Table 1.  Environmental variables and patient characteristics by case fatality

WordPress Data Table

BC – black carbon, NH4+ – ammonium, NO3 – nitrate, NRCMS – New Rural Cooperative Medical Scheme, OM – organic matter, PM2.5 – particulate matter <2.5 μm in aerodynamic diameter, SD – standard deviation, SO42− – sulphate, UEBMI – Urban Employee Basic Medical Insurance, URBMI – Urban Resident Basic Medical Insurance, x̄ – mean

Associations of PM2.5 and its chemical components with in-hospital case fatality, hospitalisation expenses, and LOS

We further explored associations of PM2.5 and its chemical components as continuous variables with in-hospital case fatality, hospitalisation expenses, and LOS (Table 2). Among the three indicators, OM and BC both showed statistically significant associations. For each increase in one IQR, OM was associated with a 13% rise in the risk of in-hospital fatality (OR = 1.13; 95% CI = 1.02, 1.26), accompanied by individual cost and LOS increases of 300.06 CNY (95% CI = 176.96, 423.16) and 0.10 days (95% CI = 0.03, 0.17), respectively. BC corresponded to a 14% increase in the risk of in-hospital fatality (OR = 1.14, 95% CI = 1.02, 1.26), with individual cost and LOS increases of 303.09 CNY (95% CI = 180.76, 425.42) and 0.16 days (95% CI = 0.08, 0.23), respectively. Regarding other pollutants, PM2.5 and SO42− were associated with increased costs and LOS, while NO3 is only related to increased costs. We found no statistically significant associations with NH4+.

Table 2.  Associations of PM2.5 and its chemical components (each IQR increment) with in-hospital case fatality, hospitalisation cost, and LOS*

WordPress Data Table

β – correlation coefficient, BC – black carbon, CI – confidence interval, IQR – interquartile range, NH4+ – ammonium, NO3 – nitrate, OM – organic matter, OR – odds ratio, PM2.5 – particulate matter <2.5 μm in aerodynamic diameter, SO42− – sulphate

*Adjusted for age, sex, ethnicity, occupation, marital status, payment methods, comorbidities, hospital level, temperature and humidity.

When using air pollutants as categorical variables categorized from lower concentration (quartile 1) to higher concentration (quartile 4), no statistically significant association with in-hospital case fatality could be found (Table 3). Regarding hospitalisation cost, the trend tests for all pollutants showed statistical significance (P-value for trend <0.05). Regarding the LOS, only PM2.5, OM and BC showed statistical significance in the trend tests.

Table 3.  Associations of PM2.5 and its chemical components with in-hospital case fatality, hospitalisation cost, and LOS, using air pollution as categorical variables*

WordPress Data Table

β – correlation coefficient, BC – black carbon, CI – confidence interval, IQR – interquartile range, LOS – length of hospital stay, NH4+ – ammonium, NO3 – nitrate, OM – organic matter, OR – odds ratio, PM2.5 – particulate matter <2.5 μm in aerodynamic diameter, ref – reference, SO42− – sulphate

*Adjusted for age, sex, ethnicity, occupation, marital status, payment methods, comorbidities, hospital level, temperature and humidity.

For hospitalisation cost, PM2.5, OM, and BC displayed a consistent positive association with higher costs as their concentrations increased. Notably, curves of these three pollutants demonstrated a rapid escalation in cost at lower concentrations, followed by a more moderate increase at higher concentrations, while curves of SO42−, NO3 and NH4+ demonstrated a distinct pattern, with rapid increase at lower concentrations but a slight decrease at higher concentrations. As for LOS, the exposure-response relationships of PM2.5, OM, and BC demonstrated an overall increasing trend, while SO42−, NO3 and NH4+ exhibited a decreasing trend at low concentrations, followed by an increase at high concentrations (Figure 2; Figure S3 in the Online Supplementary Document).

Figure 2.  Exposure-response relationships of seven-day average PM2.5 and its chemical compositions with HF LOS. Panel A. Particulate matter <2.5 μm in aerodynamic diameter (PM2.5). Panel B. Sulphate (SO42−). Panel C. Nitrate (NO3−). Panel D. Ammonium (NH4+). Panel E. Organic matter (OM). Panel F. Black carbon (BC).

Sensitivity analyses

The results of various time lags (lag 0 to lag 7) of short-term exposure to PM2.5 and its chemical components demonstrated considerable robustness. Meanwhile, our findings changed minimally for participants only with a main diagnosis of HF and remained largely consistent after excluding the top 1% of extreme values of hospitalisation expenses and LOS. The Cox regression analysis likewise showed consistent (Figures S4–6 and Tables S3–5 in the Online Supplementary Document).

DISCUSSION

Based on a sample of 58 129 HF patients from Shanxi, a region known for severe air pollution, we studied the association of PM2.5 and its chemical components with in-hospital case fatality, hospitalisation cost, and LOS. We found a positive association between the concentrations of two PM2.5 components (OM and BC) and in-hospital fatality of HF patients. Our results also indicated that higher concentrations of PM2.5 and its components (SO42−, NO3, OM, and BC) were associated with increased hospitalisation cost, while elevated concentrations of PM2.5, SO42−, OM, and BC were associated with longer LOS. We also conducted several sensitivity analyses to demonstrate the robustness of the results, where our findings regarding OM and BC showed the highest level of significance and consistency, showing statistical significance in all three indicators.

Several previous studies have explored the associations between these components and cardiovascular health. The National Particle Component Toxicity report indicated that the increased incidence of cardiovascular diseases is associated with exposure to OM [34]. Prior studies reported on the harmful impact of OM on cardiopulmonary mortality and overall mortality [35,36]. We also identified the adverse effects of OM on HF patients. However, this is inconsistent with some previous studies which no association between OM and the incidence of CVDs [37]. This difference might be due to variations in geographic regions, air pollution exposure patterns, and population characteristics, and thus warrants further exploration. The association between BC and CVDs appear to be more substantial, with several studies reporting the adverse effects of BC on cardiovascular health [3740], which aligns with our findings.

There is limited evidence on the impact of secondary inorganic ions, including SO42−, NO3 and NH4+, on cardiovascular health. A previous cohort study found that these three ions have long-term adverse effects on the incidence of CVDs [37]. However, in our study, NO3 was solely linked to increased hospitalisation cost in HF patients, aligning with prior research indicating minimal or negligible adverse health effects of nitrate at current levels, while toxicological studies utilised nitrate concentrations significantly exceeding environmental levels [41]. Furthermore, we observed no significant association between NH4+ and HF patients. A recent meta-analysis suggested that more in-depth investigations are still required to elucidate the impact of these ions as air pollutants on health [42]. Notably, exposure-response relationship curves indicate that SO42−, NO3 and NH4+ exhibit a U-shaped curve with a slight initial decrease followed by an increase for LOS, yet demonstrate an inverted U-shaped curve for hospitalisation costs. This suggests that these dose-response relationships are likely nonlinear, necessitating external data for a more comprehensive understanding.

These observed associations could be explained by several plausible biological mechanisms. Previous studies in both animals and humans have indicated that PM may damage the cardiovascular system through oxidative stress, systemic inflammation, alterations in cardiac autonomic function, and acceleration of atherosclerosis [4345]. For example, BC has been linked to adverse cardiovascular outcomes, potentially mediated by its ability to induce systemic inflammation and oxidative stress through down-regulating DNA methylation [46]. Upon penetration of lung tissues and entry into the bloodstream, BC, OM, and SO42− can trigger various physiological processes, including oxidative stress, vascular inflammation responses, and coagulation and thrombosis mechanisms [47,48]. Notably, longer-term exposure to PM over several years posed a greater risk for cardiovascular mortality compared to shorter-term exposures lasting only a few days [49], which indicated that the adverse effects of PM observed in this study may still be underestimated.

The sources of PM2.5 components have been described elsewhere [26,40,50]. SO42− mainly comes from human fossil fuel combustion and natural sources, including sulphur-containing gases emitted from the ocean and volcanoes. OM is primarily generated from a combination of emissions from combustion sources and secondary reactions, including biogenic volatile organic compounds, and BC is solely produced from incomplete combustion in various sources such as vehicular traffic, industrial emissions and residential biofuel burning. The detrimental impact of BC and OM on human health is exacerbated by their contribution to intensifying heat-absorbing greenhouse effects and contributing to global warming, akin to carbon dioxide [51,52]. This additional impact on human health emphasises the significance of addressing these pollutants in environmental regulations and public health interventions.

Our study has several major strengths. To the best of our knowledge, it is the first one conducted in China to explore the relationship between ambient fine particulate matter chemical components and HF, focusing on in-hospital mortality, hospitalisation cost, and LOS in HF patients, and thus assesses the health burden of air pollution exposure on HF from multiple perspectives. Our results can provide preliminary evidence for policymakers and regulatory authorities to establish threshold levels for these components and to implement control measures for emission sources. Compared with developed regions in the West (Europe and North America), China experiences more severe air pollution exposure, making it conducive to studying the toxicity of different ambient fine particulate matter chemical components. Moreover, the exposure assessment in this study was conducted at the individual level with high temporal (daily) and spatial (10km ×10km grid) resolution, resulting in greater accuracy in measurements.

However, few limitations should be noted. As this is a cross-sectional study, we could not establish a causal relationship between exposure and outcomes. We also only included HF patients from secondary and tertiary hospitals, which inevitably introduced selection bias. The findings were based on data from Shanxi Province, which limited the applicability to other provinces and emerging nations. We should also acknowledge that certain potential confounders, such as cigarette smoking, alcohol consumption, and physical activity, could not be completely ruled out in our study due to data constraints. However, the inclusion of surrogate variables (such as age, sex, and occupation) in the regression models might have helped mitigate potential biases. Likewise, we primarily focused on hospitalised HF patients who may also have various other coexisting conditions, some of which might similarly be affected by air pollutants. However, we could not provide explanations for the relationship between PM2.5 and its chemical components and these additional diseases due to the study design. The Tracking Air Pollution data set primarily focused on monitoring outdoor concentrations of air pollutants, so we could not assess the concentrations of indoor PM2.5 chemical components. Finally, the high degree of so between components and the absence of suitable multi-component statistical models limited our ability to explore potential interactions among various PM2.5 chemical components.

CONCLUSIONS

We found that elevated concentrations of PM2.5 chemical components, particularly OM and BC, are associated with increased hospitalisation mortality, costs, and LOS among HF patients. It is crucial to establish guidelines for PM2.5 chemical components and regulate the emissions of the most perilous constituents to enhance the health condition of this population.

Additional material

Online Supplementary Document

Acknowledgements

We are grateful to acknowledge those who participated in the data collection and management of this study.

Ethics statement: This study received ethical approval from the institutional review board of the School of Public Health, Sun Yat-sen University.

Data availability: The data from this study is available at: https://10.6084/m9.figshare.24787392. This comprises the following underlying data: Data file 1. (data) and Data file 2. (description of data.). Data are available under the terms of the Creative Commons Zero ‘No rights reserved’ data waiver (CC0 1.0 Public domain dedication).

[1] Funding: This work was supported by the by the Bill & Melinda Gates Foundation (Grant Number: INV-016826) and National Natural Science Foundation of China (82373534). Under the grant conditions of the foundation, a Creative Commons Attribution 4.0 Generic License has already been assigned to the Author Accepted Manuscript version that might arise from this submission.

[2] Authorship contributions: MC and HL contributed to acquisition of the data. ZH conducted data analysis and writing of the draft. SS, DS, JF, and GW helped with data analysis and visualization of results. JY, MC, and HL contributed to the draft edition and review. All authors have read and approved the final version of the manuscript and agree with the order of presentation of the authors.

[3] Disclosure of interest: The authors completed the ICMJE Disclosure of Interest Form (available upon request from the corresponding author) and disclose no relevant interests.

references

[1] B Ziaeian and GC Fonarow. Epidemiology and aetiology of heart failure. Nat Rev Cardiol. 2016;13:368-78. DOI: 10.1038/nrcardio.2016.25. [PMID:26935038]

[2] JJ McMurray and MA Pfeffer. Heart failure. Lancet. 2005;365:1877-89. DOI: 10.1016/S0140-6736(05)66621-4. [PMID:15924986]

[3] R Lozano, M Naghavi, K Foreman, S Lim, K Shibuya, and V Aboyans. Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380:2095-128. DOI: 10.1016/S0140-6736(12)61728-0. [PMID:23245604]

[4] J Rajadurai, HF Tse, CH Wang, NI Yang, J Zhou, and D Sim. Understanding the Epidemiology of Heart Failure to Improve Management Practices: An Asia-Pacific Perspective. J Card Fail. 2017;23:327-39. DOI: 10.1016/j.cardfail.2017.01.004. [PMID:28111226]

[5] C Cook, G Cole, P Asaria, R Jabbour, and DP Francis. The annual global economic burden of heart failure. Int J Cardiol. 2014;171:368-76. DOI: 10.1016/j.ijcard.2013.12.028. [PMID:24398230]

[6] L Li, R Liu, C Jiang, X Du, MD Huffman, and CSP Lam. Assessing the evidence-practice gap for heart failure in China: the Heart Failure Registry of Patient Outcomes (HERO) study design and baseline characteristics. Eur J Heart Fail. 2020;22:646-60. DOI: 10.1002/ejhf.1630. [PMID:31820513]

[7] KE Cosselman, A Navas-Acien, and JD Kaufman. Environmental factors in cardiovascular disease. Nat Rev Cardiol. 2015;12:627-42. DOI: 10.1038/nrcardio.2015.152. [PMID:26461967]

[8] F Liang, F Liu, K Huang, X Yang, J Li, and Q Xiao. Long-Term Exposure to Fine Particulate Matter and Cardiovascular Disease in China. J Am Coll Cardiol. 2020;75:707-17. DOI: 10.1016/j.jacc.2019.12.031. [PMID:32081278]

[9] M Cai, X Lin, X Wang, S Zhang, ZM Qian, and SE McMillin. Ambient particulate matter pollution of different sizes associated with recurrent stroke hospitalization in China: A cohort study of 1.07 million stroke patients. Sci Total Environ. 2023;856:159104. DOI: 10.1016/j.scitotenv.2022.159104. [PMID:36208745]

[10] X Lin, M Cai, K Tan, E Liu, X Wang, and C Song. Ambient particulate matter and in-hospital case fatality of acute myocardial infarction: A multi-province cross-sectional study in China. Ecotoxicol Environ Saf. 2023;268:115731. DOI: 10.1016/j.ecoenv.2023.115731. [PMID:38007949]

[11] Y Wei, Y Wang, Q Di, C Choirat, Y Wang, and P Koutrakis. Short term exposure to fine particulate matter and hospital admission risks and costs in the Medicare population: time stratified, case crossover study. BMJ. 2019;367:l6258 DOI: 10.1136/bmj.l6258. [PMID:31776122]

[12] J Rentschler and N Leonova. Global air pollution exposure and poverty. Nat Commun. 2023;14:4432 DOI: 10.1038/s41467-023-39797-4. [PMID:37481598]

[13] CK Ward-Caviness, AM Weaver, M Buranosky, ER Pfaff, LM Neas, and RB Devlin. Associations Between Long-Term Fine Particulate Matter Exposure and Mortality in Heart Failure Patients. J Am Heart Assoc. 2020;9:e012517. DOI: 10.1161/JAHA.119.012517. [PMID:32172639]

[14] X Li, H Cheng, Y Fang, Z Chen, G Qi, and R Chen. Association between fine particulate matter and heart failure hospitalizations: a time-series analysis in Yancheng, China. Environ Sci Pollut Res Int. 2021;28:26906-12. DOI: 10.1007/s11356-021-12428-5. [PMID:33501575]

[15] CK Ward-Caviness, M Danesh Yazdi, J Moyer, AM Weaver, WE Cascio, and Q Di. Long-Term Exposure to Particulate Air Pollution Is Associated With 30-Day Readmissions and Hospital Visits Among Patients With Heart Failure. J Am Heart Assoc. 2021;10:e019430. DOI: 10.1161/JAHA.120.019430. [PMID:33942627]

[16] M Wang, T Zhou, Y Song, X Li, H Ma, and Y Hu. Joint exposure to various ambient air pollutants and incident heart failure: a prospective analysis in UK Biobank. Eur Heart J. 2021;42:1582-91. DOI: 10.1093/eurheartj/ehaa1031. [PMID:33527989]

[17] FJ Kelly. Oxidative stress: its role in air pollution and adverse health effects. Occup Environ Med. 2003;60:612-6. DOI: 10.1136/oem.60.8.612. [PMID:12883027]

[18] M Sørensen, B Daneshvar, M Hansen, LO Dragsted, O Hertel, and L Knudsen. Personal PM2.5 exposure and markers of oxidative stress in blood. Environ Health Perspect. 2003;111:161-6. DOI: 10.1289/ehp.111-1241344. [PMID:12573899]

[19] M Cai, E Liu, H Tao, Z Qian, X Lin, and Z Cheng. Does Level of Hospital Matter? A Study of Mortality of Acute Myocardial Infarction Patients in Shanxi, China. Am J Med Qual. 2018;33:185-92. DOI: 10.1177/1062860617708608. [PMID:28592153]

[20] M Cai, E Liu, P Bai, N Zhang, S Wang, and W Li. The Chasm in Percutaneous Coronary Intervention and In-Hospital Mortality Rates Among Acute Myocardial Infarction Patients in Rural and Urban Hospitals in China: A Mediation Analysis. Int J Public Health. 2022;67:1604846. DOI: 10.3389/ijph.2022.1604846. [PMID:35872707]

[21] D Cao, D Li, Y Wu, ZM Qian, Y Liu, and Q Liu. Ambient PM2.5 exposure and hospital cost and length of hospital stay for respiratory diseases in 11 cities in Shanxi Province, China. Thorax. 2021;thoraxjnl-2020-215838. [PMID:34088786]

[22] D Cao, D Zheng, ZM Qian, H Shen, Y Liu, and Q Liu. Ambient sulfur dioxide and hospital expenditures and length of hospital stay for respiratory diseases: A multicity study in China. Ecotoxicol Environ Saf. 2022;229:113082. DOI: 10.1016/j.ecoenv.2021.113082. [PMID:34929503]

[23] S Liu, G Geng, Q Xiao, Y Zheng, X Liu, and J Cheng. Tracking Daily Concentrations of PM2.5 Chemical Composition in China since 2000. Environ Sci Technol. 2022;56:16517-27. DOI: 10.1021/acs.est.2c06510. [PMID:36318737]

[24] G Geng, Q Zhang, D Tong, M Li, Y Zheng, and S Wang. Chemical composition of ambient PM2. 5 over China and relationship to precursor emissions during 2005–2012. Atmos Chem Phys. 2017;17:9187-203. DOI: 10.5194/acp-17-9187-2017

[25] Tracking Air Pollution in China. Near real-time air pollutant concentrations. 2024. Available: http://tapdata.org.cn/. Accessed: 16 January 2024.

[26] M Cai, X Lin, X Wang, S Zhang, C Wang, and Z Zhang. Long-term exposure to ambient fine particulate matter chemical composition and in-hospital case fatality among patients with stroke in China. Lancet Reg Health West Pac. 2023;32:100679. DOI: 10.1016/j.lanwpc.2022.100679. [PMID:36785852]

[27] M Cai, J Wei, S Zhang, W Liu, L Wang, and Z Qian. Short-term air pollution exposure associated with death from kidney diseases: a nationwide time-stratified case-crossover study in China from 2015 to 2019. BMC Med. 2023;21:32 DOI: 10.1186/s12916-023-02734-9. [PMID:36694165]

[28] W Liu, J Wei, M Cai, Z Qian, Z Long, and L Wang. Particulate matter pollution and asthma mortality in China: A nationwide time-stratified case-crossover study from 2015 to 2020. Chemosphere. 2022;308:136316. DOI: 10.1016/j.chemosphere.2022.136316. [PMID:36084833]

[29] M Cai, S Zhang, X Lin, Z Qian, SE McMillin, and Y Yang. Association of Ambient Particulate Matter Pollution of Different Sizes With In-Hospital Case Fatality Among Stroke Patients in China. Neurology. 2022;98:e2474-86. DOI: 10.1212/WNL.0000000000200546. [PMID:35613931]

[30] M Cai, Y Xie, B Bowe, AK Gibson, MA Zayed, and T Li. Temporal Trends in Incidence Rates of Lower Extremity Amputation and Associated Risk Factors Among Patients Using Veterans Health Administration Services From 2008 to 2018. JAMA Netw Open. 2021;4:e2033953. DOI: 10.1001/jamanetworkopen.2020.33953. [PMID:33481033]

[31] A Urban, C Di Napoli, HL Cloke, J Kyselý, F Pappenberger, and F Sera. Evaluation of the ERA5 reanalysis-based Universal Thermal Climate Index on mortality data in Europe. Environ Res. 2021;198:111227. DOI: 10.1016/j.envres.2021.111227. [PMID:33974842]

[32] D Jiao, N Xu, F Yang, and K Xu. Evaluation of spatial-temporal variation performance of ERA5 precipitation data in China. Sci Rep. 2021;11:17956 DOI: 10.1038/s41598-021-97432-y. [PMID:34504211]

[33] KG Burkart, M Brauer, AY Aravkin, WW Godwin, SI Hay, and J He. Estimating the cause-specific relative risks of non-optimal temperature on daily mortality: a two-part modelling approach applied to the Global Burden of Disease Study. Lancet. 2021;398:685-97. DOI: 10.1016/S0140-6736(21)01700-1. [PMID:34419204]

[34] M Lippmann, LC Chen, T Gordon, K Ito, and GD Thurston. National Particle Component Toxicity (NPACT) Initiative: integrated epidemiologic and toxicologic studies of the health effects of particulate matter components. Res Rep Health Eff Inst. 2013;177:5-13. [PMID:24377209]

[35] S Vedal, MJ Campen, JD McDonald, TV Larson, PD Sampson, and L Sheppard. National Particle Component Toxicity (NPACT) initiative report on cardiovascular effects. Res Rep Health Eff Inst. 2013;178:5-8. [PMID:24377210]

[36] B Ostro, M Lipsett, P Reynolds, D Goldberg, A Hertz, and C Garcia. Long-term exposure to constituents of fine particulate air pollution and mortality: results from the California Teachers Study. Environ Health Perspect. 2010;118:363-9. DOI: 10.1289/ehp.0901181. [PMID:20064787]

[37] L Liu, Y Zhang, Z Yang, S Luo, and Y Zhang. Long-term exposure to fine particulate constituents and cardiovascular diseases in Chinese adults. J Hazard Mater. 2021;416:126051. DOI: 10.1016/j.jhazmat.2021.126051. [PMID:34492892]

[38] B Ostro, J Hu, D Goldberg, P Reynolds, A Hertz, and L Bernstein. Associations of mortality with long-term exposures to fine and ultrafine particles, species and sources: results from the California Teachers Study Cohort. Environ Health Perspect. 2015;123:549-56. DOI: 10.1289/ehp.1408565. [PMID:25633926]

[39] Y Chung, F Dominici, Y Wang, BA Coull, and ML Bell. Associations between long-term exposure to chemical constituents of fine particulate matter (PM2.5) and mortality in Medicare enrollees in the eastern United States. Environ Health Perspect. 2015;123:467-74. DOI: 10.1289/ehp.1307549. [PMID:25565179]

[40] EF Kirrane, TJ Luben, A Benson, EO Owens, JD Sacks, and SJ Dutton. A systematic review of cardiovascular responses associated with ambient black carbon and fine particulate matter. Environ Int. 2019;127:305-16. DOI: 10.1016/j.envint.2019.02.027. [PMID:30953813]

[41] RB Schlesinger. The health impact of common inorganic components of fine particulate matter (PM2.5) in ambient air: a critical review. Inhal Toxicol. 2007;19:811-32. DOI: 10.1080/08958370701402382. [PMID:17687714]

[42] Y Yang, Z Ruan, X Wang, Y Yang, TG Mason, and H Lin. Short-term and long-term exposures to fine particulate matter constituents and health: A systematic review and meta-analysis. Environ Pollut. 2019;247:874-82. DOI: 10.1016/j.envpol.2018.12.060. [PMID:30731313]

[43] CA Pope, RT Burnett, GD Thurston, MJ Thun, EE Calle, and D Krewski. Cardiovascular mortality and long-term exposure to particulate air pollution: epidemiological evidence of general pathophysiological pathways of disease. Circulation. 2004;109:71-7. DOI: 10.1161/01.CIR.0000108927.80044.7F. [PMID:14676145]

[44] A Fiordelisi, P Piscitelli, B Trimarco, E Coscioni, G Iaccarino, and D Sorriento. The mechanisms of air pollution and particulate matter in cardiovascular diseases. Heart Fail Rev. 2017;22:337-47. DOI: 10.1007/s10741-017-9606-7. [PMID:28303426]

[45] T Münzel, T Gori, S Al-Kindi, J Deanfield, J Lelieveld, and A Daiber. Effects of gaseous and solid constituents of air pollution on endothelial function. Eur Heart J. 2018;39:3543-50. DOI: 10.1093/eurheartj/ehy481. [PMID:30124840]

[46] X Lei, R Chen, C Wang, J Shi, Z Zhao, and W Li. Personal Fine Particulate Matter Constituents, Increased Systemic Inflammation, and the Role of DNA Hypomethylation. Environ Sci Technol. 2019;53:9837-44. DOI: 10.1021/acs.est.9b02305. [PMID:31328512]

[47] W Li, EH Wilker, KS Dorans, MB Rice, J Schwartz, and BA Coull. Short-Term Exposure to Air Pollution and Biomarkers of Oxidative Stress: The Framingham Heart Study. J Am Heart Assoc. 2016;5:e002742. DOI: 10.1161/JAHA.115.002742. [PMID:27126478]

[48] P Møller, NR Jacobsen, JK Folkmann, PH Danielsen, L Mikkelsen, and JG Hemmingsen. Role of oxidative damage in toxicity of particulates. Free Radic Res. 2010;44:1-46. DOI: 10.3109/10715760903300691. [PMID:19886744]

[49] RD Brook, S Rajagopalan, CA Pope, JR Brook, A Bhatnagar, and AV Diez-Roux. Particulate matter air pollution and cardiovascular disease: An update to the scientific statement from the American Heart Association. Circulation. 2010;121:2331-78. DOI: 10.1161/CIR.0b013e3181dbece1. [PMID:20458016]

[50] TJ Grahame and RB Schlesinger. Cardiovascular health and particulate vehicular emissions: a critical evaluation of the evidence. Air Qual Atmos Health. 2010;3:3-27. DOI: 10.1007/s11869-009-0047-x. [PMID:20376169]

[51] KR Smith, M Jerrett, HR Anderson, RT Burnett, V Stone, and R Derwent. Public health benefits of strategies to reduce greenhouse-gas emissions: health implications of short-lived greenhouse pollutants. Lancet. 2009;374:2091-103. DOI: 10.1016/S0140-6736(09)61716-5. [PMID:19942276]

[52] B Booth and N Bellouin. Climate change: Black carbon and atmospheric feedbacks. Nature. 2015;519:167-8. DOI: 10.1038/519167a. [PMID:25762278]

Correspondence to:
Hualiang Lin, PhD
Department of Epidemiology, School of Public Health, Sun Yat-sen University
Zhongshan 2rd road, Yuexiu District, Guangzhou, Guangdong
China
[email protected]