Abstract

Background

Living near waste incineration plants (WIPs) may have adverse effects on health associated with quality of life (QOL) among local residents. This study was undertaken to measure the QOL of residents living near WIPs in China, identify the association between residential distance from the WIPs and QOL, and assess the mediating effect of respiratory symptoms on the association between residential distance and QOL.

Methods

A cross-sectional study was conducted in communities surrounding three municipal WIPs in Dongguan, China. The QOL using the WHOQOL-BREF and respiratory symptoms using the Medical Research Council Questionnaire (MRCQ) on Respiratory Symptoms were measured in 881 residents (454 living 3听km of their local WIP vs. 427 living 3鈥�8听km from the WIP). Differences in respiratory symptoms and QOL in two separate items and four domains were analyzed. The associations of distance with respiratory symptoms and QOL were evaluated by multiple linear regression and logistic regression, respectively. Mediating effect of respiratory symptoms on the association between distance and QOL were tested using mediation analysis.

Results

Compared with the 3鈥�8听km group, residents in the 0鈥�3听km group had significantly lower scores in the physical health domain (尾= 鈭掆��2.02, p鈥�=鈥�0.027), social relationships domain (尾= 鈭掆��2.47, p鈥�=鈥�0.016) and environment domain (尾= 鈭掆��3.55, p鈥�&濒迟;鈥�0.001). There were no significant associations between distance and psychological domain, overall QOL, and satisfaction with health after adjusting for other variables. Residents in the 0鈥�3听km group were more likely to report respiratory symptoms than residents in the 3鈥�8听km group (OR鈥�=鈥�1.46, 95%CI 1.09鈥�1.96, p鈥�=鈥�0.011). Respiratory symptoms had a significant effect on physical health (尾= 鈭掆��7.48, p鈥�&濒迟;鈥�0.001), social relationships (尾= 鈭掆��5.02, p鈥�&濒迟;鈥�0.001) and environment domains (尾= 鈭掆��7.28, p鈥�&濒迟;鈥�0.001). Respiratory symptoms fully mediated the association of residential distance on physical health (31.7%), and partially mediated the associations of residential distance on social relationships (18.8%) and environment domains (17.3%).

Conclusions

Residential distance from a WIP was significantly associated with QOL, and the association between residential distance and QOL was mediated by respiratory symptoms. Future longitudinal studies on adverse respiratory health effects of exposure to WIP emissions are required using objective assessments.

Introduction

Global municipal solid waste has seen a rapid growth in recent decades along with increased population, expanding urbanization, and improved lifestyles [1]. Despite all the recycling technologies in use, less than 20% of the world鈥檚 waste is recycled each year, with a huge amount of waste sent to landfills [2]. Due to a scarcity of land and secondary pollution from landfills, waste incineration is the dominant method of waste disposal in some high-income countries [3], and has also become common practice in China [4]. Waste incineration plants (WIPs) have historically been associated with the emission of toxic pollutants [5], and a wide range of adverse health effects have been identified among residents living near WIPs, including significant associations with some cancers and reproductive dysfunction [6,7,8]. With advancements in incineration technology and stricter emission standards, the chemical emissions of modern WIPs are more limited, but toxic substances are still released in the atmosphere, which have a negative impact on the natural environment and the health of people living close to such facilities [9]. However, the presence of hazardous substances cannot easily be translated into useful exposure information, thus surrogate measures of exposure such as the distance of residences from a WIP are thus often used in epidemiological studies [10].

Quality of life (QOL), which is defined as 鈥渋ndividuals鈥� perception of their position in life in the context of the culture and value systems in which they live and in relation to their goals, expectations, standards and concerns鈥� [11], is commonly used as a measurement of an individual鈥檚 overall health. Previous studies on QOL focused mainly on general populations or particular populations such as AIDS patients, tuberculosis patients, refugees, and alcohol use disorders [12,13,14,15,16]. QOL is influenced by sociodemographic characteristics and baseline health status, including age, gender, socioeconomic status, residence in rural or urban areas, and health and disease status [17, 18]. However, these previous studies provided little information on health-related QOL of residents living near a WIP, and their QOL may be impacted by odors, air pollution, and truck traffic due to the WIP and reflected by their risk perception [19].

To date, few studies have examined the QOL of people living near WIPs. One study was just presented as an abstract at a conference of the International Society for Quality of Life Research which reported significantly lower QOL in exposed groups [20], and another study showed no significant QOL impacts in residents living in the vicinity of a WIP compared to the control group [19]. Two studies measured respiratory problems related to WIPs in different study populations: one study focused on wheezing symptoms among Japanese children living near a WIP [21], while another study assessed respiratory functions in incineration workers [22]. In addition, two other studies examined the impact of respiratory symptoms on QOL among tobacco smokers [23] and respiratory symptoms impairing the QOL in adults [24]. A study in Vietnam showed that individuals with low risk perception from satisfied environment generally tend to have an increasing QOL [25].

Available literature showed that both the distance from a WIP and respiratory symptoms independently affected the QOL and that respiratory symptoms were associated with living near WIPs. However, to our knowledge, no studies have explored their combined effect on the QOL of residents living near WIPs. We had a hypothesis to determine to what extent the effect of residential distance on QOL is mediated by respiratory symptoms to adjust for mediator-outcome confounding [26]. Therefore, the aims of this study were to measure the QOL of residents living near a WIP, identify the association between residential distance from WIPs and QOL, and assess the mediating effect of respiratory symptoms on the association between residential distance from the WIP and QOL.

Methods

Study design, settings and study sample

A cross-sectional study was conducted during 28 November 2021 to 1 July 2022 in communities surrounding three municipal WIPs in Dongguan, China, namely Nancheng WIP, Hengli WIP, and Machong WIP. Residents aged 18 years or above who had lived within an 8听km radius of one of the three municipal WIPs for at least one year were recruited. In order to explore the association between residential distance and QOL among residents living nearby WIPs, the residents were divided into two groups according to their distance from the WIP with 3听km as the cut-off value [27] that resulted in a 0鈥�3听km group and a 3鈥�8听km group. Definition of distance nearby WIP was not definitely recommended but 3听km was most used when assessing the environmental impact of WIPs on the surrounding residents [28] and also defined by the Dongguan Environmental Protection Authority as an environmentally sensitive area.

The formula for comparing two independent means was used for the sample size calculation considering means and standard deviations of the physical health domain of QOL of 61卤13 and 64卤11 in residents who lived near or far from the waste management facilities from a previous study [25]. Based on a 95% confidence interval (CI), 80% power, design effect of 1.5, and 10% estimation of non-response, a sample size of 422 for each group leading to a total of at least 844 participants. During the survey, the first house in each area group was randomly selected and then the houses next to the first house were visited. In each household, the household head was invited to participate in the survey until the sample size was achieved. The details of the study settings were published in a previous study [29].

Variables and measurements

QOL was measured using the World Health Organization Quality of Life Brief Version (WHOQOL-BREF), which contains 26 items. The first two stand-alone items are overall QOL and general satisfaction with health, respectively. The remaining 24 items are divided into four domains: physical (7 items), psychological (6 items), social relationships (3 items), and environment (8 items). Responses to the 26 items are on a five-point scale where 1 represents 鈥渧ery poor鈥� or 鈥渘ot at all鈥� or 鈥渧ery dissatisfied鈥� and 5 represents 鈥渧ery good鈥� or 鈥渆xtremely鈥� or 鈥渧ery satisfied鈥�. Except for items 3, 4, and 26, all of the other 23 items are in a positive direction, that is, high scores indicate a high QOL. The two stand-alone questions were analyzed separately as continuous variables. The total score of items within each of the four domains was used to calculate the domain score. Then each domain score was converted to scores ranging from 0 to 100 according to the second transformation method outlined in the WHOQOL-BREF scoring instructions [30], with a higher score indicating a better QOL. The internal consistencies of each domain of QOL were tested and ranged from 0.75 to 0.82.

The main exposure of interest was residential distance, which was divided into two categories, 0鈥�3听km and 3鈥�8听km from a WIP. Respiratory symptoms were the mediator of interest in our study and assessed by the Medical Research Council (MRC) questionnaire on cough, phlegm, periods of cough and phlegm, breathlessness, wheezing, and chest illnesses. The response to each question is 鈥測es鈥� or 鈥渘o鈥�, and then each participant鈥檚 responses to the 18 questions were then aggregated. If the respondent answered 鈥測es鈥� to any of the 18 questions, he or she was classified as symptomatic.

Other independent variables were sociodemographic characteristics, risk perception, and health-related behaviors. Sociodemographic characteristics were comprised of 6 items, age, gender, education level, marital status, annual personal income, and duration of residence. Income in yuan was calculated in USD using an exchange rate of 0.16 USD equal to 1 CNY and then categorized into three groups: low (<鈥�4,800 USD), middle (4,800鈥夆垝鈥�12,800 USD), and high (>鈥�12,800 USD) based on the Chinese standard.

Risk perception was measured using five items, environmental pollution (The WIP produces environmental pollution), health effects (The WIP affects my health), psychological stress (The WIP causes psychological stress for me), neighborhood image (The WIP damages the neighborhood鈥檚 image), and damage to property (The WIP reduces property values) [31, 32]. Each item was measured on a 5-point Likert scale ranging from 鈥渟trongly disagree鈥� scored as 1 to 鈥渟trongly agree鈥� scored as 5 (where higher scores reflected higher risk perception). The mean of these five items was calculated and then categorized into a binary scale using a cut-off value of 3 as low- or high-risk perception.

QOL may be affected by other health-related behaviors, such as smoking status, alcohol use, dietary habits, occupational exposures, and outdoor activity patterns [33,34,35]. Therefore, this study also measured these risk factors as the covariates.

Data collection

Before the field surveys, researchers visited the planned survey areas to identify the household distribution within an 8听km radius of the study WIPs. The researchers initially contacted the village or neighborhood committees to inform them of the study and get their support and approval, and these committees then informed the residents that a research team would visit them at their homes. Verbal consent was obtained from each participant. The data were collected using the structured questionnaires through face-to-face household interviews by trained research assistants who were medical students proficient in both Mandarin and Cantonese. Except for a small number of illiterate or elderly participants who needed investigators to read the items in the questionnaire to them, most of the participants completed the questionnaire independently.

Statistical analysis

Data were entered in Epidata v3.1 by double entry and analyzed using R v4.1.3 (R Core Team 2022, Vienna, Austria). The information of the sociodemographic characteristics and health-related behaviors of the study participants were descriptively presented using percentages. The differences in respiratory symptoms at different distances were analyzed by Chi-square test. For QOL, the means of two separate items and four domains were compared based on distance using the student t-test.

The mediation analysis was conducted into four steps. First, the association between distance and QOL was assessed using multiple linear regressions and only the significant QOL domains were further analyzed for the mediating effects of respiratory symptoms. Second, the association between distance and respiratory symptoms was assessed using multiple logistic regression. Third, the effect of respiratory symptoms on significant domains of QOL was evaluated using multiple linear regression. Fourth, the coefficients of the linear and logistic regression models were standardized using the Lavaan Package version 0.6鈥�16 [36] to make them comparable. The regression coefficients (尾) and p-values of direct, indirect, and total effects were presented. The proportion of indirect effects was calculated by dividing the coefficient of the total effect by the coefficient of the indirect effect. In the mediation analysis, adjustments were made for gender, age, education level, income, marital status, duration of residence, risk perception, smoking, drinking, local food consumption, occupational exposure, and outdoor activities. A p-value less than 0.05 was considered to be statistically significant.

Results

In our study, we approached 1,008 households, of which 504 were located within 3听km, and 504 were within 3鈥�8听km of the WIP. Of these, 127 households refused to participate (12.6%): 50 households within 3听km (50/504, 9.9%) and 77 households within 3鈥�8听km (77/504, 15.3%). This resulted in a total of 881 participants, with 454 living within 3听km of the WIP and 427 within 3鈥�8听km. The sociodemographic characteristics and health-related behaviors of the 881 participants in this study are shown in Table听1. The majority of them were in the working-age groups, almost two-thirds were female, and three-fourths had educational levels of high school or higher. One-fourth of them had lived in their current residence for more than 15 years, 70.0% had high risk perceptions concerning WIPs, and 42.3% reported occupational exposure. Almost half did regular outdoor activities.

Table 1 Sociodemographic characteristics and health-related behaviors of study participants

Prevalences of respiratory symptoms classified by cough, phlegm, and wheezing between those living within 3听km from their WIP versus 3鈥�8听km were 37.4% vs. 28.8%, 31.3% vs. 22.5%, and 22.2% vs. 14.8%, respectively. The respiratory symptoms and QOL between the participants who lived within 3听km and 3鈥�8听km from their WIPs are shown in Table听2. The prevalence of respiratory symptoms was 59.4%. The participants living within 3听km were significantly more likely to report respiratory symptoms (65.0% vs. 53.4%, respectively, p鈥�=鈥�0.001). Comparing the four-domain scores of QOL among the participants in the 0鈥�3听km group, the social relationships domain was the highest with a mean score of 65.46, followed by the physical health (64.75), psychological (62.64), and environment (56.76) domains. The QOL scores of the participants living in the 0鈥�3听km group were significantly lower in the two stand-alone items and in the physical health, social relationships, and environment domains than participants in the 3鈥�8听km group.

Table 2 Respiratory symptoms and QOL between the residents who lived within 3听km and 3鈥�8听km

The associations between residential distance and QOL using multivariate linear regressions are shown in Table听3 for four-domain QOL and Table听4 for two stand-alone items of QOL. Compared with the 3鈥�8听km group, participants in the 0鈥�3听km group had significantly lower QOL scores in the physical health domain (尾= 鈭掆��2.02, p鈥�=鈥�0.027), social relationships domain (尾= 鈭掆��2.47, p鈥�=鈥�0.016) and environment domain (尾= 鈭掆��3.55, p鈥�&濒迟;鈥�0.001). There were no significant associations between distance and the psychological domain (Table听3), overall QOL, and satisfaction with health (Table听4). The female participants had significantly lower scores in the physical health and environment domains and overall QOL. Those who had high annual personal incomes had significantly higher scores in the social relationships, environment domains, and overall QOL compared to those having low annual personal incomes. Those with higher risk perceptions of WIPs scored lower on the environment domain, overall QOL, and satisfaction with health. Those having higher risk perception and occupational exposure had significantly lower scores for the environment domain and overall QOL. Those who had outdoor activities had higher scores for all domains, except for satisfaction with health.

Table 3 Final models of associations between residential distance and QOL in four domains from multiple linear regression

Table 4 Final models of associations between residential distance and QOL in two stand-alone items from multiple linear regression

Table听5 shows the associations between residential distance and respiratory symptoms. After adjusting for other variables in this model, distance had a significant effect on respiratory symptoms, with participants in the 0鈥�3听km group more likely to report respiratory symptoms than participants in the 3鈥�8听km group (aOR鈥�=鈥�1.46, 95% CI 1.09鈥�1.96, p鈥�=鈥�0.011). Married participants reported more respiratory symptoms than single ones. Participants who had a higher risk perception of WIPs or thought they had occupational exposure were more likely to report respiratory symptoms.

Table 5 Final model for the associations between residential distance and respiratory symptoms from multiple logistic regression

The associations between residential distance and respiratory symptoms with QOL in the physical health, social relationships and environment domains from multiple linear regression are presented in Table听6. Respiratory symptoms had a significant effect on the physical health (尾= 鈭掆��7.48, p鈥�&濒迟;鈥�0.001), social relationships (尾= 鈭掆��5.02, p鈥�&濒迟;鈥�0.001) and environmental domains(尾= 鈭掆��7.28, p鈥�&濒迟;鈥�0.001) after controlling for distance. Distance still had a significant effect on the social relationships (尾= 鈭掆��2.05, p鈥�=鈥�0.043) and environment domains (尾= 鈭掆��2.94, p鈥�=鈥�0.001) when controlling for respiratory symptoms. Distance had no significant effect on physical health after controlling for respiratory symptoms (尾= 鈭掆��1.39, p鈥�=鈥�0.114). Gender, annual personal income, risk perception, occupational exposure, and outdoor activities were significantly associated with some aspects of QOL.

Table 6 Final models for the associations between residential distance and respiratory symptoms with physical health, social relationships, and environment domains

Figure听1 shows the mediating effects of respiratory symptoms on the associations between residential distance and the physical health (Fig.听1a), social relationships (Fig.听1b) and environmental domains (Fig.听1c). Standardized regression coefficients for indirect effects of residential distance through respiratory symptoms on physical health (尾= 鈭掆��0.026), social relationships (尾= 鈭掆��0.016) and environment domains (尾= 鈭掆��0.024) were significant. The corresponding standardized regression coefficients (尾) for the direct effects were 鈭掆��0.056 (p鈥�=鈥�0.087), 鈭掆��0.069 (p鈥�=鈥�0.035), and 鈭掆��0.115 (p鈥�&濒迟;鈥�0.001), and for the total effects were 鈭掆��0.082 (p鈥�=鈥�0.015), 鈭掆��0.085 (p鈥�=鈥�0.010), and 鈭掆��0.139 (p鈥�&濒迟;鈥�0.001), respectively. Respiratory symptoms fully mediated the association of residential distance on the physical health domain (31.7%), and partially mediated the association of residential distance on the social relationships (18.8%) and environmental domains (17.3%).

Fig. 1

Mediating effect of respiratory symptoms on the relationships between residential distance and QOL in physical health (a), social relationships (b), and environment (c) domains. Adjusted for gender, age, marital status, education level, annual personal income, risk perception, duration of residence, smoking, drinking, occupational exposure, outdoor activities, and consumption of local food; *p鈥�<鈥�0.05, **p鈥�<鈥�0.01, ***p鈥�&濒迟;鈥�0.001

Discussion

Significantly lower scores in three domains of QOL, physical health, social relationships, and environment, were found among participants living within 3听km from their WIP compared to those living within 3鈥�8听km in both the bivariate and multivariate linear regression models. Participants living in the 0鈥�3听km group had significantly lower scores in overall QOL and satisfaction with health only in univariate analysis. Participants living in the 0鈥�3听km group were significantly more likely to report respiratory symptoms than those in the 3鈥�8听km group. Respiratory symptoms fully mediated the association between residential distance and physical health, but partially mediated the association between residential distance and social relationships and environment.

Among the four QOL domains, the 0鈥�3听km group in our study scored highest on the social relationships domain and lowest on the environmental domain. This finding was similar to a previous study among a population living near a solid waste management facility [25]. One possible explanation is that WIPs are often built in the urban-rural fringe, where residents live in a community of acquaintances and have established long friendly social relationships to solve common local problems [37]. Therefore, the WIP has relatively little impact on the local social relationships domain. The lowest scores in environment domain could be explained by the fact that air pollutant emissions from WIPs have been reported to cause serious environmental pollution in China [38]. Likewise, a study conducted in Canada also reported negative impacts on air, water, and soil quality related to WIP emissions [39].

The QOL scores of all four domains among our participants living in the 0鈥�3听km were better than the scores from a previous study in Vietnam [25]. This result may be due to the fact that China has a clear policy on waste treatment technology and stricter emission standards [40]. However, compared with the findings of another study conducted in a general population in Hong Kong, China [13], the participants living in the 0鈥�3听km group in our study had relatively low scores in the physical health, psychological, and environment domains due to the perceptions on the QOL of participants living near WIPs which was different from the general population of Hong Kong.

Lower QOL scores were observed in the physical health, social relationships, and environment domains among participants in the 0鈥�3听km group compared to the 3鈥�8听km group. We found two studies from Korea conducted in 2005 [37] and 2006鈥�2009 [19] that showed different findings. The first study used the WHOQOL-100 to compare 158 residents living near a garbage dumping site to 99 residents living much farther away. The study found significance in only two QOL domains, physical health and environment [37]. The latter study used the WHOQOL-BREF with a scale of 1 to 5 among 841 persons who lived within 300听m of a WIP versus 105 residents who lived in an area without a WIP. The results showed very slightly lower scores in the residents who lived near a WIP compared to residents who lived farther away, but the difference was not significant [19]. These studies indicated the importance of study settings in terms of study participants, types of waste disposal facilities, study designs, and measurement tools on the effects of distance on QOL. Residential distance from a WIP was no longer significantly associated with overall QOL and satisfaction with health after adjusting for other variables in multiple regression models, which indicated that other variables were more important in explaining overall QOL and satisfaction with health.

Our findings on a higher proportion of respiratory symptoms in residents living within 3听km of their WIP compared to those living within 3鈥�8听km were supported by two previous studies. Those studies showed increased morbidity levels for respiratory disorders in areas with high PM鈧佲個 levels due to incinerators among men in respect to the reference group [41]. In addition, an increased prevalence of wheezing was found in children living nearer to WIPs than some control groups living more distant [21]. A possible explanation could be due to the fact that most pollutants released by WIP are stack emissions [42] and air pollution was shown to be associated with the development of respiratory illness [43]. Adverse effects on respiratory symptoms can be observed even at low levels of ambient air pollution exposure [44]. We found one study from the USA which showed no significant difference in the prevalence of chronic or acute respiratory symptoms between incinerator and comparison communities but the study was conducted between 1992 and 1994 [45].

Our study showed respiratory symptoms had a significantly negative effect on QOL in physical health, social relationships, and environment domains after controlling for distance. We could not find studies directly related to respiratory symptoms and QOL in residents living near WIPs that would be the same as our study. However, we found two studies which focused on different study populations. One study focused on adults with and without chronic obstructive pulmonary diseases [24] and the other study was done in a general population [46]. Both studies found that respiratory symptoms had a negative impact on QOL, which were the same results as our findings.

The effect of residential distance on QOL in the physical health domain was fully mediated through respiratory symptoms, which indicated that the residential distance had no direct effect on the physical health domain of QOL but through respiratory symptoms. This finding may not be surprising as people who have respiratory symptoms generally perceived this as a reflection of their physical health being measured in the QOL regardless of their residential distance from the WIP. On the other hand, respiratory symptoms only partially mediated the associations of residential distance on the social relationships and environment domains. It was clear that distance not only had a direct effect on the social relationships and environmental domains, but also indirectly affected social relationships and environmental domains through respiratory symptoms. This indicated multifactorial factors of people affecting QOL [47] which may be related to the residential distance to WIPs. To the best of our knowledge, no previous studies have been published that assessed the mediating effect of respiratory symptoms on the association between residential distance related to WIP and the QOL. Our study elucidated the mechanism of residential distance from a WIP on QOL through respiratory symptoms, particularly in physical health not social and environmental domains, using mediation effect analysis.

There were some limitations in this study. First, the nature of a cross-sectional study may be limited in its ability to explore causal relationships between the impact of WIPs and QOL. Second, the overall refusal rate to participate in this study was 12.6%. However, we did not ask why participants refused, so the reasons for refusal could not be explained. Third, we did not use random sampling for household selection, but the households within the selected radius areas were approached. Fourth, the WIPs in the study were in industrial areas where other sources of exposure to pollutants were present. However, we emphasized to the residents we wanted to assess their health effects related to the WIP in their area. Finally, the self-reported respiratory symptoms of the participants were used to measure respiratory health in our study, which involved a subjective judgement. However, the objective measurement of using lung function tests could not be used in our study due to the prevention and control policy of coronavirus disease.

In conclusion, our study found that the residential distance from a WIP was significantly associated with QOL, particularly in the physical health, social relationships, and environment domains. Residents living near a WIP were more likely to report respiratory symptoms, and the association between residential distance and QOL was mediated by respiratory symptoms. Waste incineration is the most important way for the Chinese government to dispose of municipal solid waste. Hence, it is very necessary for the operators of WIPs to adopt more advanced incineration technology and strictly follow the national pollution emission standards to reduce the harmful emissions of waste incineration. Moreover, the government should also implement long-term monitoring to determine the environmental and health effects on the residents who reside near the WIPs. Future longitudinal studies on the adverse respiratory health effects related to exposure to WIP emissions using objective assessments rather than self-reported measures among local residents are required.