BMJ...2020/08/22 · 25 Universidad Pablo de Olavide, Department of Physical, Chemical, and Natural Systems, Sevilla, Spain; 26 Institute of Epidemiology, Helmholtz Zentrum München

Confidential: For Review OnlyShort term associations of ambient nitrogen dioxide with daily total, cardiovascular, and respiratory mortality: a

global analysis in 398 cities

Journal: BMJ

Manuscript ID BMJ-2020-061413

Article Type: Research

BMJ Journal: BMJ

Date Submitted by the Author: 22-Aug-2020

Complete List of Authors: Meng, Xia; Fudan UniversityLiu, Cong; Fudan University School of Public HealthChen, Renjie; Fudan University School of Public HealthSera, Francesco ; London School of Hygiene and Tropical MedicineVicedo-Cabrera, Ana; London School of Hygiene and Tropical MedicineMilojevic, Ai; London School of Hygiene and Tropical MedicineGuo, Yuming; Monash University, School of Public Health and Preventive MedicineTong, Shilu; QUTCoelho , Micheline ; Universidade de Sao Paulo Faculdade de MedicinaSaldiva, Paulo Hilario; São Paulo University, PathologyLavigne, Eric; University of OttawaCorrea, Patricia; Universidad de los AndeOrtega, Nicolas; Universidad de los AndesGarcia, Samuel; Hospital Vista HermosaKyselý, Jan; Academy of Sciences of the Czech RepublicUrban, Ales; Academy of Sciences of the Czech RepublicOrru, Hans; Occupational and Environmental Medicine, Umeå University; Department of Public Health, University of TartuMaasikmets, Marek; Estonian Environmental Research CentreJaakkola, Jouni; University of Oulu, Institute of Health SciencesRyti, Niilo; University of Oulu, Center for Environmental and Respiratory Health Research (CERH)Huber, Veronika; Universidad Pablo de OlavideSchneider, Alexandra; Helmholtz Zentrum Muenchen, German Research Center for Environmental Health, Institute of Epidemiology IIKatsouyanni, Klea; University of Athens Medical School, Department of Hygiene, Epidemiology and Medical StatisticsAnalitis, Antonis; University of Athens Medical School, Department of Hygiene, Epidemiology and Medical StatisticsHashizume, Masahiro; Nagasaki University, Institute of Tropical MedicineHonda, Yasushi; University of TsukubaNg, Chris Fook Sheng ; Nagasaki University, School of Tropical Medicine and Global HealthNunes, Baltazar; Portuguese National Institute of HealthTeixeira, João; Portuguese National Institute of Health

https://mc.manuscriptcentral.com/bmj

BMJ

Confidential: For Review OnlyHolobaca, Iulian; Babes-Bolay UniversityFratianni, Simona; University of TorinoKim, Ho; Seoul National University, Gaduate School of Public HealthTobías, Aurelio; Spanish Council for Scientific Research, Institute of Environmental Assessment and Water Research (IDAEA)Iniguez, Carmen; University of ValenciaForsberg, Bertil; Occupational and Environmental Medicine, Umeå UniversityÅström, Christofer; Public health and clinical medicine, Umeå UniversityRagettli, Martina; University of BaselGuo, Leon; National Health Research Institutes, National Institute of Environmental Health Sciences; National Taiwan University, Environmental and Occupational MedicinePan, Shih-Chun; China Medical University, Public HealthLi, Shanshan; Monash UniversityBell, Michelle; Yale UniversityZanobetti, Antonella; Harvard UniversitySchwartz, Joel; Harvard School of Public Health, Epidemiology and Environmental HealthWu, Tangchun; Huazhong UNiv Sci and Tech, School of Public HealthGasparrini, Antonio; London School of Hygiene and Tropical MedicineKan, Haidong; Fudan University, Department of Environmental Health

Keywords: nitrogen dioxide, mortality, multi-location, time-series, meta-analysis

Page 1 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Only

1

Short term associations of ambient nitrogen dioxide with daily total,

cardiovascular, and respiratory mortality: a global analysis in 398 cities

Authors:

Xia Meng, PhD1,#, Cong Liu, PhD1,#, Renjie Chen, PhD1,2,#, Francesco Sera, PhD3, Ana

Maria Vicedo-Cabrera, PhD4,5, Ai Milojevic, PhD3, Yuming Guo, PhD6,7, Shilu Tong,

PhD8,9,10,11, Micheline de Sousa Zanotti Stagliorio Coelho, PhD12, Paulo Hilario

Nascimento Saldiva, PhD12, Eric Lavigne, PhD13,14, Patricia Matus Correa, PhD15,

Nicolas Valdes Ortega, PhD16, Samuel Osorio Garcia, PhD17, Jan Kyselý, PhD18,19,

Aleš Urban, PhD18,19, Hans Orru, PhD20,21, Marek Maasikmets, PhD22, Jouni Jaakkola,

PhD23, Niilo Ryti, PhD23, Veronika Huber, PhD24,25, Alexandra Schneider, PhD26, Klea

Katsouyanni, PhD27,28, Antonis Analitis, PhD27, Masahiro Hashizume, PhD29, Yasushi

Honda, PhD30, Chris Fook Sheng Ng, PhD31, Baltazar Nunes, PhD32, João Paulo

Teixeira, PhD32,33, Iulian Horia Holobaca, PhD34, Simona Fratianni, PhD35, Ho Kim,

PhD36, Aurelio Tobias, PhD37, Carmen Íñiguez, PhD38,39 Bertil Forsberg, PhD20,

Christofer Åström, PhD20, Martina S. Ragettli, PhD40,41, Yue-Liang Leon Guo, PhD42,

Shih-Chun Pan, PhD43, Shanshan Li, PhD6, Michelle L. Bell, PhD44, Antonella

Zanobetti, PhD45, Joel Schwartz, PhD45, Tangchun Wu, PhD46, Antonio Gasparrini,

PhD3,47,48,¶, Haidong Kan, PhD1,2,49,¶,*

# Xia Meng, Cong Liu and Renjie Chen contributed equally to this work.

¶ Antonio Gasparrini and Haidong Kan are both senior authors and contributed equally

to this work

Affiliations: 1 School of Public Health, Key Lab of Public Health Safety of the Ministry of Education

and Key Lab of Health Technology Assessment of the Ministry of Health, Fudan

University, Shanghai 200032, China;

Page 2 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


2

2 Shanghai Key Laboratory of Atmospheric Particle Pollution and Prevention (LAP3),

Fudan University, Shanghai 200030, China;

3 Department of Public Health, Environments and Society, London School of Hygiene

and Tropical Medicine, London, United Kingdom;

4 Institute of Social and Preventive Medicine, University of Bern, Bern, Switzerland;

5 Oeschger Center for Climate Change Research, University of Bern, Bern, Switzerland;

6 Department of Epidemiology and Preventive Medicine, School of Public Health and

Preventive Medicine, Monash University, Melbourne, Australia;

7 School of Public Health and Management, Binzhou Medical University, Yantai,

Shandong, China;

8 Shanghai Children’s Medical Center, Shanghai Jiao Tong University School of

Medicine, Shanghai, China;

9 School of Public Health, Institute of Environment and Population Health, Anhui

Medical University, Hefei, China;

10 School of Public Health and Social Work, Queensland University of Technology,

Brisbane, QLD, Australia;

11 Center for Global Health, School of Public Health, Nanjing Medical University,

Nanjing, China;

12 Institute of Advanced Studies, University of Sao Paulo, São Paulo, Brazil;

13 School of Epidemiology & Public Health, University of Ottawa, Ottawa, ON, Canada;

14 Air Health Science Division, Health Canada, Ottawa, ON, Canada;

15 Department of Public Health, Universidad de los Andes, Santiago, Chile;

16 School of Nursing and Obstetrics, Universidad de los Andes, Santiago, Chile;

17 Hospital Vista Hermosa, Bogotá, Colombia;

18 Institute of Atmospheric Physics, Czech Academy of Sciences, Bocni II 1401,141 00

Prague 4, Czech Republic;

19 Faculty of Environmental Sciences, Czech University of Life Sciences, Kamýcká

129, 165 21 Prague 6, Czech Republic;

20 Department of Public Health and Clinical Medicine, Umeå University, Umeå,

Page 3 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


3

Sweden;

21 Institute of Family Medicine and Public Health, University of Tartu, Tartu, Estonia.;

22 Estonian Environmental Research Centre, Tallinn, Estonia;

23 Center for Environmental and Respiratory Health Research (CERH), University of

Oulu; Medical Research Center Oulu (MRC Oulu), Oulu University Hospital and

University of Oulu;

24 Potsdam Institute for Climate Impact Research, Potsdam, Germany;

25 Universidad Pablo de Olavide, Department of Physical, Chemical, and Natural

Systems, Sevilla, Spain;

26 Institute of Epidemiology, Helmholtz Zentrum München – German Research Center

for Environmental Health (GmbH), Neuherberg, Germany;

27 Department of Hygiene, Epidemiology and Medical Statistics, School of Medicine,

National and Kapodistrian University of Athens, Athens, Greece;

28 School of Population Health & Environmental Sciences, King's College London;

29 Department of Global Health Policy, Graduate School of Medicine, The University

of Tokyo, Tokyo, Japan;

30 Faculty of Health and Sport Sciences, University of Tsukuba, Tsukuba, Japan;

31 School of Tropical Medicine and Global Health, Nagasaki University, Nagasaki,

Japan;

32 Department of Environmental Health, Portuguese National Institute of Health, Porto,

Portugal;

33 EPIUnit – Instituto de Saúde Pública, Universidade do Porto, Porto, Portugal;

34 Faculty of Geography, Babes-Bolay University;

35 Department of Earth Sciences, University of Torino;

36 Department of Public Health Science, Graduate School of Public Health & Institute

of Health and Environment, Seoul National University, Seoul, Republic of Korea;

37 Institute of Environmental Assessment and Water Research, Spanish Council for

Scientific Research, Barcelona, Spain;

38 Department of Statistics and Computational Research, Universitat de València,

Page 4 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


4

València, Spain;

39 CIBER of Epidemiology and Public Health (CIBERESP), Madrid, Spain；

40 Swiss Tropical and Public Health Institute, Basel, Switzerland;

41 University of Basel, Basel, Switzerland;

42 Environmental and Occupational Medicine, National Taiwan University (NTU)

College of Medicine and NTU Hospital, Taipei;

43 National Institute of Environmental Health Sciences, National Health Research

Institute, Miaoli, Taiwan;

44 School of the Environment, Yale University, New Haven, CT, USA;

45 Department of Environmental Health, Harvard T.H. Chan School of Public Health ,

Boston, Massachusetts, USA;

46 Key Laboratory of Environment and Health, Ministry of Education, and State Key

Laboratory of Environmental Health (Incubating), School of Public Health, Tongji

Medical College, Huazhong University of Science and Technology, Wuhan, China；47 Centre for Statistical Methodology, London School of Hygiene and Tropical

Medicine, London, UK;

48 Centre on Climate Change and Planetary Health, London School of Hygiene and

Tropical Medicine, London, UK;

49 Children's Hospital of Fudan University, National Center for Children's Health,

Shanghai 201102, China.

*Corresponding author: Dr. Haidong Kan, P.O. Box 249, 130 Dong-An Road,

Shanghai 200032, China. Tel/fax: +86 (21) 5423 7908. E-mail: [email protected].

Manuscript word count: 2798 words

Page 5 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


5

Summary boxes

What is already known on this topic?

Evidence on the short-term association between ambient nitrogen dioxide (NO2)

and mortality is limited, especially for cause-specific associations.

Previous investigations have been mostly conducted in a small number of locations

that cover limited geographical areas, and using different study designs and

modelling approaches that lead to heterogeneous results.

What this study adds?

This large multi-location study in 398 cities found that each 10 μg/m3 increase in

NO2 concentrations on lag 1 day (previous day) was significantly associated with

increased risk of total (0.46%), cardiovascular (0.37%), and respiratory (0.47%)

mortality.

The concentration-response relationships were almost linear without discernable

thresholds even below the current WHO air quality guidelines, suggesting a need

to revise the current recommendation values.

The unified analytical approach allows a direct comparison of estimates across

global regions

Page 6 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


6

Abstract

Objectives To evaluate the short-term associations between nitrogen dioxide (NO2) and

total, cardiovascular, and respiratory mortality across multiple countries/regions

worldwide, using a unified analytical protocol.

Design A two-stage time-series approach was applied, with over-dispersed generalized

linear models and multilevel meta-analysis.

Setting A multi-location study.

Participants The study included daily counts of deaths from total, cardiovascular, and

respiratory causes in 398 cities of 22 countries/regions.

Interventions Daily concentrations of NO2 were obtained from fixed-site monitoring

stations in urban areas, which were then averaged at the city level. Data on daily mean

temperature, relative humidity, and co-pollutants were also collected from local

authorities and/or monitoring stations.

Main outcome measures Daily counts of total, cardiovascular and respiratory

mortality in 398 cities covering the study period from 1973 to 2018.

Results On average, a 10 μg/m3 increase in NO2 concentration on lag 1 day (previous

day) was associated with 0.46% (95% confidence interval: 0.36% to 0.57%), 0.37% (95%

confidence interval: 0.22% to 0.51%), and 0.47% (95% confidence interval: 0.21% to

0.72%) increases in total, cardiovascular, and respiratory mortality, respectively. These

associations remained robust after adjusting for co-pollutants. The pooled

concentration–response curves for all three causes were almost linear without

discernible thresholds. The proportion of deaths attributable to NO2 concentration

above the counterfactual zero level was 1.25% (95% confidence interval 0.97% to

1.53%) across the 398 cities.

Conclusions This multi-location study provides key evidence on the independent and

linear associations between short-term exposure to NO2 and increased risk of total,

cardiovascular and respiratory mortality, suggesting that health benefits would be

achieved by tightening the guidelines and regulatory limits of NO2.

Trial registration None.

Page 7 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


7

Keywords: NO2, mortality, multi-location, time-series, meta-analysis

Page 8 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


8

Introduction

Nitrogen dioxide (NO2) is a common air pollutant that is a precursor to ozone, is

involved in the secondary formation of fine particulate matter, and has its own adverse

health impacts. Most ambient NO2 comes from anthropogenic origin such as fuel

combustion and traffic 1. There has not been a universal agreement on the adverse health

effects of short-term exposure to NO2 on human mortality or morbidity 2 3. The

Integrated Science Assessment (ISA) conducted by the US Environmental Protection

Agency (EPA) reviewed the existing evidence linking NO2 exposure with health

outcomes in 2008 4 and 2016 5, and concluded that there is a causal relationship between

short-term NO2 exposure and respiratory effects. However, the review highlighted

important gaps in knowledge that contributed to different limits for NO2 in air quality

guidelines (AQG) being adopted by various governmental and intergovernmental

institutions 1 6-8.

Several epidemiological studies have investigated the relationships between NO2

and total mortality 9, however with important limitations. First, most of them focused

on total non-accidental deaths, with only a few assessing cause-specific mortality, such

as respiratory and cardiovascular outcomes. Moreover, potential features of NO2-

mortality associations, such as non-linearity, thresholds, and lag structures remain to be

clarified. More importantly, questions remain regarding whether NO2 has an

independent effect on mortality, or if observed associations can be explained by

confounding effects of co-pollutants. Finally, the current literature mainly includes

studies performed in single cities or countries, presenting issues with publication bias

and difficulties in integrating results from different analytical approaches.

This study aimed to address these limitations by investigating the short-term

association between NO2 and mortality across multiple countries/regions worldwide,

using data from the Multi-City Multi-Country (MCC) Collaborative Research Network

(http://mccstudy.lshtm.ac.uk/). The assessment takes advantage of a common analytical

framework based on state-of-the-art epidemiological methodologies to estimate flexible

exposure-response relationships, to integrate results across cities and countries, and to

Page 9 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

http://mccstudy.lshtm.ac.uk/


9

quantify the excess mortality attributable to short-term NO2 exposure.

Methods

Data Collection

Health and environmental data were obtained from the MCC database, which has been

described in previous publications 10 11. The current analysis was limited to locations

where NO2 ground measurements were available, which included a total of 398 cities

in 22 countries and/or regions, with different study periods based on data availability

(Table 1). The geographic distribution of cities with NO2, and the corresponding

annual-mean concentrations during the study periods were visualized in eFigure 1.

Mortality data were obtained from local authorities within each country/ region. Causes

of death were classified according to the 9th or 10th version of International

Classification of Diseases (ICD) codes, where available. In each location, mortality is

represented by daily counts of either non-external causes (ICD-9: 0-799; ICD-10: A0-

R99) or, where not available, all-cause mortality. We also collected mortality data for

two main causes in 16 countries (Table 1): cardiovascular disease (ICD-10, codes I00-

I99) and respiratory disease (ICD-10, codes J00-J99) 12. This analysis used health data

at the aggregated level, no patient or individual were contacted, and thus there is no

patient and public involvement.

Daily concentrations of particulate matter and other gaseous pollutants, including

particulate matter with aerodynamic diameter ≤10 m or ≤ 2.5 m (PM10 and PM2.5,

respectively), ozone (O3), sulfur dioxide (SO2) and carbon monoxide (CO), were

obtained from fixed-site monitoring stations where available, to adjust for potential

confounding by co-pollutants. We also collected daily data on weather variables

including daily mean temperature and daily mean relative humidity.

Statistical Analysis

The associations of NO2 with daily total, cardiovascular, and respiratory mortality were

assessed using the same analytical protocol in all locations, based on an extension of a

Page 10 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


10

two-stage design widely used in previous multi-location time-series studies 13 14.

In the first stage, we estimated the city-specific associations using a time-series

quasi-Poisson generalized linear regression model featuring a natural cubic spline

function of time with 7 degrees of freedom (df) per year to control for unmeasured

temporal trends and indicators for day of the week. Temperature was adjusted using a

natural spline function with 6 df. Relative humidity with the same spline function with

3 df in cities where such data were available. We a priori selected the moving average

of lag 0-3 days to control temperature according to modelling choices of many previous

studies 14 15. We modelled the association between NO2 and mortality using distributed

lag model (DLM) with a linear lag-response function, the lag structure was inspected

on single lag day of 0 to 3 and moving average of the present and previous day (lag 0–

1) to identify the optimal lag choices.

In the second stage, we used a novel multilevel meta-analytical approach to

summarize the city-specific associations 16. Briefly, this model defines more complex

random-effects structures that properly accounts for the hierarchical structure of the

data, namely cities nested within countries and provides the best linear unbiased

predictors (BLUPs) for the NO2-mortality associations at both levels 16. We computed

the global, country, and city-specific estimates and 95% confidence intervals (CIs) as

percent change in daily mortality per 10 μg/m3 increase of NO2 concentrations.

Potential heterogeneity across cities was assessed via Cochran Q tests and I-squared

(I2) statistics. Finally, we computed the proportion of deaths attributable to NO2

exposure above the counterfactual zero level and presented the population attributable

fraction (PAF %) at the country level.

We extended the main models to assess specific features of the association. First,

we fitted two-pollutant models using the other co-pollutants, assessing the robustness

of the estimates through likelihood ratio tests. Furthermore, we conducted regional

analyses (eTable 2 and eTable 4), and explored potential effect modifiers (such as long-

term air pollution concentrations, temperature and GDP) on the associations between

NO2 and total mortality based on the main models. Finally, we explored potential non-

Page 11 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


11

linearity in the associations and pooled the concentration-response (C–R) relationship

curve using a meta-smoothing approach applied in previous studies 14 17, modelling NO2

as a nonlinear term through a natural spline function with knots at 20 and 40 μg/m3.

We conducted several sensitivity analyses to test the robustness of our estimates,

including alternative choices for controlling for temperature, alternative knots

placements for testing non-linear exposure-response between NO2 and mortality, a

comparison of models with and without adjustment for humidity, and seasonal

differences modelled through an interaction with an indicator of warm/cold season

(May-September versus October-April for Northern Hemisphere, and vice versa for

Southern Hemisphere).

We conducted all statistical analyses in R software (Version 3.3.1), using the stats

and dlnm packages for fitting first-stage models and the mixmeta package for

performing multilevel meta-analyses. We presented the percent change of mortality for

a 10 μg/m3 increase in NO2 concentrations. P-values < 0.05 were considered statistically

significant in all statistical analyses.

Results

Descriptive Statistics

This analysis included 60.1 million deaths from total or non-external causes, and there

were 19.7 and 5.5 million deaths from cardiovascular diseases and respiratory diseases,

respectively (Table 1). On average, the median annual mean NO2 concentration across

398 cities was 26.9 μg/m3 (25%-75% percentiles: 19.5 to 36.2 μg/m3), which was lower

than the Air Quality Guidelines for NO2 (annual-mean: 40 μg/m3) of the World Health

Organization (WHO) 6. The median of annual-mean temperature was 14.4°C (25%-75%

percentiles: 7.8°C to 20.6°C). The detailed summary of exposure data is provided in

eTable 1. The Spearman correlation coefficients of NO2 were 0.34 with PM10, 0.38 with

PM2.5, -0.24 with O3, 0.42 with SO2, and 0.56 with CO. On average, NO2 was negatively

correlated with mean temperature (Pearson r = -0.29) and relative humidity (Pearson r

= -0.15).

Page 12 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


12

Regression Results

Figure 1 illustrates the estimated pooled associations between NO2 and total,

cardiovascular and respiratory mortality on different lag days. Results suggested that

the magnitude of associations was increasing from the current (lag 0) to the previous

day (lag 1) and then decreased on lag 2 day for all three endpoints. In addition,

considering its causal respiratory effect 5, NO2 concentration on lag 1 day was used as

the main lag for NO2 in subsequent analysis.

Table 2 displays the country-specific and pooled estimates on the associations of

NO2 (lag 1) with total, cardiovascular and respiratory mortality. Across 398 cities, a 10

μg/m3 increase in NO2 concentrations was associated with an increase of 0.46% (0.36%

to 0.57%) in total mortality; while across 362 cities, the corresponding increases were

0.37% (0.22% to 0.51%) and 0.47% (0.21% to 0.72%) for cardiovascular and

respiratory mortality. The country-specific estimates showed considerable variations.

There were consistently positive associations of NO2 with total mortality, ranging from

0.17% to 0.69% for total mortality. Estimates were less precise for cause-specific

associations, although generally indicating an increased risk.

Figure 2 shows the results of the model with non-linear terms for estimating the

effect of the NO2–mortality association, indicating positive and almost linear

concentration-response (C–R) curves for total, cardiovascular, and respiratory mortality,

with no discernible thresholds. Thus, a counterfactual scenario at 0 µg/m3 was defined

to assess the mortality impacts in each country based on the current study sample (Table

3). Although we recognize that this level is not practical from a policy standpoint, our

analysis gives insight into the overall health burden from NO2. The PAF% ranged from

0.28% (0.05% to 0.51%) in Finland and 3.15% (1.83% to 4.48%) in Greece, with a

pooled PAF% of 1.25% (0.97% to 1.53%) across the 398 cities.

In two-pollutant models (eTable 3), the associations of NO2 with total mortality

were robust to the adjustment of co-pollutants. The estimates of NO2-mortality

associations increased slightly with adjustment of SO2, O3, and CO. Although the effect

Page 13 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


13

estimates decreased by 19% (P-value = 0.082) and 21% (P-value = 0.076) after

adjusting for PM10 and PM2.5, respectively, the associations were still positive and

statistically significant.

In the sensitivity analyses, compared with main models, the estimates for the NO2-

mortality associations were generally similar when adjusting for temperature with

different lag structures, except for a smaller association when using the lag of the

present and mean of the previous 3 days (eFigure 2). The use of alternative knots did

not substantially change the shape of the C–R curves (eFigure 3). The associations of

NO2 with total and cause-specific mortality did not change with or without adjustment

of relative humidity (eTable 5). Finally, we found no evidence of seasonal difference

in the NO2-mortality association (P-value = 0.683), with the estimate of cold season

(0.51% [0.41% to 0.60%]) similar to that of the warm season (0.43% [0.31% to 0.50%]).

Discussion

To the best of our knowledge, this work represents the largest epidemiological study on

the short-term association of NO2 with daily mortality. A key advantage of this study

is the analysis of a large dataset using the same analytical protocol, which allows a valid

comparison among countries and regions of the world. We found robust associations of

NO2 with daily mortality from total, cardiovascular and respiratory causes that were

independent from concomitant exposures to other air pollutants. More importantly, for

the first time, we pooled a C–R curve for NO2 at the global level, suggesting an almost

linear association with no discernible thresholds. This indicates that NO2 is associated

with significant health risks even at levels below health-based standards and guidelines,

including the current WHO AQGs.

At the global level, the risk of total mortality increased by 0.57% for every 10

μg/m3 increase in NO2 concentrations. The magnitude of the association is comparable

with the result from one previous systematic review, which included studies in 26 cities

worldwide and reported an estimate of 0.78% for all-cause mortality with the same NO2

increment as ours 18. The other multi-location study conducted in Europe (known as Air

Page 14 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


14

Pollution on Health: a European Approach [APHEA]-2 project) found a smaller effect

estimate with an increase of 0.30% (95%CI: 0.22%, 0.38%) in all-cause mortality

associated with a 10 μg/m3 increase in NO2 19. Our estimate is somewhat smaller than

some multi-location studies conducted in a single country. For a 10 μg/m3 increase in

NO2, a multi-location study in China observed a 0.91% (95%CI: 0.70%, 1.12%)

increase in total mortality 20; another study in Italy reported an estimate of 2.09%

(95%CI: 0.96%, 3.24%) 21. The differences in estimates do not necessarily reflect the

diverse health impact of NO2. Rather, the coverage of regions and time-periods,

population characteristics, and exposure patterns may contribute to these various

findings. Furthermore, the as found in our sensitivity analysis, different modelling

specification for temperature would introduce fluctuation in the estimated associations.

The 2016 Integrated Science Assessment conducted by the US EPA determined a

causal relationship between short-term NO2 exposure and respiratory effects; while for

total mortality and cardiovascular effects, the conclusion was “suggestive of, but not

sufficient to infer, a causal relationship”. Our findings from this global analysis add to

the supporting evidence for causal relationships between short-term NO2 exposure and

non-respiratory endpoints. Of course, experimental investigations and controlled

human exposure studies are needed to understand the non-respiratory effect of NO2.

There has been a long debate on whether NO2 exposure independently causes

health effects, or serves as a marker for a broader mixture of air pollutants, especially

those related to traffic 5. Findings from controlled human exposure and animal

toxicological studies are scarce, but are key for indicating independent relationships 22

23. Though statistically adjusting the NO2 association for another co-pollutant cannot

conclusively show an independent effect, gathering evidence may contribute toward

understanding this issue. In the current analysis, the NO2-mortality association

decreased after adjustment of PM10 and PM2.5, but remained positive and statistically

significant. A systematic review that included studies in 26 cities worldwide found the

similar conclusion that the effect estimates of NO2 on total mortality decreased from

0.78% to 0.60%, but were still statistically significant after adjusting for PM 18.

Page 15 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


15

Meanwhile, as found in our previous study, the PM-mortality association was also

influenced by NO2 15. This might indicate potential mutual confounding in the PM- and

NO2-mortality associations. On the other hand, the reason of null estimates in previous

reports of NO2 after adjustment of co-pollutants may be due to the smaller study sample,

regional difference and other uncertainties driven by study heterogeneity. The present

multi-location study and meta-analysis provided robust evidence on the independent

association between NO2 and mortality.

Characterizing the shape of the C–R relationship helps quantify the public health

impact of NO2 exposure and establish safe concentration limits. However, the shape of

the C-R relationship for health effects related to short-term NO2 exposure has only been

examined in a small number of epidemiologic studies with a limited spatial scale. This

study, for the first time, provide pooled estimates of the C–R relationship curves for

short-term NO2 exposure and total, cardiovascular and respiratory mortality across

various cities and countries. Results show that the curves are almost linear without

obvious thresholds, suggesting that the level of NO2 below the current air quality

guidelines is still hazardous to public health. We estimated that reduction in daily NO2

concentration to a counterfactual zero level would reduce 1.25% attributable deaths

across the 398 cities. Although full reductions of NO2 to zero is infeasible, our analysis

provides insight into the public health benefits of substantial NO2 reductions. This

suggests considerable health benefits from implementation of stricter control of NO2

emissions and tightening of the regulatory limits of NO2 in future revisions of WHO

AQGs.

Strengths and limitations of this study

This global study has several advantages. First, this multi-location study included 398

cities of 22 countries, providing enormous statistical power and ensuring the

generalizability of the findings. Second, we examined the associations of NO2 with not

only total, but also cause-specific mortality. Third, for the first time, we pooled C–R

curves covering wide range of ambient NO2 concentrations, the evidence of which can

Page 16 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


16

be considered generalizable. Fourth, the unified analytical approaches can aid in

integrating and comparing the results across different regions and populations. Finally,

our results contribute to the increasing evidence that supports independent health effects

of NO2 on total, cardiovascular, and respiratory mortality.

We acknowledge some limitations of the present study. First, this analysis was a

time-series design and used environmental measurements from fixed-site monitors,

which might have introduced exposure misclassification. However, these errors are

likely to be non-differential in most cases, leading to underestimation of the NO2-

mortality association 24 25. Second, the health data used in this global study covering

multiple decades is potentially subject to diagnostic or coding errors. The impacts of

these errors on our results are difficult to evaluate, presumably adding more uncertainty

on cause-specific mortality than on all-cause mortality.

Conclusions

In summary, this multi-location time-series analysis provides robust evidence on

the independent associations of short-term exposure to NO2 with increased risk of total,

cardiovascular, and respiratory mortality at the global level, and the associations

remained positive and statistically significant after adjusting for co-pollutants. The C–

R curves were linear without discernable thresholds, suggesting a need to revise and

tighten the current AQGs of NO2 for greater public health benefit, and to consider a

regulation limit for daily mean NO2 concentration.

Page 17 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


17

Contributors:

AG and HK are both senior authors and contributed equally to this work. MX, CL, AG

and HK designed the study. MX, CL, and RC are joint first authors. MX, CL and RC

coordinated the work, conducted the statistical analysis, and took the lead in drafting

the manuscript and interpreting the results. FS, AMV-C, ST, EL, IH, MH, VH, AT,

AU, YLG, CI, MB, and TW provided substantial scientific input in interpreting the

results and drafting the manuscript. AM, YMG, MS, PS, PC, NO, SG, JK, HO, MM,

JJ, NR, VH, AS, KK, AA, YH, CN, BN, JT, SF, HK, BF, CA, MR, SP, SL, AZ, and JS

provided the data and contributed to the interpretation of the results and to the submitted

version of the manuscript. HK and AG are the guarantors. The corresponding author

attests that all listed authors meet authorship criteria and that no others meeting the

criteria have been omitted.

Funding

HK was supported by the National Natural Science Foundation of China (91843302

and 91643205), and China Medical Board Collaborating Program (16-250). AG and FS

were supported by the Medical Research Council-UK (MR/M022625/1), the Natural

Environment Research Council UK (NE/R009384/1), and the European Union’s

Horizon 2020 Project Exhaustion (820655). VH was supported by the Spanish Ministry

of Science and Innovation (PCIN-2017-046), and the German Federal Ministry of

Education and Research (01LS1201A2). YH and MH were supported by the

Environment Research and Technology Development Fund (JPMEERF15S11412) of

the Environmental Restoration and Conservation Agency, Japan. JK and AU were

supported by the Czech Science Foundation (18-22125S).

Competing interests: The authors declared no conflicts of interests

Ethical approval: Not required.

Data sharing: Data have been collected within the MCC (Multi-City Multi-Country)

Collaborative Research Network (http://mccstudy.lshtm.ac.uk/) under a data sharing

Page 18 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


18

agreement and cannot be made publicly available. Researchers can refer to MCC

participants listed as coauthors for information on accessing the data for each country.

The lead authors (HK and AG) affirm that the manuscript is an honest, accurate, and

transparent account of the study being reported; that no important aspects of the study

have been omitted; and that any discrepancies from the study as planned (and, if

relevant, registered) have been explained.

Page 19 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


19

References1. EPA U. National Ambient Air Quality Standards for six criteria pollutants. 20162. Chen R, Yin P, Meng X, et al. Associations Between Ambient Nitrogen Dioxide and

Daily Cause-specific Mortality: Evidence from 272 Chinese Cities. Epidemiology 2018;29(4):482-89. doi: 10.1097/EDE.0000000000000829

3. Faustini A, Rapp R, Forastiere F. Nitrogen dioxide and mortality: review and meta-analysis of long-term studies. Eur Respir J 2014;44(3):744-53. doi: 10.1183/09031936.00114713

4. US-EPA. Integrated science assessment for oxides of nitrogen – health criteria (final report). 2008

5. EPA U. Integrated Science Assessment (ISA) For Oxides Of Nitrogen – Health Criteria (Final Report, 2016)2016.

6. WHO. Air quality guidelines for particulate matter, ozone, nitrogen dioxide and sulfur dioxide: global update 2005: World Health Organization, 2006.

7. Krzyzanowski M. WHO air quality guidelines for Europe. Journal of Toxicology and Environmental Health, Part A 2008;71(1):47-50.

8. Ministry of Ecology and Environment C. Ambient Air Quality Standards2012.9. Mills IC, Atkinson RW, Kang S, et al. Quantitative systematic review of the

associations between short-term exposure to nitrogen dioxide and mortality and hospital admissions. BMJ Open 2015;5(5):e006946. doi: 10.1136/bmjopen-2014-006946

10. Gasparrini A, Guo Y, Hashizume M, et al. Temporal Variation in Heat-Mortality Associations: A Multicountry Study. Environmental health perspectives 2015;123(11):1200-7. doi: 10.1289/ehp.1409070 [published Online First: 2015/05/02]

11. Gasparrini A, Guo Y, Hashizume M, et al. Mortality risk attributable to high and low ambient temperature: a multicountry observational study. Lancet (London, England) 2015;386(9991):369-75. doi: 10.1016/s0140-6736(14)62114-0 [published Online First: 2015/05/25]

12. WHO. International statistical classification of diseases and related health problems: World Health Organization 2004.

13. Bell ML, Dominici F, Samet JM. A meta-analysis of time-series studies of ozone and mortality with comparison to the national morbidity, mortality, and air pollution study. Epidemiology 2005;16(4):436-45.

14. Chen R, Yin P, Meng X, et al. Fine Particulate Air Pollution and Daily Mortality. A Nationwide Analysis in 272 Chinese Cities. Am J Respir Crit Care Med 2017;196(1):73-81. doi: 10.1164/rccm.201609-1862OC

15. Liu C, Chen R, Sera F, et al. Ambient Particulate Air Pollution and Daily Mortality in 652 Cities. N Engl J Med 2019;381(8):705-15. doi: 10.1056/NEJMoa1817364

16. Sera F, Armstrong B, Blangiardo M, et al. An extended mixed-effects framework for meta-analysis. Stat Med 2019;38(29):5429-44. doi: 10.1002/sim.8362 [published Online First: 2019/10/28]

Page 20 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


20

17. Samoli E, Analitis A, Touloumi G, et al. Estimating the exposure-response relationships between particulate matter and mortality within the APHEA multicity project. Environmental health perspectives 2005;113(1):88-95.

18. Mills IC, Atkinson RW, Anderson HR, et al. Distinguishing the associations between daily mortality and hospital admissions and nitrogen dioxide from those of particulate matter: a systematic review and meta-analysis. BMJ Open 2016;6(7):e010751. doi: 10.1136/bmjopen-2015-010751

19. Samoli E, Aga E, Touloumi G, et al. Short-term effects of nitrogen dioxide on mortality: an analysis within the APHEA project. Eur Respir J 2006;27(6):1129-38. doi: 10.1183/09031936.06.00143905

20. Chen R, P Y, X M, et al. Associations Between Ambient Nitrogen Dioxide and Daily Cause-specific Mortality: Evidence from 272 Chinese Cities. Epidemiology 2018;29(4):482.

21. Chiusolo M, Cadum E, Stafoggia M, et al. Short-Term Effects of Nitrogen Dioxide on Mortality and Susceptibility Factors in 10 Italian Cities: The EpiAir Study. Environmental health perspectives 2011;119(9):1233-8. doi: 10.1289/ehp.1002904

22. Riedl MA, Diaz-Sanchez D, Linn WS, et al. Allergic inflammation in the human lower respiratory tract affected by exposure to diesel exhaust. Res Rep Health Eff Inst 2012(165):5-43; discussion 45-64.

23. Thompson AM, Zanobetti A, Silverman F, et al. Baseline repeated measures from controlled human exposure studies: associations between ambient air pollution exposure and the systemic inflammatory biomarkers IL-6 and fibrinogen. Environmental health perspectives 2010;118(1):120-4. doi: 10.1289/ehp.0900550

24. Guo Y, Barnett AG, Tong S. Spatiotemporal model or time series model for assessing city-wide temperature effects on mortality? Environ Res 2013;120:55-62. doi: 10.1016/j.envres.2012.09.001

25. Zeger SL, Thomas D, Dominici F, et al. Exposure measurement error in time-series studies of air pollution: concepts and consequences. Environmental health perspectives 2000;108(5):419-26.

Page 21 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


21

Table 1. Mortality and environmental data in 398 cities of 22 countries/regions.

Number of deaths (in thousands)* Median (interquartile range)Country

/Region

No. of

CitiesPeriod

Total Cardiovascular Respiratory NO2 (μg/m3) Temperature (℃)

Australia 3 1988-2009 1178.0 NA NA 21.4 [14.1-27.9] 18.1 [14.7-21.2]

Brazil 1 1997-2011 916.2 NA NA 84.9 [62.6-115.3] 20.6 [18-22.9]

Canada 25 1986-2015 3477.6 1220.7 307.8 23.7 [15.6-33.8] 7.4 [-0.9-15.7]

Chile 3 2005-2013 316.8 NA NA 21.6 [13.9-32.4] 13.7 [10.7-17.2]

China 15 1996-2015 1201.7 468.0 164.7 46.5 [36.2-60.4] 16.3 [6.5-23.5]

Colombia 1 1998-2013 369.1 123.8 46.3 30.5 [23.3-37.9] 13.9 [13.2-14.5]

Czech Republic 1 1994-2015 270.5 145.0 13.9 30.8 [24.2-38.7] 9.2 [2.7-15.3]

Estonia 4 1997-2015 133.8 NA NA 11.4 [7.6-16.7] 6 [-0.1-13.6]

Finland 1 1994-2014 139.5 57.4 9.7 6.8 [4.3-11.7] 5.9 [0.0-13.8]

Germany 12 1993-2015 3105.9 NA NA 29.6 [21.8-38.4] 10.5 [4.8-15.9]

Greece 1 2001-2010 288.0 136.2 28.8 50.2 [39.6-61.6] 17.9 [12.9-24.9]

Japan 47 2011-2015 1885.0 496.7 296.0 16.7 [12.1-23.4] 16.1 [7.6-22.7]

Portugal 5 1990-2018 1750.7 659.8 165.9 14.9 [10.0-21.7] 15.4 [11.5-19.9]

Romania 8 1994-2016 951.1 NA NA 25.6 [17.7-36.2] 11.4 [3.4-18.9]

Page 22 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


22

South Korea 7 1992-2015 502.9 547.5 133.3 43.7 [33.0-57.5] 14.9 [5.7-21.9]

Spain 48 1990-2014 2929.6 1011.8 330.8 26.4 [20.2-33.9] 14.9 [10.2-20.7]

Sweden 1 1990-2010 201.2 91.3 15.9 26.8 [20-34.8] 6.8 [1.2-13.9]

Switzerland 8 1995-2013 235.9 90.7 16.0 32.3 [24.0-42.0] 10.7 [4.4-16.5]

Taiwan 3 1994-2014 688.4 269.4 116.5 42.2 [31.6-54.7] 24.9 [20.4-28.0]

Thailand 18 1999-2008 843.4 167.8 110.2 22.0 [16.3-30.4] 28 [26.4-29]

United Kingdom 39 1990-2016 5237.2 1978.9 799.2 25.6 [17.9-35.6] 10.4 [6.6-14.7]

United States 147 1973-2006 33502.8 12303.5 2904.9 28.7 [20.8-38.6] 14.7 [7.2-21.8]

Pooled 398 1973-2018 60125.3 19768.5 5460.1 26.9 [19.5-36.2] 14.4 [7.8-20.6]

Abbreviations: NO2, nitrogen dioxide

Notes: *Mortality data from cardiovascular and respiratory diseases were not available in Australia, Brazil, Chile, Estonia, Germany, and

Romania.

Page 23 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


23

Table 2. Percent changes (mean and 95% confidence intervals) in total, cardiovascular and respiratory mortality associated with a 10

μg/m3 increase in nitrogen dioxide on lag 1 day.

Total Mortality Cardiovascular Mortality Respiratory MortalityCountry/Region

N Estimates N Estimates N Estimates

Australia 3 0.64 (0.36 to 0.93) 0 NA 0 NA

Brazil 1 0.34 (0.22 to 0.47) 0 NA 0 NA

Canada 25 0.62 (0.47 to 0.76) 24 0.45 (0.23 to 0.67) 24 0.52 (0.10 to 0.94)

Chile 3 0.53 (0.39 to 0.67) 0 NA 0 NA

China 15 0.57 (0.31 to 0.83) 13 0.54 (0.34 to 0.73) 13 0.76 (0.41 to 1.12)

Colombia 1 0.31 (0.01 to 0.61) 1 0.28 (-0.09 to 0.66) 1 -0.05 (-0.75 to 0.67)

Czech Republic 1 0.46 (0.16 to 0.76) 1 0.37 (0.01 to 0.73) 1 0.42 (-0.34 to 1.18)

Estonia 4 0.40 (0.06 to 0.74) 0 NA 0 NA

Finland 1 0.41 (0.08 to 0.75) 1 0.39 (-0.01 to 0.79) 1 0.50 (-0.27 to 1.28)

Germany 12 0.62 (0.47 to 0.78) 0 NA 0 NA

Greece 1 0.62 (0.36 to 0.88) 1 0.47 (0.15 to 0.80) 1 0.84 (0.20 to 1.49)

Japan 47 0.36 (0.19 to 0.54) 47 0.18 (-0.10 to 0.45) 47 0.70 (0.27 to 1.12)

Portugal 5 0.52 (0.37 to 0.66) 5 0.22 (-0.07 to 0.51) 5 0.13 (-0.40 to 0.66)

Page 24 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


24

Romania 8 0.27 (0.07 to 0.48) 0 NA 0 NA

South Korea 7 0.41 (0.12 to 0.70) 7 0.48 (0.25 to 0.71) 7 0.48 (0.05 to 0.91)

Spain 48 0.70 (0.51 to 0.88) 47 0.53 (0.25 to 0.81) 47 0.88 (0.40 to 1.35)

Sweden 1 0.55 (0.27 to 0.83) 1 0.21 (-0.16 to 0.59) 1 0.41 (-0.34 to 1.15)

Switzerland 8 0.24 (-0.07 to 0.55) 8 0.52 (0.16 to 0.88) 8 0.24 (-0.47 to 0.96)

Taiwan 3 0.52 (0.28 to 0.76) 3 0.37 (0.07 to 0.67) 3 0.22 (-0.29 to 0.74)

Thailand 18 0.32 (0.13 to 0.51) 18 0.38 (0.03 to 0.74) 18 0.69 (0.10 to 1.28)

United Kingdom 39 0.17 (0.04 to 0.30) 39 0.10 (-0.09 to 0.28) 39 0.11 (-0.21 to 0.43)

United States 147 0.57 (0.46 to 0.68) 146 0.45 (0.30 to 0.61) 146 0.62 (0.34 to 0.90)

Pooled 398 0.46 (0.36 to 0.57) 362 0.37 (0.22 to 0.51) 362 0.47 (0.21 to 0.72)

Note: N, number of cities with available data within the country/region.

Page 25 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


25

Table 3. Proportion of deaths attributable to reductions in daily NO2 concentrations in each country.

Country/Region Cities

Annual deaths

(in thousands) a

Annual NO2

concentrations

(μg/m3)

Reductions in NO2

to theoretically zero b

PAF (%)

Australia 3 53.5 21.4 21.4 1.38 (0.76 to 2.00)

Brazil 1 61.1 84.9 84.9 2.95 (1.85 to 4.07)

Canada 25 115.9 23.7 23.7 1.47 (1.12 to 1.82)

Chile 3 35.2 21.6 21.6 1.14 (0.84 to 1.45)

China 15 200.3 46.5 46.5 2.68 (1.44 to 3.94)

Colombia 1 23.1 30.5 30.5 0.95 (0.04 to 1.88)

Czech Republic 1 12.3 30.8 30.8 1.41 (0.48 to 2.35)

Estonia 4 7.0 11.4 11.4 0.45 (0.07 to 0.84)

Finland 1 6.6 6.8 6.8 0.28 (0.05 to 0.51)

Germany 12 135.0 29.6 29.6 1.86 (1.39 to 2.32)

Greece 1 28.8 50.2 50.2 3.15 (1.83 to 4.48)

Japan 47 377.0 16.7 16.7 0.60 (0.31 to 0.90)

Portugal 5 44.9 14.9 14.9 0.77 (0.56 to 0.98)

Page 26 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


26

Romania 8 41.4 25.6 25.6 0.70 (0.18 to 1.23)

South Korea 7 21.0 43.7 43.7 1.79 (0.51 to 3.09)

Spain 48 117.2 26.4 26.4 1.85 (1.36 to 2.34)

Sweden 1 9.6 26.8 26.8 1.48 (0.72 to 2.24)

Switzerland 8 12.4 32.3 32.3 0.78 (-0.21 to 1.78)

Taiwan 3 32.8 42.2 42.2 2.21 (1.19 to 3.25)

Thailand 18 84.3 22 22 0.70 (0.30 to 1.11)

United Kingdom 39 194.0 25.6 25.6 0.45 (0.11 to 0.78)

United States 147 985.4 28.7 28.7 1.65 (1.33 to 1.98)

Pooled 398 2598.7 26.9 26.9 1.25 (0.97 to 1.53)

Abbreviations as in Table 1. PAF%, population attributable fraction.

Note: a Annual death numbers summed from the country-specific time-series data.

b Reductions in NO2 concentrations were calculated as the difference between country-specific annual NO2 concentrations and the theoretically

zero concentration.

Page 27 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


27

Figure 1. Relative risks of total, cardiovascular, and respiratory mortality associated with a 10 μg/m3 increase in nitrogen dioxide (NO2)

on different lag days for NO2.

Page 28 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


28

Figure 2. Concentration–response curve between NO2 concentrations (Lag 1) and total (A), cardiovascular (B), and respiratory (C)

mortality. The vertical scale can be interpreted as the relative change of the mean effect of NO2 on mortality; the fraction of the curve below zero

denotes a smaller estimate compared with the mean effect. Abbreviations: NO2, nitrogen dioxide.

Page 29 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


1

Online-only Supplements

Title: Short term associations of ambient nitrogen dioxide with daily total,

cardiovascular, and respiratory mortality: a global analysis in 398 cities

Table of contents

eMethods 1. Health and exposure data

eMethods 2. Lag structure

eMethods 3. Heterogeneity and effect modification analyses

eMethods 4. Concentration–response curves

eResults 1. Descriptive Statistics

eResults 2. Heterogeneity, regional analyses, and effect modifications

eTable 1. Descriptive statistics of annual-mean concentration (median, 25%-75%

percentiles) of nitrogen dioxide, weather conditions and other air pollutants in each

country or region throughout the study period.

eTable 2. Summary of the locations, GDP and region specification of the 22

countries/regions included in this analysis.

eTable 3. Percent changes (mean and 95% confidence intervals) in total mortality

associated with a 10 μg/m3 increase in nitrogen dioxide on lag 1 day, with and without

adjustment of co-pollutants.

eTable 4. Percentage change (mean and 95% confidence intervals) in total mortality

associated with a 10 μg/m3 increase in nitrogen dioxide (lag 1) stratified by WHO/GDP

regions and P-values for testing difference among regions.

Page 30 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


2

eTable 5. Percentage change (pooled estimate and 95% confidence intervals) in total

mortality per 10 μg/m3 increase in nitrogen dioxide (lag 1), with and without adjustment

of relative humidity.

eFigure 1. Geographic distributions of the 398 cities within the 22 countries included

in the analysis, and the corresponding annual mean NO2 concentrations (μg/m3).

eFigure 2. Percentage changes in total mortality associated with a 10 μg/m3 increase in

nitrogen dioxide (lag 1), with different modelling approaches to control for temperature.

eFigure 3. Concentration–response relationship curves between NO2 concentrations

(lag 1) and total mortality using alternative knots.

eReferences

Page 31 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


3

1. eMethods

1.1 Health and exposure data

We obtained mortality data from the Multi-City Multi-Country (MCC) database, which

has been described in previous publications 1,2. The current analysis was limited to

locations where NO2 ground measurements were available, which included a total of

398 cities in 22 countries and/or regions with different study periods, including

Australia (3 cities, 1988–2009), Brazil (one city, 1997–2011), Canada (25 cities, 1986–

2015), Chile (3 cities, 2005–2013), China (15 cities, 1996–2015), Colombia (1 city,

1998–2013), Czech Republic (1 city, 1994–2015), Estonia (4 cities, 1997–2015),

Finland (1 city, 1994–2014), Germany (12 cities, 1993-2015), Greece (1 city, 2001–

2010), Japan (47 cities, 2011–2015), Portugal (7 cities, 1990–2018), South Korea (8

cities, 1992–2015), Spain (48 cities, 1990–2014), Sweden (1 city, 1990–2010),

Switzerland (8 cities, 1995–2013), Taiwan (3 cities, 1994–2014), Thailand (18 cities,

1999–2008), United Kingdom (39 cities, 1990-2016), and United States (147 cities,

1973–2006). Mortality data were obtained from local authorities within each country/

region. Causes of death were classified according to the 9th or 10th version of

International Classification of Diseases (ICD) codes, wherever available. In each

location, mortality is represented by daily counts of either non-external causes or, where

not available, all-cause only (ICD-9: 0-799; ICD-10: A0-R99). We also collected

mortality data for two main causes: cardiovascular disease (ICD-10, codes I00-I99) and

respiratory disease (ICD-10, codes J00-J99) 3. Mortality from non-external causes were

not available in Chile, Estonia, Germany, Greece, Japan, Portugal, Romania, Sweden,

and United States. Deaths from cardiovascular and respiratory diseases were recorded

in 362 cities from 14 countries, and were not available in Australia, Brazil, Chile,

Estonia, Germany, and Romania (Table 1).

Daily data of particulate matter and other gaseous pollutants, including inhalable

particles (PM10), fine particulate matter (PM2.5), ozone (O3), sulfur dioxide (SO2) and

carbon monoxide (CO), were also obtained where available, to adjust for potential

Page 32 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


4

confounding by co-pollutants. There were 370 cities with PM10 data, 313 cities with

PM2.5 data, 364 cities with O3 data, 358 cities with SO2 data, and 298 cities with CO

data. In brief, measurements for air pollutants were obtained from fixed-site monitoring

networks operated by local authorities. The majority of monitors were located in urban

areas, only those daily measurements reporting above 75% of hourly data were included.

1.2 Lag structure

We selected and compared a list of lag structures for NO2 and temperature that

were reported in previous studies 2,4-6. The lags of NO2 include: 1) lag 0, the present

day; 2) lag 1, the previous day; 3) lag 2, the previous two day; 4) lag 3, the previous

three day; and 5) lag 0–1, the two-day moving average of the present day and the

previous day; 6) lag 0–2, the three-day moving average of the present day and the

previous two days; 7) lag 0–3, the four-day moving average of the present day and the

previous three days

For temperature, we tested the traditional single lag days and moving averages,

including 1) lag 0, the present day; 2) lag 0 and lag 1, separate terms of the present and

previous day; 3) lag 0–3, the 4-day moving average of the present and previous 3 days;

4) lag 0 and lag 1–3, separate terms of the present day and the average of the previous

3 days; 5) lag 0–7, the moving average of the present and previous 7 days; 6) lag 0–14,

the moving average of the present and previous 14 days, and 7) lag 0–21, the moving

average of the present and previous 21 days.

1.3 Heterogeneity and effect modification analyses

We explored potential effect modifiers on the associations between NO2 and total

mortality based on the main models. First, we conducted separate analyses by regions

classified by the World Health Organization (WHO) (eTable 2), which include

Western-Pacific Region (WPRO), Regional Office for the Americas (AMRO), and

Regional Office for Europe (EURO). Then, the statistical significance for differences

Page 33 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


5

among groups were determined by likelihood ratio tests. Second, using the

aforementioned approach, we conducted additional analyses by regions in terms of

Gross Domestic Product (GDP) per capita at country level (eTable 2). Third, we

assessed potential effect modification in the NO2-mortality associations by including

annual-mean levels of all air pollutants and temperature, relative humidity, latitude of

locations, region (WHO and GDP), and GDP per capita in meta-regression models all

together.

1.4 Concentration–response curves

To estimate the overall shape of the associations between NO2 concentrations and

total, cardiovascular, and respiratory mortality at the global level, we estimated the

concentration–response (C–R) relationship curves using the same approach as in

previous studies 7,8. The model specifications were the same as in the main model other

than the replacement of the linear term of NO2 with an assumed nonlinear term.

Following the distribution of annual mean NO2 concentrations at the city level, we first

used a quadratic B-spline for NO2 with two common knots at 20 μg/m3 and 40 μg/m3

(corresponding to 25% and 75% percentiles of annual mean NO2 concentrations across

all cities). This spline specification would allow enough flexibility to capture potential

non-linearity of the association in a wide range of concentrations across cities. Finally,

we combined the city-specific components of the spline estimates using random-effect

models. In the sensitivity analysis, we placed knots at different NO2 concentrations for

the curve of NO2 with total mortality, including (15, 40) μg/m3, (25, 40) μg/m3 and (25,

45) μg/m3.

Page 34 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


6

2. Results

2.1 Descriptive Statistics

This analysis included 60.1 million deaths from total or non-external causes, and there

were 19.7 and 5.5 million deaths from cardiovascular diseases and respiratory diseases,

respectively, at locations with available death classification data (Table 1). Figure 1

illustrates the locations of cities included in present analysis and the average values of

annual mean concentrations of NO2 during the periods with available NO2 ground

measurements at city level. Table 2 summarizes the descriptive statistics of

environmental data. On average, the median annual mean NO2 concentration across

398 cities was 26.9 μg/m3 (25%-75% percentiles: 19.5-36.2 μg/m3), which was lower

than the Air Quality Guidelines for NO2 (annual-mean: 40 μg/m3) of the WHO 9. The

median values of annual mean concentrations of PM10, PM2.5, O3, SO2, and CO across

all cities were 27.6 μg/m3, 11.4 μg/m3, 49.7 μg/m3, 8.7 μg/m3 and 0.6 mg/m3,

respectively. The Spearman correlation coefficients of NO2 were 0.34 with PM10, 0.38

with PM2.5, -0.24 with O3, 0.42 with SO2, and 0.56 with CO. There were marked

variations in climatic conditions among the cities, with the median of temperature of

14.4°C (25%-75% percentiles: 7.8°C to 20.6°C), and relative humidity of 71% (25%-

75% percentiles: 59%-78%). On average, NO2 was negatively correlated with mean

temperature (Pearson r = -0.29) and relative humidity (Pearson r = -0.15).

2.2 Heterogeneity, regional analyses, and effect modifications

Among all cities, heterogeneity was found in the NO2-mortality association across

country and city-specific estimates, with I2 statistics of 47.7%, and Cochran Q P-value

< .001.

In regional analyses (eTable 4), the NO2-mortality association was largest in the

AMRO with an average increment of 0.53% in total mortality per 10 μg/m3 increase of

NO2 concentrations, and was smallest in the EURO (corresponding estimate: 0.36%).

Page 35 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


7

The associations did not vary between different groups classified by GDP, with

estimates of 0.48%, 0.49%, and 0.50% corresponding to low, medium, and high GDP

areas, respectively. There were no significant effect modifications by annual levels of

air pollutants, temperature, relative humidity, GDP, region and latitude of locations.

Page 36 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


8

3. eTables

eTable 1. Descriptive statistics of annual-mean concentration (median, 25%-75% percentiles) of nitrogen dioxide, weather conditions

and other air pollutants in each country or region throughout the study period.

Country

/RegionNO2 (μg/m3) PM10 (μg/m3) PM2.5 (μg/m3) O3 (μg/m3) SO2 (μg/m3) CO (mg/m3) Temperature (℃) Humidity (%)

Australia 21.4 [14.1-27.9] 18.2 [14.5-23.2] 6.0 [4.3-8.4] 29.9 [22.1-37.9] 3.3 [1.4-5.7] 0.4 [0.2-0.7] 18.1 [14.7-21.2] 70.1 [62.5-77.2]

Brazil 84.9 [62.6-115.3] 39.6 [29.0-54.0] NA 75.4 [51.9-104.5] 10.7 [7.3-15.4] 1.9 [1.3-2.8] 20.6 [18-22.9] 74.8 [67.0-81.0]

Canada 23.7 [15.6-33.8] NA 6.7 [4.3-10.3] 41.4 [28.9-54.9] 3.5 [1.5-7.3] 0.3 [0.2-0.5] 7.4 [-0.9-15.7] 73.2 [65.6-80.9]

Chile 21.6 [13.9-32.4] 46.8 [34.3-70.7] 21.5 [14.2-38] 25.2 [13.8-33.7] NA 0.6 [0.3-1.1] 13.7 [10.7-17.2] NA

China 46.5 [36.2-60.4] 95.2 [67.2-133.6] 49.7 [30.1-80] 82.3 [49.9-121.5] 37.1 [24.9-65.3] 1.0 [0.7-1.3] 16.3 [6.5-23.5] 66.2 [54.9-77]

Colombia 30.5 [23.3-37.9] 61.3 [49-74.5] NA 24.1 [18.4-30.9] 19.8 [10.8-28.5] NA 13.9 [13.2-14.5] NA

Czech Republic 30.8 [24.2-38.7] 29.4 [19.9-44.7] NA 69.3 [47.4-95] 7.2 [4.2-15.2] NA 9.2 [2.7-15.3] 78.0 [68.0-86.0]

Estonia 11.4 [7.6-16.7] 14.4 [9.3-21.8] 6.6 [3.8-10.6] 48.9 [36.7-61.8] 1.3 [0.6-3.3] NA 6 [-0.1-13.6] 83.5 [73.8-90.8]

Finland 6.8 [4.3-11.7] 14.7 [8.9-25.5] 12.6 [7.4-21.7] 51.3 [40.0-63.0] 6.2 [3.5-11.6] 0.3 [0.2-0.3] 5.9 [0-13.8] 79.2 [67.8-87.7]

Germany 29.6 [21.8-38.4] 20.1 [14.3-28.7] 12.0 [8.0-18.8] 38.1 [21-54.6] 4.7 [3.1-8.5] 0.4 [0.3-0.6] 10.5 [4.8-15.9] NA

Greece 50.2 [39.6-61.6] 39.5 [29.5-53.1] 20.3 [15.0-26.4] 75.1 [52.8-97.5] NA 1.8 [1.3-2.6] 17.9 [12.9-24.9] 66.0 [54.0-75.4]

Page 37 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


9

Japan 16.7 [12.1-23.4] 16.7 [11.8-23.8] 12.6 [8.4-18.3] 56.2 [41.3-72.2] 4.9 [3.5-7.2] NA 16.1 [7.6-22.7] 69.8 [60.9-77.9]

Portugal 14.9 [10.0-21.7] 18.8 [12.9-28.3] 8.2 [5.4-12.8] 54.1 [41.1-66.5] 2.0 [1.3-3.5] 0.3 [0.2-0.4] 15.4 [11.5-19.9] NA

Romania 25.6 [17.7-36.2] 27.7 [18.9-40.1] 15.5 [10.2-22.5] 36.8 [24.2-50.3] 7.1 [5.1-10.0] 0.2 [0.1-0.5] 11.4 [3.4-18.9] 76.3 [66.5-85.9]

South Korea 43.7 [33.0-57.5] 46.0 [33.3-63.0] NA 41.2 [28.3-56.7] 13.9 [10.5-18.7] 0.6 [0.5-0.8] 14.9 [5.7-21.9] 65.7 [53.2-76.0]

Spain 26.4 [20.2-33.9] 26.2 [19.4-35.4] 10.0 [7.0-14.1] 51.6 [37.5-63.9] 5.0 [3.8-6.7] 0.4 [0.3-0.5] 14.9 [10.2-20.7] NA

Sweden 26.8 [20.0-34.8] 12.5 [9.3-17.9] 6.6 [4.7-9.5] 61.9 [48.9-76.0] NA 0.9 [0.6-1.5] 6.8 [1.2-13.9] 79.6 [68.4-87.6]

Switzerland 32.3 [24.0-42.0] 21.3 [14.2-31.6] 16.2 [10.6-24.7] 72.8 [47.0-98.1] 3.8 [1.8-7.5] 0.6 [0.4-0.9] 10.7 [4.4-16.5] 75.4 [66.3-83.2]

Taiwan 42.2 [31.6-54.7] 58.8 [38.7-81.5] 32.3 [20.6-44.6] 65 [44.5-91.6] 13.4 [9.6-18.7] 0.7 [0.6-0.9] 24.9 [20.4-28.0] 75.3 [70.3-80.2]

Thailand 22.0 [16.3-30.4] 44.1 [32.6-63.1] NA 31.9 [22.6-45.3] 9.1 [6.1-12.8] 0.6 [0.5-0.9] 28 [26.4-29] 74.2 [67.7-79.9]

United Kingdom 25.6 [17.9-35.6] 18.7 [13.7-25.9] 9.2 [6.4-14.5] 43.0 [30.1-55.7] 5.5 [2.7-9.9] 0.3 [0.2-0.5] 10.4 [6.6-14.7] NA

United States 28.7 [20.8-38.6] 24.1 [17-33.5] 10.9 [7.4-16] 52.1 [36.3-68.5] 10.5 [5.9-17.7] 0.8 [0.6-1.2] 14.7 [7.2-21.8] 65.3 [55.2-75.5]

Total 26.9 [19.5-36.2] 27.6 [19.7-38.6] 11.4 [7.7-17.0] 49.7 [34.9-65.1] 8.7 [5.4-14.3] 0.6 [0.4-0.9] 14.4 [7.8-20.6] 68.5 [59.0-77.6]

Abbreviations: NO2, nitrogen dioxide; PM10, particulate matter with an aerodynamic diameter less than or equal to 10 μm; PM2.5, particulate matter

with an aerodynamic diameter less than or equal to 2.5 μm; O3, ozone; SO2, sulfur dioxide; CO, carbon monoxide.

Page 38 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


10

eTable 2. Summary of the locations, GDP and region specification of the 22

countries/regions included in this analysis.

Country/Region Cities GDP per capita WHO region GDP region

Australia 3 54066 WPRO 3

Brazil 1 9925 AMRO 1

Canada 25 45149 AMRO 3

Chile 3 14999 AMRO 1

China 15 8879 WPRO 1

Colombia 1 6378 AMRO 1

Czech Republic 1 20380 EURO 1

Estonia 4 20388 EURO 1

Finland 1 46317 EURO 3

Germany 12 44350 EURO 3

Greece 1 18930 EURO 1

Japan 47 38387 WPRO 2

Portugal 5 21490 EURO 1

Romania 8 10808 EURO 1

South Korea 7 31617 WPRO 2

Spain 48 28170 EURO 2

Sweden 1 53792 EURO 3

Switzerland 8 80450 EURO 3

Taiwan 3 24283 WPRO 2

Thailand 18 6593 WPRO 1

United Kingdom 39 40361 EURO 3

United States 147 59958 AMRO 3

Abbreviations: GDP, Gross Domestic Product; WHO, World Health Organization.

Notes: Regions by GDP: Regions classified by GDP per capita in 2017 according to

World Bank (data.worldbank.org). Regions by WHO: classified by the World Health

Organization, including the Western Pacific Region (Western-Pacific Regional Office,

Page 39 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


11

WPRO), and the European Region (Regional Office for Europe, EURO).

Page 40 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


12

eTable 3. Percent changes (mean and 95% confidence intervals) in total mortality

associated with a 10 μg/m3 increase in nitrogen dioxide on lag 1 day, with and

without adjustment of co-pollutants.

Models N Estimates P* for difference

Single-pollutant 0.47 (0.36 to 0.58)

+ PM10

3700.38 (0.26 to 0.51)

0.082


+ PM2.5

3130.37 (0.25 to 0.48)

0.076


+SO2

3640.45 (0.34 to 0.55)

0.578


+O3

3580.47 (0.36 to 0.57)

0.617


+CO298

0.48 (0.37 to 0.60)0.597

Abbreviations as in Table 2. N, number of cities with available data.

* P-value for difference was calculated by evaluating a binary variable (with and

without the adjustment for co-pollutant) in likelihood ratio tests with both single- and

two- pollutant model estimates. A p-value < 0.05 was considered statistically

significant for differences.

Page 41 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


13

eTable 4. Percentage change (mean and 95% confidence intervals) in total

mortality associated with a 10 μg/m3 increase in nitrogen dioxide (lag 1) stratified

by WHO/GDP regions and P-values for testing difference among regions.

Classifications Regions Mortality P-values*

AMRO 0.53 (0.42 to 0.65)

WPRO 0.49 (0.36 to 0.61)WHO

EURO 0.36 (0.24 to 0.48)

0.007

WHO WPRO 0.88 (0.75 to 1.02) <0.0011 0.48 (0.43 to 0.54)

2 0.49 (0.44 to 0.55)GDP

3 0.50 (0.45 to 0.56)

0.791

Abbreviations as in eTable 2.

* P-values were obtained by likelihood ratio tests comparing the fit of a meta-regression

model with the region variable to the simple meta-analysis model. P-values < .05 was

considered statistically significant for regional differences.

Page 42 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


14

eTable 5. Percentage change (pooled estimate and 95% confidence intervals) in

total mortality per 10 μg/m3 increase in nitrogen dioxide (lag 1), with and without

adjustment of relative humidity.

Endpoints Adjustments N Estimates P*

RH adjusted 0.51 (0.44 to 0.59)Total

RH unadjusted290

0.54 (0.47 to 0.61)0.440

RH adjusted 0.44 (0.32 to 0.56)Cardiovascular

RH unadjusted271

0.46 (0.34 to 0.58)0.757

RH adjusted 0.63 (0.53 to 0.73)Respiratory

RH unadjusted271

0.67 (0.57 to 0.76)0.668

Abbreviations: RH, relative humidity.

* P-values for difference was calculated by evaluating a binary variable (with and

without the adjustment for humidity) in likelihood ratio tests with both model estimates.

P-values < .05 was considered statistically significant for differences.

Page 43 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


15

eFigure 1. Geographic distributions of the 398 cities within the 22 countries included in the analysis, and the corresponding annual mean

NO2 concentrations (μg/m3).

Page 44 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


16

eFigure 2. Percentage changes in total mortality associated with a 10 μg/m3 increase in nitrogen dioxide (lag 1), with different modelling

approaches to control for temperature. Lag 0, the present day; Lag 0+1, the present day and the previous day; Lag 0+1–3, the present day and

the mean of the previous 3 days; Lag 0–3, moving average of the present day and the previous 3 days; Lag 0–7, moving average of the present day

Page 45 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


17

and the previous 7 days; Lag 0–14, moving average of the present day and the previous 14 days; Lag 0–21, moving average of the present day and

the previous 21 days.

Page 46 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


18

eFigure 3. Concentration–response relationship curves between NO2

concentrations (lag 1) and total mortality using alternative knots. The y-axis can

be interpreted as the relative change from the mean effect of NO2 on mortality; the

fraction of the curve below zero denotes a smaller estimate compared with the mean

effect. (A) knots at (20, 40) μg/m3 (main analysis); (B) knots at (15, 40) μg/m3; (C)

knots at (25, 40) μg/m3; and (D) knots at (25, 45) μg/m3.

Page 47 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


19

4. eReference

1. Gasparrini A, Guo Y, Hashizume M, et al. Mortality risk attributable to high

and low ambient temperature: a multicountry observational study. Lancet

(London, England). 2015;386(9991):369-375.

2. Liu C, Chen R, Sera F, et al. Ambient Particulate Air Pollution and Daily

Mortality in 652 Cities. N Engl J Med. 2019;381(8):705-715.

3. WHO. International statistical classification of diseases and related health

problems. Vol 1: World Health Organization; 2004.

4. Vicedo-Cabrera AM, Sera F, Liu C, et al. Short term association between ozone

and mortality: global two stage time series study in 406 locations in 20 countries.

BMJ. 2020;368:m108.

5. Chen R, P Y, X M, et al. Associations Between Ambient Nitrogen Dioxide and

Daily Cause-specific Mortality: Evidence from 272 Chinese Cities.

Epidemiology. 2018;29(4):482.

6. Guo Y, Gasparrini A, Armstrong B, et al. Global variation in the effects of

ambient temperature on mortality: a systematic evaluation. Epidemiology.

2014;25(6):781-789.

7. Chen R, Yin P, Meng X, et al. Fine Particulate Air Pollution and Daily Mortality.

A Nationwide Analysis in 272 Chinese Cities. Am J Respir Crit Care Med.

2017;196(1):73-81.

8. Samoli E, Analitis A, Touloumi G, et al. Estimating the exposure-response

relationships between particulate matter and mortality within the APHEA

multicity project. Environmental health perspectives. 2005;113(1):88-95.

9. WHO. Air quality guidelines for particulate matter, ozone, nitrogen dioxide and

sulfur dioxide: global update 2005. In: World Health Organization; 2006.

Page 48 of 47


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Documents

BMJ...2020/08/22 · 25 Universidad Pablo de Olavide, Department of Physical, Chemical, and Natural Systems, Sevilla, Spain; 26 Institute of Epidemiology, Helmholtz Zentrum München