key: cord-344008-h4kc04w0
authors: Liang, Donghai; Shi, Liuhua; Zhao, Jingxuan; Liu, Pengfei; Sarnat, Jeremy A.; Gao, Song; Schwartz, Joel; Liu, Yang; Ebelt, Stefanie T.; Scovronick, Noah; Chang, Howard H.
title: Urban Air Pollution May Enhance COVID-19 Case-Fatality and Mortality Rates in the United States
date: 2020-09-21
journal: Innovation (N Y)
DOI: 10.1016/j.xinn.2020.100047
sha: 
doc_id: 344008
cord_uid: h4kc04w0

Background The novel human coronavirus disease 2019 (COVID-19) pandemic has claimed more than 600,000 lives worldwide, causing tremendous public health, social, and economic damages. While the risk factors of COVID-19 are still under investigation, environmental factors, such as urban air pollution, may play an important role in increasing population susceptibility to COVID-19 pathogenesis. Methods We conducted a cross-sectional nationwide study using zero-inflated negative binomial models to estimate the association between long-term (2010-2016) county-level exposures to NO2, PM2.5 and O3 and county-level COVID-19 case-fatality and mortality rates in the US. We used both single and multipollutant models and controlled for spatial trends and a comprehensive set of potential confounders, including state-level test positive rate, county-level healthcare capacity, phase-of-epidemic, population mobility, population density, sociodemographics, socioeconomic status, race and ethnicity, behavioral risk factors, and meteorology. Results 3,659,828 COVID-19 cases and 138,552 deaths were reported in 3,076 US counties from January 22, 2020 to July 17, 2020, with an overall observed case-fatality rate of 3.8%. County-level average NO2 concentrations were positively associated with both COVID-19 case-fatality rate and mortality rate in single-, bi-, and tri-pollutant models. When adjusted for co-pollutants, per inter-quartile range (IQR) increase in NO2 (4.6 ppb), COVID-19 case-fatality rate and mortality rate were associated with an increase of 11.3% (95% CI 4.9% to 18.2%) and 16.2% (95% CI 8.7% to 24.0%), respectively. We did not observe significant associations between COVID-19 case-fatality rate and long-term exposure to PM2.5 or O3, although per IQR increase in PM2.5 (2.6 ug/m3) was marginally associated with 14.9% (95% CI: 0.0% to 31.9%) increase in COVID-19 mortality rate when adjusted for co-pollutants. Discussion Long-term exposure to NO2, which largely arises from urban combustion sources such as traffic, may enhance susceptibility to severe COVID-19 outcomes, independent of long-term PM2.5 and O3 exposure. The results support targeted public health actions to protect residents from COVID-19 in heavily polluted regions with historically high NO2 levels. Continuation of current efforts to lower traffic emissions and ambient air pollution may be an important component of reducing population-level risk of COVID-19 case-fatality and mortality.

The novel human coronavirus disease 2019 (COVID- 19) is an emerging infectious disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) 1 . First identified in 2019 in Wuhan, the capital of Hubei Province, China, the COVID-19 pandemic has since rapidly spread throughout the world. As of July 17, 2020, there have been 3,659,828 cases and 138,552 deaths confirmed in the United States [2] [3] [4] . Despite substantial public health efforts, the observed COVID-19 case-fatality rate (i.e. the ratio of the number of COVID-19 deaths over the number of cases) in the US is estimated to be 3.8% [2] [3] [4] . Although knowledge concerning the etiology of COVID-19-related disease has grown since the outbreak was first identified, there is still considerable uncertainty concerning its pathogenesis, as well as factors contributing to heterogeneity in disease severity around the globe.

Environmental factors [5] [6] [7] [8] , such as urban air pollution, may play an important role in increasing susceptibility to severe outcomes of COVID-19. The impact of ambient air pollution on excess morbidity and mortality has been well-established over several decades [9] [10] [11] . In particular, major ubiquitous ambient air pollutants, including fine particulate matter (PM 2.5 ), nitrogen dioxide (NO 2 ), and ozone (O 3 ), may have both a direct and indirect systemic impact on the human body by enhancing oxidative stress, inflammation, and respiratory infection risk, eventually leading to respiratory, cardiovascular, and immune system dysfunction and deterioration [12] [13] [14] [15] [16] .

While the epidemiologic evidence is limited, previous findings on the outbreak of severe acute respiratory syndrome (SARS), the most closely related human coronavirus disease to COVID-19, revealed a crude positive correlation between air pollution and the SARS case-fatality rate in the Chinese population without adjustment for confounders 17 . An analysis of 213 cities in China recently demonstrated that temporal increases in COVID-19 cases were associated with short-term variations in ambient air pollution 18 . Hence, it is plausible that prolonged exposure to air pollution may have a detrimental effect on the prognosis of patients affected by COVID-19 19 . As is usual in the early literature on emerging hazards, questions remain concerning the generalizability and reproducibility of these finding, due to the lack of control for the epidemic stage-of-disease, population mobility, residual spatial correlation, and potential confounding by co-pollutants.

To address these analytical gaps and contribute towards a more complete understanding of the impact of long-term exposures to ambient air pollution on COVID-19-related health consequences, we conducted a nationwide study in the USA (3,122 counties) examining associations between multiple key ambient air pollutants, NO 2 , PM 2.5 , and O 3 , and COVID-19 case-fatality and mortality rates in both single and multi-pollutant models, with J o u r n a l P r e -p r o o f comprehensive covariate adjustment. We hypothesized that residents living in counties with higher long-term ambient air pollution levels may be more susceptible to COVID-19 severe outcomes, thus resulting in higher COVID-19 case-fatality rates and mortality rates.

We obtained the number of daily county-level COVID-19 confirmed cases and deaths that occurred from January 22, 2020, the day of first confirmed case in the US, through July 17, 2020 in the US from three databases: the New York Times 2 , USAFACTS 3 , and 1Point3Acres.com 4 . Briefly, data on county-level COVID-19 cases and deaths data was confirmed by referencing state and local health agencies directly. COVID-19 confirmed case counts include both laboratory confirmed cases and presumptive positive cases (i.e. cases diagnosed by doctors based on signs and symptoms without a test), which is in line with how the US Center for Disease Control (CDC) reports data. In these databases, cases were assigned to where the person was diagnosed as that information became available. If a state reports both location of death and the location of residency, the case was attributed to the location of residency 3 .

COVID-19 death counts include both confirmed (i.e. by meeting confirmatory laboratory evidence for COVID -19) and probable deaths (i.e. meeting clinical criteria AND epidemiologic evidence with no confirmatory laboratory testing performed for COVID-19; meeting presumptive laboratory evidence AND either clinical criteria OR epidemiologic evidence; meeting vital records criteria with no confirmatory laboratory testing performed for COVID19). After data acquisition from these sources, we compared the number of confirmed COVID-19 cases and deaths in each US county (identified by the Federal Information Processing Standards, FIPS code) across all databases for accuracy and consistency. In case of discrepancy, county-level case and death number were corrected by manually checking the data reported from the corresponding state and local health department websites.

In this analysis, the main COVID-19 death outcomes included two measures, the county-level COVID-19 case-fatality rate and mortality rate. The COVID-19 case-fatality rate was calculated by dividing the number of deaths over the number of people diagnosed in each US county with at least 1 confirmed case, which can imply the biological susceptibility towards severe COVID-19 outcomes (i.e. death). The COVID-19 mortality rate was the number of COVID-19 deaths per million population, and can reflect the severity of the COVID-19 outcomes in the general population.

J o u r n a l P r e -p r o o f Three major criteria ambient air pollutants were included in the analysis, including NO 2 , a traffic-related air pollutant and a major component of urban smog, PM 2.5 , a heterogeneous mixture of fine particles in the air, and O 3 , a common secondary air pollutant 20 . We recently estimated daily ambient NO 2 , PM 2.5 , and O 3 levels at 1 km 2 spatial resolution across the Contiguous US using an ensemble machine learning model 21-23 . We calculated the daily average for each county based on all covered 1 km 2 grid cells (i.e., we calculated the arithmetic mean of daily air pollutant concentrations at 1 km 2 grid cells whose centroids fall within the boundary of that county). We then further calculated the annual mean (2010-2016) for NO 2 and PM 2.5 , and the warm-season mean (2010-2016) for O 3 , defined as May 1 to October 31, which is a standard time window to examine the association between ozone and mortality 23 .

Although more recent exposure data were not available, county-specific mean concentrations of air pollutants across years are highly correlated 20 .

We fit zero-inflated negative binomial mixed models (ZINB) to estimate the associations between long-term exposure to NO 2 , PM 2.5 , and O 3 and COVID-19 case-fatality rates and mortality rates. The ZINB model comprises a negative binomial log-linear count model and a logit model for predicting excess zeros 24 . The former was used to describe the associations between air pollutants and COVID-19 case-fatality rate among counties with at least one reported COVID-19 case. The latter can account for excess zeros in counties that have not observed a COVID-19 death as of July 17, 2020. We fit single-pollutant, bi-pollutant, and tri-pollutant models, in order to estimate the effects of each pollutant without and with control for co-pollutants. All analyses were conducted at the county level.

For the negative binomial count component, results are presented as percent change in case-fatality rate or mortality rate per interquartile range (IQR) increase in each air pollutant concentration. IQRs were calculated based on mean air pollutant levels across all 3,122 counties. Similar results are presented as odds ratios for the excess zero component. We included a random intercept for each state because observations within the same state tended to be correlated, potentially due to similar COVID-19 responses, quarantine and testing policies, healthcare capacity, sociodemographic, and meteorological conditions. As different testing practices may bias outcome ascertainment, we adjusted for state-level COVID-19 test positive rate 4 (i.e. a high positive rate could imply that the confirmed case numbers were limited by the ability of testing, thus upward-biasing the case-fatality). To model how different counties may be at different time points of J o u r n a l P r e -p r o o f the epidemic curve (i.e., phase-of-epidemic), we adjusted for days both since the first case and since the 100 th case within a county through July 17. In addition, we adjusted for potential confounders and covariates that may also contribute to heterogeneity in the observed COVID-19 rates and thus may confound associations with long-term air pollution exposure. These include county-level healthcare capacity, population mobility, population density, sociodemographics, socioeconomic status (SES), race and ethnicity, behavior risk factors, and meteorological factors. Specifically, healthcare capacity was measured by the number of intensive care unit (ICU) beds, hospital beds, and active medical doctors per 1000 people 25 . Population travel mobility index, based on anonymized location data from smartphones, was used to account for changes in travel distance in reaction to the COVID-19 pandemic 26 .

Socioeconomic status was measured by social deprivation index 27 , a commonly used measure of area-level SES, composed of income, education, employment, housing, household characteristics, transportation, and demographics.

Sociodemographic covariates included population density, percentage of elderly (age ≥ 60), and percentage of male.

Race and ethnicity included percentage of Black and percentage of Hispanic in each county. We also obtained behavioral risk factors including population mean body mass index (BMI) and smoking rate, and meteorological variables 28 including air temperature and relative humidity. Additional information about these covariates, including data sources, are given in the Technical Appendix.

To control for potential residual spatial trends and confounding, we included spatial smoothers within the model using natural cubic splines with 5 degrees freedom for both county centroid latitude and longitude. To examine the presence of spatial autocorrelation in the residuals, we calculated Moran's I of the standardized residuals of tri-pollutant main models among counties within each state. Statistical tests were 2-sided and statistical significance and confidence intervals were calculated with an alpha of 0.05. All statistical analyses were conducted in R version 3.4.

We conducted a series of 66 sets of sensitivity analyses to test the robustness of our results to outliers, confounding adjustment, and epidemic timing (Figures 4 and 5) . Given that New York City has far higher COVID-19 cases and deaths than any other region, we excluded all five counties within New York City in one sensitivity analysis. In another, we restricted the study to the most recent 4 weeks (June 20 to July 17), when the case count and death count may be more reliable and accurate compared to earlier periods. We also conducted sensitivity analysis by using air J o u r n a l P r e -p r o o f pollution data averaged between 2000 to 2016. To assess the importance of individual confounders or covariates, we fit models by omitting a different set of covariates for each model iteration and compared effect estimates.

A total of 3,122 US counties were considered in the current analysis, with confirmed cases reported in 3,076 (98.5%) and deaths in 2,088 (66.9%). By July 17, 2020, 3,659,828 COVID-19 cases and 138,552 deaths were reported nationwide ( Table 1) . Among the counties with at least one reported COVID-19 case, the average county-level casefatality rate was 2.4 ± 3.2% (mean ± standard deviation), and the average mortality rate was 298. We observed significant positive associations between NO 2 levels and both county-level COVID-19 casefatality rate and mortality rate ( Table 2 and Figure 3 ), when controlling for covariates. In tri-pollutant models, COVID-19 case-fatality and mortality rates were associated with increases of 11.3% (95% CI 4.9% to 18.2%) and 16 .2% (95% CI 8.7% to 24.0%), respectively, per IQR (~4.6 ppb) increases in NO 2 ( Table 2 ). These results imply that one IQR reduction in long-term exposure to NO 2 level would have avoided 14,672 deaths (95% CI: 6,721 to 22,143) among those who tested positive for the virus and 44.7 deaths (95% CI: 20.5 to 67.5) per million people in the general population, as of July 17, 2020. The strength and magnitude of the associations between NO 2 and both COVID-19 case-fatality rate and mortality rate persisted across single, bi-, and tri-pollutant models ( Figure 3 ).

In contrast, PM 2.5 was not associated with COVID-19 case-fatality rate (95% CI: -6.9% to 20.0%) but was marginally associated with higher COVID-19 mortality rate in tri-pollutant models, where one IQR (2.6 ug/m 3 ) J o u r n a l P r e -p r o o f increase in PM 2.5 was associated with 14.9% (95% CI: 0.0% to 31.9%) increase in COVID-19 mortality rate (Table   2 ). Null associations were found between long-term exposure to O 3 and both COVID-19 case-fatality and mortality rates (95% CI: -8.6% to 4.2% and -8.9% to 5.1%, respectively). Similar trends persisted across single, bi-, and tripollutant models ( Figure 3 ). The Moran's I and p-values (Appendix Table S1 ) from these models suggested that most spatial correlation in the data has been accounted for.

Results remained robust and consistent across 66 sets of sensitivity analyses (Figures 4 and 5 ). When we restricted the analyses to data reported between June 20 to July 17, when COVID-19 tests were more readily available, significant associations remained between NO 2 and COVID-19 case-fatality and mortality rates, and no consistent associations were found with PM 2.5 or O 3 . We also observed similar trends pointing to associations with NO 2 when excluding New York City. In addition, we found similar results when omitting the 679 counties (21.7%) with missing behavioral risk data.

In this nationwide study, we used county-level information on long-term air pollution and corresponding health, behavioral, and demographic data to examine associations between long-term exposures to key ambient air pollutants and COVID-19 death outcomes in both single and multi-pollutant models. We observed significant positive associations between NO 2 levels and both county-level COVID-19 case-fatality rate and mortality rate, a marginal association between long-term PM 2.5 exposure and COVID-19 mortality rate, and null associations for long-term O 3 exposures in multipollutant models. These results provide additional initial support for the interpretation that long-term exposure to air pollution, especially NO 2 -a component of urban air pollution related to traffic -may enhance susceptibility to severe COVID-19 outcomes. These findings may help identify susceptible and high-risk populations, especially those living in areas with historically high NO 2 pollution, including the metropolitan areas in New York, New Jersey, California, and Arizona. Given the rapid escalation of COVID-19 spread and associated mortality in the US, swift and coordinated public health actions, including strengthened enforcement on social distancing and expanding healthcare capacity, are needed to protect these and other vulnerable populations. Although average NO 2 concentrations have decreased gradually over the past decades, it is critical to continue enforcing air pollution regulations to protect public health, given that health effects occur even at very low concentrations 29 .

Currently, there are few existing studies investigating the link between air pollution and COVID-19, the majority of which are correlation-only studies without adjustment for confounders. Among these sparse studies, our findings are consistent with a recent European study that reported 78% of the COVID-19 deaths across 66 administrative regions in Italy, Spain, France and Germany, occurred in the five most polluted regions with the highest NO 2 levels 30 . Another recent paper reported correlations between high levels of air pollution and high death rates seen in northern Italy 31 . However, major questions remain concerning the robustness and generalizability of these early findings, due to the lack of control for population mobility, multipollutant exposures, and most importantly, potential residual spatial autocorrelation.

The current analysis addresses several of these limitations. We examined two major COVID-19 death outcomes, the county-level case-fatality rate and the mortality rate. The case-fatality rate can indicate biological susceptibility to severe COVID-19 outcomes (i.e. death), while the mortality rate can offer information on the severity of COVID-19 deaths in the general population. Our study also included an assessment of three major air pollutants using high spatial resolution maps, uses recent county-level data, considers both single and multipollutant models, and controls for county-level mobility. Given that the stage of the COVID-19 epidemic might depend on the size and urbanicity of the county, we included the time of the first and 100 th case for each county in the models as covariates to minimize the possibility that the observed associations are confounded by epidemic timing due to unmeasured location and population-level characteristics. Due to the cross-sectional design, we controlled for potential spatial trends by including flexible spatial trends in the main analysis, and evaluated residual autocorrelation using Moran's I statistic. Our analyses indicated that the presence of spatial confounding was substantial, necessitating the use of spatial smoothing (Figures 4 and 5) . We performed both stratified analyses and effect modification analyses by adding interaction terms in the model to examine the effects of potential confounders including socioeconomic status. None of these potential modifying effects were significant and were not included in the final analytical modeling approach. Finally, we conducted a total of 66 sets of sensitivity analyses and observed robust and consistent results.

Although social distancing measures around the US have reduced vehicle traffic and urban air pollution in the short-term, it is plausible that long-term exposure to urban air pollutants like NO 2 may have sustained direct and indirect effects within the human body, making people more biologically susceptible to severe COVID-19 outcomes.

J o u r n a l P r e -p r o o f NO 2 can be emitted directly from combustion sources or produced from the titration of NO with O 3 . NO 2 and nitric oxide (NO) have relatively short atmospheric lifetime, thus having larger spatial heterogeneity compared to more regionally distributed pollutants such as PM 2.5 and O 3 . As a result, the spatial distribution of NO 2 represents the intensity of anthropogenic activity, especially emissions from traffic and power plants. As a reactive free radical, NO 2 plays a key role in photochemical reactions that produce other secondary pollutants, including ozone and secondary particulate matter. In our analysis of three major air pollutants, however, NO 2 showed strong and independent effects with COVID-19 case-fatality rate and mortality, meaning that the effects of NO 2 may not be mediated by PM 2.5 and O 3 . Even so, we cannot rule out the possibility that NO 2 is serving as a proxy for other traffic-related air pollutants, such as soot, trace metals, or ultrafine particles. Long-term exposures to NO 2 have been associated with acute and chronic respiratory diseases, including increased bronchial hyperresponsiveness, decreased lung function, and increased risk of respiratory infection and mortality 32-34 . In addition, as a highly reactive exogeneous oxidant, NO 2 can induce inflammation and enhance oxidative stress, generating reactive oxygen and nitrogen species, which may eventually deteriorate the cardiovascular and immune systems 13, 35 . The impact of long-term exposure to PM 2.5 on excess morbidity and mortality has also been well-established [9] [10] [11] 29 . An early unpublished report that explored the impacts of air pollution on mortality found that 1 μg/m 3 PM 2.5 was associated with 8% increase in COVID-19 mortality rates in the USA 36 . The study was conducted in a single pollutant model and did not investigate on COVID-19 case-fatality rates. Similarly, we found marginally significant associations between COVID-19 mortality rates and PM 2.5 , when controlling for co-pollutants and covariates, although the magnitude and strength of this association observed in the current analysis were weaker, mainly due to our control of the spatial trends, co-pollutants, and residual autocorrelation, which may have confounded the previous study findings. In addition, PM 2.5 was not associated with COVID-19 case-fatality rate across all single and multipollutant models, indicating that it may have less impact on biological susceptibility to severe COVID-19 outcomes compared to NO 2 .

We acknowledge that our study is limited in several key areas. First, the cross-sectional study design reduced our ability to exploit temporal variation and trends in COVID-19 deaths, an important determinant in establishing causal inference. However, an ecological (area-level) analysis may offer valuable information as part of initial public health investigations for hypothesis generation, particularly where individual-level studies may not be possible for some time until fine-scale exposure data become available. Towards this end, future time-series J o u r n a l P r e -p r o o f analyses of air pollution and COVID-19 case-fatality rates and corresponding mortality rates will be important.

Second, there may be complex case ascertainment biases in the county-level COVID-19 data, particularly during the early stages of the outbreak due to lack of reliable testing, which may greatly underestimate the actual COVID-19 case number. While the case data quality gradually improved over the past two months due to enhanced testing capacity, we repeated the analysis using the COVID data reported at several time points (by April 1, by May 1, by June 2, and by July 17), and the results still hold. Third, actual death counts are likely biased, with highly dynamic reported fatality rates, increasing from 1.8% to 5.8%, then decreasing to 3.8% in the past three months 2-4 . However, results using data from only the most recent four weeks were largely unchanged, suggesting that differential errors in reporting or testing for COVID-19 may not have exerted much influence on these findings. In this analysis, the air pollution levels were modelled between 2010-2016, which may introduce bias in the exposure assessment.

Specifically, given that the average air pollution levels in US have gradually decreased over the years, when using the exposure data between 2010-2016 rather than more recent data, we may have over-estimated the exposure levels, leading to under-estimated health effects. Although more recent exposure data were not available, county-specific mean concentrations of air pollutants across years are highly correlated 20 . Moreover, we have conducted a sensitivity analysis by using the exposure data between 2000-2016 and the results remained robust and consistent. In addition, although we controlled for many potential confounders such as population density, we cannot rule out the possibility that NO 2 might be a proxy of urbanicity. The exclusion of climate meteorological variables and SES -two factors that have received substantial attention regarding the outbreak -did not alter the main results. Due to the lack of county-level data, we could not account for the percentage of hospitalized cases or ICU use among cases or deaths, the number of available ventilators, and the underlying health conditions of cases likely to increase death risk (e.g., chronic obstructive pulmonary disease). Also, as a classic traffic related air pollutant, NO 2 can exhibit spatial variation within a county 20 , which may not be captured in our analysis. Identification of NO 2 pollution hotspots within a county may be warranted.

We found statistically significant, positive associations between long-term exposure to NO 2 and COVID-19 casefatality rate and mortality rate, independent of PM 2. As different testing practices may bias outcome ascertainment, we adjusted for state-level COVID-19 test positive rate (i.e. high positive rate might imply that the confirmed case numbers were limited by the ability of testing, and the case-fatality can be biased high). To model how different counties may be at different time points of the epidemic curve (i.e., phase-of-epidemic), we adjusted for days both since the first case and since the 100 th case (i.e., case counts reaching 100) within a county through July 17 as a measure of epidemic timing. In addition, we considered potential confounding by county-level healthcare capacity, population travel mobility index, sociodemographic, SES, behavior risk factors, and meteorological factors. Because county-specific population densities span 5 orders of magnitude, we adjusted for density using a logarithmic transformation. To control for potential residual spatial trends and confounding, we included spatial smoothers within the model using natural cubic splines with 5 degrees freedom for both county centroid latitude and longitude. We further calculated Moran's I of the standardized residuals of tri-pollutant main models for each state, to examine the presence of spatial autocorrelation in the residuals.

We also conducted a series of sensitivity analyses to test the robustness of our results to outliers, confounding adjustment, and epidemic timing (Figures 4 and 5) . Given that New York city has far higher COVID-19 cases and deaths than any other regions in the US, which can be a very influential observation, we excluded all five counties within New York city and repeated the analysis. In another set of sensitivity analyses, we restricted the study only to the most recent 4 weeks (June 20 to July 17), when the case count and death count may be more reliable and accurate than earlier periods and when COVID-19 tests were more available. We also conducted sensitivity analysis by using air pollution data averaged between 2000 to 2016. To assess the impact of potential bias of individual covariates, we fit models by omitting a different set of covariates for each model iteration while comparing effect estimates. Statistical tests were 2-sidedand statistical significance was determined with an alpha of 0.05. All statistical analyses were conducted used R version 3.4. • Long-term exposures to urban air pollutants, especially NO 2 , may enhance population susceptibility to severe COVID-19 death outcomes.

• Reduction in urban air pollution exposures would have avoided over 14,000 deaths among those who tested positive for the virus as of July 17, 2020.

• Public health actions to protect populations from COVID-19 should include considerations for areas with historically high NO 2 exposures along with other behavioral and clinical risk factors. The novel human coronavirus disease 2019 (COVID-19) pandemic has claimed more than 900,000 lives worldwide, causing tremendous public health, social, and economic damages. While the risk factors of COVID-19 are still under investigation, environmental factors, such as urban air pollution, may play an important role in increasing population susceptibility to COVID-19 pathogenesis. Major ubiquitous ambient air pollutants, including fine particulate matter (PM 2.5 ), nitrogen dioxide (NO 2 ), and ozone (O 3 ), may have both a direct and indirect systemic impact on the human body by enhancing oxidative stress, inflammation, and respiratory infection risk, eventually leading to respiratory, cardiovascular, and immune system dysfunction and deterioration. Although the epidemiologic evidence is limited, a few early crude correlation analyses conducted in China and Europe suggest a potential link between air pollution and COVID-19 outcomes. However, questions remain concerning the generalizability and validity of these finding, due to their lack of control for the epidemic timing, population mobility, residual spatial correlation, and potential confounding by co-pollutants.

To address these analytical gaps, we conducted a nationwide study in the USA (3,122 counties) examining associations between multiple key ambient air pollutants, including NO 2 , PM 2.5 , and O 3 , and COVID-19 case-fatality and mortality rates in both single and multi-pollutant models, with comprehensive covariate adjustment. We hypothesized that residents living in counties with higher long-term ambient air pollution levels may be more susceptible to COVID-19 severe outcomes, thus resulting in higher COVID-19 case-fatality rates and mortality rates.

To start this analysis, we obtained the number of daily county-level COVID-19 confirmed cases and deaths that occurred from January 22, 2020 through July 17, 2020 in the contiguous US from three independent databases. The main COVID-19 death outcomes included two measures, the case-fatality rate, which was calculated by dividing the number of deaths over the number of people diagnosed, and the other outcome is the mortality rate, which was calculated as the number of COVID-19 deaths per million population. We include three major criteria pollutants in the analysis, including NO 2 , PM 2.5 , and O 3. The high-resolution exposure datasets (daily, 1 km 2 resolution) were generated by our collaborators at Harvard using an ensemble-learning method, which integrated three machine learners and a variety of predictor variables. Based on these daily predictions from 2010-2016, we then further calculated the annual mean for NO 2 and PM 2.5 , and the warm-season mean for O 3 between May to October.

Case-fatality rate and characteristics of patients dying in relation to COVID-19 in Italy

The New York Times

Social and environmental risk factors in the emergence of infectious diseases. Nature medicine

Environmental factors on the SARS epidemic: air temperature, passage of time and multiplicative effect of hospital infection

Social, behavioural and environmental factors and their impact on infectious disease outbreaks

An imperative need for research on the role of environmental factors in transmission of novel coronavirus (COVID-19)

Ambient particulate air pollution and daily mortality in 652 cities

Estimates of the global burden of ambient PM 2.5, ozone, and NO2 on asthma incidence and emergency room visits

Ambient air pollution and emergency department visits for asthma: a multi-city assessment of effect modification by age

Air pollution and respiratory viral infection. Inhalation toxicology

Effects of nitrogen dioxide on the expression of intercellular adhesion molecule-1, neutrophil adhesion, and cytotoxicity: studies in human bronchial epithelial cells

Use of high-resolution metabolomics for the identification of metabolic signals associated with traffic-related air pollution

Perturbations of the arginine metabolome following exposures to traffic-related air pollution in a panel of commuters with and without asthma

Acute pulmonary and inflammatory response in young adults following a scripted car commute

Air pollution and case fatality of SARS in the People's Republic of China: an ecologic study

Association between short-term exposure to air pollution and COVID-19 infection: Evidence from China

Does Air Pollution Influence COVID-19 Outbreaks? In: Multidisciplinary Digital Publishing Institute

Health Effect Institute Panel on the Health Effects of Traffic-Related Air Pollution. Traffic-related air pollution: a critical review of the literature on emissions, exposure, and health effects. Health Effects Institute

The New York Times

An ensemble-based model of PM2.5 concentration across the contiguous United States with high spatiotemporal resolution

Assessing NO2 Concentration and Model Uncertainty with High Spatiotemporal Resolution across the Contiguous United States Using Ensemble Model Averaging. Environmental science & technology

Area Health Resources Files

Mobility Changes in Response to COVID-19

Mapping county-level mobility pattern changes in the United States in response to COVID-19

Social Deprivation Index

Continental-scale water and energy flux analysis and validation for the North American Land Data Assimilation System project phase 2 (NLDAS-2): 1. Intercomparison and application of model products

When adjusted for co-pollutants and confounding factors, per inter-quartile range (IQR) increase in NO 2 (4.6 ppb), COVID-19 case-fatality rate and mortality rate were associated with an increase of 11.3% (95% CI: 4.9% to 18.2%) and 16.2% (95% CI: 8.7% to 24.0%), respectively. In other words, per IQR reduction in long-term exposure to NO 2 level (4.6 ppb) would have avoided 14,672 deaths (95% CI: 6,721 to 22,143) among those who tested positive for the virus and 44.7 deaths (95% CI: 20.5 to 67.5) per million people in the US general population, as of

which largely arises from urban combustion sources such as traffic, may enhance susceptibility to severe COVID-19 death outcomes, independent of long-term PM 2.5 and O 3 exposure. Public health actions to protect populations from COVID-19 should include considerations for areas with historically high NO 2 exposures along with other behavioral and clinical risk factors. Continuation and expansion of current efforts to lower traffic emissions and ambient air pollution may be an important component of reducing population-level risk of COVID-19 case-fatality and mortality

* Descriptive statistics was conducted on 3,122 US counties using data reported as of July 17, 2020 **COVID-19 case fatality rate was calculated by the number of deaths divided by the number of cases, reported as of July 17, 2020