Effects of Environmental Air Pollution on Pulmonary Function Level of Residents in Korean Industrial Complexes

This study aims to identify environmental air pollution adversely affecting pulmonary function among a community-based general population living in Korean industrial complexes. A total of 1963 residents participated in a pulmonary function test (PFT). The sample population consisted of an exposed group (n = 1487) living within a radius of 5 km of industrial complexes and a control group (n = 476) living over a radius of 10 km from the industrial complexes in Gwangyang and Yeosu cities. PFT results were calculated for each resident of the study population. On-site questionnaire surveys with face-to-face interviews were also conducted to collect more detailed information on personal lifestyles, medical history, exposure to air pollution, and respiratory disease and related symptoms. A total of 486 measured samples were collected by eight automated air-monitoring stations installed in four counties of Gwangyang and four counties of Yeosu in South Korea from January 2006 to February 2007. Mean levels of SO2 (0.012 ppm), CO (0.648 ppm), NO2 (0.02 ppm), O3 (0.034 ppm), and PM10 (43.07 μg/m3), collected within a radius of 5 km, were significantly higher than those collected over a radius of 10 km from Gwangyang and Yeosu industrial complexes. Prevalence odds ratio (OR) of abnormal pulmonary function in the exposed group of residents (<5 km) was elevated at 1.24 (95% CI 0.71–1.96), but not statistically significant (p > 0.05). In multiple linear regression analysis, forced expiratory volume in one second (FEV1) and forced vital capacity (FVC) levels significantly declined as SO2, CO, and O3 levels increased when adjusting for age, sex, body mass index (BMI), alcohol, smoking, secondhand smoke, and respiratory disease and related symptoms (n = 1963) (p < 0.05). These results suggest that exposure to air pollution affects pulmonary function levels of residents living in Korean industrial complexes.


Introduction
Since the 1970s, a number of large-sized industrial complexes of petrochemical and steel industries were constructed and contributed to the fast economic growth in South Korea [1]. The industrial identifying associations between exposure to air pollution and adverse effects on pulmonary function of residents living in Korean industrial complexes. Therefore, we conducted PFTs and on-site questionnaire surveys with face-to-face interviews and collected more detailed information on general characteristics of personal lifestyles, medical history, respiratory disease, allergic symptoms, and related environmental factors in a large community-based population consisting of individuals either (1) living within a radius of 5 km of Gwangyang and Yeosu industrial complexes or (2) living over 10 km from Gwangyang and Yeosu industrial complexes in Korea. The objective of this study was to compare pulmonary function levels (FEV 1 , FVC, and FEV 1 /FVC) between the exposed group (≤5 km) and control group (>10 km) of the study population, to observe which environmental factor or air pollutant could adversely affect the lung function, and to determine the relationship between exposure to environmental air pollution and adverse effects on lung function among residents living in Korean industrial complexes.

Study Subject
This study was conducted by performing both PFTs and on-site questionnaire surveys for a total of 1963 residents from the general population who registered and lived in Gwangyang (n = 974) and Yeosu (n = 989) industrial complexes of Gwangyang Bay area in Korea, from May to December in 2007. We recruited a community-based population of people (ages from 5 to 70 years old) who had lived over 1 year in Gwangyang or Yeosu. All study subjects were contacted by MoE via email or phone call, and MoE provided information in advertisements on an MoE official website in 2007 ( Figure 1). We enrolled up to 1000 persons who expressed interest in participating in the study. Consent was obtained from the study population by obtaining fully signed agreements to participate in PFTs, answer questionnaire surveys, and undergo medical examination, and for the study to make use of their personal information until the end of the study period. A total of 1963 participants were classified into exposed or control groups, depending on the distance of residence from the industrial complexes. Residents classified as exposed lived within a radius of 5 km from Gwangyang (n = 674) or Yeosu (n = 813) industrial complexes. The rest of residents were classified as a control group, who lived over a radius of 10 km from Gwangyang (n = 300) or Yeosu (n = 176) industrial complexes, respectively. Figure 2 shows the map of the study protocol. This study was approved by the Institutional Review Board (IRB) of Soonchunhyang University. but predictive estimates among the target population of interest. This is because PFT results vary by age, sex, health condition, and ethnicity [25][26][27]. Despite the limitations of PFT results, the test is still useful for identifying associations between exposure to air pollution and adverse effects on pulmonary function of residents living in Korean industrial complexes. Therefore, we conducted PFTs and on-site questionnaire surveys with face-to-face interviews and collected more detailed information on general characteristics of personal lifestyles, medical history, respiratory disease, allergic symptoms, and related environmental factors in a large community-based population consisting of individuals either (1) living within a radius of 5 km of Gwangyang and Yeosu industrial complexes or (2) living over 10 km from Gwangyang and Yeosu industrial complexes in Korea. The objective of this study was to compare pulmonary function levels (FEV1, FVC, and FEV1/FVC) between the exposed group (≤5 km) and control group (>10 km) of the study population, to observe which environmental factor or air pollutant could adversely affect the lung function, and to determine the relationship between exposure to environmental air pollution and adverse effects on lung function among residents living in Korean industrial complexes.

Study Subject
This study was conducted by performing both PFTs and on-site questionnaire surveys for a total of 1963 residents from the general population who registered and lived in Gwangyang (n = 974) and Yeosu (n = 989) industrial complexes of Gwangyang Bay area in Korea, from May to December in 2007. We recruited a community-based population of people (ages from 5 to 70 years old) who had lived over 1 year in Gwangyang or Yeosu. All study subjects were contacted by MoE via email or phone call, and MoE provided information in advertisements on an MoE official website in 2007 ( Figure 1). We enrolled up to 1000 persons who expressed interest in participating in the study. Consent was obtained from the study population by obtaining fully signed agreements to participate in PFTs, answer questionnaire surveys, and undergo medical examination, and for the study to make use of their personal information until the end of the study period. A total of 1963 participants were classified into exposed or control groups, depending on the distance of residence from the industrial complexes. Residents classified as exposed lived within a radius of 5 km from Gwangyang (n = 674) or Yeosu (n = 813) industrial complexes. The rest of residents were classified as a control group, who lived over a radius of 10 km from Gwangyang (n = 300) or Yeosu (n = 176) industrial complexes, respectively. Figure 2 shows the map of the study protocol. This study was approved by the Institutional Review Board (IRB) of Soonchunhyang University.

Questionnaire Survey
The study population (n = 1963) participated in the on-site questionnaire survey and provided their personal information (birth of data, sex, home address, etc.), socioeconomic status (education, occupation, income, etc.), personal lifestyle (smoking, alcohol consumption, exercise, etc.), residence (exposure to air pollution), nutrition (diet), time-activity pattern (daily, weekly, monthly), mental health (stress, attitude, personality, etc.), and medical history (respiratory disease, allergic symptoms, etc.). This survey was conducted during the summer season from July to August in the year 2007. The survey process was as follows. Several hard copies of questionnaires were provided to all participants at the meeting rooms at their local hospitals located in Gwangyang and Yeosu industrial complexes and several professional interviewers who completed preliminary training programs developed by Ministry of Environment (MoE) arrived at the meeting rooms. They asked some questions in response to answers the respondents gave regarding their personal opinions to gather additional information necessary to complete the on-site questionnaire surveys. The process took approximately 1 h to per each group of participants (about 7-10 subjects), including face-to-face interviews.

Pulmonary Function Test
Pulmonary function tests (PFTs) were performed on the entire study population using spirometers (Spirovit model SP1, Schiller, Ottobrunn, Germany) at national hospitals in both Gwangyang and Yeosu. Prior to PFTs, nurses who completed professional PFT programs offered by the Korean Occupational Health & Safety Agency (KOSHA) assisted in providing guidance for using the spirometers. The PFT results were calculated as forced expiratory volume in one second (FEV1), forced vital capacity (FVC), and FEV1/FVC after adjusting for age, sex, height, and weight from the Korean standard population [25]. The FVC-which determines the vital capacity of maximum respiratory function-was defined as an indicator of restrictive ventilatory impairment. FEV1 was defined as the volume which has been exhaled at the end of the first forced expiration and was used to evaluate obstructive ventilatory impairment. Finally, FEV1/FVC ratios were calculated to

Questionnaire Survey
The study population (n = 1963) participated in the on-site questionnaire survey and provided their personal information (birth of data, sex, home address, etc.), socioeconomic status (education, occupation, income, etc.), personal lifestyle (smoking, alcohol consumption, exercise, etc.), residence (exposure to air pollution), nutrition (diet), time-activity pattern (daily, weekly, monthly), mental health (stress, attitude, personality, etc.), and medical history (respiratory disease, allergic symptoms, etc.). This survey was conducted during the summer season from July to August in the year 2007. The survey process was as follows. Several hard copies of questionnaires were provided to all participants at the meeting rooms at their local hospitals located in Gwangyang and Yeosu industrial complexes and several professional interviewers who completed preliminary training programs developed by Ministry of Environment (MoE) arrived at the meeting rooms. They asked some questions in response to answers the respondents gave regarding their personal opinions to gather additional information necessary to complete the on-site questionnaire surveys. The process took approximately 1 h to per each group of participants (about 7-10 subjects), including face-to-face interviews.

Pulmonary Function Test
Pulmonary function tests (PFTs) were performed on the entire study population using spirometers (Spirovit model SP1, Schiller, Ottobrunn, Germany) at national hospitals in both Gwangyang and Yeosu. Prior to PFTs, nurses who completed professional PFT programs offered by the Korean Occupational Health & Safety Agency (KOSHA) assisted in providing guidance for using the spirometers. The PFT results were calculated as forced expiratory volume in one second (FEV 1 ), forced vital capacity (FVC), and FEV 1 /FVC after adjusting for age, sex, height, and weight from the Korean standard population [25]. The FVC-which determines the vital capacity of maximum respiratory function-was defined as an indicator of restrictive ventilatory impairment. FEV 1 was defined as the volume which has been exhaled at the end of the first forced expiration and was used to evaluate obstructive ventilatory impairment. Finally, FEV 1 /FVC ratios were calculated to determine the category of pulmonary function. If the ratio was less than 0.70 (a golden standard [28]), then the predicted values of FEV 1 and FVC were lower than 80% and obstructive lung disease (OLD) was diagnosed [29].

Collection of Air-Monitoring Data
A large number of air-monitoring datasets were collected by automated monitoring stations installed in four counties of Jungma-dong, Geumho-dong, Bonggang-myeon, and Okryong-myeon in Gwangyang city as well as four counties of Samil-dong, Wolnae-dong, Gwangmu-dong, and Hwayang-myeon in Yeosu city of Jeollanam-do in South Korea from January 2006 up to February 2007. A total of 486 measurement samples consisted of five pollutants: sulfur dioxide (SO 2 ) (n = 97), carbon monoxide (CO) (n = 97), nitrogen dioxide (NO 2 ) (n = 98), ozone (O 3 ) (n = 97), and particulate matter (PM 10 ) (n = 97). Legal environmental limits of all air pollutants are shown in Table 1. The values of arithmetic mean (AM) and standard deviation (SD) of each pollutant were statistically analyzed and compared for each area (Gwangyang and Yeosu). All collected monitoring data were used for statistical analysis in this study.

Statistical Analysis
All statistical analyses were performed using SAS 9.4 software package (SAS Inc., Cary, NC, USA) after completing data translation. Descriptive statistics were calculated for results of PFTs, exposure to ambient air pollutants, and on-site questionnaire surveys. Prevalence odds ratio (OR) with 95% confidence intervals (Cis) were also estimated to observe the effects of various environmental factors-including distance from industrial complexes, personal lifestyles (alcohol and smoking), respiratory disease and allergic symptoms-on pulmonary function among residents using a logistic regression model. Multiple regression analysis was also performed to identify how environmental factors (independent) significantly affected pulmonary function (dependent) among residents living near these industrial complexes after adjusting for age, sex, BMI, alcohol, smoking, and respiratory disease and related symptoms (e.g., asthma, wheezing, exercising wheezing, nasal obstruction, allergic rhinitis). As for multicollinearity, a collinearity analysis was performed to observe if there was any correlation between independent variables. Collinearity statistics (i.e., variance inflation factors, VIFs) were also calculated in multiple regression analysis. The null hypothesis was rejected in all statistical analyses if a p-value was less than the significance level (p < 0.05).  Table 2 shows general characteristics of the individuals (from a community-based population) who participated in this study. Of 1963 subjects, female (n = 1164) residents accounted for about 60%, while school children (n = 639) aged from 10 to 19 years and the elderly (n = 599) aged over 60 years accounted for 32.6% and 30.5%, respectively. Non-response rate to the on-site questionnaire survey was 22.5% for secondhand smoke, 18% for smoking, and 10.3% for alcohol consumption. Table 3 shows summary statistics of pulmonary function levels for all population groups in Gwangyang and Yeosu industrial complexes based on sex, age, distance from industrial complexes, body mass index (BMI), alcohol consumption, smoking, and secondhand smoke exposure. Overall, arithmetic means (Ams) and standard deviations (SDs) calculated for PFT results were 101.10 ± 23.30 for FEV 1 , 98.49 ± 19.42 for FVC, and 81.27 ± 9.27 for the ratio FEV 1 /FVC in Gwangyang, and 107.08 ± 23.68 for FEV 1 , 100.74 ± 19.27 for FVC, and 83.26 ± 9.62 for the ratio FEV 1 /FVC in Yeosu city. There was also a significant difference in risks to pulmonary function levels between exposed and control groups in both cities combined. All PFT values significantly decreased by age and personal lifestyles. Ratios of FEV 1 /FVC of nonsmokers were significantly higher than those of smokers in both cities (p < 0.05). Table 4 also shows summary statistics for levels of pulmonary functions (FEV 1 /FVC) in all groups of the study population (exposed, control, and total) classified by age. Pulmonary function levels in the exposed group were significantly lower than those in the control group in both areas (p < 0.05). In Table 5, prevalence odds ratios (ORs) for abnormal pulmonary function in the exposed group were also elevated by some variables (including the distance from industrial complexes, distance from streets with automobiles, and exposure to secondhand smoke), but not statistically significant (p > 0.05). ORs were 1.24 (95% CI 0.71-1.96) for distance from industrial complexes (either ≤5 km or >10 km), 1.17 (95% CI 0.63-1.84) for distance from streets, 1.02 (95% CI 0.62-1.61) for alcohol, 0.92 (95% CI 0.49-1.72) for smoking, and 1.12 (95% CI 0.73-1.71) for secondhand smoke.

Results
Levels of five air pollutants-SO 2 (n = 97), CO (n = 97), NO 2 (n = 98), O 3 (n = 97), and PM 10 (n = 97)-were automatically measured by eight fixed air-monitoring stations located in the counties of Gwangyang and Yeosu cities. The air monitoring data are summarized in Table 1. Overall, all levels of air pollutants (SO 2 , CO, NO 2 , O 3 , and PM 10 ), which were regularly measured every month from January 2006 to February 2007, in the exposed area (≤5 km) were significantly higher than those in the control area (>10 km) (p < 0.05). All concentration levels of the five air pollutants were below the environmental limits established by MoE in 2007. Among these pollutants, CO showed the highest mean level. Mean levels of two pollutants (SO 2 and O 3 ) were significantly different between exposed and control areas in both cities (p < 0.05). Table 6 shows results from multiple linear regression analysis, which was performed to observe if any pollutant was significantly associated with pulmonary function levels in each area. Pulmonary function of residents in both cities significantly declined as SO 2 , CO, and O 3 levels increased (p < 0.05). Two pollutants (NO 2 and PM 10 ) were not related to pulmonary function levels in either area. No multicollinearity was observed between any variables in this regression model. All VIFs were less than 10 (data not shown, p < 0.05). Table 7 shows that the increase of four pollutants (SO 2 , CO, O 3 , and PM 10 ) was significantly associated with the decline in pulmonary function levels (FEV 1 , FVC, FEV 1 /FVC) in the exposed group (≤5 km) (adjusted r 2 = 0.12) (p < 0.05). However, only two pollutants (SO 2 and O 3 ) were significantly related to the decline in lung function (FEV 1 and FVC) in the control group, although the regression model for FEV 1 /FVC was not statistically significant. Most importantly, increased levels of O 3 resulted in the highest decline in pulmonary function levels among the exposed group (population living within a radius of 5 km from Gwangyang and Yeosu industrial complexes in Korea).       * Statistically significant (p < 0.05); the regression model was adjusted for age, sex, BMI, smoking, secondhand smoke, alcohol, and respiratory disease and related symptoms (e.g., asthma, wheezing, exercising wheezing, nasal obstruction, allergic rhinitis); no multicollinearity for independent variables was observed (all variance inflation factors (VIFs) < 10).

Discussion
This study used pulmonary function tests with on-site questionnaire surveys to identify if environmental air pollution adversely affects pulmonary function levels in a community-based general population (n = 1963) living within 5 km (exposed) or over 10 km (control) from Gwangyang and Yeosu industrial complex areas located in South Korea. We observed that increased levels of SO 2 and O 3 were significantly associated with a decline in pulmonary function levels of residents living in the exposed area near industrial complexes. The Ams and SDs of PFTs in the exposed group were 99.17 ± 22.29 for FEV 1 and 96.98 ± 18.86 for FVC in Gwangyang and 107.05 ± 23.52 for FEV 1 and 100.23 ± 19.46 for FVC in Yeosu, respectively. Pulmonary function levels in the exposed group were significantly lower than those in the control group in all areas (p < 0.05).
In a multiple regression model, increased levels of SO 2 and O 3 were significantly associated with the decline in pulmonary function levels of residents in both Gwangyang and Yeosu cities when adjusting for age, sex, BMI, alcohol, smoking, secondhand smoke, and respiratory disease and related symptoms (adjusted r 2 = 0.12). Our study results indicated that inhalation exposure to environmental air pollution, especially for SO 2 and O 3 , was significantly associated with lung function problems in the general population living near Gwangyang and Yeosu industrial complexes. The increased level of ozone showed the biggest impact, causing a sharp decline in pulmonary function level in the susceptible population living within 5 km from Gwangyang industrial complex, indicating that environmental air pollution played a major role in abnormal pulmonary function.
Several similar studies have been previously reported over the years in Korea since the 1990s. Lee et al. conducted a study identifying that an environment exposed to SO 2 , O 3, and total suspended particulates (TSP) significantly increased mortality in the general population in Ulsan in Korea. The increased level of 100 µg/m 3 of TSP was significantly related to elevated mortality ratios in the Seoul metropolitan city after adjusting for daily and seasonal variations in the exposure monitoring datasets [30]. In 2002, Kwon et al. reported that Asian dust events (CO, NO 2 , SO 2 , PM 10 , and O 3 ) were significantly related to elevated risk of death from cardiovascular and respiratory diseases among the general population in Seoul during the period of 1995-1998 [31]. Son et al. conducted a study that showed an association between environmental air pollutants-which included CO, NO 2 , SO 2 , PM, and O 3 -with adverse effects on lungs using a large-scale population (n = 3827) recruited from Ulsan industrial complexes in Korea during the period of 2003-2007. The authors found that FVC levels were associated with all types of air pollutants. In particular, pulmonary functions declined by 6.1% for FVC and 0.5% for FEV 1 as O 3 level increased by 11 ppb [32].
Kim et al. conducted a cross-sectional study identifying an association between environmental air pollution and allergic diseases with a total of 4545 students living near industrial complexes in Korea. They also found that increased levels of black carbon, SO 2 , and O 3 emitted from industrial complexes were significantly related to elevated ORs for allergic rhinitis and atopic dermatitis [33]. Smargiassi et al. investigated acute cardiorespiratory effects of air pollution (including SO 2 , NO 2 , PM 2.5 , benzene, and PAHs) for children aged 7-12 years (n = 72) with asthma living in proximity to an industrial complex area in Montreal, Canada. They found that a low level of personal exposure to benzene and PAHs was associated with decreased pulmonary function among children living in industrial complex areas [34]. In 2016, Wong et al. also conducted a cross-sectional study of long-term exposure to ambient air pollution, finding that it adversely affected children's lung functioning in Canada from 2007 to 2011.
The authors found that industrial emission of PM 2.5 was negatively associated with decreased lung functioning (FEV 1 and FEV 1 /FVC) among young males, contrary to females [35]. An increase of 190 tons of industrial air emissions within 25 km of residence was associated with a 1% reduction in FEV 1 , and an increase of 370 tons was related to a 1% reduction in FEV 1 /FVC. Kurt et al. reviewed a total of 53 publications and observed how environmental air pollution is associated with health concerns related to pulmonary mobility and mortality worldwide. They suggested that air pollution can be an important factor when worsening lung diseases in susceptible populations, including children, pregnant women, and elderly people [36]. The environmental exposure to NO X , PM, and O 3 was associated with elevated risk of asthma, chronic obstructive pulmonary disease (COPD), lung cancer, allergic symptoms, and respiratory infections at local, national, and global level [37]. PM and O 3 were shown to be the most widespread health threats related to cardiopulmonary diseases. With regards to association between low-level exposure to ozone and pulmonary function, our study results showed great consistency with previous studies in Korea and other countries [38].
Our study has several strengths. First, we recruited a total of 1963 residents from a large-scale community-based population, aged between 5 and 70 years old, living in Gwangyang and Yeosu cities. All study participants, recruited from both Gwangyang and Yeosu cities, voluntarily participated in PFTs to measure their lung function levels and responded to questionnaire surveys regarding their personal lifestyles, distance from the industrial complexes, and respiratory disease and related symptoms. Secondly, the large sample population included susceptible individuals, including school children and elderly people. Approximately 70% of the study population consisted of school children and elderly people, while the rest of the population (30%) were young, aged from 20 to 59 years old. This is possibly due to a younger population working at both industrial complexes. Lastly, our study significantly identified an association between low-level exposure to air pollutants (SO 2 , CO, O 3 ) and the decline in pulmonary function levels among the general and susceptible populations living in two Korean industrial complex areas.
Despite its strengths, this study had some limitations. The amount of air-monitoring data was relatively small without analyzing daily or seasonal variations, indicating that these collected datasets were not representative of true exposure to environmental air pollution in both industrial and residential areas. We only obtained a small number of air-monitoring data (n = 486) automatically collected by fixed monitoring stations located in eight counties in both Gwangyang and Yeosu cities. We did not collect personal monitoring data on daily, weekly, monthly, or seasonal bases. Furthermore, the average duration of each study participant's (exposed or control group) outside daily activities was unknown because we did not collect detailed information on their daily activities or personal lifestyles (exposure misclassification). The exposed and control groups of the study population were only classified based on the distance of their registered home address from industrial complexes, as indicated on the self-administered questionnaire surveys and copies of written documents they submitted. Furthermore, we could not recruit enough numbers from the young population (aged from 20 to 39 years old). The young population might be working as industrial manufacturers or maintenance workers employed by either petrochemical or steel manufacturing plants located in both Gwangyang and Yeosu cities. Most importantly, we did not know where those five pollutants came from: we had no specific information as to whether these five pollutants were emitted from public or private transformation or industrial complex areas. Because we collected these five pollutants (SO 2 , CO, NO 2 , O 3 , and PM 10 ) from MoE's official website, we assumed that air pollution was mostly emitted from the target industrial complexes. Due to these limitations, further studies are needed to identify the relationship between environmental air pollution emitted from industrial complex areas in Gwangyang and Yeosu cites and adverse health effects on the most susceptible populations (school children and elderly) in the future.
Our study results might have shown no association between exposure to environmental air pollution and pulmonary function level in the community-based general population if more individuals from the young population group were included in this study. That is, the young population working in the industrial complexes must be much healthier than school children or elderly populations, particularly for lung function. They could show higher pulmonary function levels, although they were more likely to be exposed to higher levels of air pollutants or other hazardous chemicals emitted from their workplaces. Further studies will demonstrate if long-term exposure to environmental air pollutants significantly affects to pulmonary function levels in all age groups of the general and susceptible populations, including the young population, and these studies will also allow to identify which environmental factor adversely affects their respiratory and cardiovascular systems by investigating how these factors could result in the decline in lung function and other related symptoms among these residents. Finally, a new longitudinal study is also needed to determine how many cases would increase after exposure to low levels of environmental air pollution in the general and susceptible populations in Korea.

Conclusions
We found that there was an association between increased levels of exposure to ambient air pollutants (SO 2 , O 3 , and CO) and decline in pulmonary function levels (FEV 1 and FVC) among residents living near Gwangyang and Yeosu industrial complexes in Korea. Risks for abnormal pulmonary function among residents were not statistically significant by independent variable. In this study, we found that increased levels of SO 2 and O 3 were significantly associated with the decline in pulmonary function levels of residents living in the exposed area near Korean industrial complexes.