
==== Front
eBioMedicine
EBioMedicine
eBioMedicine
2352-3964
Elsevier

S2352-3964(24)00306-2
10.1016/j.ebiom.2024.105270
105270
Articles
Ambient particulate matter and chronic obstructive pulmonary disease mortality: a nationwide, individual-level, case-crossover study in China
Jiang Shuo abe
Tong Xunliang ce
Yu Kexin ae
Yin Peng c
Shi Su a
Meng Xia a
Chen Renjie a
Zhou Maigeng c
Kan Haidong a
Niu Yue niuy@fudan.edu.cn
a∗∗
Li Yanming liyanming2632@bjhmoh.cn
d∗
a School of Public Health, Key Lab of Public Health Safety of the Ministry of Education and NHC Key Lab of Health Technology Assessment, Fudan University, Shanghai, China
b Shanghai Pulmonary Hospital, Tongji University School of Medicine, Shanghai, China
c National Center for Chronic Noncommunicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention, Beijing, China
d Department of Respiratory and Critical Care Medicine, Beijing Hospital, National Center of Gerontology; Institute of Geriatric Medicine, Chinese Academy of Medical, Beijing, China
∗ Corresponding author. Department of Respiratory and Critical Care Medicine, Beijing Hospital, National Center of Gerontology, Institute of Geriatric Medicine, Chinese Academy of Medical Sciences, Beijing 100730, China. liyanming2632@bjhmoh.cn
∗∗ Corresponding author. Department of Environmental Health, School of Public Health, Fudan University, Shanghai 200032, China. niuy@fudan.edu.cn
e Contributed equally to this work.

12 8 2024
9 2024
12 8 2024
107 1052703 4 2024
25 7 2024
25 7 2024
© 2024 The Author(s)
2024
https://creativecommons.org/licenses/by-nc/4.0/ This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).
Summary

Background

Short-term exposure to particulate matter air pollution has been associated with the exacerbations of COPD, but its association with COPD mortality was not fully elucidated. We aimed to assess the association between short-term particulate matter exposure and the risk of COPD mortality in China using individual-level data.

Methods

We derived 2.26 million COPD deaths from a national death registry database in Chinese mainland between 2013 and 2019. Exposures to fine particulate matter (PM2.5) and coarse particulate matter (PM2.5-10) were assessed by satellite-based models of a 1 × 1 km resolution and assigned to each individual based on residential address. The associations of PM2.5 and PM2.5-10 with COPD mortality were examined using a time-stratified case-crossover design and conditional logistic regressions with distributed lag models. We further conducted stratified analyses by age, sex, education level, and season.

Findings

Short-term exposures to both PM2.5 and PM2.5-10 were associated with increased risks of COPD mortality. These associations appeared and peaked on the concurrent day, attenuated and became nonsignificant after 5 or 7 days, respectively. The exposure-response curves were approximately linear without discernible thresholds. An interquartile range increase in PM2.5 and PM2.5-10 concentrations was associated with 4.23% (95% CI: 3.75%, 4.72%) and 2.67% (95% CI: 2.18%, 3.16%) higher risks of COPD mortality over lag 0–7 d, respectively. The associations of PM2.5 and PM2.5-10 attenuated slightly but were still significant in the mutual-adjustment models. A larger association of PM2.5-10 was observed in the warm season.

Interpretation

This individual-level, nationwide, case-crossover study suggests that short-term exposure to PM2.5 and PM2.5-10 might act as one of the environmental risk factors for COPD mortality.

Funding

This study is supported by the 10.13039/501100012166 National Key Research and Development Program of China (2023YFC3708304 and 2022YFC3702701 ), the 10.13039/501100001809 National Natural Science Foundation of China (82304090 and 82030103 ), the 3-year Action Plan for Strengthening the Construction of the Public Health System in Shanghai (GWVI-11.2-YQ31 ), and the 10.13039/501100003399 Science and Technology Commission of Shanghai Municipality (21TQ015 ).

Keywords

Fine particulate matter
Coarse particulate matter
COPD
Mortality
Case-crossover study
==== Body
pmc Research in context

Evidence before this study

We searched PubMed and Web of Science up to March 19, 2024 using search terms “(COPD OR chronic obstructive pulmonary disease) AND (mortality OR death) AND (particulate matter OR PM2.5 OR PM2.5-10) AND (short-term OR acute)” without language restrictions. Previous evidence linking short-term exposure to particulate matter exposure to COPD mortality was mainly derived from ecological time-series studies based on aggregated-level data, in which measurement errors in both exposure and outcome and unmeasured confounders can be major sources of uncertainties. Moreover, most previous investigations have only considered the acute effects of fine particulate matter on COPD mortality, while there is limited evidence on coarse particulate matter.

Added value of this study

In this individual-level, nationwide, case-crossover study including 2.26 million COPD deaths, we found that short-term exposure to both fine particulate matter and coarse particulate matter could independently increase the risk of COPD mortality. Such effects could last for 5–7 days. We further observed a stronger effect of coarse particulate matter during the warm season.

Implications of all the available evidence

Our findings underscore the role of short-term exposure to particulate matter in COPD mortality and highlight the need for ongoing COPD care and management within one week after a single episode of high particulate matter pollution.

Introduction

Chronic obstructive pulmonary disease (COPD) was the fourth- and third-ranked cause of deaths in the 50–74-year and 75-year-and-older age groups in 2019.1 As one of leading causes of death, COPD poses a substantial burden on public health in the context of population aging. According to the Global Initiative for Chronic Obstructive Lung Disease (GOLD) Scientific Committee, air pollution is one of the leading causes of COPD exacerbations, which in turn is linked to an increased risk of death among patients with COPD.2,3 The Global Burden of Disease Study (GBD) estimated that approximately 27.5% of COPD-related deaths can be attributed to ambient particulate matter (PM) pollution.4 This estimate increases to approximately 31.5% in China,5 where the annual mean population-weighted PM2.5 concentration in 2000–2019 was approximately 2–3 times higher than those in Europe and Northern America.6

Patients with COPD are susceptible to acute fluctuations in air pollution levels. A recent meta-analysis concluded that short-term exposure to particles with an aerodynamic diameter ≤ 2.5 μm (PM2.5) or ≤ 10 μm (PM10) could significantly increase the risk of COPD exacerbations,7 which is one of the causative factors of COPD mortality. However, only a few studies have directly assessed the associations between PMs and COPD mortality and the results were not fully consistent.8, 9, 10, 11, 12 Notably, the limited evidence regarding the short-term effects of PMs on COPD mortality comes almost exclusively from ecological time-series studies. In ecological studies, confounding bias is inevitable as individual-level time-invariant confounders, such as age, smoking status, and education, cannot be accounted for. Considering events and exposures at the aggregated level also introduces a considerable risk of measurement/exposure misclassification. Furthermore, most previous investigations have only considered particles of a single size,10,13,14 and few studies have attempted to differentiate between the effects of coarse (PM2.5-10) and fine (PM2.5) particles, which typically originate from different sources.

To address this knowledge gap, we conducted an individual-level, nationwide, case-crossover analysis to assess the short-term effects of PM2.5 and PM2.5-10 on COPD mortality. We also aimed to explore the exposure-response relationships and identify potentially susceptible populations. The results will contribute to the proposal of more accurate and targeted prevention strategies and early warning systems.

Methods

Study population

In this study, we used health data from the China Cause of Death Reporting System (CDRS), an Internet-based national death registry system operated by the Chinese Center for Disease Control and Prevention (CDC). The CDRS collected death data from 2844 county-level administrative units in 31 provinces in China.15,16 All deaths occurring in these counties were mandated to be registered in the CDRS by staff in the disease prevention unit of the hospitals or primary health care institutions. The underlying cause of death was coded by local CDC staff according to the International Classification of Diseases 10th Revision (ICD-10) since 2004. Regular training and supervision have been conducted at all levels of administration, from township to national level, to ensure the quality of each death certificate reported. As a routine quality control measure, a sample of death certificates was randomly selected by higher-level CDCs every month to verify the underlying cause of death. Each death record contains the date of death, the underlying cause of death, and personal information, such as age, sex, residential address, education level. Data from this system are of high quality and have been widely used in environmental epidemiological research.17,18

COPD mortality data

We retrieved anonymous personal data for all decedents whose underlying cause of death was COPD (J44.0, J44.1, J44.8 and J44.9) from January 1, 2013 to December 31, 2019. Among the extracted COPD death cases, 127,285 (5.3%) were excluded due to lack of data on residential address, and the remaining 2,260,692 cases were included in the analysis.

Study design

We adopted a time-stratified case-crossover design to assess the short-term effects of PM2.5 and PM2.5-10 on COPD mortality, in which the case day was defined as the day of death and the control day was selected as the day with the same day-of-week within the same month of death. Because of the nature of self-comparison, this design has the advantage of adjusting for potential confounding by factors that remain constant over a short period of time, such as individual-level socio-demographic characteristics, day of the week, and seasonal patterns.17

Exposure assessment

Daily concentrations (24-h average) of PM2.5 and PM10 were predicted at 1 × 1 km spatial resolution using previously constructed random forest models. In constructing the models, we used the daily PM2.5 and PM10 measurements as the dependent variable and used a variety of predictors as the independent variables, including aerosol optical depth, meteorological parameters, land use data, and other ancillary variables. The cross-validation R2 and root mean squared error between the predictions and their corresponding measurements were 0.84 and 16.09 μg/m3 for PM2.5, and 0.87 and 22.07 μg/m3 for PM10, respectively, indicating a relatively high accuracy in prediction. Further details on the exposure models were presented elsewhere.19,20 We calculated the concentrations of particles with an aerodynamic diameter between 2.5 μm and 10 μm (PM2.5-10) by subtracting the daily PM2.5 concentration from the daily PM10 concentration.21 We also predicted the daily concentrations of nitrogen dioxide (NO2) and ozone (O3) using previously constructed models with a spatial resolution of 1 × 1 km and obtained the daily concentrations of carbon oxide (CO) and sulfur dioxide (SO2) from the nearest fixed-site monitoring station.22,23 In addition, we obtained daily mean temperature and relative humidity during the study period from the European Center for Medium-Range Weather Forecasts Reanalysis Fifth Generation (ERA5) reanalysis product at 0.1° × 0.1° spatial resolution.24 We assigned the environmental data to each death according to the residential address and the date of the case day and control day.

Statistical analysis

The associations between PMs and COPD mortality were examined by comparing exposures on or before the case day with exposures on or before the corresponding control days. In brief, we used conditional logistic regression models combined with distributed lag models (DLMs) to examine the lagged effects of PMs on COPD mortality. The DLM was first applied to generate cross-bases for PM2.5 and PM2.5-10, respectively, to account for linear and lagged effects. In constructing the cross-basis, a third-degree polynomial function was used for lag space,25 and a maximum lag of 7 days was chosen based on a priori knowledge that the acute effects of air pollution are typically restricted within one week after exposure.17,26 The cross-basis function was then introduced to the conditional logistic regression models. Average temperature and relative humidity on the day of the case or control day and the previous 2 days (lag 0–2 d) were controlled for using natural cubic spline functions with 6 and 3 degrees of freedom, respectively. We also adjusted for public holidays in the models as it may affect traffic patterns, air pollutant emissions and concentrations, health-related behaviors, and availability of health-care services. In addition, to plot cumulative exposure-response curves more flexibly, we used distributed lag nonlinear models (DLNM) to construct cross-bases for PMs. Specifically, in the DLNM, both the exposure-response and lag-response functions were constructed by using natural cubic splines, with 2 equally spaced internal knots in the PM concentration range and 2 equally spaced internal knots in the log-value of the lag. The linearity of the curves was examined by comparing the non-linear models with the corresponding linear models using a likelihood ratio test.

To explore the potential heterogeneity in the PM-COPD mortality associations across subgroups, we performed stratified analyses by age (younger adults versus older adults), sex (male versus female), season (warm versus cold), and education level (junior high school and below versus high school and above). Younger and older adults were defined as those aged < 65 years and ≥ 65 years, respectively, as in previous studies.27,28 Warm and cold seasons were defined as April to September and October to March, respectively. Between-stratum differences were tested using a z-test based on the following formula:z=β1−β2SE12+SE22

where β1 and β2 are the stratum-specific coefficients of PM in the conditional logistic models, and SE1 and SE2 are their standard errors.29

In sensitivity analysis, we adjusted for other air pollutants (NO2, O3, CO, and SO2), as well as mutually adjusted for PM2.5-10 and PM2.5. The average concentrations of criteria air pollutants on the day of the case or control day and the previous 2 days (lag 0–2 d) were included in the above-mentioned model one at a time, in line with a previous study.18 Then, we also applied multiple degrees of freedom (2 and 4) for the lag-response dimension in the DLMs. All statistical analyses were conducted using R software (version 4.1.2). Effect estimates were presented as percent changes in the risk of COPD mortality associated with each interquartile range (IQR) increment in PM levels, which were calculated using the following equations:Percentchange=(eβ×IQR−1)×100%

Lower95%CI=(e(β−1.96×SE)×IQR−1)×100%

Upper95%CI=(e(β+1.96×SE)×IQR−1)×100%

where β is the regression coefficient derived from the conditional logistic regression models and SE is the corresponding standard error. Two-sided tests with p-value less than 0.05 were considered statistically significant.

Role of funders

The funders of this study had no role in the study design, data collection, analysis, interpretation of results, or manuscript drafting.

Results

Descriptive statistics

This analysis included approximately 2.26 million COPD deaths between 2013 and 2019 in Chinese mainland. As shown in Table 1, of the 2,260,692 cases of COPD deaths, males accounted for 59.2%. The majority of the decedents were aged 65 years or above (91.8%), with a mean age of 79.6 years. Only a minority had an education level of high school and above (5.4%).Table 1 Statistics of death cases due to chronic obstructive pulmonary disease.

Population characteristics	Statistics	
No. of cases, n (%)	2,260,692 (100)	
Age, years, mean (standard deviation)	79.6 (9.78)	
Age, n (%)		
 <65	184,476 (8.2)	
 ≥65	2,076,206 (91.8)	
Sex, n (%)		
 Male	1,337,757 (59.2)	
 Female	922,935 (40.8)	
Education, n (%)		
 Junior high school and below	2,1219,596 (93.9)	
 High school and above	121,032 (5.4)	

Table 2 summarizes the distribution of residential PM2.5 and PM2.5-10 exposures and meteorological factors on the day of COPD deaths. The average daily mean concentrations of PM2.5 and PM2.5-10 were 49.9 (IQR: from 27.6 to 62.4) and 35.8 (IQR: from 20.0 to 45.1) μg/m3, respectively. The daily average temperature and relative humidity levels were 14.2 °C (IQR: from 7.2 to 22.1) and 68.3% (IQR: from 57.4% to 81.9%), respectively. There were moderate positive correlations between PM2.5 and PM2.5-10, and between them and gaseous air pollutants, except for O3, with Pearson correlation coefficients ranging from 0.42 to 0.72 (Supplementary Table S1). PM2.5 and PM2.5-10 concentrations were negatively correlated with O3 concentrations, ambient temperature, and relative humidity.Table 2 Statistics for particulate matter and meteorological parameters on the day of death.

Environmental exposure	Mean	SD	5th	25th	50th	75th	95th	
Particulate matter								
 PM2.5 (μg/m3)	49.9	32.4	16.7	27.6	41.1	62.4	112.7	
 PM2.5-10 (μg/m3)	35.8	23.6	11.5	20.0	29.9	45.1	78.5	
Meteorological parameters								
 Temperature (°C)	14.2	9.7	−1.8	7.2	14.9	22.1	28.3	
 Relative humidity (%)	68.3	15.5	33.7	57.4	71.7	81.9	91.5	
SD, standard deviation; 5th, 25th, 50th, 75th, and 95th, the 5th, 25th 50th, 75th and 95th percentiles; PM2.5, fine particulate matter; PM2.5-10, coarse particulate matter.

Lagged effects of PM2.5 and PM2.5-10 on COPD mortality

Fig. 1 shows the percent changes in COPD mortality risk associated with an IQR increment in PM2.5 and PM2.5-10 concentrations during different lag periods. The effect of PM2.5 on COPD mortality appeared and peaked on the concurrent day of exposure, then decreased and persisted for 7 days, whereas the effect of PM2.5-10 demonstrated a similar lag pattern but with a shorter lag period, i.e., the effect became nonsignificant after 5 days of exposure. We therefore used lag 0–7 d in estimating the cumulative effects of PM2.5 and PM2.5-10 to keep the results comparable. At this lag, the risk of COPD mortality was 4.23% [95% confidence interval (CI): 3.75%, 4.72%] higher per IQR (34.9 μg/m3) increase in PM2.5 concentrations, and 2.67% (95% CI: 2.18%, 3.16%) higher per IQR (25.1 μg/m3) increase in PM2.5-10 concentrations.Fig. 1 Lag patterns for the associations of PM2.5 (a) and PM2.5-10 (b) concentrations with COPD mortality. Abbreviation: PM2.5, fine particulate matter; PM2.5–10, coarse particulate matter; COPD, chronic obstructive pulmonary disease. The black solid lines correspond to the average percent changes in the risk of COPD mortality associated with each interquartile range increase in particulate matter concentrations and the blue areas are their 95% confidence intervals.

Fig. 2 exhibits the cumulative exposure-response curves for the association between short-term PM exposures and COPD mortality over lag 0–7 d. In general, the risk of death from COPD increases consistently with increasing exposure to both PM2.5 and PM2.5-10. The two concentration–response curves were approximately linear without thresholds (p-values for the likelihood ratio test: 0.443 for PM2.5 and 0.514 for PM2.5-10). Compared with PM2.5-10, the cumulative effect of PM2.5 was slightly greater.Fig. 2 Cumulative exposure-response curves for the associations of PM2.5 (a) and PM2.5-10 (b) concentrations with COPD mortality over lag 0–7 d. Abbreviation: PM2.5, fine particulate matter; PM2.5–10, coarse particulate matter; COPD, chronic obstructive pulmonary disease. The black solid lines correspond to the average percent change in the risk of COPD mortality and the blue areas are their 95% confidence intervals.

Table 3 presents the relationship between PM2.5 and PM2.5-10 concentration and COPD mortality in different subgroups of the population. The associations of PM2.5 and PM2.5-10 with COPD mortality were slightly larger in females and those who aged ≥ 65 years but the between-stratum differences did not reach statistical significance. The effects of PMs on COPD mortality were relatively smaller in the more educated population, and the effect of PM2.5-10 even tended to be null among those with high school education or above. Moreover, the effect of PM2.5-10 was significantly higher in the warm season than in the cold season (3.34% and 1.92% per IQR increase, respectively; p-value for between-stratum difference < 0.01), while the effect of PM2.5 did not significantly differ between seasons.Table 3 Risks of COPD mortality associated with each interquartile range increase in PM2.5 and PM2.5-10 concentrations over lag 0–7 d, stratified by sex, age, education level, and season.

	PM2.5	PM2.5-10	
Overall	4.23 (3.75, 4.72)	2.67 (2.18, 3.16)	
Subgroup			
 Sex			
 Male	3.86 (3.22, 4.52)	2.43 (1.80, 3.07)	
 Female	5.05 (4.27, 5.83)	3.12 (2.34, 3.91)	
 Age, year			
 <65	3.18 (1.50, 4.89)	1.87 (1.70, 3.59)	
 ≥65	4.46 (3.94, 4.98)	2.78 (2.27, 3.29)	
 Education level			
 Junior high school and below	4.53 (4.01, 5.05)	2.72 (2.21, 3.23)	
 High school and above	2.75 (0.81, 4.72)	1.77 (−0.29, 3.88)	
 Season			
 Warm	4.68 (3.86, 5.50)	3.34 (2.67, 4.03)a	
 Cold	4.39 (3.76,5.03)	1.92 (1.28, 2.57)a	
Abbreviation: PM2.5, fine particulate matter; PM2.5–10, coarse particulate matter; COPD, chronic obstructive pulmonary disease.

Note: interquartile range was 34.9 μg/m3 for PM2.5 and 25.1 μg/m3 for PM2.5–10.

a Significant between-stratum difference.

In our sensitivity analyses, the risk estimates of COPD mortality related to short-term PM2.5 and PM2.5-10 exposure remained robust. Specifically, the associations of PM2.5 and PM2.5-10 with COPD mortality remained statistically significant in the two-pollutant models, but the magnitude of the associations was attenuated after mutually adjusting for each other and additionally adjusting for NO2 concentrations (Fig. 3). Moreover, altering degrees of freedom for the lag-response dimension in the DLMs did not change the lag pattern of the associations (Supplementary Fig. S1).Fig. 3 Percent changes in the risk of COPD mortality associated with each interquartile range increase in PM2.5 (a) and PM2.5-10 (b) concentrations over lag 0–7 d in single- and two-pollutant models. Abbreviation: PM2.5, fine particulate matter; PM2.5–10, coarse particulate matter; NO2, nitrogen dioxide; CO, carbon monoxide; O3, ozone; SO2, sulfur dioxide; COPD, chronic obstructive pulmonary disease.

Discussion

In this individual-level, nationwide, time-stratified case-crossover study, we found that short-term exposures to PM2.5 and PM2.5-10 were independently associated with higher risks of COPD mortality, and the effects lasted for 7 and 5 days, respectively, after exposure. The exposure-response curves demonstrated approximately linear associations with no apparent thresholds. We further observed a stronger effect of PM2.5-10 during the warm season.

To date, evidence on the associations of PMs and the risk of COPD mortality is scarce and the results are inconsistent. For PM2.5, a number of studies have reported its positive associations with COPD mortality,8,10,12,14,30 whereas a few studies have found no significant associations.11,31,32 The inconsistencies observed might be attributed to the fact that most previous studies have been limited to a single city or a few cities. Moreover, the majority of prior investigations employed an aggregated-level time-series design, which might lead to ecological fallacies. A recent individual-level case-crossover study reported that short-term PM2.5 exposure was associated with an elevated risk of mortality in patients with COPD, but the study was limited to the winter seasons.33 In contrast, we used a time-stratified case-crossover design and conducted analyses based on a large, nationally representative population sample, as well as more precise exposure assessments. We found a higher risk of COPD mortality associated with increasing PM2.5 exposure, providing compelling evidence on the relationships between PM2.5 air pollution and COPD.

Compared with PM2.5, the evidence on PM2.5-10 and COPD is rather limited. In this study, we found that short-term exposure to PM2.5-10 was also associated with a higher risk of COPD mortality. The results remained almost unchanged after adjustments for PM2.5, suggesting independent effects of PM2.5-10 from PM2.5. Differentiating and comparing the effects of PM of different particle sizes is of great significance to public health. However, it is difficult to make such comparisons directly based on existing evidence, because sociodemographic characteristics, exposure assessment methodologies, and statistical models vary considerately across studies. Our study examined the effects of both PM2.5 and PM2.5-10 using the same statistical model, thus allowing us to directly compare their effects. We found a relatively larger effect of PM2.5 on COPD mortality than PM2.5-10, which was similar to another time-series study that also analyzed the impacts of both PM2.5 and PM2.5-10 on COPD mortality.34 This finding is supported by ample evidence that PMs of a smaller aerodynamic diameter exert more hazardous health effects, as smaller particles can penetrate more deeply to deposit on the respiratory tract at an increasing rate.35,36

In addition, we explored the lagged effect of PMs on COPD mortality. Previous ecological studies have considered different lags, including single lags and cumulative lags. A study in Australia observed a longer lag of 0–14 d, while two studies in China and the United States found shorter lags (lag 0–1 to lag 0–3 d).8,12,33 In our study, the detrimental effect of PMs on COPD mortality was not only present on the concurrent day but persisted until 5–7 days after the exposure, revealing important time windows for prevention on days of poor air quality. Ongoing COPD care and management is therefore recommended within one week after a single episode of high PMs pollution.

Our findings have biological plausibility. Epidemiological evidence indicated that acute PM exposure could trigger exacerbations of pre-existing COPD, leading to an elevated risk of mortality.37, 38, 39, 40 Moreover, previous studies suggested that deposition patterns in the lungs of patients with COPD may be altered by airway anatomy, leaving patients vulnerable to the adverse effects of PMs and more likely to deteriorate.38 Experimental studies also support that acute PM exposure induces detrimental effects through pulmonary inflammatory responses and epithelial injury.41 Specifically, this process may involve the release of tumor necrosis factor α and interleukin 1β from alveolar macrophages,42 as well as the release of interleukin 1β and interleukin 8 from lung epithelial cells.43 On the other hand, human studies have also shown an increase in inflammatory cytokines and chemokines in the airways following PM exposure.44,45

In subgroup analyses, we observed a larger effect of short-term exposure to PM2.5-10 on COPD mortality in the warm season than in the cold season. This finding is similar to a number of previous studies that revealed stronger effects of PMs on health outcomes during the warm season.46,47 One possible explanation might be the interaction between heatwaves and PM, with increased frequency of heatwaves leading to changes in aeroallergens, which in turn interact with PM to trigger COPD exacerbations.48 Previous studies have consistently shown a synergistic effect of air pollution and high temperatures on human health.48, 49, 50 Our finding highlights the continued need for targeted preventive measures for patients with COPD and control of particulate matter air pollution in the summer, which will benefit the public respiratory health. We also found that females and older adults appeared to be more susceptible to the adverse effect of PM on COPD mortality. The findings were consistent with previous research that found ambient PM pollution to be the most significant risk factor for COPD mortality in females.51 The differences between sex may be partly due to different smoking prevalence and airway deposition patterns between males and females.52 As for age, the observed susceptibility may be attributed to pre-existing respiratory or cardiovascular comorbidities. Furthermore, we observed relatively higher vulnerability to PM2.5 among the less educated, although the between-stratum difference was not significant. The larger effects in less-educated subgroup may be due to higher prevalence of concurrent chronic diseases, lower socioeconomic levels, and fewer healthcare services.53 However, considering the relatively small proportion of study population with higher education, the effect estimates should be interpreted with caution.

Our study has notable strengths. First, we utilized an authoritative national death registry database to assess short-term effects of PMs on COPD mortality at the individual level. The large sample size, nationwide coverage, and high quality of the mortality data provided validity and credibility to the effect estimates and exposure-response relationships. Second, our exposure assessment was markedly improved by the use of exposure models with high spatiotemporal resolution, which allowed coverage of suburban and rural areas where few fixed-site monitoring sites are available.

Nevertheless, the limitations of this study should also be addressed. First, individual information on comorbidity and medication use were unavailable in the death records. However, neither is likely to confound the associations because the case-crossover design could control for confounding by variables that remain unchanged over a short period of time (i.e., comorbidity) and also because changes in medication-use are randomized and independent of PM exposure. Second, as with most studies using registry data, coding errors and underreporting of deaths may occur. However, we considered that both misreporting and underreporting in this database is acceptable because all deaths were mandated to be registered in the CDRS and this system has a series of quality control or assurance measures to ensure the timeliness of death registration as well as the completeness and accuracy of entries. Moreover, such coding errors are likely to be randomized and therefore would only bias the results towards the null.54 Third, residual confounding due to time-varying risk factors (physical activity, alcohol consumption, etc.) may still exist. Lastly, we only assessed the level of ambient PMs instead of personal exposure, thus exposure misclassifications were inevitable.

In conclusion, this individual-level, nationwide, case-crossover study demonstrated that short-term exposure to PM2.5 and PM2.5-10 could increase the risk of COPD mortality in China. Our findings reinforce the evidence on modifiable environmental risk factors for COPD and thus carry significant implications for the management of COPD. During episodes of high levels of PMs, special attention and preventive strategies are warranted to mitigate the burden of respiratory mortality associated with PM2.5 and PM2.5-10.

Contributors

Shuo Jiang, Xunliang Tong, Kexin Yu, Yue Niu and Yanming Li designed the study. The underlying data were verified by Shuo Jiang, Xunliang Tong, Kexin Yu, Yue Niu and Yanming Li. Shuo Jiang, Xunliang Tong, and Kexin Yu analysed the data. Shuo Jiang, Xunliang Tong, and Kexin Yu wrote the first draft of the manuscript, which was revised by Renjie Chen, Maigeng Zhou, Haidong Kan, Yue Niu, Yanming Li. All authors interpreted data, provided critical review and revision of the text, and approved the final version of the manuscript. All authors had access to the data underlying the study and accept responsibility for the decision to submit for publication.

Data sharing statement

Health data will be made available upon reasonable request and with permission of the National Center for Chronic Non-communicable Disease Control and Prevention. Environmental data will be made available upon reasonable request.

Declaration of interests

All authors declare no potential conflicts of interest.

Appendix ASupplementary data

Supplementary Fig. S1 and Table S1

Acknowledgements

This study is supported by the 10.13039/501100012166 National Key Research and Development Program of China (2023YFC3708304 and 2022YFC3702701 ), the 10.13039/501100001809 National Natural Science Foundation of China (82304090 and 82030103 ), the 3-year Action Plan for Strengthening the Construction of the Public Health System in Shanghai (GWVI-11.2-YQ31 ), and the 10.13039/501100003399 Science and Technology Commission of Shanghai Municipality (21TQ015 ).

Appendix A Supplementary data related to this article can be found at https://doi.org/10.1016/j.ebiom.2024.105270.
==== Refs
References

1 GBD 2019 Diseases and Injuries Collaborators Global burden of 369 diseases and injuries in 204 countries and territories, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019 Lancet 396 10258 2020 1204 1222 33069326
2 Sin D.D. Doiron D. Agusti A. Air pollution and COPD: GOLD 2023 committee report Eur Respir J 61 5 2023 2202469
3 Berry C.E. Wise R.A. Mortality in COPD: causes, risk factors, and prevention COPD 7 5 2010 375 382 20854053
4 GBD 2019 Risk Factors Collaborators Global burden of 87 risk factors in 204 countries and territories, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019 Lancet 396 10258 2020 1223 1249 33069327
5 Safiri S. Carson-Chahhoud K. Noori M. Burden of chronic obstructive pulmonary disease and its attributable risk factors in 204 countries and territories, 1990-2019: results from the Global Burden of Disease Study 2019 BMJ 378 2022 e069679
6 Yu W. Ye T. Zhang Y. Global estimates of daily ambient fine particulate matter concentrations and unequal spatiotemporal distribution of population exposure: a machine learning modelling study Lancet Planet Health 7 3 2023 e209 e218 36889862
7 Li J. Sun S. Tang R. Major air pollutants and risk of COPD exacerbations: a systematic review and meta-analysis Int J Chron Obstruct Pulmon Dis 11 2016 3079 3091 28003742
8 Chen R. Yin P. Meng X. Fine particulate air pollution and daily mortality. A nationwide analysis in 272 Chinese cities Am J Respir Crit Care Med 196 1 2017 73 81 28248546
9 Zhu F. Chen L. Qian Z.M. Acute effects of particulate matter with different sizes on respiratory mortality in Shenzhen, China Environ Sci Pollut Res Int 28 28 2021 37195 37203 33715123
10 Valdes A. Zanobetti A. Halonen J.I. Cifuentes L. Morata D. Schwartz J. Elemental concentrations of ambient particles and cause specific mortality in Santiago, Chile: a time series study Environ Health 11 2012 82 23116481
11 Samoli E. Stafoggia M. Rodopoulou S. Which specific causes of death are associated with short term exposure to fine and coarse particles in Southern Europe? Results from the MED-PARTICLES project Environ Int 67 2014 54 61 24657768
12 Neuberger M. Moshammer H. Rabczenko D. Acute and subacute effects of urban air pollution on cardiopulmonary emergencies and mortality: time series studies in Austrian cities Int J Environ Res Public Health 10 10 2013 4728 4751 24157504
13 Doubleday A. Schulte J. Sheppard L. Mortality associated with wildfire smoke exposure in Washington state, 2006-2017: a case-crossover study Environ Health Glob 19 1 2020 4
14 Neuberger M. Rabczenko D. Moshammer H. Extended effects of air pollution on cardiopulmonary mortality in Vienna Atmos Environ 41 38 2007 8549 8556
15 Zeng X. Adair T. Wang L. Measuring the completeness of death registration in 2844 Chinese counties in 2018 BMC Med 18 1 2020 176 32615965
16 Liu S. Wu X. Lopez A.D. An integrated national mortality surveillance system for death registration and mortality surveillance, China Bull World Health Organ 94 1 2016 46 57 26769996
17 Yin P. Luo H. Gao Y. Criteria air pollutants and diabetes mortality classified by different subtypes and complications: a nationwide, case-crossover study J Hazard Mater 460 2023 132412
18 He Q.L. Liu Y.N. Yin P. Differentiating the impacts of ambient temperature on pneumonia mortality of various infectious causes: a case-crossover EBioMedicine 98 2023 104854
19 Shi S. Wang W. Li X. Optimizing modeling windows to better capture the long-term variation of PM(2.5) concentrations in China during 2005-2019 Sci Total Environ 854 2023 158624
20 Meng X. Liu C. Zhang L. Estimating PM(2.5) concentrations in Northeastern China with full spatiotemporal coverage, 2005-2016 Remote Sens Environ 253 2021 112203
21 Chen R. Yin P. Meng X. Associations between coarse particulate matter air pollution and cause-specific mortality: a nationwide analysis in 272 Chinese cities Environ Health Perspect 127 1 2019 17008
22 Meng X. Wang W. Shi S. Evaluating the spatiotemporal ozone characteristics with high-resolution predictions in mainland China, 2013-2019 Environ Pollut 299 2022 118865
23 Li X. Wang P. Wang W. Mortality burden due to ambient nitrogen dioxide pollution in China: application of high-resolution models Environ Int 176 2023 107967
24 Xin Y. Yang Y. Chen X. Yue X. Liu Y. Yin C. Evaluation of IMERG and ERA5 precipitation products over the Mongolian Plateau Sci Rep 12 1 2022 21776
25 Jiang Y. Yi S. Gao C. Cold spells and the onset of acute myocardial infarction: a nationwide case-crossover study in 323 Chinese cities Environ Health Perspect 131 8 2023 87016
26 Chen R. Pan G. Kan H. Ambient air pollution and daily mortality in Anshan, China: a time-stratified case-crossover analysis Sci Total Environ 408 24 2010 6086 6091 20889186
27 Yang K. Yang X. Jin C. Global burden of type 1 diabetes in adults aged 65 years and older, 1990-2019: population based study BMJ 385 2024 e078432
28 Niu Y. Niu H. Meng X. Associations between air pollution and the onset of acute exacerbations of chronic obstructive pulmonary disease: a time-stratified case-crossover study in China Chest 2024 10.1016/j.chest.2024.05.030
29 Altman D.G. Bland J.M. How to obtain the P value from a confidence interval BMJ 343 2011 d2304
30 Schwartz J. Dockery D.W. Neas L.M. Is daily mortality associated specifically with fine particles? J Air Waste Manag Assoc 46 10 1996 927 939
31 Atkinson R.W. Kang S. Anderson H.R. Mills I.C. Walton H.A. Epidemiological time series studies of PM2.5 and daily mortality and hospital admissions: a systematic review and meta-analysis Thorax 69 7 2014 660 665 24706041
32 Lepeule J. Laden F. Dockery D. Schwartz J. Chronic exposure to fine particles and mortality: an extended follow-up of the Harvard Six Cities study from 1974 to 2009 Environ Health Perspect 120 7 2012 965 970 22456598
33 Aron J. Baldomero A.K. Rau A. Fiecas M.B. Wendt C.H. Berman J.D. Individual risk factors of PM(2.5) associated with wintertime mortality in urban patients with COPD Chest 165 4 2023 825 835 37858719
34 Tian F. Qi J. Wang L. Differentiating the effects of ambient fine and coarse particles on mortality from cardiopulmonary diseases: a nationwide multicity study Environ Int 145 2020 106096
35 Kim K.H. Kabir E. Kabir S. A review on the human health impact of airborne particulate matter Environ Int 74 2015 136 143 25454230
36 Brown J.S. Gordon T. Price O. Asgharian B. Thoracic and respirable particle definitions for human health risk assessment Part Fibre Toxicol 10 2013 12 23575443
37 Schraufnagel D.E. Balmes J.R. Cowl C.T. Air pollution and noncommunicable diseases: a review by the forum of international respiratory societies’ environmental committee, Part 2: air pollution and organ systems Chest 155 2 2019 417 426 30419237
38 Ling S.H. van Eeden S.F. Particulate matter air pollution exposure: role in the development and exacerbation of chronic obstructive pulmonary disease Int J Chron Obstruct Pulmon Dis 4 2009 233 243 19554194
39 Newman J.D. Bhatt D.L. Rajagopalan S. Cardiopulmonary impact of particulate air pollution in high-risk populations: JACC state-of-the-art review J Am Coll Cardiol 76 24 2020 2878 2894 33303078
40 Sint T. Donohue J.F. Ghio A.J. Ambient air pollution particles and the acute exacerbation of chronic obstructive pulmonary disease Inhal Toxicol 20 1 2008 25 29 18236218
41 Li X.Y. Gilmour P.S. Donaldson K. MacNee W. Free radical activity and pro-inflammatory effects of particulate air pollution (PM10) in vivo and in vitro Thorax 51 12 1996 1216 1222 8994518
42 van Eeden S.F. Yeung A. Quinlam K. Hogg J.C. Systemic response to ambient particulate matter: relevance to chronic obstructive pulmonary disease Proc Am Thorac Soc 2 1 2005 61 67 16113470
43 Fujii T. Hayashi S. Hogg J.C. Interaction of alveolar macrophages and airway epithelial cells following exposure to particulate matter produces mediators that stimulate the bone marrow Am J Respir Cell Mol Biol 27 1 2002 34 41 12091243
44 Ghio A.J. Kim C. Devlin R.B. Concentrated ambient air particles induce mild pulmonary inflammation in healthy human volunteers Am J Respir Crit Care Med 162 3 Pt 1 2000 981 988 10988117
45 Salvi S.S. Nordenhall C. Blomberg A. Acute exposure to diesel exhaust increases IL-8 and GRO-alpha production in healthy human airways Am J Respir Crit Care Med 161 2 Pt 1 2000 550 557 10673199
46 Hennig F. Quass U. Hellack B. Ultrafine and fine particle number and surface area concentrations and daily cause-specific mortality in the ruhr area, Germany, 2009-2014 Environ Health Persp 126 2 2018 e027008
47 Jiang Y.X. Chen R.J. Peng W.H. Hourly ultrafine particle exposure and acute myocardial infarction onset: an individual-level case-crossover study in Shanghai, China, 2015-2020 Environ Sci Technol 57 4 2023 1701 1711 36668989
48 De Sario M. Katsouyanni K. Michelozzi P. Climate change, extreme weather events, air pollution and respiratory health in Europe Eur Respir J 42 3 2013 826 843 23314896
49 Wang Q. Li B. Benmarhnia T. Independent and combined effects of heatwaves and PM2.5 on preterm birth in guangzhou, China: a survival analysis Environ Health Perspect 128 1 2020 17006
50 Zhou W. Wang Q. Li R. Combined effects of heatwaves and air pollution, green space and blue space on the incidence of hypertension: a national cohort study Sci Total Environ 867 2023 161560
51 Yin P. Wu J. Wang L. The burden of COPD in China and its provinces: findings from the global burden of disease study 2019 Front Public Health 10 2022 859499
52 Sunyer J. Schwartz J. Tobias A. Macfarlane D. Garcia J. Anto J.M. Patients with chronic obstructive pulmonary disease are at increased risk of death associated with urban particle air pollution: a case-crossover analysis Am J Epidemiol 151 1 2000 50 56 10625173
53 Kan H. London S.J. Chen G. 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 Perspect 116 9 2008 1183 1188 18795161
54 Funk M.J. Landi S.N. Misclassification in administrative claims data: quantifying the impact on treatment effect estimates Curr Epidemiol Rep 1 4 2014 175 185 26085977
