Respiratory and Irritant Health Effects of a Population Living in a
Petrochemical-Polluted Area in Taiwan
Chun-Yuh Yang,* Jung-Der Wang,† Chang-Chuan Chan,† Pao-Chung Chen,† Jing-Shiang Huang,‡ and Ming-Fen Cheng§
*School of Public Health, Kaohsiung Medical College, Kaohsiung, Taiwan; †Institute of Occupational Medicine and Industrial Hygiene, College of Public Health, National Taiwan University, Taipei, Taiwan; ‡Institute of Statistical Science, Academia
Sinica, Taipei, Taiwan; and §School of Public Health, Kaohsiung Medical College, Kaohsiung, Taiwan Received January 6, 1997
Reported herewith are the results from an ongo-ing study of outdoor air pollution and the health of persons living in the communities in close proxim-ity to petrochemical industrial complexes. To deter-mine if there is an excess of adverse health out-comes in the population exposed to petrochemical industrial emissions, a health survey was under-taken in 1996 in this area and in one reference area which has no local industrial emissions. The sub-jects were 436 adults (30–64 years of age) living in the Sanwei area (exposed area) and 488 in Taicei (reference area). For several indicators of respira-tory health, including cough, wheezing, and chronic bronchitis, the prevalence rates were not signifi-cantly different between the study and the control populations. Acute irritative symptoms (eye irrita-tion, nausea, throat irritairrita-tion, and chemical odor perception) were significantly more common in the exposed area, particularly perception of chemical odors (84.6% vs 2.1%). It is concluded that exposure to petrochemical air emissions may be associated with increased rates of acute irritative symptoms. Future studies are needed to identify the potential role of petrochemical industrial emissions (particu-larly volatile organic compounds) in the genesis of acute irritative symptoms in a nearby petrochemi-cal industrial area. © 1997 Academic Press
INTRODUCTION
Development of the basic petrochemical industry in Taiwan started some 25 years ago in 1968. Ac-cording to the pollution reports compiled by the En-vironmental Protection Administration of Taiwan, the Republic of China, since 1983, nine serious air pollution episodes have occurred in petrochemical counties between 1971 and 1990 (EPA/ROC, 1984– 1992). The pollutants emitted by these
petrochemi-cal industries included not only the vinyl chloride monomer and polycyclic aromatic hydrocarbons (PAHs) which have been recognized as environmen-tal carcinogens but also large quantities of criteria pollutants such as sulfur dioxide (SO2), nitrogen
di-oxide (NO2), and particulate matter (PM10). An
ex-cess rate for liver cancer among males has been noted among those living in petrochemical indus-trial counties (PIC) (Yang et al., 1997), while the respiratory and irritant health effects of petrochemi-cal industry air emissions on surrounding commu-nities have not been well studied.
Petrochemical industries are large, usually iso-lated, sources of air pollution. Taiwan has 19 such petrochemical industrial complexes on the periphery of small communities (Yang et al., 1997). The small communities in the southern Taiwan in close prox-imity to petrochemical industrial complexes provide unique opportunities to evaluate the health effects of a single source of pollution. The petrochemical industrial complexes are the only large pollution sources in these communities with little air pollu-tion from motor vehicle activity.
The objective of the present study was to deter-mine whether there was an excess of adverse respi-ratory and irritant health outcomes in population who live close to petrochemical manufacturing facili-ties. This report characterizes respiratory and irri-tant symptoms of population who lived in Sanwei (petrochemical exposed area) and in one reference area (Taicei) that was not in the proximity of petro-chemical or other industrial emissions.
MATERIALS AND METHODS Design
A cross-sectional survey was designed to deter-mine the prevalence of respiratory and irritative
ARTICLE NO. ER973762
145
0013-9351/97 $25.00 Copyright © 1997 by Academic Press All rights of reproduction in any form reserved.
symptoms in the population living near the petro-chemical works (exposed area) compared with a con-trol area in Taiwan.
Defining the Study Population
Lin-Yuan county is one of the largest petrochemi-cal manufacturing centers in Taiwan. A prior hy-pothesis was that exposure to petrochemical indus-try emissions would be inversely related to the dis-tance of primary schools from the petrochemical industrial complexes. The elementary school was chosen as a surrogate for residential location be-cause elementary schools in Taiwan are community-based and serve very small areas (usually of less than a 2-km radius). The exposed area analyzed here consisted of three communities which are lo-cated in close proximity to the elementary school (Sanwei primary school) which has the shortest dis-tance between the school and the geographic center of major petrochemical industry complexes in Lin-Yuan. The exposed population consisted of all resi-dents aged 30–64 years living in the exposed area. Hence, the exposed population resided in close prox-imity to the major petrochemical manufacturing in-dustries in Lin-Yuan and with intense air pollution caused by the petrochemical emissions.
The control area were at Taicei county which was a relatively unpolluted area, about 120 km north of Lin-Yuan. This area is basically free of local sources of industrial air pollution and is felt to be remote from the influence of any significant source of air pollution by site visits. However, the Taiwan Plastic Group (the biggest petrochemical industry company in Taiwan) has announced plans to invest in the Taicei area, creating an another petrochemical manufacturing center. This announcement gives rise to an opportunity to compare the health effects of a new environmental air pollution setting. When the present survey was carried out, baseline health and environment data were collected to provide a framework for a prospective epidemiological moni-toring program in the near future, but in this study, the data were collected as a reference basis in order to evaluate whether or not there is an increased prevalence of respiratory and irritant symptoms in a petrochemical industrial area. The control area ana-lyzed in this study consisted of three communities which are located in close proximity to an elemen-tary school (Taicei primary school) and where an air quality monitoring station has been established by the Environmental Protection Administration (EPA), a central governmental agency. The control population consisted of all residents aged 30–64 years living in close proximity to the primary school.
Questionnaire
A modified version of the American Thoracic So-ciety Questionnaire (adult version) (Ferris, 1978) was administered to all of the study subjects by trained interviewers. For this study, some questions were added concerning acute irritative symptoms that had the potential to cause irritation. Interview-ers had been trained prior to the study through lec-ture sessions as well as practice administration of the questionnaire. Each interviewer worked in both the exposed and control areas. The survey, including the administration of questionnaires, was carried out between April 1996 to June 1996.
The health questionnaire inquired about chronic respiratory symptoms such as coughing, phlegm production, wheezing, shortness of breath, and chronic bronchitis. The chronic respiratory symp-toms concerned were defined as follows: chronic cough, coughing part of the day or the entire day for at least 3 months per year; phlegm production, phlegm production during a part of a day or for an entire day for at least 3 months per year; wheezing, a condition of causing a wheezy or whistling sound on inspiration at least occasionally, apart from that caused by a cold or acute upper respiratory infection; chronic bronchitis, a cough and/or phlegm on most days for 3 months or more out of a year; dyspnea, having to stop for breath when walking at one’s own pace on level ground.
A second set of questions focused on acute eye, nose, and throat irritation, nausea symptoms, and chemical odor perception potentially effected by ex-posure to irritants. These questions concerned the occurrence of symptoms during the 1-month period prior to the date on which the questionnaire was completed.
Air Pollution Measurement
Air pollution measurements were carried out by the EPA air quality monitoring stations at each of the two primary schools. The monitoring stations were fully automatic. The air pollution data (SO2, NO2, and PM10) were retrieved from the air quality monitoring data banks of the EPA for the whole year 1995. SO2was measured by means of a UV
fluores-cence instrument, NO2 by a chemiluminescent
ap-paratus, and PM10by a beta attenuation method.
Statistics
The prevalence of respiratory symptoms according to place of residence were analyzed by means of the
x2
variables. To explore the relations between respira-tory symptoms and the exposure areas, multiple lo-gistic regression models (Hosmer and Lemeshow, 1989) were used in which the potential confounding factors were controlled. The adjusted odds ratios and their 95% confidence intervals were computed. All statistical analyses were performed using the Statistical Analysis System.
RESULTS
There were 2036 and 1976 residents 30–64 years of age who were identified from the electorate lists of 1995 for the exposed and control area, respectively. Subjects were numbered and used for random sam-pling in each area. A random sample (one-third of the population) for the exposed and control area numbered 679 and 659, respectively. The response rates were higher in control area—64.2% (436/679) of the representative sample in the exposed area and 74.1% (488/659) in the control area. In both areas, the respondents were questioned by door-to-door in-terviews.
As shown in Table 1, subjects in the exposed area and the control area were similar in mean age, dis-tribution of current smoking habits, and years of residence at the current address but varied in edu-cation (lower in exposed area) and sex (more male subjects in exposed area). The subjects living in the exposed area reported a higher rate of occupation exposure to dust/fumes than the subjects living in the control area (21.1% vs 10.3%).
The subjects living in exposed area had higher rates of selected respiratory symptoms (cough, phlegm, wheezing, and chronic bronchitis) but had a lower rate of dyspnea. The differences in phlegm production and dyspnea were significant (P < 0.05). There were significant differences of acute irritative symptoms (eye irritation, nausea, and chemical odor perception) between the exposed and the control area (Table 2).
A number of potential confounding factors which could have influenced an association between air pollution and respiratory symptoms and acute irri-tative symptoms were examined in this study. These included age, sex, educational level, smoking status, length of residence in the respondent’s current home, and self-reported occupational exposures. The results of the multiple logistic regression analyses are shown in Table 3. All models controlled for the potential confounders mentioned above. The ad-justed odds ratios were mostly greater than the cor-responding crude odds ratios, but were similar to the crude odds ratios. Phlegm production remained sig-nificantly more prevalent in the exposed area. Rates of acute irritative symptoms showed an association with the living area except for acute nose symptoms. The odds ratios calculated for acute irritative symp-toms in exposed area are between 1.52 for throat symptom and 403.43 for chemical perception, as compared with 1.0 for control subjects.
The monitoring station in the exposed area
TABLE 1
Demographic Characteristics of Study Subjects from Sanwei (Exposed Area) and Taicei (Control Area)
Demographic characteristics Taicei (n4 488) Sanwei (n4 436) Mean age in years (SD)a 52.4 (14.4) 49.4 (10.7)
Sex (%)
Male 43.2 52.3*
Female 56.8 47.7
Education >high school (%) 26.0 15.8* Smoker (%) Current smoker 66.6 63.0 Ex-smoker 4.1 3.9 Never 29.3 33.1 Years of residence in current address >10 years 92.6 94.3 ø10 years 7.4 5.7 Occupational dust/fumes (%) Exposed 10.3 21.1* Nonexposed 89.7 78.9 aSD, standard deviation. * P < 0.05. TABLE 2
Prevalences of Selected Health Outcomes and Unad-justed Odds Ratios for Study Subjects from Sanwei (Ex-posed Area) and Taicei (Control Area)
Symptom Taicei (%) Sanwei (%) Odds ratio (95% CIa) Chronic respiratory symptoms Cough 7.8 11.2 1.50 (0.96–2.33) Phlegm 5.9 10.8 1.91 (1.18–3.10)* Wheezing 4.3 6.7 1.58 (0.89–2.82) Dyspnea 6.8 3.2 0.39 (0.25–0.60)* Chronic bronchitis 1.6 2.5 1.55 (0.61–3.90) Acute irritative symptoms Eyes 10.9 18.9 1.90 (1.31–2.76)* Nose 18.0 17.4 0.96 (0.68–1.35) Throat 13.3 17.9 1.42 (0.99–2.03) Nausea 4.5 10.8 2.56 (1.52–4.32)* Chemical odor perception 2.1 84.6 263.26 (133.61–518.70)* a95% confidence interval. * P < 0.05.
showed consistently higher concentrations of SO2,
NO2, and PM10than the control area (Table 4).
DISCUSSION
Researchers have used many approaches in deter-mining the health effects of air pollution. Holland et al. (1980) suggested that evaluation of the effects of individual sources of pollutions is the epidemiologic approach that best eliminates the effects of other variables. Unfortunately, such an approach is im-possible in urban areas. The present study employs this approach in small communities of rural Taiwan. Lebowitz (1981) recommended spatiotemporal de-signs as useful strategies in surveillance of respira-tory effects of point sources of pollution. In this study, an attempt to estimate health effects in the two areas with different pollution levels was made. A spatial approach was used in which multivariate statistical analyses were performed in order to con-trol for possible confounding factors. As in other en-vironmental studies (Bhopal et al., 1994; Dales et al., 1989; Symington et al., 1992), only estimates of
ex-posure for residents, based on community monitor-ing, were available.
The results show that in many respiratory re-spects the population living close to petrochemical manufacturing complexes (Sanwei) had a similar re-spiratory health experience to that in the control area (Taicei). However, for a number of acute tive symptoms (mainly eye irritation, throat irrita-tion, nausea, and chemical odor perception), there were large differences. However, due to the limited population studied and the cross-sectional study de-sign, the evidence cannot be considered conclusive.
In a previous study (Yang et al., 1997), it was dem-onstrated that schoolchildren residing in petro-chemical industrial area (Sanwei) have significantly more respiratory symptoms compared with a cohort living in a control area (Taicei). Thus, the current findings suggest that chronic exposure to petro-chemical industrial emissions have a biologic effect on the respiratory system great enough to cause symptoms among children, who may be sensitive in-dicators of environmental pollution, but not among adults.
Since no other major industrial sources of air pol-lution existed close to the petrochemical industrial area or in the control area, it can be assumed that the petrochemical industrial complexes were the single contributor to local air pollution. In this con-text it is inferred that acute symptom differences between the study and the control areas were not caused by another external source of air pollution.
The stability of the population, evidenced by length of residence (almost all respondents resided in current home more than 10 years), indicates that exposure has indeed been of long duration. However, the choice of residential location reflects factors re-lated to health. Families with higher incomes may avoid living near a petrochemical plant, and persons adversely affected may move away. In a cross-sectional study, the bias introduced by personal choices as to place of residence cannot be fully de-tected or controlled.
Odor perception in the exposed area may signal exposure to toxicologically significant concentra-tions of hazardous materials. In a pilot study con-ducted by Chan (1995) in the exposed area, for ex-ample, airborne levels of benzene and other volatile organic compounds (VOCs) could be distinguished from background levels in a neighborhood city (a city near the exposed area). While the mechanisms are not well understood (Wallace, 1991), these VOC compounds may cause symptoms through direct ir-ritation and neurotoxic effects (Nielsen and Alaire, 1982; Harving et al., 1991). Experimental human TABLE 3
Adjusted Odds Ratios for the Assocation between Selected Health Outcomes and Living Area
Health indicator
Adjusted odds
ratioa 95% CIb
Chronic respiratory symptoms
Cough 1.62 (0.99–2.64)
Phlegm 2.01 (1.19–3.42)*
Wheezing 1.65 (0.88–3.10)
Dyspnea 0.42 (0.26–0.68)*
Chronic bronchitis 1.63 (0.67–2.59)
Acute irritative symptoms
Eyes 2.16 (1.43–3.25)*
Nose 0.97 (0.68–1.38)
Throat 1.52 (1.05–2.20)*
Nausea 2.51 (1.45–4.35)*
Chemical odor perception 403.43 (181.27–880.07)*
aTaicei (control area) as a reference group. b95% confidence interval.
* p < 0.05.
TABLE 4
Summary of Air Pollution Data in Sanwei Primary School (Exposed Area) and Taicei Primary School (Con-trol Area)
Pollutantsa Taicei Sanwei P value
SO2(ppb) 2.59 (2.72) 10.60 (3.15) <0.05
NO2(ppb) 8.29 (1.64) 17.43 (2.25) <0.05
PM10(mg/m3) 62.79 (1.82) 93.57 (1.92) <0.05
exposures to mixtures of volatile organic compounds also produce symptoms of eye, mucosal, and airway irritation (Molhave et al., 1986). Thus, the findings that irritative symptoms (i.e., ‘‘eye soreness or irri-tation’’ and ‘‘throat soreness or irriirri-tation’’) and au-tonomic symptoms (nausea) were associated with residence exposure to petrochemical industrial emissions are biologically plausible and are consis-tent with the limited available evidence.
In conclusion, the results of this study show that residents in Sanwei, a rural area exposed to petro-chemical industrial air pollution, have statistically significant more acute symptoms (eye irritation, nausea, throat irritation, and odor perception) than those in Taicei, a low-pollution rural area. The local community’s observations that emissions from the petrochemical industrial complexes affected their health are probably correct. The excess of acute symptoms observed in those living close to the pet-rochemical industrial complexes can be best ex-plained as a result of their exposure to its emissions, especially the VOCs, although the concentration of VOCs is lacking in this study. Therefore, it is rec-ommended that future studies seek to identify the potential role of VOCs in the genesis of acute irrita-tive symptoms near petrochemical industrial areas.
ACKNOWLEDGMENT
This study was supported by a grant from the National Science Council, Executive Yuan, Taiwan (NSC-85-2621-P-037-001).
REFERENCES
Bhopal, R. S., Phillimore, P., Moffatt, S., and Foy C. (1994). Is living near a coking works harmful to health? A study of in-dustrial air pollution. J. Epidemiol. Community Health 48, 237–247.
Chan, C. C. (1993). ‘‘The Monitoring of Ambient Volatile Organic
Compounds: A Pilot Study.’’ Environmental Protection Admin-istration. Taipei [EPA-82-L104-09-07]
Dales, R. E., Spitzer, W. O., Suissa, S., Schechter, M. T., Tousig-nant, P., and Steinmetz, N. (1989). Respiratory health of a population living downwind from natural gas refineries. Am. Rev. Respir. Dis. 139, 595–600.
Environmental Protection Administration, Republic of China (1984–1992). ‘‘Annual Reports of Air pollution in Taiwan, 1983–1990.’’ Environmental Protection Administration, Taipei. Ferris, B. G. (1978). Epidemiology standardization project. Am.
Rev. Respir. Dis. 118, 36–47.
Harving, H., Dahi, R., and Molhave, L. (1991). Lung function and bronchial reactivity in asthmatics during exposure to volatile organic compounds. Am. Rev. Respir. Dis. 143, 751–754. Holland, W. W., Bennett, A. E., Cameron, I. R., Florey, C. D. V.,
Leeder, S. R., Swan, A. V., and Waller, R. E. (1980). Epidemio-logic evidence and the United States air quality standards. Am. J. Epidemiol. 111, 458–459.
Hosmer, D. W., and Lemeshow, S. (1989). ‘‘Applied Logistic Re-gression.’’ Wiley, New York.
Lebowitz, M. D. (1981). Respiratory indicators. Environ. Res. 25, 225–235.
Molhave, L., Bach B., and Pedersen, O. F. (1986). Human reac-tions to low concentrareac-tions of volatile organic compounds. En-viron. Int. 12, 167–175.
Nielsen, G. D., and Alaire, Y. (1982). Sensory irritation and re-spiratory stimulation by airborne benzene and alkyl-benzenes. Toxicol. Appl. Pharmacol. 65, 459–477.
Symington, P., Coggon, D., and Holgate, S. (1992). Respiratory symptoms in children at schools near a foundry. Br. J. Indust. Med. 48, 588–591.
Wallace, L. A. (1991). Comparison of risks from outdoor and in-door exposure to toxic chemicals. Environ. Health Perspect. 95, 7–13.
Yang, C. Y., Wang, J. D., Chan, C. C., Hwang, J. S., and Chen, P. C. (1997). Respiratory symptoms of primary school children living in a petrochemical polluted area in Taiwan. Pediatr. Pulm.
Yang, C. Y., Chiu, H. F., Chiu, J. F., Kao, W. Y., Tsai, S. S., and Lan, S. J. (1997). Cancer mortality and residence near petro-chemical industries in Taiwan. J. Toxicol. Environ. Health 50, 265–273.
