Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Dec 7.
Published in final edited form as: Environ Int. 2023 Oct 10;181:108258. doi: 10.1016/j.envint.2023.108258

Joint effect of heat and air pollution on mortality in 620 cities of 36 countries

Massimo Stafoggia a,*, Paola Michelozzi a, Alexandra Schneider b, Ben Armstrong c, Matteo Scortichini a, Masna Rai b, Souzana Achilleos d, Barrak Alahmad e, Antonis Analitis f, Christofer Åström g, Michelle L Bell h, Neville Calleja i, Hanne Krage Carlsen j, Gabriel Carrasco k, John Paul Cauchi l, Micheline DSZS Coelho m, Patricia M Correa n, Magali H Diaz o, Alireza Entezari p, Bertil Forsberg g, Rebecca M Garland q, Yue Leon Guo r, Yuming Guo s, Masahiro Hashizume t, Iulian H Holobaca u, Carmen Íñiguez v, Jouni JK Jaakkola w, Haidong Kan x, Klea Katsouyanni f,y, Ho Kim z, Jan Kyselý aa,ab, Eric Lavigne ac,ad, Whanhee Lee h, Shanshan Li s, Marek Maasikmets ae, Joana Madureira af,ag,ah, Fatemeh Mayvaneh p, Chris Fook Sheng Ng t, Baltazar Nunes ai, Hans Orru aj, Nicolás V Ortega n, Samuel Osorio ak, Alfonso DL Palomares al, Shih-Chun Pan am, Mathilde Pascal an, Martina S Ragettli ao, Shilpa Rao al, Raanan Raz ap, Dominic Roye aq,ar, Niilo Ryti w, Paulo HN Saldiva m, Evangelia Samoli f, Joel Schwartz e, Noah Scovronick as, Francesco Sera c,at, Aurelio Tobias au, Shilu Tong av, César DLC Valencia o, Ana Maria Vicedo-Cabrera aw,ax, Aleš Urban aa,ab, Antonio Gasparrini c, Susanne Breitner ay, Francesca K de’ Donato a
PMCID: PMC10702017  NIHMSID: NIHMS1945853  PMID: 37837748

Abstract

Background:

The epidemiological evidence on the interaction between heat and ambient air pollution on mortality is still inconsistent.

Objectives:

To investigate the interaction between heat and ambient air pollution on daily mortality in a large dataset of 620 cities from 36 countries.

Methods:

We used daily data on all-cause mortality, air temperature, particulate matter ≤ 10 μm (PM10), PM ≤ 2.5 μm (PM2.5), nitrogen dioxide (NO2), and ozone (O3) from 620 cities in 36 countries in the period 1995–2020. We restricted the analysis to the six consecutive warmest months in each city. City-specific data were analysed with over-dispersed Poisson regression models, followed by a multilevel random-effects meta-analysis. The joint association between air temperature and air pollutants was modelled with product terms between non-linear functions for air temperature and linear functions for air pollutants.

Results:

We analyzed 22,630,598 deaths. An increase in mean temperature from the 75th to the 99th percentile of city-specific distributions was associated with an average 8.9 % (95 % confidence interval: 7.1 %, 10.7 %) mortality increment, ranging between 5.3 % (3.8 %, 6.9 %) and 12.8 % (8.7 %, 17.0 %), when daily PM10 was equal to 10 or 90 μg/m3, respectively. Corresponding estimates when daily O3 concentrations were 40 or 160 μg/m3 were 2.9 % (1.1 %, 4.7 %) and 12.5 % (6.9 %, 18.5 %), respectively. Similarly, a 10 μg/m3 increment in PM10 was associated with a 0.54 % (0.10 %, 0.98 %) and 1.21 % (0.69 %, 1.72 %) increase in mortality when daily air temperature was set to the 1st and 99th city-specific percentiles, respectively. Corresponding mortality estimate for O3 across these temperature percentiles were 0.00 % (−0.44 %, 0.44 %) and 0.53 % (0.38 %, 0.68 %). Similar effect modification results, although slightly weaker, were found for PM2.5 and NO2.

Conclusions:

Suggestive evidence of effect modification between air temperature and air pollutants on mortality during the warm period was found in a global dataset of 620 cities.

Keywords: Air temperature, Air pollution, Effect modification, Epidemiology, Mortality

1. Introduction

Air pollution and climate change are closely linked: many ambient air pollutants contribute to climate change and changes in climate have impacts on air quality (Kinney, 2018). Several air pollutants are responsible for the health burden, specifically particulate matter (PM), nitrogen dioxide (NO2), and ozone (O3) (WHO, 2021; Vicedo-Cabrera et al., 2020; Liu et al., 2019; Orellano et al., 2020; Dominski et al., 2021; Meng et al., 2021). PM is a mixture of solid and liquid particles originating from different sources, with both anthropogenic (vehicular traffic, domestic heating, industry) and natural (wildfires, desert dust) sources causing adverse effects on human health (WHO, 2021). NO2 is a gaseous air pollutant originated from burning fossil fuels (coal, oil, gas or diesel) at high temperatures. Its largest sources are motor vehicles and industrial plants, therefore NO2 concentrations are highest in urban and industrial areas (WHO, 2021). O3 is a highly reactive secondary pollutant originating from the reaction between anthropogenic and biogenic precursors such as nitrogen oxides (NOx) and volatile organic compounds (VOCs) in the presence of sunlight (WHO, 2021). Numerous epidemiological studies have shown that short-term exposure to PM, NO2, and O3 are associated with adverse health outcomes, including increased daily mortality and morbidity (Vicedo-Cabrera et al., 2020; Liu et al., 2019; Orellano et al., 2020; Dominski et al., 2021; Meng et al., 2021; Stafoggia et al., 2010).

The association between extreme ambient temperatures, especially during the warm season, and daily mortality has been extensively documented in the epidemiological literature (Gasparrini et al., 2015; Anderson and Bell, 2009; Basu, 2009; Song et al., 2017). Most studies report immediate (i.e., up to 3 days) associations with non-linear effects of high summer temperatures (Gasparrini et al., 2015). Such associations are observed in different geographical areas, with some heterogeneity related to several factors such as climatic conditions to which local populations are acclimatized, local population characteristics, and the diverse vulnerability of the underlying population, among others (Gasparrini et al., 2015; Anderson and Bell, 2009; Basu, 2009; Song et al., 2017; Gasparrini et al., 2015; Lavigne et al., 2014; Zhao et al., 2021; Sera et al., 20192019).

The interaction between heat and ambient air pollution on human health is less investigated. Most of the studies have been conducted in single cities or multiple cities from the same region, often using heterogeneous methodological approaches to adjust for time-varying confounders or to test for the presence of effect modification between air pollutants and temperatures on human health (Chen et al., 2017; Anenberg et al., 2020; Li et al., 2017). Furthermore, most studies have considered either air pollution as an effect modifier of the air temperature-mortality relationship, or vice versa (Chen et al., 2017; Anenberg et al., 2020; Li et al., 2017; Rai et al., 2023).

The objective of this multicentre analysis was to investigate the joint short-term effects of heat and air pollutants on all-cause mortality on a global scale. We used the Multi-Country Multi-City (MCC) Collaborative Research Network dataset (Gasparrini et al., 2015) and applied a comprehensive, consistent modelling framework to estimate the health risks and compare the associations at the global and continental level.

2. Methods

2.1. Data collection

Daily time-series data on mortality counts, mean air temperature, and air pollution for 620 cities across 36 countries were retrieved from the database of the MCC Network, a voluntary-based collaborative research network where investigators provide daily data on mortality counts, air pollutants concentrations and air temperature to be used for environmental epidemiology investigations (Gasparrini et al., 2015). The years of data available differed by city. We restricted the study periods to start at 1995 and selected only the warm seasons, defined for each city as the six consecutive warmest months for the whole study period, based on air temperature data. Furthermore, we only included cities with at least three years of data.

Mortality data was represented by daily counts of deaths due to non-external causes (International Classification of Diseases codes 0–799 [9th revision] or codes A00–R99 [10th revision]), or by all-cause deaths when data on non-external causes were not available.

For each city, data on daily mean air temperature from stations included in local monitoring networks were considered. Daily average concentrations of PM10, PM2.5 and NO2, and daily maximum 8-hour moving average O3 concentrations, were collected from urban and suburban air quality monitoring stations for subsets of cities: 372 cities had data on PM10, 486 on PM2.5, 386 on O3, and 411 on NO2. For NO2 and O3, when data were available in parts per billion (ppb), they were converted into micrograms per cubic meter (μg/m3) using 1 ppb = 1.88 μg/m3 and 1 ppb = 1.96 μg/m3 conversion factors, respectively. Details on data collection are provided in the appendix (p 2).

2.2. Statistical analysis

We adopted a two-stage design, where city-specific data were analysed in the first stage, and pooled results were obtained in the second stage.

For each city, we assumed an over-dispersed Poisson distribution for the daily mortality counts, and applied time-series regression adjusted for long-term and seasonal time trends and the day of the week. Air temperature, PM10, NO2 and O3 were modelled at lag 0–1 (average of current and previous day exposures) based on previous publications using the same data (Vicedo-Cabrera et al., 2020; Liu et al., 2019; Meng et al., 2021; Gasparrini et al., 2015). PM2.5 was modelled as a single-day lag (lag 0) because approximately 100 cities (mostly located in the U.S.) had data only every third day. Long-term and seasonal time trends were adjusted for by fitting year-specific natural splines with 4 degrees of freedom (d.f.) and a natural spline of calendar year with one knot every three years, and day of the week was modelled as a categorical variable. Firstly, we analysed each exposure individually, by modelling air temperature with a natural spline with four d.f. and air pollutants with linear terms. Models for air temperature were fit with and without adjustment for air pollutants, while models for air pollutants (one per each pollutant) were always fit with adjustment for air temperature. In each city, effects were estimated as a % change in mortality, with 95 % confidence intervals (95 % CI), per an increase in mean temperature from the 75th to the 99th percentile. The effect of air pollutants was estimated per a 10 μg/m3 increment in the exposure. This choice was motivated by comparability with the existing literature, although we acknowledge that 10 μg/m3 captures different amounts of daily variability both across locations and across pollutants.

Secondly, in each city, we modelled the interaction between mean air temperature and air pollution on mortality by defining a product term between the natural spline of air temperature and the linear term of air pollutant. From this model, we calculated the % change in mortality associated with 75th-99th percentile increase in temperature, for increasing levels of air pollutants concentrations: from daily averages of 10 μg/m3 to 90 μg/m3 (for PM10), from 1 μg/m3 to 40 μg/m3 (for PM2.5), from 40 μg/m3 to 160 μg/m3 (for O3), and from 1 μg/m3 to 60 μg/m3 (for NO2). These ranges were chosen based on inspection of city-specific air pollutants distributions on warm-season days with air temperature between 75th and 99th distribution (Figure S1 of the appendix). Similarly, from the same joint model, we calculated the % change in pollutant-related mortality for increasing levels of mean air temperature, from the 1st to the 99th percentile of city-specific distributions. Since we modelled the interaction using a product term in the log-linear model, it is implicit that we modelled a “multiplicative” interaction, rather than an additive one. We acknowledge that also the latter was of interest for our study hypothesis, however it raised methodological complexities which were beyond the scope of the paper. More details on the methodological approach are reported in the appendix (pp 34).

In the second stage, we pooled the city-specific estimates of the main effects, and the estimates of level-specific effects from the effect modification analysis, with multilevel random-effects meta-analyses, where the countries and cities were modelled as nested random effects (Vicedo-Cabrera et al., 2020; Liu et al., 2019; Gasparrini et al., 2015). Similarly, we pooled the city-specific estimates of the product terms between air temperature and each air pollutant as a formal test of interaction. See the appendix (p 4) for further details.

We carried out a series of sensitivity analyses to check the robustness of our main findings to different modelling choices and definitions. Since our main focus was on heat, we restricted the “warm” season to the three warmest consecutive months instead of six for each city (as done in the main analysis). Secondly, we estimated the main effect of air temperature or air pollutants using the alternative lags 0–3 (average of same-day and previous 3-days exposure) or 0–10 (average of same-day and previous-10 days exposure, only for air temperature), in order to capture possible cumulative effects on multiple days. Furthermore, we checked the robustness of our results concerning the time trend adjustment, by modelling it with natural splines with two or six d.f. per year, instead of four. Finally, we used the 50th percentile of air temperature as reference point, instead of the 75th, to estimate the association between air temperature and mortality. More details are reported in the appendix (p 4).

All analyses were conducted using the R statistical software, version 4.1.2 (The R Foundation for Statistical Computing, Vienna, Austria), using the mgcv, spline and dlnm packages in the first-stage analysis and the mixmeta package in the second-stage analysis.

3. Results

A map of the geographical distribution of the 620 cities is shown in Fig. 1 (air temperature data), and in Figure S2 (four panels reporting data for the individual pollutants); country-specific descriptive statistics are reported in Table 1, while those for individual cities are reported in the appendix, Table S1. We analysed 22,630,598 deaths from non-external causes (284 cities) or all-causes (336 cities) occurring in 36 countries. The warm season average ambient air temperature ranged from 9 °C in Reykjavik (Iceland) to 36 °C in Kuwait City (Kuwait), with a temperature difference of 4–6 °C between the 75th and the 99th percentile for most cities. PM10 and PM2.5 warm-season average concentrations varied greatly across cities, with lower values observed in several U.S., Canadian and Scandinavian cities, and highest concentrations detected in “hot spot” areas, such as Kuwait City and Shanghai (China). O3 concentrations also varied across the 376 cities with available data, ranging from 31 μg/m3 in Sidney (Australia) to 175 μg/m3 in the Valley of Mexico. NO2 mean concentrations ranged between 4 μg/m3 (in two cities from Japan and the U.S.) and 87 μg/m3 (in Teheran, Iran).

Fig. 1.

Fig. 1.

Distribution of the cities with data on air temperature and either of the four pollutants. Colour shading represents the average of daily ambient temperature during the warm period in the studied cities.

Table 1.

Environmental and mortality data relative to the warm period only (defined as the six warmest consecutive months in each city): data by country.

Country N.cities Period
Mortality
Air temperature
PM10*
PM2.5
O3
NO2§
N. deaths daily mean daily mean 75th-99th percentiles daily mean SD daily mean SD daily mean SD daily mean SD

Australia 3 2000 2009 241,720 46 21 24 29 20 10 7 5 33 11 17 9
Brazil 1 1997 2011 433,127 171 22 24 27 37 14 85 38
Canada 25 1995 2015 1,172,352 13 15 19 26 19 12 8 7 81 31 21 13
Chile 3 2004 2014 144,477 32 17 19 25 45 26 20 13 18 15
China 14 1996 2015 567,084 48 23 27 32 90 55 50 34 45 23
Colombia 1 1998 2013 183,207 63 14 15 16 62 20 30 10
Cyprus 5 2005 2019 33,777 3 26 28 33 41 41 18 24 23 10
Czech Republic 1 1995 2009 95,671 35 15 19 25 30 16 96 27 28 9
Ecuador 1 2014 2018 21,987 24 16 16 19 15 5 25 7
Estonia 4 2002 2020 54,359 5 13 17 24 16 11 6 4 52 16 10 9
Finland 1 1995 2014 60,750 17 13 17 24 17 14 14 12 8 6
France 20 2000 2017 914,028 14 18 21 28 20 9 12 7 20 11
Germany 12 1995 2015 1,341,855 29 16 19 26 23 13 12 6 74 32 27 12
Greece 1 2001 2010 137,232 75 25 29 33 45 19 23 10 93 22 51 17
Iceland 1 2002 2018 10,364 3 9 11 15 17 14
Iran 1 2002 2015 315,677 132 26 29 35 88 51 87 73
Israel 4 2000 2020 184,415 12 25 28 31 43 59 17 17 19 12
Italy 18 2006 2015 376,961 12 22 25 31 27 12 92 27
Japan 49 1995 2019 2,531,598 26 23 27 31 31 17 13 8 84 32 15 9
Kuwait 1 2010 2016 15,903 12 36 39 43 220 202
Malta 1 2006 2019 18,816 8 24 27 31 39 15 18 8 36 15
Mexico 9 2000 2012 857,491 58 21 23 31 50 24 27 11 117 48
Norway 1 2000 2018 36,656 11 12 15 22 18 8 9 4
Peru 1 2010 2014 88,703 97 21 22 24 82 26 25 11
Portugal 6 1995 2018 461,206 24 20 23 30 24 16 10 7 79 24 16 13
Puerto Rico 1 2009 2016 13,241 9 28 29 30 30 18
Romania 8 2008 2016 174,385 14 19 23 29 28 14 13 7 22 12
South Africa 7 2004 2013 478,780 55 21 23 27 44 28 25 18 73 29
South Korea 7 1999 2015 810,954 37 22 25 30 46 24 74 31 40 18
Spain 51 2001 2014 768,533 7 21 24 30 29 15 13 8 78 23 26 12
Sweden 1 1995 2010 71,764 24 14 17 23 14 7 8 5 69 19 27 11
Switzerland 8 1995 2013 110,620 4 16 20 26 21 11 15 8 91 36 29 15
Taiwan 3 1995 2014 523,614 47 28 29 31 50 23 27 13 115 42 36 14
Thailand 18 1999 2008 404,853 13 29 30 33 44 24 21 12
United Kingdom 123 1995 2018 2,138,448 7 15 17 22 20 10 11 4 24 13
United States 209 1995 2006 6,835,990 16 21 26 32 28 15 13 8 93 35 27 16
*

Canada 8 cities, Colombia 2001–2013, Mexico 7 cities, Portugal 5 cities, Romania 3 cities, South Africa 6 cities, Spain 42 cities 2001–2014, UK 31 cities, United States 89 cities.

China 3 cities 2013–2015, Chile 2008–2014, Cyprus 2 cities, Estonia 3 cities 2008–2018, Germania 11 cities 2004–2015, Greece 2007–2010, Israel 3 cities, Japan 48 cities, Mexico 2 cities 2003–2012, Portugal 4 cities 2004–2018, Romania 7 cities, South Africa 5 cities, Spain 11 cities, Sweden 2001–2010, Switzerland 4 cities, Taiwan 2007–2014, UK 119 cities, United States 203 cities.

Italy 14 cities, South Africa 6 cities, Spain 49 cities, United States 189 cities. For Japan, ozone data was derived from the measurements of photochemical oxidant, which is primarily ozone (≥90 %), followed by others such as peroxyacetyl nitrate (PAN), hydrogen peroxide (H2O2) and organic hydroperoxides.

§

France 18 cities, Spain 48 cities, UK 36 cities, United States 130 cities.

The pooled estimates of the associations between each environmental exposure and mortality are reported in Table 2. Overall, an increase in air temperature from the 75th to the 99th percentile of the city-specific distribution was associated on average with an 8.9 % (95 % confidence interval [95 % CI]: 7.1 %, 10.7 %) increase in mortality. 10 μg/m3 increases in lag 0–1 PM10, lag 0 PM2.5, lag 0–1 O3 and lag 0–1 NO2 daily concentrations were associated with changes in mortality of 0.41 % (95 % CI: 0.28 %, 0.53 %), 0.61 % (95 % CI: 0.40 %, 0.82 %), 0.26 % (95 % CI: 0.15 %, 0.36 %), and 0.57 % (95 % CI: 0.38 %, 0.77 %), respectively. Sensitivity analyses showed that a different definition of the warm season or different lags and model adjustments did not substantially alter the main findings (Table 2).

Table 2.

Association between daily mean air temperature, air pollutants and all-cause mortality in the warm season: % change in mortality, and 95% confidence intervals, at the specified increment of the exposure. Meta-analytical results of the main model and of the sensitivity analyses*.

Exposure Increment (percentiles) Model N. cities % change 95 % CI

Air temperature 75th-99th Main model 620 8.89 7.12 10.68
Adj. lag 0–1 PM10 372 8.56 6.99 10.16
Adj. lag 0 PM2.5 486 8.00 6.16 9.88
Adj. lag 0–1 O3 386 8.76 6.28 11.29
Adj. lag 0–1 NO2 411 8.87 7.25 10.51
Warm season as 3 months 620 9.02 7.07 11.01
Time trends 2 d.f./year 620 8.79 7.03 10.59
Time trends 6 d.f./year 620 8.59 6.86 10.35
Lag 0–3 620 8.66 6.59 10.76
Lag 0–10 620 5.84 4.01 7.70
50th-99th 50th pct as reference 620 10.65 8.30 13.05
PM10 10 μg/m3 Main model 372 0.41 0.28 0.53
Warm season as 3 months 369 0.52 0.29 0.74
Time trends 2 d.f./year 372 0.40 0.25 0.55
Time trends 6 d.f./year 372 0.41 0.28 0.54
Lag 0–3 369 0.25 0.13 0.37
PM2.5 10 μg/m3 Main model 486 0.61 0.40 0.82
Warm season as 3 months 482 0.58 0.31 0.85
Time trends 2 d.f./year 486 0.64 0.43 0.86
Time trends 6 d.f./year 486 0.55 0.34 0.77
Lag 0–3 389 0.34 0.05 0.62
O3 10 μg/m3 Main model 386 0.26 0.15 0.36
Warm season as 3 months 386 0.23 0.11 0.35
Time trends 2 d.f./year 386 0.21 0.09 0.33
Time trends 6 d.f./year 386 0.26 0.15 0.37
Lag 0–3 386 0.26 0.14 0.39
NO2 10 μg/m3 Main model 411 0.57 0.38 0.77
Warm season as 3 months 411 0.54 0.37 0.70
Time trends 2 d.f./year 411 0.52 0.32 0.73
Time trends 6 d.f./year 411 0.57 0.38 0.76
Lag 0–3 411 0.54 0.32 0.75
*

“Main model”: exposure modelled with a natural spline with 4 d.f. (air temperature) or a linear term (air pollutants) at lag 0–1, warm season defined as the 6 consecutive warmest months; “Adj. poll”: model adjusted for the specified air pollutant (at the specified lag), with a linear term; “Warm season as 3 months”: warm season defined as the 3 consecutive warmest months; “time trends 2.d.f./year”: time trend modelled with 2 d.f./year instead of 4; “time trends 6.d.f./year”: time trend modelled with 6 d.f./year instead of 4; “lag 0–3”: exposure modelled with a lag 0–3 term (instead of lag 0–1); “lag 0–10”: air temperature modelled with a lag 0–10 term (instead of lag 0-.

Overall, we found higher average associations between mean ambient air temperature and mortality on days with high air pollution concentrations (Fig. 2, Table S2 and Figure S3). Increments in mortality when temperature increased from the 75th to the 99th percentile ranged from 5.3 % (95 % CI: 3.8 %, 6.9 %) to 12.8 % (95 % CI: 8.7 %, 17.0 %) when daily mean PM10 concentrations were 10 and 90 μg/m3, respectively. Estimates of temperature-related mortality for concentrations of max-8 h O3 equal to 40 and 160 μg/m3 were 2.9 % (95 % CI: 1.1 %, 4.7 %) and 12.5 % (95 % CI: 6.9 %, 18.5 %). Similarly, increments in air temperature between 75th and 99th percentiles were associated to 3.9 % (95 % CI: 2.7 %, 5.1 %) and 12.3 % (95 % CI: 8.6 %, 16.1 %) increases in daily mortality when daily mean PM2.5 was equal to 1 or 40 μg/m3, and to 5.4 % (95 % CI: 1.9 %, 8.9 %) and 11.0 % (95 % CI: 8.4 %, 13.8 %) when daily mean NO2 was equal to 1 or 60 μg/m3. Estimates of association between air temperature and mortality increased, on average, steadily from lower to higher pollutants concentrations, were statistically different across levels of air pollutants (Fig. 2, Table S2 and Figure S3), and presented substantial differences across macro-regions, with more pronounced effect modification in European and Australian cities, and little to no effect modification in North American and South African cities (Figure S4).

Fig. 2.

Fig. 2.

Association between daily mean air temperature and all-cause mortality by levels of daily mean air pollutants: % change in mortality, and 95% confidence intervals, per increments of air temperature from the 75th to the 99th percentile of city-specific distributions, for different daily mean concentrations of the four pollutants. Results of the random-effects meta-analysis.

PM10, PM2.5 and O3 associations with mortality changed little with temperature until about the 80th percentile, but then increased, so that they were highest on hottest days (Fig. 3 and Table S3: mortality increased on average by 1.21 % (95 % CI: 0.69 %, 1.72 %), 1.11 % (95 % CI: 0.27 %, 1.95 %) and 0.53 % (95 % CI: 0.38 %, 0.68 %) per 10 μg/m3 increments in PM10, PM2.5 and O3, respectively, on days when air temperature was at its 99th percentile. Corresponding estimates on days at the 1st percentile of air temperature were 0.54 % (95 % CI: 0.10 %, 0.98 %), −0.41 % (95 % CI: −1.59 %, 0.79 %) and 0.00 % (95 % CI: −0.44 %, 0.44 %). Continent-specific estimates were largely heterogeneous, with a suggestion of a stronger effect modification in European and Australian cities (Figure S5). We found no clear effect modification of air temperature in the NO2-mortality association (Fig. 3 and Table S3).

Fig. 3.

Fig. 3.

Association between daily mean air pollutants and all-cause mortality by percentiles of daily mean air temperature: % change in mortality, and 95 % confidence intervals, per 10 μg/m3 increments of air pollutants, for different percentiles of air temperature city-specific distributions. Results of the random-effects meta-analysis.

Overall, we found strong evidence of interaction (on a multiplicative scale) between air temperature and each air pollutant: the p-values of the meta-analytical estimates of the product terms were: 3.6e-10 for temperature*PM10, 3.2e-05 for temperature*PM2.5, 1.6e-08 for temperature*O3, and 5.0e-04 for temperature*NO2 (reported in footnotes of Tables S2 and S3).

4. Discussion

To the best of our knowledge, this is the first epidemiological study reporting the joint effects of high air temperature and air pollution exposures, considering PM, NO2 and O3, on daily mortality in countries across all continents. We found evidence of higher heat-related mortality effects with higher levels of daily PM, NO2 and O3, as well as increased PM- and O3-related mortality for higher levels of mean air temperature during the warm months (but not for NO2). Results were largely heterogeneous across different geographical regions, mostly driven by results in the European and United States cities, and were robust when alternative model adjustments and definitions of the warm season were considered.

The acute effects of heat (Basu, 2009; Song et al., 2017) and air pollution (WHO, 2021; Vicedo-Cabrera et al., 2020; Liu et al., 2019; Orellano et al., 2020; Dominski et al., 2021; Meng et al., 2021) are well established and relatively consistent throughout the literature. Guo et al. investigated the association between non-optimal air temperature and all-cause mortality in 306 communities from 12 countries, and estimated increments in mortality, when temperature increased from optimal values to 99th location-specific percentile, ranging from 4 % in the United States and 30 % in Italy, consistent with the estimate we report in Table 2, despite the substantial differences in terms of data and methods between their study and ours (Guo et al., 2014). Previous analyses of the association between all-cause mortality and daily PM (Liu et al., 2019), O3; (Vicedo-Cabrera et al., 2020) and NO2 concentrations (Meng et al., 2021) using data from the MCC collaborative network also provided results very consistent with the ones presented in our Table 2, despite differences in the study locations, all-year versus warm season analysis, and other methodological choices.

However, given that people are simultaneously exposed to multiple environmental risk factors, such as air pollutants and extreme heat, it is important to expand the knowledge basis on the interactive effects of these exposures on health outcomes in order to define appropriate mitigation and response measures. The evidence of interactive effects of air pollution and temperatures has grown in recent years (Chen et al., 2017; Anenberg et al., 2020; Li et al., 2017; Rai et al., 2023; Chen et al., 2018; Analitis et al., 2014; Jhun et al., 2014; Ren et al., 2008; Scortichini et al., 2018; Shi et al., 2020). However, it is still inconclusive, with some areas of the world remaining unstudied and several studies focusing on one-way interactions and/or single pollutant investigation. A recent review on the joint effects of heat and air pollution reported that 19 of the 39 studies carried out in Europe, the United States, Canada, Russia, Taiwan, South Korea, India, Hong Kong, and China showed positive interactive effects on the human health outcomes studied, with the strongest evidence between heat and exposure to O3 and PM2.5 (Anenberg et al., 2020). Findings from our study can be compared with existing evidence in the literature on the positive interactive effects of heat and air pollution on mortality (Rai et al., 2023; Chen et al., 2018; Analitis et al., 2014; Jhun et al., 2014; Ren et al., 2008; Scortichini et al., 2018; Shi et al., 2020). A meta-analysis found a statistically significant modification of the acute effects of PM10 or O3 on total and cardiovascular disease mortality by temperature (Anenberg et al., 2020). Two multi-centre European studies found significant interactions between temperature and air pollution (considering both PM and O3) and comparable results for this region (Chen et al., 2018; Analitis et al., 2014). A review on the interaction between PM10 and air temperature found that most studies reported that temperature modifies the associations between PM and mortality and results on the interactive effect of PM and temperature seem to be robust (Li et al., 2017). On the other hand, results on the interactive effect of air temperature and O3 seem to be less consistent across regions, countries and cities, showing both positive and negative associations as well as no interaction (Anenberg et al., 2020; Li et al., 2017). A study conducted in 97 U.S. cities using the National Morbidity Mortality Air Pollution Study (NMMAPS) database found that the interaction between O3 and temperature was not statistically significant. However, there was a suggestive indication that high temperatures may exacerbate physiological responses to short-term O3 exposure (Jhun et al., 2014).

Personal exposure to ambient air pollutants and outdoor temperatures may be greater in warmer conditions because people tend to spend more time outdoors and open windows more often, especially in countries and periods with limited use of air conditioning (Li et al., 2017; Scortichini et al., 2018). Furthermore, it has been shown that the source, composition and oxidative potential of PM vary seasonally, and some research suggested that more toxic components of PM are higher during the summer season and in the presence of high temperatures (Zhang et al., 2010).

The physiological mechanisms underlying the synergistic association between temperature and air pollutants on mortality are not yet clearly defined; however, a few hypotheses have been proposed as they act on common pathways. High temperatures can increase thermoregulatory stress and alter the physiological response to toxicants, leading to a higher susceptibility to air pollution effects as the uptake, and distribution of air pollutants in the human body is enhanced by the increase in ventilation rate (Li et al., 2017; Gordon, 2003). Heat may also promote thrombosis through increasing blood viscosity and cholesterol levels secondary to dehydration and salt depletion (Bouchama et al., 2007). It has also been suggested that exposure to PM is associated with systemic and pulmonary inflammation and increased risk of coagulability by increasing blood levels of C-reactive protein and fibrinogen levels (Rückerl et al., 2011). O3 and NO2 also increase oxidative stress causing inflammation of the airways and increased permeability of the lung lining, thus impairing host defences against respiratory infections as well as fibrinolysis, thus reducing the efficiency of preventing clot formation and clearance (Anenberg et al., 2020; Li et al., 2017; Chen et al., 2018).

Several strengths should be acknowledged. Firstly, the study included 620 cities from 36 countries across the globe, with very diverse ambient air temperature and air pollution levels in the warm season, and applied common protocols for statistical analysis, representing the largest study on this topic to date to the best of our knowledge. This allowed us to compare results across locations by removing those sources of heterogeneity stemming from different study designs. Secondly, we applied flexible non-linear three-dimensional functions to estimate mortality increments corresponding to joint variations in air pollutants and high temperatures. This made the effect modification results (of air pollutants on temperature-related mortality and vice versa) comparable, as they were obtained from the same joint relationship. Thirdly, we included four key pollutants (PM10, PM2.5, NO2 and O3) in our analysis, each with a solid background of harmful short-term effects on mortality, and interactive effects with high ambient temperatures. Finally, the extensive sensitivity analyses supported our main results and provided evidence of the robustness of our findings.

The study also has some limitations. Despite the large number of cities included in the analysis, these are non-representative of the entire world population. In fact, there still are areas with limited coverage (the Middle East, Latin America, Australia) or no coverage at all (Northern and Central Africa, Northern Asia). Even within the most represented areas, some countries or regions contributed with data from a limited number of cities (for some countries, just one city), making the study representative of the 620 included cities, rather than the urban populations of the 36 represented countries. Further, rural populations are not represented. Future studies will focus on trying to address this issue by extending the collaborative network to these under-represented areas or countries with few cities, retrieving data, where available and possibly including mortality records from sub-urban and rural areas to investigate different population characteristics, activity patterns, built environment, and air pollution composition. A second limitation of the study is the ecological approach, which assumed constant exposure within the city on a given day, with exposure estimated using averages of limited sets of monitoring stations, which might not fully represent the study areas, and induce some exposure measurement error Unfortunately, we did not have information on the location of the deceased subjects within each city, nor exposure data at a spatial scale finer than the city itself. However, since the focus of the study was on day-to-day variability, and not on fine-scale spatial contrasts, we consider that as a minor limitation with negligible consequences on the overall interpretation, although we recognize that sub-scale heterogeneity in effects, as well as potential residual bias due to exposure measurement error, may exist. Finally, we only analysed natural-cause (or all-cause) mortality data and not cause-specific mortality. A recent review by Anenberg et al. looking at the interactive effect of air pollution and heat found consistent evidence for both total and cause-specific (cardio-respiratory) health outcomes, although the number of the latter studies was limited (Anenberg et al., 2020). We have recently filled this gap by analysing cause-specific mortality data available from the MCC collaborative network: we reported suggestive evidence of effect modification of air pollutants in the relationship between daily air temperature and cardiorespiratory mortality in 482 cities from 24 countries (Rai et al., 2023).

In conclusion, this multicentre study produced new and compelling evidence of the joint effects of high ambient temperatures and air pollution on daily mortality on the global scale. Climate change will increase both average and extreme temperatures (Romanello et al., 2021; IPCC. Climate Change, 2022), as well as indirectly impact air pollution levels by increasing the frequency of stagnation events, enhancing photochemical production of secondary pollutants and increasing “natural” gaseous and PM emissions influenced by warmer and drier conditions having a detrimental impact on human health. (IPCC. Climate Change, 2022; Chen et al., 2020; Vicedo-Cabrera et al., 2021) Public health interventions in response to climate change should consider the synergistic health effects of heat and air pollution focusing on adaptation actions for vulnerable subgroups and promoting mitigation measures that account for both exposures.

Supplementary Material

1

Acknowledgments

Massimo Stafoggia, Francesca K. de’ Donato, Masna Rai and Alexandra Schneider were partially supported by the European Union’s Horizon 2020 Project Exhaustion (Grant ID: 820655). Jan Kyselý and Aleš Urban were supported by the Czech Science Foundation project (22–24920S). Joana Madureira was supported by the Fundação para a Ciência e a Tecnologia (FCT) (grant SFRH/BPD/115112/2016). Masahiro Hashizume was supported by the Japan Science and Technology Agency (JST) as part of SICORP, Grant Number JPMJSC20E4. Noah Scovronick was supported by the NIEHS-funded HERCULES Center (P30ES019776). South African Data were provided by Statistics South Africa, which did not have any role in conducting the study. Antonio Gasparrini was supported by the Medical Research Council-UK (Grants ID: MR/V034162/1 and MR/R013349/1), the Natural Environment Research Council UK (Grant ID: NE/R009384/1), and the European Union’s Horizon 2020 Project Exhaustion (Grant ID: 820655).

Footnotes

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary material

Supplementary data to this article can be found online at https://doi.org/10.1016/j.envint.2023.108258.

Data availability

Data will be made available on request.

References

  1. Analitis A, Michelozzi P, D’Ippoliti D, de’Donato F, Menne B, Matthies F, Atkinson RW, Iñiguez C, Basagaña X, Schneider A, Lefranc A, Paldy A, Bisanti L, Katsouyanni K, 2014. Effects of heat waves on mortality: effect modification and confounding by air pollutants. Epidemiology 25 (1), 15–22. [DOI] [PubMed] [Google Scholar]
  2. Anderson BG, Bell ML, 2009. Weather-related mortality: how heat, cold, and heat waves affect mortality in the United States. Epidemiology 20, 205–213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Anenberg SC, Haines S, Wang E, Nassikas N, Kinney PL, 2020. Synergistic health effects of air pollution, temperature, and pollen exposure: a systematic review of epidemiological evidence. Environ. Health 19, 130. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Basu R, 2009. High ambient temperature and mortality: a review of epidemiologic studies from 2001 to 2008. Environ. Health 8, 40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bouchama A, Dehbi M, Mohamed G, Matthies F, Shoukri M, Menne B, 2007. Prognostic factors in heat wave related deaths: a meta-analysis. Arch. Intern. Med. 167, 2170–2176. [DOI] [PubMed] [Google Scholar]
  6. Chen F, Fan Z, Qiao Z, Cui Y, Zhang M, Zhao X, Li X, 2017. Does temperature modify the effect of PM10 on mortality? A systematic review and meta-analysis. Environ. Pollut. 224, 326–335. [DOI] [PubMed] [Google Scholar]
  7. Chen K, Wolf K, Breitner S, Gasparrini A, Stafoggia M, Samoli E, Andersen ZJ, Bero-Bedada G, Bellander T, Hennig F, Jacquemin B, Pekkanen J, Hampel R, Cyrys J, Peters A, Schneider A, 2018. Two-way effect modifications of air pollution and air temperature on total natural and cardiovascular mortality in eight European urban areas. Environ. Int. 116, 186–196. [DOI] [PubMed] [Google Scholar]
  8. Chen K, Vicedo-Cabrera AM, Dubrow R, 2020. Projections of ambient temperature- and air pollution-related mortality burden under combined climate change and population aging scenarios: a review. Curr. Environ. Heal Reports 7 (3), 243–255. [DOI] [PubMed] [Google Scholar]
  9. Dominski FH, Lorenzetti Branco JH, Buonanno G, Stabile L, Gameiro da Silva M, Andrade A, 2021. Effects of air pollution on health: A mapping review of systematic reviews and meta-analyses. Environ. Res. 201, 111487. [DOI] [PubMed] [Google Scholar]
  10. Gasparrini A, Guo Y, Hashizume M, Lavigne E, Zanobetti A, Schwartz J, Tobias A, Tong S, Rocklöv J, Forsberg B, Leone M, De Sario M, Bell ML, Guo Y-L, Wu C. f., Kan H, Yi S-M, de Sousa Zanotti Stagliorio Coelho M, Saldiva PHN, Honda Y, Kim H.o., Armstrong B, 2015. Mortality risk attributable to high and low ambient temperature: a multicountry observational study. Lancet 386 (9991), 369–375. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Gasparrini A, Guo Y, Hashizume M, Kinney PL, Petkova EP, Lavigne E, Zanobetti A, Schwartz JD, Tobias A, Leone M, Tong S, Honda Y, Kim H.o., Armstrong BG, 2015. Temporal variation in heat-mortality associations: a multicountry study. Environ. Health Perspect. 123 (11), 1200–1207. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Gordon CJ, 2003. Role of environmental stress in the physiological response to chemical toxicants. Env Res 92 (1), 1–7. [DOI] [PubMed] [Google Scholar]
  13. Guo Y, Gasparrini A, Armstrong B, Li S, Tawatsupa B, Tobias A, Lavigne E, de Sousa Zanotti Stagliorio Coelho M, Leone M, Pan X, Tong S, Tian L, Kim H.o., Hashizume M, Honda Y, Guo Y-L, Wu C-F, Punnasiri K, Yi S-M, Michelozzi P, Saldiva PHN, Williams G, 2014. Global variation in the effects of ambient temperature on mortality: a systematic evaluation. Epidemiology 25 (6), 781–789. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. IPCC. Climate Change 2022: Impacts, Adaptation and Vulnerability. Technical Summary. 2022. https://report.ipcc.ch/ar6wg2/pdf/IPCC_AR6_WGII_FinalDraft_TechnicalSummary.pdf. [Google Scholar]
  15. Jhun I, Fann N, Zanobetti A, Hubbell B, 2014. Effect modification of ozone-related mortality risks by temperature in 97 US cities. Environ. Int. 73, 128–134. [DOI] [PubMed] [Google Scholar]
  16. Kinney PL, 2018. Interactions of climate change, air pollution, and human health. Curr Environ Health Reports 5 (1), 179–186. [DOI] [PubMed] [Google Scholar]
  17. Lavigne E, Gasparrini A, Wang X, Chen H, Yagouti A, Fleury MD, Cakmak S, 2014. Extreme ambient temperatures and cardiorespiratory emergency room visits: assessing risk by comorbid health conditions in a time series study. Env Health 13 (1), 5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Li J, Woodward A, Hou X-Y, Zhu T, Zhang J, Brown H, Yang J, Qin R, Gao J, Gu S, Li J, Xu L, Liu X, Liu Q, 2017. Modification of the effects of air pollutants on mortality by temperature: A systematic review and meta-analysis. Sci Total Env 575, 1556–1570. [DOI] [PubMed] [Google Scholar]
  19. Liu C, Chen R, Sera F, Vicedo-Cabrera AM, Guo Y, Tong S, Coelho MSZS, Saldiva PHN, Lavigne E, Matus P, Valdes Ortega N, Osorio Garcia S, Pascal M, Stafoggia M, Scortichini M, Hashizume M, Honda Y, Hurtado-Díaz M, Cruz J, Nunes B, Teixeira JP, Kim H.o., Tobias A, Íñiguez C, Forsberg B, Åström C, Ragettli MS, Guo Y-L, Chen B-Y, Bell ML, Wright CY, Scovronick N, Garland RM, Milojevic A.i., Kyselý J, Urban A, Orru H, Indermitte E, Jaakkola JJK, Ryti NRI, Katsouyanni K, Analitis A, Zanobetti A, Schwartz J, Chen J, Wu T, Cohen A, Gasparrini A, Kan H, 2019. Ambient particulate air pollution and daily mortality in 652 cities. N. Engl. J. Med. 381 (8), 705–715. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Meng X, Liu C, Chen R, et al. Short term associations of ambient nitrogen dioxide with daily total, cardiovascular, and respiratory mortality: multilocation analysis in 398 cities. BMJ 2021; 372: n534. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Orellano P, Reynoso J, Quaranta N, Bardach A, Ciapponi A, 2020. Short-term exposure to particulate matter (PM10 and PM2.5), nitrogen dioxide (NO2), and ozone (O3) and all-cause and cause-specific mortality: systematic review and meta-analysis. Environ. Int. 142, 105876. [DOI] [PubMed] [Google Scholar]
  22. Rai M, Stafoggia M, de’Donato F, Scortichini M, Zafeiratou S, Vazquez Fernandez L, Zhang S, Katsouyanni K, Samoli E, Rao S, Lavigne E, Guo Y, Kan H, Osorio S, Kyselý J, Urban A, Orru H, Maasikmets M, Jaakkola JJK, Ryti N, Pascal M, Hashizume M, Fook Sheng Ng C, Alahmad B, Hurtado Diaz M, De la Cruz Valencia C, Nunes B, Madureira J, Scovronick N, Garland RM, Kim H.o., Lee W, Tobias A, Íñiguez C, Forsberg B, Åström C, Maria Vicedo-Cabrera A, Ragettli MS, Leon Guo Y-L, Pan S-C, Li S, Gasparrini A, Sera F, Masselot P, Schwartz J, Zanobetti A, Bell ML, Schneider A, Breitner S, 2023. Heat-related cardiorespiratory mortality: effect modification by air pollution across 482 cities from 24 countries. Environ. Int. 174, 107825. [DOI] [PubMed] [Google Scholar]
  23. Ren C, Williams GM, Morawska L, Mengersen K, Tong S, 2008. Ozone modifies associations between temperature and cardiovascular mortality: analysis of the NMMAPS data. Occup. Env. Med. 65 (4), 255–260. [DOI] [PubMed] [Google Scholar]
  24. Romanello M, McGushin A, Di Napoli C, Drummond P, Hughes N, Jamart L, Kennard H, Lampard P, Solano Rodriguez B, Arnell N, Ayeb-Karlsson S, Belesova K, Cai W, Campbell-Lendrum D, Capstick S, Chambers J, Chu L, Ciampi L, Dalin C, Dasandi N, Dasgupta S, Davies M, Dominguez-Salas P, Dubrow R, Ebi KL, Eckelman M, Ekins P, Escobar LE, Georgeson L, Grace D, Graham H, Gunther SH, Hartinger S, He K, Heaviside C, Hess J, Hsu S-C, Jankin S, Jimenez MP, Kelman I, Kiesewetter G, Kinney PL, Kjellstrom T, Kniveton D, Lee JKW, Lemke B, Liu Y, Liu Z, Lott M, Lowe R, Martinez-Urtaza J, Maslin M, McAllister L, McMichael C, Mi Z, Milner J, Minor K, Mohajeri N, Moradi-Lakeh M, Morrissey K, Munzert S, Murray KA, Neville T, Nilsson M, Obradovich N, Sewe MO, Oreszczyn T, Otto M, Owfi F, Pearman O, Pencheon D, Rabbaniha M, Robinson E, Rocklöv J, Salas RN, Semenza JC, Sherman J, Shi L, Springmann M, Tabatabaei M, Taylor J, Trinanes J, Shumake-Guillemot J, Vu B, Wagner F, Wilkinson P, Winning M, Yglesias M, Zhang S, Gong P, Montgomery H, Costello A, Hamilton I, 2021. The 2021 report of the Lancet Countdown on health and climate change: code red for a healthy future. Lancet 398 (10311), 1619–1662. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Rückerl R, Schneider A, Breitner S, Cyrys J, Peters A, 2011. Health effects of particulate air pollution: A review of epidemiological evidence. Inhal. Toxicol. 23 (10), 555–592. [DOI] [PubMed] [Google Scholar]
  26. Scortichini M, De Sario M, de’Donato F, Davoli M, Michelozzi P, Stafoggia M, 2018. Short-Term Effects of Heat on Mortality and Effect Modification by Air Pollution in 25 Italian Cities. Int. J. Environ. Res. Public Health 15 (8), 1771. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Sera F, Armstrong B, Tobias A, Vicedo-Cabrera AM, Åström C, Bell ML, Chen B-Y, de Sousa Zanotti Stagliorio Coelho M, Matus Correa P, Cruz JC, Dang TN, Hurtado-Diaz M, Do Van D, Forsberg B, Guo YL, Guo Y, Hashizume M, Honda Y, Iñiguez C, Jaakkola JJK, Kan H, Kim H.o., Lavigne E, Michelozzi P, Ortega NV, Osorio S, Pascal M, Ragettli MS, Ryti NRI, Saldiva PHN, Schwartz J, Scortichini M, Seposo X, Tong S, Zanobetti A, Gasparrini A, 2019. How urban characteristics affect vulnerability to heat and cold: a multi-country analysis. Int. J. Epidemiol. 48 (4), 1101–1112. [DOI] [PubMed] [Google Scholar]
  28. Shi W, Sun Q, Du P, Tang S, Chen C, Sun Z, Wang J, Li T, Shi X, 2020. Modification effects of temperature on the ozone-mortality relationship: a nationwide multicounty study in China. Environ. Sci. Tech. 54 (5), 2859–2868. [DOI] [PubMed] [Google Scholar]
  29. Song X, Wang S, Hu Y, Yue M, Zhang T, Liu Y.u., Tian J, Shang K, 2017. Impact of ambient temperature on morbidity and mortality: An overview of reviews. Sci. Total Environ. 586, 241–254. [DOI] [PubMed] [Google Scholar]
  30. Stafoggia M, Forastiere F, Faustini A, Biggeri A, Bisanti L, Cadum E, Cernigliaro A, Mallone S, Pandolfi P, Serinelli M, Tessari R, Vigotti MA, Perucci CA, 2010. Susceptibility factors to ozone-related mortality: A population-based case-crossover analysis. Am. J. Respir. Crit. Care Med. 182 (3), 376–384. [DOI] [PubMed] [Google Scholar]
  31. 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. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Vicedo-Cabrera AM, Scovronick N, Sera F, Royé D, Schneider R, Tobias A, Astrom C, Guo Y, Honda Y, Hondula DM, Abrutzky R, Tong S, Coelho M.d.S. Z.S., Saldiva PHN, Lavigne E, Correa PM, Ortega NV, Kan H, Osorio S, Kyselý J, Urban A, Orru H, Indermitte E, Jaakkola JJK, Ryti N, Pascal M, Schneider A, Katsouyanni K, Samoli E, Mayvaneh F, Entezari A, Goodman P, Zeka A, Michelozzi P, de’Donato F, Hashizume M, Alahmad B, Diaz MH, Valencia CDLC, Overcenco A, Houthuijs D, Ameling C, Rao S, Di Ruscio F, Carrasco-Escobar G, Seposo X, Silva S, Madureira J, Holobaca IH, Fratianni S, Acquaotta F, Kim H, Lee W, Iniguez C, Forsberg B, Ragettli MS, Guo YLL, Chen BY, Li S, Armstrong B, Aleman A, Zanobetti A, Schwartz J, Dang TN, Dung DV, Gillett N, Haines A, Mengel M, Huber V, Gasparrini A, 2021. The burden of heat-related mortality attributable to recent human-induced climate change. Nat. Clim. Chang. 11 (6), 492–500. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. WHO. WHO global air quality guidelines. Particulate matter (PM2.5 and PM10), ozone, nitrogen dioxide, sulfur dioxide and carbon monoxide. 2021; published online Oct. DOI: 10.1016/S0140-6736(06)69530-5. [DOI] [PubMed] [Google Scholar]
  34. Zhang Y, Bi P, Hiller JE, 2010. Climate variations and Salmonella infection in Australian subtropical and tropical regions. Sci. Total Environ. 408 (3), 524–530. [DOI] [PubMed] [Google Scholar]
  35. Zhao Q.i., Guo Y, Ye T, Gasparrini A, Tong S, Overcenco A, Urban A, Schneider A, Entezari A, Vicedo-Cabrera AM, Zanobetti A, Analitis A, Zeka A, Tobias A, Nunes B, Alahmad B, Armstrong B, Forsberg B, Pan S-C, Íñiguez C, Ameling C, De la Cruz Valencia C, Åström C, Houthuijs D, Dung DV, Royé D, Indermitte E, Lavigne E, Mayvaneh F, Acquaotta F, de’Donato F, Di Ruscio F, Sera F, Carrasco-Escobar G, Kan H, Orru H, Kim H.o., Holobaca I-H, Kyselý J, Madureira J, Schwartz J, Jaakkola JJK, Katsouyanni K, Hurtado Diaz M, Ragettli MS, Hashizume M, Pascal M, de Sousa Zanotti Stagliorio Coélho M, Valdés Ortega N, Ryti N, Scovronick N, Michelozzi P, Matus Correa P, Goodman P, Nascimento Saldiva PH, Abrutzky R, Osorio S, Rao S, Fratianni S, Dang TN, Colistro V, Huber V, Lee W, Seposo X, Honda Y, Guo YL, Bell ML, Li S, 2021. Global, regional, and national burden of mortality associated with non-optimal ambient temperatures from 2000 to 2019: a three-stage modelling study. Lancet Planet Health 5 (7), e415–e425. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1

Data Availability Statement

Data will be made available on request.

RESOURCES