Skip to main content
NCBI home page
Elsevier - PMC COVID-19 Collection logoLink to Elsevier - PMC COVID-19 Collection
. 2023 Apr 18;229:115904. doi: 10.1016/j.envres.2023.115904

Environmental factors are associated to hospital outcomes in COVID-19 patients during lockdown and post-lockdown in 2020: A nationwide study

Laura Sánchez-de Prada a,b,1, José María Eiros-Bachiller c, Álvaro Tamayo-Velasco b,d,e, Marta Martín-Fernández b,e,f,∗, F Javier Álvarez b,e,f, Carlos Giner-Baixauli g, Eduardo Tamayo b,e,h,1, Salvador Resino b,e,i,2, Alejandro Alvaro-Meca e,j,2
PMCID: PMC10112945  PMID: 37080281

Abstract

Objective

This study analyzed, at a postcode detailed level, the relation-ship between short-term exposure to environmental factors and hospital ad-missions, in-hospital mortality, ICU admission, and ICU mortality due to COVID-19 during the lockdown and post-lockdown 2020 period in Spain.

Methods

We performed a nationwide population-based retrospective study on 208,744 patients admitted to Spanish hospitals due to COVID-19 based on the Minimum Basic Data Set (MBDS) during the first two waves of the pandemic in 2020. Environmental data were obtained from Copernicus Atmosphere Monitoring Service. The association was assessed by a generalized additive model.

Results

PM2.5 was the most critical environmental factor related to hospital admissions and hospital mortality due to COVID-19 during the lockdown in Spain, PM10, NO2, and SO2and also showed associations. The effect was considerably reduced during the post-lockdown period. ICU admissions in COVID-19 patients were mainly associated with PM2.5, PM10, NO2, and SO2 during the lockdown as well. During the lockdown, exposure to PM2.5 and PM10 were the most critical environmental factors related to ICU mortality in COVID-19.

Conclusion

Short-term exposure to air pollutants impacts COVID-19 out-comes during the lockdown, especially PM2.5, PM10, NO2, and SO2. These pollutants are associated with hospital admission, hospital mortality and ICU admission, while ICU mortality is mainly associated with PM2.5 and PM10. Our findings reveal the importance of monitoring air pollutants in respiratory infectious diseases.

Keywords: COVID-19, Hospitalizations, Hospital mortality, Intensive care unit, Air pollution, Respiratory virus

1. Introduction

In December 2019, a new betacoronavirus (SARS-CoV2) emerged in Wuhan, changing life radically as we used to know it (Lu et al., 2020; Xu et al., 2020). In Spain, there were two epidemic waves and more than 2 million cases. Hospitalizations reached up to 200.000 admissions overloading hospital services; mainly, the capacity of Intensive Care Units (ICU) had to be increased to admit more than 18.000 severe cases. More than 42.000 people died in Spain that first year (Equipo COVID-19. Red Nacional de Vigilancia Epidemiológica (RENAVE). Centro Nacional de Microbiología (CNM), 2020; Equipo COVID-19. Red Nacional de Vigilancia Epidemiológica (RENAVE). Centro Nacional de Microbiología (CNM), 2021).

The first measure taken to prevent the virus spread was the confinement of the population that started in Spain on March 15th and lasted until May 10th, 2020 (Boletín Oficial del Estado (67): 25390–25400, 2020). One of the main consequences was the reduction of person-to- person contact, drastically impacting the virus's transmission (Kraemer et al., 2020). Another direct consequence was the reduced traffic, as many people started working remotely from home. This had implications on the levels of environmental conditions, as anthropogenic emission of nitrogen oxides comes mainly from fossil fuels.

In the past two years, many studies on pollution and COVID-19 outcomes have been performed (Zang et al., 2022; Sarmadi et al., 2021; Martelletti and Martelletti, 2020). The main findings suggest short and long-term exposure to NO2, PM2.5, and SO2 was associated with higher COVID- 19 incidence and long-term exposure to PM2.5 with increased COVID-19 mortality. A recent study suggested short and long-term exposure to NO2, and PM2.5 increased COVID-19 hospitalizations and ICU admissions (Chen et al., 2022).

This study analyzed, at a detailed postcode level, the relationship between short-term exposure to environmental factors and hospital admissions, in-hospital mortality, ICU admission, and ICU mortality due to COVID-19 in Spain during the lockdown and post-lockdown 2020 periods.

2. Materials and methods

2.1. Study design

We conducted a nationwide population-based retrospective study in patients hospitalized due to COVID-19 in Spain in 2020. Clinical and administrative data of all patients were collected from the Spanish Minimum Basic Data Set (MBDS), an administrative database provided by the Ministry of Health, which has an estimated coverage of 99.5% for both public and private Spanish hospitals discharges (Subdirección General de Información Sanitaria, 2016). The database includes encrypted patient identification numbers, gender, birth date, hospital admission, and discharge, and admission to Intensive Critical care Units (ICU), and postcode. It also

Includes clinical data, 20 diagnoses, and 20 procedures codes according to the International Classification of Diseases 10th Revision, Clinical Modification (ICD-10-CM), as well as the outcome at discharge (ICD-10-CM, 2022). The MBDS is validated for data quality and overall methodology by the Spanish Ministry of Health, establishing protocols and periodic audits. The data were treated with complete confidentiality according to Spanish legislation. Thus, given the anonymous and mandatory nature of the data, informed consent was not required or necessary. This study was approved by the Ethics Committee of Valladolid East Health Area under the code PI 22–2855.

2.2. Study variables and outcomes

Diagnosis codes included in the MBDS, differentiate between primary and secondary diagnoses at discharge and whether they were present on admission (POA). Hospitalization due to COVID-19 was defined as any hospitalization with codes B97.29 and U07.1, as the principal diagnosis present on admission, from January 1st to December 31st, 2020. Outcomes considered in this study included COVID-19 severity, defined as: a) hospital admission, b) in-hospital mortality, c) ICU admission, and d) ICU mortality. In Spain, the first wave of COVID-19 was marked by the lockdown from March 15th to May 10th that year (Boletín Oficial del Estado (67): 25390–25400, 2020). That had multiple effects, including a considerable reduction in traffic flow, among others (Donzelli et al., 2021). Therefore, the study was divided into two periods: the first, from COVID-19 introduction in Spain until May 10th, 2020; and the second, from May 11th, 2020 until December 31st, 2020.

2.3. Air pollution data

Air pollutant data from January 1st, 2020, to December 31st, 2020 was obtained from Copernicus Atmosphere Monitoring Service (CAMS) European air quality forecasts (METEO FRANCE, 2022). CAMS registers hourly analysis for the European Region at a level of (0.1° × 0.1°) approx. 10 km2. For the seven main air pollutants, daily averages were calculated and used for analysis, including carbon monoxide (CO) in μg/m3, nitrogen monoxide (NO) in μg/m3, nitrogen dioxide (NO2) in μg/m3, sulfur dioxide (SO2) in μg/m3, ozone (O3) in μg/m3, particulate matter < 2.5 μm (PM2.5) in μg/m3, and particulate matter < 10 μm (PM10) in μg/m3.

2.4. Meteorological data

Temperature and relative humidity were considered the two main meteorological effects. Data was obtained from Copernicus Climate Change Service, ERA5-Land hourly data from 1950 to the present (Muñoz Sabater, 2022). ERA-5 produces hourly analysis in a regular grid of (0.1° × 0.1°). As Copernicus does not provide precipitation data that might influence pollution levels, relative humidity was used as an indirect method to account for precipitation. To measure air humidity, we computed the daily average temperature of air at 2 m above the ground surface, and the temperature to which the air, at 2 m above the ground surface, would have to be cooled for saturation to occur. That data, combined with temperature and pressure, can be used to calculate the relative humidity by (Alduchov and Eskridge, 1996):

RH=100exp((17.625*TD)/(243.04+TD))exp((17.625*T)/(243.04+T))

With T D as dew point temperature (◦C) at 2 m above the surface and T as temperature (◦C) of air at 2m above the surface.

2.5. Link environmental data with clinical data

A critical task in these studies is the linkage of air pollutant exposures to individuals in the data set. Having the data on a fine grid over Spain, the assignment was performed as follows: first, each patient's centroid postcode was calculated; secondly, the nearest position of the centroid to the grid was searched; and finally, the environmental data were linked according to the date of admission.

2.6. Statistical analysis

A descriptive study of each environmental factor was carried out in each wave defined. Spearman's rank correlation test was used to study the bivariate relationships between the environmental factors and the outcomes. The incubation period of COVID-19 ranges from 1 to 7 days, so we used a moving average approach to account for the cumulative effect of environ-mental factors. Therefore, the cumulative effects were examined by modeling the moving average lag effect (lag3, lag5, lag7) on the mean environmental factors on daily hospital admission of COVID-19. For example, lag0 represented the concentration of the day of hospital admission. Lag3 represented a 3-day moving average exposure, which was calculated as the average concentration of the day of hospital admission and the three previous days, and so on. We performed separate models by lockdown and post-lockdown period, using a generalized additive model (GAM) (Hastie and Tibshirani, 1990) with a Poisson family distribution and log-link function to estimate the association between the moving average air pollutant concentrations and the daily hospital admission, hospital mortality, ICU admission, and ICU mortality. GAMs were adjusted by temperature, relative humidity, and day of the week. Because there were higher correlations between air pollutants (see Appendix A; Table A1), each model only contains one of them to avoid collinearity. Therefore, the model is defined as follows:

log(yit) = Xil + Tempil + RHil + s(dayi) + εit

where i is the postcode, t is for the date of admission, and log(y it) is the log-transformed cases of hospital admissions, hospital mortality, ICU admission, and ICU mortality for postcode i and date t. X il represents the moving average term (lag0-l) of daily air pollution in postcode i. We controlled by daily mean temperature (Temp il), relative humidity, (RH il) and day of the week as s(day i), which is the basis spline to smooth the data with 4 degrees of freedom. We obtained the percentage of change (%) by exponentiating the effects estimates, subtracting 1, and multiplying by 100. We controlled over-dispersion using quasi-Poisson distribution (McCullagh and Nelder, 1989).

Finally, a sensitivity study was performed on the pollutants that showed an association to establish values at which the pollutants showed a more significant impact on the outcomes studied. For this purpose, and due to the high correlation between different outcomes, we used a multi-output decision tree regression (Dumont et al., 2009) with a minimum of 1000 observations per sheet. As regressors, we used the moving averages of each of the pollutants. This was performed with a MANOVA test which allows the degree of correlation between outputs to be taken into account in the segmentation process. The outputs considered were hospital admissions, in-hospital mortality, ICU admission, and ICU mortality. Thus, in each tree, we can find those pollutant values at lag3, lag5, and lag7 for which the mean number of outputs is different. A secondary sensitivity analysis was also carried out to assess environmental factors’ impact on the clinical events analyzed, stratifying the population ac-cording to the previous chronic lower respiratory diseases (CLRD) with the ICD-10 codes J40 to J47, which encompasses four major diseases: chronic obstructive pulmonary disease (COPD), chronic bronchitis, emphysema, and asthma.

All analyses were performed using python (version 3.9) and R statistical software (version 4.2.1) with the mgcv package (Wood, 2011) for GAM analysis. All statistical analyses were evaluated using two-sided tests at the 0.05 level of significance. False discovery rates were calculated using the Benjamini-Hochberg method for multiple comparisons.

3. Results

3.1. Population characteristics

A total of 3,114,793 hospital admissions were recorded in Spanish MDBS during 2020, of which 251,417 were admitted with a COVID-19 diagnosis and 217,106 of them with the principal diagnosis and present on hospital admission. Finally, 208,744(96%) hospital admissions with full postcodes were selected Fig. 1 .

Fig. 1.

Fig. 1

Open in a new tab

Diagram of patient selection.

Table 1 shows all patients’ clinical and epidemiological characteristics stratified by waves. Overall, the median age was 69 years, and 56% were men. The hospital stay was seven days, in-hospital mortality was 15.9%, ICU admission was 8.8%, and ICU mortality was close to 30%. A difference in in-hospital mortality between waves was found, with the epidemiological characteristics remaining constant.

Table 1.

Summary of the epidemiological and clinical characteristics of patients with a COVID-19 hospital admission in Spain. Abbreviations: No: Number of patients; ICU: Intensive Care Unit. Values are expressed as number (%) for categorical variable and median (interquartile range) for quantitative variable.

No, Total
 Lockdown period
 Post-lockdown period
 p-value
208,744 103,154 105,590
Gender (male) 117,620 (56.35) 58,217 (56.44) 59,403 (56.26) 0.413
mean Age (years) 69.0 (26.0) 69.0 (25.0) 69.0 (27.0) 0.770
Length of stay (days) 8.0 (8.0) 8.0 (9.0) 8.0 (7.0) <0.001
In-hospital mortality 34,867 (16.7) 19,317 (18.7) 15,550 (14.7) <0.001
Charlson Index 1.0 (2.0) 1.0 (2.0) 1.0 (2.0) <0.001
ICU
ICU 18,915 (9.1) 9094 (8.8) 9821 (9.3) <0.001
ICU death 6030 (31.9) 3057 (33.6) 2973 (30.3) 0.270
ICU length of stay 10.0 (16.0) 12.0 (19.0) 9.0 (13.0) <0.001
Open in a new tab

3.2. Environmental conditions in 2020

During the lockdown in 2020, NO, NO2, and SO2 levels were lower and CO, O3 PM10, and PM2.5 levels were higher compared to post-lockdown (Appendix A, Fig. 1, Fig. 2 ). In order to look for a seasonal pattern, the previous year and the year after were also studied. Compared to those years, overall pollution was lower in 2020. However, differences corresponding to lock-down and post-lockdown periods are equally observed. Therefore, a separate analysis of 2020 would be appropriate to discard the seasonal effect.

Fig. 2.

Fig. 2

Open in a new tab

Summary of the association between environmental factors and hospital admissions due to COVID-19. Abbreviations: PC: Percentage of change (%), computed by GAM adjusted for temperature, humidity, and day of the week. (*) Increases of 0.1 μg/m3 for NO and SO2. CI95%, 95% of the confidence interval. Lag: Moving average lag effect at 3, 5 and 7 days. q-value: False discovery rate q-value. Note that the x-axes have different scales.

3.3. Environmental conditions related to COVID-19 hospital admissions

Fig. 2 shows the association between environmental factors and the number of hospital admission. During the lockdown period (Fig. 2a), all air pollutants were positively associated with hospital admissions, except for O3. PM2.5 had the most significant impact on hospital admission, with a percentage of change (PC) greater than 8% for increments of 1 μg/m3. Exposures to NO2, PM10, and SO2 also impacted hospital admissions, with an increase around 5% for every 1 μg/m3 increase for the first two pollutants and 0.1 μg/m3 for the third, in the moving average (Fig. 2a).). Although the negative impact of some air pollution was maintained in the post-lockdown period, it was not as high as in the lockdown period. The PC for PM2.5 was around 3–4% for increases of 1 μg/m3 in the moving average (Fig. 2b). Besides, a similar pattern was observed for NO2 and PM10. However, SO2 considerably decreased in this period (Fig. 2b).

3.4. Environmental conditions related to COVID-19 hospital mortality

During the lockdown, the negative impact of PM2.5 was present in the hospital mortality with a PC around 5% for increases of 1 μg/m3 in the moving average at lag 3, 5, and 7 days. Also PM10,NO2, and SO2, showed a negative impact on hospital mortality with a PC of at least 2% for increases of 1 μg/m3 (PM10 and NO2), and 0.1 μg/m3 (SO2) in the moving average at 3, 5, and 7 days (Fig. 3 a). Also, NO presented a PC of at least 1% for increases of 0.1 μg/m3 in the moving average at lag 3, 5, and 7 days. On.

Fig. 3.

Fig. 3

Open in a new tab

Summary of the association between environmental factors and hospital mortality due to COVID-19. Abbreviations: PC: Percentage of change (%), computed by GAM adjusted for temperature, humidity, and day of the week. (*) Increases of 0.1 μg/m3 for NO and SO2. CI95%, 95% of the confidence interval. Lag: Moving average lag effect at 3, 5 and 7 days. q-value: False discovery rate q-value. Note that the x-axes have different scales.

The contrary, a lesser impact of environmental conditions was observed on hospital mortality during the post-lockdown period (Fig. 3b).

3.5. Environmental conditions related to COVID-19 ICU admissions

The effect of environmental conditions during lockdown showed PM10 and PM25 had the strongest association with ICU admission, with a PC between 2 and 4% for increases of 1 μg/m3 in the moving average at lag 3,5 and 7 days. Likewise, NO2 and SO2 were associated with ICU admission. By contrast, the impact of pollutants in the post-lockdown was lesser and the strongest association was present for PM2.5 with a Pc of 1% for increases of 1 μg/m3 in the moving average at lag 3,5 and 7 days (Fig. 4 a and b).

Fig. 4.

Fig. 4

Open in a new tab

Summary of the association between environmental factors and ICU admissions due to COVID-19. Abbreviations: PC: Percentage of change (%), computed by GAM adjusted for temperature, humidity, and day of the week. (*) Increases of 0.1 μg/m3 for NO and SO2. CI95%, 95% of the confidence interval. Lag: Moving average lag effect at 3,5 and 7 days. q-value: False discovery rate q-value. Note that the x-axes have different scales.

3.6. Environmental conditions related to COVID-19 ICU mortality

Fig. 5 shows the environmental effect on ICU mortality, but this was evident only during the lockdown for PM2.5 and PM10 (Fig. 5a). During post-lockdown (Fig. 5b), a different picture from the one previously detected was observed. Only exposures to PM2.5 and PM10 showed a slight effect on ICU mortality with PC around 0.25% for increases of 1 μg/m3 in the moving average at lag 3, and 7 days (Fig. 5b).

Fig. 5.

Fig. 5

Open in a new tab

Summary of the association between environmental factors and ICU mortality due to COVID-19. Abbreviations: PC: Percentage of change (%), computed by GAM adjusted for temperature, humidity, and day of the week. (*) Increases of 0.1 μg/m3 for NO and SO2. CI95%, 95% of the confidence interval. Lag: Moving average lag effect at 3,5 and 7 days. q-value: False discovery rate q-value. Note that the x-axes have different scales.

3.7. Sensitivity analysis

Using the environmental factors that presented an association with the events studied, the sensitivity analysis determined that in the lockdown period, close to 65% of admissions, hospital mortality, ICU admissions, and ICU mortality happened when the moving average at 3 days was over 162 μg/m3 for CO, over 0.9 μg/m3 for SO2, 3.1 μg/m3 for NO2, 0.2 μg/m3 for NO, 9.8 μg/m3 for PM10, and 8.5 μg/m3 for PM2.5 (Fig. 6 ).

Fig. 6.

Fig. 6

Open in a new tab

Summary of the distribution of each pollutant with its cut-off value for each outcome, obtained by multi-output decision tree regression at lag3 days during lockdown.

In the post-lockdown period, the environmental values associated with significant differences in the means of each outcome were similar to those obtained in the lockdown period. However, although the segmentation performed provided cut-offs for which the mean of the number of events was significantly different, the distribution of the events during the post-lockdown is more uniform than that obtained in the lockdown period (see Appendix A Figure A3). Similar patterns were found for lag5 and lag7 days (data not shown).

We stratified the patients according to previous CLRD to analyze the as-sociation between environmental factors and clinical events. When patients did not have CLRD, the association of environmental factors with the out-comes studied remained in the same direction as when the population was not stratified (see Appendix A Table A2). However, when the previous CLRD was considered, the effect of environmental factors was reduced in all clinical events analyzed, particularly in ICU-related clinical events (admission and death) (see Appendix A Table A3).

4. Discussion

This report analyzes the influence of environmental factors on hospital admission, hospital mortality, ICU admission, and ICU mortality in COVID- 19 patients during the lockdown and post-lockdown periods in 2020. Those outcomes were significantly affected mainly by PM2.5, NO2, PM10, and SO2 specially in the lockdown. Additionally, we determined the values at which an increase in admissions, in-hospital mortality, ICU admissions, and ICU mortality would be observed three days later.

The lockdown in many countries worldwide during the spring of 2020 to prevent the spread of COVID-19 significantly impacted the quality of life in most areas (Choi et al., 2021). As one of the countries most affected by the COVID-19 pandemic, Spain implemented one of the most strict confinement measures (Domínguez-Amarillo et al., 2020). Most studies agree on a global decrease in pollution during the lockdown (Srivastava, 2021). However, a few studies found no differences from previous years (Schiermeier, 2020; Jia et al., 2020; Varotsos et al., 2021). In our study, although 2020 was less polluted, the differences between the lockdown and post-lockdown periods were not as significant as we initially

Expected. Different factors are probably involved in those results. First, after the lockdown, some cities returned rapidly to their usual pollution levels, while in others, levels remained relatively low for a while (Jevtic et al., 2021). And second, meteorological conditions affecting pollution levels are usually repeated year after year.

When comparing both periods (lockdown and post-lockdown), the short-term impact of air pollutants is higher during the lockdown. Still, there are many factors that we must take into consideration. First, the traffic, aviation, industrial activities, and shipping reduction did not recover immediately. Second, home-based work was prolonged for more than a year in many sectors, 2020 summer holiday trips remained local, and mobility restrictions were imposed in Spain in the autumn of 2020 as cases increased dramatically (Gobierno de España, 2020). The use of masks outdoors was also implemented in Spain at the end of June 2020. And third, the confinement from March to May 2020 in Spain involved an intense use of home heating systems, which has previously been described from data all over Europe (Menut et al., 2020).

The difficult task of relating COVID-19 disease with air pollutants has been attempted since the early days of the pandemic (Zang et al., 2022). Those studies have initially established relationships between air pollutants and COVID- 19 transmission, particularly PM2.5 (Wang et al., 2020; Tateo et al., 2022), but also PM10, CO, NO2, and O3 (Zhu et al., 2020; Copat et al., 2020). Many mortality studies have been performed worldwide (Bozack et al., 2022; Coker et al., 2020; Hendryx and Luo, 2020; Wu et al., 2020), but only a few studies contemplating a short term analysis (Jiang and Xu, 2021; Khorsandi et al., 2021). Those agree that PM2.5 is related to higher mortality. Also, PM2.5, PM10 and O3 were associated with higher mortality and hospitalization rates (Khorsandi et al., 2021). Our results are aligned with previous findings in which PM2.5 was associated with hospitalization, remarking the additional impact of PM10 and SO2 on mortality. Additionally, our results show that NO should be considered in future studies since it has a critical and explainable impact on COVID-19 outcomes. NO and small amounts of NO2 are generated by traffic, heating, and industrial processes, but the latter mainly derives from NO conversion in the atmosphere. Therefore, it is essential not to limit studies to the influence of NO2 but to broaden them to NOx (Ayuntamiento de Valladolid, 2022).

It is well-known air pollutants have a short-term and long-term impact on human health. On the one hand, they alter the functions of lung cells, increasing oxidative stress and inflammation and altering the immune responses which favor viral infections (Boningari and Smirniotis, 2016; Copat et al., 2020; Ray and Kim, 2014). Also, a relationship between acute exposure and cardiovascular problems, including stroke, cardiac arrest, and thrombosis, has been described (Robertson and Miller, 2018). On the other hand, those processes

Lead to fibrosis that reduces pulmonary function, mediating the development, maintenance, and exacerbation of obstructive airway diseases and favoring infectious diseases (Feng et al., 2016). The relation between the aforementioned mechanisms and COVID-19 has also been proposed (Bourdrel et al., 2021; Woodby et al., 2020). In fact, those effects have been similarly described for different respiratory viruses (Domingo and Rovira, 2020). Additionally, it has been suggested that pollutant particles could be acting as transporters for SARS-CoV-2 (Martelletti and Martelletti, 2020),a theory that should be carefully studied.

Although not many studies include SO2 among the air pollutants studied regarding COVID-19, to our knowledge, no evidence has been described before that a positive correlation exists with COVID-19 outcomes. In contrast, a negative correlation between SO2 and mortality has been found in a couple of studies (Jiang and Xu, 2021; Zhu et al., 2020). Interestingly, the short-term effects of SO2 were described 25 years ago (Katsouyanni et al., 1997; Stieb et al., 2002, 2003; Sunyer et al., 2003a, b) as a potent irritant contributing to airway inflammation. However, its implications are still controversial, suggesting it could contribute as a co-factor of other pollutants (Kan et al., 2010).

In addition, the cut-off values provided in our study could help us in two directions. On the one hand, it can help us predict when hospital demands will increase if those limits are exceeded the previous 3 days. On the other hand, they can be set as limits of pollution allowed to prevent these events from happening. Although, this has not been performed before, we consider this should be further studied.

A potential bias to consider is that there were changes in patient management during the study period, especially during the first wave. The overflow of patients and the increase in medical personnel with little experience in the management of critical patients could have influenced the outcome of patients with COVID-19. The MBDS database does not have information on the medical staff who care for the patients and their previous experience, so we cannot analyze their real influence on the outcome of these patients.

Overall, we showed associations between four air pollutants (PM2.5, PM10, NO2, and SO2) and different clinical outcomes, with similar patterns, despite the impact of the post-lockdown period. The short-term exposure with lags (3, 5, 7d) was assessed because the incubation period of COVID-19 ranges from 1 to 7 days. Thus, the relevant role of PM2.5, PM10, NO2, and SO2 could be due to their direct pathological effects on the lower respiratory tract, which could increase the severity of COVID-19. Moreover, three air pollutants (CO, NO, and O3) did not show a significant short-term impact, but a possible direct or indirect long-term impact cannot be ruled out.

CLRD is a major predictor of severe outcomes in COVID-19 patients.

(Beltramo et al., 2021; Gerayeli et al., 2021). We conducted a sensitivity analysis to evaluate the impact of previous CLRD on the relationship between environmental factors and clinical events, finding previous CLRD diluted this association with the clinical events analyzed, particularly in the ICU. It is possible that the impact of environmental contaminants on the clinical outcomes of COVID-19 is more evident in the absence of CLRD and that the presence of CLRD dilutes or cancels this association because it is already a risk factor for severe COVID- 19.

4.1. Limitations of the study

The main limitations are: (i) The retrospective design could introduce biases. (ii) There was no relevant clinical information to interpret the COVID- 19 infection (iii) The accuracy of the MBDS for COVID-19 diagnosis was not evaluated, generating a confusion bias. (iv) Records of ICU admitted patients only include length of stay but not whether it was at the time of admission or later on. (v) We did not have data on indoor air contaminants, which may also influence susceptibility to COVID-19 infection.

Our study also has several strengths that must be considered: (i) This nationwide study covers around 47 million population and all postcodes including all hospitalizations due to COVID-19 in 2020, unlike studies in individual regions or hospitals. (ii) We use data at the postcode level, and exposures were linked in a fine 10 km2 grid rather than a few stations.

5. Conclusions

Short-term exposure to air pollutants impacts COVID-19 outcomes. During the lockdown, PM2.5, PM10, NO2, but also SO2, significantly impacted hospital admission, hospital mortality, and ICU admission. ICU mortality was mainly associated with PM2.5 and PM10 during the same period. The influence of pollutants in COVID-19 outcomes during the post-lockdown period was much lower. Our findings reveal the importance of monitoring air pollutants in respiratory infectious diseases.

Credit author statement

Conceptualization: ALEJANDRO ÁLVARO-MECA and LAURA SÁNCHEZ-DE PRADA. Formal Analysis: ALEJANDRO ÁLVARO-MECA and CARLOS GINER-BAIXAULI.Funding Acquisition: EDUARDO TAMAYO and SALVADOR RESINO. Investigation: ALEJANDRO ÁLVARO-MECA, LAURA SÁNCHEZ-DE PRADA, and SALVADOR RESINO. Resources: ALEJANDRO ÁLVARO-MECA and LAURA SÁNCHEZ-DE PRADA. Supervision: SALVADOR RESINO and EDUARDO TAMAYO. Visualization: ALEJANDRO ÁLVARO-MECA and LAURA SÁNCHEZ-DE PRADA. Writing – Original Draft Preparation: ALEJANDRO ÁLVARO-MECA and LAURA SÁNCHEZ-DE, PRADA. Writing, Review & Editing: SALVADOR RESINO, ÁLVARO TAMAYO-VELASCO, F.JAVIER ÁLVAREZ, MARTA MARTÍN-FERNÁNDEZ and JOSÉ MARÍA EIROS- BACHILLER.

All authors read and approved the final manuscript.

Consent for publication

Not applicable.

Funding

This research was supported by CIBER -Consorcio Centro de Investigación Biomédica en Red-(CB, 2021), Instituto de Salud Carlos III, Ministerio de Ciencia e Innovación and Unión Europea – Next Generation EU (CB21/13/00044 and CB21/13/00051). L. Sánchez-de Prada received a Río Hortega grant (CM20/00138) from Instituto Carlos III (Co-funded by European Regional Development Fund/European Social Fund “A way to make Europe”/“Investing in your future”).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

We wish to thank the Spanish MHCSW for providing the records of the MBDS.

Handling Editor: Jose L Domingo

Appendix A.

Table A1.

Spearman's correlation matrix among environmental factors. Abbreviations: CO: Carbon monoxide; NO: Nitrogen monoxide; NO2: Nitrogen dioxide; SO2: Sulfur dioxide; O3: Ozone; PM10: Particulate matter < 10 μm; PM2.5:Particulate matter < 2.5 μm; RH: Relative Humidity; Temp: Temperature above 2m surface; *: p-value < 0.001

CO NO NO2 SO2 PM10 PM2.5 O3 RH Temp
CO 1.00
NO 0.56∗ 1.00
NO2 0.63∗ 0.96∗ 1.00
SO2 0.51∗ 0.73∗ 0.76∗ 1.00
PM10 0.29∗ 0.47∗ 0.51∗ 0.48∗ 1.00
PM2.5 0.40∗ 0.54∗ 0.59∗ 0.51∗ 0.94∗ 1.00
O3 −0.63∗ −0.43∗ −0.48∗ −0.28∗ −0.17∗ −0.28∗ 1.00
RH 0.37∗ 0.00 0.10∗ 0.00 0.05 0.09 −0.60 1.00
Temp −0.55∗ −0.17∗ −0.21∗ −0.05∗ 0.12∗ 0.03∗ 0.67∗ −0.60∗ 1.00
Open in a new tab

Table A.2.

Summary of the association between environmental factors and clinical out-comes due to COVID-19 in patients without chronic lower respiratory disease. Abbreviations: PC: Percentage of change (%), computed by GAM adjusted for temperature, humidity, and day of the week. (*) Increases of 0.1 μg/m3 for NO and SO2. CI95%, 95% of the confidence interval. Lag: Moving average lag effect at 3, 5 and 7 days. q-value: False discovery rate q-value
Hospital Admission
 Hospital Mortality
First Wave
 Second Wave
 First Wave
 Second Wave
Enviromental factor
 Lag
 N.° events
 PC (CI95%)
 q-value
 N.°events
 PC (CI95%)
 q-value
 N.° events
 PC (CI95%)
 q-value
 N.° events
 PC (CI95%)
 q-value
CO Lag 3 88,905 1.12 (1.03–1.20) <0.001 91,280 0.28 (0.25–0.31) <0.001 16,154 0.44 (0.36–0.51) <0.001 12,950 0.07 (0.05–0.09) <0.001
NO∗ Lag 3 88,905 2.04 (1.87–2.22) <0.001 91,280 0.25 (0.21–0.29) <0.001 16,154 0.94 (0.76–1.12) <0.001 12,950 0.09 (0.05–0.14) <0.001
NO2 Lag 3 88,905 5.40 (4.98–5.81) <0.001 91,280 2.13 (2.01–2.25) <0.001 16,154 2.13 (1.77–2.50) <0.001 12,950 0.53 (0.42–0.65) <0.001
SO2∗ Lag 3 88,905 4.25 (3.96–4.55) <0.001 91,280 0.79 (0.71–0.87) <0.001 16,154 1.81 (1.55–2.07) <0.001 12,950 0.20 (0.12–0.27) <0.001
O3 Lag 3 88,905 0.00 (−0.22–0.23) >0.999 91,280 0.00 (−0.06–0.06) >0.999 16,154 0.07 (−0.12–0.26) >0.999 12,950 −0.02 (−0.07–0.04) >0.999
PM10 Lag 3 88,905 4.38 (3.97–4.79) <0.001 91,280 1.84 (1.68–2.00) <0.001 16,154 2.34 (1.97–2.70) <0.001 12,950 0.49 (0.36–0.63) <0.001
PM25 Lag 3 88,905 8.38 (7.68–9.09) <0.001 91,280 3.01 (2.78–3.25) <0.001 16,154 3.89 (3.28–4.51) <0.001 12,950 0.71 (0.51–0.92) <0.001
CO Lag 5 88,905 1.22 (1.12–1.31) <0.001 91,280 0.30 (0.28–0.33) <0.001 16,154 0.47 (0.39–0.56) <0.001 12,950 0.08 (0.06–0.10) <0.001
NO∗ Lag 5 88,905 1.97 (1.80–2.13) <0.001 91,280 0.27 (0.23–0.31) <0.001 16,154 1.00 (0.81–1.18) <0.001 12,950 0.11 (0.06–0.16) <0.001
NO2 Lag 5 88,905 5.44 (5.02–5.87) <0.001 91,280 2.28 (2.16–2.41) <0.001 16,154 2.25 (1.87–2.62) <0.001 12,950 0.59 (0.46–0.71) <0.001
SO2∗ Lag 5 88,905 4.33 (4.03–4.64) <0.001 91,280 0.83 (0.74–0.91) <0.001 16,154 1.88 (1.61–2.15) <0.001 12,950 0.20 (0.12–0.28) <0.001
O3 Lag 5 88,905 −0.25 (−0.49–0.00) 0.124 91,280 0.01 (−0.06–0.07) >0.999 16,154 −0.12 (−0.33–0.09) 0.670 12,950 −0.01 (−0.06–0.05) >0.999
PM10 Lag 5 88,905 5.30 (4.83–5.79) <0.001 91,280 2.20 (2.02–2.38) <0.001 16,154 2.90 (2.48–3.32) <0.001 12,950 0.57 (0.42–0.73) <0.001
PM25 Lag 5 88,905 9.73 (8.93–10.54) <0.001 91,280 3.46 (3.20–3.72) <0.001 16,154 4.72 (4.03–5.42) <0.001 12,950 0.80 (0.57–1.02) <0.001
CO Lag 7 88,905 1.33 (1.22–1.43) <0.001 91,280 0.32 (0.29–0.35) <0.001 16,154 0.54 (0.44–0.63) <0.001 12,950 0.08 (0.06–0.11) <0.001
NO∗ Lag 7 88,905 2.10 (1.92–2.28) <0.001 91,280 0.29 (0.25–0.33) <0.001 16,154 1.08 (0.89–1.26) <0.001 12,950 0.11 (0.05–0.16) <0.001
NO2 Lag 7 88,905 5.54 (5.11–5.98) <0.001 91,280 2.40 (2.27–2.53) <0.001 16,154 2.44 (2.06–2.82) <0.001 12,950 0.59 (0.46–0.72) <0.001
SO2∗ Lag 7 88,905 4.45 (4.13–4.77) <0.001 91,280 0.84 (0.76–0.93) <0.001 16,154 2.00 (1.73–2.28) <0.001 12,950 0.19 (0.11–0.27) <0.001
O3 Lag 7 88,905 −0.50 (−0.76–0.23) 0.001 91,280 0.01 (−0.06–0.08) >0.999 16,154 −0.33 (−0.56–0.10) 0.012 12,950 0.01 (−0.05–0.07) >0.999
PM10 Lag 7 88,905 6.57 (6.04–7.10) <0.001 91,280 2.51 (2.32–2.70) <0.001 16,154 3.66 (3.19–4.12) <0.001 12,950 0.63 (0.46–0.79) <0.001
PM25 Lag 7 88,905 11.19 (10.32–12.08) <0.001 91,280 3.84 (3.57–4.12) <0.001 16,154 5.67 (4.93–6.43) <0.001 12,950 0.83 (0.59–1.08) <0.001
ICU Admission ICU Mortality
CO Lag 3 7977 0.26 (0.15–0.38) <0.001 8543 0.05 (0.02–0.08) 0.008 2623 0.13 (0.05–0.21) 0.004 2478 0.00 (−0.02–0.03) >0.999
NO∗ Lag 3 7977 0.65 (0.38–0.92) <0.001 8543 0.10 (0.05–0.16) <0.001 2623 0.31 (0.12–0.49) 0.004 2478 0.02 (−0.03–0.06) >0.999
NO2 Lag 3 7977 1.30 (0.76–1.85) <0.001 8543 0.77 (0.62–0.92) <0.001 2623 0.65 (0.28–1.03) 0.003 2478 0.15 (0.02–0.27) 0.114
SO2∗ Lag 3 7977 1.01 (0.61–1.41) <0.001 8543 0.25 (0.16–0.35) <0.001 2623 0.55 (0.28–0.83) <0.001 2478 0.03 (−0.04–0.10) >0.999
O3 Lag 3 7977 0.27 (−0.02–0.55) 0.171 8543 0.03 (−0.04–0.11) 0.984 2623 0.21 (0.02–0.41) 0.087 2478 −0.03 (−0.09–0.02) >0.999
PM10 Lag 3 7977 2.04 (1.47–2.61) <0.001 8543 0.65 (0.46–0.83) <0.001 2623 1.01 (0.60–1.42) <0.001 2478 0.21 (0.07–0.35) 0.060
PM25 Lag 3 7977 3.12 (2.21–4.04) <0.001 8543 0.98 (0.71–1.25) <0.001 2623 1.55 (0.89–2.20) <0.001 2478 0.28 (0.07–0.48) 0.075
CO Lag 5 7977 0.26 (0.13–0.39) <0.001 8543 0.04 (0.01–0.07) 0.068 2623 0.11 (0.02–0.20) 0.054 2478 0.01 (−0.02–0.03) >0.999
NO∗ Lag 5 7977 0.64 (0.37–0.92) <0.001 8543 0.12 (0.06–0.18) <0.001 2623 0.32 (0.13–0.51) 0.004 2478 0.02 (−0.03–0.07) >0.999
NO2 Lag 5 7977 1.33 (0.77–1.90) <0.001 8543 0.83 (0.67–0.99) <0.001 2623 0.66 (0.27–1.05) 0.004 2478 0.16 (0.03–0.29) 0.129
SO∗2 Lag 5 7977 1.01 (0.60–1.42) <0.001 8543 0.26 (0.15–0.36) <0.001 2623 0.56 (0.27–0.84) 0.001 2478 0.02 (−0.05–0.10) >0.999
O3 Lag 5 7977 0.19 (−0.13–0.50) 0.640 8543 0.04 (−0.04–0.12) 0.794 2623 0.16 (−0.06–0.38) 0.427 2478 −0.03 (−0.09–0.03) >0.999
PM10 Lag 5 7977 2.50 (1.83–3.17) <0.001 8543 0.80 (0.59–1.00) <0.001 2623 1.15 (0.67–1.64) <0.001 2478 0.20 (0.04–0.36) 0.129
PM25 Lag 5 7977 3.67 (2.62–4.72) <0.001 8543 1.16 (0.86–1.46) <0.001 2623 1.71 (0.95–2.46) <0.001 2478 0.27 (0.04–0.50) 0.132
CO Lag 7 7977 0.28 (0.14–0.42) <0.001 8543 0.02 (−0.01–0.06) 0.506 2623 0.10 (0.01–0.20) 0.116 2478 0.01 (−0.01–0.04) >0.999
NO∗ Lag 7 7977 0.68 (0.40–0.96) <0.001 8543 0.12 (0.06–0.19) <0.001 2623 0.35 (0.16–0.54) 0.001 2478 0.02 (−0.03–0.08) >0.999
NO2 Lag 7 7977 1.42 (0.85–2.00) <0.001 8543 0.85 (0.69–1.02) <0.001 2623 0.72 (0.32–1.11) 0.001 2478 0.18 (0.05–0.32) 0.100
SO2∗ Lag 7 7977 1.05 (0.63–1.47) <0.001 8543 0.25 (0.14–0.35) <0.001 2623 0.59 (0.30–0.88) <0.001 2478 0.03 (−0.05–0.11) >0.999
O3 Lag 7 7977 0.08 (−0.26–0.42) >0.999 8543 0.05 (−0.04–0.13) 0.760 2623 0.08 (−0.16–0.32) >0.999 2478 −0.03 (−0.10–0.03) >0.999
PM10 Lag 7 7977 3.10 (2.35–3.86) <0.001 8543 0.93 (0.71–1.15) <0.001 2623 1.45 (0.91–2.00) <0.001 2478 0.21 (0.04–0.38) 0.100
PM25 Lag 7 7977 4.29 (3.15–5.44) <0.001 8543 1.30 (0.98–1.63) <0.001 2623 1.99 (1.17–2.81) <0.001 2478 0.31 (0.06–0.56) 0.100
Open in a new tab

Table A.3.

Summary of the association between environmental factors and clinical out-comes due to COVID-19 in patients with chronic lower respiratory disease. Abbreviations:PC: Percentage of change (%), computed by GAM adjusted for temperature, humidity, and day of the week. (*) Increases of 0.1 μg/m3 for NO and SO2. CI95%, 95% of the confidence interval. Lag: Moving average lag effect at 3, 5 and 7 days. q-value: False discovery rate q-value
Hospital Admission
 Hospital Mortality
First Wave
 Second Wave
Enviromental factor Lag N.° events PC (CI95%) Enviromental factor Lag N.° events PC (CI95%) Enviromental factor Lag N.° events PC (CI95%) Enviromental factor Lag
CO Lag 3 14,249 0.41 (0.32–0.49) <0.001 14,310 0.08 (0.05–0.11) <0.001 3163 0.12 (0.05–0.18) 0.002 2600 0.03 (0.00–0.05) 0.059
NO∗ Lag 3 14,249 0.36 (0.27–0.46) <0.001 14,310 0.10 (0.06–0.14) <0.001 3163 0.26 (0.11–0.40) 0.002 2600 0.04 (0.00–0.08) 0.092
NO2 Lag 3 14,249 2.00 (1.59–2.42) <0.001 14,310 0.91 (0.78–1.05) <0.001 3163 0.60 (0.29–0.90) <0.001 2600 0.28 (0.18–0.39) <0.001
SO2∗ Lag 3 14,249 1.72 (1.42–2.02) <0.001 14,310 0.38 (0.29–0.47) <0.001 3163 0.55 (0.33–0.78) <0.001 2600 0.13 (0.05–0.20) 0.003
O3 Lag 3 14,249 0.11 (−0.11–0.33) 0.821 14,310 0.03 (−0.04–0.10) >0.999 3163 0.01 (−0.15–0.17) >0.999 2600 −0.00 (−0.05–0.05) >0.999
PM10 Lag 3 14,249 2.53 (2.11–2.96) <0.001 14,310 1.06 (0.88–1.23) <0.001 3163 1.10 (0.79–1.42) <0.001 2600 0.31 (0.18–0.44) <0.001
PM25 Lag 3 14,249 4.19 (3.49–4.90) <0.001 14,310 1.58 (1.33–1.83) <0.001 3163 1.55 (1.04–2.07) <0.001 2600 0.43 (0.24–0.63) <0.001
CO Lag 5 14,249 0.44 (0.34–0.54) <0.001 14,310 0.09 (0.05–0.12) <0.001 3163 0.11 (0.04–0.18) 0.010 2600 0.03 (0.01–0.06) 0.024
NO∗ Lag 5 14,249 0.43 (0.32–0.54) <0.001 14,310 0.12 (0.07–0.16) <0.001 3163 0.25 (0.10–0.40) 0.004 2600 0.05 (0.01–0.09) 0.065
NO2 Lag 5 14,249 2.05 (1.63–2.47) <0.001 14,310 0.98 (0.84–1.12) <0.001 3163 0.60 (0.29–0.91) 0.001 2600 0.30 (0.19–0.41) <0.001
SO2∗ Lag 5 14,249 1.75 (1.44–2.07) <0.001 14,310 0.40 (0.30–0.50) <0.001 3163 0.54 (0.31–0.77) <0.001 2600 0.13 (0.06–0.21) 0.002
O3 Lag 5 14,249 0.03 (−0.21–0.27) >0.999 14,310 0.04 (−0.03–0.12) 0.662 3163 −0.06 (−0.23–0.12) >0.999 2600 0.00 (−0.05–0.06) >0.999
PM10 Lag 5 14,249 3.14 (2.64–3.64) <0.001 14,310 1.21 (1.01–1.40) <0.001 3163 1.36 (0.99–1.74) <0.001 2600 0.34 (0.19–0.49) <0.001
PM25 Lag 5 14,249 5.08 (4.27–5.90) <0.001 14,310 1.78 (1.51–2.06) <0.001 3163 1.87 (1.28–2.46) <0.001 2600 0.47 (0.26–0.69) <0.001
CO Lag 7 14,249 0.50 (0.39–0.60) <0.001 14,310 0.09 (0.06–0.12) <0.001 3163 0.12 (0.04–0.20) 0.008 2600 0.04 (0.01–0.06) 0.009
NO∗ Lag 7 14,249 0.52 (0.41–0.64) <0.001 14,310 0.12 (0.08–0.17) <0.001 3163 0.29 (0.14–0.44) 0.001 2600 0.05 (0.01–0.09) 0.050
NO2 Lag 7 14,249 2.14 (1.71–2.57) <0.001 14,310 1.04 (0.89–1.18) <0.001 3163 0.68 (0.36–1.00) <0.001 2600 0.32 (0.20–0.43) <0.001
SO2∗ Lag 7 14,249 1.82 (1.50–2.14) <0.001 14,310 0.41 (0.31–0.51) <0.001 3163 0.57 (0.34–0.81) <0.001 2600 0.13 (0.06–0.21) 0.002
O3 Lag 7 14,249 −0.09 (−0.34–0.17) >0.999 14,310 0.05 (−0.03–0.13) 0.501 3163 −0.17 (−0.36–0.02) 0.199 2600 0.02 (−0.04–0.07) >0.999
PM10 Lag 7 14,249 3.86 (3.31–4.42) <0.001 14,310 1.34 (1.13–1.55) <0.001 3163 1.68 (1.27–2.10) <0.001 2600 0.35 (0.20–0.51) <0.001
PM25 Lag 7 14,249 5.98 (5.10–6.87) <0.001 14,310 1.93 (1.63–2.23) <0.001 3163 2.24 (1.60–2.89) <0.001 2600 0.48 (0.25–0.71) <0.001
ICU Admission ICU Mortality
CO Lag 3 1117 0.03 (−0.06–0.11) >0.999 1278 0.00 (−0.02–0.03) >0.999 434 −0.04 (−0.12–0.03) >0.999 495 −0.01 (−0.03–0.01) >0.999
NO∗ Lag 3 1117 0.01 (−0.09–0.11) >0.999 1278 0.04 (0.00–0.09) 0.143 434 0.02 (−0.12–0.16) >0.999 495 0.00 (−0.06–0.06) >0.999
NO2 Lag 3 1117 0.13 (−0.26–0.52) >0.999 1278 0.24 (0.12–0.37) 0.003 434 0.07 (−0.24–0.38) >0.999 495 0.01 (−0.11–0.13) >0.999
SO2∗ Lag 3 1117 0.17 (−0.12–0.46) >0.999 1278 0.11 (0.03–0.19) 0.037 434 0.05 (−0.18–0.28) >0.999 495 0.05 (−0.02–0.12) >0.999
O3 Lag 3 1117 0.24 (0.02–0.46) 0.208 1278 −0.02 (−0.08–0.04) >0.999 434 −0.03 (−0.20–0.14) >0.999 495 −0.03 (−0.09–0.03) >0.999
PM10 Lag 3 1117 0.78 (0.34–1.22) 0.010 1278 0.27 (0.12–0.41) 0.004 434 0.34 (−0.03–0.72) >0.999 495 0.08 (−0.06–0.22) >0.999
PM25 Lag 3 1117 1.06 (0.35–1.77) 0.032 1278 0.37 (0.15–0.59) 0.006 434 0.35 (−0.21–0.92) >0.999 495 0.07 (−0.14–0.28) >0.999
CO Lag 5 1117 0.04 (−0.06–0.14) >0.999 1278 0.01 (−0.02–0.04) >0.999 434 −0.02 (−0.10–0.06) >0.999 495 0.00 (−0.02–0.03) >0.999
NO∗ Lag 5 1117 0.02 (−0.09–0.13) >0.999 1278 0.04 (−0.01–0.09) 0.311 434 0.04 (−0.10–0.19) >0.999 495 0.00 (−0.06–0.07) >0.999
NO2 Lag 5 1117 0.21 (−0.20–0.61) >0.999 1278 0.24 (0.11–0.38) 0.008 434 0.14 (−0.19–0.46) >0.999 495 0.03 (−0.10–0.16) >0.999
SO2∗ Lag 5 1117 0.22 (−0.08–0.52) 0.704 1278 0.10 (0.02–0.19) 0.083 434 0.09 (−0.14–0.33) >0.999 495 0.05 (−0.03–0.13) >0.999
O3 Lag 5 1117 0.22 (−0.03–0.46) 0.491 1278 0.00 (−0.06–0.07) >0.999 434 −0.06 (−0.24–0.13) >0.999 495 −0.02 (−0.08–0.04) >0.999
PM10 Lag 5 1117 1.04 (0.52–1.57) 0.002 1278 0.26 (0.09–0.43) 0.023 434 0.47 (0.04–0.90) 0.536 495 0.07 (−0.09–0.24) >0.999
PM25 Lag 5 1117 1.49 (0.68–2.30) 0.003 1278 0.35 (0.10–0.60) 0.036 434 0.62 (−0.02–1.26) 0.536 495 0.07 (−0.17–0.31) >0.999
CO Lag 7 1117 0.05 (−0.05–0.16) 0.948 1278 0.01 (−0.02–0.04) >0.999 434 −0.01 (−0.10–0.07) >0.999 495 0.01 (−0.02–0.04) >0.999
NO∗ Lag 7 1117 0.03 (−0.09–0.16) >0.999 1278 0.04 (−0.01–0.09) 0.431 434 0.05 (−0.10–0.21) >0.999 495 0.00 (−0.06–0.07) >0.999
NO2 Lag 7 1117 0.28 (−0.14–0.69) 0.692 1278 0.25 (0.10–0.39) 0.012 434 0.15 (−0.18–0.48) >0.999 495 0.04 (−0.09–0.18) >0.999
SO2∗ Lag 7 1117 0.27 (−0.04–0.58) 0.555 1278 0.10 (0.01–0.19) 0.110 434 0.10 (−0.15–0.34) >0.999 495 0.06 (−0.02–0.14) >0.999
O3 Lag 7 1117 0.18 (−0.07–0.44) 0.692 1278 0.01 (−0.06–0.08) >0.999 434 −0.07 (−0.28–0.13) >0.999 495 −0.03 (−0.09–0.04) >0.999
PM10 Lag 7 1117 1.26 (0.68–1.85) <0.001 1278 0.26 (0.08–0.45) 0.052 434 0.58 (0.10–1.06) 0.336 495 0.09 (−0.08–0.27) >0.999
PM25 Lag 7 1117 1.73 (0.88–2.60) 0.001 1278 0.36 (0.09–0.63) 0.055 434 0.62 (−0.05–1.30) 0.635 495 0.13 (−0.13–0.39) >0.999
Open in a new tab

Fig. A.1.

Fig. A.1

Open in a new tab

Mean of environmental effects in the Lockdown 2020 period (center) compared with the same period in 2019 (left) and 2021 (right)

Fig. A.2.

Fig. A.2

Open in a new tab

Mean of environmental effects in the Post-lockdown 2020 period (center) compared with the same period in 2019 (left) and 2021 (right)

Fig. A.3.

Fig. A.3

Open in a new tab

Summary of the distribution of each pollutant with its cut-off value for each outcome, obtained by multi-output decision tree regression at lag3 days during post-lockdown

Data availability

Data will be made available on request.

References

  1. Alduchov O.A., Eskridge R.E. Improved magnus form approximation of saturation vapor pressure. J. Appl. Meteorol. Climatol. 1996;35(4):601–609. doi: 10.1175/1520-0450(1996)035&#x0003c;0601:IMFAOS&#x0003e;2.0.CO;2. [DOI] [Google Scholar]
  2. Ayuntamiento de Valladolid, óxidos de Nitrógeno NO/NO2. 2022. https://www.valladolid.es/es/rccava/contaminantes/oxidos-nitrogeno-no2 Available at:
  3. Beltramo G., Cottenet J., Mariet A.-S., Georges M., Piroth L., Tubert-Bitter P., Bonniaud P., Quantin C. Chronic respiratory diseases are predictors of severe outcome in Covid-19 hospitalised patients: a nationwide study. Eur. Respir. J. 2021;58(6) doi: 10.1183/13993003.04474-2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Boletín Oficial del Estado (67): 25390–25400 14 de marzo de 2020. ISSN 0212-033X, Real Decreto 463/2020, de 14 de marzo, por el que se declara el estado de alarma para la gestión de la situación de crisis sanitaria ocasionada por el COVID-19. 2020. https://www.boe.es/buscar/doc.php?id=BOE-A-2020-3692 Available at.
  5. Boningari T., Smirniotis P.G. Impact of nitrogen oxides on the environment and human health: Mn-based materials for the NO x abatement. Curr. Opion. Chem. Eng. 2016;13:133–141. doi: 10.1016/j.coche.2016.09.004. [DOI] [Google Scholar]
  6. Bourdrel T., Annesi-Maesano I., Alahmad B., Maesano C.N., Bind M.-A. The impact of outdoor air pollution on covid-19: a review of evidence from “in vitro”. Anim Hum stud, Euro- pean Respir.Rev. 2021;30 doi: 10.1183/16000617.0242-2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bozack A., Pierre S., DeFelice N., Colicino E., Jack D., Chillrud S.N., Rundle A., Astua A., Quinn J.W., McGuinn L., Yang Q., Johnson K., Masci J., Lukban L., Maru D., Lee A.G. Long-term air pollution exposure and covid-19 mortality: a patient-level analysis from New York city. Am. J. Respir. Crit. Care Med. 2022;205:651–662. doi: 10.1164/rccm.202104-0845OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Chen Z., Sidell M.A., Huang B.Z., Chow T., Eckel S.P., Martinez M.P., Gheissari R., Lurmann F., Thomas D.C., Gilliland F.D., Xiang A.H. Ambient air pollutant exposures and covid-19 severity and mortality in a cohort of patients with covid-19 in southern California. Am. J. Respir. Crit. Care Med. 2022;206:440–448. doi: 10.1164/RCCM.202108-1909OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Choi E.P.H., Hui B.P.H., Wan E.Y.F., Kwok J.Y.Y., Tam T.H.L., Wu C. Covid-19 and health-related quality of life: a community based online survey in Hong Kong. Int. J. Environ. Res. Publ. Health. 2021;18:3228. doi: 10.3390/ijerph18063228. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Coker E.S., Cavalli L., Fabrizi E., Guastella G., Lippo E., Maria, Parisi L., Pontarollo N., Rizzati M., Varacca A., Vergalli S. The effects of air pollution on Covid-19 related mortality in northern Italy. Environ. Resour. Econ. 2020;76:611–634. doi: 10.1007/s10640-020-00486-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Copat C., Cristaldi A., Fiore M., Grasso A., Zuccarello P., Signorelli S.S., Conti G.O., Ferrante M. The role of air pollution (PM and NO2) in Covid-19 spread and lethality: a systematic review. Environ. Res. 2020;191 doi: 10.1016/j.envres.2020.110129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Domingo J.L., Rovira J. Effects of air pollutants on the transmission and severity of respiratory viral infections. Environ. Res. 8 2020;187 doi: 10.1016/j.envres.2020.109650. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Domínguez-Amarillo S., Fernández-Agüera J., Cesteros-García S., González-Lezcano R.A. Bad air can also kill: residential indoor air quality and pollutant exposure risk during the Covid-19 crisis. Int. J. Environ. Res. Publ. Health. 2020;17:7183. doi: 10.3390/ijerph17197183. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Donzelli G., Cioni L., Cancellieri M., Llopis-Morales A., Morales- Suárez-Varela M. Relations between air quality and covid-19 lockdown measures in Valencia, Spain. Int. J. Environ. Res. Publ. Health. 2021;18(5) doi: 10.3390/ijerph18052296. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Dumont M., Marée R., Wehenkel L., Geurts P. vol. 2. 2009. Fast Multi-Class Image Annotation with Random Subwindows and Multiple Output Randomized Trees; pp. 196–203. [Google Scholar]
  16. Equipo COVID-19. Red Nacional de Vigilancia Epidemiológica (RENAVE). Centro Nacional de Microbiología (CNM) vol. 5. Instituto de investigación Carlos III (ISCIII); 2020. https://www.isciii.es/QueHacemos/Servicios/VigilanciaSaludPublicaRENAVE/EnfermedadesTransmisibles/Paginas/Informes_Previos_Covid-12_2020.aspx (Informe n° 32. situación de covid-19 en españa). Available at. [Google Scholar]
  17. Equipo COVID-19. Red Nacional de Vigilancia Epidemiológica (RENAVE). Centro Nacional de Microbiología (CNM) vol. 1. Instituto de investigación Carlos III (ISCIII); 2021. https://www.isciii.es/QueHacemos/Servicios/VigilanciaSaludPublicaRENAVE/EnfermedadesTransmisibles/Paginas/Informes_Previos_Covid-12_2021.aspx (Informe n° 60. situación de covid-19 en españa). Available at. [Google Scholar]
  18. Feng S., Gao D., Liao F., Zhou F., Wang X. The health effects of ambient pm2.5 and potential mechanisms. Ecotoxicol. Environ. Saf. 2016;128:67–74. doi: 10.1016/j.ecoenv.2016.01.030. [DOI] [PubMed] [Google Scholar]
  19. Gerayeli F.V., Milne S., Cheung C., Li X., Yang C.W.T., Tam A., Choi L.H., Bae A., Sin D.D. vol. 33. 2021. Copd and the Risk of Poor Outcomes in Covid-19: A Systematic Review and Meta-Analysis. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Gobierno de España Crisis sanitaria COVID-19: normativa e información útil. Estado de alarma 25 de octubre 2020. 2020. https://administracion.gob.es/pag_Home/atencionCiudadana/Crisis-sanitaria-COVID-19.html Available at.
  21. Hastie T., Tibshirani R. Wiley Online Library; 1990. Generalized Additive Models. [DOI] [PubMed] [Google Scholar]
  22. Hendryx M., Luo J. Covid-19 prevalence and fatality rates in association with air pollution emission concentrations and emission sources. Environ. Pollu. 2020;265(10) doi: 10.1016/J.ENVPOL.2020.115126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. ICD-10-CM Diagnosis codes. 2022. www.icd10data.com Available at:
  24. Jevtic M., Matkovic V., van den Hazel P., Bouland C. Environment lockdown, air pollution and related diseases: could we learn something and make it last? Eur. J. Publ. Health. 2021;31 doi: 10.1093/eurpub/ckab157. iv36–iv39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Jia C., Fu X., Bartelli D., Smith L. Insignificant impact of the “stay- at-home” order on ambient air quality in the Memphis metropolitan area. u.s.a., Atmosphere. 2020;11(2020):630. doi: 10.3390/ATMOS11060630. Page 630 11. [DOI] [Google Scholar]
  26. Jiang Y., Xu J. The association between covid-19 deaths and short-term ambient air pollution/meteorological condition exposure: a retrospective study from Wuhan, China. Air Qual. Atmos. Health. 2021;14:1–5. doi: 10.1007/s11869-020-00906-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kan H., Wong C.-M., Vichit-Vadakan N., Qian Z., the PAPA Project Teams. Vajanapoom N., Wong H.K.C., Thach T., Chau P., Chan K., Chung R., Ou C., Yang L., Thomas G., Lam T., Hedley A., Wong T., Kan S.H., Chen B., Zhao N., Zhang Y., Yang N., Wuhan D.Z. Short-term association between sulfur dioxide and daily mortality: the public health and air pollution in Asia (papa) study † project teams by location: Bangkok-n. vichit-vadakan and. Environ. Res. 2010;110:258–264. doi: 10.1016/j.envres.2010.01.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Katsouyanni K., Touloumi G., Spix C., Schwartz J., Balducci F., Medina S., Rossi G., Wojtyniak B., Sunyer J., Bacharova L., Schouten J.P., Ponka A., Anderson H.R. Short term effects of ambient sulphur dioxide and particulate matter on mortality in 12 European cities: results from time series data from the aphea project. BMJ. 1997;314:1658. doi: 10.1136/bmj.314.7095.1658. 1658. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Khorsandi B., Farzad K., Tahriri H., Maknoon R. Association between short-term exposure to air pollution and Covid-19 hospital admission/mortality during warm seasons. Environ. Monit. Assess. 2021;193:426. doi: 10.1007/s10661-021-09210-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kraemer M.U.G., Yang C.-H., Gutierrez B., Wu C.-H., Klein B., Pigott D.M., null null, du Plessis L., Faria N.R., Li R., Hanage W.P., Brownstein J.S., Layan M., Vespignani A., Tian H., Dye C., Pybus O.G., Scarpino S.V. The effect of human mobility and control measures on the covid-19 epidemic in China. Science. 2020;368(6490):493–497. doi: 10.1126/science.abb4218. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Lu R., Zhao X., Li J., Niu P., Yang B., Wu H., Wang W., Song H., Huang B., Zhu N., Bi Y., Ma X., Zhan F., Wang L., Hu T., Zhou H., Hu Z., Zhou W., Zhao L., Chen J., Meng Y., Wang J., Lin Y., Yuan J., Xie Z., Ma J., Liu W.J., Wang D., Xu W., Holmes E.C., Gao G.F., Wu G., Chen W., Shi W., Tan W. Genomic characterisation and epidemiology of 2019 novel coronavirus: implications for virus ori- gins and receptor binding. Lancet. 2020;395(10224):565–574. doi: 10.1016/S0140-6736(20)30251-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Martelletti L., Martelletti P. Air pollution and the novel Covid-19 disease: a putative disease risk factor. SN Compr Clin. Med. 2020;2:383–387. doi: 10.1007/s42399-020-00274-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. McCullagh P., Nelder J. Chapman & Hall CRC; 1989. Generalized Linear Models. [DOI] [Google Scholar]
  34. Menut L., Bessagnet B., Siour G., Mailler S., Pennel R., Cholakian A. Impact of lockdown measures to combat covid-19 on air quality over western europe. Sci. Total Environ. 2020;741(11) doi: 10.1016/J.SCITOTENV.2020.140426. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Meteo France. Institut national de l’environnement industriel et des risques (Ineris), Aarhus University, Norwegian Meteorological Institute (MET Norway), Jülich Institut für Energie- und Klimaforschung (IEK), Institute of Environmental Protection National Research Institute (IEP-NRI), Koninklijk Nederlands Meteorologisch Instituut (KNMI), Nederlandse Organisatie voor toegepast natuurwetenschappelijk onderzoek (TNO), Swedish Meteorological and Hydrological Institute (SMHI), Finnish Meteorological Institute (FMI), Italian National Agency for New Technologies, Energy and Sustainable Economic Development (ENEA) and Barcelona Supercomputing Center (BSC), CAMS European air quality forecasts, ENSEMBLE data. Copernicus Atmosphere Monitoring Service (CAMS) Atmosphere Data Store (ADS); 2022. Available at: [Google Scholar]
  36. Muñoz Sabater J. 2022. ERA5-Land Hourly Data from 1950 to 1980. Copernicus Climate Change Service (C3S) Climate Data Store (CDS) [DOI] [Google Scholar]
  37. Ray S., Kim K.-H. The pollution status of sulfur dioxide in major urban areas of Korea between 1989 and 2010. Atmos. Res. 2014;147–148:101–110. doi: 10.1016/j.atmosres.2014.05.011. [DOI] [Google Scholar]
  38. Robertson S., Miller M.R. Ambient air pollution and thrombosis. Part. Fibre Toxicol. 2018;15 doi: 10.1186/s12989-017-0237-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Sarmadi M., Moghanddam V.K., Dickerson A.S., Martelletti L. Association of Covid-19 distribution with air quality, sociodemographic factors, and comorbidities: an ecological study of us states. Air Qual. Atmos.Health. 2021;14:455–465. doi: 10.1007/s11869-020-00949-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Schiermeier Q. Why pollution is plummeting in some cities but not others. Nature. 2020;580:313. doi: 10.1038/d41586-020-01049-6. 313. [DOI] [PubMed] [Google Scholar]
  41. Srivastava A. Covid-19 and air pollution and meteorology-an intricate relationship: a review. Chemosphere. 2021;263(1) doi: 10.1016/J.CHEMOSPHERE.2020.128297. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Stieb D.M., Judek S., Burnett R.T. Meta-analysis of time-series studies of air pollution and mortality: effects of gases and particles and the influence of cause of death, age, and season. J. Air Waste Manag. Assoc. 2002;52:470–484. doi: 10.1080/10473289.2002.10470794. [DOI] [PubMed] [Google Scholar]
  43. Stieb D.M., Judek S., Burnett R.T. Meta-analysis of time-series studies of air pollution and mortality: update in relation to the use of generalized additive models. J. Air Waste Manag. Assoc. 2003;53:258–261. doi: 10.1080/10473289.2003.10466149. [DOI] [PubMed] [Google Scholar]
  44. Subdirección General de Información Sanitaria e Innovación, Registro de Actividad de Atención Especializada RAE-CMBD. 2016. https://www.sanidad.gob.es/estadEstudios/estadisticas/cmbdhome.htm Available at.
  45. Sunyer J., Atkinson R., Ballester F., Tertre A.L., Ayres J.G., Forastiere F., Forsberg B., Vonk J.M., Bisanti L., Anderson R.H., Schwartz J., Katsouyanni K. Respiratory effects of sulphur dioxide: a hierarchical multicity analysis in the aphea 2 study. Occup. Environ. Med. 2003;60:2. doi: 10.1136/oem.60.8.e2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Sunyer J., Ballester F., Tertre A.L., Atkinson R., Ayres J.G., Forastiere F., Forsbergg B., Vonk J.M., Bisanti L., Tenías J.M., Medina S., Schwartz J., Katsouyanni K. The association of daily sulfur dioxide air pollution levels with hospital admissions for cardiovascular diseases in Europe (the aphea-ii study) Eur. Heart J. 2003;24:752–760. doi: 10.1016/S0195-668X(02)00808-4. [DOI] [PubMed] [Google Scholar]
  47. Tateo F., Fiorino S., Peruzzo L., Zippi M., De Biase D., Lari F., Melucci D. Effects of environmental parameters and their interactions on the spreading of sars-cov-2 in north Italy under different social restrictions. a new approach based on multivariate analysis. Environ. Res. 2022;210 doi: 10.1016/j.envres.2022.112921. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Varotsos C., Christodoulakis J., Kouremadas G.A., Fotaki E.-F. The signature of the coronavirus lockdown in air pollution in Greece. Water, Air, Soil Pollut. 2021;232:119. doi: 10.1007/s11270-021-05055-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Wang B., Liu J., Li Y., Fu S., Xu X., Li L., Zhou J., Liu X., He X., Yan J., Shi Y., Niu J., Yang Y., Li Y., Luo B., Zhang K. Airborne particulate matter, population mobility and covid-19: a multi- city study in China. BMC Publ. Health. 2020;20(12) doi: 10.1186/s12889-020-09669-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Wood S.N. Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models. J. Roy. Stat. Soc. B. 2011;73(1):3–36. doi: 10.1111/j.14679868.2010.00749.x. [DOI] [Google Scholar]
  51. Woodby B., Arnold M.M., Valacchi G. Sars-cov-2 infection, covid-19 pathogenesis, and exposure to air pollution: what is the connection? Ann. N. Y. Acad. Sci. 2020 doi: 10.1111/nyas.14512. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Wu X., Nethery R.C., Sabath M.B., 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(11) doi: 10.1126/sciadv.abd4049. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Xu X., Chen P., Wang J., Feng J., Zhou H., Li X., Zhong W., Hao P. Evolution of the novel coronavirus from the ongoing Wuhan outbreak and modeling of its spike protein for risk of human trans- mission. Sci. China Life Sci. 2020;63(3):457–460. doi: 10.1007/s11427-020-1637-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Zang S.-T., Luan J., Li L., Yu H.-X., Wu Q.-J., Chang Q., Zhao Y.-H. Ambient air pollution and covid-19 risk: evidence from 35 observational studies. Environ. Res. 2022;204 doi: 10.1016/j.envres.2021.112065. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Zhu Y., Xie J., Huang F., Cao L. Association between short-term exposure to air pollution and covid-19 infection: evidence from China. Sci. Total Environ. 7 2020;727 doi: 10.1016/j.scitotenv.2020.138704. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Data will be made available on request.

Articles from Environmental Research are provided here courtesy of Elsevier

RESOURCES