key: cord-0993261-6bbatpqf
authors: Shao, Longyi; Cao, Yaxin; Jones, Tim; Santosh, M.; Silva, Luis F.O.; Ge, Shuoyi; da Boit, Kátia; Feng, Xiaolei; Zhang, Mengyuan; BéruBé, Kelly
title: COVID-19 mortality and exposure to airborne PM2.5: A lag time correlation
date: 2021-10-29
journal: Sci Total Environ
DOI: 10.1016/j.scitotenv.2021.151286
sha: cffc5d0901580b961c8af480eeb799dcd3a70830
doc_id: 993261
cord_uid: 6bbatpqf

COVID-19 has escalated into one of the most serious crises in the 21st Century. Given the rapid spread of SARS-CoV-2 and its high mortality rate, here we investigate the impact and relationship of airborne PM2.5 to COVID-19 mortality. Previous studies have indicated that PM2.5 has a positive relationship with the spread of COVID-19. To gain insights into the delayed effect of PM2.5 concentration (μgm−3) on mortality, we focused on the role of PM2.5 in Wuhan City in China and COVID-19 during the period December 27, 2019 to April 7, 2020. We also considered the possible impact of various meteorological factors such as temperature, precipitation, wind speed, atmospheric pressure and precipitation on pollutant levels. The results from the Pearson's correlation coefficient analyses reveal that the population exposed to higher levels of PM2.5 pollution are susceptible to COVID-19 mortality with a lag time of >18 days. By establishing a generalized additive model, the delayed effect of PM2.5 on the death toll of COVID-19 was verified. A negative correction was identified between temperature and number of COVID-19 deaths, whereas atmospheric pressure exhibits a positive correlation with deaths, both with a significant lag effect. The results from our study suggest that these epidemiological relationships may contribute to the understanding of the COVID-19 pandemic and provide insights for public health strategies.

Since December 2019, the coronavirus disease outbreak has resulted in a global health catastrophe caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). COVID-19 was first officially reported in Wuhan, but the source of the virus is unknown (Special Expert Group for

Medicine, 2020). The disease spread rapidly with much higher infection levels and speed than comparable epidemics such as Middle East Respiratory Syndrome (MERS) and Severe Acute Respiratory Syndrome (SARS) (Ganesh et al., 2021) , disrupting normal lifestyles and social frameworks (Candido et al., 2020) . The SARS-CoV-2 variant strains currently prevalent are Alpha, Beta, Gamma, Delta, Epsilon, Theta, Kappa, Lambda, and Mu, making the epidemic more serious (Chen, Wang, & Wei, 2021) . The mortality rates of varies in different regions based on the local conditions and medical facilities, with an average of 2.21% as of March 15, 2021 (https://covid19.who.int/). Factors affecting mortality are widely discussed, including air quality, meteorological conditions (Rahman et al., 2021) , travel habits (B. Wang et al., 2020) and social distance (Neto et al., 2021) . Among them, air quality and meteorological conditions are relatively uncontrollable factors. PM 2.5 is airborne particulate matter with an aerodynamic diameter of less than 2.5 μm (Sciences & Centre, 2012) , which can be inhaled into the distal regions of the lung, and has a protagonist role (Mehmood et al., 2021) , as well as the ability to carry into the respiratory system hazardous elements and organic and organometallic toxic substances; potentially resulting in lung and cardiovascular diseases (Kesic, Meyer, Bauer, & Jaspers, 2012; Mehmood et al., 2021; Xu et al., 2021) . It has been found that SARS-CoV-2 RNA exists in urban environmental PM samples (Kayalar et al., 2021) . Chronic exposure to certain air pollutants may lead to more severe and lethal COVID-19 outcomes (Domingo, Marques, & Rovira, 2020) . Numerous studies have now shown that the inhalation of particulate matter pollution has a strong correlation with the prevalence and mortality of COVID-19 (De Angelis et al., 2021;  SanJuan-Reyes, Gomez-Olivan, & Islas-Flores, 2021; Sasidharan, Singh, Torbaghan, & Parlikad, 2020) , and could even lead to increased susceptibility to the disease (Chakrabarty et al., 2021; Coccia, 2020; Milicevic et al., 2021) .

In an Italian case study, it was shown that the daily COVID-19 cases were directly related to the mobility habits of a person(s) 21 days prior to infection (Carteni, Francesco, & Martino, 2020) . It was also noted that a small increase in air pollution led to a large increase in COVID-19 infectivity and mortality in England (Travaglio et al., 2021) . A study in California found that PM 2.5 pollution caused by wildfires was correlated with mortality rates of COVID-19 (Meo et al., 2021) . However, most studies have not considered the lag effect of PM 2.5 on pollutants and human disease (Seposo, Ueda, Sugata, Yoshino, & Takami, 2020) , including COVID-19 (Choi, Peters, & Mueller, 2010; Daniele & Francesco, 2020) . The formula for the correlation coefficient was as follows:

Where and are the i-th sample values of variables X and Y respectively. ̅ , ̅ are the mean value of variables X and Y correspondingly. R is a dimensionless value with a range [-1, 1] . If the R value is positive the two variables are positively correlated, and negative R values show a negative correlation. The larger the |R| value (the closer to 1 or -1), the stronger the correlation between the two variables. If the R value is +1 or -1, it indicates that the two variables have a strict linear relationship (Choi et al., 2010) .

PCA (Principal Component Analysis) uses dimensionality reduction to transform multiple indicators into a few Principal components. By simplifying the data structure, the index load and variance contribution rate on the principal components are used to calculate the index weight, so as to achieve comprehensive evaluation (Eder, Bash, Foley, & Pleim, 2014; Groth, Hartmann, Klie, & Selbig, 2013) . PCA is widely used in the study of atmospheric pollutants and meteorological factors (Mor et al., 2021; Nguyen et al., 2021; Xiao et al., 

GAM (Generalized additive model) represents a method of fitting a smooth relationship between two or more variables. GAMs are useful when the relationship between the variables is expected to be of a complex form (Verbeke, 2007) . The GAM (Generalized additive model) analysis was performed by the 'mgcv' package (v 1.8-31) in R v4.0.3. The main advantage of this model is its flexibility to allow non-parametric fittings with relaxed assumptions on the actual association between response and predictor that provides the potential for better fitting to data than purely parametric models.

This model has been widely used in air pollution research and epidemiology (Manoj, Satheesh Kumar, Valsaraj, Sivan, & Vijayan, 2020; Rahman et al., J o u r n a l P r e -p r o o f 2021). The GAM is used in this study to analyze the additive relationship between PM 2.5 and meteorological factors COVID-19. The core GAM equation is:

( ( | 1 , 2 , ⋯ )) = 0 + 1 ( 1 ) + 2 ( 2 ) + ⋯ + ( ) Where, Y is the dependent variable and X is the independent variable ( ), = 1,2, ⋯ are smooth functions which are obtained by the backfitting algorithm. In this study, was set as PM 2.5 , temperature, air pressure, FAC1, FAC2 and FAC3 according to the linear correlation.

Wuhan is in the east of Hubei Province, China, at the intersection of the 

Since January 23 th , 2020, Wuhan City has officially counted and released the number of new daily deaths caused by the COVID-19 epidemic. As shown in Figure 1 , the deaths in Wuhan from January 23 th , 2020 to April 7 th , 2020, presented a quasi-normal distribution. According to the changing trends, three stages were identified, including (1) (trial version fifth) has added "clinical diagnosis cases" in the case diagnosis classification of Hubei (China & Medicine, 2020) , so that patients receive standardized treatment as early as possible. On February 12 th , 13436 new confirmed cases were reported in Wuhan, and the corresponding new deaths on February 12 th increased to 216.

The dominant wind direction in Wuhan throughout the year is from the north. Figure 2 shows the variation of several meteorological factors during COVID-19 in Wuhan. Table 1 

Although some studies have shown that the level of air pollutants was lower during COVID-19 lockdown compared with the period before the onset of showing an upward trend overall ( increased the proportion and risk of secondary aerosol PM 2.5 .

The formation efficiencies of secondary aerosols were enhanced during the lockdown due to the increase of atmospheric oxidation capacity, as also noted in another recent study (Tian et al., 2021) . As shown in Figure 3 , these pollutants do not display significant correlations with the daily death toll. temperature and air pressure are independent of each other. This also provides a basis for selecting variables in the next "GAM" analysis.

The principal component analysis of air pollutants and meteorological factors was carried out, and KMO and Bartlett spherical test were selected to confirm the applicability of the data. KOM is 0.638 and greater than 0.6. The p value of Bartlett's spherical test was 0.00, which was less than the significance test limit of 0.05. Therefore, the input data are suitable for principal component analysis. In this study, the maximum variance method is used to rotate the initial load matrix, and the factor with load value greater than 0.5 is selected as 

The generalized additive model is used to study the additive correlation and influence of various factors on the death toll of COVID-19. We used the PM 2.5 , temperature, and atmospheric pressure of lag 18 days to establish a generalized additive model. The family is set to "quasipoisson", and the smoothing parameter "k" is set to 3. The degree of freedom (EDF) of AP is 1 ( had no additive correlation with the number of deaths. FAC3 related to AP, T and SO 2 was negatively additive correlated with the number of deaths. We used the subgroup analysis method for sensitivity analysis. The analysis results show that (Table S4) , the sensitivity is low and the model is relatively stable. Exposure to ambient PM may also reduce the resistance to infection in the population (Maleki, Anvari, Hopke, Noorimotlagh, & Mirzaee, 2021 ). People exposed to higher levels of PM 2.5 pollution are more likely to suffer from cardiopulmonary diseases (Manojkumar & Srimuruganandam, 2021) ,

including IHD (Ischemic Heart Disease) (Nirel et al., 2021) , COPD (Chronic Obstructive Pulmonary Disease) (Guo et al., 2021) , lung cancer (Hvidtfeldt et al., 2021) , and strokes (Niu, Liu, Yu, Wu, & Xiang, 2021) . The adverse health outcomes are more significant for persons who have these pre-existing diseases Mahmood et al., 2021; W. T. Zhu et al., 2021) . A recent study suggested that long-term exposure to poor air quality may aggravate the clinical symptoms of COVID-19 (Al-Kindi et al., 2021) .

Airborne particles can act as the possible carriers of the SARS-CoV-2 virus into the human body, resulting in increased morbidity and mortality Maleki et al., 2021; Nguyen Thanh et al., 2021; Nor et al., 2021) . It has been revealed that the interaction of SARS-CoV-2 with PM is possible in moist environments. After drying, PM can serve as a carrier for transmission of SARS-CoV-2 immobilized on their surface (Borisova & Komisarenko, 2020).

The SARS-CoV-2 virus has been detected in hospitals, buses, subways and other environments (Hadei et al., 2021; Moreno et al., 2021; Nor et al., 2021;  J o u r n a l P r e -p r o o f Journal Pre-proof Yarahmadi et al., 2021) .

The increase of secondary aerosols may also lead to an increase in the death toll of COVID-19. Some studies have noted that during the lockdown, O 3 in Wuhan has increased by more than two times compared with the values before the lockdown (C. W. Lian et al., 2020) , which can boost atmospheric oxidizing capacity and further enhance the formation of secondary organic aerosols (X. Le et al., 2020; Meng et al., 2021 ) . An increase in the formation efficiency of secondary aerosols represented by nitrate and secondary OA was observed in Wuhan and many other areas during the COVID-19 lockdown Tian et al., 2021;  Z. L. Zheng et al., 2020) . It has been found that secondary aerosols can carry toxic bacteria (Jiang, Xia, & Liu, 2021) and are more harmful to humans than primary aerosols (Lin et al., 2016) . Therefore, the increase of secondary aerosol may be an important factor in PM 2.5 role in the mortality of COVID-19.

In addition, PM 2.5 can upregulate ACE-2 (Du et al., 2020) , the receptor of the SARS-CoV-2 virus (Baildya, Ghosh, & Chattopadhyay, 2021) , and increase the chance of viral RNA entering cells (Nguyen Thanh et al., 2021) . Thus, the potential role of fine particles in the transmission of COVID-19 is of increased importance. The findings in this study support the urgent need to implement environmental mitigation strategies for reducing airborne particulate pollution.

There is a lag time between the inhalation of PM 2.5 and the onset of adverse respiratory responses (Dong, Wang, Wang, & Bao, 2021; X. L. Zhu et al., 2021) . A Mexican study, at an individual level, did not find robust evidence for short-term PM 2.5 exposure increasing the chances of dying from COVID-19 (Lopez-Feldman, Heres, & Marquez-Padilla, 2021) . This supports findings that COVID-19 mortality seems to be driven mainly by longer-term (i.e., lagging, chronic or cumulative) rather than the short-term (i.e., acute) factors. Our results also suggest that the COVID-19 epidemic was established in Wuhan before any mobility restrictions were implemented.

The influence of meteorological factors has also been widely studied. The longevity of SARS-CoV-2 outside hosts decreases at high temperature and under sunlight (Yap, Liu, Shveda, & Preston, 2020) . Respiratory-related mortality will increase with decreasing temperatures (Dadbakhsh, Khanjani, Bahrampour, & Haghighi, 2017) . In a cold environment, the susceptibility of the host may be higher due to slower mucociliary clearance or decreased immune function under these conditions (Ficetola & Rubolini, 2021; Lowen & Steel, 2014 (Crema, 2021) .

Although the lag time correlation between the COVID-19 mortality and exposure to airborne PM 2.5 has been proposed, the limitation of this study still exists. Firstly, we have not considered impacts from personal behaviors like social distancing and personal hygiene (Bang et al., 2021; Magnan, Gibson, & Bryan, 2021) . Secondly, the current study only considered the Wuhan city, lack of research in the neighboring regions which might have a trans-regional influence for the COVID-19 transmissions. Finally, the model needs to be tested in other high incidence areas in the world.

In conclusion, the concentration of PM 2.5 is related to the risk of contracting COVID-19, and its delayed effect on the mortality of COVID-19 is J o u r n a l P r e -p r o o f identified to be more than 18 days. COVID-19 mortality in Wuhan was driven mainly by the longer-term rather than the short-term factors. Relevant administration departments and policy makers should consider that the incubation period of the SARS-CoV-2 is longer than 18 days, which is even longer than the public think of 14 days. The temperature has a strong negative correlation with COVID-19 deaths in Wuhan, while atmospheric pressure has a positive correlation.

Source code, with full documentation and examples, are freely available under the GNU General Public License on the WeirauchLab GitHub page:

https://github.com/Fossette-x/COVID-19-and-PM2.5-in-Wuhan.git Air rainfall -0.564* -0.161 -0.258 -0.253* Significance codes: 0.01 "**", 0.05 "*". The related sig. value (two tails), covariance and standard deviation are in Table S1 . 1.00 5.7*10-5 *** Significance codes: 0 "***", 0.001 "**", 0.01 "*". a ：EDF represents the estimated degree of freedom. Higher EDF values represent more complex splines. b

Evolution and epidemic spread of SARS-CoV-2 in Brazil

Multiple relationships between aerosol and COVID-19: A framework for global studies

How mobility habits influenced the spread of the COVID-19 pandemic: Results from the Italian case study

Ambient PM2.5 exposure and rapid spread of COVID-19 in the United States

Review of the mechanisms of SARS-CoV-2 evolution and transmission

Air quality characteristics in Wuhan (China) during the 2020 COVID-19 pandemic

Lower-than-standard particulate matter air pollution reduced life expectancy in Hong Kong: A time-series analysis of 8.5 million years of life lost

The New Coronavirus pneumonia diagnosis and treatment plan (trial version fifth

Correlational analysis of ordinal data: from Pearson's r to Bayesian polychoric correlation

Factors determining the diffusion of COVID-19 and suggested strategy to prevent future accelerated viral infectivity similar to COVID

The SARS-COV-2 outbreak around the Amazon rainforest: The relevance of the airborne transmission

Death from respiratory diseases and temperature in

Role of the chronic air pollution levels in the COVID-19 outbreak risk in Italy

COVID-19 incidence and mortality in Lombardy, Italy: An ecological study on the role of air pollution, meteorological factors, demographic and socioeconomic variables

Influence of airborne transmission of SARS-CoV-2 on COVID-19 pandemic. A review

Association between atmospheric PM 2.5 and daily outpatient visits for children's respiratory diseases in Lanzhou

Ambient fine particulate matter induced the elevation of blood pressure through ACE2/Ang (1-7) pathway: The evidence from urine metabolites

Incorporating principal component analysis into air quality model evaluation

Containment measures limit environmental effects on COVID-19 early outbreak dynamics

Epidemiology and pathobiology of SARS-CoV-2 (COVID-19) in comparison with SARS, MERS: An updated overview of current knowledge and future perspectives

Principal components analysis

Determining the effects of socioeconomic and environmental determinants on chronic obstructive pulmonary disease (COPD) mortality using geographically and temporally weighted regression model across Xi'an during 2014-2016

Presence of SARS-CoV-2 in the air of public places and transportation

Study on the variation of air pollutant concentration and its formation mechanism during the COVID-19 period in Wuhan

Enhanced secondary pollution offset during COVID-19 lockdown in China

Econometric analysis of COVID-19 cases, deaths, and meteorological factors in

Theoretical analysis for bacteria participating in atmospheric nucleation

Existence of SARS-CoV-2 RNA on ambient particulate matter samples: A nationwide study in Turkey

Exposure to Ozone Modulates Human Airway Protease/Antiprotease Balance Contributing to Increased Influenza an Infection

Independent association of meteorological characteristics with initial spread of COVID-19 in India

Unexpected air pollution with marked emission reductions during the COVID-19 outbreak in China

Impact of city lockdown on the air quality of COVID-19-hit of Wuhan city

Wintertime haze deterioration in Beijing by industrial pollution deduced from trace metal fingerprints and enhanced health risk by heavy metals

Substantial decreases of light absorption, concentrations and relative contributions of fossil fuel to light-absorbing carbonaceous aerosols attributed to the COVID-19 lockdown in east China

Impact of meteorological conditions and air pollution on COVID-19 pandemic transmission in Italy

Air pollution exposure and COVID-19: A look at mortality in Mexico City using individual-level data

Roles of Humidity and Temperature in Shaping Influenza Seasonality

Cognitive and Affective Risk Beliefs and their Association with Protective Health Behavior in Response to the Novel Health Threat of COVID-19

Evaluation of the susceptibility and fatality of lung cancer patients towards the COVID-19 infection: A systemic approach through analyzing the ACE2, CXCL10 and their co-expressed genes

An updated systematic review on the association between atmospheric particulate matter pollution and prevalence of SARS-CoV-2

Potential link between compromised air quality and transmission of the novel corona virus (SARS-CoV-2) in affected areas

Health benefits of achieving fine particulate matter standards in India -A nationwide assessment

Spatiotemporal variability of COVID-19 pandemic in relation to air pollution, climate and socioeconomic factors in Pakistan

Enhanced photochemical formation of secondary organic aerosols during the COVID-19 lockdown in Northern China

Effect of environmental pollutants PM 2.5 , carbon monoxide, and ozone on the incidence and mortality of SARS-COV-2 infection in ten wildfire affected counties in California

PM 2.5 as a major predictor of COVID-19 basic reproduction number in the USA

Impact of COVID-19 lockdown on air quality in Chandigarh, India: Understanding the emission sources during controlled anthropogenic activities

Tracing surface and airborne SARS-CoV-2 RNA inside public buses and subway trains

Particulate matter and SARS-CoV-2: A possible model of COVID-19 transmission

Impact of COVID-19 partial lockdown on PM 2.5 , SO 2 , NO 2 , O 3 , and trace elements in PM 2.5 in Hanoi

Concentration-response relationships between hourly particulate matter and ischemic events: A case-crossover analysis of effect modification by season and air-mass origin

Association between exposure to ambient air pollution and hospital admission, incidence, and mortality of stroke: an updated systematic review and meta-analysis of more than 23 million participants

How air quality and COVID-19 transmission change under different lockdown scenarios? A case from Dhaka city

COVID-19 in the environment

A vulnerability-based approach to human-mobility reduction for countering COVID-19 transmission in London while considering local air quality

Ambient air quality standards

Short-term effects of air pollution on daily single-and co-morbidity cardiorespiratory outpatient visits

Meteorological factors, COVID-19 cases, and deaths in top 10 most affected countries: an econometric investigation

Special Expert Group for Control of the Epidemic of Novel C, Pneumonia of the Chinese Preventive Medicine A. 2020. An update on the epidemiological characteristics of novel coronavirus pneumonia (COVID-19)

COVID-19 pandemic in Wuhan: Ambient air quality and the relationships between criteria air pollutants and meteorological variables before, during, and after lockdown

A chemical cocktail during the COVID-19 outbreak in Beijing, China: Insights from six-year aerosol particle composition measurements during the Chinese New Year holiday

Impacts of primary emissions and secondary aerosol formation on air pollution in an urban area of China during the COVID-19 lockdown

Links between air pollution and COVID-19 in England

Generalized additive models: an introduction with R

Airborne particulate matter, population mobility and COVID-19: a multi-city study in China

Nonlinear impact of COVID-19 on pollutions -Evidence from Wuhan

Weakened Aerosol-PBL Interaction During COVID-19 Lockdown in Northern China

Characteristics and Sources of Metals in TSP and PM 2.5 in an Urban Forest Park at

Diverse bacterial populations of PM 2.5 in urban and suburb Shanghai

Association of particulate matter pollution and case fatality rate of COVID-19 in 49 Chinese cities

A predictive model of the temperature-dependent inactivation of coronaviruses

Occupational exposure of health care personnel to SARS-CoV-2 particles in the intensive care unit of Tehran hospital

The association between ambient temperature and mortality of the coronavirus disease 2019 (COVID-19) in Wuhan, China: a time-series analysis

Long-term exposure to fine particulate matter relates with incident myocardial infarction (MI) risks and post-MI mortality: A metaanalysis

Acute effects of personal exposure to fine particulate matter on salivary and urinary biomarkers of inflammation and oxidative stress in healthy adults