PreprintPDF Available

Exposure to air pollution and COVID-19 mortality in the United States: A nationwide cross-sectional study

Authors:

Abstract and Figures

Objectives: United States government scientists estimate that COVID-19 may kill tens of thousands of Americans. Many of the pre-existing conditions that increase the risk of death in those with COVID-19 are the same diseases that are affected by long-term exposure to air pollution. We investigated whether long-term average exposure to fine particulate matter (PM2.5) is associated with an increased risk of COVID-19 death in the United States. Design: A nationwide, cross-sectional study using county-level data. Data sources: COVID-19 death counts were collected for more than 3,000 counties in the United States (representing 98% of the population) up to April 22, 2020 from Johns Hopkins University, Center for Systems Science and Engineering Coronavirus Resource Center. Main outcome measures: We fit negative binomial mixed models using county-level COVID-19 deaths as the outcome and county-level long-term average of PM2.5 as the exposure. In the main analysis, we adjusted by 20 potential confounding factors including population size, age distribution, population density, time since the beginning of the outbreak, time since state’s issuance of stay-at-home order, hospital beds, number of individuals tested, weather, and socioeconomic and behavioral variables such as obesity and smoking. We included a random intercept by state to account for potential correlation in counties within the same state. We conducted more than 68 additional sensitivity analyses. Results: We found that an increase of only 1 μg/m3 in PM2.5 is associated with an 8% increase in the COVID-19 death rate (95% confidence interval [CI]: 2%, 15%). The results were statistically significant and robust to secondary and sensitivity analyses. Conclusions: A small increase in long-term exposure to PM2.5 leads to a large increase in the COVID-19 death rate. Despite the inherent limitations of the ecological study design, our results underscore the importance of continuing to enforce existing air pollution regulations to protect human health both during and after the COVID-19 crisis. The data and code are publicly available so our analyses can be updated routinely.
Content may be subject to copyright.
1
Exposure to air pollution and COVID-19 mortality in the United States: A nationwide
cross-sectional study
Xiao Wu, Rachel C Nethery, M Benjamin Sabath, Danielle Braun, Francesca Dominici
Xiao Wu, doctoral student; Rachel C Nethery, assistant professor; Benjamin Sabath, data
scientist; Danielle Braun, research scientist; Francesca Dominici, Clarence James Gamble
professor of biostatistics, population, and data science, Department of Biostatistics, Harvard
T.H. Chan School of Public Health, Boston, MA, 02115, USA
Lead authors: Xiao Wu and Rachel C. Nethery
Correspondence to: Francesca Dominici, PhD (ORCiD 0000-0001-5964-0756)
Clarence James Gamble Professor of Biostatistics, Population and
Data Science
Harvard T.H. Chan School of Public Health
Co-Director Harvard Data Science Initiative
677 Huntington Avenue
Boston, MA 02115
410.258.5886
Email: fdominic@hsph.harvard.edu
Updated April 24, 2020
2
Abstract
Objectives: United States government scientists estimate that COVID-19 may kill tens of
thousands of Americans. Many of the pre-existing conditions that increase the risk of death in
those with COVID-19 are the same diseases that are affected by long-term exposure to air
pollution. We investigated whether long-term average exposure to fine particulate matter (PM2.5)
is associated with an increased risk of COVID-19 death in the United States.
Design: A nationwide, cross-sectional study using county-level data.
Data sources: COVID-19 death counts were collected for more than 3,000 counties in the United
States (representing 98% of the population) up to April 22, 2020 from Johns Hopkins University,
Center for Systems Science and Engineering Coronavirus Resource Center.
Main outcome measures: We fit negative binomial mixed models using county-level COVID-19
deaths as the outcome and county-level long-term average of PM2.5 as the exposure. In the main
analysis, we adjusted by 20 potential confounding factors including population size, age
distribution, population density, time since the beginning of the outbreak, time since state’s
issuance of stay-at-home order, hospital beds, number of individuals tested, weather, and
socioeconomic and behavioral variables such as obesity and smoking. We included a random
intercept by state to account for potential correlation in counties within the same state. We
conducted more than 68 additional sensitivity analyses.
Results: We found that an increase of only 1 𝜇g/m3 in PM2.5 is associated with an 8% increase in
the COVID-19 death rate (95% confidence interval [CI]: 2%, 15%). The results were statistically
significant and robust to secondary and sensitivity analyses.
Conclusions: A small increase in long-term exposure to PM2.5 leads to a large increase in the
COVID-19 death rate. Despite the inherent limitations of the ecological study design, our results
3
underscore the importance of continuing to enforce existing air pollution regulations to protect
human health both during and after the COVID-19 crisis. The data and code are publicly available
so our analyses can be updated routinely.
4
Summary Box
What is already known on this topic
1. Long-term exposure to PM2.5 is linked to many of the comorbidities that have been
associated with poor prognosis and death in COVID-19 patients, including cardiovascular
and lung disease.
2. PM2.5 exposure is associated with increased risk of severe outcomes in patients with certain
infectious respiratory diseases, including influenza, pneumonia, and SARS.
3. Air pollution exposure is known to cause inflammation and cellular damage, and evidence
suggests that it may suppress early immune response to infection.
What this study adds
1. This is the first nationwide study of the relationship between historical exposure to air
pollution exposure and COVID-19 death rate, relying on data from more than 3,000
counties in the United States. The results suggest that long-term exposure to PM2.5 is
associated with higher COVID-19 mortality rates, after adjustment for a wide range of
socioeconomic, demographic, weather, behavioral, epidemic stage, and healthcare-related
confounders.
2. This study relies entirely on publicly available data and fully reproducible, public code to
facilitate continued investigation of these relationships by the broader scientific community
as the COVID-19 outbreak evolves and more data become available.
A small increase in long-term PM2.5 exposure was associated with a substantial increase in the
county’s COVID-19 mortality rate up to April 22, 2020.
5
Introduction
The scale of the COVID-19 public health emergency is unmatched in our lifetime and will have
grave social and economic consequences. The suddenness and global scope of this pandemic has
raised urgent questions that require coordinated investigation in order to slow the disease’s
devastation. A critically important public health objective is to identify key modifiable
environmental factors that may contribute to the severity of the health outcomes (e.g., ICU
hospitalization and death) among individuals with COVID-19. Data from China and Italy show
that a majority of COVID-19 deaths occurred in adults aged ≥60 years1 and in persons with serious
underlying health conditions.2-4 Early age-stratified COVID-19 death rates in the United States,
reported by the Centers for Disease Control and Prevention (CDC),5 also suggest that persons aged
≥65 are at highest risk. Additional factors associated with severe disease include male sex and the
presence of comorbidities including hypertension, obesity, diabetes mellitus, cardiovascular
disease, and chronic lung disease.6 7 Severe COVID-19 infection is characterized by a high
inflammatory burden, and it can cause viral pneumonia with additional extrapulmonary
manifestations and complications including acute respiratory distress syndrome (ARDS),8-13 which
has a mortality rate ranging from 27% to 45%.14 Studies have also documented high rates of heart
damage,11 15 cardiac arrhythmias,12 and blood clots16 in COVID-19 patients. Patients with severe
disease can suffer respiratory failure and failure of other vital systems, leading to death.
Although the epidemiology of COVID-19 is evolving, there is a large overlap between causes of
death in COVID-19 patients and the conditions caused and/or exacerbated by long-term exposure
to fine particulate matter (PM2.5). PM2.5 contains microscopic solids or liquid droplets small
enough that they can be inhaled and cause serious health problems. The Global Burden of Disease
6
Study identified air pollution as a risk factor for total and cardiovascular disease mortality, and it
is believed to have contributed to nearly 5 million premature deaths worldwide in 2017 alone.17
On Thursday, March 26, 2020 the US EPA announced a sweeping relaxation of environmental
rules in response to the coronavirus pandemic, allowing power plants, factories and other facilities
to determine for themselves if they are able to meet legal requirements on reporting air and water
pollution. The association between PM2.5 and health, including both infectious and chronic
respiratory diseases, cardiovascular diseases, neurocognitive disease, and pregnancy outcomes in
the United States and worldwide is well established.18-24 A recent study by our group also
documented a statistically significant association between long-term exposures to PM2.5 and ozone
and risk of ARDS among older adults in the United States.25 Numerous scientific studies reviewed
by the United States Environmental Protection Agency (US EPA) have linked PM2.5 to a variety
of health concerns including premature death in people with heart or lung disease, non-fatal heart
attacks, irregular heartbeats, aggravated asthma, decreased lung function, and increased respiratory
symptoms such as inflammation, airway irritations, coughing, or difficulty breathing.26
We hypothesize that because long-term exposure to PM2.5 adversely affects the respiratory and
cardiovascular systems and increases mortality risk,27-29 it also exacerbates the severity of COVID-
19 infection symptoms and worsens the prognosis of COVID-19 patients. In this study, we
quantified the impact of long-term PM2.5 exposure on COVID-19 mortality rates in United States
counties. Our study includes 3,087 counties in the United States, covering 98% of the population.
We leveraged our previous efforts that focused on estimating the long-term effects of PM2.5 on
mortality among 60 million United States’ Medicare enrollees.20 30 31 We used a well-tested
research data platform that gathers, harmonizes, and links nationwide air pollution data, census
7
data, and other potential confounding variables with health outcome data. We augmented this
platform with newly collected COVID-19 data from authoritative data sources.32 All data sources
used in these analyses, along with fully reproducible code, are publicly available to facilitate
continued investigation of these relationships as the COVID-19 outbreak evolves and more data
become available.
Methods
Table 1 summarizes our data sources and their provenance, including links where the raw data
can be extracted directly.
COVID-19 deaths
We obtained COVID-19 death counts for each county in the United States from Johns Hopkins
University, Center for Systems Science and Engineering Coronavirus Resource Center.32 This
source provides the most comprehensive county-level COVID-19 data to date reported by the CDC
and state health departments, including the number of new and cumulative deaths and confirmed
cases reported in each county across the United States, updated daily. We collected the cumulative
number of deaths for each county up to and including April 22, 2020. County-level COVID-19
mortality rates were defined for our analyses as the ratio of COVID-19 deaths to county level
population size. While individual-level data would have allowed a more rigorous statistical
analyses, individual-level data on COVID-19 death is currently not available.
Exposure to air pollution
8
We calculated county-level long-term exposure to PM2.5 (averaged from 2000 to 2016) from an
established exposure prediction model.33 The PM2.5 exposure levels were estimated monthly at
0.01° × 0.01° grid resolution across the entire continental United States by combining satellite,
modeled, and monitored PM2.5 data in a geographically weighted regression. These estimates
have been extensively cross-validated.33 We aggregated these levels spatially by averaging the
values for all grid points within a zip code and then averaging across zip codes within a county.
We obtained temporally averaged PM2.5 values (20002016) at the county level by averaging
estimated PM2.5 values within a given county. We computed the average 2016 PM2.5 exposure
analogously for each county to use in sensitivity analyses.
Potential confounders
In the main analysis, we considered the following 19 county-level variables and one state-level
variable as potential confounders (see also Table 2): days since first COVID-19 case reported (a
proxy for epidemic stage), population density, percent of population 65 years of age, percent of
the population 45-64 years of age, percent of the population 15-44 years of age, percent living in
poverty, median household income, percent black, percent Hispanic, percent of the adult
population with less than a high school education, median house value, percent of owner-occupied
housing, percent obese, percent current smokers, number of hospital beds per unit population, and
average daily temperature and relative humidity for summer (June-September) and winter
(December-February) for each county, and days since issuance of stay-at-home order for each
state. Note that publicly available daily COVID-19 case counts at the county level were only
available starting March 22, 2020, so that the measure of days since first COVID-19 case reported
9
was truncated by this date. Additional detail on the creation of all variables used in the analysis is
available in the Supplementary Materials.
Statistical methods
We fit a negative binomial mixed model34-36 using COVID-19 deaths as the outcome and PM2.5 as
the exposure of interest to estimate the association between COVID-19 mortality rate and long-
term PM2.5 exposure, adjusted by covariates. The model included a population size offset and was
adjusted for all the potential confounders listed above. We also included a random intercept by
state to account for potential correlation in counties within the same state, due to similar socio-
cultural, behavioral, and healthcare system features and similar COVID-19 response and testing
policies. Additional modeling details are provided in the Supplementary Materials. We report
mortality rate ratios (MRR), i.e., exponentiated parameter estimates from the negative binomial
model, and 95% CI. The MRR for PM2.5 can be interpreted as the relative increase in the COVID-
19 mortality rate associated with a 1 𝜇g/m3 increase in long-term average PM2.5 exposure. We
carried out all analyses in R statistical software and performed model fitting using the lme4
package.37 38
Quantifying unmeasured confounding bias
Because this study is observational and the contributing factors to COVID-19 spread and severity
remain largely unknown at this early stage of the pandemic, unmeasured confounding is a concern
in our analyses. The E-value is a commonly used metric to evaluate the potential impact of
unmeasured confounding on results from an observational study.39 For a pre-specified exposure
variable of interest (long-term exposure to PM2.5), the E-value quantifies the minimum strength of
10
association that an unmeasured confounder must have, with both the outcome (COVID-19
mortality rate) and exposure (long-term exposure to PM2.5) conditional to all of the potential
confounders included in the regression model, to explain away the estimated exposure-outcome
relationship. We report the E-value for the MRR estimate for PM2.5 under the main model with 20
potential confounders.
Secondary analyses
In addition to the main analysis, we conducted six secondary analyses to assess the robustness of
our results to the confounder set used, outliers, and the model form specification.
First, because the New York metropolitan area has experienced the most severe COVID-19
outbreak in the United States to date, we anticipated that it would strongly influence our analysis.
As a result, we repeated the analysis excluding the counties comprising the New York metropolitan
area, as defined by the Census Bureau.
Second, although in our main analysis we adjusted for days since first COVID-19 case reported to
capture the size of an outbreak in a given county, this measure is imprecise. To further investigate
the potential for residual confounding bias (i.e., if counties with high PM2.5 exposure also tend to
have large outbreaks relative to the population size, then their death rates per unit population could
appear differentially elevated, inducing a spurious correlation with PM2.5), we also conducted
analyses excluding counties with fewer than 10 confirmed COVID-19 cases.
11
Third, we omitted an anticipated strong confounder, days since first COVID-19 case reported,
from the model. Fourth, we additionally adjusted our models for the number of tests performed at
the state level (see Table 1 for data source) to evaluate how state-level differences in testing
policies might impact our results. Fifth, we additionally adjusted our models for county-level
estimated percentage of people with COVID-19 symptoms (see Table 1 for data source) to evaluate
how the size of the outbreak in each county might impacts our results. Sixth, we introduced PM2.5
into our models as a categorical variable, categorized at the empirical quintiles, to assess the
sensitivity of our results to the assumption of a linear effect of PM2.5 on COVID-19 mortality rates.
Sensitivity analyses
We conducted 68 sensitivity analyses to assess the robustness of our results to data and modeling
choices. First, we repeated all the analyses using alternative methods to estimate exposure to
PM2.5.31 Second, we fit the models, modifying the adjustment for confounders, such as using a log
transformation or categorized versions of some of the covariates. Third, because our study relies
on observational data, our results could be sensitive to modeling choices (e.g., distributional
assumptions or assumptions of linearity). We evaluated sensitivity to such choices by considering
alternative model specifications and by fitting models stratified by county urban-rural status.
Additional detail about the sensitivity analyses and the results are provided in the Supplementary
Materials.
Results
Our study utilized data from 3,087 counties, of which 1,799 (58.3%) had reported zero COVID-
19 deaths at the time of this analysis. Table 2 describes the data used in our analyses. All COVID-
12
19 death counts (a total of 45,817 deaths) are cumulative up to April 22, 2020. Figure 1 illustrates
the spatial variation of long-term average exposure to PM2.5 and COVID-19 death rates (per 1
million population) by county. Visual inspection suggests higher COVID-19 death rates in the
Mid-Atlantic, upper Midwest, and Gulf Coast regions. These spatial patterns in COVID-19 death
rates generally mimic patterns in both high population density and high PM2.5 exposure areas. In
the Supplementary Materials, we provide additional data diagnostics that justify the use of the
negative binomial model for our analyses.
In Table 3, we report the estimated regression coefficients for each of the covariates included in
our main analysis, including PM2.5. We found that the estimated MRR for PM2.5 is 1.08 (1.02,
1.15). That is, we found that an increase of only 1 𝜇g/m3 in long-term average PM2.5 is associated
with a statistically significant 8% increase in the COVID-19 death rate. Importantly, we also found
that population density, days since first COVID-19 case reported, rate of hospital beds, median
household income, percent with less than a high school education, and percent Black are important
predictors of COVID-19 death rate. Our results are consistent with previously reported findings
that Black Americans are at higher risk of COVID-19 mortality than other groups,40 we found a
45% (32%, 60%) increase in COVID-19 mortality rate associated with a 1-standard deviation (per
14.2%) increase in percent Black residents.
For our main analysis, the E-value for the estimated MRR for PM2.5 was 1.37. That is, in order for
an unmeasured confounder to fully account for the estimated effect of PM2.5 on the COVID-19
mortality rate, it would have to be associated with both long-term PM2.5 exposure and COVID-19
mortality by a risk ratio of at least 1.37-fold each, through pathways independent of all covariates
13
already included in the model. If we were to include such a confounder in our models, along with
all other confounders considered, the estimated MRR for PM2.5 mortality would become 1 (the
null value). To get a sense of the magnitude of the required confounding effect, we also computed
the E-value for some of our key measured confounders for comparison. The E-values for days
since first COVID-19 case reported (1.16), the weather variables (1.02), number of hospital beds
(1.04) and the behavioral risk factors (1.02) were significantly smaller than the reported E-values
for the required unmeasured confounder. This suggests that any unmeasured confounder would
need to have a confounding effect substantially larger than any of our observed confounders in
order to explain away the relationship between PM2.5 and COVID-19 mortality rate.
In Figure 2, we report the MRR and 95% CI for PM2.5 from all secondary analyses. In these
analyses, we separately (a) omitted New York metropolitan area; (b) excluded counties with fewer
than 10 confirmed COVID-19 cases; (c) omitted time since first reported COVID-19 case from
the model; (d) additionally adjusted the model for number of tests performed; (e) additionally
adjusted the model for estimated percentage of people with COVID-19 symptoms; and (f) treated
PM2.5 as a categorical variable. The results of these analyses were consistent with the main
analysis. For the analysis of the PM2.5 categorized into quintiles, the MRR for the kth can be
interpreted as the increase in COVID-19 mortality rate associated with a change from the first
quintile to the kth quintile in long-term PM2.5 exposure. The MRR estimates from this model
monotonically increased as PM2.5 increased, supporting the assumption of a linear relationship
between PM2.5 and COVID-19 mortality rates. The results of all sensitivity analyses are provided
in the Supplementary Materials.
14
Discussion
This is the first nationwide study in the United States to estimate the relationship between long-
term exposure to PM2.5 and COVID-19 death rates. The results indicate that long-term exposure
to air pollution increases vulnerability to the most severe COVID-19 outcomes. We found
statistically significant evidence that an increase of 1 𝜇g/m3 in long-term PM2.5 exposure is
associated with an 8% increase in the COVID-19 mortality rate. Our results were adjusted for a
large set of socioeconomic, demographic, weather, behavioral, epidemic stage, social isolation
measures, and healthcare-related confounders and demonstrated robustness across a wide range of
sensitivity analyses.
In our previous study20 of 60 million Americans older than 65 years of age, we found that a 1
𝜇g/m3 in long-term PM2.5 exposure is associated with a 0.73% increase in the rate of all-cause
mortality. Therefore, the same small increase in long-term exposure to PM2.5 led to an increase in
the COVID-19 death rate of a magnitude 11 times that estimated for all-cause mortality.
Our results are consistent with previous findings that air pollution exposure increases severe
outcomes during infectious disease outbreaks. Ciencewicki and Jaspers19 provide a review of the
epidemiologic and experimental literature linking air pollution to infectious disease. During the
2003 outbreak of Severe Acute Respiratory Syndrome (SARS), a type of coronavirus closely
related to COVID-19, Cui et al41 reported that locations in China with a moderate or high long-
term air pollution index (API) had SARS case fatality rates 126% and 71% higher, respectively,
than locations with low API. Long-term particulate matter exposure has been associated with
hospitalizations for pneumonia in the well-controlled quasi-experimental conditions provided by
15
the closing of the Utah Valley Steel Mill,42 and a link between long-term PM2.5 exposure and
pneumonia and influenza deaths was reported in a well-validated cohort study.28 Several studies
have reported associations between short-term PM2.5 exposure and poor infectious disease
outcomes,43 44 including higher hospitalization rates or increased medical encounters for influenza,
pneumonia, and acute lower respiratory infections. In these studies and in the literature on the
association between air pollution and chronic disease outcomes, relationships with long-term
pollution exposure tend to be stronger than relationships with short-term exposure,20 45 46 and the
large effect estimate in our study is consistent with this trend.
Relationships have also been detected between pollution exposures and severe outcomes in the
context of past pandemics. Studies found particulate matter exposure to be associated with the
mortality during the H1N1 influenza pandemic in 2009.47 48 Recent studies have even used historic
data to show a relationship between air pollution from coal burning and mortality in the 1918
Spanish influenza pandemic.49 50
Although our study design cannot provide insight into the mechanisms underlying the relationship
between PM2.5 and COVID-19 mortality, prior studies have shed light on the potential biological
mechanisms that may explain the relationship between air pollution and viral outcomes.19 PM2.5
exposure is known to be associated with many of the cardiovascular and respiratory comorbidities
that dramatically increase the risk of death in COVID-19 patients. We hypothesize that the effects
captured here are largely mediated by these comorbidities and pre-existing PM-related
inflammation and cellular damage,46 51 as suggested by a recent commentary.52 Experimental
studies19 53-56 also suggest that exposure to pollution can suppress early immune responses to the
16
infection, leading to later increases in inflammation and worse prognosis, which may also explain
our findings. Some studies57-59 have suggested that air pollution can also proliferate the
transmission of infectious disease. If COVID-19 spread is indeed impacted by air pollution levels,
which is not yet known, some of the effects detected in our study could be mediated by this factor
as well.
This analysis provides a timely characterization of the relationship between historical exposure to
air pollution and COVID-19 deaths in the United States. Research on how modifiable factors may
exacerbate COVID-19 symptoms and increase mortality risk is essential to guide policies and
behaviors to minimize fatality related to the outbreak. Our analysis relies on up-to-date population-
level COVID-19 data and well-validated air pollution exposure measures.
Strengths of this analysis include adjusting for a wide range of potential confounders and a
demonstrated robustness of results to different model choices. Moreover, the analyses rely
exclusively on data and code that are publicly available. This provides a platform for the scientific
community to continue updating and expanding these analyses as the pandemic evolves and data
accumulate.
It is important to acknowledge that this study has limitations, mainly due to the fact that this is an
ecological study with data available at the county level and that this is a cross-sectional study.
High quality, nationwide individual-level COVID-19 outcome data are unavailable at this time
and for the foreseeable future, thus necessitating the use of an ecologic study design for these
analyses. Due to the potential for ecologic bias, our results should be interpreted in the context of
17
this design and should not be used to make individual-level inferential statements. Also,
unmeasured confounding bias is a threat to the validity of our conclusions. Unfortunately, in the
midst of a pandemic it is not feasible to design a study and collect the data at the ideal level of
spatial and temporal resolution to minimize all sources of bias. Yet, conditional on the data
available, we have endeavored to adjust for confounding bias by all of the most important factors,
including population density, time since the beginning of the outbreak, social isolation measures,
behavior, weather, age structure, ethnicity, access to health care, and socio-economic factors. We
also conducted 68 additional analyses to assess the robustness of the results to many modelling
choices. Furthermore, we computed the E-value to demonstrate that the confounding effect of any
unmeasured confounder would need to be much stronger than that of any of our observed
confounders in order to explain away the relationship between PM2.5 exposure and COVID-19
mortality rate. The calculation of the E-value provided reassurance that the presence of a strong
unmeasured confounder is unlikely; however, this possibility cannot be ruled out completely.
The inability to accurately quantify the number of COVID-19 cases due to limited testing capacity
presents another potential limitation. We instead used total population size as the denominator for
our mortality rates, and we additionally adjusted our models for numerous anticipated proxies of
outbreak size, including time since first reported COVID-19 case, time since stay-at-home order
was issued, and population density.
To conduct the most rigorous possible studies of air pollution and health using ecologic data, it is
critical to utilize areal units that minimize within-area exposure variability and maximize between-
area exposure variability.60 61 We anticipated that our use of counties satisfies this criterion,
18
because counties generally represent meaningful boundaries between urban, suburban, and rural
areas. These population density-related delineations also often correspond to steep gradients in air
pollution levels, thus maximizing across-unit exposure variability while minimizing within-unit
variability. We also note that the use of long-term county-level exposure data in our study likely
led to some degree of exposure misclassification. However, previous literature has found that using
sub-county scale PM2.5 exposure in studies of mortality tends to either have no impact or to increase
the strength of the associations between PM2.5 and mortality from various causes.62
Because of the many limitations, this study also provides justification for expanded follow-up
investigations as more and higher-quality COVID-19 data become available. Such studies would
include validation of our findings with other data sources and study types, as well as studies of
biological mechanisms, impacts of PM2.5 exposure timing, and relationships between PM2.5 and
other COVID-19 outcomes such as hospitalization. The results of this study also underscore the
importance of continuing to enforce existing air pollution regulations. Based on our results, we
anticipate a failure to do so could potentially increase the long-term COVID-19 death toll and
hospitalizations, as well as further burden our healthcare system with other PM2.5-related death
and disease that would draw resources away from COVID-19 patients.
Acknowledgments: The computations in this paper were run on (1) the Odyssey cluster supported
by the FAS Division of Science, Research Computing Group at Harvard University; and (2) the
Research Computing Environment supported by the Institute for Quantitative Social Science in
the Faculty of Arts and Sciences at Harvard University. The authors would like to thank Lena
Goodwin and Stacey Tobin for editorial assistance in the preparation of this manuscript.
19
Funding: This work was made possible by the support from the NIH grant R01 ES024332-01A1,
P50MD010428, ES024012, ES026217, ES028033; MD012769, HEI grant 4953-RFA14-3/16-4,
and USEPA grants 83587201-0, RD-83479801, 1R01ES030616, 1R01AG066793-01R01. The
funding sources did not participate in the design or conduct of the study; collection, management,
analysis, or interpretation of the data; or preparation, review, or approval of the manuscript.
20
Table 1: Publicly available data sources used in the analysis
Source
Data
Outcome: COVID-19
Deaths
Johns Hopkins University the
Center for Systems Science and
Engineering (JHU-CSSE)
Coronavirus Resource Center
(https://coronavirus.jhu.edu/)
County-level COVID-19
death count up to and
including April 22, 2020
Exposure: PM2.5
concentrations
Atmospheric Composition
Analysis Group
(https://sites.wustl.edu/acag/)
0.01° × 0.01° grid resolution
PM2.5 prediction, averaged
across the period 20002016
and averaged across grid cells
in each county
Confounders for main
analysis
US Census/American
Community Survey
(https://www.census.gov/progra
ms-surveys/acs/data.html)
County-level socioeconomic
and demographic variables
for 20122016
Robert Wood Johnson
Foundation County Health
Rankings
(https://www.countyhealthranki
ngs.org/)
County-level behavioral risk
factor variables for 2020
JHU-CSSE Coronavirus
Resource Center
Time since first reported
COVID-19 case
Raifman et al, Boston
University School of Public
Health, COVID-19 United
States state policy database
(www.tinyurl.com/statepolicies)
Time since issuance of stay-
at-home order
Homeland Infrastructure
Foundation-Level Data (HIFLD)
(https://hifld-
geoplatform.opendata.arcgis.co
m/datasets/hospitals)
County-level number of
hospital beds in 2019
Gridmet via Google Earth
engine
(https://developers.google.com/e
arth-
engine/datasets/catalog/IDAHO
_EPSCOR_GRIDMET)
4 km × 4 km temperature and
relative humidity predictions,
summer and winter averaged
across the period 20002016
and averaged across grid cells
in each county
21
Additional confounders for
secondary analyses
The COVID tracking project
(https://covidtracking.com/)
State level number of
COVID-19 tests performed
up to and including April 22,
2020
Carnegie Mellon University
Delphi Research Center
(https://covid-
survey.dataforgood.fb.com/)
Estimated percentage of
people with COVID-19
symptoms, based on survey
data
22
Table 2: Characteristics of the study cohort up to and including April 22, 2020, mean
(standard deviation)
Total
3,087 counties
PM2.5 <8 𝜇g/m3
1,217 counties
PM2.58 𝜇g/m3
1,870 counties
3.4 (10.6)
1.6 (5.7)
4.7 (12.7)
8.4 (2.5)
5.7 (1.4)
10.1 (1.2)
242 (391.9)
300 (515.2)
204.2 (278)
23.6 (10.7)
19 (12.6)
26.5 (7.9)
18.3 (12.4)
16.7 (13.6)
19.2 (11.4)
17.4 (3.5)
15.8 (3.1)
18.5 (3.4)
32.9 (5.4)
31.2 (5.1)
34 (5.3)
10.5 (5.7)
9.7 (5.7)
11.1 (5.6)
21.2 (10.4)
16.5 (8.7)
24.2 (10.3)
74.2 (8.8)
76 (7.7)
73.1 (9.3)
7.6 (12.3)
9.7 (13.7)
6.3 (11.1)
8.2 (14.2)
1 (1.8)
12.9 (16.5)
16 (4.1)
17.4 (4.5)
15 (3.4)
26.4 (3)
26.9 (3.8)
26.1 (2.4)
37.6 (6.5)
35.2 (8.2)
39.2 (4.5)
406.7 (1732.6)
132.6 (430.7)
585.1 (2180.6)
49 (13.1)
50.5 (10.9)
48 (14.3)
136 (89.4)
140.4 (87.3)
133.1 (90.6)
86 (5.7)
83.7 (6.7)
87.4 (4.4)
45.1 (11.9)
39.4 (11.5)
48.7 (10.7)
89 (9.6)
83.2 (11.5)
92.8 (5.5)
87.5 (4.8)
87.9 (5.6)
87.2 (4.1)
23
Table 3: Mortality rate ratios (MRR), 95% confidence intervals (CI), and P-values for all
variables in the main analysis.
MRR
95% CI
P-value
PM2.5 (𝜇g/m3)
1.08
(1.02, 1.15)
0.01
Population density (Q2)
0.86
(0.60, 1.23)
0.40
Population density (Q3)
0.58
(0.40, 0.82)
0.00
Population density (Q4)
0.47
(0.33, 0.68)
0.00
Population density (Q5)
0.52
(0.35, 0.77)
0.00
% Poverty
1.02
(0.93, 1.13)
0.65
log(Median house value)
1.17
(0.99, 1.39)
0.06
log(Median household income)
1.28
(1.09, 1.51)
0.00
% Owner-occupied housing
1.12
(1.02, 1.23)
0.18
% Less than high school education
1.36
(1.21, 1.52)
0.00
% Black
1.45
(1.32, 1.60)
0.00
% Hispanic
1.00
(0.89, 1.12)
0.99
% 65 years of age
1.15
(0.99, 1.33)
0.07
% 15-44 years of age
0.93
(0.74, 1.17)
0.54
% 45-64 years of age
0.96
(0.83, 1.12)
0.62
Days since stay-at-home order
1.28
(0.97, 1.70)
0.08
Days since first case
2.96
(2.50, 3.51)
0.00
Rate of hospital beds
1.12
(1.02, 1.23)
0.01
% Obese
0.94
(0.86, 1.02)
0.14
% Smokers
1.08
(0.92, 1.26)
0.36
Average summer temperature (°F)
0.96
(0.79, 1.16)
0.68
Average winter temperature (°F)
1.18
(0.90, 1.53)
0.22
Average summer relative humidity (%)
0.84
(0.71, 1.01)
0.07
Average winter relative humidity (%)
1.00
(0.89, 1.13)
0.99
24
Fig 1: Maps show (a) county-level 17-year long-term average of PM2.5 concentrations (2000
25
2016) in the United States in 𝜇g/m3, and (b) county-level number of COVID-19 deaths per 1
million population in the United States up to and including April 22, 2020.
26
Fig 2: Mortality Risk Ratios (MRR) and 95% confidence intervals. Upper panel, MRR can be
interpreted as percentage increase in the COVID-19 death rate associated with a 1 𝜇g/m3 increase
in long-term average PM2.5 exposure. The MRR from the main analysis was adjusted for 20
potential confounders. In addition to the main analysis, results are shown for secondary analyses
(a) excluding the counties in New York metropolitan area, (b) excluding counties with fewer than
Main Analysis Exclude NY Metro Exclude Confirmed < 10 Omit Day since 1st Case Add # Tested Add COVID−like Symptoms
0.9
1.0
1.1
1.2
Mortality Rate Ratios
Q1 (0−5.79) Q2 (5.79−8.06) Q3 (8.06−9.54) Q4 (9.54−10.74) Q5 (10.74+)
1.0
1.5
2.0
2.5
3.0
Mortality Rate Ratios
27
10 confirmed COVID-19 cases, (c) omitting time since first reported COVID-19 case from the
model, (d) adding state-level number of tests performed to the model, (e) adding county-level
estimated percentage of people with COVID-19 symptoms to the model, and (f) using PM2.5
exposure categorized at quintiles. All COVID-19 death counts are cumulative counts up to and
including April 22, 2020. Lower panel, MRR can be interpreted as the percentage increase in the
COVID-19 death rate associated with each empirical quintile of long-term average PM2.5 exposure
compared to the baseline quintile (Q1).
28
References
1. Onder G, Rezza G, Brusaferro S. Case-fatality rate and characteristics of patients dying in
relation to COVID-19 in Italy. JAMA 2020 doi: 10.1001/jama.2020.4683 [published
Online First: 2020/03/24]
2. Centers for Disease Control and Prevention Covid-19 Response Team. Severe outcomes
among patients with Coronavirus Disease 2019 (COVID-19) - United States, February
12-March 16, 2020. MMWR Morb Mortal Wkly Rep 2020;69(12):343-46. doi:
10.15585/mmwr.mm6912e2 [published Online First: 2020/03/28]
3. Guan WJ, Ni ZY, Hu Y, et al. Clinical characteristics of coronavirus disease 2019 in
China. N Engl J Med 2020 doi: 10.1056/NEJMoa2002032 [published Online First:
2020/02/29]
4. Wu JT, Leung K, Bushman M, et al. Estimating clinical severity of COVID-19 from the
transmission dynamics in Wuhan, China. Nat Med 2020;26(4):506-10. doi:
10.1038/s41591-020-0822-7 [published Online First: 2020/04/15]
5. Centers for Disease Control and Prevention. National Vital Statistics System. Provisional
Death Counts for Coronavirus Disease (COVID-19) Available from:
https://www.cdc.gov/nchs/nvss/vsrr/COVID19/.
6. Garg S, Kim L, Whitaker M, et al. Hospitalization rates and characteristics of patients
hospitalized with laboratory-confirmed Coronavirus Disease 2019 - COVID-NET, 14
States, March 1-30, 2020. MMWR Morb Mortal Wkly Rep 2020;69(15):458-64. doi:
10.15585/mmwr.mm6915e3 [published Online First: 2020/04/17]
7. Zhou F, Yu T, Du R, et al. Clinical course and risk factors for mortality of adult
inpatients with COVID-19 in Wuhan, China: a retrospective cohort study. Lancet
29
2020;395(10229):1054-62. doi: 10.1016/S0140-6736(20)30566-3 [published Online
First: 2020/03/15]
8. Chen L, Liu HG, Liu W, et al. [Analysis of clinical features of 29 patients with 2019
novel coronavirus pneumonia]. Zhonghua Jie He He Hu Xi Za Zhi 2020;43(0):E005. doi:
10.3760/cma.j.issn.1001-0939.2020.0005 [published Online First: 2020/02/07]
9. Huang C, Wang Y, Li X, et al. Clinical features of patients infected with 2019 novel
coronavirus in Wuhan, China. Lancet 2020;395(10223):497-506. doi: 10.1016/S0140-
6736(20)30183-5 [published Online First: 2020/01/28]
10. Mehta P, McAuley DF, Brown M, et al. COVID-19: consider cytokine storm syndromes
and immunosuppression. Lancet 2020;395(10229):1033-34. doi: 10.1016/S0140-
6736(20)30628-0 [published Online First: 2020/03/21]
11. Ruan Q, Yang K, Wang W, et al. Clinical predictors of mortality due to COVID-19 based
on an analysis of data of 150 patients from Wuhan, China. Intensive Care Med 2020 doi:
10.1007/s00134-020-05991-x [published Online First: 2020/03/04]
12. Wang D, Hu B, Hu C, et al. Clinical characteristics of 138 hospitalized patients with
2019 novel coronavirus-infected pneumonia in Wuhan, China. JAMA 2020 doi:
10.1001/jama.2020.1585 [published Online First: 2020/02/08]
13. Xu Z, Shi L, Wang Y, et al. Pathological findings of COVID-19 associated with acute
respiratory distress syndrome. Lancet Respir Med 2020;8(4):420-22. doi: 10.1016/S2213-
2600(20)30076-X [published Online First: 2020/02/23]
14. Diamond M, Peniston Feliciano HL, Sanghavi D, et al. Acute Respiratory Distress
Syndrome (ARDS) [Updated 2020 Jan 5]. StatPearls [Internet]. Treasure Island, FL:
30
StatPearls Publishing 2020. Available from
https://www.ncbi.nlm.nih.gov/books/NBK436002/.
15. Shi S, Qin M, Shen B, et al. Association of cardiac injury with mortality in hospitalized
patients with COVID-19 in Wuhan, China. JAMA Cardiol 2020 doi:
10.1001/jamacardio.2020.0950 [published Online First: 2020/03/27]
16. Klok FA, Kruip M, van der Meer NJM, et al. Incidence of thrombotic complications in
critically ill ICU patients with COVID-19. Thromb Res 2020 doi:
10.1016/j.thromres.2020.04.013 [published Online First: 2020/04/16]
17. Health Effects Institute. State of Global Air 2019 2019 Available from:
www.stateofglobalair.org accessed April 23 2020.
18. Brook RD, Franklin B, Cascio W, et al. Air pollution and cardiovascular disease: a
statement for healthcare professionals from the Expert Panel on Population and
Prevention Science of the American Heart Association. Circulation 2004;109(21):2655-
71. doi: 10.1161/01.CIR.0000128587.30041.C8 [published Online First: 2004/06/03]
19. Ciencewicki J, Jaspers I. Air pollution and respiratory viral infection. Inhal Toxicol
2007;19(14):1135-46. doi: 10.1080/08958370701665434 [published Online First:
2007/11/08]
20. Di Q, Wang Y, Zanobetti A, et al. Air pollution and mortality in the Medicare population.
N Engl J Med 2017;376(26):2513-22. doi: 10.1056/NEJMoa1702747 [published Online
First: 2017/06/29]
21. Dominici F, Peng RD, Bell ML, et al. Fine particulate air pollution and hospital
admission for cardiovascular and respiratory diseases. JAMA 2006;295(10):1127-34. doi:
10.1001/jama.295.10.1127 [published Online First: 2006/03/09]
31
22. Puett RC, Hart JE, Yanosky JD, et al. Chronic fine and coarse particulate exposure,
mortality, and coronary heart disease in the Nurses' Health Study. Environ Health
Perspect 2009;117(11):1697-701. doi: 10.1289/ehp.0900572 [published Online First:
2010/01/06]
23. Sram RJ, Binkova B, Dejmek J, et al. Ambient air pollution and pregnancy outcomes: a
review of the literature. Environ Health Perspect 2005;113(4):375-82. doi:
10.1289/ehp.6362 [published Online First: 2005/04/07]
24. Wellenius GA, Burger MR, Coull BA, et al. Ambient air pollution and the risk of acute
ischemic stroke. Arch Intern Med 2012;172(3):229-34. doi:
10.1001/archinternmed.2011.732 [published Online First: 2012/02/15]
25. Rhee J, Dominici F, Zanobetti A, et al. Impact of long-term exposures to ambient PM2.5
and ozone on ARDS risk for older adults in the United States. Chest 2019;156(1):71-79.
doi: 10.1016/j.chest.2019.03.017 [published Online First: 2019/03/31]
26. United States Environmental Protection Agency. Integrated Science Assessment (ISA)
for Particulate Matter (Final Report, 2019). EPA/600/R-19/188. Washington, DC: US
EPA 2019.
27. Brook RD, Rajagopalan S, Pope CA, et al. Particulate matter air pollution and
cardiovascular disease. Circulation 2010;121(21):2331-78. doi:
doi:10.1161/CIR.0b013e3181dbece1
28. Pope CA, Burnett RT, Thurston GD, et al. Cardiovascular mortality and long-term
exposure to particulate air pollution. Circulation 2004;109(1):71-77. doi:
doi:10.1161/01.CIR.0000108927.80044.7F
32
29. Pope CA, Coleman N, Pond ZA, et al. Fine particulate air pollution and human mortality:
25+ years of cohort studies. Environ Res 2020;183:108924. doi:
https://doi.org/10.1016/j.envres.2019.108924
30. Kioumourtzoglou MA, Schwartz JD, Weisskopf MG, et al. Long-term PM2.5 exposure
and neurological hospital admissions in the Northeastern United States. Environ Health
Perspect 2016;124(1):23-9. doi: 10.1289/ehp.1408973 [published Online First:
2015/05/16]
31. Zanobetti A, Dominici F, Wang Y, et al. A national case-crossover analysis of the short-
term effect of PM2.5 on hospitalizations and mortality in subjects with diabetes and
neurological disorders. Environ Health 2014;13(1):38. doi: 10.1186/1476-069X-13-38
[published Online First: 2014/06/03]
32. Dong E, Du H, Gardner L. An interactive web-based dashboard to track COVID-19 in
real time. Lancet Infect Dis 2020 doi: 10.1016/S1473-3099(20)30120-1 [published
Online First: 2020/02/23]
33. van Donkelaar A, Martin RV, Li C, et al. Regional estimates of chemical composition of
fine particulate matter using a combined geoscience-statistical method with information
from satellites, models, and monitors. Environ Sci Technol 2019;53(5):2595-611. doi:
10.1021/acs.est.8b06392 [published Online First: 2019/01/31]
34. Booth JG, Casella G, Friedl H, et al. Negative binomial loglinear mixed models. Stat
Modelling 2003;3(3):179-91. doi: 10.1191/1471082X03st058oa
35. R Core Team. R: A language and environment for statistical computing. Vienna, Austria:
R Foundation for Statistical Computing 2020. Available from: https://www.R-
project.org/.
33
36. Zhang X, Mallick H, Tang Z, et al. Negative binomial mixed models for analyzing
microbiome count data. BMC Bioinformatics 2017;18(1):4. doi: 10.1186/s12859-016-
1441-7 [published Online First: 2017/01/05]
37. Bates D, Machler M, Bolker B, et al. Fitting linear mixed-effects models using lme4. J
Stat Software 2015;67(1) doi: 10.18637/jss.v067.i01
38. Bolker B. glmer.nb Fitting Negative Binomial GLMMs Available from:
https://www.rdocumentation.org/packages/lme4/versions/1.1-23/topics/glmer.nb.
39. VanderWeele TJ, Ding P. Sensitivity analysis in observational research: introducing the
E-value. Ann Intern Med 2017;167(4):268-74. doi: 10.7326/M16-2607 [published Online
First: 2017/07/12]
40. Yancy CW. COVID-19 and African Americans. JAMA 2020 doi:
10.1001/jama.2020.6548 [published Online First: 2020/04/16]
41. Cui Y, Zhang ZF, Froines J, et al. Air pollution and case fatality of SARS in the People's
Republic of China: an ecologic study. Environ Health 2003;2(1):15. doi: 10.1186/1476-
069X-2-15 [published Online First: 2003/11/25]
42. Pope CA, 3rd. Respiratory disease associated with community air pollution and a steel
mill, Utah Valley. Am J Public Health 1989;79(5):623-8. doi: 10.2105/ajph.79.5.623
[published Online First: 1989/05/01]
43. Croft DP, Zhang W, Lin S, et al. Associations between source-specific particulate matter
and respiratory infections in New York state adults. Environ Sci Technol 2020;54(2):975-
84. doi: 10.1021/acs.est.9b04295 [published Online First: 2019/11/23]
44. Horne BD, Joy EA, Hofmann MG, et al. Short-term elevation of fine particulate matter
air pollution and acute lower respiratory infection. Am J Respir Crit Care Med
34
2018;198(6):759-66. doi: 10.1164/rccm.201709-1883OC [published Online First:
2018/04/14]
45. Di Q, Dai L, Wang Y, et al. Association of short-term exposure to air pollution with
mortality in older adults. JAMA 2017;318(24):2446-56. doi: 10.1001/jama.2017.17923
[published Online First: 2017/12/28]
46. Tsai DH, Riediker M, Berchet A, et al. Effects of short- and long-term exposures to
particulate matter on inflammatory marker levels in the general population. Environ Sci
Pollut Res Int 2019;26(19):19697-704. doi: 10.1007/s11356-019-05194-y [published
Online First: 2019/05/13]
47. Morales KF, Paget J, Spreeuwenberg P. Possible explanations for why some countries
were harder hit by the pandemic influenza virus in 2009 - a global mortality impact
modeling study. BMC Infect Dis 2017;17(1):642. doi: 10.1186/s12879-017-2730-0
[published Online First: 2017/09/28]
48. Xu Z, Hu W, Williams G, et al. Air pollution, temperature and pediatric influenza in
Brisbane, Australia. Environ Int 2013;59:384-8. doi: 10.1016/j.envint.2013.06.022
[published Online First: 2013/08/06]
49. Clay K, Lewis J, Severnini E. Pollution, infectious disease, and mortality: evidence from
the 1918 Spanish influenza pandemic. J Econ Hist 2018;78(4):1179-209. doi:
10.1017/S002205071800058X [published Online First: 2018/10/02]
50. Clay K, Lewis J, Severnini E. What explains cross-city variation in mortality during the
1918 influenza pandemic? Evidence from 438 U.S. cities. Econ Hum Biol 2019;35:42-50.
doi: 10.1016/j.ehb.2019.03.010 [published Online First: 2019/05/10]
35
51. Pope CA, 3rd, Bhatnagar A, McCracken JP, et al. Exposure to fine particulate air
pollution is associated with endothelial injury and systemic inflammation. Circ Res
2016;119(11):1204-14. doi: 10.1161/CIRCRESAHA.116.309279 [published Online
First: 2016/10/27]
52. Conticini E, Frediani B, Caro D. Can atmospheric pollution be considered a co-factor in
extremely high level of SARS-CoV-2 lethality in Northern Italy? Environ Pollut
2020:114465. doi: 10.1016/j.envpol.2020.114465 [published Online First: 2020/04/10]
53. Becker S, Soukup JM. Exposure to urban air particulates alters the macrophage-mediated
inflammatory response to respiratory viral infection. J Toxicol Environ Health A
1999;57(7):445-57. doi: 10.1080/009841099157539 [published Online First: 1999/09/24]
54. Kaan PM, Hegele RG. Interaction between respiratory syncytial virus and particulate
matter in guinea pig alveolar macrophages. Am J Respir Cell Mol Biol 2003;28(6):697-
704. doi: 10.1165/rcmb.2002-0115OC [published Online First: 2003/05/23]
55. Lambert AL, Mangum JB, DeLorme MP, et al. Ultrafine carbon black particles enhance
respiratory syncytial virus-induced airway reactivity, pulmonary inflammation, and
chemokine expression. Toxicol Sci 2003;72(2):339-46. doi: 10.1093/toxsci/kfg032
[published Online First: 2003/03/26]
56. Lambert AL, Trasti FS, Mangum JB, et al. Effect of preexposure to ultrafine carbon black
on respiratory syncytial virus infection in mice. Toxicol Sci 2003;72(2):331-8. doi:
10.1093/toxsci/kfg031 [published Online First: 2003/03/28]
57. Chen PS, Tsai FT, Lin CK, et al. Ambient influenza and avian influenza virus during dust
storm days and background days. Environ Health Perspect 2010;118(9):1211-6. doi:
10.1289/ehp.0901782 [published Online First: 2010/05/04]
36
58. Peng L, Zhao X, Tao Y, et al. The effects of air pollution and meteorological factors on
measles cases in Lanzhou, China. Environ Sci Pollut Res Int 2020;27(12):13524-33. doi:
10.1007/s11356-020-07903-4 [published Online First: 2020/02/08]
59. Ye Q, Fu JF, Mao JH, et al. Haze is a risk factor contributing to the rapid spread of
respiratory syncytial virus in children. Environ Sci Pollut Res Int 2016;23(20):20178-85.
doi: 10.1007/s11356-016-7228-6 [published Online First: 2016/07/22]
60. Elliott P, Savitz DA. Design issues in small-area studies of environment and health.
Environ Health Perspect 2008;116(8):1098-104. doi: 10.1289/ehp.10817 [published
Online First: 2008/08/19]
61. Kunzli N, Tager IB. The semi-individual study in air pollution epidemiology: a valid
design as compared to ecologic studies. Environ Health Perspect 1997;105(10):1078-83.
doi: 10.1289/ehp.105-1470382 [published Online First: 1997/12/24]
62. Crouse DL, Erickson AC, Christidis T, et al. Evaluating the sensitivity of PM2.5-
mortality associations to the spatial and temporal scale of exposure assessment.
Epidemiology 2020;31(2):168-76. doi: 10.1097/EDE.0000000000001136 [published
Online First: 2019/11/07]
... Early in the pandemic, researchers reported associations of air pollutants with COVID-19 mortality using ecological study designs (Konstantinoudis et al., 2021;Lipsitt et al., 2021;Ogen, 2020;Wu et al., 2020aWu et al., , 2020b. Ecological studies are valid for hypothesis generating purposes but cannot be used to make inferences about individual risks (Villeneuve and Goldberg, 2020). ...
Article
Full-text available
Individual-level studies with adjustment for important COVID-19 risk factors suggest positive associations of long-term air pollution exposure (particulate matter and nitrogen dioxide) with COVID-19 infection, hospitalisations and mortality. The evidence, however, remains limited and mechanisms unclear. We aimed to investigate these associations within UK Biobank, and to examine the role of underlying chronic disease as a potential mechanism. UK Biobank COVID-19 positive laboratory test results were ascertained via Public Health England and general practitioner record linkage, COVID-19 hospitalisations via Hospital Episode Statistics, and COVID-19 mortality via Office for National Statistics mortality records from March-December 2020. We used annual average outdoor air pollution modelled at 2010 residential addresses of UK Biobank participants who resided in England (n = 424,721). We obtained important COVID-19 risk factors from baseline UK Biobank questionnaire responses (2006-2010) and general practitioner record linkage. We used logistic regression models to assess associations of air pollution with COVID-19 outcomes, adjusted for relevant confounders, and conducted sensitivity analyses. We found positive associations of fine particulate matter (PM 2.5) and nitrogen dioxide (NO 2) with COVID-19 positive test result after adjustment for confounders and COVID-19 risk factors, with odds ratios of 1.05 (95% confidence intervals (CI) = 1.02, 1.08), and 1.05 (95% CI = 1.01, 1.08), respectively. PM 2.5 and NO 2 were positively associated with COVID-19 hospitalisations and deaths in minimally adjusted models, but not in fully adjusted models. No associations for PM 10 were found. In analyses with additional adjustment for pre-existing chronic disease, effect estimates were not substantially attenuated, indicating that underlying chronic disease may not fully explain associations. We found some evidence that long-term exposure to PM 2.5 and NO 2 was associated with a COVID-19 positive test result in UK Biobank, though not with COVID-19 hospitalisations or deaths.
... Besides, several studies Chakrabarty et al. 2020;Laxmipriya and Narayanan 2021) have tried to find out the direct relation between PM 2.5 and COVID-19-positive cases in different areas. According to the estimation of Wu et al. (2020), about 8% COVID-19 death rate has been increased with the only rise of only 1 µg/m 3 PM 2.5 . PM 2.5 also has a great impact on visibility and put adverse effects on the environment and also poses a great impact on human health especially affecting the lung and heart of the human body, making lung cancer like deadly disease (Hu and Jiang 2014) that can increase the mortality and morbidity by weakening the immune system (Glencross et al. 2020). ...
Article
Full-text available
Modern human civilization has suffered from the disastrous impact of COVID-19, but it teaches us the lesson that the environment can restore its stability without human activity. The government of India (GOI) has launched many strategies to prevent the situation of COVID-19, including a lockdown that has a great impact on the environment. The present study focuses on the analysis of PM2.5 concentration levels in pre-locking, lockdown, and unlocking phases across ten major 10 cities of Maharashtra (MH) that were the COVID hotspot of India during the COVID-19 outbreak; Phase-wise and year-wise (2018-2020) hotspot analysis, box diagram and line graph methods were used to assess spatial variation in PM2.5 across MH cities. Our study showed that The PM2.5 concentration level was severe at pre lockdown stage (January- March) and it decrease dramatically at the lockdown stage, later it also increase in its previous position at the unlocking stages i.e. PM2.5 decreased dramatically (59%) during the lockdown period compared to the pre-lockdown period due to the shutdown of outdoor activities. It returns to its previous position due to the unlocking situation and increases (70%) compared to the lockdown period which illustrated the ups and downs of PM2.5 and ensures the position of different cities in the AQI categories at different times. In the pre-lockdown phase maximum PM2.5 concentration was in NAV (358) and MUM (338), PUN (335), NAS (325) subsequently whereas at the last of the lockdown phase it becomes CHN (82), NAG (76), and SOL (45) subsequently. Hence, the restoration of the environment during the lockdown phase was temporary rather than permanent. Therefore, our findings propose that several effective policies of government such as relocation of polluting industries, short-term lockdown, odd-even vehicle number, installation of air purifiers, and government strict initiatives are needed in making a sustainable environment. Keywords: PM2.5; Sustainable; Hotspot; lockdown; COVID-19
... Nevertheless, the findings of this study support the notion that air pollution may play a role in SARS-CoV-2 infections [204]. The findings of this study, which has been replicated in Italy (Conticini et al., 2020) and in the United States , suggest that PM 2.5 leads to a large increase in COVID death rate, suggesting that persistent exposure to air pollution hinders recovery and leads to more severe and deadly types of illness [203,205]. Coccia (2020) investigated the mechanisms of COVID-19 transmission dynamics in the environment to determine a feasible approach for dealing with future epidemics comparable to coronavirus infections. Their research focused on a case study of Italy, which has one of the highest rates of mortality in the world. ...
Article
Full-text available
Several epidemiologic and toxicological studies have commonly viewed ambient fine particulate matter (PM2.5), defined as particles having an aerodynamic diameter of less than 2.5 µm, as a significant potential danger to human health. PM2.5 is mostly absorbed through the respiratory system, where it can infiltrate the lung alveoli and reach the bloodstream. In the respiratory system, reactive oxygen or nitrogen species (ROS, RNS) and oxidative stress stimulate the generation of mediators of pulmonary inflammation and begin or promote numerous illnesses. According to the most recent data, fine particulate matter, or PM2.5, is responsible for nearly 4 million deaths globally from cardiopulmonary illnesses such as heart disease, respiratory infections, chronic lung disease, cancers, preterm births, and other illnesses. There has been increased worry in recent years about the negative impacts of this worldwide danger. The causal associations between PM2.5 and human health, the toxic effects and potential mechanisms of PM2.5, and molecular pathways have been described in this review.
... De esta manera, una sindemia puede deberse a la interacción toxicocinética entre dos o más sustancias presentes en una mezcla de contaminantes o la interacción entre dos o más amenazas, ya sean físicas, químicas o biológicas. Pongamos el caso de la interacción entre las partículas PM2.5 con el SARS-CoV-2, la cual genera una sindemia que afecta las vías respiratorias, ya que las PM2.5 incrementan la gravedad de la COVID-19 13 . Esta sindemia se ha extendido a químicos y población infantil 14 , aumentando la vulnerabilidad debido a la relación que existe entre el virus y las comorbilidades como la obesidad, la diabetes y la hipertensión. ...
Article
Actualmente, el mundo enfrenta una pandemia y retos globales como el cambio climático, la pobreza, el hambre, la pérdida de la biodiversidad, la acidificación de los océanos y las desigualdades. En este contexto, se infiere que las sociedades hayan olvidado un problema antiguo y preocupante, la contaminación en todas sus formas. Es así entonces, que debe establecerse un plan para afrontar los efectos de la contaminación química, física y biológica. Sin embargo, dicha estrategia no debe ignorar la sindemia causada por la interacción de la contaminación con el resto de los problemas ambientales, sociales, políticos, económicos y sanitarios que enfrenta la humanidad. Asimismo, debe contemplar el reto de iniciar la intervención desde la escala local, es decir, desde la comunidad. En este trabajo, presentamos las conclusiones de una serie de colaboraciones que hemos publicado en esta revista. Así como una estrategia generada para intervenir procesos contaminantes a nivel comunitario en regiones de alta vulnerabilidad que hemos definido como Escenarios Humanitarios, incluyendo no solamente a las comunidades humanas, sino al resto de los seres vivos de la naturaleza (esquema de Salud Total). Esta iniciativa ha sido denominada como Sistema de Vigilancia Integrada para Comunidades Contaminadas (SIVICCO), ya que incorpora un conjunto de conceptos y herramientas para prevenir los efectos negativos de la contaminación en todas sus formas y en todas las comunidades. SIVICCO representa una estrategia integrada, pero sobre todo, la expresión de un nuevo civismo basado en la participación activa del colectivo comunitario con una perspectiva de derechos humanos.
Article
Recent literature has suggested a link between poor air quality and worse COVID-19 outcomes. In the United States, this link is particularly noteworthy because of residential sorting along ethnic lines within the US population; minorities are disproportionately exposed to health hazards, including air pollution. The impacts of the COVID-19 pandemic have also been disproportionately concentrated amongst minorities. We explore the association between air quality and COVID-19 outcomes, using county level data for the United States from the first wave of the pandemic in 2020, and test whether exposure to more polluted air can account for some of the observed disparities in COVID-19 outcomes among minorities.
Conference Paper
Full-text available
Since the early days of the Covid-19 crisis, the scientific community is constantly trying to clarify various issues such as the mechanisms of the spread of this virus, its environmental and economic impacts, and the necessary policies for recovery and adaptation. Due to the high concentration of population and economic activities in cities, they are often centers prone to contamination and infection with Covid-19. Therefore, many researchers are trying to explore the dynamics of the epidemic in urban areas in order to understand the effects of Covid-19 on cities. On the other hand, the covid-19 pandemic has made the importance of "human resource management" (HRM) in organizations and companies, especially, face the new challenges posed by the corona virus. With an uncertain future and a high probability of job loss. or salary deductions, employees need to be supported more than ever; therefore, employers depend more on their HR department to achieve the right HR strategy that can meet new challenges. This unknown virus in addition to the bitter series Deaths themselves have an impact on the environment.This paper will try to analyze the effects of the covid-19 epidemic on natural and human resources by looking at the available statistics and evidence.
Experiment Findings
Full-text available
The nationwide lockdown in India announced from March 25-May 31, 2020 in four phases to prevent the rapid spread of pandemic COVID-19 has resulted in the reduction of anthropogenic emission sources (vehicular and industries mainly) to a great extent. Present study reports change in air quality during the lockdown period at Dayalbagh, Agra, a semi-urban site in Northern India. The concentration of surface O3, NOx and CO were measured through continuously operating O3 (Thermo Fischer Model 49i), NOx (Thermo Fischer Model 42i) and CO (Teledyne T300) analysers, respectively. The data showed a remarkable reduction in the mean pollutant levels influenced by traffic emissions, that is, Nitric oxide (NO) by 54.1%, Nitrogen Dioxide (NO2) by 86.7% (64.6% (NOx)) during lockdown over same period in the previous year (2019). Comparatively, a lower reduction of Carbon monoxide (CO) 8.6% is attributed to the dominance of natural atmospheric chemical regulation, biogenic sources in addition to anthropogenic contributions and long-life span. An enhancement of secondary pollutant viz. Ozone (O3) 5.1% was observed during lockdown over same period in the previous year. O3 showed same diurnal pattern during lockdown phase as in other phases, while the bi-modal peaks of NO, NO2 (NOx) were supressed due to less vehicular emission and other anthropogenic activities, however CO showed prominent bi-modal peaks during lockdown. The concentration of NO, NO2, NOx and CO reduced by 38.0%, 71.9%, 48.6% and 34.8% respectively during lockdown period in comparison to pre-lockdown period (2020), on contrast O3 concentration increase by 24.5%. While, concentration of NO, NO2, NOx and CO increased by 44.4%, 79.7%, 72.0% and 3.5% respectively during unlock period in comparison to lockdown period, on contrast O3 concentration decreased by 16.2%.
Article
The COVID-19 pandemic has profoundly affected human society on a global scale. COVID-19 pandemic control measures have led to significant changes in nighttime light (NTL) and air quality. Four cities that were severely impacted by the pandemic and that implemented different pandemic control measures, namely, Wuhan (China), Delhi (India), New York (United States), and Rome (Italy), were selected as study areas. The Visible Infrared Imaging Radiometer Suite (VIIRS) and air quality data were used to study the variation characteristics of NTL and air quality in the four cities in 2020. NTL brightness in Wuhan, Delhi, New York, and Rome decreased by 8.88%, 17.18%, 8.21%, and 6.33%, respectively, compared with pre-pandemic levels; in the resumption phase Wuhan and Rome NTL brightness recovered by 13.74% and 3.38%, but Delhi and New York decreased by 16.23% and 4.99%. Nitrogen dioxide (NO2) concentrations in the lockdown periods of Wuhan, Delhi, New York, and Rome decreased by 65.07%, 68.75%, 55.59%, and 56.81%, respectively; PM2.5 decreased by 49.25%, 69.40%, 52.54%, and 66.67%. Air quality improved, but ozone (O3) concentrations increased significantly during the lockdown periods. The methods presented herein can be used to investigate the impact of pandemic control measures on urban lights and air quality.
Article
Full-text available
Hava kirliliği, temelde atmosferin doğal özelliklerinde olumsuz yönde yaşanan değişim şeklinde ifade edilmektedir. Bu değişime yol açan kimyasal, fiziksel ve biyolojik maddeler ise hava kirletici olarak adlandırılmaktadır. Söz konusu maddeleri ortaya çıkaran çeşitli doğal faktörler bulunsa da, bu maddeler ağırlıklı olarak beşeri faaliyetlerden kaynaklanmaktadır. Covid-19 salgını kapsamında alınan tedbirlerin, bu faaliyetlerde bir daralmaya yol açtığı düşünülmektedir. Dolayısıyla bu çalışma, Türkiye’nin hava kalitesinde Covid-19 salgını döneminde bir değişim olup olmadığını incelemeyi amaçlamaktadır. Bu amaç doğrultusunda Covid-19 öncesi dönem ve Covid-19 dönemini kapsayan Nisan 2019-Mart 2021 tarih aralığında Türkiye’de seçilen illerin hava kalitesi; PM10, O3 , NO2 ve SO2 değişkenleri kullanılarak TOPSIS yöntemi ile analiz edilmiştir. Söz konusu tarih aralığı altı aylık dört dönem şeklinde ele alınmıştır. Değişkenler açısından dönemler arasındaki farklılık, Kruskall-Wallis H ve Tamhane’s T2 testleri ile kontrol edilmiştir. Yapılan analiz sonucunda seçilmiş illerin Covid-19 döneminde hava kalitelerinde kısmi bir iyileşme olduğu tespit edilmiştir. Bu iyileşme üzerinde mevsimsel değişimlerin yanı sıra Covid-19 salgını kapsamında alınan önlemlerin de etkili olduğu sonucuna ulaşılmıştır.
Experiment Findings
Full-text available
The nationwide lockdown in India announced from March 25-May 31, 2020 in four phases to prevent the rapid spread of pandemic COVID-19 has resulted in the reduction of anthropogenic emission sources (vehicular and industries mainly) to a great extent. Present study reports change in air quality during the lockdown period at Dayalbagh, Agra, a semiurban site in Northern India. The concentration of surface O3 , NOx and CO were measured through continuously operating O3 (Thermo Fischer Model 49i), NOx (Thermo Fischer Model 42i) and CO (Teledyne T300) analysers, respectively. The data showed a remarkable reduction in the mean pollutant levels influenced by traffic emissions, that is, Nitric oxide (NO) by 54.1%, Nitrogen Dioxide (NO2) by 86.7% (64.6% (NOx )) during lockdown over same period in the previous year (2019). Comparatively, a lower reduction of Carbon monoxide (CO) 8.6% is attributed to the dominance of natural atmospheric chemical regulation, biogenic sources in addition to anthropogenic contributions and long-life span. An enhancement of secondary pollutant viz. Ozone (O3 ) 5.1% was observed during lockdown over same period in the previous year. O3 showed same diurnal pattern during lockdown phase as in other phases, while the bi-modal peaks of NO, NO2 (NOx ) were supressed due to less vehicular emission and other anthropogenic activities, however CO showed prominent bi-modal peaks during lockdown. The concentration of NO, NO2, NOx and CO reduced by 38.0%, 71.9%, 48.6% and 34.8% respectively during lockdown period in comparison to prelockdown period (2020), on contrast O3 concentration increase by 24.5%. While, concentration of NO, NO2, NOx and CO increased by 44.4%, 79.7%, 72.0% and 3.5% respectively during unlock period in comparison to lockdown period, on contrast O3 concentration decreased by 16.2%.
Article
Full-text available
Since SARS-CoV-2, the novel coronavirus that causes coronavirus disease 2019 (COVID-19), was first detected in December 2019 (1), approximately 1.3 million cases have been reported worldwide (2), including approximately 330,000 in the United States (3). To conduct population-based surveillance for laboratory-confirmed COVID-19-associated hospitalizations in the United States, the COVID-19-Associated Hospitalization Surveillance Network (COVID-NET) was created using the existing infrastructure of the Influenza Hospitalization Surveillance Network (FluSurv-NET) (4) and the Respiratory Syncytial Virus Hospitalization Surveillance Network (RSV-NET). This report presents age-stratified COVID-19-associated hospitalization rates for patients admitted during March 1-28, 2020, and clinical data on patients admitted during March 1-30, 2020, the first month of U.S. surveillance. Among 1,482 patients hospitalized with COVID-19, 74.5% were aged ≥50 years, and 54.4% were male. The hospitalization rate among patients identified through COVID-NET during this 4-week period was 4.6 per 100,000 population. Rates were highest (13.8) among adults aged ≥65 years. Among 178 (12%) adult patients with data on underlying conditions as of March 30, 2020, 89.3% had one or more underlying conditions; the most common were hypertension (49.7%), obesity (48.3%), chronic lung disease (34.6%), diabetes mellitus (28.3%), and cardiovascular disease (27.8%). These findings suggest that older adults have elevated rates of COVID-19-associated hospitalization and the majority of persons hospitalized with COVID-19 have underlying medical conditions. These findings underscore the importance of preventive measures (e.g., social distancing, respiratory hygiene, and wearing face coverings in public settings where social distancing measures are difficult to maintain)† to protect older adults and persons with underlying medical conditions, as well as the general public. In addition, older adults and persons with serious underlying medical conditions should avoid contact with persons who are ill and immediately contact their health care provider(s) if they have symptoms consistent with COVID-19 (https://www.cdc.gov/coronavirus/2019-ncov/symptoms-testing/symptoms.html) (5). Ongoing monitoring of hospitalization rates, clinical characteristics, and outcomes of hospitalized patients will be important to better understand the evolving epidemiology of COVID-19 in the United States and the clinical spectrum of disease, and to help guide planning and prioritization of health care system resources.
Article
Full-text available
As of 29 February 2020 there were 79,394 confirmed cases and 2,838 deaths from COVID-19 in mainland China. Of these, 48,557 cases and 2,169 deaths occurred in the epicenter, Wuhan. A key public health priority during the emergence of a novel pathogen is estimating clinical severity, which requires properly adjusting for the case ascertainment rate and the delay between symptoms onset and death. Using public and published information, we estimate that the overall symptomatic case fatality risk (the probability of dying after developing symptoms) of COVID-19 in Wuhan was 1.4% (0.9–2.1%), which is substantially lower than both the corresponding crude or naïve confirmed case fatality risk (2,169/48,557 = 4.5%) and the approximator1 of deaths/deaths + recoveries (2,169/2,169 + 17,572 = 11%) as of 29 February 2020. Compared to those aged 30–59 years, those aged below 30 and above 59 years were 0.6 (0.3–1.1) and 5.1 (4.2–6.1) times more likely to die after developing symptoms. The risk of symptomatic infection increased with age (for example, at ~4% per year among adults aged 30–60 years). An estimation of the clinical severity of COVID-19, based on the data available so far, can help to inform the public health response during the ongoing SARS-CoV-2 pandemic.
Article
Full-text available
Background: Since December 2019, when coronavirus disease 2019 (Covid-19) emerged in Wuhan city and rapidly spread throughout China, data have been needed on the clinical characteristics of the affected patients. Methods: We extracted data regarding 1099 patients with laboratory-confirmed Covid-19 from 552 hospitals in 30 provinces, autonomous regions, and municipalities in China through January 29, 2020. The primary composite end point was admission to an intensive care unit (ICU), the use of mechanical ventilation, or death. Results: The median age of the patients was 47 years; 41.9% of the patients were female. The primary composite end point occurred in 67 patients (6.1%), including 5.0% who were admitted to the ICU, 2.3% who underwent invasive mechanical ventilation, and 1.4% who died. Only 1.9% of the patients had a history of direct contact with wildlife. Among nonresidents of Wuhan, 72.3% had contact with residents of Wuhan, including 31.3% who had visited the city. The most common symptoms were fever (43.8% on admission and 88.7% during hospitalization) and cough (67.8%). Diarrhea was uncommon (3.8%). The median incubation period was 4 days (interquartile range, 2 to 7). On admission, ground-glass opacity was the most common radiologic finding on chest computed tomography (CT) (56.4%). No radiographic or CT abnormality was found in 157 of 877 patients (17.9%) with nonsevere disease and in 5 of 173 patients (2.9%) with severe disease. Lymphocytopenia was present in 83.2% of the patients on admission. Conclusions: During the first 2 months of the current outbreak, Covid-19 spread rapidly throughout China and caused varying degrees of illness. Patients often presented without fever, and many did not have abnormal radiologic findings. (Funded by the National Health Commission of China and others.).
Article
Introduction COVID-19 may predispose to both venous and arterial thromboembolism due to excessive inflammation, hypoxia, immobilisation and diffuse intravascular coagulation. Reports on the incidence of thrombotic complications are however not available. Methods We evaluated the incidence of the composite outcome of symptomatic acute pulmonary embolism (PE), deep-vein thrombosis, ischemic stroke, myocardial infarction or systemic arterial embolism in all COVID-19 patients admitted to the ICU of 2 Dutch university hospitals and 1 Dutch teaching hospital. Results We studied 184 ICU patients with proven COVID-19 pneumonia of whom 23 died (13%), 22 were discharged alive (12%) and 139 (76%) were still on the ICU on April 5th 2020. All patients received at least standard doses thromboprophylaxis. The cumulative incidence of the composite outcome was 31% (95%CI 20-41), of which CTPA and/or ultrasonography confirmed VTE in 27% (95%CI 17-37%) and arterial thrombotic events in 3.7% (95%CI 0-8.2%). PE was the most frequent thrombotic complication (n = 25, 81%). Age (adjusted hazard ratio (aHR) 1.05/per year, 95%CI 1.004-1.01) and coagulopathy, defined as spontaneous prolongation of the prothrombin time > 3 s or activated partial thromboplastin time > 5 s (aHR 4.1, 95%CI 1.9-9.1), were independent predictors of thrombotic complications. Conclusion The 31% incidence of thrombotic complications in ICU patients with COVID-19 infections is remarkably high. Our findings reinforce the recommendation to strictly apply pharmacological thrombosis prophylaxis in all COVID-19 patients admitted to the ICU, and are strongly suggestive of increasing the prophylaxis towards high-prophylactic doses, even in the absence of randomized evidence.
Article
This paper investigates the correlation between the high level of Severe Acute Respiratory Syndrome CoronaVirus 2 (SARS-CoV-2) lethality and the atmospheric pollution in Northern Italy. Indeed, Lombardy and Emilia Romagna are Italian regions with both the highest level of virus lethality in the world and one of Europe's most polluted area. Based on this correlation, this paper analyzes the possible link between pollution and the development of acute respiratory distress syndrome and eventually death. We provide evidence that people living in an area with high levels of pollutant are more prone to develop chronic respiratory conditions and suitable to any infective agent. Moreover, a prolonged exposure to air pollution leads to a chronic inflammatory stimulus, even in young and healthy subjects. We conclude that the high level of pollution in Northern Italy should be considered an additional co-factor of the high level of lethality recorded in that area.
Article
Importance Coronavirus disease 2019 (COVID-19) has resulted in considerable morbidity and mortality worldwide since December 2019. However, information on cardiac injury in patients affected by COVID-19 is limited. Objective To explore the association between cardiac injury and mortality in patients with COVID-19. Design, Setting, and Participants This cohort study was conducted from January 20, 2020, to February 10, 2020, in a single center at Renmin Hospital of Wuhan University, Wuhan, China; the final date of follow-up was February 15, 2020. All consecutive inpatients with laboratory-confirmed COVID-19 were included in this study. Main Outcomes and Measures Clinical laboratory, radiological, and treatment data were collected and analyzed. Outcomes of patients with and without cardiac injury were compared. The association between cardiac injury and mortality was analyzed. Results A total of 416 hospitalized patients with COVID-19 were included in the final analysis; the median age was 64 years (range, 21-95 years), and 211 (50.7%) were female. Common symptoms included fever (334 patients [80.3%]), cough (144 [34.6%]), and shortness of breath (117 [28.1%]). A total of 82 patients (19.7%) had cardiac injury, and compared with patients without cardiac injury, these patients were older (median [range] age, 74 [34-95] vs 60 [21-90] years; P < .001); had more comorbidities (eg, hypertension in 49 of 82 [59.8%] vs 78 of 334 [23.4%]; P < .001); had higher leukocyte counts (median [interquartile range (IQR)], 9400 [6900-13 800] vs 5500 [4200-7400] cells/μL) and levels of C-reactive protein (median [IQR], 10.2 [6.4-17.0] vs 3.7 [1.0-7.3] mg/dL), procalcitonin (median [IQR], 0.27 [0.10-1.22] vs 0.06 [0.03-0.10] ng/mL), creatinine kinase–myocardial band (median [IQR], 3.2 [1.8-6.2] vs 0.9 [0.6-1.3] ng/mL), myohemoglobin (median [IQR], 128 [68-305] vs 39 [27-65] μg/L), high-sensitivity troponin I (median [IQR], 0.19 [0.08-1.12] vs <0.006 [<0.006-0.009] μg/L), N-terminal pro-B-type natriuretic peptide (median [IQR], 1689 [698-3327] vs 139 [51-335] pg/mL), aspartate aminotransferase (median [IQR], 40 [27-60] vs 29 [21-40] U/L), and creatinine (median [IQR], 1.15 [0.72-1.92] vs 0.64 [0.54-0.78] mg/dL); and had a higher proportion of multiple mottling and ground-glass opacity in radiographic findings (53 of 82 patients [64.6%] vs 15 of 334 patients [4.5%]). Greater proportions of patients with cardiac injury required noninvasive mechanical ventilation (38 of 82 [46.3%] vs 13 of 334 [3.9%]; P < .001) or invasive mechanical ventilation (18 of 82 [22.0%] vs 14 of 334 [4.2%]; P < .001) than those without cardiac injury. Complications were more common in patients with cardiac injury than those without cardiac injury and included acute respiratory distress syndrome (48 of 82 [58.5%] vs 49 of 334 [14.7%]; P < .001), acute kidney injury (7 of 82 [8.5%] vs 1 of 334 [0.3%]; P < .001), electrolyte disturbances (13 of 82 [15.9%] vs 17 of 334 [5.1%]; P = .003), hypoproteinemia (11 of 82 [13.4%] vs 16 of 334 [4.8%]; P = .01), and coagulation disorders (6 of 82 [7.3%] vs 6 of 334 [1.8%]; P = .02). Patients with cardiac injury had higher mortality than those without cardiac injury (42 of 82 [51.2%] vs 15 of 334 [4.5%]; P < .001). In a Cox regression model, patients with vs those without cardiac injury were at a higher risk of death, both during the time from symptom onset (hazard ratio, 4.26 [95% CI, 1.92-9.49]) and from admission to end point (hazard ratio, 3.41 [95% CI, 1.62-7.16]). Conclusions and Relevance Cardiac injury is a common condition among hospitalized patients with COVID-19 in Wuhan, China, and it is associated with higher risk of in-hospital mortality.
Article
Background Since December, 2019, Wuhan, China, has experienced an outbreak of coronavirus disease 2019 (COVID-19), caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Epidemiological and clinical characteristics of patients with COVID-19 have been reported but risk factors for mortality and a detailed clinical course of illness, including viral shedding, have not been well described. Methods In this retrospective, multicentre cohort study, we included all adult inpatients (≥18 years old) with laboratory-confirmed COVID-19 from Jinyintan Hospital and Wuhan Pulmonary Hospital (Wuhan, China) who had been discharged or had died by Jan 31, 2020. Demographic, clinical, treatment, and laboratory data, including serial samples for viral RNA detection, were extracted from electronic medical records and compared between survivors and non-survivors. We used univariable and multivariable logistic regression methods to explore the risk factors associated with in-hospital death. Findings 191 patients (135 from Jinyintan Hospital and 56 from Wuhan Pulmonary Hospital) were included in this study, of whom 137 were discharged and 54 died in hospital. 91 (48%) patients had a comorbidity, with hypertension being the most common (58 [30%] patients), followed by diabetes (36 [19%] patients) and coronary heart disease (15 [8%] patients). Multivariable regression showed increasing odds of in-hospital death associated with older age (odds ratio 1·10, 95% CI 1·03–1·17, per year increase; p=0·0043), higher Sequential Organ Failure Assessment (SOFA) score (5·65, 2·61–12·23; p<0·0001), and d-dimer greater than 1 μg/L (18·42, 2·64–128·55; p=0·0033) on admission. Median duration of viral shedding was 20·0 days (IQR 17·0–24·0) in survivors, but SARS-CoV-2 was detectable until death in non-survivors. The longest observed duration of viral shedding in survivors was 37 days. Interpretation The potential risk factors of older age, high SOFA score, and d-dimer greater than 1 μg/L could help clinicians to identify patients with poor prognosis at an early stage. Prolonged viral shedding provides the rationale for a strategy of isolation of infected patients and optimal antiviral interventions in the future. Funding Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences; National Science Grant for Distinguished Young Scholars; National Key Research and Development Program of China; The Beijing Science and Technology Project; and Major Projects of National Science and Technology on New Drug Creation and Development.