Effects of Household Clean Fuel Combustion on the Physical and Mental Health of the Elderly in Rural China

Abstract

Against the backdrop of the construction of an ecological civilization and the “Healthy China 2030” initiative, access to clean fuels is crucial for achieving optimal health and wellbeing, as well as sustainable social development. The purpose of this study is to investigate the effect of household clean fuel combustion (HCFC) on multiple dimensions of health among older adults while shedding light on the potential mechanisms. We performed a cross-sectional study of the data from the 2018 China Health and Retirement Longitudinal Survey, and we surmounted the underlying issues of endogeneity with the application of propensity score matching and the instrumental variable strategy. The results revealed that HCFC has positive effects on the health of older adults, particularly by improving their psychological wellbeing. The adoption of clean fuels among the elderly was associated with a significant increase in SRH by 3.06% to 3.42% and a decrease in CES-D by 7.96% to 8.28%. These positive environmental health effects became stronger among the elderly under the age of 75, as well as among those who were educated and had chronic diseases. Moreover, the results demonstrated that HCFC significantly alleviated chronic pain and increased social interaction among older adults, highlighting two potential pathways for promoting their wellbeing. Given that a significant number of rural households in China rely on polluting fuels, targeted strategies are crucial for promoting the use of clean fuels, particularly for vulnerable populations.

1. Introduction

The comprehensive implementation of clean fuels, such as electricity and natural gas, is yet to be realized in developing countries. Shockingly, about 2.9 billion people worldwide continue to rely on polluting fuels for their domestic energy needs [1]. The utilization of fossil fuels releases greenhouse gases, leading to detrimental environmental effects, such as global warming and acid rain. Developing more efficient methods to make clean fuels more attractive and feasible has emerged as a key area of focus among scholars [2]. Furthermore, incomplete combustion of solid fuels such as biomass for cooking and space heating also leads to high emissions of air pollutants, including CO, CO₂, PM_2.5, SO₂, and NO_X [3]. In an indoor kitchen with no partitions, the 24 h average PM_2.5 concentration can reach a maximum of 758 μg/m³ [4]. When combined with a simple stove and inadequate ventilation, pollutants are unable to be effectively diluted and dispersed, resulting in serious household air pollution (HAP), which is strongly associated with health hazards [5]. According to a recent report by the World Health Organization (WHO), an estimated 3.2 million people died prematurely in 2020 from illnesses attributable to HAP [6]. Additionally, individuals incur significant expenses each year due to HAP-related health expenditures [7] and insurance premiums [8]. Therefore, it is crucial to consider the most favorable and optimal fuels while taking both economic and environmental factors into account to achieve sustainable development in developing countries [9].

The United Nations proposed that ensuring access to affordable and clean fuels for cooking is an important pathway for addressing ambient and HAP issues, leading to good health and overall wellbeing [10]. The guidance and support provided by the Chinese government are essential in promoting the development of clean fuels in rural areas. On the one hand, the government has closely linked the development of clean fuels with ecological environment management, exemplified by its promotion of biogas projects and encouragement of waste-to-energy incineration. On the other hand, the government has formulated appropriate preferential policies in finance and taxation to ensure the widespread use of clean fuels, such as subsidies for energy-efficient and environmentally friendly stoves and subsidies to winter heating electricity prices. Notably, the clean energy heating program bestowed benefits upon roughly 23 million rural households by the end of 2018, showcasing the economic and environmental advantages of this initiative [11]. Despite the decrease in HAP exposure in China from 54% to 36% over the last decade, a significant population of over 450 million people—mainly consisting of farmers—remains exposed to HAP [12]. Therefore, curbing HAP exposure remains a persistent challenge in China.

Another crucial factor to consider is the rapidly increasing rate of aging in rural China. Currently, it stands at a staggering 23.81% [13]. With the advent of urbanization, industrialization, and modernization, over 200 million young individuals have migrated to cities for employment opportunities, including an increasing number of rural women. Meanwhile, the arduous task of gathering and preparing polluting fuels now rests on the shoulders of rural elders. As evidenced by a previous study, the severity of the health repercussions suffered by occupants correlates with the concentrations of pollutants and length of exposure [14]. Daily exposure to incomplete combustion products greatly increases the susceptibility of rural older adults to the environmental health risks caused by HAP. Therefore, given the context of the household energy transition in rural China, clean fuels hold significant potential for significantly improving the health of older adults.

Therefore, our study centers on the integration of the construction of an ecological civilization and the “Healthy China 2030” initiative while reassessing the health benefits of household clean fuel combustion (HCFC) for older adults in rural China. In this regard, our study has several important contributions. Firstly, we establish a theoretical framework for the effects of HCFC on health based on the Grossman model and examine the effects of HCFC on multiple dimensions of health among older adults. This comprehensive approach enables us to appreciate the heightened significance of the advantages for psychological wellbeing that HCFC provides. Secondly, we perform a heterogeneity analysis while focusing on the disparities among subgroups with distinct levels of education and chronic disease status, unlike in prior studies that emphasized intergenerational and gender inequalities in environmental threats. Lastly, we delved into the possible mechanisms of the health benefits of HCFC regarding chronic pain and social interaction, which are intimately related to the lives of older adults.

The remaining sections of our study are arranged as follows: Section 2 provides a comprehensive review of the literature and establishes the theoretical framework. Our data and identification strategies are expounded upon in Section 3, followed by the presentation of the primary empirical findings in Section 4. Section 5 is dedicated to the ensuing discussions, while Section 6 serves as the conclusion and recommendation to our research.

2. Literature Review and Theoretical Framework

2.1. Literature Review

The Global Burden of Disease (GBD) study assessed the health risk attributed to HAP [15]. Further supported by cohort studies, associations between pollutants such as PM_2.5 and mortality [16], particularly in terms of cardiovascular disease [17] and lung cancer [18], were confirmed. The wealth of public data and natural experiments conducted across various nations have furnished robust evidence of the detrimental connections between HAP exposure and life expectancy [19], as well as the occurrence of chronic diseases [20] and impaired physical functioning [21]. Households using polluting fuels are impelled by frequent health issues to consult medical professionals and undergo medical treatments, thereby incurring greater medical expenses [22]. It has been approximated that HAP accounts for an increase of 1.4–1.9% in healthcare expenditure, with rural areas and less-educated households being particularly vulnerable to its detrimental impacts [23]. Moreover, air pollution poses a significant threat to individuals’ mental health, with reports of increased adult depression risk, even at pollutant concentrations below the stipulated standards [24]. Studies further indicated that there is a positive correlation between PM_2.5 and rising urban anxiety indexes, with every unit increase leading to a 0.157 rise [25]. Exposure to other ambient air contaminants escalates the likelihood of outpatient visits for anxiety disorders [26], in addition to increasing the probability of being prescribed antidepressants [27]. In low- and middle-income countries, particularly in India and China, where the use of solid fuels for cooking and space heating is prevalent, individuals aged 50 and over who live in rural settings have recorded a marked surge in depression rates [28].

Given the government’s active push towards structural reforms for energy, the relationship between clean fuel combustion and health has gained increasing scholarly attention in recent years. For instance, the Nepalese government’s introduction of biogas plants in previous years has been shown to effectively mitigate the incidences of eye infections, headaches, cough, and respiratory diseases in the population [29]. In Turkey, natural gas services have been found to make a significant contribution to reducing mortality rates among the elderly [30]. Furthermore, a provincial Chinese study revealed that clean energy usage reduced household medical expenditure, affording greater resources for strengthening energy consumption patterns and ultimately alleviating energy poverty [31]. Environmental enhancement has also been shown to confer significant benefits to mental health, with the utilization of air purifiers, for example, resulting in improved sleep quality due to enhanced indoor air quality [32]. The use of kitchen ventilation likewise serves to ameliorate the association between prolonged cooking with solid fuels and the prevalence of depression and anxiety symptoms [33]. Several studies have conducted valuable research on the health effects surrounding clean fuel combustion in rural Chinese households. For instance, clean fuel users exhibited a marked reduction in the incidence of arthritis [34]. Additionally, the utilization of non-solid fuels has been shown to significantly augment the independent activities of daily living (ADLs) by approximately 3.02% to 3.40% [35], which is a significant predictor of a reduced risk of disability among the elderly.

Health issues related to HAP have been extensively discussed. Furthermore, the economic and health advantages of implementing national or provincial clean energy promotion policies and modern technologies have been well documented in recent scholarship. Taking a microscopic view, research on the association between HCFC and the wellbeing of elderly populations in rural Chinese regions has prioritized physical function, affliction incidence, mortality rates, and healthcare expenditures. In contrast, there has been a relative paucity of attention given to the perceived physical and mental health outcomes of this population. Furthermore, previous research has primarily focused on the physiological mechanisms underlying the negative health effects (e.g., the propagation of lung inflammation) caused by the combustion of polluting fuels. Consequently, a comprehensive discussion of the effects of HCFC on physical and mental health improvement while taking the unique lifestyles and social relationships of rural older adults into account is necessary. To address this gap in knowledge, our study utilized the latest national survey data from 2018 to investigate the potential benefits of HCFC for cooking and heating among rural elderly, in line with the establishment of an ecological civilization and the “Healthy China 2030” initiative.

2.2. Theoretical Framework and Potential Mechanism

Smith distinctly described the pathway through which polluting fuels used for domestic purposes affect human health [36]. The environmental health pathway begins with the combustion of biomass fuels and the consequent release of pollutants, moves to the concentration of pollutants in the environment and their accumulation in the body via daily exposure, and culminates in health effects. The Grossman model, which is a significant theoretical model for analyzing health capital formation, further reveals the mechanism by which HAP influences the health of residents [37]. Assuming that the initial stock of health capital is exogenously determined, the constraints on health capital investment are as follows:

$$\Delta {\mathrm{H}}_{\mathrm{t}+1}={\mathrm{H}}_{\mathrm{t}+1}-{\mathrm{H}}_{\mathrm{t}}={\mathrm{I}}_{\mathrm{t}}-{{\mathsf{\delta }}_{\mathrm{t}}\mathrm{H}}_{\mathrm{t}}$$

(1)

where $\Delta {\mathrm{H}}_{\mathrm{t}+1}$ is the change in resident health capital stock in period $\mathrm{t}+1$ and period $\mathrm{t}$. ${\mathrm{I}}_{\mathrm{t}}$ is the flow of resident investment in health capital in period $\mathrm{t}$. ${\mathsf{\delta }}_{\mathrm{t}}$ is the depreciation rate of health capital in period $\mathrm{t}$. It was proven that the present rate at which residents’ health capital depreciates is positively correlated with the current level of HAP [38]. For the same stock of health capital, more intense HAP leads to a faster depletion of the population’s health capital. Conversely, the depletion of the population’s health capital is slower with lower HAP. Notably, those inhabiting regions with inferior environmental conditions frequently experience an accelerated decline in their health capital stock [39]. As such, HCFC in rural areas can effectively mitigate the release of pollutants at the source, consequently reducing levels of harmful airborne particles and slowing the rate of depreciation of health capital, particularly among the elderly.

Our study further proposes the potential mechanisms underlying the impact of HCFC on the health of the rural elderly (please see Figure 1). In rural regions, particularly those situated in hills and mountains, transportation options are limited, and straw and firewood are used. The process of collecting and drying solid fuels must be performed manually, thus placing high physical demands on individuals. This further increases the risk of back pain in the elderly [40], especially for women who engage in prolonged periods of heavy housework with awkward postures. Chronic pain not only impairs mobility but leads to other consequences, such as sleep disorders [41]. On average, 5% to 85% of patients with pain experience depression [42]. Furthermore, due to the low combustion efficiency of these fuels, cooking, and heating necessitate frequent additions to the stove, further reducing leisure time. Participating in social activities serves as a bridge for individuals to form close relationships with family members, neighbors, and friends [43], thus contributing to their overall health [44]. In rural areas, better social capital buffers the negative effects of low socioeconomic status on health [45]. Frequent and diverse social participation can also improve the levels of emotional and social support for the elderly, thereby reducing the symptoms of depression [46].

Research has shown that HCFC not only benefits the environment but also provides convenience, efficiency, and safety. Although the adoption of clean fuels may require some financial resources from households, more and more rural families have come to realize the cost-effectiveness of electricity and have made it their preferred choice, thanks to government policy support [47]. Through the use of clean fuels, older adults can conserve significant amounts of time and effort that were previously spent on the laborious tasks of cooking and heating every day. This reduction in frequency and intensity of these activities may alleviate the symptoms of chronic pain. Additionally, HCFC saves opportunity costs, and the elderly can use leisure time to participate in social activities. This not only relieves daily stress but also promotes behaviors for good health, thereby improving the perception of health and providing protection against anxiety and depression [48]. Based on our analysis, we suggest that HCFC may efficiently reduce labor intensity and time, thus positively impacting the physical and mental health of older adults through two key channels: alleviating the occurrence of chronic pain and increasing opportunities for social interaction.

3. Materials and Methods

3.1. Materials

The China Health and Retirement Longitudinal Survey (CHARLS) is a thorough inquiry carried out by the National Development Institute of Peking University and is funded by the National Natural Science Foundation of China. The survey instruments are harmonized with more than 25 health and retirement surveys worldwide, and the survey results are internationally comparable [49]. The interview response rate and data quality of CHARLS are among the highest in the world for similar projects, and the data are widely used and recognized in the academic field. Our study used data from the latest CHARLS in 2018, covering 28 provinces in China that comprised 450 communities in 150 counties, with an overall aggregate of 12,400 households and 19,000 respondents. CHARLS 2018 inquired exhaustively about both the health information and household fuel choices of the elderly in rural China. Furthermore, the comprehensive evaluation of demographic backgrounds, family structure, health insurance, and income status presented the ability to analyze the relationship between HCFC and health outcomes. Based on the study content, our analysis only encompasses data from respondents aged 50 years and older residing in the rural regions of China. Upon discarding all extraneous and missing samples, our study investigated a sample size of 7959 respondents, of which 4344 used clean fuels for household cooking and heating, and 3615 used polluting fuels.

3.2. Variable Selection

3.2.1. Dependent Variables

Self-rated health (SRH): SRH is a comprehensive measure of a respondent’s own health, and it was deemed to be a valid indicator of health status with a strong relation to mortality [50]. This measure emphasizes the importance of perceived health, as a clinical diagnosis may devalue the importance of what patients say [51]. Referencing Giles [52], SRH is measured according to the response to the following question: “Would you say your health is very good, good, fair, poor or very poor?” The answer uses a 5-point Likert scale ranging from 1 = very poor to 5 = very good. A higher score indicates a better SRH status.

The Center for Epidemiological Studies Depression Brief Scale (CES-D): The CES-D, which was designed to assess the current level of depressive symptomatology in the general population, has been proven to be a valid measure of mental health in older Chinese adults [53]. Participants were asked to recall their feelings and behaviors over the last week and report the frequency of their symptoms (please see

Table A1. Questions for the measurement of CES-D score.

Variables	Questions
CES-D score 1. Rarely or none of the time (<1 day). 2. Some or a little of the time (1–2 days). 3. Occasionally or a moderate amount of the time (3–4 days). 4. Most or all of the time (5–7 days).	1. I was bothered by things that don’t usually bother me.
	2. I had trouble keeping my mind on what I was doing.
	3. I felt depressed.
	4. I felt everything I did was an effort.
	5. I felt hopeful about the future.
	6. I felt fearful.
	7. My sleep was restless.
	8. I was happy.
	9. I felt lonely.
	10. I could not get “going”.

Note: Our study reverse-codes two of these positive emotion items (item 5 and item 8).

in the Appendix A). The responses on the CES-D use a four-scale metric, from rarely or none of the time (<1 day) to most or all of the time (5–7 days). According to Lenore [54], responses to items with negative emotion are scored based on a scale of 0 for rarely to 3 for most of the time. Meanwhile, responses to items with positive emotions are reverse-coded from 0 (most of the time) to 3 (rarely). Subsequently, the scores of the 10 items are summed to obtain each senior’s CES-D score; these scores range from 0 to 30, with higher scores indicating more depressive symptoms and suboptimal mental health.

3.2.2. Explanatory Variable

Household clean fuel combustion (HCFC): According to the WHO, clean fuels include electricity, LPG, piped natural gas, biogas, solar energy, and alcohol fuels. Polluting fuels include biofuels (animal dung, wood, charcoal, firewood, crop waste), coal, kerosene, and paraffin [55]. Whenever a polluting fuel is employed for household cooking or heating, regardless of its significance as the primary household fuel, it is counted as household polluting fuel combustion. In this case, HCFC is defined as 0. For respondents who exclusively employed clean fuels for cooking and heating, HCFC is defined as 1.

3.2.3. Control Variables

Consistently with the prior literature [56], we incorporated a range of variables in our model as controls to account for individual demographics, socioeconomic variables, health risks, and household-level variables. The demographic variables included age, gender, and marital status. The socioeconomic variables included education, medical insurance, household consumption, and Dibaohu. The health risk variables included health during childhood, ADLs, and smoking. The household-level variables included daughter number and flushable toilets. The control variables used in our study were defined as follows:

Age: age of respondents in years.

Gender: a dummy variable that is equal to 1 if the respondent is male and 0 otherwise.

Marital status: a dummy variable that is equal to 1 if the respondent is married and 0 otherwise.

Education: years of schooling of the respondents (judged according to the highest level of education obtained by the respondents).

Medical insurance: a dummy variable that is equal to 1 if the respondent participates in basic medical insurance and 0 otherwise.

Household consumption: the annual consumption expenditure of a household (excluding medical expenditure).

Dibaohu: a dummy variable that is equal to 1 if the respondent’s housing or income is significantly lower than the local low-income standard for residents and receives the minimum subsistence allowance, and 0 otherwise.

Health during childhood: health status before the age of 15: from 1 = very poor to 5 = very good.

ADLs: a dummy variable that is equal to 1 if the respondent is in difficulty and needs help in any daily behaviors because of physical, mental, emotional, or memory problems, and 0 otherwise.

Smoke: a dummy variable that is equal to 1 if the respondent has the habit of smoking and 0 otherwise.

Daughter number: the total number of daughters.

Toilet flushable: a dummy variable that is equal to 1 if a household uses a flushing sanitary toilet and 0 otherwise.

3.2.4. Mediating Variables

Our proposed mechanism of HCFC improving health status relies on reduced labor intensity and time, unlike the focus of previous studies. Though CHARLS did not directly inquire about the labor intensity and time required for collecting and using household fuels, it fortuitously included inquiries into chronic pain symptoms and leisure-time socialization. As such, we derived indicators for chronic pain and social interaction in the elderly by employing relevant survey items to conduct our empirical analysis.

Feeling pain: the degree of trouble due to any physical pain: from 1 = none to 5 = very.

Arm pain: a dummy variable that is equal to 1 if the respondent feels pain in the arm and 0 otherwise.

Back pain: a dummy variable that is equal to 1 if the respondent feels pain in the back and 0 otherwise.

Knees pain: a dummy variable that is equal to 1 if the respondent feels pain in the knees and 0 otherwise.

Social interaction: a dummy variable that is equal to 1 if the respondent engaged in any kind of social activities in the last month and 0 otherwise.

Playing Ma-Jong/chess/cards: a dummy variable that is equal to 1 if the respondent played Ma-Jong/chess/cards in the last month and 0 otherwise.

Interacting with friends: a dummy variable that is equal to 1 if the respondent interacted with friends in the last month and 0 otherwise.

Going to a club: a dummy variable that is equal to 1 if the respondent went to a sport, social, or other kind of club in the last month and 0 otherwise.

3.3. Identification Strategy

The following ordinary least squares (OLS) model was used to examine whether the health of the rural elderly was related to HCFC:

$${\mathrm{H}\mathrm{e}\mathrm{a}\mathrm{l}\mathrm{t}\mathrm{h}}_{\mathrm{i}}={\mathsf{\alpha }}_0+{\mathsf{\beta }}_0{\mathrm{H}\mathrm{C}\mathrm{F}\mathrm{C}}_{\mathrm{i}}+{\mathsf{\gamma }}_0{\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}_{\mathrm{i}}+{\mathsf{\epsilon }}_{\mathrm{i}}$$

(2)

where ${\mathrm{H}\mathrm{e}\mathrm{a}\mathrm{l}\mathrm{t}\mathrm{h}}_{\mathrm{i}}$ is the dependent variable, which includes the SRH score and CES-D score. ${\mathrm{H}\mathrm{C}\mathrm{F}\mathrm{C}}_{\mathrm{i}}$ represents the respondents’ use of clean fuels for cooking and heating. ${\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}_{\mathrm{i}}$ is a set of control variables that include the demographic and economic status of the respondent $\mathrm{i}$. ${\mathsf{\epsilon }}_{\mathrm{i}}$ is an error term.

Endogeneity issues are commonly observed in research on air pollution and health. On the one hand, the use of clean fuels may positively affect the health of the elderly, and healthier household heads may be more likely to adopt clean energy sources [57], leading to potential two-way causation. On the other hand, various socioeconomic and demographic factors affect the determinants of fuel choices [58], and if these factors are correlated with the health status of the elderly, the study results may suffer from selection bias. To address this critical issue, we employed the instrumental variable (IV) and the propensity score matching (PSM) methods in our study to mitigate endogeneity issues and more accurately evaluate the effects of HCFC on the health of the elderly. Firstly, we employed the community clean fuel utilization (CCFU) as an IV for individual HCFC [59]. Secondly, we focused specifically on the average treatment effect on the treated (ATT) [60]. The PSM estimators we used to be as follows:

$${\mathsf{\tau }}_{\mathrm{A}\mathrm{T}\mathrm{T}}^{\mathrm{P}\mathrm{S}\mathrm{M}}={\mathrm{E}}_{\mathrm{p}\left(\mathrm{X}\right)\left|\mathrm{D}=1\right.}\left[\mathrm{E}\left({\mathrm{H}\mathrm{e}\mathrm{a}\mathrm{l}\mathrm{t}\mathrm{h}}_1\left|\mathrm{D}=1\right.,\mathrm{p}\left(\mathrm{X}\right)\right)-\mathrm{E}\left({\mathrm{H}\mathrm{e}\mathrm{a}\mathrm{l}\mathrm{t}\mathrm{h}}_0\left|\mathrm{D}=0\right.,\mathrm{p}\left(\mathrm{X}\right)\right)\right]$$

(3)

where ${\mathrm{H}\mathrm{e}\mathrm{a}\mathrm{l}\mathrm{t}\mathrm{h}}_1$ and ${\mathrm{H}\mathrm{e}\mathrm{a}\mathrm{l}\mathrm{t}\mathrm{h}}_0$ represent the health outcomes for the group using clean fuels and the group not using clean fuels, respectively.

To further investigate the potential mechanisms of HCFC on the health of the rural elderly, our study establishes the following model to test the possible mediating effects in Model (2):

$${\mathrm{M}\mathrm{e}\mathrm{d}\mathrm{i}\mathrm{a}\mathrm{t}\mathrm{o}\mathrm{r}}_{\mathrm{i}}={\mathsf{\alpha }}_1+{\mathsf{\beta }}_1{\mathrm{H}\mathrm{C}\mathrm{F}\mathrm{C}}_{\mathrm{i}}+{\mathsf{\gamma }}_1{\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}_{\mathrm{i}}+{\mathsf{\mu }}_{\mathrm{i}}$$

(4)

where ${\mathrm{M}\mathrm{e}\mathrm{d}\mathrm{i}\mathrm{a}\mathrm{t}\mathrm{o}\mathrm{r}}_{\mathrm{i}}$ represents the chronic pain and social interaction of the respondent $\mathrm{i}$. Model (4) examines the effect of HCFC on the mediating variables. If the coefficient ${\mathsf{\beta }}_1$ is significant, it could indicate that HCFC influences health status by affecting chronic pain and socialization to some extent.

4. Results

4.1. Descriptive Statistics

Table A2. Sample summary statistics for key variables.

Variables	Full Sample	With HCFC	Without HCFC	p-Value
	(1)	(2)	(3)	(4)
SRH score, mean (SD)	2.985 (1.032)	3.068 (1.030)	2.884 (1.026)	0.184 ***
CES-D score, mean (SD)	9.336 (6.652)	8.653 (6.485)	10.157 (6.758)	−1.504 ***
Age, mean (SD)	62.227 (8.361)	61.334 (8.253)	63.299 (8.364)	−1.966 ***
Gender (%) (1 if male)	0.509 (0.500)	0.509 (0.500)	0.509 (0.500)	−0.000
Male	4048 (50.86%)	2209 (50.85%)	1839 (50.87%)
Female	3911 (49.14%)	2135 (49.15%)	1776 (49.13%)
Marital status (%) (1 if married)	0.873 (0.333)	0.873 (0.333)	0.872 (0.334)	0.000
Married	6946 (87.27%)	3792 (87.29%)	3154 (87.25%)
Others	1013 (12.73%)	552 (12.71%)	461 (12.75%)
Education, mean (SD)	4.252 (4.360)	4.772 (4.417)	3.628 (4.208)	1.144 ***
Basic medical insurance (%) (1 if yes)	0.959 (0.198)	0.961 (0.193)	0.957 (0.203)	0.004
Yes	7724 (97.05%)	4231 (97.40%)	3493 (96.63%)
No	235 (2.95%)	113 (.60%)	122 (3.37%)
Household consumption, mean (SD)	9023.89 (30970.08)	12323.94 (39509.62)	5058.372 (14398.69)	7.3 × 103 ***
Dibaohu (%) (1 if yes)	0.088 (0.284)	0.059 (0.236)	0.124 (0.329)	−0.065 ***
Yes	703 (8.83%)	256 (5.89%)	447 (12.37%)
No	7256 (91.17%)	4088 (94.11%)	3168 (87.63%)
Health during childhood, mean (SD)	3.273 (1.123)	3.286 (1.117)	3.256 (1.130)	0.029
ADL (%) (1 if ADL disability)	0.057 (0.232)	0.042 (0.201)	0.075 (0.264)	−0.033 ***
Yes	455 (5.72%)	183 (4.21%)	272 (7.52%)
No	7504 (94.28%)	4161 (95.79%)	3343 (92.48%)
Smoke (%) (1 if yes)	0.456 (0.498)	0.450 (0.498)	0.465 (0.499)	−0.015
Daughter number, mean (SD)	1.307 (1.093)	1.288 (1.080)	1.329 (1.107)	−0.041 *
Toilet flushable (%) (1 if yes)	0.511 (0.500)	0.627 (0.484)	0.372 (0.483)	0.255 ***
Yes	4066 (1.09%)	2722 (62.66%)	1344 (37.18%)
No	3893 (48.91%)	1622 (37.34%)	2271 (62.82%)

Note: *** p < 0.01; * p < 0.1. The last column shows the results of the t-Test.

in the Appendix A details the health outcomes and characteristics of the subgroups of individuals with and without HCFC, as well as the entire sample. It shows a marginally higher average SRH score among the elderly using clean fuels relative to those using polluting fuels. Consistently, their average CES-D score was lower than that of those using polluting fuels. Statistically significant differences were found between the group with HCFC and the group without HCFC (significant at the 1% level of error). Figure 2 reveals that in households using polluting fuels, 32.89% of individuals self-reported poor or very poor health, and close to 40% had severe CES-D scores. In households using clean fuels, nearly 25% of individuals self-reported good or very good health, and most of them had lower CES-D scores. These results tend to support the existence of crude health benefits of HCFC.

In addition to health differences, we also found that the respondents using clean fuels tended to be younger and better educated with a higher household income and living in a better residential environment compared with those using polluting fuels. Respondents with poorer socioeconomic conditions were more likely not to use clean fuels, which was not surprising considering that household fuel choices in rural China are primarily based on policy support and individual economic status.

4.2. Basic Regression Analysis

This subsection examines the effects of HCFC on the health of the rural elderly by using OLS regression (please see

Table 1. Results of OLS regression of SRH score and CES-D score for rural elderly.

	SRH Score			CES-D Score
	(1)	(2)	(3)	(4)	(5)	(6)
HCFC	0.160 *** (0.023)	0.104 *** (0.023)	0.080 *** (0.023)	−1.460 *** (0.147)	−0.949 *** (0.146)	−0.799 *** (0.150)
Age	−0.012 *** (0.075)	−0.006 *** (0.001)	−0.005 *** (0.002)	0.022 ** (0.009)	−0.032 *** (0.009)	−0.034 *** (0.010)
Gender	0.165 *** (0.023)	0.172 *** (0.036)	0.167 *** (0.036)	−2.555 *** (0.146)	−2.298 *** (0.220)	−2.271 *** (0.220)
Married	0.075 ** (0.037)	0.022 (0.036)	0.025 (0.036)	−1.928 *** (0.241)	−1.481 *** (0.235)	−1.492 *** (0.235)
Education		0.011 *** (0.003)	0.011 *** (0.003)		−0.146 *** (0.017)	−0.148 *** (0.017)
Insurance		−0.131 ** (0.056)	−0.119 ** (0.057)		0.355 (0.347)	0.298 (0.348)
Consumption		0.008 ** (0.004)	0.006 * (0.004)		−0.091 *** (0.026)	−0.080 *** (0.026)
Dibaohu		−0.234 *** (0.041)	−0.223 *** (0.041)		2.377 *** (0.272)	2.318 *** (0.272)
Childhood health		0.107 *** (0.010)	0.108 *** (0.010)		−0.553 *** (0.062)	−0.557 *** (0.062)
ADL		−0.836 *** (0.045)	−0.826 *** (0.045)		5.110 *** (0.330)	5.062 *** (0.331)
Smoke		−0.094 *** (0.034)	−0.089 *** (0.034)		0.460 ** (0.210)	0.435 ** (0.210)
Daughter number			−0.023 ** (0.010)			0.058 (0.066)
Toilet flushes			0.108 *** (0.023)			−0.655 *** (0.146)
R−square	0.025	0.085	0.088	0.064	0.131	0.133
Observation	7959	7959	7959	7959	7959	7959

Note: *** p < 0.01; ** p < 0.05; * p < 0.1; robust standard errors are provided within parentheses.

). Based on Model (2), we gradually introduced individual demographic, socioeconomic, health risk, and household-level variables. The baseline model showed that HCFC had a statistically significant effect on both the SRH score and the CES-D score among the users of clean fuels. Specifically, the estimated coefficients of HCFC on the SRH score decreased from 0.160 to 0.080 after the gradual introduction of the control variables; nevertheless, a positive correlation remained evident and was significant at the 1% level of error. The effects of HCFC on mental health were markedly pronounced: The coefficients of HCFC on the CES-D score changed from −1.460 to −0.799, and this negative effect remained significant at the 1% level of error. These results reinforce the findings of previous studies that showed that using clean fuels for domestic energy has health benefits. The results for the effects of the other variables on health outcomes were consistent with those of existing studies. To be specific, older adults with formal education, better childhood health, and a good residential environment tended to have good physical and mental health.

4.3. Endogeneity Tests and Robustness Checks

4.3.1. Results of the Propensity Score Matching Method

To avoid the endogeneity problem caused by sample selection bias, we employed PSM for our analysis. Figure 3 illustrates the kernel density functions of individual propensity scores for the group using clean fuels and the group using polluting fuels both before and after matching. It is shown that after matching, the difference between the kernel density equation curves of the two groups decreased, and the trends tended to be the same. Based on this analysis, it was found that applying the PSM method indeed reduced the difference in the distributions of explanatory variables between the two groups and eliminated the estimation bias caused by sample self-selection.

We have documented the estimated ATT based on Model (3) in

Table 2. Results of PSM of physical and mental health for rural elderly.

	(1)	(2)	(3)	(4)
Panel 1: physical health
SRH score	0.104 *** (0.028)	0.094 *** (0.027)	0.104 *** (0.028)	0.099 *** (0.026)
Observations	7959	7959	7959	7959
SRH-good	0.033 *** (0.011)	0.032 *** (0.011)	0.032 *** (0.011)	0.034 *** (0.011)
Observations	7959	7959	7959	7959
Panel 2: mental health
CES-D score	−0.748 *** (0.184)	−0.780 *** (0.173)	−0.748 *** (0.184)	−0.780 *** (0.170)
Observations	7959	7959	7959	7959
Depression	−0.048 *** (0.013)	−0.050 *** (0.013)	−0.048 *** (0.013)	−0.049 *** (0.012)
Observations	7959	7959	7959	7959
Loneliness	−0.074 ** (0.029)	−0.083 *** (0.028)	−0.074 ** (0.029)	−0.078 *** (0.027)
Observations	7959	7959	7959	7959

Note: *** p < 0.01; ** p < 0.05. Column (1) shows estimates by k-nearest neighbor matching; Column (2) shows estimates by radius matching; Column (3) shows estimates by nearest-neighbor matching within caliper; Column (4) shows estimates by kernel matching.

. In order to enhance the persuasiveness of the study, we report four sets of results, namely, k-nearest neighbor matching, radius matching, nearest-neighbor matching within calipers, and kernel matching. As shown in Panel 1, HCFC increased the SRH score of the rural elderly by 0.094–0.104, corresponding to an increase of 3.06–3.42% over the level of the group that used polluting fuels. Further, as revealed in Panel 2, HCFC significantly improved the mental health of the elderly. The CES-D score declined by 0.748–0.780, which was equivalent to a 7.96–8.28% reduction with respect to the level of the group using polluting fuels. All of the coefficients of ATT were significant at the 1% level of error. Overall, the estimates were robust to each matching technique, both quantitatively and qualitatively. We thus concluded that the group with HCFC experienced a significantly larger health gain than that of the group without HCFC, indicating a positive effect of clean fuel combustion on health.

4.3.2. Results of the Instrument Variable Strategy

Another potential threat to the validity of the findings above is that there might have been a two-way causal relationship between clean fuel combustion and individual health. To overcome this potential threat, we conducted IV regression for the analysis. As evidenced by the estimation results of the Hausman test, the original hypothesis was significantly rejected, indicating that HCFC was an endogenous explanatory variable. We then used CCFU as an IV for endogeneity testing to illustrate the validity of the IV in several aspects. As shown in

Table 3. Results of IV regression of physical and mental health for rural elderly.

	HCFC	Physical Health		Mental Health
	(1)	(2)	(3)	(4)	(5)	(6)
HCFC (IV: CCFU)	0.923 *** (0.015)	0.159 *** (0.046)	0.070 *** (0.019)	−1.347 *** (0.290)	−0.103 *** (0.021)	−0.206 *** (0.047)
Control variables	Yes	Yes	Yes	Yes	Yes	Yes
R−square	0.343	0.087	0.042	0.132	0.092	0.085
Observation	7959	7959	7959	7959	7959	7959

Note: *** p < 0.01. Column (1) shows the results of the first step of the 2SLS estimation, i.e., the estimate of CCFU on HCFC; the last five Columns show the results of the 2SLS estimation of the CCFU on SRH score, SRH-good, CES-D score, Depression and Loneliness, respectively.

, the IV had a significant positive effect with a coefficient of 0.923, indicating that CCFU had a good explanatory effect on the endogenous variable HCFC. By using the limited information maximum likelihood estimate (LIML), which is more insensitive to weak IVs, it was found that the LIML coefficient estimates were very close to 2SLS, indicating that CCFU was not a weak IV in our study.

We conducted a 2SLS regression based on Model (2), and the results are reported in Table 3. Using CCFU as the IV, we found that the effect of HCFC on physical and mental health did not significantly change, and both were significant at the 1% level of error. Specifically, compared to the results in Table 1, the estimated coefficient of HCFC for the SRH score increased to 0.159, and the estimated coefficient for the CES-D score decreased to −1.347. The results indicated that the explanatory power of HCFC for the health of the elderly was significantly enhanced after the endogeneity problem of the model was mitigated with the IV strategy. To some extent, this also suggested that our finding that the use of clean fuels was beneficial to the health of older adults was stable and robust.

4.3.3. Results of Replacing the Estimation Variables

An additional robustness check was performed to ensure the reliability of the above results. Using alternative measures, we next recoded the SRH score and CES-D score into new variables at different levels. According to the self-reported health status of the respondents, SRH-good was recorded as 1 if the answer was “very good” or “good”; otherwise, it was recorded as 0. Considering that depression is an important component of mental health, with reference to Gao [61], depression was defined as a dichotomous variable, where CES-D score $>10$ was classified as 1, signifying a depressive tendency, and 0 indicated the absence of this tendency. Moreover, we additionally concentrated on the sensation of loneliness, which is pervasive among the elderly and manifests as an unpleasant and distressing subjective psychological experience. In keeping with the CES-D score assessment method, this indicator was based on the frequency of respondents’ experience of loneliness over the last week, ranging from 0 for rarely to 3 for most of the time.

As shown in Table 2, we replaced the explanatory variables of the SRH score and CES-D score. The results of PSM showed that the elderly who chose to use clean fuels were more likely to have good perceptions of health, and the coefficients of ATT were significantly increased by 0.032–0.034. In addition, HCFC significantly decreased the risk of having depression by 4.8–5.0 percentage points, and it also significantly reduced the frequency of feelings of loneliness in the last week. As shown in Table 3, the results of IV regression remained unchanged, and the estimated coefficients were significant at the 1% level of error. As it turns out, all of the robustness checks were basically identical to what we found previously, which reinforced the finding that the health effects of HCFC among older adults in rural China were significantly positive.

4.4. Heterogeneous Effects of Household Clean Fuel Combustion

On the basis of the regression results of PSM,

Table 4. Results of PSM of SRH score and CES-D score for different population subgroups.

	SRH Score				CES-D Score
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Panel 1: by age
Age < 75 (n = 7247)	0.115 *** (0.030)	0.099 *** (0.028)	0.115 *** (0.030)	0.106 *** (0.027)	−0.913 *** (0.194)	−0.814 *** (0.180)	−0.912 *** (0.194)	−0.784 *** (0.178)
Age ≥ 75 (n = 712)	0.026 (0.096)	0.040 (0.091)	0.025 (0.094)	0.004 (0.088)	−0.715 (0.745)	−0.674 (−0.605)	−0.530 (0.629)	−0.706 (0.585)
Panel 2: by education level
Illiterate (n = 3793)	0.062 (0.041)	0.059 (0.038)	0.062 (0.041)	0.054 (0.037)	−0.732 *** (0.278)	−0.918 *** (0.258)	−0.712 *** (0.278)	−0.869 *** (0.251)
Educated (n = 4166)	0.145 *** (0.039)	0.126 *** (0.036)	0.145 *** (0.039)	0.129 *** (0.036)	−0.755 *** (0.239)	−0.795 *** (0.222)	−0.753 *** (0.239)	−0.752 *** (0.220)
Panel 3: by chronic diseases status
None (n = 1532)	0.038 (0.063)	0.042 (0.061)	0.033 (0.063)	0.070 (0.058)	−0.419 (0.375)	−0.430 (0.359)	−0.474 (0.376)	−0.4481 (0.342)
Yes (n = 6427)	0.072 ** (0.030)	0.084 *** (0.028)	0.072 ** (0.030)	0.091 *** (0.028)	−0.822 *** (0.208)	−0.831 *** (0.194)	−0.825 *** (0.208)	−0.851 *** (0.190)

Note: *** p < 0.01; ** p < 0.05. Columns (1) and (5) show estimates by k-nearest neighbor matching; Columns (2) and (6) show estimates by radius matching; Columns (3) and (7) show estimates by nearest-neighbor matching within caliper; Columns (4) and (8) show estimates by kernel matching.

presents the estimated health outcomes attributed to HCFC for the subgroups of different ages, education levels, and chronic disease statuses.

In accordance with the WHO’s age classification criteria for old age (please see GBD), we estimated the impact of HCFC across different age groups. As illustrated in Panel 1, HCFC significantly increased the SRH score and decreased the CES-D score among those below 75 years old. However, the health effect of HCFC on those aged 75 years and older was not significant. As Alem found a positive correlation between the educational level and the likelihood of selecting clean fuels [62], our study further analyzed the effect of HCFC on the elderly with different levels of education. Panel 2 reveals that HCFC contributed to a significant enhancement in the SRH score and a decrease in the CES-D score for educated older adults. The findings for the illiterate were mixed, with evidence of mental health benefits but no significant physical health improvements. Considering the high prevalence of chronic diseases among older adults and the attendant health risks, our study estimated the difference in the health effects of HCFC on those with and without chronic diseases. Panel 3 shows that HCFC significantly promoted the physical and mental health of older adults with chronic diseases, with an increase in the SRH score and a decrease in the CES-D score. However, the health effects of HCFC were not significant for those without chronic diseases.

4.5. The Effects of Household Clean Fuel Combustion on Chronic Pain and Social Interaction

In order to obtain a clearer picture of the underlying channels through which using clean fuels may affect the health of older adults, we undertook a further analysis based on Model (4). As mentioned in the theoretical framework, we attempted to test both effects below, including chronic pain and social interaction (please see

Table 5. Results of PSM of chronic pain and social interaction for rural elderly.

	(1)	(2)	(3)	(4)
Panel 1: chronic pain
Feeling pain	−0.173 *** (0.036)	−0.190 *** (0.034)	−0.173 *** (0.036)	−0.197 *** (0.034)
Arm pain	−0.041 ** (0.017)	−0.050 *** (0.015)	−0.041 ** (0.017)	−0.054 *** (0.015)
Back pain	−0.051 *** (0.016)	−0.059 *** (0.015)	−0.051 *** (0.016)	−0.062 *** (0.015)
Knees pain	−0.035 ** (0.017)	−0.038 ** (0.016)	−0.035 ** (0.017)	−0.052 *** (0.014)
Panel 2: social interaction
Social activities	0.052 *** (0.014)	0.044 *** (0.013)	0.052 *** (0.014)	0.047 *** (0.013)
Playing Ma−Jong/chess/cards	0.051 *** (0.010)	0.051 *** (0.009)	0.051 *** (0.010)	0.052 *** (0.009)
Interacting with friends	0.023 * (0.013)	0.022 * (0.012)	0.023 * (0.013)	0.022 * (0.012)
Going to a club	0.016 *** (0.005)	0.015 *** (0.005)	0.016 *** (0.005)	0.016 *** (0.005)

Note: *** p < 0.01; ** p < 0.05; * p < 0.1. Column (1) shows estimates by k-nearest neighbor matching; Column (2) shows estimates by radius matching; Column (3) shows estimates by nearest-neighbor matching within caliper; Column (4) shows estimates by kernel matching. Standard errors are reported in parentheses.

).

Panel 1 reports the PSM estimates of the effects of HCFC on chronic pain. The results revealed a significant difference in chronic pain perception between older adults who used clean fuels and those who relied on polluting fuels. HCFC significantly reduced the overall perception of chronic pain, with an ATT effect estimate of −0.197–−0.173. Specifically, HCFC led to a decrease in arm, back, and knee pain, with ATT effects of −0.054–−0.041, −0.062–−0.051, and −0.052–−0.035, respectively. The findings indicated that HCFC considerably alleviated several chronic pains that might be associated with physically demanding domestic labor that is commonly performed by elderly individuals dwelling in rural areas. In Panel 2, we report the effects of HCFC on social interaction. Overall, HCFC promoted the participation of older adults in social activities. The PSM results showed that HCFC increased the likelihood of the elderly engaging in recreational activities, such as playing Ma-Jong/chess/cards, interacting with friends, and going to a club by 5.1–5.2%, 2.2–2.3%, and 1.5–1.6%, respectively. These results imply that using clean fuels may enable older adults to have more leisure time to engage in social interaction.

5. Discussion

As China prioritizes the construction of an ecological civilization and the realization of “Healthy China 2030”, HCFC has become an essential aspect of achieving these goals due to its multiple health benefits. Our study employed the latest CHARLS data from 2018 to investigate the influence of HCFC on the health of the rural elderly. We adopted a rigorous empirical approach by using PSM, IV, and multiple other methods to ensure the robustness of our results. Through a heterogeneity analysis, we identified distinct beneficiary groups and facilitated the development of targeted policies. We further explored the potential channels that rendered the health benefits of HCFC, thereby providing a comprehensive assessment of the effects and value of the clean energy transition.

After controlling for other variables, our study revealed that the utilization of clean fuels among rural older adults had a health-promoting effect, corroborating the findings of previous studies [63]. In addition, Yang discerned that, compared to other clean fuels, renewable energy sources, such as solar energy and biogas, hold greater potential for improving the health of residents [64]. Our study sheds light on the impact of HCFC on the psychological wellbeing of older people, which appears to be even more significant. The use of clean fuels led to a 3.06% to 3.42% increase in SRH scores and a 7.96% to 8.28% decrease in the CES-D scores among the elderly. Research on the association between fuel types and mental health has become increasingly salient and has revealed that clean fuel combustion can lead to improved sleep quality [65]. Moreover, He found that there are spillover health benefits of cooking energy [66], which explains why rural elderly people derived health advantages from clean fuels even when they did not cook.

The use of clean fuels can lead to meaningful health improvements, especially for those under the age of 75, who are classified as the “young old” by the WHO. These findings concur with those of Song [67], who concluded that clean fuel combustion resulted in greater health benefits for a younger group. The “young old” group is in a stage of life in which physical decline is not apparent, and HCFC may assist in sustaining their wellbeing. However, adults aged 75 or older display more significant physical functional limitations and accelerated physical degeneration [68], which impart certain constraints on their utilization of clean fuels. Hence, while implementing energy transition measures in rural households, efforts should be made to minimize the detrimental health effects on these older adults at an advanced age. The hazards of HAP demonstrate a more pronounced impact among those with less formal education [69], whereas our study indicates that the health benefits of HCFC are notably superior for older people who have received a formal education. This can be partly attributed to the fact that HCFC necessitates a certain level of knowledge and skills. For instance, technical obstacles may curtail the usage of clean fuels among the illiterate and may even compromise the safety of gas stoves. In contrast, older people with formal education not only possess better knowledge regarding the health benefits of modern energy sources [70] but also demonstrate a relatively proficient mastery of clean-burning appliances. Therefore, it is critical to ensure the safety and effectiveness of employing clean fuels among the elderly through measures such as the implementation of lectures and the provision of on-site instruction. Individuals suffering from chronic diseases are uniquely vulnerable to harmful gases and particulate matter [71]. In addition, the financial burden of healthcare costs renders them beholden to a frugal lifestyle, exacerbating their susceptibility to the cycle of “energy–health–poverty” [72]. HCFC engenders a healthier living environment for those afflicted with chronic diseases and concomitantly affords the alleviation of fuel-related “poverty traps”. Given that chronic diseases inflict a grave public health challenge in China, with 190 million older adults contending with chronic diseases, the economic and social benefits of promoting a household fuel transition are boundless.

The mechanism by which smoke generated by the combustion of polluting fuels affects the human body is comparable to the insidious pathogenic process of tobacco smoke. The inflammation present in the lungs spills over into the systemic circulation, resulting in damage to other organs [73]. The inhalation of PM_2.5, which is associated with the impairment of cognitive and emotional functions in the brain, has been suggested as an explanation for the relationship between air pollution and depression [74]. Furthermore, our study explores the social implications of HCFC and suggests that it can alleviate chronic pain symptoms and increase social interaction as potential mechanisms for enhancing health in older adults. The prevalence of chronic pain in developing countries is 4% higher than that in developed countries, with a greater incidence of headaches and back pain [75]. In China, chronic pain affects nearly 50% of individuals aged 49 years or older [76]. As a fundamental component of daily life, HCFC offers a promising solution for markedly reducing the prevalence of chronic pain, particularly arm, back, and knee pain. Additionally, our research reveals that the use of clean fuels has the potential to provide older adults with more leisure time to devote to activities, especially the traditional Chinese pastime of playing Ma-Jong. Significantly, this culturally rich and time-honored activity has been found to serve as an effective intervention tool for alleviating and preventing feelings of loneliness among the elderly [77]. Other indirect pathways, as opposed to a direct one involving inflammation of organs or brain function, have likewise been identified. Liu observed that the use of clean fuels led to a decrease in the frequency of absenteeism stemming from health problems, thus concomitantly alleviating concerns regarding job-related stress and joblessness [78].

Some limitations need to be acknowledged. Firstly, HAP may arise from cooking fuels, tobacco smoke exposure, and daily incense burning exposure, all of which have been shown to be damaging to health [79]. Due to data restrictions, we may not be able to precisely recognize the total health implications of improved household air quality. Secondly, beyond the fuel type, several unobserved indicators, such as cooking frequency, kitchen type, and ventilation frequency, can also influence HAP levels. Again, we were unable to appropriately address these concerns due to the absence of information in the dataset. Furthermore, with China having put measures for promoting the transition to clean energy in rural neighborhoods into practice in recent years, the macro-level health impacts of such policies necessitate further exploration.

6. Conclusions

Our study explores the effectiveness of HCFC in improving health, particularly in terms of the notable enhancement in mental health among older adults in rural China. We also reveal differences in the health benefits of HCFC among various demographic groups. Furthermore, the results demonstrate that HCFC can effectively alleviate chronic pain symptoms and enhance the quality of life for the elderly. Concurrently, older adults have more leisure time to engage in social activities, thus augmenting their societal contributions. Thus, it has become evident that the transition towards clean energy sources generates numerous environmental, health, and social benefits. We used PSM and IV methods to overcome potential endogeneity issues in our study, which are widely applied to address causal relationships and can provide guidance for research in other fields.

Our findings underscore the close relationship between the utilization of clean fuels and the living environment, health, and lifestyle of rural residents. From a policy perspective, our results are of significant importance. Firstly, the use of energy sources in rural households is diverse. The government should strengthen the process of extracting solid fuels to reduce the production of harmful substances. Additionally, measures should be implemented to enhance the affordability of clean energy, such as subsidizing the use of appliances (e.g., air conditioning) to promote their usage. Secondly, policies should be tailored to benefit vulnerable groups, ensuring that they have equitable access to the health benefits of clean fuels. Lastly, the construction of social activity venues in rural areas should be strengthened to encourage social participation and contribute to the active aging process in both natural and social environments.