Effects of air pollution on daily hospital admissions for cardiovascular diseases in Castilla-La Mancha, Spain: a region with moderate air quality

Adverse impacts of air pollution on human health have been well documented in Spain; however, very few have been conducted in the Spanish region of Castilla-La Mancha (CLM). CLM is an extensive region with a low population density, little industrial production, and moderate air pollution level. The aim of the study was to assess the relationship between the risk of hospital admission for cardiovascular disease (CVD) and exposure to PM2.5, PM10, and NO2 in CLM during 2006–2015. Daily air pollution concentration, temperature, and relative humidity were monitored from the air quality monitoring stations in CLM. A time-series analysis with generalized linear model was used to examine the effects of air pollution on hospital admissions by controlling for long-term trend and other potential confounders. The effect modifications by sex and age (15–64; ≥ 65 years) were examined. Lagging exposure concept was used to analyze a possible latency period in cumulative exposure-pollution analyses. Relative risks (RR) of CVD admissions at the same day (lag 0) and to 7 days after exposure (lag 7) were calculated. Relative risks of CVD admissions (for an increase of 10 µg m−3 in concentration of air pollutant) were calculated. The elderly group (≥ 65 years) were the most susceptible group to the effect of air pollution, whereas the estimated effect by sex was significantly different depending on the age group. The potential utility of these results may help for syndromic surveillance during future similar air pollutant concentrations, where an increase in the use of health services among the vulnerable groups could be unexpected by the emergency department.


Introduction
Cardiovascular diseases (CVD), a class of disorders involving the heart and blood vessels, contribute to about onethird of deaths worldwide (WHO 2017). CVD have also been ranked as the leading cause of death and disability in European countries, as well as in Spain (Amor et al. 2014). Patients with CVD share many risk factors such as obesity, hyperlipidemia, hypertension, smoking, poor nutrition, and inactive lifestyle (Riegel et al. 2017). The elderly population is the group who present the highest susceptibility in CVD risk factors (North et al. 2012). However, sex differences lead to discrepancies in CVD risk factors which may be attributed to sex hormones and their associated receptors (García et al. 2016). Air pollution may also be considered a risk factor and differs from the previous risk factors because exposure to air pollution, for the large majority of the world's population, is unavoidable. Therefore, even though the individual risk estimates for exposure to air pollution are relatively small compared to the cardiovascular risk factors presented previously, since exposure to air pollution is ubiquitous, the overall population attributable risk and subsequent burden are significant (Nawrot et al. 2011).
Particulate matter (PM) and nitrogen oxides (NOx) are two of the most prominent air pollutants. PM is a complex mixture of extremely small particles and liquid droplets made up of acids, organic chemicals, metals, and soil or dust particles (Anderson et al. 2012). PM tends to be categorized, measured, and regulated in relation to particle size, called coarse particles (PM 10 , particles with a diameter of 10 μm or less) and fine particles (PM 2.5 , diameter of 2.5 μm or less). Generally, the scientific studies about health effects have been based on PM 10 , since PM 2.5 has been always monitored later and less frequently, even though it has been suggested that PM 2.5 is more harmful to human health than PM 10 (Hassanvand et al. 2017). Moreover, only a limit has been set for its annual concentration, so a standard for a daily PM 2.5 concentration is still missing in Europe (European Directive 2008/50/CE).
The sources of PM may be natural, such as forest fires and mineral dust released, or the result of anthropogenic activities, such as industrial facilities, power plants, vehicular traffics, and incinerator (Adams et al. 2015), while major sources of NO 2 are usually anthropogenic emission such as vehicles, power plants, and other forms of fuel burning (Richmont-Bryant et al. 2017;Constantin et al. 2020).
Although several studies have been conducted to assess the effects of air pollution on human health in different regions of Spain (Medrano et al. 2005;Ballester et al. 2006;de Pablo et al. 2006;Gabriel et al. 2008;Amor et al. 2015;Santurtun et al. 2017;Borge et al. 2018), very few have been carried out in Castilla-La Mancha (CLM), Spain (de Pablo et al. 2013;Monsalve et al. 2013). Earlier Spanish studies indicated the substantial variability in cardiovascular risk factors prevalence between autonomous communities (Medrano et al. 2005;Gabriel et al. 2008). Recently, the Mortality Atlas of Castilla-la Mancha that has just been published showed that CVD were the leading cause of death in the region during 2003(Gomez and Palmí 2020. So, an analysis was conducted in our study to assess the overall and sex/age-specific associations between air pollutants (PM 2.5 , PM 10 , and NO 2 ) and hospital emergency admission for CVD in this Spanish autonomous community during 2006-2015. To the best of our knowledge, this is the first study to examine the associations between exposures to air pollutants and emergency hospital admissions in CLM due to CVD. Unlike most previous studies, where the effects of CVD in large cities with high pollution were evaluated, this work covers cities of the region of CLM with moderate air quality. Therefore, this study aims to detect the acute effects of exposure to PM 2.5 , PM 10 , and NO 2 on cardiovascular hospital admissions within a 10-year period in four cities of CLM.

Study area
The autonomous community of CLM is located in the middle/southeast of Spain ( Fig. 1) and it is bordered by Madrid, the capital of Spain, to the north and the west. CLM is delimited by the Central Mountain Chain in the north, by the Iberian Mountain Chain in the northeast, and by the Mountains of Toledo. The region is made up of five provinces: Albacete, Ciudad Real, Cuenca, Guadalajara, and Toledo ( Fig. 1). CLM covers an area of 79,463 km 2 , which in 2015 was home to 2,048,900 inhabitants (Eurostat 2018), and it is the most sparsely populated autonomous community of Spain (representing approximately 4.4% of the country's population). CLM has a continental Mediterranean climate, so the most important factors of this aspect are extreme temperatures, with very hot summers and very cold winters. The summer season is dry, and the temperature often exceeds 30.0 °C. In contrast, in winter the temperatures frequently fall below 0 °C (reaching even -20.0 °C) with sporadic snowfalls (Nájera et al. 2009). Rainfall is not copious, and it shows a pronounced gradient from the central region (around 400 mm per year) to mountainous zones (more than 1000 mm per year).

Data
Data from the occurred CVD events over a period of 10 years (2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015) were collected from hospital admission records by the called Minimum Basic Data Set (MBDS). MBDS, supported by the Ministry of Health of Spain, is the largest administrative dataset on hospitalized patients, as well as the main source of information on morbidity in patients receiving care. This system used the International Classification of Diseases (ICD-9-CM) until December 2015 and the ICD10ES (a Spanish translation of the ICD10CM) thereafter. So, for this study, the selection was made with those patients who attended the emergency department and were admitted for cardiovascular disease according to the ICD-9-MC: heart diseases (410 acute myocardial infarction; 413 angina pectoris; 427 cardiac dysrhythmias; 428 heart failure) and cerebrovascular diseases or stroke (430-432 hemorrhagic stroke; 433-436 ischemic stroke).
The air pollution monitoring network of CLM registers levels of the air pollutants and meteorological variables (temperature and relative humidity (RH)) (JCCM 2020a). The air pollution recording stations used for this study are sited in the cities Toledo, Albacete, Guadalajara, and Ciudad Real (Fig. 1). The suburban background stations of Toledo, Albacete, and Ciudad Real and the urban background stations of Guadalajara are located in residential and commercial areas, so, consequently, the main source of emission that affects these stations are traffic and combustion in residential and commercial sectors. PM 2.5 is only measured in Toledo and Albacete air quality monitoring stations. PM 10 and NO 2 are measured in all stations (JCCM 2020b). The instruments for real-time monitoring of pollutants are chemiluminescence (NO 2 ) and beta ray attenuation monitor (PM 10 and PM 2.5 ) (JCCM 2020b). The daily average and 95 th percentile values were calculated from the hourly air pollutant and meteorological data through the Openair package 2.6-1 (Carslaw and Ropkins 2012), only in those days in which at least 75% of the hourly values were available; otherwise, it was identified as a missing value and it was not a part of the dataset for analysis.

Statistical analysis
In this study, the hypothesized association between the 95 th percentile of daily air pollutant concentrations as the predictor variable, and the number of cardiovascular hospital admission in CLM were analyzed using quasi-Poisson regression in single-pollutant generalized linear model (GLM) with Distributed Lag Model (DLM) at different lags (from 0 to 7 days) (Bhaskaran et al. 2013). Hospital admission data are Poisson distributed and Poisson regression provides an estimation of the relative risk (RR) as RR = exp(β), where β is the regression coefficient associated with a unit increment in a pollutant. The advantage of the Poisson regression model is the ability to estimate the RR of the health effect with the increase of exposure by unit. Air pollution and hospital admission relationships were conducted for sex (men and women) and age groups (adults 15-64 years, and elderly ≥ 65 years). Moreover, in order to address the possible exposure misclassification while using pollution data from one central monitoring station, we did a sensitivity analysis by restricting the CVD hospital admission to the nearest hospital to the air quality monitoring station in the case of Toledo City.
A penalized smoothing spline was used for filtering out seasonal patterns and long-term trends in daily hospitalization, as well as daily mean temperature, daily mean RH, day of week, and public holidays which were included in the model as dummy variables. To reduce the problems associated with multiple testing and model selection strategies, we followed some previous time series studies to select a priori model specification and the degree of freedom (df) for time trend and the meteorological variables (Qiu et al. 2013). A df of 7 per year for time trend, a df of 6 for mean temperature of the current day, and a df of 3 for the current day humidity were used.
The differences in the number of daily admissions between groups, defined according to if the daily 95 th percentile pollutant concentration of the pollutant exceeding or not the limit value defined by WHO (WHO 2016), were performed by the nonparametric Mann-Whitney U test. WHO guideline values (WHO 2016) are more stringent than European Union (EU) air quality Directive (2008/50/EU). Furthermore, the existence of the differences between the number of cardiovascular hospital admissions and pollutant concentration decile was carried out by means of an analysis of variance (ANOVA, Tukey post hoc test).
The analysis was carried out in the R environment, version 4.0.2, for statistical computing and visualization (Gasparrini 2011). Functions in splines and dlmn package of R 1 3 (Gasparrini 2011) were used to build GLMs as described (Bhaskaran et al. 2013). Graphics of smooth trends were estimated with Openair 2.6-1 (Carslaw and Ropkins 2012).

Results
A total of 109,974 cardiovascular hospital admissions derived from the emergency department were recorded in CLM, from January 1, 2006, to December 31, 2015 (Table 1). Table 1 also shows the average values, standard deviation, and range of the daily admissions values and their distribution according to sub-diagnoses (cardiac disease or stroke), age, and sex.
On the other hand, the 10 years daily mean pollution concentrations for PM 2.5 , PM 10 , and NO 2 and meteorological conditions of temperature and RH in Toledo, Albacete, Guadalajara, and Ciudad Real were 10.7 and 13.8 μg m −3 for PM 2.5 (only are measured in Toledo and Albacete); 29.1, 35.9, 25.1, and 27.1 μg m −3 for PM 10 ; 22.8, 14.6, 25.2, and 10.6 μg m −3 for NO 2 (Table S1). The average pollutant concentrations were not above the WHO standard air quality threshold (WHO 2016). The daily ambient temperatures were 15.9, 16.2, 15.6, and 16.7 °C and the RH was 57.8, 60.6, 57.1, and 57.2%, respectively (Table S1). Figure S1 shows the evolution of NO 2 measured by the different stations during the study period. As can be observed in Figure S1 for NO 2 , the Guadalajara and Toledo stations provided similar information and with slightly higher values than those reported by the Ciudad Real and Albacete stations. Regarding the trend of PM 10 , Figure S2 shows the evolution during the study period for each station. Although similar trends were observed in the evolution of this pollutant, there were differences between the different stations. Since 2010, differences between stations below 10 µg m −3 were usually observed. However, as can be seen in Figure S3, a large difference in the average PM 2.5 values was observed between the measurements collected in the Albacete and Toledo stations during the 2013-2015 period. This abnormal behavior could not be explained, and it was also described by a MITECO air quality report (MITECO 2013).
To test the statistical relationship between daily air pollutant concentrations, temperature, and RH in the different cities, the non-parametric bootstrap test of the Pearson correlation was employed (Table 2). PM 10 , PM 2.5 , and NO 2 were positively correlated with each other. The highest and closest correlation between PM 10 and PM 2.5 (0.625-0.547, strong-moderate) in Toledo and Albacete, respectively, could confirm that these pollutants have a common source. However, weak and very weak correlations were found between PM 2.5 and NO 2 and PM 10 and NO 2 in the four cities. Correlations between the other variables were negligible.
Exposure to pollutants, even for a few hours, can trigger cardiovascular disease and related deaths from heart attack, stroke, arrhythmia, sudden cardiac arrest, and heart failure (Nel 2005;Mustafic et al. 2012;Gardner et al. 2014). For this reason, we took into account the 95 th percentile daily pollutant concentrations to assess the relationship between CVD and each air pollutant. Figures 2, 3, and 4 provide the relationship between daily admissions at 10 quantile groups of the 95 th percentile value of each air pollutant at lag 0. As can be seen in Fig. 2, there was a clear trend to increase the number of hospital admissions when increasing the concentration of PM 2.5 in Toledo, both when admissions are only in the Toledo hospital and for the whole of CLM. The pairwise comparisons show significant differences between the last and the rest of deciles (p < 0.001).
Concerning PM 10 (Fig. 3), again in Toledo, there was an increasing trend in the number of admissions when the 1 3 concentration of this pollutant increased. Regarding Ciudad Real and Albacete, there was not a clear trend. Only in Guadalajara was there a significant difference between the 6 th and 10 th deciles and the first five deciles, so at a higher concentration of PM 10 (46.5-48.9 µg m −3 ), there was a significant increase in hospital admissions.
In the case of NO 2 (Fig. 4), the trend was similar in all the studied cities. Furthermore, every NO 2 concentration value was assigned a decile in order to find the threshold at which cardiovascular admissions increased significantly in CLM. From the analysis of Fig. 4, significant differences were observed for values close to 60 µg m −3 , in Toledo and Guadalajara (Toledo: (54.7, 61.4); Guadalajara: (58.7, 63.4)) and 25 µg m −3 in Ciudad Real and Albacete (Ciudad Real: (26.3,29.3); Albacete: (25.6, 28.8)).
We carried out a division of the 95 th percentile daily concentrations according to the WHO guide value (WHO 2016). Between these two groups, the number of daily hospital admissions was studied. Thus, in these two established groups, the distribution of the number of admissions was studied (see box plot, Figures S4, S5, S6, and S7). The existence of significant differences between the two groups was analyzed using the Wilcoxon non-parametric test.
The number of hospital admissions was higher the days in which the 95 th percentile of the pollutants exceeded the WHO guideline values. The Wilcoxon test confirmed the existence of statistically significant differences between daily hospital admission and the mean values of pollutants, except in Albacete for PM 2.5 and PM 10 . The value of the test criterion W, in the one-sample Wilcoxon signed-rank test, and the p-values are summarized in Table S2.
The link between CLM hospital admission and air pollutants over the total population (women, men, and two age groups) explained with the models at lag 0 and lags of up to 7 days (lag 7) is shown in Figure S8 and Table 3. The RR and 95% confidence intervals (CI) associated with an increase of 10 μg m −3 in pollutant concentration were obtained for each pollutant after adjusting seasonality and long-term trends, day of the week, public holiday, and weather variables. Table 3 also shows both, the admissions exclusively from the Toledo reference hospital, to address possible exposure misclassification when using pollution data from the closest monitoring station, and the whole of CLM. As can be observed, the results obtained were both similar.
In general (except for the measurement of PM 2.5 at the Albacete station), in all cities and for the whole pollutants, the effect on CVD admissions was more immediate, strongest, on the current day, and with close similar RR behavior. For PM 10 and NO 2 , the effect is stronger again at lags 2 and 3 and lag 6 or 7 (Table 3). At lag Table 2 Pearson correlation coefficients between atmospheric variables (i.e., air temperature and RH) and the air pollutants (PM 0, for every 10 µg m −3 increase in PM 2.5 , the RR of the hospital admission was 1.028 (CI: 1.009-1.048) in Toledo station, and 1.013 (CI: 0.997-1.028) in Albacete station, then the RR decreases in both cities until lag 4, with RR values of 1.013 (CI: 0.993-1.34) and 1.008 (CI: 0.991-1.025), respectively. For PM 10 , an increase between 1.013 and 1.026 was found for each 10 μg m −3 of the pollutant increases. NO 2 was associated with the highest effect at lag 0 in the four stations (especially in Albacete and Ciudad Real stations) with a 1.042 and 1.035 per each 10 μg m −3 , respectively. In addition, at lags 2-3, the RR decreased and reached a close value (between 0.990 and 0.995) for the studied cities. This significant effect observed with NO 2 , not always with PM 10 , and never with PM 2.5 disappeared when only the admission at the nearest hospitals was analyzed (Table 3).
Moreover, according to our results, the RR of PM 2.5 , PM 10 , and NO 2 increase with advancing age (aged ≥ 65 age), in both sexes (Tables 4, 5, and 6). With regard to the association between PM 2.5 and hospital admissions for CVD (Table 4) at lag 0, the RR showed a higher value to the elderly (≥ 65). For PM 10 (Table 5), the differences of the RR between age and sex were less pronounced than PM 2.5 . Moreover, when the RR were significant, it reached lower values than in the case of PM 2.5 . With regard to NO 2 (Table 6), the significant RR values were also lower than those reached by PM 2.5 and with hardly any differences by age and sex.

Discussion
CVD have been among the diseases with the highest mortality during 2003-2014 in the region, which recently has been published in the Mortality Atlas of Castilla-La Mancha (Gomez and Palmí 2020). This 10-year study has demonstrated the significant association between daily hospital admissions for CVD and daily concentrations of PM 2.5 , PM 10 , and NO 2 . CLM may be assumed to have a moderate air quality in comparison with other parts of Spain such as Madrid (Borge et al. 2018) (Table S1) with high levels of air pollution (PM 2.5 up to 23 µg m −3 in Albacete, and 19 µg m −3 in Toledo). Despite CLM's seemingly relatively low pollution levels measured, ambient concentrations of PM 2.5 , PM 10 , and NO 2 have been correlated with negative health effects in our study. So, it should be advisable to adapt the current EU air quality standards to WHO guidelines which is stricter.

or Barcelona
The effect of air pollution on hospital admission in CLM was statistically significant at some, not all, lag structures (Tables 3, 4, 5, and 6). Studied pollutants were significantly associated with increased risk of cardiovascular hospital admissions, especially at current day (lag 0). However, the risk increased again after lag 4 in men and lag 6 in women. Regarding age groups, adult men again showed effect at lags 3 and 5; however, women did not show effect until lag 7. Moreover, adult men seemed to have a higher risk of being admitted to the hospital for CVD than adult women since they had generally higher RR values.
In the case of the measurement of NO 2 , there was higher uniformity in the trends shown by the measurement of the different stations and the number of hospital admissions in the region. However, there was a discrepancy between the PM values in Albacete and the rest of CLM. There are two possible explanations. First, it could be due to the technique used in the determination of PM (beta attenuation). Second, the behavior between PM and chemical pollutant (NO 2 ) is different (e.g., dispersion and reactivity); it is more homogeneous in the case of the chemical pollutant. Moreover, it should be noted that, as observed in Albacete and Ciudad Real, it seems that even at values lower than those established as a guide by the WHO (WHO 2016), a significant increase in the number of hospital admissions was observed.
There is a substantial body of epidemiologic literature showing a clear and consistent association between concentrations of ambient PM 2.5 , PM 10 , and NO 2 and CVD (Gold et al. 2000;Ballester et al. 2006;Miller et al. 2007; Pope  Cesaroni et al. 2014;Santurtun et al. 2017;Liu et al. 2015;Vahedian et al. 2017;Collart et al. 2018;Sanyal et al. 2018;Liu et al. 2019;Phosri et al. 2019;Amsalu et al. 2019;Franco et al. 2020). The association between these pollutants and CVD was reported worldwide for high levels of pollutants or even in a setting where pollutant levels were within ambient air quality standards European Directive 2008/50/EC or WHO guideline (WHO 2016), such as in our work. However, there is evidence that PM 2.5 and NO 2 have a steeper exposure-response curve at low concentrations, indicating that increases in low concentrations intensify negative health effects more than equivalent increases at higher concentrations (Burnett et al. 2018;Hanigan et al. 2019).
If we compare our results with some of those studies, our findings are consistent with the existing literature. Studies in Beijing, China, where the mean daily PM 2.5 concentration was 76.9 μg m −3 , found that for every 10 μg m −3 increase in the PM 2.5 concentration from the previous day to the current (lags 0-1), there was a significant increase in total CVD admissions (0.30, 95% CI: 0.20-0.39%), with a strong association for older adults (aged ≥ 65 years) (Amsalu et al. 2019). Moreover, Phosri et al. (2019) reported in their study in Bangkok (Thailand) that exposure to PM 10 and NO 2 poses a significant risk of cardiovascular admission with estimates of 1.28% (95%, CI: 0.87-1.69) and 1.04% (95%, CI: 0.68-1.41), for each 10 μg m −3 increase in the pollutant, respectively. In this study, the daily mean concentration values reached were 41.6 µg m −3 for PM 10 and 41.8 µg m −3 for NO 2 .
Within the ESCAPE Project (Cesaroni et al. 2014), the study of long-term exposure to PM 2.5 and incidence of acute coronary events in 11 European cities reveals a positive association below the current annual European limit value of 25 μg m −3 for PM 2.5 (1.18, 1.01-1.39, for 5 μg m −3 increase in PM 2.5 ) and below 40 μg m −3 for PM 10 (1.12, 1.00-1.27, for 10 μg m −3 increase in PM 10 ). In the Spanish multicity study within the EMECAS Project (Ballester et al. 2006), the evaluation of the short-term effect of NO 2 and PM 10 on cardiovascular admissions, with mean concentrations from 23.4 µg m −3 (Castellón) to 76.2 μg m −3 (Valencia) for NO 2 and from 32.8 µg m −3 (Zaragoza) to 43.2 µg m −3 (Granada) for PM 10 , showed a  consistent association at lags 0 and 1 with CVD hospital admission. Moreover, studies in Boston (USA) with mean 4-h PM 2.5 levels ranging from 3 to 49 μg m −3 have demonstrated a significant increase in CVD, from 0.5 to 1.5%, for every 5 ~ 6 μg m −3 increase in PM 2.5 (Gold et al. 2000). NO 2 has been also found to be associated with hospital admissions for CVD in Belgium (Collart et al. 2018). In this work, the average annual concentration of NO 2 was stable for the 4 years of analysis (around 20.5 μg m −3 ). For all ages combined, the strongest association between this pollutant and the number of hospital admissions was observed at lag 0, except for hemorrhagic stroke where the strongest association occurred at lag 2. A recent systematic evaluation of time-series studies of air pollution from 652 cities in 24 countries or regions found, on average, an increase over the current and previous day of 0.36% (95% CI: 0.30-0.43) and 0.55% (95% CI: 0.45-0.66) in daily cardiovascular mortality per 10 μg m −3 increase in PM 10 and PM 2.5 concentration, respectively . In this same study, for the period 1986-2011, China had the highest level of annual-mean PM 10 pollution (89.2 μg m −3 ) and one of its cities had the highest level of PM 2.5 (116.9 μg m −3 ). This study found in Spain, for the studied period 2001-2014, a daily annual average of 27.8 μg m −3 for PM 10 and 11.3 μg m −3 for PM 2.5 , similar values to the concentrations measured in CLM in our study (from 2006 to 2015).
Our study could be a first step for implementing cardiovascular diseases alert protocols depending on air pollution levels in CLM and other Spanish communities. Inevitably, there are some limitations in our work. Firstly, we extracted hospital admission data for people residing and admitting to the hospitals in the study area. Therefore, people who were admitted to the hospitals could not reside in CLM. Secondly, socioeconomic status and with underlying diseases of the patients were not explored in this study, which could skew the results. Thirdly, we did not have data on people who went to the emergency department for CVD and who were not finally admitted in the hospital. Finally, chemical composition of PM has not been considered and harmful health effects can vary by its chemical constituents. Therefore, an additional study taking into account the above limitations should be carried out to help strengthen our findings on the effects of air pollution on cardiovascular hospital admissions.

Conclusions
This study reported the effects of air pollutants (PM 2.5 , PM 10 , and NO 2 ) on the daily hospital admission for CVD in Toledo, Albacete, Guadalajara, and Ciudad Real (CLM, Spain) for a period of 2006-2015. In general, the concentrations of the pollutants were lower than the ambient air quality standards European Directive 2008/50/EC or WHO guideline. The annual EU limit values for NO 2 and PM 2.5 there were not exceedances, only the PM 10 limit value exceeded punctually due to Saharan dust intrusions. However, negative effects of air pollution on human health have been found in the whole pollutants studied here. Moreover, we also found that elderly (≥ 65 years) held the greater effects of air pollution on hospital admissions compared to the adult group (aged 15-64), which is relatively important for planning and implementing intervention for this population group since the CLM population is going to experience aging in the near future. Moreover, it is known the emergency department is one of the most congested units in any hospital that faces greater pressure in terms of patient load and healthcare resources as compared to other departments of the healthcare system. So, the potential utility of these results may help the emergency department during similar air pollutant concentration, where an increase in the use of this health service among the vulnerable groups could be expected. The use of the app Avis@ * Only in Toledo: Exclusively admissions from Toledo Hospital, the closest location to the measuring station (JCCM 2020c), which informs citizens interested in the quality of the aur they breathe, could also be a powerful tool in the emergency department to alert of real-time information on air quality in CLM.

Supplementary Information
The online version contains supplementary material available at https:// doi. org/ 10. 1007/ s11869-021-01144-1. review of the effects of particulate matter air pollution on human Table 6 RRs and 95% CI of daily hospital admission stratified by age and sex group associated with 10 µg m −3 increase of NO 2 concentrations Sub-group