key: cord-0737271-sh3r084f
authors: Bhaskar, A.; Chandra, J.; Braun, D.; Cellini, J.; Dominici, F.
title: Air pollution, SARS-CoV-2 transmission, and COVID-19 outcomes: A state-of-the-science review of a rapidly evolving research area
date: 2020-08-20
journal: nan
DOI: 10.1101/2020.08.16.20175901
sha: b78f7b9c9ca9db98edfcc832ef120a166c94429c
doc_id: 737271
cord_uid: sh3r084f

Background: As the coronavirus pandemic rages on, 692,000 (August 7, 2020) human lives and counting have been lost worldwide to COVID-19. Understanding the relationship between short- and long-term exposure to air pollution and adverse COVID-19 health outcomes is crucial for developing solutions to this global crisis. Objectives: To conduct a scoping review of epidemiologic research on the link between short- and long-term exposure to air pollution and COVID-19 health outcomes. Method: We searched PubMed, Web of Science, Embase, Cochrane, MedRxiv, and BioRxiv for preliminary epidemiological studies of the association between air pollution and COVID-19 health outcomes. 28 papers were finally selected after applying our inclusion/exclusion criteria; we categorized these studies as long-term studies, short-term time-series studies, or short-term cross-sectional studies. One study included both short-term time-series and a cross-sectional study design. Results: 27 studies of the 28 reported evidence of statistically significant positive associations between air pollutant exposure and adverse COVID-19 health outcomes; 11 of 12 long-term studies and all 16 short-term studies reported statistically significant positive associations. The 28 identified studies included various confounders, spatial and temporal resolutions of pollution concentrations, and COVID-19 health outcomes. Discussion: We discuss methodological challenges and highlight additional research areas based on our findings. Challenges include data quality issues, ecological study design limitations, improved adjustment for confounders, exposure errors related to spatial resolution, geographic variability in testing, mitigation measures and pandemic stage, clustering of health outcomes, and a lack of publicly available data and code.

Laboratory's medRxiv, and Cold Spring Harbor Laboratory's bioRxiv. The final search date was June 9, 2020. "Air pollution" and "COVID-19" were used as the keywords for the searches, along with their synonyms, including "PM2.5," "PM10," "particulate," "NO2," "ozone," "O3," etc., for air pollution, and "coronavirus," "SARS-CoV-2," "SARSCov19," etc., for COVID-19. The full search queries used for each database are provided in Supplemental Materials Table 1 .

Study eligibility criteria and study selection: Inclusion/exclusion criteria were designed to identify original research describing the effects of air pollution on COVID-19 health outcomes ( Figure 1 ). We included both published and preprint studies due to the quickly evolving nature of the topic. Only studies published in English up until June 9, 2020 were included.

Two screeners (AB and JC) reviewed each identified paper from the database searches in the Covidence online platform. Duplicate papers were removed. The studies were then put through a coarse-grain selection process in which studies were excluded based on a review of the title and abstract. Papers that did not discuss air pollution and COVID-19 health outcomes were excluded. In the second round, full texts were reviewed. Studies that discussed the effects of short-or long-term exposure to air pollution on COVID-19 health outcomes were included.

Studies could report various COVID-19 health outcomes, including case incidence, case-fatality rate, mortality rate, case hospitalizations, death rate, and secondary cases produced by an initially infected individual. Studies were neither restricted based on the temporal and spatial resolution of air pollution data or the COVID-19 health data nor on the basis of confounder selections.

Any studies investigating the effects of COVID-19 and the lockdowns caused by COVID-19 on air pollution were excluded as these studies are not the focus of this specific review. Many early . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 20, 2020 . . https://doi.org/10.1101 studies described the relationship between COVID-19 cases and air pollution without significant quantitative or statistical analysis. These studies concluded that air pollution is related to coronavirus spread because early cases in the pandemic were concentrated in the highly polluted Po Valley and other highly polluted areas of China (Frontera et al. 2020; Martelletti and Martelletti 2020) . These studies, which provided only cursory analyses, were also excluded.

Data extraction: For each selected study, we extracted information about the title, author, and which journal/preprint repository it was housed in, as well as geographical area of study, research question, health outcome, exposure, statistical methods, results, covariates or possible confounders, and code availability.

Study quality: One study did not describe the statistical methods used but still reported a significant p-value ( Baron et al. 2020) and another study used an unspecified air quality metric (Bashir et al. 2020) .

The search yielded 825 articles across all of the databases (Figure 1 ). After excluding studies that did not discuss both air pollution and COVID-19, 80 articles were eligible for full text review.

Of these, only 28 were full studies published in English that investigated the effects of air pollution on COVID-19 health outcomes. Only 3 of the 28 articles included open access of their code.

Tables 1-3 present a summary of the 28 selected articles. Table 1 summarizes cross-sectional studies that focus on the long-term effects of air pollution on COVID-19 health outcomes (n=12 studies) (Andree 2020; Coccia 2020; Fattorini and Regoli 2020; Liang et al. 2020; Lippi et al. . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 20, 2020 . . https://doi.org/10.1101 .08.16.20175901 doi: medRxiv preprint 2020 Ogen 2020; Pansini and Fornacca 2020a, b; Tian H et al. 2020; Tian T et al. 2020; Travaglio et al. 2020; Wu et al. 2020) . These studies examined the impact of air pollution exposure over a long period of time prior to the SARS-CoV-2 outbreak. The underlying hypotheses of these studies is that exposure to air pollution over long periods of time negatively impacts respiratory health, thereby increasing the susceptibility to SARS-CoV-2 infection and more severe COVID-19 health outcomes. All of these studies relied on a cross-sectional study design. Table 2 summarizes time-series studies that focus on the short-term effects of air pollution on COVID-19 health outcomes (n=9 studies) (Bashir et al. 2020; Han et al. 2020; Jiang et al. 2020; Li et al. 2020; Wang B et al. 2020; Xu H et al. 2020; Yao et al. 2020a, b; Zhu et al. 2020) . These studies compared day-to-day variations in air pollution with day-to-day variations in COVID-19 health outcomes. Table 3 summarizes cross-sectional studies that focus on the short-term effects of air pollution on COVID-19 health outcomes (n=8 studies) ( Baron 2020; Frontera et al. 2020; Fronza et al. 2020; Setti et al. 2020; Yao et al. 2020a; Yao et al. 2020c; Zhang T et al. 2020; Zhou et al. 2020) . These studies compared air pollution levels during the outbreak or shortly before the outbreak with COVID-19 health outcomes. The underlying hypothesis of both time-series and cross-sectional studies is that short-term air pollution exposure increases the transmission of the virus or the severity of COVID-19.

Note that one of the 28 studies included both short-term time-series and short-term crosssectional study designs (Yao et al. 2020a ); this study is included in both Tables 2 and 3.

. CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 20, 2020 . . https://doi.org/10.1101 Of the 12 long-term studies, 11 were published in MedRxiv and 1 was published in a peerreviewed journal. Of the 16 short-term studies, 10 were published in MedRxiv and 6 were published in a peer-reviewed journal.

The 28 studies examined the effects of air pollution on COVID-19 health outcomes across 9 different countries (China, Italy, USA, France, Iran, Spain, Germany, United Kingdom, and the Netherlands). Many of the studies focused on cities and provinces in China, regions and provinces in Italy, and counties in the United States, all regions where the SARS-CoV-2 outbreak occurred early and was prominent. Among the short-term studies, 12 were conducted in China, 3 in Italy, 1 in the United States, 1 in France, 1 in Germany, and 1 in Spain. Among the long-term studies, 6 were conducted in Italy, 5 in the United States, 3 in China, 2 in Spain, 2 in France, 2 in Germany, 2 in the United Kingdom, 1 in Iran, and 1 in the Netherlands.

5 different COVID-19 health outcomes were included across the 12 long-term studies and 7 different COVID-19 health outcomes were included across the 16 short-term studies. Among the 12 long-term studies, 8 used number of cases, 7 used number of deaths, 1 used casehospitalizations, 1 used case-fatality rate, and 1 used percent of severe infection. Among the 16 short-term studies, 12 used number of cases, 3 used number of deaths, 2 used case-fatality rate, 2 used basic reproduction number, 1 used ICU admissions, 1 used hospitalized cases, and 1 used a binary variable for epidemic escalation.

The 28 studies also used a variety of air pollution exposures. 11 of the 12 long-term studies collected air pollution data over extended periods of time ranging from 1 to 16 years. These studies used data after 1999 and prior to 2020. One study that self-classified as long-term . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 20, 2020 . . https://doi.org/10.1101 collected air pollution data from January and February 2020 (Ogen 2020 . While this study included air pollution data prior to the outbreak, it does not fully represent long-term exposure, as the data were collected over a period of only 2 months. Ten of the long-term studies obtained air pollution concentrations or air quality index values, and 3 studies used the number of daily limit exceedances.

The short-term time series studies used daily air pollution data collected over relatively shorter periods of time, ranging from 17 days to 80 days. These studies used daily air pollution data during the outbreak with a short lag time to account for incubation period and/or time from exposure to death. However, Bashir et al. (2020) did not discuss the use of a lag period in their analyses, which should be taken into consideration when interpreting the results from that study.

Eight of the short-term time-series studies used air pollutant concentrations or air quality index values, and one study used an unspecified air quality metric provided by the National Weather Service (Bashir et al. 2020). Short-term cross-sectional studies used aggregated air pollution data collected before or during the outbreak. These studies collected air pollution data over periods of time ranging from 1 to 3 months. One of the short-term cross-sectional studies collected air pollution concentration data based on the number of daily limit exceedances over a period of time (Setti et al. 2020) . The other 7 studies used air pollutant concentrations or air quality index values.

Overall, 21 studies examined PM2.5 (9 long-term, 12 short-term); 16 studied PM10 (7 long-term, 9 short-term); 14 studied NO2 (7 long-term, 7 short-term); 12 studied O3 (8 long-term, 4 shortterm); 7 studied CO (3 long-term, 4 short-term); and 7 studied SO2 (4 long-term, 3 short-term).

Two short-term studies examined Air Quality Index (AQI), 1 long-term study examined NO, 1 long-term study examined aerosols, 1 long-term study examined HCHO, 1 short-term study . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 20, 2020 . . https://doi.org/10.1101 looked at NH3, and 1 short-term study included an unspecified air quality metric provided by the National Weather Service.

5 of the 12 long-term studies (Table 1) did not consider covariates and confounders in their analyses. 3 of the 9 short-term time-series studies ( Table 2 ) and 3 of the 8 short-term crosssectional studies (Table 3 ) did not consider covariates and confounders in their analyses.

There were 8 different statistical models used across the 12 long-term studies and 9 different statistical models used across the 16 short-term studies. Among the 12 long-term studies, 3 used linear regression, 2 used negative binomial regression, 2 used negative binomial mixed models, 2 used rank correlation, 1 used descriptive analyses, 1 used non-parametric penalized linear regression, 1 used Poisson regression, and 1 used Partitioning Around the Medoids Clustering Algorithm. Among the 16 short-term studies, 8 used linear regression, 2 used Poisson regression, 2 used rank correlation, 1 used binomial regression, 1 used unspecified analyses, 1 used artificial neural networks (ANN), 1 used negative binomial regression, 1 used a generalized additive model with a Gaussian distribution, and 1 used a generalized additive model with a quasi-Poisson distribution.

Of the 28 studies, 27 showed a statistically significant positive association between air pollution and adverse COVID-19 health outcomes ( Figure 2 ). The only long-term study that did not show a statistically significant association conducted a descriptive analysis of NO2 and COVID-19 health deaths (Ogen 2020). All of the 16 short-term studies found statistically significant positive associations between air pollution and adverse COVID-19 outcomes. While 27 of the 28 papers . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 20, 2020 . . https://doi.org/10.1101 found a statistically significant positive association between an air pollutant and adverse COVID-19 health outcomes, not every air pollutant studied was found to be statistically significant or positively associated with every adverse COVID-19 health outcome studied in every location studied for every lag period analyzed.

Among the long-term studies, those that found a statistically significant positive association between a specific air pollutant and adverse COVID-19 health outcomes are as follows: 8 of 9 studies of PM2.5; 5 of 8 studies of O3; 6 of 7 studies of NO2; 6 of 7 studies of PM10; 1 of 4 studies of SO2; and 2 of 3 studies of CO. The only long-term studies that included HCHO and NO found a statistically significant positive association. The only long-term study of aerosols did not find a statistically significant positive association.

Among the short-term studies, those that found a statistically significant positive association between a specific air pollutant and adverse COVID-19 health outcomes are as follows: all 12 studies of PM2.5; 8 of 9 studies of PM10; 5 of 7 studies of NO2; 3 of 4 studies of O3; and 3 of 4 studies of CO. Of the 3 short-term studies of SO2, none found a statistically significant positive association with adverse COVID-19 health outcomes, whereas both of the 2 short-term papers studying AQI found a statistically significant positive association. The only papers studying NH3 and an unspecified air quality metric provided by the National Weather Service also found statistically significant positive associations with adverse COVID-19 health outcomes.

We searched the literature from January 1, 2020 to June 9, 2020 to identify studies that examined the relationship between air pollution and adverse COVID-19 health outcomes. The query yielded papers published in 2020, and 28 papers satisfied our exclusion/inclusion criteria.

These papers used different statistical models, different datasets, and different confounders.

. CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 20, . . https://doi.org/10.1101 Nonetheless, 27 of the 28 included studies reported some form of statistically significant association between exposure to air pollution and adverse COVID-19 health outcomes.

While researchers want to disseminate their results rapidly to accelerate knowledge about this pandemic as it unfolds, this practice presented us with many challenges when trying to understand the results of the selected studies and their implications. Yet, even in pre-pandemic times, research on the health effects of air pollution is challenging. It is well known that assessing causality in air pollution epidemiology is complicated by the fact that most studies are observational with a large number of confounding factors (Dominici and Zigler 2017). There is also the consistent threat of unmeasured confounding bias. Exposure and outcome misclassification present further challenges.

When the goal is to determine a clear association between air pollution and its effects on health outcomes, such as adverse COVID-19 health outcomes, some well-known challenges become even more serious, and a number of additional complications emerge. These additional challenges are mostly related to the reality that the data on SARS-CoV-2 and COVID-19 are constantly evolving, come from a variety of sources, and are primarily available only at the aggregate level (e.g., county, regional, or country level). Indeed, the quality of the data and the lack of available well-validated electronic health record data at the national level (particularly in the United States) renders the general challenges in evaluating causality enumerated above even more difficult to address. We summarize below the most important methodological challenges, which in turn highlight critically important areas of research that will need to be pursued.

The validity of the health outcome data are questionable, as there is no uniform case definition of a COVID-19 death and there are also diagnostic errors in COVID-19 cases . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 20, . . https://doi.org/10.1101 .08.16.20175901 doi: medRxiv preprint (Gandhi and Singh 2020 . This could potentially contribute to a high degree of over-or underreporting. Many infected people may have died without ever having been tested for SARS-CoV-2, while in other cases, SARS-CoV-2 might have been secondary to the cause of an individual's death. Moreover, in the early phases of the SARS-CoV-2 pandemic, and even in the current phase of the pandemic, testing was not universally available due to shortages in test kits and the relatively late adoption of mass-scale testing. Additionally, only a subset of the population has been tested, and we cannot account for those who are asymptomatic but carry the virus.

Ecological fallacy: 27 of the 28 studies included were ecological studies as opposed to individual-level studies. The only study conducted at the individual level was conducted by a group in England, where links between air pollution and adverse COVID-19 health outcomes were studied by looking at patient health records (Travaglio et al. 2020) . Ecological fallacy is a formal fallacy in the interpretation of statistical data that occurs when inferences about the nature of individuals are deduced from inferences about the group to which those individuals belong (see for example (Jackson et al. 2008) ). While ecological studies are often very useful at generating preliminary evidence in data-scarce settings, increasing the scientific rigor of research in this area requires access to nationally representative, individual-level data on adverse COVID-19 health outcomes, including information about patients' residential address, demographics, and individual-level confounders. This is an enormous challenge that will require many privacy, legal, and ethical trade-offs (Sittig and Singh 2020).

Other determinants of COVID-19: Ecological data typically do not allow for adjustment by individual risk factors such as age, gender, ethnicity, or occupation. Age is one of the strongest predictors of survival for most conditions, including COVID-19. Gender-based differences in time-activity patterns contribute to different levels of air pollution exposure between men and women, and women have been shown to be more susceptible to several environmental . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 20, . . https://doi.org/10.1101 exposures (Clougherty 2010). Occupation is also an important factor in this pandemic, as those who provide medical care and other essential workers, such as those working in meat packing plants, are at increased risk of developing SARS-CoV-2 infection (Mutambudzi et al. 2020) .

Accounting for difference and socioeconomic status: It is also important to recognize that disadvantaged people (e.g., those without health insurance, undernourished, and with poorly managed underlying health conditions, such as cardiovascular conditions and/or diabetes) have a greater susceptibility for both contracting SARS-CoV-2 and dying from COVID-19. Myriad social and economic factors contribute to high rates of infection and put individuals at higher risk from the sequelae of COVID-19, and disparities in COVID-19-related outcomes may be due to social deprivation rooted in long-standing racial and socio-economic inequities. This issue has been raised by a number of authors (see for example (Chowkwanyun and Reed 2020; Yancy 2020)).

In most of the studies reviewed, both long-term and short-term, the same level of air pollution exposure was assigned to everyone living in large geographical areas. Therefore, spatial differences in exposure were not captured. Several statistical approaches have been developed to propagate the different sources of uncertainty associated with exposure error into the statistical model to estimate health effects (see for example (Dominici et al. 2000; Gryparis et al. 2009; Szpiro et al. 2011) ). These approaches have not yet been implemented in studies of air pollution exposure and adverse COVID-19 health outcomes.

Physical distancing: The implementation of public health policies, which can vary widely by jurisdiction, has been shown to be successful in reducing SARS-CoV-2 transmission and flattening the epidemic curve. For example, in Georgia, areas that did not adopt physical distancing practices experienced higher incidence and mortality from COVID-19 when . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 20, 2020. . https://doi.org/10.1101/2020.08.16.20175901 doi: medRxiv preprint compared to other areas in the state that did. Cities in California that tend to have higher levels of fine particulate matter adopted stay-at-home policies earlier than other regions. Because these policies differ by regional air pollution levels, including rural and urban areas within the same county, they can distort the observed associations between air pollution and adverse COVID-19 health outcomes.

Timing on the epidemic curve: There will be temporal differences in the number of incident COVID-19 cases, and by extension in COVID-19-related deaths, by region. US counties were at very different stages on the epidemic curves, especially in early April. Larger cities are more populous and tend to have increased travel to and from international locations, providing increased opportunity for the spread of COVID-19 early in the pandemic. These larger cities also tend to have higher concentrations of air pollution. The practical implications are that there will be a greater number of incident cases and deaths in those cities that are further along on the epidemic curve, which further confounds analysis.

Unlike studies of long-term exposure to air pollution and chronic diseases where deaths can reasonably be assumed to be independent, COVID-19 cases and deaths tend to occur in clusters. This has been widely reported, such as the now famous choir practice where a large portion of attendees became ill or the tragic events in congregant settings such as retirement homes and long-term care facilities. Although some of the studies' authors included random effects to account for clustering, without individual-level data, it is simply not possible to account for this clustering.

To discover crucial linkages between air pollution and adverse COVID-19 outcomes in a more definitive, causal manner, both the data used for the analyses as well as the code should be made publicly available. Transparency and shared resources will assist in . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 20, 2020. . https://doi.org/10.1101/2020.08.16.20175901 doi: medRxiv preprint the global push towards uncovering the solutions to the pandemic and instituting public policies that will protect the health of people worldwide. Only two of the 28 selected papers included publicly-available code. Many of the selected papers were in preprint format, and access to code will be critical to validate results and build upon the conclusions. It is difficult to reproduce results and continue work on these areas without code availability.

Assessing the short-and long-term effects of air pollution on adverse COVID-19 health outcomes is a rapidly evolving area of research. Most of these studies are still in the preprint stage and many more studies will likely be published in the next few weeks and months. Therefore, our aim was to identify this topic as a critically important area of research, identify the numerous methodological challenges, and inform future research. Despite the preliminary nature of the evidence summarized in this scoping review, our findings underscore the need to hold governments accountable for the installation of environmental protections that will permanently maintain safe levels of air pollution to protect human health, rather than removing those environmental protections at the behest of the industries that pollute our environment. In the era of climate change, and now global pandemics, studies like these are among many calling for action to protect the earth systems that are so deeply intertwined with human systems, particularly systems of human health.

A primary goal of this scoping review was to begin paving the way toward overcoming the many methodological challenges that are inherent in studies of environmental health epidemiology.

Some of these challenges are common to all epidemiological studies of air pollution and health outcomes (e.g., exposure error, measured and unmeasured bias), while others are exacerbated in the study of adverse COVID-19 health outcomes (e.g., outcome misclassification, evolving . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 20, 2020. . https://doi.org/10.1101/2020.08.16.20175901 doi: medRxiv preprint number of new cases and deaths) and global pandemics (e.g., accounting for different testing practices, different stages of the pandemic).

With respect to these new challenges, it is just as important to highlight new opportunities. For example, the extreme measures implemented during the lockdown are providing new research opportunities to investigate important questions regarding achievable reductions in air pollution exposure and health effects. Indeed, in order to mitigate the effects of the pandemic and to protect lives, countries across the globe instituted temporary, and in many cases, continued closures of all but essential businesses and services, instituting lockdowns to encourage people to stay at home and prevent the spread of disease. This has resulted in significantly lower levels of vehicular, train, aircraft, and industry related emissions (Le Quéré et al. 2020; Shilling and Waetjen 2020; Zhang R et al. 2020) . These lockdown measures provide a unique opportunity to exploit the features of a quasi-experimental design to assess the extent to which different pollutants have declined and estimate the potential "beneficial" effects of these declines on health outcomes. . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 20, 2020. . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 20, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 20, 2020. Pansini R, Fornacca D. 2020a. Initial evidence of higher morbidity and mortality due to SARS-CoV-2 in regions with lower air quality. medRxiv: 2020.2004.2004.20053595, 10.1101/2020.04.04.20053595 Pansini R, Fornacca D. 2020b. Higher virulence of COVID-19 in the air-polluted regions of eight severely affected countries. medRxiv: 2020.2004.2030.20086496, 10.1101/2020.04.30.20086496 Rhee Evidence-based Research Hypotheses. medRxiv: 2020.2004.2011.20061713, 10.1101/2020.04.11.20061713 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 20, 2020. . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 20, 2020. . https://doi.org/10.1101/2020.08.16.20175901 doi: medRxiv preprint . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 20, 2020. . https://doi.org/10.1101/2020.08.16.20175901 doi: medRxiv preprint Zhang T, Zhao G, Luo L, Li Y, Shi W. 2020. Associations between ambient air pollutants exposure and case fatality rate of COVID-19: a multi-city ecological study in China. medRxiv: 2020.2005.2006.20088682, 10.1101/2020.05.06.20088682 Zhou J, Qin L, Liu N. 2020. Ambient air pollutants, meteorological factors and their interactions affect confirmed cases of COVID-19 in 120 Chinese cities. medRxiv: 2020.2005.2027.20111542, 10.1101/2020.05.27.20111542 Zhu Y, Xie J, Huang F, Cao L. 2020. Association between short-term exposure to air pollution and COVID-19 infection: Evidence from China. Science of The Total Environment 727:138704, https://doi.org/10.1016/j.scitotenv. 2020.138704 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 20, 2020. . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 20, 2020. . https://doi.org/10.1101/2020.08.16.20175901 doi: medRxiv preprint Figure 1 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 20, 2020. . https://doi.org/10.1101/2020.08.16.20175901 doi: medRxiv preprint

. CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 20, 2020. . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 20, 2020. . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 20, 2020. . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 20, 2020. TOPIC: ("novel coronavirus" OR 2019-ncov OR coronavirus OR "coronavirus disease-9" OR coronavirus* OR coronovirus* OR "2019-nCoV" OR COVID-19 OR Covid OR ncov OR SARS-CoV-2 OR SARSCov19 OR "SARS-Covid" OR "covid 19" OR "Coronaviridae" OR "Coronaviridae infection" OR "coronavirus disease 2019") AND TOPIC: ((((pollut* OR qualit*) AND air) AND ("air pollution" OR "pollution and pollution related phenomena" OR "particulate matter" OR "particulate matter 2 5" OR "particulate matter 10" OR "airborne particle" OR "air quality" OR "wood smoke" OR "air pollutant" OR "air pollution" OR "PM 2.5" OR PM5 OR PM10 OR "Air pollutant*" OR particle* OR particulate* OR "air quality" OR (air AND (quality* OR pollut*)) OR air-quality OR smog OR soot OR gases OR "wood smoke" OR ozone OR hydrocarbon* OR "polyaromatic hydrocarbon*" OR PAH OR "greenhouse gas*" OR "carbon dioxide" OR methane OR hydrofluorocarbon* OR HFC OR "volatile organic compound*" OR VOC OR CO2 OR combustion OR emission* OR "Nitrogen oxide*" OR NOx OR NO2 OR "Sulphur dioxide" OR SO2 OR O3))) 201 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 20, 2020. 165 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 20, . . https://doi.org/10.1101 

Template: COVID-19 Pandemic Data

Exposure to air pollution and COVID-19 mortality in the United States: A nationwide cross-sectional study. medRxiv : the preprint server for health sciences:2020

Possible environmental effects on the spread of COVID-19 in China

Pathological findings of COVID-19 associated with acute respiratory distress syndrome

COVID-19 and African Americans

Ambient nitrogen dioxide pollution and spread ability of COVID-19 in Chinese cities. medRxiv:2020

Temporal Association Between Particulate Matter Pollution and Case Fatality Rate of COVID-19 in Wuhan, China. medRxiv:2020

Spatial Correlation of Particulate Matter Pollution and Death Rate of COVID-19. medRxiv:2020

NOx Emission Reduction and Recovery during COVID-19 in East China

OR 'nitrogen oxide*':ab,ti OR nox:ab,ti OR no2:ab,ti OR 'sulphur dioxide':ab,ti OR so2:ab,ti OR o3:ab,ti)

ti OR coronavirus*:ab,ti OR coronovirus*:ab,ti OR '2019-ncov':ab,ti OR 'covid 19':ab,ti OR covid:ab,ti OR ncov:ab,ti OR 'sars cov 2':ab,ti OR sarscov19:ab,ti OR 'sars-covid':ab,ti OR

We gratefully acknowledge support from the 2020 Star-Friedman Challenge for Promising Scientific Research which supported Dr. Dominici and Dr. Braun.