Next Article in Journal
Re-Thinking the Role of Government Information Intervention in the COVID-19 Pandemic: An Agent-Based Modeling Analysis
Previous Article in Journal
Iron Oxide Particles Alter Bacterial Uptake and the LPS-Induced Inflammatory Response in Macrophages
Previous Article in Special Issue
Feasibility of Indonesia Family Life Survey Wave 5 (IFLS5) Data for Air Pollution Exposure–Response Study in Indonesia
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
IJERPH
Volume 18
Issue 1
10.3390/ijerph18010144
ijerph-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Short-Term Exposure to Fine Particulate Matter and Hospitalizations for Acute Lower Respiratory Infection in Korean Children: A Time-Series Study in Seven Metropolitan Cities

by
Jongmin Oh
1,†,
Changwoo Han
2,†,
Dong-Wook Lee
3,
Yoonyoung Jang
3,
Yoon-Jung Choi
3,4,
Hyun Joo Bae
5,
Soontae Kim
6,
Eunhee Ha
1,
Yun-Chul Hong
3,4,* and
Youn-Hee Lim
7
1
Department of Occupational and Environmental Medicine, School of Medicine, Ewha Womans University, Seoul 120-750, Korea
2
Department of Preventive Medicine, Chungnam National University College of Medicine, Daejeon 301-747, Korea
3
Department of Preventive Medicine, Seoul National University College of Medicine, Seoul 110-799, Korea
4
Environmental Health Center, Seoul National University College of Medicine, Seoul 03080, Korea
5
Korea Environment Institute, Sejong 30147, Korea
6
Department of Environmental and Safety Engineering, Ajou University, Suwon 16499, Korea
7
Department of Public Health, University of Copenhagen, 1599 Copenhagen, Denmark
*
Author to whom correspondence should be addressed.
†
Contribute equally as first author.
Int. J. Environ. Res. Public Health 2021, 18(1), 144; https://doi.org/10.3390/ijerph18010144
Submission received: 30 October 2020 / Revised: 9 December 2020 / Accepted: 23 December 2020 / Published: 28 December 2020
(This article belongs to the Special Issue Air Pollution, Climate Change and Respiratory Health)
Browse Figures
Review Reports Versions Notes

Abstract

:
Although several studies have evaluated the association between fine particulate matter (PM2.5) and acute lower respiratory infection (ALRI) in children, their results were inconsistent Therefore, we aimed to evaluate the association between short-term exposure to PM2.5 and ALRI hospitalizations in children (0–5 years) living in seven metropolitan cities of Korea. The ALRI hospitalization data of children living in seven metropolitan cities of Korea from 2008 to 2016 was acquired from a customized database constructed based on National Health Insurance data. The time-series data in a generalized additive model were used to evaluate the relationship between ALRI hospitalization and 7-day moving average PM2.5 exposure after adjusting for apparent temperature, day of the week, and time trends. We performed a meta-analysis using a two-stage design method. The estimates for each city were pooled to generate an average estimate of the associations. The average PM2.5 concentration in 7 metropolitan cities was 29.0 μg/m3 and a total of 713,588 ALRI hospitalizations were observed during the 9-year study period. A strong linear association was observed between PM2.5 and ALRI hospitalization. A 10 μg/m3 increase in the 7-day moving average of PM2.5 was associated with a 1.20% (95% CI: 0.71, 1.71) increase in ALRI hospitalization. While we found similar estimates in a stratified analysis by sex, we observed stronger estimates of the association in the warm season (1.71%, 95% CI: 0.94, 2.48) compared to the cold season (0.31%, 95% CI: −0.51, 1.13). In the two-pollutant models, the PM2.5 effect adjusted by SO2 was attenuated more than in the single pollutant model. Our results suggest a positive association between PM2.5 exposure and ALRI hospitalizations in Korean children, particularly in the warm season. The children need to refrain from going out on days when PM2.5 is high.

1. Introduction

Many studies have reported that fine particulate matter (PM2.5) is associated with respiratory morbidity and the mortality of children [1,2,3]. Children’s respiratory systems can be more susceptible to PM2.5 due to their fast breathing rate, narrow bronchi, and immature immune system [4,5]. Exposure to PM2.5 in children has been linked to respiratory diseases such as asthma, allergic inflammation, and acute lower respiratory infection (ALRI) [6,7,8,9].
ALRI is an ordinary respiratory disease in young children, and its clinical symptoms are characterized by fever, shortness of breath, sore throat, and cough. Acute bronchitis and bronchiolitis, influenza, and pneumonia are included in ALRI [10]. Globally, ALRI is the fifth major cause of child death, accounting for 30% to 50% of child hospitalizations [7,11].
The Global Burden of Disease group suggested that air pollution, including PM2.5, is a risk factor for ALRI in children [7,11,12]. Respiratory infections in children or newborns can be fatal, and they also put a considerable burden on parents. Some previous studies reported evidence supporting the biological plausibility of an association between PM2.5 and ALRI. PM2.5 is comprised of diverse heavy metals and organic components and may persuade oxidative stress and inflammation in the lung tissues [13,14]. Toxicological research studies have shown that PM2.5 exposure may lead to impaired immune responses by altering the phagocytic activities of alveolar macrophages and reducing antioxidant levels [7,15].
Previous studies showed inconsistent results regarding PM2.5 exposure and ALRI in children. Although several epidemiology studies in China, the US and Australia showed positive associations [1,10,16], some studies have found no association between PM2.5 and ALRI in children [17,18].
Korea has relatively high PM2.5 levels (annual mean concentration of 25 µg/m3 in 2016) among the countries of the Organization for Economic Co-operation and Development [19] and a high prevalence of pediatric ALRI hospitalization (20–30% prevalence) [20]. However, a recent Korean study showed no relationship between PM2.5 and nationwide childhood respiratory hospitalization in 2013–2015 [21]. Given the high concentration levels of PM2.5 and the prevalence of ALRI in metropolitan cities, additional studies of any association, particularly in urban areas, are urgently needed. In some previous studies, only certain cities were considered to be within the scope of the research area [16,18,22]. In other cases, the study period was relatively short [21]. The inconsistency of prior research may be due to differences in exposure by region, which may result in regional differences, and the short duration of the study may also be a factor. In this study, we investigated the relationship between PM2.5 and children’s ALRI hospitalization using the Korea’s National Health Insurance data over a longer period (2008–2016). The aim of our study was to evaluate the impact of short-term exposure to PM2.5 on ALRI in children. We also investigated sex- and season-specific estimates.

2. Methods

2.1. Number of Daily Children’s ALRI Hospitalization

In this study, we utilized the children’s ALRI hospitalization data archived from the National Health Insurance Database (NHID) in South Korea. The NHID contains all of the medical information of the insured persons, including hospital visits (date of admission, date of discharge, and diagnostic codes) and personal information (sex, birth year, and residential address). Detailed information on NHID and its customizing processes for researchers have been described elsewhere [23]. Since this data is big data, it can represent the characteristics of the population. The National Health Information Service (NHIS) prepares queries and provides data at the request of researchers. We claimed information on ALRI hospitalization for children living in seven metropolitan cities (Seoul, Busan, Daegu, Incheon, Daejeon, Gwangju, and Ulsan) of South Korea from 2008 to 2016 (Figure 1). The ALRI hospitalization was determined by the International Classification of Disease, 10th revision diagnostic codes: either primary or secondary diagnostic codes of J20–J22 (J20: Acute bronchitis, J21: Acute bronchiolitis, J22: Unspecified acute lower respiratory infection). Daily counts of hospitalization for ALRI were calculated in each metropolitan city.
This study uses the unidentifiable data prepared by the NHIS and follows the guidelines for the protection of personal information provided by the NHIS. Therefore, our research was exempted from deliberation by the Institutional Review Board (E-1807-038-956) of Seoul National University Hospital.

2.2. PM2.5 Modeling Data

Since the nationwide monitoring system for PM2.5 levels has only been in operation since 2015, we used the Community Multiscale Air Quality (CMAQ, version 4.7.1) modeling system to estimate PM2.5 concentrations in areas throughout the study period (2008–2016). These modeling data are designed by the US Environment Protection Agency. Several previous studies have used this modelled data [24,25]. A detailed description of the CMAQ PM2.5 data is given in Supplementary Materials Test S1. Exposure values were estimated for each of 74 administrative districts in seven cities. We calculated the modeled data by administrative districts on a daily average by city.

2.3. Air Pollution Monitoring Data and Environmental Variables

We also collected data on other air pollutants to construct multiple pollutant models, considering that a single pollutant can be biased or misleading [26].
Information on PM10, sulfur dioxide (SO2), nitrogen dioxide (NO2), carbon monoxide (CO) and ozone (O3) was archived from the AirKorea database (https://www.airkorea.or.kr/web). This database includes hourly data collected from the stationary monitoring stations located in each metropolitan area (number of monitoring stations in 2018, Seoul: 25, Busan: 17–20, Daegu: 13, Incheon: 17, Gwangju: 7, Daejeon: 10, Ulsan: 16). Measurement methods for each air pollutant are well described in previous studies. [27]. Using the hourly data of air pollutants in multiple monitoring stations, we computed daily average values for each region.
We collected the daily environmental variables (mean temperature, relative humidity, dew point temperature etc.). These data were prepared from the Korea Meteorological Administration database. We can use the data for research purposes on this website (http://www.kma.go.kr), and we have obtained the data by each city. We wanted to simplify the model by using variables containing temperature, humidity and dew point information. So, we used daily mean apparent temperature (AT) [28,29]. The equation for AT is described in the Supplementary Materials.

2.4. Statistical Analysis

We analyzed the time-series data for daily ALRI hospitalization count. A two-stage design was considered to investigate the relationship between PM2.5 and ALRI hospitalization. Briefly, we evaluated the generalized additive model (GAM) analysis to estimate city-specific associations. Then, we used random-effects meta-analysis to provide an effect estimate for the whole city.
In the first stage, we constructed a GAM to estimate the association between PM2.5 exposure and ALRI hospitalizations for each city. The daily ALRI hospitalization is count data and assumes a Poisson distribution in the model. Exposure-response curves in the metropolitan cities showed linear associations (Figure 2). Hereafter, we made an assumption that the association was linear between PM2.5 exposure and ALRI hospitalization. The detailed linear model is as follows:
Log(μt) = Intercept + β × PM2.5 + s(Time, df = 7 × 9) +s(ATt, df = 6) + γ × Day of week
The μt is the daily ALRI hospitalization count in each metropolitan city on day t and β is the estimate of the association between PM2.5 and ALRI hospitalization, ATt is daily AT in a metropolitan city on day t, Day of week (DOW) indicates the categorical variable with a reference day of Sunday (from Monday to Sunday), γ means the estimate of the DOW and s is a penalized spline term. The spline function for Time was applied to control for unmeasured confounder (e.g., seasonality, trends) of daily ALRI hospitalization counts [30]. Since air pressure, wind speed and other meteorological variables did not improve the model, they were excluded from the main model. We estimated the associations of single-lag (lag 0, 1, …, 7) and moving average-lag (Lag 01, 02, …, 07) days of PM2.5 with ALRI hospitalization. To focus on the short-term effects of PM2.5 on ALRI hospitalization, we considered the following PM2.5 exposure windows: single-lag day exposure from current day (lag 0) to 7 days before hospitalization (lag 7) and moving average exposure by averaging PM2.5 levels for 2 days (current day and 1 previous day before hospitalization, lag 01) to 8 days (current day and 7 previous days before hospitalization, lag 07). The main results are evaluated as a percentage change (%) of the odds ratio of daily ALRI hospitalization associated with a 10 μg/m3 increase in 7-day moving average exposure to PM2.5. The percentage change was calculated using the following formula: ( e β * 10 − 1) × 100.
To select degrees of freedom (df) for time trends, we performed the analysis by allocating the df per year from 1 to 9 (i.e., df = 1 × 9, …, 9 × 9) and evaluated the patterns of the effects estimates in each metropolitan city. The df for AT was selected by referring to previous papers [31].
We pooled the city-specific estimates of the associations between PM2.5 and ALRI hospitalizations in urban areas. We used the random effect model to account for the heterogeneity of the estimates across regions [32,33]. We evaluated the heterogeneity of the model through I2, H2 and Cochran’s Q statistics. If the I2 statistic is less than 25%, there is low heterogeneity, medium for 25 to 50%, and high heterogeneity for greater than 50% [34].
When the H2 statistic is close to 1, this indicates homogeneity [35]. The Cochran’s Q statistic is based on a chi-square distribution, and a larger value means greater variation across studies rather than within subjects within a study [36]. We presented the meta-analysis results in a forest plot.
Subgroup analyses were conducted by sex (boys and girls) and season (warm season (April to September) and cold season (October to March)).

2.5. Two-Pollutant Model

We constructed a two-pollutant model to control for the effect of other pollutants on the ALRI hospitalization [37] because risk estimates can be affected by the accompanying air pollutants in the two-pollutant model [26]. We used the Variance Inflation Factor (VIF) for each pollutant pair to evaluate the multi-collinearity between the two pollutants (i.e., PM2.5 and other pollutants). If the VIF value of the other pollutant was less than 10, we included the pollutant in the two-pollutant models [38].
The software used for all data pre-processing and analysis are SAS Ver. 9.4 (SAS Institute Inc., Cary, NC, USA) and R Programming Language (Ver. 4.0.0, R Foundation for Statistical Computing, Vienna, Austria).

3. Results

3.1. Main Results

From 2008 to 2016, there were 713,588 cases of children (0–5 years) hospitalized in seven metropolitan cities of South Korea (Seoul: 147,225, Busan: 213,336, Daegu: 90,329, Incheon: 60,606, Gwangju: 112,196, Daejeon: 34,944, and Ulsan: 54,952) (Table 1). The number of daily ALRI hospitalizations was slightly higher in boys (n = 397,333 (55.7%)) than in girls (n = 316,255 (44.3%)). The daily ALRI hospitalization distribution shows the shape of a right-skewed distribution (Figure S1). Table 2 describes the summary statistics for daily air pollutants and environmental variables. The average (standard deviation, SD) daily PM2.5 concentration during the 2008 to 2016 was 29.0 (7.6) μg/m3 (Gwangju being the lowest, 24.8 (7.9) μg/m3 and Seoul being the highest, 30.6 (8.3) μg/m3) (Figure 1 and Table 2). The average (SD) daily SO2, NO2 and O3 concentrations (ppm) were 0.005 (0.002), 0.024 (0.009) and 0.025 (0.010), respectively.
We found slightly greater estimations of the associations when we applied 1 to 4 df per year (Figure S2). Therefore, we selected 7 df per year (63 df = 7 per year × 9 years) for Time (Figure S2). The exposure–response relationship between PM2.5 and the risk of ALRI hospitalizations are presented in Figure 2. In Figure 2, we showed a strong linear association between PM2.5 and the risk of ALRI hospitalization. Figure 3 shows the city-specific estimates and pooled estimates between PM2.5 exposure and ALRI hospitalization risk. A 7-day moving average PM2.5 was associated with a 1.20% (95% confidence interval (CI): 0.71, 1.71) increase in ALRI hospitalization risk.

3.2. Subgroup Analysis

In the stratified analysis by sex, we found a similar association of PM2.5 exposure with ALRI hospitalization (boys: 1.31%, 95% CI: 0.75, 1.86; girls: 1.02%, 95% CI: 0.45, 1.58). However, we also observed differences in the associations by the season. The association in the warm season was shown (warm season: 1.71%, 95% CI: 0.94, 2.48; cold season: 0.31%, 95% CI: −0.51, 1.13).

3.3. Two-Pollutants Model

There were positive correlations between the modeled PM2.5 and the monitored pollutants: PM10 [correlation coefficient r =  0.96], SO2 (r = 0.76), NO2 (r =  0.68), and CO (r = 0.75). However, there was no correlation between modeled PM2.5 and monitored O3 (r = −0.03) (Figure 4). In the two-pollutant model, we considered SO2, NO2 and O3. After adjusting for NO2 and O3, the model was similar to that of the single pollutant model (Table 3). However, after it was SO2 adjusted, the association with PM2.5 was unrelated (0.69%, 95% CI: −0.34, 1.74).

3.4. Sensitivity Analysis

The results for the single-day and moving average-day lag models are shown in Figure 5, Tables S1 and S2. In a single-day lag model, we observed decreasing estimates with increasing lag days (Lag 0: 0.58%, 95% CI: 0.41, 0.76; Lag 7: −0.01, 95% CI: −0.18, 0.16). In the moving average lag model, the estimated association with the 7-day moving average exposure to PM2.5 was the highest (Lag 06: 1.20%, 95% CI: 0.71, 1.71). The residual plots and histograms for our main model in the first stage showed no specific patterns (Figure S3). In an additional analysis to validate the CMAQ Modeled PM2.5 concentration levels, we observed higher correlation coefficients between the CMAQ modeled and monitored PM2.5 data levels in each of the seven cities (0.96 or higher) (Figure S4).

4. Discussion

We used the time series data for 9 years on children’s ALRI hospitalization. In our study, we evaluated the association between PM2.5 and ALRI hospitalization in urban areas of South Korea. We found a 1.2% increase in ALRI hospitalization per 10 μg/m3 increase in the 7-day moving average exposure to PM2.5. A strong linear association was observed between PM2.5 and ALRI hospitalizations, and the association was stronger in the warm season. However, we found no association in the cold season. The associations of PM2.5 with ALRI hospitalization were slightly attenuated after adjusting for NO2, O3, and SO2 in the two-pollutant model.
Previous short-term studies have evaluated the relationship between air pollution and ALRI in children and have mostly conducted their analyses with time-series data or in a case-crossover design. Time-series data for 94,315 children under age 15 in China revealed the association between PM2.5 and ALRI hospitalizations (≤1:1.26% (0.03, 2.50; 2–4 years: 1.24% (0.35, 2.14); 5–14 years old: 1.72% (0.34, 3.12)) [16]. In a study with time-series data from Brazil, increases of PM10 was associated with an increase in hospitalizations for pediatric or adolescent respiratory diseases, including upper respiratory tract acute infections, acute bronchitis, bronchiolitis, asthma, pneumonia, bronchopneumonia and other respiratory diseases [39]. In another case-crossover study with time-series data of 50,857 children under 3 years old in Brazil, a 4-day moving average PM10 was associated with a 4% increase in bronchiolitis hospitalizations [40]. In a case-crossover study for multi cities, exposures to PM2.5, PM10, NO2, and SO2 were associated with hospitalization of children with pneumonia and acute bronchitis [41]. In a case-crossover study with 112,467 children in the US, exposure to PM2.5 within a week was related to ALRI hospitalization aged 0–2 years, 3–17 years and over 18 years [10]. A Malaysia study evaluated a positive association between primary air pollutants (PM10, NO2 and O3) and ALRI hospitalization of children in cities [42].
A Vietnam study showed that exposure to PM2.5 increased ALRI hospitalization for children aged 0–4 (3.51%, 95% CI: 0.96, 6.12). In this time-series study, boys were higher risk of PM2.5 exposure (male: 5.22%, 95% CI: 0.86, 9.78; female: −5.84, −4.30) [22]. In our study, the subgroup analysis by sex was similar, but the risk for boys was slightly higher. This is probably because boys are more presumably exposed to ambient PM2.5 than girls [22]. In Korea, there was a study showing increased respiratory mortality in infants with exposure to high PM10 levels [43]. Another nationwide study in Korea in children aged 0-4 reported a stronger association between PM2.5 and ALRI hospitalization in the cold season [21]. The discrepancy between the two studies by Kim et al. and our study may come from the different definitions of ALRI hospitalization (primary diagnostic codes of ALRI in Kim et al. vs. primary or secondary diagnostic codes in our study), study periods (2013–2015 in Kim et al., vs. 2006–2015 in our study), study areas (nine provinces and seven metropolitan cities in Kim et al., vs. only seven metropolitan cities in our study), and analysis methods (a difference-in-differences approach in Kim et al., vs. a Poisson regression model in our study).
There are a few studies reporting contradictory results. A study with time-series data of US emergency room hospitalizations has revealed an association of PM2.5 with pneumonia and upper respiratory diseases, but not with bronchiolitis hospitalizations [17]. In a study with 17,118 children under 18 years old, the association between hospitalization due to pneumonia was associated with PM2.5, but bronchitis and asthma were not related to PM [44]. Another case-crossover study with 19,901 infants in California found no link between PM2.5 and bronchitis [18].
There were several previous studies considering the two-pollutants model. Zheng et al. showed that the estimate of the association with PM2.5 was not significant in the two-pollutant model after controlling for SO2 and NO2 (PM2.5: 1.50, 95% CI: 0.35, 2.66; PM2.5 adjusted for SO2: 0.17, 95% CI: −2.55, 0.43; PM2.5 adjusted for NO2: 0.17, 95% CI: −1.47, 1.85) [16]. Barnett et al. argued that PM2.5 and SO2 concentration levels are a proxy measure of the same emission source (e.g., motor vehicles), and they may impact each other [41]. They showed that SO2 attenuated the effects of PM2.5 in the adjusted two-pollutant model, which is similar to our findings.
Most air pollutants, including fine dust, are caused by traffic, industries and other sources of combustion [37,41]. These air pollutants are highly correlated because they are affected by climatic conditions (temperature, humidity, wind speed, air pressure, and etc.). In this study, air pollutants excluding ozone were highly correlated with PM2.5 (Figure 4). Therefore, it is important to note that models containing multiple air pollutants may have collinearity.
We observed a greater estimate of the link between PM2.5 and ALRI hospitalization in the warm season. The association between PM2.5 and ALRI in children aged 1–4 was more outstanding in warm seasons in Australia and New Zealand in a previous study, which is consistent with our study [41]. A high temperature is well-known to have undesirable health effects; a high temperature exacerbates pre-existing health conditions [45] and is related to the death of susceptible population group (e.g., young children, elderly) [46]. Exposure to PM2.5 is associated with decreased heart rate variability (HRV), suggesting it triggers a change in the autonomic nervous system [47,48]. Interestingly, exposure to high temperature was associated with decrease in HRV during the warm season [49]. This suggests that high levels of PM2.5 and ambient temperature could have interaction effects on autonomic nervous system [50], which could interact with the immune system [51]. On the other hand, another study in San Paulo showed greater effects of PM10 during the cold season [52]; however, the average ambient temperature in San Paulo occurs during the warm season, and 13–22 °C is considered the cold season, which is very different from the significant seasonal changes in ambient temperature in Korea. Discrepancies of the season-specific associations may come from the geographical and climatic characteristics of PM2.5 [52,53]. The composition of PM2.5 may vary by season, and different compositions of PM2.5 may cause various adverse effects, particularly in urban areas with heavy traffic volumes [54]. In cold seasons, while the concentration of fine dust is much higher than in warm seasons, the contributions of secondary sulfate salts and nitrate salts are greater during the summer in Beijing, China [55]. Although the PM2.5 concentration was also higher in the cold season in Seoul as well, the concentration of NH4+ was higher in the cold season while SO42− was higher during the warm season [56]. Different study periods, geographic coverages, and definitions of outcomes may also impact the study results. In addition, exposure to high temperatures and fine dust can affect the lung function of the elderly as well as in ALRI on children [57,58]. The interaction of high temperatures and PM2.5 may be the cause, and these effects need to be addressed in future studies.
Some studies have reported PM2.5 and ALRI relevance, explaining oxidative stress and inflammatory reactions as causal mechanisms. PM2.5 can cause oxidant stress to the lung tissue and induce systemic inflammation [59,60]. PM2.5 stimulates the expression of inflammatory cytokines, which may cause systematic inflammation [61].The PM2.5 inhaled in the body mainly affects macrophages and epithelial cells in the respiratory tract. In vitro and in vivo studies have described that PM2.5 reduces the phagocytic activities of alveolar macrophages, making people vulnerable to air pollutants [62,63]. Because of immature lung development, children are vulnerable to health effects if they inhale PM2.5. In addition, children breathe more air pollutants because they breathe faster than adults.
Young children may be diagnosed with ALRI for respiratory syncytial virus (RSV) infection and bronchitis. If children with RSV or bronchitis are exposed to PM2.5, the ALRI risk may increase [10].
This study has several limitations. First, we were unable to use monitored data of PM2.5 due to the short period of nationwide monitoring of PM2.5. So, we compared modeled PM2.5 levels with monitored data in 2015 and 2016 to validate the modeled concentration levels of PM2.5. We found a high correlation between modeled and monitored exposure levels in years with available data for both PM2.5 levels (2015 and 2016) (Figure S4). There are several studies that have estimated and evaluated the CMAQ PM2.5 concentrations [64,65]. These studies found that CMAQ PM2.5 concentrations can be affected by weather variables (temperature, wind, etc.), emissions inventory, etc. In addition, during the 2013–2014 period, CMAQ PM2.5 concentrations may be underestimated due to high concentrations of air pollution and high emissions [64]. High concentrations were included and emissions were high in 2013 and 2014, so that the reproducibility of the simulation was somewhat lower than that of other years. However, the level of model performance evaluation was not low, so we concluded that the CMAQ PM2.5 data were suitable for our study. Second, we were unable to gather individual-level data such as socioeconomic status, parental education levels, and secondary smoking status due to the health insurance nature of the data used in our study. Third, we could not control the unmeasured confounding effects [66]. However, the time trends in the model may control for short and long-term temporal fluctuations, which may reduce unmeasured confounding effects [67]. Fourth, exposure data were applied as the daily average in each region. Therefore, potential misclassification may have been caused by exposure values. Finally, this research was evaluated in metropolitan cities of South Korea. Thus, our results cannot be generalized to rural areas or other countries with different PM2.5 levels.

5. Conclusions

Our results have revealed a link between PM2.5 and ALRI hospitalization among Korean children. The increase in PM2.5 is related to the increased ALRI hospitalization for Korean children. In particular, exposure to PM2.5 in warm seasons was highly related to ALRI. Children need to refrain from outdoor activity when PM2.5 is high. The results of our research can be used for children’s health care and health policies for children’s respiratory disease. Future research needs to investigate the link between air pollution, temperature and children’s respiratory health.
Since PM2.5 is mainly caused by traffic and industrial emissions, government control is necessary. The South Korean government announced its 2nd Seoul Metropolitan Air Quality Control Master Plan from 2015 to 2024. It includes regulations such as reducing eco-friendly cars and carbon emissions and strengthening management around life. In the future, managing air pollution is expected to reduce children’s health effects caused by exposure to air pollution.

Supplementary Materials

The following are available online at https://www.mdpi.com/1660-4601/18/1/144/s1, Figure S1: The histogram of daily ALRI hospitalization in seven cities, in Korea, Figure S2: Estimation of log relative risk according to the degrees of freedom for time per year. The model was adjusted for apparent temperature, day of weeks and time trends, Figure S3: Residuals plots for fitted models for each region (left), and residual histograms for fitted models for each region (right), Figure S4: Correlation between daily CMAQ PM2.5 concentration and Monitoring PM2.5 concentration by region (2015–2016). Table S1: The association between single lag PM2.5 concentrations and ALRI hospital admissions among children the seven major cities in Korea, 2018–2016, Table S2: The association between moving average PM2.5 concentrations and ALRI hospital admissions among children the seven major cities in Korea, 2018–2016. The moving average is the average from the current day to the previous day. Time-series model adjusted for same covariates (time, AT, day of weeks). The percentage change is calculated as the change per PM2.5 10 μg/m3 increase.

Author Contributions

Conceptualization: J.O., C.H., Y.-C.H. and Y.-H.L.; Methodology: J.O., C.H., D.-W.L., H.J.B., E.H., Y.-C.H. and Y.-H.L.; Formal analysis/Visualization: J.O. and C.H.; Investigation: J.O., C.H., Y.J., Y.-J.C. and S.K.; Data curation: J.O., C.H., S.K.; Writing—original draft: All authors; Writing—review and editing: J.O., C.H., D.-W.L., Y.J., Y.-J.C., Y.-C.H. and Y.-H.L. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by the National Strategic Project–Fine Particle of the National Research Foundation of Korea, funded by the Ministry of Science, Information, Communications, and Technology, the Ministry of Environment, and the Ministry of Health and Welfare (grant nos. 2017M3D8A1092008, 2017M3D8A1092009), research fund of Chungnam National University, as well as the Center for Environmental Health through the Ministry of Environment.

Institutional Review Board Statement

The study was approved by the Institutional Review Board (IRB) of Seoul National University (E-1807-038-956).

Informed Consent Statement

This data was used de-identified data provided by National Health Insurance Service (NHIS) and complied with the NHIS personal information protection guideline. Therefore, the requirement for informed consent was waived by the IRB.

Data Availability Statement

The data used in this study is available from the NHIS. However, this data requires deliberation by the NHIS and analysis is conducted in a limited closed network. Only limited analysis results can be obtained.

Acknowledgments

This study was conducted using data from the National Health Insurance Service of Korea (NHIS-2019-1-023).

Conflicts of Interest

The authors declare they have no actual or potential competing financial interests.

Abbreviations

PM2.5particulate matter with a diameter equal to or less than 2.5 μm
ALRIacute lower respiratory infection
GAMgeneralized additive model
RRrelative risk
CIconfidence interval

References

  1. Apte, J.S.; Marshall, J.D.; Cohen, A.J.; Brauer, M. Addressing Global Mortality from Ambient PM2.5. Environ. Sci. Technol. 2015, 49, 8057–8066. [Google Scholar] [CrossRef] [PubMed]
  2. Ostro, B.; Roth, L.; Malig, B.; Marty, M. The effects of fine particle components on respiratory hospital admissions in children. Environ. Health Perspect. 2009, 117, 475–480. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  3. Koenig, J.Q.; Larson, T.V.; Hanley, Q.S.; Rebolledo, V.; Dumler, K.; Checkoway, H.; Wang, S.Z.; Lin, D.; Pierson, W.E. Pulmonary function changes in children associated with fine particulate matter. Environ. Res. 1993, 63, 26–38. [Google Scholar] [CrossRef] [PubMed]
  4. Grigg, J. Particulate matter exposure in children: Relevance to chronic obstructive pulmonary disease. Proc. Am. Thorac. Soc. 2009, 6, 564–569. [Google Scholar] [CrossRef] [PubMed]
  5. Tecer, L.H.; Alagha, O.; Karaca, F.; Tuncel, G.; Eldes, N. Particulate matter (PM(2.5), PM(10-2.5), and PM(10)) and children’s hospital admissions for asthma and respiratory diseases: A bidirectional case-crossover study. J. Toxicol. Environ. Health A 2008, 71, 512–520. [Google Scholar] [CrossRef]
  6. Guarnieri, M.; Balmes, J.R. Outdoor air pollution and asthma. Lancet 2014, 383, 1581–1592. [Google Scholar] [CrossRef] [Green Version]
  7. Mehta, S.; Shin, H.; Burnett, R.; North, T.; Cohen, A.J. Ambient particulate air pollution and acute lower respiratory infections: A systematic review and implications for estimating the global burden of disease. Air Qual. Atmos. Health 2013, 6, 69–83. [Google Scholar] [CrossRef] [Green Version]
  8. Colley, J.R.; Douglas, J.W.; Reid, D.D. Respiratory disease in young adults: Influence of early childhood lower respiratory tract illness, social class, air pollution, and smoking. Br. Med. J. 1973, 3, 195–198. [Google Scholar] [CrossRef] [Green Version]
  9. Spears, D.; Dey, S.; Chowdhury, S.; Scovronick, N.; Vyas, S.; Apte, J. The association of early-life exposure to ambient PM2.5 and later-childhood height-for-age in India: An observational study. Environ. Health 2019, 18, 62. [Google Scholar] [CrossRef] [Green Version]
  10. Horne, B.D.; Joy, E.A.; Hofmann, M.G.; Gesteland, P.H.; Cannon, J.B.; Lefler, J.S.; Blagev, D.P.; Korgenski, E.K.; Torosyan, N.; Hansen, G.I.; et al. Short-Term Elevation of Fine Particulate Matter Air Pollution and Acute Lower Respiratory Infection. Am. J. Respir. Crit. Care Med. 2018, 198, 759–766. [Google Scholar] [CrossRef]
  11. GBD. Estimates of the global, regional, and national morbidity, mortality, and aetiologies of lower respiratory tract infections in 195 countries: A systematic analysis for the Global Burden of Disease Study 2015. Lancet Infect. Dis. 2015, 17, 1133–1161. [Google Scholar] [CrossRef] [Green Version]
  12. Bruce, N. Indoor air pollution from unprocessed solid fuel use and pneumonia risk in children aged under five years: A systematic review and meta-analysis. Bull. World Health Organ. 2008, 86, 390–398. [Google Scholar] [CrossRef] [PubMed]
  13. Rahman, I.; MacNee, W. Role of oxidants/antioxidants in smoking-induced lung diseases. Free Radic. Biol. Med. 1996, 21, 669–681. [Google Scholar] [CrossRef]
  14. Greenwell, L.L.; Moreno, T.; Jones, T.P.; Richards, R.J. Particle-induced oxidative damage is ameliorated by pulmonary antioxidants. Free Radic. Biol. Med. 2002, 32, 898–905. [Google Scholar] [CrossRef]
  15. Donaldson, K.; Beswick, P.H.; Gilmour, P.S. Free radical activity associated with the surface of particles: A unifying factor in determining biological activity? Toxicol. Lett. 1996, 88, 293–298. [Google Scholar] [CrossRef]
  16. Zheng, P.W.; Wang, J.B.; Zhang, Z.Y.; Shen, P.; Chai, P.F.; Li, D.; Jin, M.J.; Tang, M.L.; Lu, H.C.; Lin, H.B.; et al. Air pollution and hospital visits for acute upper and lower respiratory infections among children in Ningbo, China: A time-series analysis. Environ. Sci. Pollut. Res. Int. 2017, 24, 18860–18869. [Google Scholar] [CrossRef]
  17. Darrow, L.A.; Klein, M.; Flanders, W.D.; Mulholland, J.A.; Tolbert, P.E.; Strickland, M.J. Air pollution and acute respiratory infections among children 0-4 years of age: An 18-year time-series study. Am. J. Epidemiol. 2014, 180, 968–977. [Google Scholar] [CrossRef] [Green Version]
  18. Karr, C.; Lumley, T.; Shepherd, K.; Davis, R.; Larson, T.; Ritz, B.; Kaufman, J. A case-crossover study of wintertime ambient air pollution and infant bronchiolitis. Environ. Health Perspect. 2006, 114, 277–281. [Google Scholar] [CrossRef]
  19. Health Effects Institute. State of Global Air 2017: A Special Report on Global Exposure to Air Pollution and Its Disease Burden; Health Effect Institute: Boston, MA, USA, 2017. [Google Scholar]
  20. Park, J.-E.; Lee, S.-Y.; Kang, J.H. Common Disease Codes in Pediatric Inpatients (1997–2008). Korean J. Pediatric Infect. Dis. 2011, 18. [Google Scholar] [CrossRef]
  21. Kim, K.N.; Kim, S.; Lim, Y.H.; Song, I.G.; Hong, Y.C. Effects of short-term fine particulate matter exposure on acute respiratory infection in children. Int. J. Hyg. Environ. Health 2020, 229, 113571. [Google Scholar] [CrossRef]
  22. Luong, L.T.M.; Dang, T.N.; Thanh Huong, N.T.; Phung, D.; Tran, L.K.; Van Dung, D.; Thai, P.K. Particulate air pollution in Ho Chi Minh city and risk of hospital admission for acute lower respiratory infection (ALRI) among young children. Environ. Pollut. 2020, 257. [Google Scholar] [CrossRef] [PubMed]
  23. Seong, S.C.; Kim, Y.Y.; Khang, Y.H.; Heon Park, J.; Kang, H.J.; Lee, H.; Do, C.H.; Song, J.S.; Hyon Bang, J.; Ha, S.; et al. Data Resource Profile: The National Health Information Database of the National Health Insurance Service in South Korea. Int. J. Epidemiol. 2017, 46, 799–800. [Google Scholar] [CrossRef] [Green Version]
  24. Oh, J.; Lee, S.; Kim, M.H.; Kwag, Y.; Kim, H.S.; Kim, S.; Ye, S.; Ha, E. The impact of PM2.5 on acute otitis media in children (aged 0–3): A time series study. Environ. Int. 2020, 145, 106133. [Google Scholar] [CrossRef] [PubMed]
  25. Han, C.; Kim, S.; Lim, Y.H.; Bae, H.J.; Hong, Y.C. Spatial and Temporal Trends of Number of Deaths Attributable to Ambient PM2.5 in the Korea. J. Korean Med Sci. 2018, 33, e193. [Google Scholar] [CrossRef]
  26. Tolbert, P.E.; Klein, M.; Peel, J.L.; Sarnat, S.E.; Sarnat, J.A. Multipollutant modeling issues in a study of ambient air quality and emergency department visits in Atlanta. J. Expo. Sci. Environ. Epidemiol. 2007, 17, S29–S35. [Google Scholar] [CrossRef] [Green Version]
  27. Seo, J.; Kim, J.Y.; Youn, D.; Lee, J.Y.; Kim, H.; Lim, Y.B.; Kim, Y.; Jin, H.C. On the multiday haze in the Asian continental outflow: The important role of synoptic conditions combined with regional and local sources. Atmos. Chem. Phys. 2017, 17, 9311–9332. [Google Scholar] [CrossRef] [Green Version]
  28. Zanobetti, A.; Schwartz, J. Temperature and Mortality in Nine US Cities. Epidemiology 2008, 19, 563–570. [Google Scholar] [CrossRef] [Green Version]
  29. Nguyen, J.L.; Schwartz, J.; Dockery, D.W. The relationship between indoor and outdoor temperature, apparent temperature, relative humidity, and absolute humidity. Indoor Air 2014, 24, 103–112. [Google Scholar] [CrossRef] [Green Version]
  30. Bhaskaran, K.; Gasparrini, A.; Hajat, S.; Smeeth, L.; Armstrong, B. Time series regression studies in environmental epidemiology. Int. J. Epidemiol. 2013, 42, 1187–1195. [Google Scholar] [CrossRef]
  31. Alessandrini, E.; Sajani, S.Z.; Scotto, F.; Miglio, R.; Marchesi, S.; Lauriola, P. Emergency ambulance dispatches and apparent temperature: A time series analysis in Emilia-Romagna, Italy. Environ. Res. 2011, 111, 1192–1200. [Google Scholar] [CrossRef]
  32. Rucker, G.; Schwarzer, G.; Carpenter, J.R.; Schumacher, M. Undue reliance on I(2) in assessing heterogeneity may mislead. BMC Med. Res. Methodol. 2008, 8, 79. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  33. Borenstein, M.; Hedges, L.V.; Higgins, J.P.; Rothstein, H.R. A basic introduction to fixed-effect and random-effects models for meta-analysis. Res. Synth. Methods 2010, 1, 97–111. [Google Scholar] [CrossRef] [PubMed]
  34. Melsen, W.G.; Bootsma, M.C.; Rovers, M.M.; Bonten, M.J. The effects of clinical and statistical heterogeneity on the predictive values of results from meta-analyses. Clin. Microbiol. Infect. 2014, 20, 123–129. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  35. Dwivedi, S.N. Which is the Preferred Measure of Heterogeneity in Meta-Analysis and Why? A Revisit. Biostat. Biom. Open Access J. 2017, 1. [Google Scholar] [CrossRef] [Green Version]
  36. Kulinskaya, E.; Dollinger, M.B. An accurate test for homogeneity of odds ratios based on Cochran’s Q-statistic. BMC Med. Res. Methodol. 2015, 15, 49. [Google Scholar] [CrossRef] [Green Version]
  37. Ito, K.; Thurston, G.D.; Silverman, R.A. Characterization of PM2.5, gaseous pollutants, and meteorological interactions in the context of time-series health effects models. J. Expo. Sci. Environ. Epidemiol. 2007, 17, S45–S60. [Google Scholar] [CrossRef] [Green Version]
  38. Vatcheva, K.P.; Lee, M.; McCormick, J.B.; Rahbar, M.H. Multicollinearity in Regression Analyses Conducted in Epidemiologic Studies. Epidemiology (Sunnyvale) 2016, 6. [Google Scholar] [CrossRef] [Green Version]
  39. Braga, A.L.; Saldiva, P.H.; Pereira, L.A.; Menezes, J.J.; Conceicao, G.M.; Lin, C.A.; Zanobetti, A.; Schwartz, J.; Dockery, D.W. Health effects of air pollution exposure on children and adolescents in Sao Paulo, Brazil. Pediatr. Pulmonol. 2001, 31, 106–113. [Google Scholar] [CrossRef]
  40. Segala, C.; Poizeau, D.; Mesbah, M.; Willems, S.; Maidenberg, M. Winter air pollution and infant bronchiolitis in Paris. Environ. Res. 2008, 106, 96–100. [Google Scholar] [CrossRef]
  41. Barnett, A.G.; Williams, G.M.; Schwartz, J.; Neller, A.H.; Best, T.L.; Petroeschevsky, A.L.; Simpson, R.W. Air pollution and child respiratory health: A case-crossover study in Australia and New Zealand. Am. J. Respir. Crit. Care Med. 2005, 171, 1272–1278. [Google Scholar] [CrossRef]
  42. Ismail, S.N.S.; Abdul Rahman, S.R.; Sahani, M.; Firuz, R.; Latif, M.T. A case crossover analysis of primary air pollutants association on acute respiratory infection (ARI) among children in urban region of Klang valley, Malaysia. Ann. Trop. Med. Public Health 2017, 10. [Google Scholar] [CrossRef]
  43. Ha, E.H.; Lee, J.T.; Kim, H.; Hong, Y.C.; Lee, B.E.; Park, H.S.; Christiani, D.C. Infant Susceptibility of Mortality to Air Pollution in Seoul, South Korea. Pediatrics 2003, 111, 284–290. [Google Scholar] [CrossRef] [PubMed]
  44. Nhung, N.T.T.; Schindler, C.; Dien, T.M.; Probst-Hensch, N.; Perez, L.; Kunzli, N. Acute effects of ambient air pollution on lower respiratory infections in Hanoi children: An eight-year time series study. Environ. Int. 2018, 110, 139–148. [Google Scholar] [CrossRef] [PubMed]
  45. Gordon, C.J. Role of environmental stress in the physiological response to chemical toxicants. Environ. Res. 2003, 92, 1–7. [Google Scholar] [CrossRef]
  46. Kalkstein, L.S.; Greene, J.S. An evaluation of climate/mortality relationships in large U.S. cities and the possible impacts of a climate change. Environ. Health Perspect. 1997, 105, 84–93. [Google Scholar] [CrossRef]
  47. Gold, D.R.; Litonjua, A.; Schwartz, J.; Lovett, E.; Larson, A.; Nearing, B.; Allen, G.; Verrier, M.; Cherry, R.; Verrier, R. Ambient Pollution and Heart Rate Variability. Circulation 2000, 101, 1267–1273. [Google Scholar] [CrossRef]
  48. Creason, J.; Neas, L.; Walsh, D.; Williams, R.O.N.; Sheldon, L.; Liao, D.; Shy, C. Particulate matter and heart rate variability among elderly retirees: The Baltimore 1998 PM study. J. Expo. Sci. Environ. Epidemiol. 2001, 11, 116–122. [Google Scholar] [CrossRef] [Green Version]
  49. Ren, C.; O’Neill, M.S.; Park, S.K.; Sparrow, D.; Vokonas, P.; Schwartz, J. Ambient Temperature, Air Pollution, and Heart Rate Variability in an Aging Population. Am. J. Epidemiol. 2011, 173, 1013–1021. [Google Scholar] [CrossRef] [Green Version]
  50. Ren, C.; Tong, S. Temperature modifies the health effects of particulate matter in Brisbane, Australia. Int. J. Biometeorol. 2006, 51, 87–96. [Google Scholar] [CrossRef]
  51. Kenney, M.J.; Ganta, C.K. Autonomic Nervous System and Immune System Interactions. Compr Physiol 2014, 4, 1177–1200. [Google Scholar]
  52. Gouveia, N.; Fletcher, T. Time series analysis of air pollution and mortality: Effects by cause, age and socioeconomic status. J. Epidemiol. Community Health 2000, 54, 750–755. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  53. Cox, L.A.T.; Liu, X.; Shi, L.; Zu, K.; Goodman, J. Applying Nonparametric Methods to Analyses of Short-Term Fine Particulate Matter Exposure and Hospital Admissions for Cardiovascular Diseases among Older Adults. Int. J. Environ. Res. Public Health 2017, 14, 1051. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  54. Cesari, D.; De Benedetto, G.E.; Bonasoni, P.; Busetto, M.; Dinoi, A.; Merico, E.; Chirizzi, D.; Cristofanelli, P.; Donateo, A.; Grasso, F.M.; et al. Seasonal variability of PM2.5 and PM10 composition and sources in an urban background site in Southern Italy. Sci. Total Environ. 2018, 612, 202–213. [Google Scholar] [CrossRef] [PubMed]
  55. Chen, R.; Cheng, J.; Lv, J.; Wu, L.; Wu, J. Comparison of chemical compositions in air particulate matter during summer and winter in Beijing, China. Environ. Geochem. Health 2016, 39, 913–921. [Google Scholar] [CrossRef] [PubMed]
  56. Park, J.; Park, E.H.; Schauer, J.J.; Yi, S.-M.; Heo, J. Reactive oxygen species (ROS) activity of ambient fine particles (PM2.5) measured in Seoul, Korea. Environ. Int. 2018, 117, 276–283. [Google Scholar] [CrossRef] [PubMed]
  57. Li, J.; Woodward, A.; Hou, X.Y.; Zhu, T.; Zhang, J.L.; Brown, H.; Yang, J.; Qin, R.N.; Gao, J.H.; Gu, S.H.; et al. Modification of the effects of air pollutants on mortality by temperature: A systematic review and meta-analysis. Sci. Total Environ. 2017, 575, 1556–1570. [Google Scholar] [CrossRef] [PubMed]
  58. Lepeule, J.; Litonjua, A.A.; Gasparrinie, A.; Koutrakis, P.; Sparrows, D.; Vokonas, P.S.; Schwartz, J. Lung function association with outdoor temperature and relative humidity and its interaction with air pollution in the elderly. Environ. Res. 2018, 165, 110–117. [Google Scholar] [CrossRef]
  59. Zhu, L.; Ge, X.; Chen, Y.; Zeng, X.; Pan, W.; Zhang, X.; Ben, S.; Yuan, Q.; Xin, J.; Shao, W.; et al. Short-term effects of ambient air pollution and childhood lower respiratory diseases. Sci. Rep. 2017, 7, 4414. [Google Scholar] [CrossRef] [Green Version]
  60. Lodovici, M.; Bigagli, E. Oxidative Stress and Air Pollution Exposure. J. Toxicol. 2011, 2011, 1–9. [Google Scholar] [CrossRef]
  61. Sigaud, S.; Goldsmith, C.A.; Zhou, H.; Yang, Z.; Fedulov, A.; Imrich, A.; Kobzik, L. Air pollution particles diminish bacterial clearance in the primed lungs of mice. Toxicol. Appl. Pharmacol. 2007, 223, 1–9. [Google Scholar] [CrossRef] [Green Version]
  62. Miyata, R.; van Eeden, S.F. The innate and adaptive immune response induced by alveolar macrophages exposed to ambient particulate matter. Toxicol. Appl. Pharmacol. 2011, 257, 209–226. [Google Scholar] [CrossRef] [PubMed]
  63. Jalava, P.I.; Salonen, R.O.; Pennanen, A.S.; Sillanpaa, M.; Halinen, A.I.; Happo, M.S.; Hillamo, R.; Brunekreef, B.; Katsouyanni, K.; Sunyer, J.; et al. Heterogeneities in inflammatory and cytotoxic responses of RAW 264.7 macrophage cell line to urban air coarse, fine, and ultrafine particles from six European sampling campaigns. Inhal. Toxicol. 2007, 19, 213–225. [Google Scholar] [CrossRef] [PubMed]
  64. Bae, M.; Kim, B.-U.; Kim, H.C.; Kim, S. A Multiscale Tiered Approach to Quantify Contributions: A Case Study of PM2.5 in South Korea during 2010–2017. Atmosphere 2020, 11, 141. [Google Scholar] [CrossRef] [Green Version]
  65. Kim, J.-H.; Choi, D.-R.; Koo, Y.-S.; Lee, J.-B.; Park, H.-J. Analysis of Domestic and Foreign Contributions using DDM in CMAQ during Particulate Matter Episode Period of February 2014 in Seoul. J. Korean Soc. Atmos. Environ. 2016, 32, 82–99. [Google Scholar] [CrossRef] [Green Version]
  66. VanderWeele, T.J.; Arah, O.A. Bias Formulas for Sensitivity Analysis of Unmeasured Confounding for General Outcomes, Treatments, and Confounders. Epidemiology 2011, 22, 42–52. [Google Scholar] [CrossRef] [Green Version]
  67. Peng, R.D.; Dominici, F. Statistical Methods for Environmental Epidemiology with R: A Case Study in Air Pollution and Health; Springer: Berlin/Heidelberg, Germany, 2008. [Google Scholar]
Ijerph 18 00144 g001 550
Figure 1. Map of Korea and average fine particulate matter (PM2.5) concentrations during the study period (2008–2016). The names of seven metropolitan cities are indicated with orange color.
Figure 1. Map of Korea and average fine particulate matter (PM2.5) concentrations during the study period (2008–2016). The names of seven metropolitan cities are indicated with orange color.
Ijerph 18 00144 g001
Ijerph 18 00144 g002 550
Figure 2. Exposure–response relationship between short-term PM2.5 concentrations and ALRI hospital admissions in children in seven metropolitan cities in Korea. The gray shade indicates the 95% confidence interval. The figure below shows the ALRI hospital admission distribution over PM2.5 concentrations. All models used the 7-day moving average exposure to PM2.5 μg/m3. The percentage change is calculated as the change per PM2.5 10 μg/m3 increase.
Figure 2. Exposure–response relationship between short-term PM2.5 concentrations and ALRI hospital admissions in children in seven metropolitan cities in Korea. The gray shade indicates the 95% confidence interval. The figure below shows the ALRI hospital admission distribution over PM2.5 concentrations. All models used the 7-day moving average exposure to PM2.5 μg/m3. The percentage change is calculated as the change per PM2.5 10 μg/m3 increase.
Ijerph 18 00144 g002
Ijerph 18 00144 g003 550
Figure 3. Meta-analysis of the association between PM2.5 and ALRI hospital admissions in seven major cities in Korea (Percentage change (%) per 10 μg/m3). All models were used the 7-day moving average exposure to PM2.5 μg/m3. Time-series model adjusted for same covariates (time, AT, day of weeks)
Figure 3. Meta-analysis of the association between PM2.5 and ALRI hospital admissions in seven major cities in Korea (Percentage change (%) per 10 μg/m3). All models were used the 7-day moving average exposure to PM2.5 μg/m3. Time-series model adjusted for same covariates (time, AT, day of weeks)
Ijerph 18 00144 g003
Ijerph 18 00144 g004 550
Figure 4. Spearman’s correlation analysis of modeled PM2.5 and monitored air pollutants PM10, SO2, NO2, CO, and O3) and apparent temperature during the study period. All correlation coefficients were statistically significant (p-value < 0.05). Statistically insignificant correlations are indicated by X-marks.
Figure 4. Spearman’s correlation analysis of modeled PM2.5 and monitored air pollutants PM10, SO2, NO2, CO, and O3) and apparent temperature during the study period. All correlation coefficients were statistically significant (p-value < 0.05). Statistically insignificant correlations are indicated by X-marks.
Ijerph 18 00144 g004
Ijerph 18 00144 g005 550
Figure 5. Percentage change (%) for ALRI hospitalizations following single lag days (lag 0 to lag 7) and moving average (lag 01 to lag 07) PM2.5 exposure in the meta-analysis. Time-series model adjusted for the same covariates (time, AT, day of weeks). The percentage change is calculated as the change per PM2.5 10 μg/m3 increase. The blue line means the border line. The red circle means the highest percentage change.
Figure 5. Percentage change (%) for ALRI hospitalizations following single lag days (lag 0 to lag 7) and moving average (lag 01 to lag 07) PM2.5 exposure in the meta-analysis. Time-series model adjusted for the same covariates (time, AT, day of weeks). The percentage change is calculated as the change per PM2.5 10 μg/m3 increase. The blue line means the border line. The red circle means the highest percentage change.
Ijerph 18 00144 g005
Table
Table 1. Number of acute lower respiratory infection (ALRI) hospitalizations by sex and season for the seven metropolitan cities of Korea from 2008 to 2016. The warm season is April–September and the cold season is January–March, October–December.
Table 1. Number of acute lower respiratory infection (ALRI) hospitalizations by sex and season for the seven metropolitan cities of Korea from 2008 to 2016. The warm season is April–September and the cold season is January–March, October–December.
CategoryTotalSeoulBusanDaeguIncheonGwangjuDaejeonUlsan
Sex
Boys397,33319,85585,099116,53861,45634,35849,81530,212
Girls316,25515,05962,12696,79850,74026,24840,51424,740
Seasons
Warm334,76666,075102,54841,26929,09552,88716,56626,326
Cold378,82281,150110,78849,06031,51159,30918,37828,626
Total713,588147,225213,33690,32960,606112,19634,94454,952
Table
Table 2. Descriptive statistics for daily air pollutants and environmental variables in seven metropolitan cities in Korea, 2008–2016. The warm season is April–September and the cold season is January–March, October–December.
Table 2. Descriptive statistics for daily air pollutants and environmental variables in seven metropolitan cities in Korea, 2008–2016. The warm season is April–September and the cold season is January–March, October–December.
CategoryTotalSeasonCities
WarmColdSeoulBusanDaeguIncheonGwangjuDaejeonUlsan
PM2.5 (μg/m3)29.0 (7.6)24.2 (13.4)30.1 (17.2)30.6 (8.3)25.3 (6.1)26.4 (6.9)29.1 (8.3)24.8 (7.9)28.6 (9.2)25.6 (6.2)
PM10 (μg/m3)46.9 (23.7)43.1 (23.9)50.6 (28.5)47.7 (29.4)47.3 (24.0)47.0 (25.7)52.6 (28.9)43.0 (26.7)43.1 (24.8)47.4 (24.9)
SO2 (ppm)0.005 (0.002)0.005 (0.003)0.006 (0.002)0.005 (0.002)0.006 (0.002)0.004 (0.002)0.007 (0.002)0.004 (0.001)0.004 (0.002)0.008 (0.003)
NO2 (ppm)0.024 (0.009)0.021 (0.009)0.028 (0.011)0.033 (0.012)0.021 (0.007)0.023 (0.011)0.028 (0.012)0.020 (0.008)0.021 (0.009)0.023 (0.008)
CO (ppm)0.500 (0.163)0.419 (0.114)0.581 (0.221)0.544 (0.210)0.398 (0.108)0.481 (0.202)0.588 (0.218)0.501 (0.180)0.492 (0.212)0.494 (0.141)
O3 (ppm)0.025 (0.010)0.031 (0.011)0.019 (0.009)0.021 (0.011)0.028 (0.010)0.025 (0.012)0.024 (0.011)0.027 (0.012)0.024 (0.013)0.026 (0.010)
Mean apparent temperature (°C)13.9 (11.5)22.8 (7.7)4.9 (6.8)12.7 (11.9)15.2 (10.2)14.5 (11.2)12.7 (12.2)14.3 (11.9)13.1 (12.2)14.5 (10.7)
Mean temperature (°C)13.9 (9.7)21.4 (5.2)6.3 (6.9)12.8 (10.8)15.1 (8.3)14.7 (9.7)12.6 (10.1)14.4 (9.6)13.2 (10.2)14.4 (8.8)
Mean relative humidity (%)64.6 (16.7)69.7 (15.3)59.5 (16.4)60.2 (14.9)62.4 (18.3)57.8 (16.9)71.7 (16.2)67.5 (13.7)67.8 (14.4)64.9 (17.6)
Mean dew point (°C)6.5 (11.6)14.9 (7.3)−1.9 (8.7)4.5 (12.1)7.2 (11.6)5.3 (11.8)7.1 (11.8)7.7 (10.7)6.6 (11.4)7.1 (11.6)
Table
Table 3. Percentage change in PM2.5 single pollutant model and two-pollutant model. The percentage change is calculated as the change per PM2.5 10 μg/m3 increase. All models used the 7-day moving average exposure to PM2.5 μg/m3. The time-series model was adjusted for the same covariates (time, AT, day of weeks).
Table 3. Percentage change in PM2.5 single pollutant model and two-pollutant model. The percentage change is calculated as the change per PM2.5 10 μg/m3 increase. All models used the 7-day moving average exposure to PM2.5 μg/m3. The time-series model was adjusted for the same covariates (time, AT, day of weeks).
CitiesSingle Pollutant (95% C.I)Two-Pollutant (95% C.I)
PM2.5PM2.5 + SO2PM2.5 + NO2PM2.5 + O3
Seoul1.60 (1.00, 2.20)1.58 (0.80, 2.37)1.84 (1.09, 2.60)1.56 (0.96, 2.16)
Busan0.81 (0.17, 1.45)−0.96 (−1.71, −0.19)0.31 (−0.48, 1.10)0.94 (0.28, 1.60)
Daegu1.64 (0.70, 2.58)1.60 (0.45, 2.77)1.15 (0.00. 2.32)1.55 (0.60, 2.50)
Incheon0.57 (−0.45, 1.61)−0.98 (−2.26. 0.32)0.62 (−0.61, 1.87)0.45 (−0.59, 1.50)
Gwangju0.28 (−0.51, 1.08)−0.09 (−1.04, 0.86)0.14 (−0.79, 1.08)−0.01 (−0.84, 0.83)
Daejeon2.09 (0.83, 3.37)2.70 (1.09, 4.34)3.01 (1.55, 4.49)1.74 (0.44, 3.05)
Ulsan1.92 (0.71, 3.15)1.45 (0.00, 2.92)1.40 (−0.16, 2.99)1.91 (0.66, 3.17)
Total1.20 (0.71, 1.71)0.69 (−0.34, 1.74)1.14 (0.43, 1.85)1.11 (0.60, 1.64)
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Oh, J.; Han, C.; Lee, D.-W.; Jang, Y.; Choi, Y.-J.; Bae, H.J.; Kim, S.; Ha, E.; Hong, Y.-C.; Lim, Y.-H. Short-Term Exposure to Fine Particulate Matter and Hospitalizations for Acute Lower Respiratory Infection in Korean Children: A Time-Series Study in Seven Metropolitan Cities. Int. J. Environ. Res. Public Health 2021, 18, 144. https://doi.org/10.3390/ijerph18010144

AMA Style

Oh J, Han C, Lee D-W, Jang Y, Choi Y-J, Bae HJ, Kim S, Ha E, Hong Y-C, Lim Y-H. Short-Term Exposure to Fine Particulate Matter and Hospitalizations for Acute Lower Respiratory Infection in Korean Children: A Time-Series Study in Seven Metropolitan Cities. International Journal of Environmental Research and Public Health. 2021; 18(1):144. https://doi.org/10.3390/ijerph18010144

Chicago/Turabian Style

Oh, Jongmin, Changwoo Han, Dong-Wook Lee, Yoonyoung Jang, Yoon-Jung Choi, Hyun Joo Bae, Soontae Kim, Eunhee Ha, Yun-Chul Hong, and Youn-Hee Lim. 2021. "Short-Term Exposure to Fine Particulate Matter and Hospitalizations for Acute Lower Respiratory Infection in Korean Children: A Time-Series Study in Seven Metropolitan Cities" International Journal of Environmental Research and Public Health 18, no. 1: 144. https://doi.org/10.3390/ijerph18010144

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop