As a starting point, this study utilized the findings for a group of children between the ages of 0 and 10 years who may be directly affected by air pollution with airborne particulate matter that primarily originated from metallurgical processes, urban traffic, and domestic heating. The analysis considered the medical aspects of 111 children diagnosed with respiratory diseases associated with wheezing. Children’s respiratory symptoms that were potentially determined or aggravated by atmospheric pollution included upper respiratory airways symptoms (cough, rhinorrhea, and sore throat) and lower airways symptoms (dyspnea, wheezing, and thoracic pain).
The location where the symptoms were likely to occur was spatially described using the children’s addresses to assess the potential correlations with PM2.5 levels. The locations of the affected areas (kindergartens, playgrounds, and schools) where children might be exposed to PM2.5 levels were also added to the digital map to facilitate the establishment of critical areas. Figure 1 shows the spatial positioning of 72 children selected from a total of 111 children, who presented high levels of IgE (normal value <60 units/ml) and a high eosinophil count (normal value = 0.1–3 % from leukocytes), as well as wheezing episodes.
The spatial analysis revealed three groups of children: (A) northwest group—61 children, (B) center-northeast group—20 children, and (C) southeast group—30 children. The specific layer was included in the environmental mapping system after the selection of children who manifested an allergenic response: 38 children (A), 7 (B), and 27 (C), which were usually more sensitive to PM2.5 pollution (Figs. 1 and 2). The main pollution point sources and roads with heavy traffic were overlapped to provide a comprehensive image of the emissions’ impact on the sensitive receptors.
Model simulations were performed by grouping the stationary sources by cardinal directions, i.e., north, east, south, and west. The most important contribution of PM2.5 was estimated to originate from the industrial metallurgical facility located south of city. Figure 2 shows the average concentration highlighting that the location where group C is located was the most heavily impacted by PM2.5 air pollution (10–20 μg m−3 PM2.5). A monitoring plan was developed based on the receptor modeling results. Figure 3 presents the location of the PM2.5 sampling points that were established using a top-down approach.
Our results of an association between PM2.5 levels, meteorological factors, and an increased number of hospital admissions corroborate previous findings (e.g., Basagana et al. 2015; Neuberger et al. 2004).
Geolocated study on selected groups of children
A geolocated study was developed by positioning each of the 72 children with high levels of IgE and eosinophils based on their residential address to link the airborne PM2.5 levels with the potential adverse health effects in children (Fig. 1). The majority of children had also experienced wheezing episodes. The group of 111 selected children who were vulnerable to air pollution consisted of 60 males (54 %) and 51 females (46 %) who were born between 2004 and 2012. The ranking of the age categories were as follows: 5 years (18 cases), 7 years (17), 2 years (14), 6 years (13), 4 years (13), 3 years (11), 8 years (8), 9 years (7), 1 year (7), and 10 years (3). The categories between 2 and 7 years accounted for 77.5 % of the group, which suggests that children in this age range are the most vulnerable to the occurrence of wheezing-related diseases (i.e., asthma attacks, bronchitis, and recurrent wheezing). More than 80 % of children with asthma exhibited distinct symptoms before the age of 5 years (Hay et al. 2014). The main symptoms experienced by the selected group were chest pain, prolonged cough, intolerance to physical effort, breathing difficulties with varying frequency and intensity, recurrent bronchitis, and pneumonia. Table 3 shows the descriptive statistics associated with the group of 111 children. Of particular interest were the numbers of children who exceeded the normal thresholds of IgE (56 children) and eosinophils (41 children). Many children presented with IgE values at the upper limit of the normal interval (e.g., mode was 60 units/ml). The statistical results during the study period of 2 years indicated median values of 7 wheezing episodes, 2 hospitalizations/child, 153 U/ml IgE, and 3.8 % for eosinophils, whereas the maximum values were 50 episodes, 10 hospitalizations/child, 2500 U/ml IgE, and 26 % eosinophils.
After performing the geolocation of children, a comparison of the geometrical means showed that the highest values of IgE (187.5 U/ml) and eosinophils (4.2 %) were recorded for the group A, followed by the group C (126.5 U/ml IgE and 1.6 % eosinophils) and the group B (66.6 U/ml IgE and 2.9 % eosinophils). No significant differences (p < 0.05) were observed when performing multiple range tests (LSD) between groups for wheezing episodes, IgE level, or eosinophil count. Higher values were recorded for all considered variables in the group A, which was located in northwestern Targoviste City. Another objective of the study was to determine whether the concentrations of PM2.5 are correlated with the spatial distributions of respiratory diseases in children.
PM2.5 levels and spatial correlations with wheezing occurrence
The PM2.5 multiannual average of measured concentrations ranged from 4.6 to 22.5 μg m−3 with a coefficient of variation (CV) of 57.3 %, and the maximum concentrations ranged from 13.1 to 102 μg m−3 (CV = 81.3 %), depending on the sampling point (Table 4). The average of the maximum absolute values was 187.1 μg m−3 (CV = 175.29 %). The thematic maps with isolines of concentrations showed high levels of PM2.5 in the western and northwestern parts of the city, which were correlated with intense heavy traffic and neighboring active industries from the northwest (Figs. 4 and 5). The impact of emissions generated by the southern metallurgical facility was less evident due to lower emissions because of the economic recession since 2009, which have affected the industrial production. In situ measurements showed a different pattern compared with the dispersion modeling results (Table 5), which revealed higher annual average concentrations (10–22.5 μg m−3) in the group A area; the concentrations of the group C area ranged from 6 to 14 μg m−3 (Figs. 4 and 5). The PM2.5 map showed that the group B area recorded the lowest concentrations (≤6 μg m−3). Highly significant correlations (p < 0.01) were observed between the locations of the children with high number of wheezing episodes and hospitalizations and the PM2.5 multiannual average (r = 0.985), PM2.5 maximum values (r = 0.813), and PM2.5 momentary peak values (r = 0.802). This spatial correlation supports the hypothesis that the respiratory health impact of increased PM2.5 concentrations was more pronounced in areas with presumptively higher exposure (e.g., playgrounds, schoolyards, etc.), having a direct influence on wheezing-related symptoms and asthma attacks in the analyzed group of children.
Heavy metal concentration of PM2.5
The laboratory analyses indicate that PM2.5 contained the following concentrations of heavy metals (multiannual averages) in descending order: Fe (3.1–5.8 ng m−3, CV = 20.9 %), Pb (0.8–2.8 ng m−3, CV = 44.7 %), Ni (0.5–1.16 ng m−3, CV = 23.5 %), Cd (0.01–0.25 ng m−3, CV = 49.9 %), and Cr (0.01–0.09 ng m−3, CV = 68.59 %). PM2.5 composition was similar to the rankings observed in other studies performed near steel-related sites (Dai et al. 2015; Querol et al. 2007; Taiwo et al. 2014) as shown in Table 4. The results are consistent with the concentrations of heavy metals recorded in certain urban US areas for corresponding mass concentrations (Chow and Watson 1998).
Our findings show that nickel (Ni) concentrations were more consistent compared with other metals. No statistical significances between children’s locations and any determined metal concentrations (r = −0.007–0.21) were observed. This result suggests that other air pollutants of concern in the ambient air and/or compounds (organic compounds and salts) in PM2.5 may have an immediate adverse effect on children’s respiratory health.
Lead (Pb) is present in paved road dust due to deposition from previous emissions of leaded-gasoline vehicle exhaust (Lu et al. 2014; Wei et al. 2015). The highest concentrations were recorded in areas with intense traffic (i.e., central market and city exits of the main roads; see points TGV 7, 3, 2 and 1). Cadmium (Cd) primarily originates from steel production, pigment facilities, and tire wear (Tian et al. 2012). The highest Cd concentrations were observed in the western and northwestern sections of the city (i.e., points TGV 1, 2, 3, and 6). Chromium (Cr) occurs in soluble forms from fossil fuel combustion and vehicle emissions (Abuduwailil et al. 2015). The highest values were recorded in the city center and toward the west and northwest (points TGV 9, 6, 3, and 1). Nickel (Ni) was emitted in the area from steel production and coal/oil combustion. The highest values were recorded at points TGV 3, 1, 9, 8, and 2 due to dust resuspension and residential heating. Iron (Fe) was the most abundant element in PM2.5; it primarily originated from steel dust accumulation in the area. High concentrations of this element also occurred in suburban areas. The highest multiannual averages of Fe concentrations were observed at the TGV 1, 9, 2, and 8 sampling points. The bivariate relationships between metal concentrations were only significant for the following pairs: Cd–Ni (r = 0.69; p < 0.05), Cr–Ni (r = 0.76; p < 0.05), and Cr–Fe (r = 0.68; p < 0.05). Consequently, Fe, Ni, Cd, and Cr may be regarded as main marker elements of emissions from specialized steel production and metalworking in the Targoviste area.
Table 5 shows the grouping of children with respiratory issues in three spatial groups and the corresponding PM2.5 averages and associated heavy metal content in Targoviste City. The differences among the concentrations of heavy metals of the groups were small, which suggests that the impact of metallurgical activities affects larger areas. However, significant differences among the groups were observed for the PM2.5 averages that were recorded at the sampling points.
Factorial analysis applied to the medical and air pollutant datasets
The relevant factor loadings (>0.55) were considered for each factor (Table 6). The rotated matrix showed that the eosinophil count, age of the child, and PM2.5 air pollution form the first factor (PC1), wheezing episodes and hospitalizations form the second factor (PC2), and IgE and gender form the third factor (PC3) (Fig. 6). These factors accounted for a cumulative variance of 64.5 % of the total variability in the dataset. The interpretation of the first factor loadings suggests that PM2.5 concentrations affect mainly small children having a major influence on eosinophils increasing, which are actively involved in inflammatory processes and allergy-like patterns of response. The health effect components, i.e., wheezing episodes and hospital admissions, showed high factor loadings. Consequently, factor analysis allowed the comparison of health effect estimates based on single pollutant metrics suggesting that air pollution with fine particulates may be a potential trigger of the asthma attacks leading to increases in hospital admissions, and the fact that the male subjects are potentially more susceptible to PM2.5 air pollution.
To support the FA results, an association between PM2.5 high concentrations and several physiological changes and clinical symptoms in children was observed in our study: alteration of lung function, symptoms of the upper and lower respiratory tracts, bronchial asthma, and rhinitis. Toxicological and clinical studies regarding the effects of combustion-derived particles showed that peak exposures of short duration (ranging from less than an hour to a few hours) lead to immediate physiological changes (WHO 2013). Hence, short-term exposure to peak and maximum levels of PM2.5 (Table 4) during the outdoor program of children in Targoviste has impacted the triggering of asthma exacerbations, especially in infants and preschoolers, which has increased the number of wheezing episodes and maintained elevated levels of allergic indicators (eosinophils and IgE) despite the use of controller medication for asthma.