Air pollution impact on the Covid-19 mortality in Iran considering the comorbidity (obesity, diabetes, and hypertension) correlations.

Since the rise of the Covid-19 pandemic, several researchers stated the possibility of a positive relationship between Covid-19 spread and climatic parameters. An ecological study in 12 Iranian cities using the report of daily deaths from Covid-19 (March to August 2020) and validated data on air pollutants, considering average concentrations in each city in the last year used to analyze the association between chronic exposure to air pollutants and the death rate from Covid-19 in Iran. Poisson regression models were used, with generalized additive models and adjustment variables. A significant increase of 2.7% (IC(95%) 2.6-4.4) was found in the mortality rate due to Covid-19 due to an increase of 1 μg/m3 of NO2. The results suggest an association between Covid-19 mortality and NO2 exposure. As a risk approximation associated with air pollution, more precise analysis is done. The results also show a good consistency with studies from other regions; this paper's results can be useful for the public health policymakers and decision-making to control the Covid-19 spread.

Introduction

On Dec 31, 2019, cases of pneumonia with unknown etiology were presented in the municipality of Wuhan, in the province of Hubei in China, and on Jan 9, 2020, the SARS-CoV-2 virus was identified as the causal agent of the Severe Acute Respiratory Syndrome (SARS) (Keni et al., 2020). This virus belonging to the coronavirus family causes a new infectious disease (Covid-19), with rhinorrhea, sore throat, cough, fever, diarrhea, anosmia, and myalgia. Its most severe form manifests as pneumonia with dyspnea, disseminated coagulation, and eventually death (Adams, 2020). The virus is transmitted by respiratory droplets when an infected person coughs, speaks, sneezes, or directly contacts people carrying the virus or contaminated surfaces (Bashir et al., 2020). Due to its potential for transmissibility, the SARS-CoV-2 virus managed to spread worldwide until reaching Iran, where the first case was confirmed on Mar 2, 2020; 180 days later, by Aug 31, 2020, the figure of 375,212 confirmed cases was reached (Iranian Statistics Organization). In Iran, the fatality rate for Covid-19 up to that date was 6.3%, almost 1.5 times higher than that registered in the rest of the world, of 4.2% with groups of greater vulnerability made up of people with some comorbidity such as hypertension, diabetes, cardiovascular diseases and diseases of the respiratory system (Bell et al., 2004), as well as the population over 60 years of age (Bontempi, 2020).

Several diverse pieces of research have been done to investigate essential parameters impacting the spread of SARS-CoV-2 disease to control the transmission of the COVID-19 pandemic. Many premier pieces of research have determined that person-to-person contact could enhance the danger of infection (Chen et al., 2020; Gilbert et al., 2020; Lu et al., 2020; Sohrabi et al., 2020; Kraemer et al., 2020). Besides these human-based factors mentioned by Kraemer (Xie and Zhu, 2020), Xie and Zhu (Ali et al., 2021) mentioned that the climate condition could also be a very important parameter in the Covid-19 spread. They mentioned that mean temperature increase was positively associated with newly confirmed Covid-19 cases. Also, in their results, they mentioned a 4.861% increase in the daily covid cases for each 1 °C rise in the mean temperature. Similarly, the World Health Organization (WHO) reported that in European and North American cities with high concentrations of nitrogen dioxide (NO2), there is a decrease in lung function (Hendryx and Luo, 2020). All these results opened a discussion to the field of climatic parameters such as air quality impacts on the Covid-19 pandemic aspects. Several significant papers studied association between air quality and covid-19 pandemic. Conticini et al. (2020) investigated the correlation between the high level of Severe Acute Respiratory Syndrome CoronaVirus 2 (SARS-CoV-2) lethality and the atmospheric pollution in Northern Italy. They concluded that the high level of pollution in Northern Italy should be considered an additional co-factor of the high level of lethality recorded in that area. Cole et al. (2020) estimated the relationship between long term air pollution exposure and Covid-19 in 355 municipalities in the Netherlands. Using detailed data they found compelling evidence of a positive relationship between air pollution, and particularly PM2.5 concentrations, and Covid-19 cases, hospital admissions and deaths. Setti et al. (2020) presented a position paper proposing a research hypothesis concerning the association between higher mortality rates due to COVID-19 observed in Northern Italy and average concentrations of PM10 exceeding a daily limit of 50 μg/m3. Barnett-Itzhaki and Levi (2021) examined the association between populations’ exposure to air pollution and the morbidity and mortality rates from COVID-19. According to their results and comparison with previously published studies, it appears that long-term exposure to air pollutants concentrations exceeding WHO guidelines, such as PM2.5 and NOx, might exacerbate morbidity and mortality rates from COVID-19. These results should raise a red flag globally among decision makers about the urgent need to reduce air pollution and its harmful effects. Jiang and Xu (2020) used the Pearson and Poisson regression models accordingly to understand the association between COVID-19 deaths and each risk factor. The results suggested that PM2.5 and diurnal temperature range are tightly associated with increased COVID-19 deaths. Also Stieb et al. (2020) observed a positive association between COVID-19 incidence and long-term PM2.5 exposure in Canadian health regions. Another study by Wu et al. (2020) also found that higher historical PM2.5 exposures are positively associated with higher county-level COVID-19 mortality rates after accounting for many area-level confounders. Yao et al., 2020a, 2020b explored the temporal correlation between the case fatality rate (CFR) of COVID-19 and particulate matter (PM) in Wuhan. They observed a higher CFR of COVID-19 with increasing concentrations of inhalable particulate matter (PM) with an aerodynamic diameter of 10 μm or less (PM10) and fine PM with an aerodynamic diameter of 2.5 μm or less (PM2.5) in the temporal scale. Zoran et al., 2020a, 2020b investigates the correlation between the high level of coronavirus SARS-CoV-2 infection accelerated transmission and lethality, and surface air pollution in Milan metropolitan area, Lombardy region in Italy. The results show positive correlation of daily averaged O3, PM2.5 and PM10 with air temperature and inversely correlations with relative humidity and precipitation rates. Nearly all these paper concluded a significant association between air pollution and Covid-19 pandemic transition or mortality rate. This conclusion is also considered by two important reviews done by Marquès et al. (Marquès and Domingo, 2021; Marquès et al., 2021) in which they state: “Taking together the previous results and those of most investigations now reviewed, we have concluded that there is a significant association between chronic exposure to various outdoor air pollutants: PM2.5, PM10, O3, NO2, SO2 and CO, and the incidence/risk of COVID-19 cases, as well as the severity/mortality of the disease” and with this statement concluded from a literature review of hundreds of papers in this field a firm confirmation of this theory is made in this papers.

Considering the previous studies have found a relationship between exposure to air pollutants and increased mortality and morbidity from Covid-19 (Coker et al., 2020). However, it is unknown whether these impacts are similar in the different regions of the world, making it necessary to carry out specific analyses by country and city (Iqair ranking (2021) and ava, 2021). Iranian cities have higher levels of pollutants than other cities, such as Europe and North America (Safari-Faramani et al., 2019). On the other hand, the Iranian population has high prevalences of comorbidities such as diabetes and hypertension that should be considered in studies on exposure to pollutants and Covid-19 (Contini and Costabile, 2020). The role that chronic exposure to air pollutants plays in increasing the risk of bacterial and respiratory infections is known. The chronic inflammation they produce, added to the immune response effect, explains this effect. During the influenza pandemic, it was observed that the most polluted cities had higher incidence rates (Deek, 2020). However, no large-scale and reliable study on this field was done in Iran; a limited number of researchers considered this issue. Ali & Islam (Ali and Islam, 2020) stated that available data also indicate that exposure to air pollution may influence COVID-19 transmission. Faridi et al. (2020) found a significant association between coronavirus disease (COVID-19) lethality and exposure to ambient air pollution, the rise in airborne PM2.5 during this outbreak may increase the mortality rate of SARS-CoV-2 in the city of Tehran. Khorsandi et al., 2021a, 2021b concluded that the short-term exposure to air pollution (especially O3) may increase the population's susceptibility infected with COVID-19 and, therefore, increase hospital admissions and mortality rates even during the warm seasons.

It is estimated that outdoor air pollution, mainly by fine particulate matter PM2.5 (particulate matter), leads to 3.3 million premature deaths per year globally (Domingo and Rovira, 2020). Furthermore, it is recognized that the potential of these particles to generate a toxic effect depends on factors such as frequency and duration of exposure, in addition to their composition (Fattorini and Regoli, 2020). Various epidemiological studies, systematic reviews, and meta-analyzes have associated chronic exposure to PM2.5 and PM10 with conditions such as ischemic heart disease (Frontera et al., 2020), cerebrovascular disease (Nouri et al., 2021), lower respiratory tract infections, chronic obstructive pulmonary disease, lung cancer (Cole et al., 2020), and type 2 diabetes in the case of PM2.5. In the same way, the acute effects of these particles have been reported (Khorsandi et al., 2021a; Iranian meteorological organization). As the review of the early literature shows, no comprehensive research is done for the Iranian case. This study is aimed to study the relationship between concentrations of three pollutants with major healthcare impacts (PM2.5, PM10, and NO2) and daily confirmed COVID-19 cases and mortality in 12 Iranian cities to supply useful implications for the control and prevention of this novel coronavirus disease. These cities were selected to cover all minor climate conditions of the country, which can be a main solidification point for the results of this paper with correcting the climate impact on the results, which is not considered in the previous researches.

Material and methods

An ecological study with retrospective analysis was carried out from Mar 1, 2019 to Aug 31, 2020, in 12 Iranian cities: Tehran, Tabriz, Isfahan, Shiraz, Mashhad, Ahvaz, Karaj, Yazd, Kermanshah, Bandar-Abbas, Zahedan, and Rasht. The inclusion criteria were that these cities had a continuous air quality monitoring system and sufficient and validated data to estimate exposure and confirmed mortality cases from SARS-CoV-2 in the period study. Also, one of the other main reasons was their diverse climate characteristics to validate the research samples. And the study period was only limited to access to reliable data (see Fig. 1 ).

Fig. 1

Location of studied cities in Iran's map.

Data collection, processing Air quality, the data on suspended particles (PM10 and PM2.5) and nitrogen dioxide (NO2) were obtained from the National Air Quality Information System (Iranian meteorological organization). Also, the results of monitoring stations in all cities were only included in the processing data after being validated under sufficiency (data of at least two of the pollutants be available) and consistency (data be gathered in at least two-thirds of each month) standards. After validating the gathered data, the data preparation process was implemented and prepared for further study stages.

The daily data of cases and the percentage of them with some comorbidity (obesity, diabetes, and hypertension) were obtained from the database of the official coronavirus site enabled by the Ministry of Health (Food and Drug association). Next, the cumulative confirmed cases and mortality rates per 100,000 inhabitants for the analysis period were estimated to be used in the data analysis.

The information on the prevalence of comorbidities in the general population was obtained from the National Survey of Nutrition and Health databases in the reports of the years 2020 and 2021 (Bajgain et al., 2021). The report's number of people classified with diabetes and hypertension was divided by the total number of respondents for each city for the prevalence of comorbidities. For the prevalence of obesity, the body mass index (BMI) was used, and the population was classified according to international standards, considering the value of BMI ≥30 as obesity.

The investigation resolution for all parameters is on a daily scale. However, the air quality and comorbidities investigation period is 365 days longer (545 days) than the Covid-19 data investigation period (180 days). The difference between observations of covid-19 cases and the other cases and data observations is due to studying the historical trend of other parameters. This extended investigation period shows the natural trend of air pollution in comorbidity cases, which by neutralizing this natural trend from the post-Covid-19 trend, more accurate results can be obtained.

This study consists of three main stages (as shown in Fig. 2 ). The first stage is data gathering, in which the needed data is gathered from accessible sources in the Iranian health system and Air quality control datasets. Then a data preparation and classification process are done. In this process, the data is made ready for processing and analysis. Also, age classification and then a Comorbidity classification is done in this stage. Four age categories are used in this paper, Children (Under 18), Adulthood (18–35), Middle Age (35–55), and Elderly (Above 55). Considering these categories, observations are first classified under each class, and then the comorbidity classification is implemented (with three classes of Hypertension, Diabetes, and Obesity) (Bajgain et al., 2021; Asadi and Norouzi, 2021). A correlative analysis is implemented on the results of these classifications to determine the impact of each parameter on the Covid-19 and air pollution impact on each of these parameters. This strategy estimates the indirect impact of the air quality on the Covid-19 mortality and infected cases (see Table 1 and Table 1A).

Fig. 2

Schematics of the methodology implemented in this study.

Table 1A

Correlative analysis of study variables

	X1	X2	X3	X4	X5	X6	X7	X8	X9	X10	X11	X12	X13	X14	X15	X16	X17	X18	X19	X20	X21	X22	X23	X24	X25	X26	X27	X28	X29	X30	X31	X32
X1	1
X2	0	1
X3	0	0	1
X4	0	0	0	1
X5	0.11	0.03	0.58	0.58	1
X6	0.13	0.19	0.89	0.89	1	1
X7	0.21	0.57	0.69	0.69	0.55	0.88	1
X8	0.19	0.54	0.92	0.92	0.74	0.76	0.84	1
X9	0.34	0.39	0.67	0.56	0.94	0.84	0.58	0.91	1
X10	0.92	0.78	0.79	0.79	0.45	0.67	0.55	0.82	0.91	1
X11	0.66	0.65	0.47	0.47	0.93	0.76	0.95	0.52	0.19	0.22	1
X12	0.58	0.4	0.78	0.78	0.75	0.17	0.13	0.06	0.78	0.07	0.21	1
X13	0.58	0.56	0.16	0.16	0.54	0.73	0.05	0.15	0.72	0.96	0.02	0.31	1
X14	0.62	0.83	0.7	0.7	0.49	0.55	0.43	0.36	0.31	0.83	0.5	0.81	0.47	1
X15	0.48	0.42	0.25	0.25	0.95	0.87	0.18	0.75	0.48	0.04	0.59	0.85	0.41	0.02	1
X16	0.41	0.71	0.84	0.84	0.61	0.13	0.35	0.45	0.85	0.38	0.82	0.23	0.07	0.12	0.62	1
X17	0.74	0.68	0.87	0.87	0.18	0.52	0.96	0.14	0.68	0.39	0.25	0.12	0.78	0.17	0.76	0.79	1
X18	0.91	0.51	0.98	0.98	0.48	0.02	0.99	0.5	0.89	0.54	0.71	0.86	0.39	0.08	0.16	0.68	0.71	1
X19	0.49	0.22	0.59	0.59	0	0.67	0.86	0.56	0.36	0.51	0.84	0.52	0.13	0.53	0.84	0.83	0.67	0.28	1
X20	0.63	0.9	0.03	0.03	0.77	0.76	0.22	0.06	0.88	0.47	0.43	0.26	0.42	0.26	0.57	0.58	0.88	0.76	0.46	1
X21	0.32	0.31	0.55	0.55	0.94	0.62	0.8	0.49	0.22	0.24	0.6	0.13	0.25	0.49	0.06	0.37	0.25	0.65	0.13	0.11	1
X22	0.78	0.36	0.03	0.03	0.59	0.97	0.1	0.34	0.26	0.6	0.75	0.68	0.58	0.48	0.38	0.72	0.39	0.81	0.19	0.77	0.41	1
X23	0.04	0.65	0.26	0.26	0.33	0.53	0.21	0.42	0.41	0.72	0.69	0.97	0.51	0.87	0.76	0.71	0.25	0.17	0.23	0.35	0.2	0.6	1
X24	0.7	0.86	0.88	0.88	0.22	0.1	0.12	0.28	0.79	0.16	0.35	0.62	0.13	0.78	0.91	0.08	0.47	0.77	0	0.34	0.24	0.77	0.34	1
X25	0	0	0.59	0.59	1	1	0.71	0.79	0.67	0.4	0.17	0.3	0.44	0.02	0.93	0.31	0.42	0.61	0.34	0.83	0.3	0.77	0.89	0.09	1
X26	0	0	0.93	0.93	1	1	0.28	0.58	0.53	0	0.85	0.78	0.25	0.11	0.45	0.84	0.62	0.6	0.51	0.05	0.51	0.48	0.03	0.69	0	1
X27	0	0	0.8	0.8	1	1	0.78	0.93	0.12	0.04	0.58	0.25	0.38	0.54	0.48	0.91	0.27	0.55	0.02	0.28	0.56	0.73	0.35	0.87	0	0	1
X28	0	0	0.41	0.41	1	1	0.59	0.21	0.29	0.69	0.63	0.71	0.68	0.18	0.99	0.22	0.96	0.65	0.01	0.74	0.7	0.83	0.73	0.59	0	0	0	1
X29	0	0	088	0.88	1	1	0.98	0.73	0.22	0.9	0.1	0.32	0.37	0.3	0.01	0.1	0.71	0.31	0.14	0.07	0.14	0.4	0.81	0.39	1	1	1	1	1
X30	0	0	0.48	0.48	1	1	0.59	0.02	0.18	0.72	0.11	0.1	0.67	0.86	0.06	0.59	0.44	0.85	0.56	0.8	0.28	0.45	0.96	0.51	1	1	1	1	0	1
X31	0	0	0.68	0.68	1	1	0.02	0.39	0.34	0.93	0.17	0.98	0.48	0.75	0.33	0.09	0.85	0.39	0.16	0.98	0.36	0.48	0.18	0.9	1	1	1	1	0	0	1
X32	0	0	0.71	0.71	1	1	0.12	0.47	0.31	0.18	0.77	0.39	0.25	0.83	0.68	0.66	0.55	0.9	0.58	0.31	0.73	0.11	0.29	0.08	1	1	1	1	0	0	0	1

Then in the third stage, an ecological analysis was carried out in three steps. First, the distribution, measures of central tendency, and correlation of the information in each city were estimated and observed. Then, using simple bivariate Poisson regression, the implicit risk of the association between mortality and air quality (PM and NO2) was independently explored. Finally, a Poisson multivariate association model was developed, with robust variance adjustment, including available adjustment covariates (prevalence of hypertension, diabetes, obesity, and days elapsed after the first case); the statistical model assumes that the data of the dependent variable Y (cases) follows a Poisson distribution of the form (Coker et al., 2020):where μ is the occurrence rate or the expected number of times, an event will occur in a period, and y is a random variable indicating the number of times an event occurred (cases). Sometimes the event will occur fewer times than the average rate or will not occur at all, and other times it will occur more frequently. The relationship between the expected count (μ) and the probability of examining any observed count (y) is specified by the Poisson distribution (Ali and Islam, 2020). The Wald test was used to measure the level of significance of each coefficient, and both in simple and multivariate regressions were with a significance level of 5% in the described model. The analysis was performed with Eviews software version 10.1.

Results and discussion

Analysis and results

During the period considered, the average of PM10 was 45.4 μg/m3, and that of PM2.5 was 19.5 μg/m3 in the 12 cities included; Kermanshah presented the minimum concentration of PM2.5 (7.5 μg/m3) and Tehran the maximum with 38.4 μg/m3. Concerning NO2, a minimum concentration of 7.9 μg/m3 were recorded in Zahedan, a maximum of 45.3 μg/m3 in Tehran, and an average of 26.03 μg/m3 (Table 1, Table 2 ).

Table 1

Descriptive analysis of the air pollution and comorbidity from Mar 1, 2019 to Aug 31, 2020 (for 12 mentioned cities).

	Variable	Observationsa,b	Mean	Std.Dev.	Min	Max
X1	PM10 (μg/m3)	6540	45.4	10.7	26.9	65.1
X2	PM2.5 (μg/m3)	6540	19.5	7.0	7.4	38.4
X3	NO2 (μg/m3)	6540	26.3	11.7	8.0	47.8
X4	NO2 (ppm)	6540	–	–	0.004	0.144
X5	Covid-19 Confirmed Cases	2160	1953.0	352.9	584.0	3574.0
X6	Covid-19 Death Cases	2160	143.0	11.7	126.0	179.0
X7	Hypertension (%)	6540	24.3	9.7	11.8	52.8
X8	Diabetes (%)	6540	13.2	5.1	5.1	24.1
X9	Obesity (%)	6540	51.0	10.8	42.9	63.2
X10	Hypertension in confirmed cases of Covid-19 (%)	2160	18.4	6.0	7.1	35.0
X11	Diabetes in Covid-19 cases (%)	2160	16.1	5.1	9.2	28.2
X12	Obesity in Covid-19 cases (%)	2160	27.3	5.9	17.8	37.5
X13	Elderly with diabetes (%)	6540	54.1	1.3	45.3	60.3
X14	Adults with diabetes (%)	6540	4.2	1.9	4.0	11.0
X15	Middle ages with diabetes (%)	6540	40.6	2.8	39.9	46.9
X16	Children with diabetes (%)	6540	1.1	0.7	1.0	1.8
X17	Elderly with obesity (%)	6540	33.2	1.9	24.3	38.7
X18	Adults with obesity (%)	6540	37.8	1.5	29.4	42.6
X19	Middle ages with obesity (%)	6540	45.2	2.8	37.1	48.6
X20	Children with obesity (%)	6540	19.1	1.9	7.3	21.2
X21	Elderly with hypertension (%)	6540	42.2	3.5	33.5	42.5
X22	Adults with hypertension (%)	6540	4.3	1.3	2.4	5.4
X23	Middle ages with hypertension (%)	6540	22.3	2.3	14.0	23.4
X24	Children with hypertension (%)	6540	1.1	0.9	0.6	3.4
X25	Elderly Covid cases (%)	2160	22.2	3.4	21.8	25.5
X26	Adults Covid cases (%)	2160	29.9	3.6	29.4	35.1
X27	Middle ages Covid cases (%)	2160	17.6	3.5	17.3	22.7
X28	Children Covid cases (%)	2160	4.8	1.1	4.4	7.4
X29	Elderly Covid mortality (%)	2160	13.1	2.8	12.5	16.2
X30	Adults Covid mortality (%)	2160	0.5	0.1	0.4	0.6
X31	Middle ages Covid mortality (%)	2160	1.4	0.5	0.9	1.9
X32	Children Covid mortality (%)	2160	0.4	0.3	0.3	1.0

Observations are reported based on the total data for each parameter. As an example, 2160 data in Covid-19 confirmed cases is equal to covid cases of 12 cities × 180 days of the investigation period (=2160) or 6540 data in hypertension is equal to hypertension cases in 12 cities × 545 days (a year before first Covid-19 case and start of covid-19 investigation period) of the investigation period (=6540).

The difference between observations of covid-19 cases and the other cases and data observations is due to studying the historical trend of other parameters. This extended investigation period shows the natural trend of air pollution in comorbidity cases, which by neutralizing this natural trend from the post-Covid-19 trend, more accurate results can be obtained.

Table 2

Mean concentration of air pollution and mortality rate by city, from Mar 1, 2019 to Aug 31, 2020

City	Population	PM2.5 (μg/m3)	PM10 (μg/m3)	NO2 (μg/m3)	Cases	Deaths	Mortality rate (per 100,000 inhabitants)	Mortality rate ratio
Tehran	9,135,000	38.4	36.7	45.3	73,080	5580	56.7	2.7
Tabriz	1,558,693	19.7	18.8	23.2	37402.3	2855.8	29.9	1.4
Isfahan	2,132,000	26.9	25.7	31.7	51,156	3906	39.7	1.9
Yazd	529,673	15.4	14.7	18.1	29,232	2232	22.7	1.1
Shiraz	1,565,572	19.2	18.4	22.7	36,540	2790	28.3	1.3
Mashhad	3,208,000	28.8	27.5	34	54,810	4185	42.5	2.2
Ahvaz	1,244,000	15.7	25.7	18.6	29962.8	2287.8	23.2	1.1
Karaj	1,592,492	20	19.1	23.6	38001.6	2901.6	29.5	1.4
Kermanshah	1,026,000	17.3	16.5	20.4	32,886	2511	25.5	1.2
Bandar-Abbas	526,648	12.7	12.1	14.9	24116.4	1841.4	18.7	0.9
Zahedan	610,000	13.4	12.8	15.9	25,578	1953	19.8	0.9
Rasht	713,000	18.4	8.1	21.7	35078.4	2678.4	27.2	1.3

In the cities considered, the estimated prevalence of hypertensive patients was 24.3% of the general population and 18.44% in the confirmed cases of Covid-19. The prevalence of diabetes was 13.17% in the general population and 16.12% in the cases of Covid-19. Likewise, the prevalence of obesity in the general population of these cities was 51.02% and 27.3% in the cases of Covid-19 (Table 1).

Tehran presented the highest number of confirmed cases (73,080) and deaths (5,580) from Covid-19. The lowest mortality and fatality rates were registered in Kermanshah in this same period, while the highest mortality rate was registered in Ahvaz. In that period, the national mortality rate ratio was 1.29. The mortality rate ratio shows that Isfahan, Karaj, Tehran, and Mashhad present a mortality rate higher than the national average (Table 2).

The statistical model results indicate an association between NO2 concentrations and an increase with the IRRNO2 mortality rate in which the incidence risk rate of NO2 (IRRNO2) impact on the Covid-19 mortality rate is estimated at 1.051 (IC(95%) 1.031–1.068). Furthermore, the comorbidities of diabetes and obesity were also significant. The results of PM2.5 indicate an incidence risk rate for mortality of 1.024; however, it is classified as not significant according to results (IC(95%) 0.992–1.045) (Table 3 ). Also, the PM10 was excluded from the models due to an insignificant direct association with the mortality rate from Covid-19.

Table 3

Mean relative risk for mortality from Covid-19 per unit of increase in the concentration of PM2.5 and NO2 during the period from Mar 1, 2019, to Aug 31, 2020,*.

Variable	Incidence risk rate (IRR) (%)	P-value	IC95%
			Min	Max
PM2.5 (μg/m3)	1.024 (1.5)	0.092	0.992	1.045
NO2 (μg/m3)	1.051 (2.7)	0.000	1.031	1.068
Days (from first case)	1.015 (1.4)	0.002	1.009	1.024
Hypertension (%)	0.893 (0.5)	0.669	0.872	1.011
Diabetes (%)	1.079 (4.2)	0.012	1.046	1.089
Obesity case (%)	1.066 (5.4)	0.017	1.019	1.098

*Poisson multivariate model, adjusted for two pollutants and comorbidities.

Discussion

In this study, it was found that there is an association between chronic concentrations of air pollutants and an increase in mortality rates from Covid-19. According to the results presented, NO2 concentrations are those that contribute most significantly to this association.

These results are similar to those found in other studies (Zoran et al., 2020b; Marquès and Domingo, 2021). In studies of chronic exposure and controlled by covariates, a significant association similar to that of the present study, between mortality due to Covid-19 and levels of NO2 were found; and reported an increase of 1 μg/m3 in the long-term average of NO2 is significantly associated with a 2.5% increase in mortality from Covid-19 (Zoran et al., 2020a). As in this study, these authors found that the increase in PM2.5 was not a significant predictor of Covid-19 mortality, although it was a significant predictor of the number of cases (12% increase in cases per 1 μg/m3 increase) (Yao et al., 2020b). Similarly, other researchers found a significant association between NO2 levels and fatality and mortality rates from Covid-19, as well as a marginal association between exposure to PM2.5 and the mortality rate; they report an increase of 11.3 and 16%, respectively, due to an increase in the interquartile range (IQR) of NO2 (∼4.6 ppb) (Wu et al., 2020; Yao et al., 2020a). These authors conclude that a long-term reduction of an IQR in NO2 levels would have prevented 14,672 deaths from positive Covid-19 cases and 44.7 deaths per million in the general population. Likewise, the association between exposure to NO2 and the mortality rate due to Covid-19 in the Iranian cities of the present study coincides with that reported by studies in Italy, Spain, France, and Germany, where they conclude that 83% of the fatal cases presented exposure to high concentrations of NO2 (Contini and Costabile, 2020; Deek, 2020; Ali and Islam, 2020). In the same sense, another study reviewing acute and chronic exposures shows that chronic exposure to NO2 and PM2.5 contributes to an increase in fatality from Covid-19, more significantly than other criteria air pollutants (Marquès et al., 2021).

Our results also indicate an increase (although not significant) in mortality from Covid-19 per unit increase in the concentration of PM2.5 and PM10. In other studies, this association was significant, and an 8% increase in the mortality rate for each μg/m3 of PM2.5 was found (Iqair ranking (2021) and ava, 2021). Other researchers in the same field found a significant association between mortality rate and PM2.5, but specifically due to diesel particulate matter (Conticini et al., 2020); likewise, an increase in the probability of dying from Covid-19 of 0.43% for each unit of increase in PM2.5, equivalent to a 3.3% increase in the mortality rate from each 1 μg/m3 of PM2.5 (Wu et al., 2020). Similarly, a significant association between fatality from Covid-19 and PM2.5 and PM10 particles concentrations was found in other studies (Setti et al., 2020; Fattorini and Regoli, 2020).

Regarding acute exposure studies, the author(s) concluded that NO2 is one of the key elements for developing Covid-19 and ozone and respirable particles (Domingo and Rovira, 2020). Finally, another study, which considered the levels of contamination of CO, NO2, SO2, O3, PM10, and PM2.5 in the air and the number of cases and deaths from Covid-19, found a statistically significant correlation between concentrations of CO, O3, and PM2.5 and the number of cases and deaths from Covid-19 (Khorsandi et al., 2021a). However, unlike the present work, it has the limitations of being a correlation study without adjusting covariates. A possible explanation for the variation in the results related to respirable particles between the studies is their composition, which can vary from city to city and modifies their toxic effects (Faridi et al., 2020). In the present analysis, NO2 was found to contribute significantly to the risk of mortality. NO2 is released into the environment due to incomplete combustion in vehicle engines, coal, oil, and natural gas combustion processes. As is known, exposure to this pollutant is related to an increased risk of mortality from respiratory and heart conditions (Momtazmanesh et al., 2020), mainly in highly populated cities with high industrial and vehicular activity (Deek, 2020; Ali and Islam, 2020). The toxicological basis that explains these effects is because chronic exposure to NO2 generates alterations in the respiratory and immune system since the airways and the inflammatory response are compromised, which facilitates the presence of complications in the event of infections by Covid-19 (Domingo and Rovira, 2020). Specifically, persistent exposure to NO2 can cause lung changes, including alveolar hyperplasia, increased fibrin, and accumulation of inflammatory cells in the alveoli. Within the pathophysiology of Covid-19, these factors contribute to the “Cytokine storm” phenomenon that generates severe complications of the disease. Furthermore, it is recognized that NO2 contributes to the decrease in oxygen saturation since, when combined with the hemoglobin, it produces Methemoglobin or Nitrosomyoglobin (Yao et al., 2020b).

In the multivariate model, the prevalence of obesity and diabetes was included as factors contributing to the risk of worsening Covid-19; both diseases were associated with increased mortality rates. Adjustment for comorbidities (hypertension, type 2 diabetes mellitus, and obesity) was explored in models with the prevalence of the general population and models with the prevalence of confirmed cases; the results of both versions of the analysis were similar in statistical significance and order of magnitude. Among the limitations in this study is the length of the time series that includes only the warm seasons in the country and the lack of adjustment for other confusing variables, such as the transmission rate and accessibility to medical services. Other factors that may influence this association include socio-economic variables, for example, in economic crises. On the time series, the averages of pollutants were obtained considering 18 months with daily measurements. This period is similar to other studies that used more than one year to estimate exposure; others used shorter periods (Domingo and Rovira, 2020). However, it is considered that the period contemplated is sufficient to have an estimate of the chronic exposure to the included pollutants, especially if it is taken into account that during the last five years, the concentrations have not changed significantly in the cities considered (Yao et al., 2020a). Regarding other unmeasured variables, it can be said that the situation was the same for all cities, so it is considered that in any case, the effect is underestimated. It is recommended that later studies include sociodemographic variables more specifically and conduct studies with better geographic resolution.

Also, the results show a direct impact of NO2 concentration in the atmosphere on the Covid-19 mortality rate. Results of age classification in Table 4 show that the age classes vulnerable to the lower air quality are at a higher risk of disease or even mortality with decreased air quality. Results showed that the Elderly (with 3.10%) and Children (with 2.90%) have the greatest vulnerability toward the air quality effect on the Covid-19 mortality rate, and Middle age and Adults showed the least vulnerability (Marquès and Domingo, 2021; Marquès et al., 2021).

Table 4

The direct and indirect impact of the air quality on the Covid-19 mortality rate in different age classes.

	Air Quality	Elderly		Middle Age		Adulthood		Children
		Direct	Indirect	Direct	Indirect	Direct	Indirect	Direct	Indirect
Covid-19 mortality rate	NO2	3.10%	2.04%	2.80%	1.85%	2.10%	1.42%	2.90%	1.97%
	PM2.5	insignificant	2.10%	insignificant	1.91%	insignificant	1.47%	insignificant	2.04%
	PM10	insignificant	1.92%	insignificant	1.75%	insignificant	1.35%	insignificant	1.88%

Conclusion

An ecological study in 12 Iranian cities is done using the report of daily deaths from Covid-19 (March to August 2020) and validated daily data on air pollutants, considering average daily concentrations in each city in the last year to study the possible association of the Covid-19 mortality rate and Air quality. A Poisson regression model with generalized additive methods and adjustment variables is used to analyze the possibility and intensity of this assumed relationship. This paper showed that the NO2 atmospheric concentration is significantly associated with the increased mortality rate of Covid-19, even adjusting for comorbidities. These results were aligned and confirmed by most of the similar studies in the literature, and it is observed that improved air quality could reduce the risk of mortality of Covid-19. Also, it is recommended that the socio-economic parameters (i.e., level of education, economic status, social class, etc.) be considered a classification parameter since it might impact the covid-19 mortality rate.

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.