Abstract
Objectives
The role of the environment and climate in the transmission and case fatality rates of SARS-CoV-2 is still being investigated a year into the pandemic. Elevation and air quality are believed to be significant factors in the development of the pandemic, but the influence of additional environmental factors remains unclear.
Methods
We explored the relationship between the cumulative number of infections and mortality cases with climate (temperature, precipitation, solar radiation, water vapor pressure, wind), environmental data (elevation, normalized difference vegetation index or NDVI, particulate matter at 2.5 μm or PM2.5 and NO2 concentration), and population density in Peru. We use confirmed cases of infection from 1,287 districts and mortality in 479 districts, we used Spearman’s correlations to assess the bivariate correlation between environmental and climatic factors with cumulative infection cases, cumulative mortality and case-fatality rate. We explored district cases within the ecozones of coast, sierra, high montane forest and lowland rainforest.
Results
Multiple linear regression models indicate elevation, mean solar radiation, air quality, population density and green vegetation cover, as a socioeconomic proxy, are influential factors in the distribution of infection and mortality of SARS-CoV-2 in Peru. Case-fatality rate was weakly associated with elevation.
Conclusions
Our results also strongly suggest that exposure to poor air quality is a significant factor in the mortality of individuals below the age of 30. We conclude that environmental and climatic factors do play a significant role in the transmission and case fatality rates in Peru, however further study is required to see if these relationships are maintained over time.
Introduction
As SARS-CoV-2 spread around the world within a few months of detection, several groups of investigators are looking into the different factors that could be related to the distribution of the infection and the severity of the disease. The incidence of infection in different countries and cities suggests that several factors influence the rate of infection, which are not only related to the virus and the immune system. In the case of Peru, the cities that have the most cases are generally located near the coast and have poor air quality throughout the year (https://www.datosabiertos.gob.pe/). In addition, population density seems to be a key factor in the spread of the virus, where overcrowded cities increase the probability of contact with infected individuals, increasing the number of infection and mortality cases in a short amount of time. Therefore, population density has been taken into account as one of the cofactors in SARS-CoV-2 distribution.
Although more studies are necessary, the rate of infection and the severity of the diseases seems different for people living in cities at high altitudes, where not only hypoxia is a major factor, but other factors such as air quality, solar radiation, and population density, could play a role in SARS-CoV-2 person-to-person transmission. Arias-Reyes et al. [1] suggested that there exists less rate of infection at high altitude possibly due to a lower level of expression of ACE2 compared to sea level. This observation is important, particularly since SARS-CoV-2 uses ACE2 as a point of infection for the cells [2]. Yao et al. [3] explored the rate of SARS-CoV-2 infection across China, and concluded that there was no correlation between temperature, UV, and the rate of infection. However, they recommended future studies using more complex models and environmental factors. Finch et al. [4] explored the role of pollution exposure in the development of different diseases, where heart and respiratory diseases indicated a detrimental role of pollution on the endothelial integrity, and the changes in the levels of Endothelin I could be observed in people that were exposed to air pollution [5].
Environmentals factors related to mechanics of different viruses spread have been studied by several groups, before the pandemic. Pica et al. [6] studied the environmental factors related to the spread of seasonal influenza and concluded that some meteorological factors play a central role in the person-to-person transmission of the virus, besides the sociodemographic factors. It is well known that UV radiation can decontaminate surfaces and air, and UV incidence increases at higher elevations and cities will have different exposures depending on their location. Kowalski [7] demonstrated that the UV light can destroy a bacteria, many viruses, and fungi, but additional studies are necessary to understand the impact of UV on SARS-CoV-2 in the natural environment, since no amount of germicidal UV-C and only 5% of UV-B reaches below the Earth’s surface [8].
A growing body of research indicate that environmental and climatic factors, specifically temperature and air pollution, may indeed be factors in the heterogeneous distribution and rate of infection and mortality of SARS-CoV-2 in many parts of the world. Lower average temperatures and relative humidity seem to contribute to the increase the number of total infections in various cities from a range of climates, such as the temperate to subtropical climates in China [9], Mediterranean climate of Italy [10] and tropical and subtropical climates in Brazil [11]. Similarly, increasing air pollution density, particularly with particulate matter at 2.5 μm (PM2.5) and nitrogen dioxide (NO2), has been associated with increasing SARS-CoV-2 infection and mortality in different parts of Europe [12–14], Asia [15–17], and the America [18, 19]. However, it is less clear which factors may be contributing to the lower rates of infection at high altitudes, principally due to confounding factors and lack of data [8].
Understanding all the factors that are related to the spread of SARS-CoV-2 will be an important part of public policy, particularly toward the implementation of focalized quarantines in cities and districts where the cases of infection are high and the implementation of effective vaccination distribution. Therefore, this study aims to explore the different cofactors that could increase or decrease the possibility of person-to-person transmission of SARS-CoV-2. The main objectives of this study was to explore the relationship between SARS-CoV-2 infection and mortality cases, case-fatality rates with a set of climate (temperature, precipitation, solar radiation, water vapor pressure, and wind), environmental data (elevation, normalized difference vegetation index or NDVI, particulate matter at 2.5 μm or PM2.5 and NO2 concentration), and population density in Peru. We also divide the distribution of infection, mortality, and case-fatality rates by four ecozones (coast, sierra, high montane forest, and lowland rainforest) and subset mortality cases by age groups and gender.
Materials and methods
We obtained SARS CoV two confirmed cases by district, of the 1873 districts in Peru, from official reports provided by the Peruvian Ministry of Health (MINSA), through an official government open data portal (https://www.datosabiertos.gob.pe/). We used the cumulative number of confirmed cases of infection from 1,287 districts and confirmed cases of mortality in 479 districts, where at least one confirmed case was registered as of June 27, 2020 (Figure 1a). The first case of SARS-CoV-2 was registered on March 5, 2020. Under this criteria, the number of positive SARS-CoV-2 cases analyzed were 263,743 and 7,877 deaths by June 27, 2020. Available data of infection included date of confirmation, while data on mortality included age, sex, and date registered. The first confirmed case of SARS-CoV-2 in Peru was on March 6, 2020. Infection and mortality values were log-transformed to meet statistical assumptions in relation to residuals and added one to avoid taking the logarithm of 0 [9, 16, 17, 20].
We also assessed case-fatality rates, which implies the severity of the condition by estimating the proportion of cases that die of a given condition [21]. The case-fatality rate was estimated by dividing the number of cause-specific deaths among the incidence cases by the total number of incident cases*100 [22].
We labeled districts based on four ecozones, including coast, sierra, high montane forest, and lowland rainforest (Figure 1d), which was originally developed based on elevation and biogeography and used in national and international reporting, and have a significant effect on agricultural activity, population density, population connectivity, and other socioeconomic activities ([23]; see Supplementary Materials for more details).
Climate and environmental data
A set of climate metrics were obtained from WorldClim version 2.1 [24], which presents a historical baseline from the years 1979–2000 and includes monthly temperature, rainfall, wind speed, and solar radiation. Satellite-based environmental data was obtained from a variety of satellite sensors that detect environmental variables, including elevation, vegetation cover, and air quality. Vegetation cover was estimated through NDVI (Normalized Difference Vegetation Index), a commonly used spectral index used to quantify green cover, including agricultural extent, forest cover, and green cover in urban settings. We infer air quality through the metrics of two health-relevant air pollutants, which include particulate matter at 2.5 μm (PM2.5) and nitrogen dioxide (NO2; Figure 1c). Population density was obtained from the WorldPop model of population density (people km−2) for Peru adjusted to match official United Nations population estimates for the year 2020 (www.worldpop.org; Figure 1b). Further details on climate data and satellite-based environmental data are provided in the online Supplementary Materials.
Data analysis
We extracted the zonal statistics by district (i.e. average value per district polygon) of each of the climate, remotely sensed, and population density layers using ArcPro (verison 2.2). We used SPSS 25.0 (IBM, USA) for all statistical analysis of the extracted values. We used the one-sample Kolmogorov-Smirnov parametric test to evaluate the distribution of residuals obtained for each district from each of the original 36 data layers used. Bivariate covariance within the subsets of climate and environmental variables were estimated with Spearman’s Rho correlation, (following [16, 17], where correlation of the order of 0.9 or larger were determined to have a high covariance. As a result, the final reduced data set used in this study were 21 data layers, including 14 climatic layers, six environmental layers and the population density model (Table S1). The climatic layers included four temperature, four rainfall, two solar radiation, two wind speed, and 2 water vapor metrics. All six environmental layers and the population density model had a relatively low level of covariance and were kept for further analysis. The climate and environmental data were also normalized (following [16, 17], due to the different units used in each variable, through the following method:
Where X is the normalized data, X n is the raw data for the variable, X an is the mean value of the raw data, and S n is the standard deviation of X n .
We used stepwise linear regression to explore the association of climatic and environmental variables with the log-transformed data of cumulative number of confirmed cases of infection and confirmed cases of mortality, across Peru and within ecozones. The numbers of cases were used as dependent variables, and the environmental and climatic were selected as independent variables. The resulting statistically significant predictive models use one or multiple variables that best explain the dependent variable. The adjusted R2 from model fitting indicates the percentage of all factors that explain the distribution of cases and the magnitude of the standardized β reflects the influence of the corresponding variable in the predictive model. Collinearity diagnostics, resulting from the regression, were also used to identify model variables that were highly correlated. Significant differences between means were evaluated using one-way ANOVA followed by a Tukey test.
Results
The district of San Juan de Lurigancho, of the department of Lima and in the coastal ecozone, had the most number of confirmed infected cases and case fatalities of SARS-CoV-2, with 13,724 total infections and 394 fatalities, up until the date of the data obtained for this study. The average number of confirmed cases of infection and fatalities was 205 (SD ± 938) and 16.4 (SD ± 42.2) per district, respectively. Across ecozones, the number of registered male fatalities was 5,601 (71.1%) and that of females were 2,276 (28.9%). The average age of total male fatalities was 64.2 (SD ± 15.6) and females was 66.3 (SD ± 15.9). There was also a significant difference between ecozones for mean age of fatalities and female only cases, but not in male only cases (see Table S2). We found no significant difference between the mean case-fatality rates by ecozones (F=0.34, p=0.80).
Correlation analysis indicated a positive correlation of total infections with PM2.5 (⍴=0.449) and a negative correlation with elevation (⍴=−0.476), where the number of infections correlates with increased air pollution, while there are less number of infections with elevation Figure S1). Similarly, correlation analysis indicated a high positive correlation coefficient between mean tropospheric NO2 (⍴=0.497, p<0.001), PM2.5 (⍴=0.506; p<0.001) and registered mortality across sexes and in the male population with ⍴=0.504 and ⍴=0.486, respectively.
In stepwise linear regression model fitting, the three most important parameters that explain the distribution of cumulative number of infections across the country (R2=0.46; p-value=0.038) were NDVI (β=−0.392), elevation (β=−0.311), and mean solar radiation (β=−0.244) (Table 1). The negative correlation indicates a decrease in the cumulative number of infections, with an increase in surrounding green cover (NDVI). In addition, the cumulative number of infections decreases with increasing elevation and increasing mean solar radiation. Within each ecozone, parameters importance change indicating the importance of population density (β=0.474) in the coast, NDVI (β=−0.359) in the sierra, maximum NO2 (β=0.352) in high montane forest, and elevation (β=−0.441) in lowland rainforest.
Dependent | Predictors | Standardized β | R2 | Adj. R2 | p-Value |
---|---|---|---|---|---|
Cumulative inf. | NDVI | −0.392 | 0.46 | 0.46 | 0.04 |
Elevation | −0.311 | ||||
Radiation mean | −0.244 | ||||
NO2 max | 0.171 | ||||
NO2 mean | 0.057 | ||||
PM2.5 | 0.129 | ||||
Pop. density | 0.150 | ||||
Wind mean | −0.081 | ||||
Radiation (SD) | −0.077 | ||||
Coast | Pop. density | 0.474 | 0.35 | 0.34 | 0.03 |
Elevation | −0.199 | ||||
PM2.5 | 0.163 | ||||
Sierra | NDVI | −0.359 | 0.50 | 0.50 | 0.02 |
Radiation mean | −0.272 | ||||
Elevation | −0.244 | ||||
NO2 mean | 0.191 | ||||
PM2.5 | 0.171 | ||||
Pop. density | 0.131 | ||||
Wind SD | −0.077 | ||||
High montane rainforest | NO2 max | 0.325 | 0.40 | 0.39 | <0.001 |
PM2.5 | 0.266 | ||||
Elevation | −0.253 | ||||
Lowland rainforest | Elevation | −0.441 | 0.37 | 0.35 | 0.001 |
Pop. density | 0.327 |
Model fitting for cumulative mortality cases across the Peru indicated the negative correlation of NDVI (β=−0.299), elevation (β=−0.245), and mean solar radiation (β=−0.230), and the positive correlation with mean NO2 (β=0.237) and population density (β=0.215) (Table 2). Population density was the most influential factor for mortality in the coastal ecozone, lowland rainforest, and male and female mortality across the country. It is worth noting that adjusted R2 indicated that the models included at least 28% of the factors that affect the difference in mortality.
Predictors | Standardized β | R2 | Adj. R2 | p-Value | |
---|---|---|---|---|---|
Cumulative mortality | NDVI | −0.299 | 0.48 | 0.48 | <0.001 |
Elevation | −0.245 | ||||
NO2 mean | 0.237 | ||||
Radiation mean | −0.230 | ||||
Pop. density | 0.215 | ||||
Male | Pop. density | 0.250 | 0.46 | 0.45 | <0.001 |
NDVI | −0.245 | ||||
NO2 mean | 0.229 | ||||
Elevation | −0.223 | ||||
Radiation mean | −0.200 | ||||
Female | Pop. density | 0.379 | 0.33 | 0.33 | 0.015 |
NO2 mean | 0.191 | ||||
Elevation | −0.136 | ||||
Coast | Pop. density | 0.502 | 0.37 | 0.36 | <0.001 |
PM2.5 | 0.335 | ||||
Sierra | Elevation | −0.274 | 0.51 | 0.50 | 0.005 |
NDVI | −0.273 | ||||
Pop. density | 0.253 | ||||
NO2 mean | 0.193 | ||||
Radiation mean | −0.162 | ||||
High montane rainforest | NO2 max | 0.576 | 0.55 | 0.53 | 0.003 |
PM2.5 | 0.139 | ||||
Lowland rainforest | Pop. density | 0.548 | 0.30 | 0.28 | <0.001 |
Regarding age groups, mean NO2 was the most influential factor of difference of mortality for individuals with an age range of 0–17 (β=0.654) and 18–29 (β=0.422) (Table 3). Population density was the most influential factor for the remaining age ranges of 30–49 (β=0.306), 50–70 (β=0.483), and above 80 (β=0.309). However, it is worth noting that mean NO2 was indicated as an influential factor for mortality for the age ranges of 30–49 (β=0.236) and above 80 (β=0.258).
Age range | Predictors | Standardized β | R2 | Adj. R2 | p-Value |
---|---|---|---|---|---|
0 to 17 | NO2 mean | 0.654 | 0.43 | 0.40 | <0.001 |
18 to 29 | NO2 mean | 0.422 | 0.18 | 0.16 | 0.001 |
30 to 49 | Pop. density | 0.306 | 0.28 | 0.27 | 0.004 |
NO2 mean | 0.236 | ||||
Elevation | −0.14 | ||||
50 to 79 | Pop. density | 0.483 | 0.28 | 0.27 | 0.034 |
Elevation | −0.14 | ||||
≥80 | Pop. density | 0.309 | 0.36 | 0.35 | 0.046 |
NO2 mean | 0.258 | ||||
PM2.5 | 0.164 | ||||
Solar radiation mean | −0.113 |
A positive correlation from model fitting was found between case fatality rates and elevation (β=0.318), although a relatively low explanatory factor (R2=0.12; p-value=0.015) compared to most predictive models performed in this study (Table 4). Indeed, case fatality rate models had the lowest R2, although statistically significant (p<0.05), compared to models for infection and mortality. Elevation was the most influential factor for the coast (β=0.281), sierra (β=0.303), and high montane forest (β=0.466). Mean diurnal range was the main factor for case fatality rates in high montane forests (β=0.466) and the only climate variable to appear as a factor for model fitting in this study.
Predictors | Standardized β | R2 | Adj. R2 | p-Value | |
---|---|---|---|---|---|
Peru | Elevation | 0.318 | 0.12 | 0.12 | 0.015 |
NDVI | 0.106 | ||||
Coast | Elevation | 0.281 | 0.08 | 0.06 | 0.022 |
Sierra | Elevation | 0.303 | 0.12 | 0.11 | 0.029 |
NDVI | 0.117 | ||||
High montane forest | Elevation | 0.466 | 0.22 | 0.20 | <0.001 |
Lowland rainforest | Mean diurnal range | 0.460 | 0.21 | 0.19 | 0.002 |
Discussion
Our study found that several environmental factors are influential in the cumulative number of SARS-CoV-2 infection and mortality, with particular factors having apparently significant roles in specific scenarios. Generally, the cumulative number of infections is reduced with increased elevation and solar radiation, but other factors such as poor air quality and population density have significant roles. Surprisingly, NDVI, as a measure of green vegetation cover and socioeconomic level in urban settings [25], was a strong predictive factor of SARS-CoV-2 infection (see Table 1). Previous studies have found that more affluent suburbs in Peru tend to be less populated, have more green space that reflects higher property values [25], and therefore, residents may be more likely to have access to private health measures and are physically distant, which could decrease rates of infection in the longer term.
Once we separate infection and mortality by ecozones, the influence of environmental factors is similar to looking at Peru as a whole, with a few notable exceptions. Solar radiation is a much more influential factor in the Sierra, where there are strong geographic and heterogeneous patterns particularly in the south, and UV radiation is known to be strong in these drier environments (see Figure 1d). If we take into account the UV radiation as a possible factor that could decrease the survival of the virus in the air, this could account for the decrease of infected people in the sierra and in the south of the country. However, we do acknowledge that further research is necessary to understand how the specific UV band is effecting SARS-CoV-2, as noted by Pun et al. [8]. Our study does provide the most numerous number of districts with infection and mortality data at high elevation and geospatial data on UV radiation to begin to understand that differences in UV-B concentration in different regions may be indicative of limiting conditions for infection other than temperature, which was not a significant factor in our study.
Mean NO2 concentration is an overriding factor for the cumulative number of mortality for ages below 30. Higher mean NO2 concentration is indicative of anthropogenic activity, such as fossil fuel consumption and biomass burning, which occur in more population dense districts, but it is only moderately correlated with population density in Peru. Mean NO2 concentration in 2019 was highest in the metropolitan cities of Lima, Arequipa and the small southern city of Moquegua, where NO2 density of above 250 μmol m−2 were found year round. Seasonal high concentrations, due to agricultural and other biomass burning, are found in cities in the Amazon basin during the dry season. The overall detrimental effects of air pollution on the health and mortality within populations, in accordance to type and length of exposure is known [26], along with other studies exploring its effects on the current pandemic [27].
Unlike male mortality, we found a significant difference in the mean age of women who died across ecozones in Peru, with a higher average age for women in the Sierra. To our knowledge, this is the first study to indicate the multiple environmental factors that seem to influence female mortality, including a negative correlation with elevation and mean solar radiation, but a larger positive correlation with population density, NDVI and mean NO2 density. Female mortality from SARS-CoV-2 is lower than males in all the countries possibly due to the expression of the ACE2 (angiotensin-converting enzyme 2) receptor regulated by female hormones. However, protection is lost in older women, as the levels of hormones decrease. The probability of infection in older women becomes similar to their male counterpart, but there is still a higher rate of fatalities in men than women independent of age [28].
Our study also found elevation to be an influential factor in case fatality rate for the coast, sierra, high montane forest, lowland rainforest, and the country as a whole when looking at the data at the district level. This is contrary to the findings of [21]; which studied cases at the provincial level, instead of the district level, and found that the case fatality rate was not modified by elevation. Our analysis indicates a rather weak but statistically significant correlation. However, we acknowledge that further studies still need to be conducted to see if this relationship continues throughout the course of the pandemic.
Ultimately, our study provides a different perspective to understanding the factors effecting SARS-CoV-2 infection and mortality due to the greater number of environmental and climate variables available for analysis. Increasing elevation were found to be influential factors, as described in [1]. However, no temperature metric, including lower average temperature identified as important factors in refs. [9, 11]; were statistically significant contributors to the predictive models. When evaluating a larger set of factors, environmental variables, such as air quality, solar radiation, and elevation, appear to be more significant factors than either temperature or precipitation, which may reflect the more environmentally and climatically heterogeneous region. Notwithstanding, our research adds to the growing body of studies that strongly indicate that poor air quality is among the most significant factor of number of infection and mortality in any region of the world.
The limitations of this study are indicative of the asymptomatic nature of SARS-CoV-2 for many patients. It is currently unclear the magnitude of underestimation occurring at the present time and accurate numbers may not become available until widespread molecular testing is performed. As of December 2020, the Peruvian government continues to resist deploying a widespread molecular testing regime. Also, news reports indicate that the cumulative number of mortality may also be significantly underestimated due to the lack of testing and patient care in overwhelmed urban and rural hospitals, particularly during the peaks of SARS-CoV-2 infection.
Conclusion
Elevation is one of several factors that has determined the number of infections and mortality. Other significant factors include population density, air quality, solar radiation and NDVI, as a measure of both green vegetation cover and socioeconomic level. Poor air quality was the single most important factor to determine mortality below the age of 30. We also found that case fatality rate is modified, albeit weakly, by elevation, which is contrary to previously published findings. As more data becomes available, this study can be replicated to see if the relationship between SARS-CoV-2 and climatic and environmental factors are maintained over time.
Acknowledgement
The authors wish to thank Dr. Michal Horowitz for her critical reading and discussion of the manuscript and the two anonymous reviewers for their comments.
-
Research funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
-
Author contributions: All authors have accepted responsibility for the entire content of this manuscript and approved its submission.
-
Competing interests: Authors state no conflict of interest.
-
Informed consent: Informed consent was obtained from all individuals included in this study.
-
Ethical approval: Not applicable.
References
1. Arias-Reyes, C, Zubieta-DeUrioste, N, Poma-Machicao, L, Aliaga-Raudan, F, Carvajal-Rodriguez, F, Dutschmann, M, et al.. Does the pathogenesis of SAR-CoV-2 virus decrease at high-altitude? Respir Physiol Neurobiol 2020:103443. https://doi.org/10.1016/j.resp.2020.103443.Search in Google Scholar PubMed PubMed Central
2. Ren, LL, Wang, YM, Wu, ZQ, Xiang, ZC, Guo, L, Xu, T, et al.. Identification of a novel coronavirus causing severe pneumonia in human: a descriptive study. Chinese Med J 2020;133:1015–24. https://doi.org/10.1097/cm9.0000000000000722.Search in Google Scholar PubMed PubMed Central
3. Yao, Y, Pan, J, Liu, Z, Meng, X, Wang, W, Kan, H, et al.. No Association of COVID-19 transmission with temperature or UV radiation in Chinese cities. Eur Respir J 2020;55. https://doi.org/10.1183/13993003.00517-2020.Search in Google Scholar PubMed PubMed Central
4. Finch, J, Conklin, DJ. Air pollution-induced vascular dysfunction: potential role of endothelin-1 (ET-1) system. Cardiovasc Toxicol 2016;16:260–75. https://doi.org/10.1007/s12012-015-9334-y.Search in Google Scholar PubMed PubMed Central
5. Briet, M, Collin, C, Laurent, S, Tan, A, Azizi, M, Agharazii, M, et al.. Endothelial function and chronic exposure to air pollution in normal male subjects. Hypertension 2007;50:970–6. https://doi.org/10.1161/hypertensionaha.107.095844.Search in Google Scholar
6. Pica, N, Bouvier, NM. Environmental factors affecting the transmission of respiratory viruses. Curr Opin Virol 2012;2:90–5. https://doi.org/10.1016/j.coviro.2011.12.003.Search in Google Scholar PubMed PubMed Central
7. Kowalski, W. Ultraviolet germicidal irradiation handbook: UVGI for air and surface disinfection, 1st ed. Berlin, Heidelberg: Springer Science & Business Media; 2009.10.1007/978-3-642-01999-9Search in Google Scholar
8. Pun, M, Turner, R, Strapazzon, G, Brugger, H, Swenson, ER. Lower incidence of COVID-19 at high Altitude: facts and confounders. High Alt Med Biol 2020;21:217–22. https://doi.org/10.1089/ham.2020.0114.Search in Google Scholar PubMed
9. Liu, J, Zhou, Ji, Yao, J, Zhang, X, Li, L, Xu, X, et al.. Impact of meteorological factors on the COVID-19 transmission: a multicity study in China. Sci Total Environ 2020;726. https://doi.org/10.1016/j.scitotenv.2020.138513.Search in Google Scholar PubMed PubMed Central
10. Perone, G. The determinants of COVID-19 case fatality rate (CFR) in the Italian regions and provinces: an analysis of the environmental demographic, and health factors. Sci Total Environ 2021;755:142523. https://doi.org/10.1016/j.scitotenv.2020.142523.Search in Google Scholar PubMed PubMed Central
11. Prata, DN, Rodrigues, W, Bernejo, PH. Temperature significantly changes COVID-19 transmission in (sub) tropical cities of Brazil. Sci Total Environ 2020;729:138862. https://doi.org/10.1016/j.scitotenv.2020.138862.Search in Google Scholar PubMed PubMed Central
12. Cole, MA, Ozgen, C, Strobl, E. Air pollution exposure and COVID-19 in Dutch municipalities. Environ Resour Econ 2020;76:581–610. https://doi.org/10.1007/s10640-020-00491-4.Search in Google Scholar PubMed PubMed Central
13. Coker, ES, Cavalli, L, Fabrizi, E, Guastella, G, Lippo, E, Parisi, ML, et al.. The effects of air pollution on COVID-19 related mortality, in Northern Italy. Environ Resour Econ 2020;76:611–34. https://doi.org/10.1007/s10640-020-00486-1.Search in Google Scholar PubMed PubMed Central
14. Lolli, S, Chen, YC, Wang, SH, Vivone, G. Impact of meteorological conditions and air pollution on COVID-19 pandemic transmission in Italy. Sci Rep 2020;10:16213. https://doi.org/10.1038/s41598-020-73197-8.Search in Google Scholar PubMed PubMed Central
15. Rao Kolluru, SS, Kumae Patra, A, Nazneen, Shiva Nagendra, SM. Association of air pollution and meteorological variables with COVID-19 incidence: evidence from five megacities in India. Environ Res 2021;195:110854. https://doi.org/10.1016/j.envres.2021.110854.Search in Google Scholar PubMed PubMed Central
16. Zhu, L, Liu, X, Huang, H, Avellán-Llaguno, RD, Lazo, MML, Gaggero, A, et al.. Meteorological impact on the COVID-19 pandemic: a study across eight severely affected regions in South America. Sci Total Environ 2020;744:140881. https://doi.org/10.1016/j.scitotenv.2020.140881.Search in Google Scholar PubMed PubMed Central
17. Zhu, H, Xie, J, Huang, F, Cao, L. Association between short-term exposure to air pollution and COVID-19 infection: evidence from China. Sci Total Environ 2020;727:138704. https://doi.org/10.1016/j.scitotenv.2020.138704.Search in Google Scholar PubMed PubMed Central
18. Bolano-Ortiz, TR, Camargo-Caicedo, Y, Puliafito, SE, Ruggeri, MF, Bolano-Diaz, S, Pascual-Flores, R, et al.. Spread of SARS-CoV-2 through Latina America and the Caribbean región: a look from its economic conditions, climate and air pollution indicators. Environ Res 2020;191:1009938. https://doi.org/10.1016/j.envres.2020.109938.Search in Google Scholar PubMed PubMed Central
19. Wu, X, Nethery, RC, Sabath, MB, Braun, D, Dominici, F. Air pollution and COVID-19 mortality in the United States: strengths and limitations of an ecological regression analysis. Sci Adv 2020;6:eabd4049. https://doi.org/10.1126/sciadv.abd4049.Search in Google Scholar PubMed PubMed Central
20. Xie, J, Zhu, Y. Association between ambient temperature and COVID-19 infection in 122 cities from China. Sci Total Environ 2020;724:138201. https://doi.org/10.1016/j.scitotenv.2020.138201.Search in Google Scholar PubMed PubMed Central
21. Segovia-Juarez, J, Castagnetto, JM, Gonzales, GF. High altitude reduces infection rate of COVID-19 but not case-fatality rate. Respir Physiol Neurobiol 2020;281:103494. https://doi.org/10.1016/j.resp.2020.103494.Search in Google Scholar PubMed PubMed Central
22. CDC. Section 3: mortality frequency measures. In: Principles of epidemiology in public health practice. An introduction to applied epidemiology and biostatistics, 3rd ed. Atlanta, GA, USA: US Department of Health and Human Services; 2019. https://www.cdc.gov/csels/dsepd/ss1978/lesson3/section3.html.Search in Google Scholar
23. MINAM. Peru’s submission of a Forest Reference Emission Level (FREL) for reducing emissions from deforestation in the Peruvian Amazon. Peru: Ministry of the Environment; 2016.Search in Google Scholar
24. Fick, SE, Hijmans, RJ. WorldClim 2: new 1‐km spatial resolution climate surfaces for global land areas. Int J Climatol 2017;37:4302–15. https://doi.org/10.1002/joc.5086.Search in Google Scholar
25. Loret de Mola, U, Ladd, B, Duarte, S, Borchard, N, Anaya La Rosa, R, Zutta, B. On the use of hedonic price indices to understand ecosystem service provision from urban green space in five Latin American megacities. Forests 2017;8:478. https://doi.org/10.3390/f8120478.Search in Google Scholar
26. Carey, IM, Atkinson, RW, Kent, AJ, Van Staa, T, Cook, DG, Anderson, HR. Mortality associations with long-term exposure to outdoor air pollution in a national English cohort. Am J Respir Crit Care Med 2013;187:1226–33. https://doi.org/10.1164/rccm.201210-1758oc.Search in Google Scholar PubMed PubMed Central
27. Fattorini, D, Regoli, F. Role of the chronic air pollution levels in the Covid-19 outbreak risk in Italy. Environl Pollut 2020:114732. https://doi.org/10.1016/j.envpol.2020.114732.Search in Google Scholar PubMed PubMed Central
28. Jin, JM, Bai, P, He, W, Wu, F, Liu, XF, Han, DM, et al.. Gender differences in patients with COVID-19: focus on severity and mortality. Front Publ Health 2020;8:152. https://doi.org/10.3389/fpubh.2020.00152.Search in Google Scholar PubMed PubMed Central
Supplementary Material
The online version of this article offers supplementary material (https://doi.org/10.1515/jbcpp-2021-0007).
© 2021 Walter de Gruyter GmbH, Berlin/Boston