Ambient PM2.5, residential greenspace, and household healthcare expenditure in Shandong, China

Ambient PM2.5, residential greenspace, and household healthcare expenditure in Shandong, China | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Ambient PM2.5, residential greenspace, and household healthcare expenditure in Shandong, China Siyuan Wang, Zhiwei Xu, Gian Luca Di Tanna, Raksha Pandya-Wood, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9378849/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 7 You are reading this latest preprint version Abstract Background Environmental factors such as air pollution and access to greenspace are increasingly recognised as important determinants of population health and healthcare expenditure. This study examined the relationship between ambient PM2.5, urban greenspace, and household healthcare expenditure in China, and assessed how healthcare spending was concentrated across these gradients of environmental exposure. Methods We conducted a cross-sectional analysis using data from the 6th Health Services Survey in Shandong Province, linking household healthcare expenditure to residential greenspace (NDVI) and ambient PM2.5 at the village level. We used Generalised Linear Mixed Models with village-level random effects to estimate the relationship between household healthcare expenditure and environmental exposures. To examine inequalities in healthcare expenditure, we calculated concentration indices (CIs), quantifying the distribution of healthcare expenditure across levels of greenspace and air pollution. Results A total of 27,603 individuals were included in the analysis. Higher NDVI exposure was associated with lower household medical expenditure (Q2: β = −0.21, p < 0.01; Q3: β = −0.21, p < 0.01; Q4: β = −0.26, p < 0.01), whereas higher ambient PM2.5 concentrations were linked to increased expenditure (Q2: β = 0.13, p < 0.01; Q3: β = 0.17, p < 0.01; Q4: β = 0.30, p < 0.01). In both models, older age (45–65 and ≥ 65 years) and underweight status were associated with higher costs (NDVI: β = 0.13–0.32, p < 0.01; PM2.5: β = 0.15–0.35, p < 0.01), as were households earning above 60% of median income (NDVI: β = 0.10, p < 0.01; PM2.5: β = 0.11, p < 0.01). Health insurance was linked to lower expenditure in the NDVI model (β = −0.10, p < 0.001) but not in the PM2.5 model. Inequality analyses indicated that household healthcare expenditure was disproportionately concentrated among residents with lower greenspace (NDVI: PCI = − 0.04, 95% CI: −0.05 to − 0.03) and higher PM2.5 exposure (PCI = 0.06, 95% CI: 0.05 to 0.07). Conclusion Our findings highlight the importance of integrating environmental equity into public health and policy interventions to reduce healthcare costs and inequalities. Air pollution Greenspace Household healthcare expenditure Health inequality Figures Figure 1 1. Introduction Environmental factors are increasingly recognised as important determinants of population health and health-related economic outcomes, with growing evidence linking environmental degradation to higher household medical expenditures and widening disparities in health outcomes across populations. 1 – 3 Exposure to environmental determinants, including air pollution and natural environments, has been associated with a wide range of chronic diseases as well as both adverse and beneficial health outcomes. 4 In particular, ambient fine particulate matter (PM2.5) is a major environmental risk factor contributing to cardiometabolic diseases, respiratory conditions, and premature mortality. 5 In contrast, exposure to residential greenspace has been associated with a range of beneficial health outcomes, including improvements in mental health, reduced cardiovascular risk, and decreased all-cause mortality. 6 These environmental determinants not only affect healthcare costs but also contribute to inequities in health burdens. Exposure to harmful environments such as air pollution can lead to higher healthcare utilisation, increased out-of-pocket expenditures, and greater pressure on public health resources. 7 , 8 These burdens are often disproportionately borne by socioeconomically disadvantaged populations, who are more likely to reside in areas with higher pollution levels and limited access to health-promoting environments such as greenspace. 9 As a result, environmental inequalities can exacerbate existing health disparities, reinforcing cycles of poor health and financial hardship. China has experienced rapid economic development and urbanisation over recent decades, accompanied by substantial environmental changes. The burden of chronic diseases has increased substantially, leading to growing healthcare needs and rising healthcare costs. 10 Household out-of-pocket healthcare spending represents a major component of total health expenditure in China and can impose considerable financial pressure on families, particularly among vulnerable populations. 11 To date, several studies have assessed the relationship between household healthcare expenditure and environmental exposures in China for air pollution and urban greenspace. Zhang et al. constructed an environmental quality index encompassing multiple indicators, including PM2.5 concentrations, water quality, and noise levels, to examine the effects of environmental conditions on older adults in China, and found that improvements in environmental quality were associated with a significant reduction in elderly healthcare expenditures. 12 An analysis of the China Health and Retirement Longitudinal Study indicated that air pollution substantially increased inpatient care utilisation among individuals aged 60 and older. 13 Additional studies have reported that higher air pollution significantly elevates household medical costs, disproportionately affecting low-income households and contributing to health inequalities. 8 , 14 , 15 In contrast, empirical evidence on the relationship between greenspace and household medical expenditure remains relatively limited. Evidence from Chengdu suggested that a higher degree of urbanisation, was associated with reductions in household medical expenses. 16 Similarly, another study reported that greater urban greenspace quality, as measured by user satisfaction, was significantly inversely related to healthcare expenditure in Shanghai. 17 International evidence also supports the link between environmental exposures and healthcare costs. 17 In the United States, exposure to fine particulate matter has been associated with increased hospital admissions and higher Medicare expenditures. 18 Studies across European countries have demonstrated that elevated air pollution levels are linked to higher healthcare spending relating to cardiovascular and respiratory diseases. 19 In South and Southeast Asia, exposure to air pollution was associated with increased out-of-pocket healthcare expenditure, disproportionately burdening lower-income households. 20 More importantly, the distributional aspects of household health expenditure across these environmental determinants of health remain poorly understood. Previous studies have assessed environmental exposure inequalities by socioeconomic or income groups, using measures such as the Gini coefficient or concentration index to quantify distributions of PM2.5 and urban greenspace. For example, Song et al. applied the Gini index to quantify inequality in greenspace exposure across 303 cities in China, finding that the majority of cities experienced substantial disparities, with 207 cities exhibiting Gini coefficients greater than 0.6. 21 Cao et al. used the Palma coefficient, as an alternative inequality measure, to assess disparities in residential greenspace exposure within urban areas, reporting that populations with the highest exposure experienced nearly four times the levels observed in the least exposed groups. 22 While these studies provide important evidence on inequalities in environmental exposure across socioeconomic cohorts, they do not address how health-related outcomes are distributed across environmental exposure gradients. To bridge these gaps, this study assesses the relationship between air pollution, urban greenspace, and household healthcare expenditure, and further evaluates the distribution of healthcare expenditure across gradients of air pollution and urban greenspace. 2. Methods 2.1 Study design and setting This study was a cross-sectional analysis using population health data collected in Shandong Province through the sixth-round of the Health Services Survey (HSS) in 2018. In brief, the HSS is a government-administered household survey that collects routine information on household individual’s health status, health care utilisation, quality of life, and household healthcare expenditures. The survey used a stratified multistage cluster sampling design, in which administrative units were stratified by geographic region and urban–rural status. All eligible members of selected households were invited to participate in face-to-face interviews administered by trained professionals. For the 6th HSS in Shandong Province, twenty districts were selected, and within each district, two villages were sampled from each of five townships, yielding 10 villages per district. Sixty households were selected from each village, resulting in a total of 35,262 individuals from 12,938 households interviewed across more than 150 counties and cities. Individuals with missing values were excluded from the analysis. We chose Shandong Province due to its large and demographically diverse population of over 100 million residents, spanning a wide range of urban and rural settings that reflect substantial socioeconomic and environmental heterogeneity. Supplementary Table 1 provides a summary of the socioeconomic and environmental indicators for the prefectural-level cities sampled in the 6th Health Services Survey of Shandong Province. Despite a general declining trend, annual average PM2.5 concentrations in Shandong Province exhibited pronounced spatial variation, with the highest levels (> 40 µg/m³) concentrated in western plains such as Liaocheng, Heze, Dezhou, and Jining, moderately high levels (33–40 µg/m³) in central, southern, and northern cities, and the lowest levels (< 30 µg/m³) in the Jiaodong Peninsula. 23 The province has a high urbanisation level (67%) and urban population density (647 people/km²), with above-average urban greening, including 78,676 hectares of parks and 326,340 hectares of total urban green space (1.7% of the province’s land area). 24 Further, the province’s landscape is characterised by diverse geographic features, including plains, hills, and coastal regions, contributing to substantial variation in land cover and greenspace density. Heterogeneity in land cover creates substantial variation in coastal–inland differentiation of greenspace morphology, cover, and density, with fractional vegetation cover exhibiting a clear inland‑to‑coast gradient. 25 Together, this provides an opportunity to assess the relationship between environmental exposures and household medical expenditure, as well as inequalities in the distribution of healthcare costs across these factors. To ensure transparent and standardised reporting, this analysis was conducted following the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) checklist for observational studies (Supplementary table 2). 2.2 Health status and household medical expenditure Self-reported household medical expenditure was collected from respondents based on household spending over the previous year, including out-of-pocket payments for inpatient and outpatient visits, medications, medical devices, diagnostic tests and other healthcare services, but excluding payments for health care products and non-medical services. Health status information was collected through self-reported responses, where participants were asked whether they had been diagnosed by a medical doctor, or undergone treatment for specific conditions within the past six months. Reported conditions were subsequently classified according to the questionnaire’s prespecified list, including chronic conditions such as cardiometabolic diseases (CMDs), cancer, respiratory diseases (RDs), and neuropsychiatric disorders (NPDs). A full list of conditions is provided in Supplementary table 3. 2.3 Residential greenspace and air pollution We used Normalised Difference Vegetation Index (NDVI) measured from the Terra Moderate Resolution Imaging Spectroradiometer Vegetation Indices (MOD13Q1), which provides repeated measures of surface vegetation at 250 m spatial resolution at 16-day intervals. 26 NDVI is a validated proxy for vegetation cover and has been widely applied in population environmental exposure studies to quantify greenspace exposure. 27 Values range from − 1 to 1, with higher values indicating denser vegetation. Residential greenness exposure for each participant was quantified by calculating the three-year average of NDVI values within a 1 km radius of their geo-coded village-level residence. Air pollution exposure was estimated using ground-level concentrations of PM2.5 derived from the China High Air Pollutants (CHAP) dataset. 28 Briefly, The dataset provides PM2.5 measurements at a spatial resolution of 0.01°, generated using a space–time extremely randomised trees (STET) model. This model integrates multiple data sources, including satellite-derived aerosol optical depth (AOD), meteorological variables, land-use and topographic data, and emission inventories. 28 Using a similar method to NDVI, residential PM2.5 exposure were estimated as the mean concentration within a 3 km buffer around each respondent’s geo-coded village over the five years prior to the survey. 2.4 Covariates To control for potential confounding at the individual level, models were adjusted for a set of sociodemographic and health-related characteristics. These included age ( 65 years), sex (males or females), educational attainment (primary school, junior high school, high school or above ), residential setting (urban or rural), marital status (never married, married or other), annual household income (categorised as above or below 60% of the 2018 median annual income of 24,336 CNY, representing relative poverty 29 ), insurance status (insured or not insured), and body mass index (BMI; <18.5 kg/m², ≥ 18.5 kg/m² and < 25 kg/m², and ≥ 25 kg/m²). 2.5 Statistical analysis We employed Generalised Linear Mixed Models (GLMMs) with village-level random effects to assess the relationship between environmental exposures (NDVI and PM2.5) and household medical expenditure. To allow for potential nonlinear exposure–response relationships, both NDVI and PM2.5 were categorised into quartiles. GLMMs have been widely applied in studies to assess the relationships between environmental exposures, such as air pollution and greenness, and health outcomes. Village-level random effects were specified to account for the hierarchical sampling structure and potential clustering within villages. Specifically, we fitted two GLMMS: Model 1 assessed the association between healthcare expenditure and residential greenspace, whereas Model 2 assessed the association between healthcare expenditure and ambient air pollution. Both models were adjusted for individual-level covariates, including age, sex, urban/rural residence, marital status, insurance status, education, income, and BMI. To examine inequalities in household healthcare expenditure in relation to residential greenspace and air pollution, we applied the concept of the Concentration Index (CI). CI has been widely used in health economics to quantify the degree of socioeconomic-related inequality in health outcomes. 30 In our analysis, each individual was assigned a fractional rank based on their mapped NDVI and PM2.5 values. The CI was then calculated as: $$\:CI=\frac{2}{\mu\:}\:Cov(y,R)$$ Where 𝑦 is self-reported healthcare expenditure, R is the fractional rank of the population by NDVI or PM2.5 exposure, and \(\:\mu\:\) is the mean of 𝑦. The CI is interpreted as twice the area between the Concentration Curve (CC), which plots the cumulative share of household healthcare expenditure against the cumulative share of the population ranked by environmental exposure, and the line of equality (the 45-degree line). It ranges from − 1 to 1, with − 1 indicating total concentration among the least exposed, 1 indicating total concentration among the most exposed, and 0 indicating perfect equality. As the standard CI is unadjusted and may be influenced by differences in population characteristics, Partial Concentration Indices (PCI) were computed to assess inequality of expenditure across environmental exposures after accounting for covariates. The PCI quantifies inequality in a health outcome with respect to a ranking variable after removing the influence of standardising variables (e.g., age and sex) that are correlated with the ranking variable but are not themselves policy amenable. 31 We employed the direct standardisation concept as described by Gravelle et al,. in which the outcome is adjusted using a regression model that includes both the ranking variable and standardising covariates. 31 Subgroup analyses were conducted to assess how inequalities in household medical expenditure across environmental exposures varied by chronic disease class, as defined in the HSS (Supplementary Table 3). The analysis focused on respiratory diseases, cardiometabolic diseases, and neuropsychiatric disorders, which were selected due to their high prevalence and the well-established links between these conditions and environmental factors such as residential greenspace and air pollution. 32 – 35 In addition, sensitivity analyses were conducted to assess the robustness of the associations between household medical expenditure and environmental exposures, using annual PM2.5 concentrations measured at 4 km, 5 km, and 10 km buffers, and NDVI across 2–5 km buffers. 3. Results 3.1 Sample characteristics A total of 27,603 individuals were included in this analysis. The flowchart of study participants selection is presented in Supplementary Fig. 1. Table 1 presents the characteristics of the study population stratified by chronic disease status. Chronic disease prevalence was higher among participants aged ≥ 65 years, females, those who were married, those who have no health insurance, individuals with lower educational attainment or household income, and those with higher body mass index. Residential greenness, measured by NDVI, ranged from 0.081 to 0.420 (mean 0.229, SD 0.082), while ambient PM2.5 concentrations ranged from 18.39 to 45.77 µg/m³ (mean 36.88, SD 6.47). Median household medical expenditure was 3,000 RMB (IQR 5,000 RMB) and was higher among participants with chronic conditions than among those without. Overall, 7,783 participants (28% of the study population) had at least one chronic condition, with 3,083 experiencing cardiometabolic diseases, 401 experiencing respiratory diseases, and 261 experiencing neuropsychiatric disorders. Table 1 General characteristics of the sample population Age Total (N = 27,603) Chronic disease (N = 7,783) No chronic disease (N = 19,820) P value* < 0.01 Below 45 years 10,948 683 10,265 45 to 65 years 11,454 4,061 7,393 Above 65 years 5,201 3,039 2,162 Sex < 0.01 Male 13,273 3,640 9,633 Female 14,330 4,143 10,187 Marital status < 0.01 Never Married 2,412 123 2,289 Married and other 25,191 7,660 17,531 Education < 0.01 Primary school 8,820 3,973 4,847 Junior high school 9,815 2,386 7,429 High school or above 8,968 1,424 7,544 Annual household income < 0.01 Less than 60% of median household income (14,601 CNY) 5,521 2,534 2,987 More than 60% of median household income (14,601 CNY) 22,082 5,249 16,833 Body mass index (BMI) < 0.01 Under weight ( 25 kg/m²) 10,864 3,793 7,071 Household healthcare expenditure < 0.01 RMB 5,000 7,669 3,057 4,612 Insurance status < 0.01 Insured 13,963 3,733 10,230 Not insured 13,640 4,050 9,590 NDVI < 0.01 \(\:{NDVI}_{Q1}\) [0.081–0.153) 6,869 1,766 5,103 \(\:{NDVI}_{Q2}\) [0.153–0.238) 6,901 1,788 5,113 \(\:{NDVI}_{Q3}\) [0.238–0.290) 6,978 2,008 4,970 \(\:{NDVI}_{Q4}\) [0.290–0.420] 6,855 2,221 4,634 PM2.5 < 0.01 \(\:{PM2.5}_{Q1}\:\) [18.39–35.15) 7,282 1,995 5,287 \(\:{PM2.5}_{Q2}\:\) [35.15–38.86) 6,669 1,822 4,847 \(\:{PM2.5}_{Q3}\:\) [38.86–41.27) 6,843 1,922 4,921 \(\:{PM2.5}_{Q4}\:\) [41.27–45.77] 6,809 2,044 4,765 * P-values were obtained under the Χ² test. 3.2 Household medical expenditure, residential greenspace and air pollution Table 2 presents the associations between self-reported annual medical expenditure and residential greenspace. After adjusting for age, sex, marital status, region, income, education, insurance status, and BMI, higher residential greenspace exposure was associated with lower annual household medical expenditure. Compared with participants in the lowest NDVI quartile (Q1), those in Q2, Q3, and Q4 had significantly lower healthcare expenditure (β = −0.21, p < 0.001; β = −0.21, p < 0.001; and β = −0.26, p < 0.001, respectively. Older age was associated with higher medical expenditure, with participants aged 45–65 years (β = 0.13, p 65 years (β = 0.32, p < 0.001) incurring greater costs compared with those aged < 45 years. Households with income above the relative poverty line incurred higher healthcare expenditure (β = 0.10, p < 0.001). Compared with normal-weight individuals, underweight (β = 0.17, p < 0.001) and overweight or obese individuals (β = 0.03, p = 0.05) had higher medical expenditure. Being married was associated with higher household healthcare expenditure (β = 0.07, p = 0.01), whereas residing in a rural area was associated with lower medical expenditure (β = −0.05, p = 0.04), as was having health insurance (β = −0.10, p < 0.001). Table 2 Association between greenspace exposure and self-reported household medical expenditure Variables β (95% CI) P value Sex (base = Female) Male -0.01 0.51 Age (base = < 45 years) 45–65 years 0.13 65 years 0.32 < 0.001 Marriage (base = Never married) Married 0.07 0.01 Region (base = Urban) Rural -0.05 0.04 Annual household income (base = Less than 60% of median household income) More than 60% of median income 0.10 < 0.001 Education (base = Primary school) Junior high school -0.03 0.08 High school or above 0.03 0.21 BMI (base = Normal weight, 18.5–24.9 kg/m²) Underweight (< 18.5 kg/m²) 0.17 < 0.001 Overweight or obese (≥ 25.0 kg/m²) 0.03 0.05 Insurance status (base = No insurance) Insured -0.10 < 0.001 NDVI (base = Q1, [0.081–0.153)) Q2 [0.153–0.238) -0.21 < 0.001 Q3 [0.238–0.290) -0.21 < 0.001 Q4 [0.290–0.420] -0.26 < 0.001 Table 3 summarises the association between annual household medical expenditure and PM2.5 exposure. After adjustment for the same set of covariates, higher PM2.5 exposure was consistently associated with increased healthcare expenditure. Compared with the lowest quartile, participants in the second, third, and fourth PM2.5 quartiles had progressively higher expenditures (β = 0.13, p < 0.001; β = 0.17, p < 0.001; and β = 0.30, p < 0.001, respectively). Older age was associated with increased expenditure, with participants aged 45–65 years (β = 0.15, p 65 years (β = 0.35, p < 0.001) having higher costs compared with those aged < 45 years. Household income above 60% of the median was associated with higher expenditure (β = 0.11, p < 0.001). Underweight individuals had higher medical expenditure (β = 0.16, p < 0.001), and overweight or obese individuals had a smaller but significant increase (β = 0.03, p = 0.03). Being married was associated with higher expenditure (β = 0.06, p = 0.04), whereas residing in rural areas was associated with lower expenditure (β = −0.07, p = 0.001). Male sex, education level, and health insurance coverage were not significantly associated with household healthcare expenditure. Sensitivity analyses using alternative spatial buffers indicated that the association between residential greenspace and household medical expenditure was generally robust across NDVI buffers of 2–4 km, with the exception of the highest quartile at the 4 km buffer. Relative to the lowest quartile (Q1), higher levels of residential greenspace were consistently associated with lower annual household medical expenditure across these NDVI measurements. However, at the 5km radius buffer, the association was not statistically significant (Supplementary table 4). In contrast, sensitivity analyses under alternative PM2.5 buffer sizes suggested no statistically significant differences in household medical expenditure across exposure quartiles, except for Q3 in the 10 km buffer model (Supplementary table 5). Table 3 Association between PM2.5 and self-reported household medical expenditure Variables β (95% CI) P value Sex (base = Female) Male -0.02 0.23 Age (base = < 45 years) 45–65 years 0.15 65 years 0.35 < 0.001 Marriage (base = Never married) Married 0.06 0.04 Region (base = Urban) Rural -0.07 0.001 Annual household income (base = Less than 60% of median household income) More than 60% of median income 0.11 < 0.001 Education (base = Primary school) Junior high school -0.02 0.30 High school or above 0.06 0.01 BMI (base = Normal weight, 18.5–24.9 kg/m²) Underweight (< 18.5 kg/m²) 0.16 < 0.001 Overweight or obese (≥ 25.0 kg/m²) 0.03 0.03 Insurance status (base = No insurance) Insured 0.01 0.89 PM2.5 (base = Q1, [18.39–35.15)) Q2 [35.15–38.86) 0.13 < 0.001 Q3 [38.86–41.27) 0.17 < 0.001 Q4 [41.27–45.77] 0.30 < 0.001 3.3 Concentration of annual household medical expenditure by environmental exposures Figure 1 presents the concentration curve of healthcare expenditure ranked by levels of greenspace and PM2.5 exposure. Healthcare expenditure was marginally disproportionately concentrated among residents with lower greenspace exposure (CI = − 0.02; 95% CI: −0.04, − 0.01) and higher PM2.5 exposure (CI = 0.05; 95% CI: 0.03, 0.06). Supplementary Figs. 2–4 present the results of subgroup analyses by disease category. For CMDs, annual medical expenditure was more concentrated among individuals with lower greenspace (CI = − 0.04; 95% CI: −0.07, − 0.02) and higher PM2.5 exposure (CI = 0.03; 95% CI: 0.01, 0.06). For NPDs, healthcare costs were disproportionately concentrated among residents with lower greenspace (CI = − 0.08; 95% CI: −0.15, − 0.01) but did not exhibit a clear pattern with PM2.5 exposure (CI = − 0.02; 95% CI: −0.13, 0.06). In contrast, medical costs among individuals with respiratory disease showed no clear distributional pattern across levels of residential greenspace or PM2.5 exposure (CI = − 0.02; 95% CI: −0.10, 0.07 and CI = 0.02; 95% CI: −0.07, 0.12, respectively). After adjusting for covariates using the direct standardisation method, the concentration of inequality was marginally greater than the unadjusted model. Specifically, household medical expenditure remained disproportionately concentrated among individuals with lower greenspace (PCI = − 0.04; 95% CI: −0.05, − 0.03) and higher PM2.5 exposure (PCI = 0.06; 95% CI: 0.05, 0.07). For CMDs, expenditures were concentrated among residents with lower greenspace exposure (PCI = − 0.04; 95% CI: −0.05, − 0.03) and higher PM2.5 (PCI = 0.03; 95% CI: 0.02, 0.04). NPD medical expenditures remained concentrated among individuals with lower greenspace exposure (PCI = − 0.06; 95% CI: −0.08, − 0.03) but not for PM2.5 exposure (PCI = 0.01; 95% CI: −0.02, 0.01). For respiratory diseases, medical expenditure was not strongly concentrated across either environmental exposure, with a partial concentration index of − 0.01 (95% CI: −0.03, 0.01) for greenspace exposure and 0.01 (95% CI: −0.02, 0.02) for PM2.5 exposure. 4. Discussion This study examined the relationships between environmental exposures and household healthcare expenditure and assessed how medical costs were distributed across gradients of air pollution and urban greenspace. We found that higher levels of medical expenditure were significantly associated with greater exposure to PM2.5 and lower levels of residential greenness. Inequality analyses based on concentration indices indicated a statistically significant but marginal distribution of healthcare expenditure across environmental gradients. Specifically, higher medical expenditure was concentrated among individuals experiencing higher PM2.5 concentrations and lower levels of residential greenspace exposure, after accounting for potential confounding factors. When examined by disease category, household healthcare expenditure for people reporting cardiometabolic diseases were concentrated among residents experiencing both lower levels of greenspace and higher exposure to PM2.5. Medical expenditures among individuals reporting neuropsychiatric disorders were disproportionately concentrated in areas with lower residential greenspace, whereas no clear pattern was observed across PM2.5 exposure levels. In contrast, respiratory disease expenditures did not exhibit a clear distributional pattern with either environmental exposure. While the observed associations between environmental exposures and healthcare expenditure are broadly consistent with existing evidence linking air pollution and greenspace to medical expenditure, this study extends prior research in several aspects. First, a major strength of this study is the integration of individual-level health data with high-resolution environmental exposure data, enabling a more comprehensive assessment of how environmental factors relate to household medical expenditure within an equity framework. We examined the distribution of household medical expenditure across gradients of PM2.5 and residential greenspace exposure, an approach that has not been applied in previous studies. We observed a modest but statistically significant concentration of healthcare expenditure among individuals exposed to higher levels of PM2.5 and lower levels of greenspace, after adjustment for sociodemographic and health-related confounders. This pattern likely reflects the cumulative health impacts of environmental disadvantage, whereby populations living in areas with elevated air pollution and limited greenspace are more susceptible to chronic conditions, particularly cardiometabolic and neuropsychiatric disorders, which necessitate ongoing medical care and contribute to higher healthcare costs. Second, while previous research on air pollution and healthcare expenditure has largely focused on older populations, 13, 36, 37 the present study examined these relationships in a broader population-based sample, providing evidence that air pollution may influence healthcare costs across the general population. Our findings suggested that individuals exposed to higher PM2.5 levels incurred progressively greater medical expenditure, with those in the second, third, and highest quartiles experiencing approximately 14%, 19%, and 35% higher costs, respectively, compared with individuals in the lowest quartile. When assuming a linear relationship, our results are comparable to previous national evidence, which reported 2.9% and 5.7% increases in medical expenditure per unit increase in PM2.5 concentrations. 8 , 38 Third, this study contributes to the limited evidence on the relationship between greenspace and medical expenditure in China. One previous study assessed the relationship between urban greenspace coverage and medical expenditure, finding that a 10% increase in urban green space was associated with a 0.012% reduction in residents’ medical costs. 39 However, the study measured greenspace exposure as aggregate area (hectares), which inadequately reflects vegetation greenness, density, or health characteristics. NDVI and other spatially resolved metrics have been widely used in epidemiological greenspace studies to address this limitation. 40 Another study conducted in Chengdu found a statistically significant association between greenspace, measured using NDVI, and inpatient medical costs. 16 This study builds on previous research by addressing key limitations in how greenspace exposure is measured and extends the analysis to examine its relationship with household healthcare expenditure. These results reinforce the economic rationale for public investment in greenspace. Beyond its well-documented health and environmental benefits, our findings demonstrate that strategic investment in urban greening can generate measurable economic returns by reducing household medical expenditures, while also promoting equitable access to environmental resources and protecting vulnerable populations from catastrophic healthcare costs. This underscores the potential of urban greening as a cost-effective strategy to enhance population health, while promoting environmental equity and household financial security. Our findings provide empirical evidence supporting China’s ongoing policy efforts to improve air quality and expand access to urban greenspace. China has demonstrated a strong commitment to improving environmental quality through a series of national and regional policy initiatives. Measures such as the Clean Air Action Plan and the Three-Year Blue Sky Defence Plan have established stricter standards for major air pollutants, contributing to measurable improvements in population health as well as economic benefits, including reductions in household healthcare expenditures. 41 – 43 Similarly, greenspace planning has been progressively incorporated into China’s national urban development framework, with key targets for greenspace coverage and per-capita availability embedded within the national Green Space System Planning framework. 44 These policy initiatives reflect a growing recognition of the importance of environmental determinants in influencing health and healthcare utilisation. Further, our results suggest that future policy development could be benefited by stronger considerations for equity. While existing policies have primarily emphasised overall improvements in air quality and greenspace provision, less attention has been given to the distribution of environmental benefits and burdens across populations. In particular, greenspace policies have prioritised aggregate coverage targets over equality or accessibility. 44 As China continues its efforts to fulfil its commitment to the Sustainable Development Goals, such as providing equitable access to quality greenspace and substantially reducing deaths and illnesses from air pollution, our findings highlight the importance of addressing environmental determinants of health to reduce the financial burden of healthcare. As exposure —or lack of exposure —to these environmental determinants can contribute to the exacerbation of major chronic and acute illnesses, resulting in catastrophic healthcare expenditures, improving air quality and access to greenspace could help reduce financial risks for households, support poverty alleviation, and advance China’s goals for universal health coverage. Our finding suggest that future urban environmental planning should place greater emphasis on the quality, accessibility, and spatial distribution of greenspace, alongside continued efforts to reduce air pollution, to reduce disparities in healthcare expenditure associated with unequitable environmental exposures. This study has several limitations. First, the cross-sectional design limited any inference of causality between environmental exposures and medical expenditure, as well as limiting the understanding of the long-term health impacts of environmental exposures on medical expenditures. Second, although we adjusted for key demographic and socioeconomic factors, residual confounding by unmeasured individual characteristics, such as underlying health status, local healthcare infrastructure, or healthcare-seeking behaviours, as well as other environmental factors, including noise, urban heat, or traffic density, could influence the observed associations. Third, our analysis examined the relationship and distribution of medical expenditure with air pollution and greenspace separately, without accounting for their potential joint or interactive effects. This was mainly due to spatial correlation and multicollinearity between air pollution and greenspace, leading to unstable estimates and limiting the ability to examine their independent contributions. Fourth, we relied on self-reported medical expenditure, which, despite the HSS being carried out by trained professionals, may be subject to recall bias and reporting errors. This limitation also applies to health conditions, particularly for under-diagnosed diseases such as diabetes and hypertension. Under-reporting of these conditions may be more pronounced among rural residents or populations with limited access to healthcare, potentially attenuating observed associations between environmental exposures, such as greenspace, and health outcomes. Consequently, the true magnitude of inequities in healthcare expenditure across environmental gradients may be even greater than estimated in this study. Fifth, our findings were based on analysing a representative cohort for Shandong province, a region with relatively higher socioeconomic development compared with the national average. Consequently, the observed associations and distributional patterns of medical expenditure may not be generalisable to the national population and warrants further research. 5. Conclusion Our study demonstrated that environmental factors are closely linked to household healthcare expenditure in China, with greater residential greenspace associated with lower costs and higher air pollution linked to increased costs. Medical spending is disproportionately borne by populations living in areas with less greenspace and higher PM2.5 levels, reflecting inequities in health and economic burden created by geographical inequities in environmental risks. These findings highlight the need for public health and policy strategies that incorporate environmental quality improvements, focusing particularly on regions which experience the dual burden of higher environmental risks and health care costs. Abbreviations AOD Aerosol Optical Depth BMI Body Mass Index CC Concentration Curve CHAP China High Air Pollutants CIs Confidence Intervals CMD Cardiometabolic Disease GLMMs Generalised Linear Mixed Models Health Services Survey Health Services Survey MOD13Q1 Moderate Resolution Imaging Spectroradiometer Vegetation Indices NDVI Normalised Difference Vegetation Index NPD Neuropsychiatric Disorders PCI Partial Concentration Indices PM2.5 Particulate Matter ≤ 2.5 µm RDs Respiratory Diseases STET Space – Time Extremely Randomised Trees STROBE Strengthening the Reporting of Observational Studies in Epidemiology Declarations Ethics approval and consent to participate This study was a secondary analysis of de-identified data from the Shandong Health Services Survey and informed consent was obtained from all participants. The study was approved by the UNSW Human Research Ethics Committee (HC230209). Availability of data and materials The data that support the findings of this study are available from the corresponding author upon reasonable request. Competing interests The authors report no conflict of interest. The authors alone are responsible for the content and writing of the paper. Funding This study was funded by the National Natural Science Foundation of China (grant number: 71874086, 72174093, 72474109). Authors’ contributions Conceptualization: SW, LS, SJ Data curation: SW, LS Formal analysis: SW, LS Methodology: SW, LS, SJ Resources: MC, LS Supervision: LS Writing – original draft: SW Writing – reviewing & editing: SW, LS, MC, ZX, RP, LD, GLDT, SJ References Jerrett M, Eyles J, Dufournaud C, Birch S. Environmental influences on healthcare expenditures: an exploratory analysis from Ontario, Canada. J Epidemiol Community Health. 2003; 57:334–8. Anwar A, Hyder S, Bennett R, Younis M. Impact of Environmental Quality on Healthcare Expenditures in Developing Countries: A Panel Data Approach. Healthcare (Basel). 2022; 10. Szymańska A. The relationship between health expenditure, income, and environmental degradation: Evidence from OECD economies. Economic Analysis and Policy. 2025; 87:2183–201. Wu C, Liu J, Li Y, Qin L, Gu R, Feng J, et al. Association of residential air pollution and green space with all-cause and cause-specific mortality in individuals with diabetes: an 11-year prospective cohort study. EBioMedicine. 2024; 108:105376. Sangkham S, Phairuang W, Sherchan SP, Pansakun N, Munkong N, Sarndhong K, et al. An update on adverse health effects from exposure to PM2.5. Environmental Advances. 2024; 18:100603. Yang B-Y, Zhao T, Hu L-X, Browning MHEM, Heinrich J, Dharmage SC, et al. Greenspace and human health: An umbrella review. The Innovation. 2021; 2:100164. Bailie CR, Ghosh JKC, Kirk MD, Sullivan SG. Effect of ambient PM(2.5) on healthcare utilisation for acute respiratory illness, Melbourne, Victoria, Australia, 2014-2019. J Air Waste Manag Assoc. 2023; 73:120–32. Yang J, Zhang B. Air pollution and healthcare expenditure: Implication for the benefit of air pollution control in China. Environment International. 2018; 120:443–55. Hajat A, Hsia C, O'Neill MS. Socioeconomic Disparities and Air Pollution Exposure: a Global Review. Curr Environ Health Rep. 2015; 2:440–50. Liu H, Yin P, Qi J, Zhou M. Burden of non-communicable diseases in China and its provinces, 1990-2021: Results from the Global Burden of Disease Study 2021. Chin Med J (Engl). 2024; 137:2325–33. Qiu T, Shi L, Zhou W, Deng J, Sun G. From individual burden to social risk pooling: drivers of the declining out-of-pocket health expenditure share in China (2009-2024). Int J Equity Health. 2026; 25:33. Zhang Y, Chen S, Liu D. A measurement study of the environmental quality and medical expenditures of elderly individuals: causal inference based on machine learning. Archives of Public Health. 2024; 82:195. Jin B, Li Z. Air pollution, healthcare use, and inequality: Evidence from China. Economic Modelling. 2024; 141:106905. Wang E, Zhu M, Lin Y, Xi X. Multilevel medical insurance mitigate health cost inequality due to air pollution: Evidence from China. International Journal for Equity in Health. 2024; 23:153. Zhou L, Zhong Q, Yang J. Air Pollution and Household Medical Expenses: Evidence From China. Frontiers in Public Health. 2022; Volume 9 - 2021. Seyler BC, Luo H, Wang X, Zuoqiu S, Xie Y, Wang Y. Assessing the impact of urban greenspace on physical health: An empirical study from Southwest China. Front Public Health. 2023; 11:1148582. Zhang L, Wu Y. Negative Associations between Quality of Urban Green Spaces and Health Expenditures in Downtown Shanghai. Land [serial on the Internet]. 2022; 11(8). Wei Y, Wang Y, Di Q, Choirat C, Wang Y, Koutrakis P, et al. Short term exposure to fine particulate matter and hospital admission risks and costs in the Medicare population: time stratified, case crossover study. Bmj. 2019; 367:l6258. Zhang C, Zhang L. The relationship between toxic air pollution, health expenditure, and economic growth in the European Union: fresh evidence from the PMG-ARDL model. Environmental Science and Pollution Research. 2024; 31:21107–23. Jaafar H, Razi NA, Azzeri A, Isahak M, Dahlui M. A systematic review of financial implications of air pollution on health in Asia. Environ Sci Pollut Res Int. 2018; 25:30009–20. Song Y, Chen B, Ho HC, Kwan M-P, Liu D, Wang F, et al. Observed inequality in urban greenspace exposure in China. Environment International. 2021; 156:106778. Cao Y, Li G. Extensive inequality of residential greenspace exposure within urban areas in China. Science of The Total Environment. 2024; 948:174625. Shandong Provincial Government. Shandong Province ecological environment status report. Shandong Provincial Bureau of Statistics. Statistical Yearbook. Available from: http://tjj.shandong.gov.cn/col/col6279/index.html. Yu Y, Liu D, Hu S, Shi X, Tang J. Spatiotemporal Heterogeneity of Vegetation Cover Dynamics and Its Drivers in Coastal Regions: Evidence from a Typical Coastal Province in China. Remote Sensing [serial on the Internet]. 2025; 17(5). Xiong C, Ma H, Liang S, He T, Zhang Y, Zhang G, et al. Improved global 250 m 8-day NDVI and EVI products from 2000-2021 using the LSTM model. Sci Data. 2023; 10:800. Gascon M, Cirach M, Martínez D, Dadvand P, Valentín A, Plasència A, et al. Normalized difference vegetation index (NDVI) as a marker of surrounding greenness in epidemiological studies: The case of Barcelona city. Urban Forestry & Urban Greening. 2016; 19:88–94. Wei J, Li Z, Lyapustin A, Wang J, Dubovik O, Schwartz J, et al. First close insight into global daily gapless 1 km PM2.5 pollution, variability, and health impact. Nature Communications. 2023; 14:8349. Li Q, Dong L, Zhang L. Have pensions reduced the relative poverty? ----- empirical analysis from China CHARLS data. Heliyon. 2023; 9:e22711. Erreygers G, Van Ourti T. Measuring socioeconomic inequality in health, health care and health financing by means of rank-dependent indices: a recipe for good practice. J Health Econ. 2011; 30:685–94. Gravelle H. Measuring income related inequality in health: standardisation and the partial concentration index. Health Econ. 2003; 12:803–19. Twohig-Bennett C, Jones A. The health benefits of the great outdoors: A systematic review and meta-analysis of greenspace exposure and health outcomes. Environmental Research. 2018; 166:628–37. Wang S, Sun J, Xu Z, Di Tanna GL, Chen M, Downey LE, et al. Residential greenspace and multiple chronic health conditions in China: a cross-sectional study. J Glob Health. 2025; 15:04218. Dominski FH, Lorenzetti Branco JH, Buonanno G, Stabile L, Gameiro da Silva M, Andrade A. Effects of air pollution on health: A mapping review of systematic reviews and meta-analyses. Environmental Research. 2021; 201:111487. Werder E, Lawrence K, Deng X, Braxton Jackson W, Christenbury K, Buller I, et al. Residential air pollution, greenspace, and adverse mental health outcomes in the U.S. Gulf Long-term Follow-up Study. Science of The Total Environment. 2024; 946:174434. Zhao Y, Yang S, Zhang X. What does poor air quality lead to? - The influence of air pollution on elderly medical expenditures in China. Cities. 2026; 169:106543. Pi T, Wu H, Li X. Does Air Pollution Affect Health and Medical Insurance Cost in the Elderly: An Empirical Evidence from China. Sustainability [serial on the Internet]. 2019; 11(6). Hou Z, Kim MJ, Zhang N. Air pollution, defensive behaviors, and medical expenditures. Journal of Asian Economics. 2025; 98:101931. Bi Y, Wang Y, Yang D, Mao J, Wei Q. Urban green spaces and resident health: an empirical analysis from data across 30 provinces in China. Frontiers in Public Health. 2024; Volume 12 - 2024. Sadeh M, Brauer M, Dankner R, Fulman N, Chudnovsky A. Remote sensing metrics to assess exposure to residential greenness in epidemiological studies: A population case study from the Eastern Mediterranean. Environment International. 2021; 146:106270. Weng Z, Tong D, Wu S, Xie Y. Improved air quality from China's clean air actions alleviates health expenditure inequality. Environ Int. 2023; 173:107831. ZHANG J, JIANG H, ZHANG W, MA G, WANG Y, LU Y, et al. Cost-benefit analysis of China's Action Plan for Air Pollution Prevention and Control. ENGINEERING Management. 2019; 6:524–37. Li S, Wu D, Liu L, Yang L, Wang Y, Cao S, et al. Is it worth implementing the Blue Sky Defense Battle initiative? A cost-benefit analysis of the Chengdu case. Integrated Environmental Assessment and Management. 2025; 21:425–41. Zhou Q, Chen J, van den Bosch CCK, Zhang W, Zhu L, Vera YMR, et al. Constructing an Aims-Indicators-Methods framework for Green Space System Planning in China. Urban Forestry & Urban Greening. 2022; 67:127437. Additional Declarations No competing interests reported. Supplementary Files supplementarmaterials.docx Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 07 May, 2026 Reviews received at journal 06 May, 2026 Reviewers agreed at journal 22 Apr, 2026 Reviewers invited by journal 22 Apr, 2026 Editor assigned by journal 15 Apr, 2026 Submission checks completed at journal 15 Apr, 2026 First submitted to journal 10 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9378849","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":628869171,"identity":"dc74ce73-d903-4f15-8cee-c0213bd3195a","order_by":0,"name":"Siyuan Wang","email":"","orcid":"","institution":"George Institute for Global Health","correspondingAuthor":false,"prefix":"","firstName":"Siyuan","middleName":"","lastName":"Wang","suffix":""},{"id":628869174,"identity":"b9b4ea22-8945-4a89-83fc-3a6284a006fa","order_by":1,"name":"Zhiwei Xu","email":"","orcid":"","institution":"School of Medicine and Dentistry, Griffith University","correspondingAuthor":false,"prefix":"","firstName":"Zhiwei","middleName":"","lastName":"Xu","suffix":""},{"id":628869176,"identity":"d0a50fba-7465-4d1d-819b-466eacd110be","order_by":2,"name":"Gian Luca Di Tanna","email":"","orcid":"","institution":"Department of Business Economics, Health and Social Care, University of Applied Sciences and Arts of Southern Switzerland","correspondingAuthor":false,"prefix":"","firstName":"Gian","middleName":"Luca Di","lastName":"Tanna","suffix":""},{"id":628869179,"identity":"497ff5be-d479-4dff-b7c1-c88e4901cb0a","order_by":3,"name":"Raksha Pandya-Wood","email":"","orcid":"","institution":"Public Health, School of Medicine and Translational Health Research Institute (THRI), Faculty of Health, Western Sydney University","correspondingAuthor":false,"prefix":"","firstName":"Raksha","middleName":"","lastName":"Pandya-Wood","suffix":""},{"id":628869180,"identity":"f1803a32-bd06-4a80-875a-98b5e924c453","order_by":4,"name":"Mingsheng Chen","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAArElEQVRIiWNgGAWjYHACxgcMPGCGAdFamA1I1sImAWUQqYV/2vFn1QUy2xIb2Ju3STDU3CGsReJ2jtntGTy3Ext4jpVJMBx7RliLgXQO220ekBaJHDMJxobDxGhJf1YM1iL/hmgtCWbMEFt4iNQC9IuxNFCLcRtPWrFFwjEitPDPTn/4mbfntmw/++GNNz7UEKEFDBh7gLEDYiQQqQEIfhCvdBSMglEwCkYgAAB7MDOUWnk6rwAAAABJRU5ErkJggg==","orcid":"","institution":"School of Health Policy and Management, Nanjing Medical University","correspondingAuthor":true,"prefix":"","firstName":"Mingsheng","middleName":"","lastName":"Chen","suffix":""},{"id":628869182,"identity":"4d5a0c25-bfc8-4a9b-ae14-1e002acf5d77","order_by":5,"name":"Laura Downey","email":"","orcid":"","institution":"George Institute for Global Health","correspondingAuthor":false,"prefix":"","firstName":"Laura","middleName":"","lastName":"Downey","suffix":""},{"id":628869183,"identity":"966b0c9d-6fd7-4c64-87fd-dc199f625daa","order_by":6,"name":"Stephen Jan","email":"","orcid":"","institution":"George Institute for Global Health","correspondingAuthor":false,"prefix":"","firstName":"Stephen","middleName":"","lastName":"Jan","suffix":""},{"id":628869185,"identity":"48ee1e52-a419-466f-8833-3186bbf627d0","order_by":7,"name":"Lei Si","email":"","orcid":"","institution":"School of Health Sciences, Faculty of Health, Western Sydney University","correspondingAuthor":false,"prefix":"","firstName":"Lei","middleName":"","lastName":"Si","suffix":""}],"badges":[],"createdAt":"2026-04-10 11:23:46","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9378849/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9378849/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108349893,"identity":"c4fd2dc1-91e0-4cd3-8b72-23b6cd5100f8","added_by":"auto","created_at":"2026-05-03 09:57:07","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":102467,"visible":true,"origin":"","legend":"\u003cp\u003eConcentration curve of annual household medical expenditure by residential greenspace and particulate matter exposure\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-9378849/v1/2807c7829ea6cc0be712a2ab.png"},{"id":108349894,"identity":"158f762e-701f-4103-bbec-412770fcda48","added_by":"auto","created_at":"2026-05-03 09:57:12","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":525997,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9378849/v1/970f31b3-3930-4971-bb9a-8914fa6ae7b3.pdf"},{"id":108349892,"identity":"df8041bb-0117-442d-a478-beaf54610ae3","added_by":"auto","created_at":"2026-05-03 09:57:07","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":492881,"visible":true,"origin":"","legend":"","description":"","filename":"supplementarmaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-9378849/v1/c90cb27f58644abc66c1c3bb.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Ambient PM2.5, residential greenspace, and household healthcare expenditure in Shandong, China","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eEnvironmental factors are increasingly recognised as important determinants of population health and health-related economic outcomes, with growing evidence linking environmental degradation to higher household medical expenditures and widening disparities in health outcomes across populations.\u003csup\u003e\u003cspan additionalcitationids=\"CR2\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e Exposure to environmental determinants, including air pollution and natural environments, has been associated with a wide range of chronic diseases as well as both adverse and beneficial health outcomes.\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e In particular, ambient fine particulate matter (PM2.5) is a major environmental risk factor contributing to cardiometabolic diseases, respiratory conditions, and premature mortality.\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e In contrast, exposure to residential greenspace has been associated with a range of beneficial health outcomes, including improvements in mental health, reduced cardiovascular risk, and decreased all-cause mortality.\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e These environmental determinants not only affect healthcare costs but also contribute to inequities in health burdens. Exposure to harmful environments such as air pollution can lead to higher healthcare utilisation, increased out-of-pocket expenditures, and greater pressure on public health resources.\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e These burdens are often disproportionately borne by socioeconomically disadvantaged populations, who are more likely to reside in areas with higher pollution levels and limited access to health-promoting environments such as greenspace.\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e As a result, environmental inequalities can exacerbate existing health disparities, reinforcing cycles of poor health and financial hardship.\u003c/p\u003e \u003cp\u003eChina has experienced rapid economic development and urbanisation over recent decades, accompanied by substantial environmental changes. The burden of chronic diseases has increased substantially, leading to growing healthcare needs and rising healthcare costs.\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e Household out-of-pocket healthcare spending represents a major component of total health expenditure in China and can impose considerable financial pressure on families, particularly among vulnerable populations.\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e To date, several studies have assessed the relationship between household healthcare expenditure and environmental exposures in China for air pollution and urban greenspace. Zhang et al. constructed an environmental quality index encompassing multiple indicators, including PM2.5 concentrations, water quality, and noise levels, to examine the effects of environmental conditions on older adults in China, and found that improvements in environmental quality were associated with a significant reduction in elderly healthcare expenditures.\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e An analysis of the China Health and Retirement Longitudinal Study indicated that air pollution substantially increased inpatient care utilisation among individuals aged 60 and older.\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e Additional studies have reported that higher air pollution significantly elevates household medical costs, disproportionately affecting low-income households and contributing to health inequalities.\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e In contrast, empirical evidence on the relationship between greenspace and household medical expenditure remains relatively limited. Evidence from Chengdu suggested that a higher degree of urbanisation, was associated with reductions in household medical expenses.\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e Similarly, another study reported that greater urban greenspace quality, as measured by user satisfaction, was significantly inversely related to healthcare expenditure in Shanghai.\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e International evidence also supports the link between environmental exposures and healthcare costs. \u003csup\u003e17\u003c/sup\u003e In the United States, exposure to fine particulate matter has been associated with increased hospital admissions and higher Medicare expenditures.\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e Studies across European countries have demonstrated that elevated air pollution levels are linked to higher healthcare spending relating to cardiovascular and respiratory diseases.\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e In South and Southeast Asia, exposure to air pollution was associated with increased out-of-pocket healthcare expenditure, disproportionately burdening lower-income households.\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eMore importantly, the distributional aspects of household health expenditure across these environmental determinants of health remain poorly understood. Previous studies have assessed environmental exposure inequalities by socioeconomic or income groups, using measures such as the Gini coefficient or concentration index to quantify distributions of PM2.5 and urban greenspace. For example, Song et al. applied the Gini index to quantify inequality in greenspace exposure across 303 cities in China, finding that the majority of cities experienced substantial disparities, with 207 cities exhibiting Gini coefficients greater than 0.6.\u003csup\u003e21\u003c/sup\u003e Cao et al. used the Palma coefficient, as an alternative inequality measure, to assess disparities in residential greenspace exposure within urban areas, reporting that populations with the highest exposure experienced nearly four times the levels observed in the least exposed groups.\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e While these studies provide important evidence on inequalities in environmental exposure across socioeconomic cohorts, they do not address how health-related outcomes are distributed across environmental exposure gradients. To bridge these gaps, this study assesses the relationship between air pollution, urban greenspace, and household healthcare expenditure, and further evaluates the distribution of healthcare expenditure across gradients of air pollution and urban greenspace.\u003c/p\u003e"},{"header":"2. Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Study design and setting\u003c/h2\u003e \u003cp\u003eThis study was a cross-sectional analysis using population health data collected in Shandong Province through the sixth-round of the Health Services Survey (HSS) in 2018. In brief, the HSS is a government-administered household survey that collects routine information on household individual\u0026rsquo;s health status, health care utilisation, quality of life, and household healthcare expenditures. The survey used a stratified multistage cluster sampling design, in which administrative units were stratified by geographic region and urban\u0026ndash;rural status. All eligible members of selected households were invited to participate in face-to-face interviews administered by trained professionals. For the 6th HSS in Shandong Province, twenty districts were selected, and within each district, two villages were sampled from each of five townships, yielding 10 villages per district. Sixty households were selected from each village, resulting in a total of 35,262 individuals from 12,938 households interviewed across more than 150 counties and cities. Individuals with missing values were excluded from the analysis.\u003c/p\u003e \u003cp\u003eWe chose Shandong Province due to its large and demographically diverse population of over 100\u0026nbsp;million residents, spanning a wide range of urban and rural settings that reflect substantial socioeconomic and environmental heterogeneity. Supplementary Table\u0026nbsp;1 provides a summary of the socioeconomic and environmental indicators for the prefectural-level cities sampled in the 6th Health Services Survey of Shandong Province. Despite a general declining trend, annual average PM2.5 concentrations in Shandong Province exhibited pronounced spatial variation, with the highest levels (\u0026gt;\u0026thinsp;40 \u0026micro;g/m\u0026sup3;) concentrated in western plains such as Liaocheng, Heze, Dezhou, and Jining, moderately high levels (33\u0026ndash;40 \u0026micro;g/m\u0026sup3;) in central, southern, and northern cities, and the lowest levels (\u0026lt;\u0026thinsp;30 \u0026micro;g/m\u0026sup3;) in the Jiaodong Peninsula.\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e The province has a high urbanisation level (67%) and urban population density (647 people/km\u0026sup2;), with above-average urban greening, including 78,676 hectares of parks and 326,340 hectares of total urban green space (1.7% of the province\u0026rsquo;s land area).\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e Further, the province\u0026rsquo;s landscape is characterised by diverse geographic features, including plains, hills, and coastal regions, contributing to substantial variation in land cover and greenspace density. Heterogeneity in land cover creates substantial variation in coastal\u0026ndash;inland differentiation of greenspace morphology, cover, and density, with fractional vegetation cover exhibiting a clear inland‑to‑coast gradient.\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e Together, this provides an opportunity to assess the relationship between environmental exposures and household medical expenditure, as well as inequalities in the distribution of healthcare costs across these factors.\u003c/p\u003e \u003cp\u003eTo ensure transparent and standardised reporting, this analysis was conducted following the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) checklist for observational studies (Supplementary table 2).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Health status and household medical expenditure\u003c/h2\u003e \u003cp\u003eSelf-reported household medical expenditure was collected from respondents based on household spending over the previous year, including out-of-pocket payments for inpatient and outpatient visits, medications, medical devices, diagnostic tests and other healthcare services, but excluding payments for health care products and non-medical services. Health status information was collected through self-reported responses, where participants were asked whether they had been diagnosed by a medical doctor, or undergone treatment for specific conditions within the past six months. Reported conditions were subsequently classified according to the questionnaire\u0026rsquo;s prespecified list, including chronic conditions such as cardiometabolic diseases (CMDs), cancer, respiratory diseases (RDs), and neuropsychiatric disorders (NPDs). A full list of conditions is provided in Supplementary table 3.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Residential greenspace and air pollution\u003c/h2\u003e \u003cp\u003eWe used Normalised Difference Vegetation Index (NDVI) measured from the Terra Moderate Resolution Imaging Spectroradiometer Vegetation Indices (MOD13Q1), which provides repeated measures of surface vegetation at 250 m spatial resolution at 16-day intervals.\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e NDVI is a validated proxy for vegetation cover and has been widely applied in population environmental exposure studies to quantify greenspace exposure.\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e Values range from \u0026minus;\u0026thinsp;1 to 1, with higher values indicating denser vegetation. Residential greenness exposure for each participant was quantified by calculating the three-year average of NDVI values within a 1 km radius of their geo-coded village-level residence.\u003c/p\u003e \u003cp\u003eAir pollution exposure was estimated using ground-level concentrations of PM2.5 derived from the China High Air Pollutants (CHAP) dataset.\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e Briefly, The dataset provides PM2.5 measurements at a spatial resolution of 0.01\u0026deg;, generated using a space\u0026ndash;time extremely randomised trees (STET) model. This model integrates multiple data sources, including satellite-derived aerosol optical depth (AOD), meteorological variables, land-use and topographic data, and emission inventories.\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e Using a similar method to NDVI, residential PM2.5 exposure were estimated as the mean concentration within a 3 km buffer around each respondent\u0026rsquo;s geo-coded village over the five years prior to the survey.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Covariates\u003c/h2\u003e \u003cp\u003eTo control for potential confounding at the individual level, models were adjusted for a set of sociodemographic and health-related characteristics. These included age (\u0026lt;\u0026thinsp;45 years, 45\u0026ndash;65 years, and \u0026gt;\u0026thinsp;65 years), sex (males or females), educational attainment (primary school, junior high school, high school or above ), residential setting (urban or rural), marital status (never married, married or other), annual household income (categorised as above or below 60% of the 2018 median annual income of 24,336 CNY, representing relative poverty\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e), insurance status (insured or not insured), and body mass index (BMI; \u0026lt;18.5 kg/m\u0026sup2;, \u0026ge;\u0026thinsp;18.5 kg/m\u0026sup2; and \u0026lt;\u0026thinsp;25 kg/m\u0026sup2;, and \u0026ge;\u0026thinsp;25 kg/m\u0026sup2;).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Statistical analysis\u003c/h2\u003e \u003cp\u003eWe employed Generalised Linear Mixed Models (GLMMs) with village-level random effects to assess the relationship between environmental exposures (NDVI and PM2.5) and household medical expenditure. To allow for potential nonlinear exposure\u0026ndash;response relationships, both NDVI and PM2.5 were categorised into quartiles. GLMMs have been widely applied in studies to assess the relationships between environmental exposures, such as air pollution and greenness, and health outcomes. Village-level random effects were specified to account for the hierarchical sampling structure and potential clustering within villages. Specifically, we fitted two GLMMS: Model 1 assessed the association between healthcare expenditure and residential greenspace, whereas Model 2 assessed the association between healthcare expenditure and ambient air pollution. Both models were adjusted for individual-level covariates, including age, sex, urban/rural residence, marital status, insurance status, education, income, and BMI.\u003c/p\u003e \u003cp\u003eTo examine inequalities in household healthcare expenditure in relation to residential greenspace and air pollution, we applied the concept of the Concentration Index (CI). CI has been widely used in health economics to quantify the degree of socioeconomic-related inequality in health outcomes.\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e In our analysis, each individual was assigned a fractional rank based on their mapped NDVI and PM2.5 values. The CI was then calculated as:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:CI=\\frac{2}{\\mu\\:}\\:Cov(y,R)$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eWhere \u0026#119910; is self-reported healthcare expenditure, R is the fractional rank of the population by NDVI or PM2.5 exposure, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\mu\\:\\)\u003c/span\u003e\u003c/span\u003e is the mean of \u0026#119910;. The CI is interpreted as twice the area between the Concentration Curve (CC), which plots the cumulative share of household healthcare expenditure against the cumulative share of the population ranked by environmental exposure, and the line of equality (the 45-degree line). It ranges from \u0026minus;\u0026thinsp;1 to 1, with \u0026minus;\u0026thinsp;1 indicating total concentration among the least exposed, 1 indicating total concentration among the most exposed, and 0 indicating perfect equality. As the standard CI is unadjusted and may be influenced by differences in population characteristics, Partial Concentration Indices (PCI) were computed to assess inequality of expenditure across environmental exposures after accounting for covariates. The PCI quantifies inequality in a health outcome with respect to a ranking variable after removing the influence of standardising variables (e.g., age and sex) that are correlated with the ranking variable but are not themselves policy amenable.\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e We employed the direct standardisation concept as described by Gravelle et al,. in which the outcome is adjusted using a regression model that includes both the ranking variable and standardising covariates.\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eSubgroup analyses were conducted to assess how inequalities in household medical expenditure across environmental exposures varied by chronic disease class, as defined in the HSS (Supplementary Table\u0026nbsp;3). The analysis focused on respiratory diseases, cardiometabolic diseases, and neuropsychiatric disorders, which were selected due to their high prevalence and the well-established links between these conditions and environmental factors such as residential greenspace and air pollution.\u003csup\u003e\u003cspan additionalcitationids=\"CR33 CR34\" citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e In addition, sensitivity analyses were conducted to assess the robustness of the associations between household medical expenditure and environmental exposures, using annual PM2.5 concentrations measured at 4 km, 5 km, and 10 km buffers, and NDVI across 2\u0026ndash;5 km buffers.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Sample characteristics\u003c/h2\u003e \u003cp\u003eA total of 27,603 individuals were included in this analysis. The flowchart of study participants selection is presented in Supplementary Fig.\u0026nbsp;1. Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e presents the characteristics of the study population stratified by chronic disease status. Chronic disease prevalence was higher among participants aged\u0026thinsp;\u0026ge;\u0026thinsp;65 years, females, those who were married, those who have no health insurance, individuals with lower educational attainment or household income, and those with higher body mass index. Residential greenness, measured by NDVI, ranged from 0.081 to 0.420 (mean 0.229, SD 0.082), while ambient PM2.5 concentrations ranged from 18.39 to 45.77 \u0026micro;g/m\u0026sup3; (mean 36.88, SD 6.47). Median household medical expenditure was 3,000 RMB (IQR 5,000 RMB) and was higher among participants with chronic conditions than among those without. Overall, 7,783 participants (28% of the study population) had at least one chronic condition, with 3,083 experiencing cardiometabolic diseases, 401 experiencing respiratory diseases, and 261 experiencing neuropsychiatric disorders.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eGeneral characteristics of the sample population\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eAge\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTotal (N\u0026thinsp;=\u0026thinsp;27,603)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eChronic disease (N\u0026thinsp;=\u0026thinsp;7,783)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eNo chronic disease (N\u0026thinsp;=\u0026thinsp;19,820)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eP value*\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBelow 45 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e10,948\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e683\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e10,265\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e45 to 65 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e11,454\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4,061\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e7,393\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAbove 65 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e5,201\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3,039\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2,162\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSex\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e13,273\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3,640\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9,633\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e14,330\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4,143\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e10,187\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMarital status\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNever Married\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2,412\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e123\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2,289\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMarried and other\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e25,191\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e7,660\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e17,531\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEducation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePrimary school\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e8,820\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3,973\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4,847\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eJunior high school\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e9,815\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2,386\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e7,429\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHigh school or above\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e8,968\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1,424\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e7,544\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAnnual household income\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLess than 60% of median household income (14,601 CNY)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e5,521\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2,534\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2,987\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMore than 60% of median household income (14,601 CNY)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e22,082\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5,249\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e16,833\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBody mass index (BMI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUnder weight (\u0026lt;\u0026thinsp;18.5 kg/m\u0026sup2;)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1,301\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e313\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e988\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNormal weight (18.5\u0026ndash;24.9 kg/m\u0026sup2;)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e15,438\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3,677\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e11,761\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOverweight+ (\u0026gt;\u0026thinsp;25 kg/m\u0026sup2;)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e10,864\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3,793\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e7,071\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHousehold healthcare expenditure\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026lt;=RMB 1,500\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e9,381\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1,692\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e7,689\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRMB 1,500-5,000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e10,553\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3,034\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e7,519\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026gt;RMB 5,000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e7,669\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3,057\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4,612\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInsurance status\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInsured\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e13,963\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3,733\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e10,230\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNot insured\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e13,640\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4,050\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9,590\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNDVI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{NDVI}_{Q1}\\)\u003c/span\u003e\u003c/span\u003e [0.081\u0026ndash;0.153)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6,869\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1,766\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e5,103\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{NDVI}_{Q2}\\)\u003c/span\u003e\u003c/span\u003e [0.153\u0026ndash;0.238)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6,901\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1,788\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e5,113\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{NDVI}_{Q3}\\)\u003c/span\u003e\u003c/span\u003e [0.238\u0026ndash;0.290)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6,978\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2,008\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4,970\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{NDVI}_{Q4}\\)\u003c/span\u003e\u003c/span\u003e [0.290\u0026ndash;0.420]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6,855\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2,221\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4,634\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePM2.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{PM2.5}_{Q1}\\:\\)\u003c/span\u003e\u003c/span\u003e[18.39\u0026ndash;35.15)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e7,282\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1,995\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e5,287\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{PM2.5}_{Q2}\\:\\)\u003c/span\u003e\u003c/span\u003e[35.15\u0026ndash;38.86)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6,669\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1,822\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4,847\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{PM2.5}_{Q3}\\:\\)\u003c/span\u003e\u003c/span\u003e[38.86\u0026ndash;41.27)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6,843\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1,922\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4,921\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{PM2.5}_{Q4}\\:\\)\u003c/span\u003e\u003c/span\u003e[41.27\u0026ndash;45.77]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6,809\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2,044\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4,765\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"5\"\u003e* P-values were obtained under the Χ\u0026sup2; test.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Household medical expenditure, residential greenspace and air pollution\u003c/h2\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e presents the associations between self-reported annual medical expenditure and residential greenspace. After adjusting for age, sex, marital status, region, income, education, insurance status, and BMI, higher residential greenspace exposure was associated with lower annual household medical expenditure. Compared with participants in the lowest NDVI quartile (Q1), those in Q2, Q3, and Q4 had significantly lower healthcare expenditure (β = \u0026minus;0.21, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; β = \u0026minus;0.21, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; and β = \u0026minus;0.26, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001, respectively. Older age was associated with higher medical expenditure, with participants aged 45\u0026ndash;65 years (β\u0026thinsp;=\u0026thinsp;0.13, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and \u0026gt;\u0026thinsp;65 years (β\u0026thinsp;=\u0026thinsp;0.32, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) incurring greater costs compared with those aged\u0026thinsp;\u0026lt;\u0026thinsp;45 years. Households with income above the relative poverty line incurred higher healthcare expenditure (β\u0026thinsp;=\u0026thinsp;0.10, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Compared with normal-weight individuals, underweight (β\u0026thinsp;=\u0026thinsp;0.17, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and overweight or obese individuals (β\u0026thinsp;=\u0026thinsp;0.03, p\u0026thinsp;=\u0026thinsp;0.05) had higher medical expenditure. Being married was associated with higher household healthcare expenditure (β\u0026thinsp;=\u0026thinsp;0.07, p\u0026thinsp;=\u0026thinsp;0.01), whereas residing in a rural area was associated with lower medical expenditure (β = \u0026minus;0.05, p\u0026thinsp;=\u0026thinsp;0.04), as was having health insurance (β = \u0026minus;0.10, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eAssociation between greenspace exposure and self-reported household medical expenditure\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariables\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eβ (95% CI)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eP value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSex (base\u0026thinsp;=\u0026thinsp;Female)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.51\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge (base\u0026thinsp;=\u0026thinsp;\u0026lt;\u0026thinsp;45 years)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e45\u0026ndash;65 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;65 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMarriage (base\u0026thinsp;=\u0026thinsp;Never married)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMarried\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRegion (base\u0026thinsp;=\u0026thinsp;Urban)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRural\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.04\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAnnual household income (base\u0026thinsp;=\u0026thinsp;Less than 60% of median household income)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMore than 60% of median income\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEducation (base\u0026thinsp;=\u0026thinsp;Primary school)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eJunior high school\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.08\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHigh school or above\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.21\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBMI (base\u0026thinsp;=\u0026thinsp;Normal weight, 18.5\u0026ndash;24.9 kg/m\u0026sup2;)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUnderweight (\u0026lt;\u0026thinsp;18.5 kg/m\u0026sup2;)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOverweight or obese (\u0026ge;\u0026thinsp;25.0 kg/m\u0026sup2;)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.05\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInsurance status (base\u0026thinsp;=\u0026thinsp;No insurance)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInsured\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNDVI (base\u0026thinsp;=\u0026thinsp;Q1, [0.081\u0026ndash;0.153))\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eQ2 [0.153\u0026ndash;0.238)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eQ3 [0.238\u0026ndash;0.290)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eQ4 [0.290\u0026ndash;0.420]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e summarises the association between annual household medical expenditure and PM2.5 exposure. After adjustment for the same set of covariates, higher PM2.5 exposure was consistently associated with increased healthcare expenditure. Compared with the lowest quartile, participants in the second, third, and fourth PM2.5 quartiles had progressively higher expenditures (β\u0026thinsp;=\u0026thinsp;0.13, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; β\u0026thinsp;=\u0026thinsp;0.17, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; and β\u0026thinsp;=\u0026thinsp;0.30, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001, respectively). Older age was associated with increased expenditure, with participants aged 45\u0026ndash;65 years (β\u0026thinsp;=\u0026thinsp;0.15, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and \u0026gt;\u0026thinsp;65 years (β\u0026thinsp;=\u0026thinsp;0.35, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) having higher costs compared with those aged\u0026thinsp;\u0026lt;\u0026thinsp;45 years. Household income above 60% of the median was associated with higher expenditure (β\u0026thinsp;=\u0026thinsp;0.11, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Underweight individuals had higher medical expenditure (β\u0026thinsp;=\u0026thinsp;0.16, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), and overweight or obese individuals had a smaller but significant increase (β\u0026thinsp;=\u0026thinsp;0.03, p\u0026thinsp;=\u0026thinsp;0.03). Being married was associated with higher expenditure (β\u0026thinsp;=\u0026thinsp;0.06, p\u0026thinsp;=\u0026thinsp;0.04), whereas residing in rural areas was associated with lower expenditure (β = \u0026minus;0.07, p\u0026thinsp;=\u0026thinsp;0.001). Male sex, education level, and health insurance coverage were not significantly associated with household healthcare expenditure.\u003c/p\u003e \u003cp\u003eSensitivity analyses using alternative spatial buffers indicated that the association between residential greenspace and household medical expenditure was generally robust across NDVI buffers of 2\u0026ndash;4 km, with the exception of the highest quartile at the 4 km buffer. Relative to the lowest quartile (Q1), higher levels of residential greenspace were consistently associated with lower annual household medical expenditure across these NDVI measurements. However, at the 5km radius buffer, the association was not statistically significant (Supplementary table 4). In contrast, sensitivity analyses under alternative PM2.5 buffer sizes suggested no statistically significant differences in household medical expenditure across exposure quartiles, except for Q3 in the 10 km buffer model (Supplementary table 5).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eAssociation between PM2.5 and self-reported household medical expenditure\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariables\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eβ (95% CI)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eP value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSex (base\u0026thinsp;=\u0026thinsp;Female)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.23\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge (base\u0026thinsp;=\u0026thinsp;\u0026lt;\u0026thinsp;45 years)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e45\u0026ndash;65 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;65 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.35\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMarriage (base\u0026thinsp;=\u0026thinsp;Never married)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMarried\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.06\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.04\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRegion (base\u0026thinsp;=\u0026thinsp;Urban)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRural\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAnnual household income (base\u0026thinsp;=\u0026thinsp;Less than 60% of median household income)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMore than 60% of median income\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEducation (base\u0026thinsp;=\u0026thinsp;Primary school)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eJunior high school\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHigh school or above\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.06\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBMI (base\u0026thinsp;=\u0026thinsp;Normal weight, 18.5\u0026ndash;24.9 kg/m\u0026sup2;)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUnderweight (\u0026lt;\u0026thinsp;18.5 kg/m\u0026sup2;)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOverweight or obese (\u0026ge;\u0026thinsp;25.0 kg/m\u0026sup2;)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.03\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInsurance status (base\u0026thinsp;=\u0026thinsp;No insurance)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInsured\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.89\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePM2.5 (base\u0026thinsp;=\u0026thinsp;Q1, [18.39\u0026ndash;35.15))\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eQ2 [35.15\u0026ndash;38.86)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eQ3 [38.86\u0026ndash;41.27)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eQ4 [41.27\u0026ndash;45.77]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Concentration of annual household medical expenditure by environmental exposures\u003c/h2\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e presents the concentration curve of healthcare expenditure ranked by levels of greenspace and PM2.5 exposure. Healthcare expenditure was marginally disproportionately concentrated among residents with lower greenspace exposure (CI\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.02; 95% CI: \u0026minus;0.04, \u0026minus;\u0026thinsp;0.01) and higher PM2.5 exposure (CI\u0026thinsp;=\u0026thinsp;0.05; 95% CI: 0.03, 0.06). Supplementary Figs.\u0026nbsp;2\u0026ndash;4 present the results of subgroup analyses by disease category. For CMDs, annual medical expenditure was more concentrated among individuals with lower greenspace (CI\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.04; 95% CI: \u0026minus;0.07, \u0026minus;\u0026thinsp;0.02) and higher PM2.5 exposure (CI\u0026thinsp;=\u0026thinsp;0.03; 95% CI: 0.01, 0.06). For NPDs, healthcare costs were disproportionately concentrated among residents with lower greenspace (CI\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.08; 95% CI: \u0026minus;0.15, \u0026minus;\u0026thinsp;0.01) but did not exhibit a clear pattern with PM2.5 exposure (CI\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.02; 95% CI: \u0026minus;0.13, 0.06). In contrast, medical costs among individuals with respiratory disease showed no clear distributional pattern across levels of residential greenspace or PM2.5 exposure (CI\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.02; 95% CI: \u0026minus;0.10, 0.07 and CI\u0026thinsp;=\u0026thinsp;0.02; 95% CI: \u0026minus;0.07, 0.12, respectively). After adjusting for covariates using the direct standardisation method, the concentration of inequality was marginally greater than the unadjusted model. Specifically, household medical expenditure remained disproportionately concentrated among individuals with lower greenspace (PCI\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.04; 95% CI: \u0026minus;0.05, \u0026minus;\u0026thinsp;0.03) and higher PM2.5 exposure (PCI\u0026thinsp;=\u0026thinsp;0.06; 95% CI: 0.05, 0.07). For CMDs, expenditures were concentrated among residents with lower greenspace exposure (PCI\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.04; 95% CI: \u0026minus;0.05, \u0026minus;\u0026thinsp;0.03) and higher PM2.5 (PCI\u0026thinsp;=\u0026thinsp;0.03; 95% CI: 0.02, 0.04). NPD medical expenditures remained concentrated among individuals with lower greenspace exposure (PCI\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.06; 95% CI: \u0026minus;0.08, \u0026minus;\u0026thinsp;0.03) but not for PM2.5 exposure (PCI\u0026thinsp;=\u0026thinsp;0.01; 95% CI: \u0026minus;0.02, 0.01). For respiratory diseases, medical expenditure was not strongly concentrated across either environmental exposure, with a partial concentration index of \u0026minus;\u0026thinsp;0.01 (95% CI: \u0026minus;0.03, 0.01) for greenspace exposure and 0.01 (95% CI: \u0026minus;0.02, 0.02) for PM2.5 exposure.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eThis study examined the relationships between environmental exposures and household healthcare expenditure and assessed how medical costs were distributed across gradients of air pollution and urban greenspace. We found that higher levels of medical expenditure were significantly associated with greater exposure to PM2.5 and lower levels of residential greenness. Inequality analyses based on concentration indices indicated a statistically significant but marginal distribution of healthcare expenditure across environmental gradients. Specifically, higher medical expenditure was concentrated among individuals experiencing higher PM2.5 concentrations and lower levels of residential greenspace exposure, after accounting for potential confounding factors. When examined by disease category, household healthcare expenditure for people reporting cardiometabolic diseases were concentrated among residents experiencing both lower levels of greenspace and higher exposure to PM2.5. Medical expenditures among individuals reporting neuropsychiatric disorders were disproportionately concentrated in areas with lower residential greenspace, whereas no clear pattern was observed across PM2.5 exposure levels. In contrast, respiratory disease expenditures did not exhibit a clear distributional pattern with either environmental exposure.\u003c/p\u003e \u003cp\u003eWhile the observed associations between environmental exposures and healthcare expenditure are broadly consistent with existing evidence linking air pollution and greenspace to medical expenditure, this study extends prior research in several aspects. First, a major strength of this study is the integration of individual-level health data with high-resolution environmental exposure data, enabling a more comprehensive assessment of how environmental factors relate to household medical expenditure within an equity framework. We examined the distribution of household medical expenditure across gradients of PM2.5 and residential greenspace exposure, an approach that has not been applied in previous studies. We observed a modest but statistically significant concentration of healthcare expenditure among individuals exposed to higher levels of PM2.5 and lower levels of greenspace, after adjustment for sociodemographic and health-related confounders. This pattern likely reflects the cumulative health impacts of environmental disadvantage, whereby populations living in areas with elevated air pollution and limited greenspace are more susceptible to chronic conditions, particularly cardiometabolic and neuropsychiatric disorders, which necessitate ongoing medical care and contribute to higher healthcare costs. Second, while previous research on air pollution and healthcare expenditure has largely focused on older populations,\u003csup\u003e13, 36, 37\u003c/sup\u003e the present study examined these relationships in a broader population-based sample, providing evidence that air pollution may influence healthcare costs across the general population. Our findings suggested that individuals exposed to higher PM2.5 levels incurred progressively greater medical expenditure, with those in the second, third, and highest quartiles experiencing approximately 14%, 19%, and 35% higher costs, respectively, compared with individuals in the lowest quartile. When assuming a linear relationship, our results are comparable to previous national evidence, which reported 2.9% and 5.7% increases in medical expenditure per unit increase in PM2.5 concentrations.\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e Third, this study contributes to the limited evidence on the relationship between greenspace and medical expenditure in China. One previous study assessed the relationship between urban greenspace coverage and medical expenditure, finding that a 10% increase in urban green space was associated with a 0.012% reduction in residents\u0026rsquo; medical costs.\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e However, the study measured greenspace exposure as aggregate area (hectares), which inadequately reflects vegetation greenness, density, or health characteristics. NDVI and other spatially resolved metrics have been widely used in epidemiological greenspace studies to address this limitation.\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e Another study conducted in Chengdu found a statistically significant association between greenspace, measured using NDVI, and inpatient medical costs.\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e This study builds on previous research by addressing key limitations in how greenspace exposure is measured and extends the analysis to examine its relationship with household healthcare expenditure. These results reinforce the economic rationale for public investment in greenspace. Beyond its well-documented health and environmental benefits, our findings demonstrate that strategic investment in urban greening can generate measurable economic returns by reducing household medical expenditures, while also promoting equitable access to environmental resources and protecting vulnerable populations from catastrophic healthcare costs. This underscores the potential of urban greening as a cost-effective strategy to enhance population health, while promoting environmental equity and household financial security.\u003c/p\u003e \u003cp\u003eOur findings provide empirical evidence supporting China\u0026rsquo;s ongoing policy efforts to improve air quality and expand access to urban greenspace. China has demonstrated a strong commitment to improving environmental quality through a series of national and regional policy initiatives. Measures such as the Clean Air Action Plan and the Three-Year Blue Sky Defence Plan have established stricter standards for major air pollutants, contributing to measurable improvements in population health as well as economic benefits, including reductions in household healthcare expenditures.\u003csup\u003e\u003cspan additionalcitationids=\"CR42\" citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e Similarly, greenspace planning has been progressively incorporated into China\u0026rsquo;s national urban development framework, with key targets for greenspace coverage and per-capita availability embedded within the national Green Space System Planning framework.\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e These policy initiatives reflect a growing recognition of the importance of environmental determinants in influencing health and healthcare utilisation. Further, our results suggest that future policy development could be benefited by stronger considerations for equity. While existing policies have primarily emphasised overall improvements in air quality and greenspace provision, less attention has been given to the distribution of environmental benefits and burdens across populations. In particular, greenspace policies have prioritised aggregate coverage targets over equality or accessibility.\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e As China continues its efforts to fulfil its commitment to the Sustainable Development Goals, such as providing equitable access to quality greenspace and substantially reducing deaths and illnesses from air pollution, our findings highlight the importance of addressing environmental determinants of health to reduce the financial burden of healthcare. As exposure \u0026mdash;or lack of exposure \u0026mdash;to these environmental determinants can contribute to the exacerbation of major chronic and acute illnesses, resulting in catastrophic healthcare expenditures, improving air quality and access to greenspace could help reduce financial risks for households, support poverty alleviation, and advance China\u0026rsquo;s goals for universal health coverage. Our finding suggest that future urban environmental planning should place greater emphasis on the quality, accessibility, and spatial distribution of greenspace, alongside continued efforts to reduce air pollution, to reduce disparities in healthcare expenditure associated with unequitable environmental exposures.\u003c/p\u003e \u003cp\u003eThis study has several limitations. First, the cross-sectional design limited any inference of causality between environmental exposures and medical expenditure, as well as limiting the understanding of the long-term health impacts of environmental exposures on medical expenditures. Second, although we adjusted for key demographic and socioeconomic factors, residual confounding by unmeasured individual characteristics, such as underlying health status, local healthcare infrastructure, or healthcare-seeking behaviours, as well as other environmental factors, including noise, urban heat, or traffic density, could influence the observed associations. Third, our analysis examined the relationship and distribution of medical expenditure with air pollution and greenspace separately, without accounting for their potential joint or interactive effects. This was mainly due to spatial correlation and multicollinearity between air pollution and greenspace, leading to unstable estimates and limiting the ability to examine their independent contributions. Fourth, we relied on self-reported medical expenditure, which, despite the HSS being carried out by trained professionals, may be subject to recall bias and reporting errors. This limitation also applies to health conditions, particularly for under-diagnosed diseases such as diabetes and hypertension. Under-reporting of these conditions may be more pronounced among rural residents or populations with limited access to healthcare, potentially attenuating observed associations between environmental exposures, such as greenspace, and health outcomes. Consequently, the true magnitude of inequities in healthcare expenditure across environmental gradients may be even greater than estimated in this study. Fifth, our findings were based on analysing a representative cohort for Shandong province, a region with relatively higher socioeconomic development compared with the national average. Consequently, the observed associations and distributional patterns of medical expenditure may not be generalisable to the national population and warrants further research.\u003c/p\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eOur study demonstrated that environmental factors are closely linked to household healthcare expenditure in China, with greater residential greenspace associated with lower costs and higher air pollution linked to increased costs. Medical spending is disproportionately borne by populations living in areas with less greenspace and higher PM2.5 levels, reflecting inequities in health and economic burden created by geographical inequities in environmental risks. These findings highlight the need for public health and policy strategies that incorporate environmental quality improvements, focusing particularly on regions which experience the dual burden of higher environmental risks and health care costs.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003e\u003cspan\u003eAOD \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Aerosol Optical Depth\u003c/span\u003e\u003c/p\u003e\n\u003cp\u003e\u003cspan\u003eBMI\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Body Mass Index\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n\u003cp\u003e\u003cspan\u003eCC\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Concentration Curve\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n\u003cp\u003e\u003cspan\u003eCHAP\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;China High Air Pollutants\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n\u003cp\u003e\u003cspan\u003eCIs\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Confidence Intervals\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n\u003cp\u003e\u003cspan\u003eCMD\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Cardiometabolic Disease\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n\u003cp\u003e\u003cspan\u003eGLMMs\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Generalised Linear Mixed Models\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n\u003cp\u003e\u003cspan\u003eHealth Services Survey\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Health Services Survey\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n\u003cp\u003e\u003cspan\u003eMOD13Q1\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Moderate Resolution Imaging Spectroradiometer Vegetation Indices\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n\u003cp\u003e\u003cspan\u003eNDVI\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Normalised Difference Vegetation Index\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n\u003cp\u003e\u003cspan\u003eNPD\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Neuropsychiatric Disorders\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n\u003cp\u003e\u003cspan\u003ePCI\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Partial Concentration Indices\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n\u003cp\u003e\u003cspan\u003ePM2.5 Particulate Matter\u0026nbsp;\u003c/span\u003e\u003cspan\u003e\u0026le;\u003c/span\u003e\u003cspan\u003e2.5 \u0026micro;m\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n\u003cp\u003e\u003cspan\u003eRDs\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Respiratory Diseases\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n\u003cp\u003e\u003cspan\u003eSTET Space\u003c/span\u003e\u003cspan\u003e\u0026ndash;\u003c/span\u003e\u003cspan\u003eTime Extremely Randomised Trees\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n\u003cp\u003e\u003cspan\u003eSTROBE \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Strengthening the Reporting of Observational Studies in Epidemiology\u003c/span\u003e\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was a secondary analysis of de-identified data from the Shandong Health Services Survey and informed consent was obtained from all participants. The study was approved by the UNSW Human Research Ethics Committee (HC230209).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data that support the findings of this study are available from the corresponding author upon reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors report no conflict of interest. The authors alone are responsible for the content and writing of the paper.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was funded by the National Natural Science Foundation of China (grant number: 71874086, 72174093, 72474109).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualization: SW, LS, SJ\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eData curation: SW, LS\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFormal analysis: SW, LS\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMethodology: SW, LS, SJ\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eResources: MC, LS\u003c/p\u003e\n\u003cp\u003eSupervision: LS\u003c/p\u003e\n\u003cp\u003eWriting \u0026ndash; original draft: SW\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWriting \u0026ndash; reviewing \u0026amp; editing: SW, LS, MC, ZX, RP, LD, GLDT, SJ\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eJerrett M, Eyles J, Dufournaud C, Birch S. Environmental influences on healthcare expenditures: an exploratory analysis from Ontario, Canada. J Epidemiol Community Health. 2003; 57:334\u0026ndash;8.\u003c/li\u003e\n\u003cli\u003eAnwar A, Hyder S, Bennett R, Younis M. Impact of Environmental Quality on Healthcare Expenditures in Developing Countries: A Panel Data Approach. Healthcare (Basel). 2022; 10.\u003c/li\u003e\n\u003cli\u003eSzymańska A. The relationship between health expenditure, income, and environmental degradation: Evidence from OECD economies. Economic Analysis and Policy. 2025; 87:2183\u0026ndash;201.\u003c/li\u003e\n\u003cli\u003eWu C, Liu J, Li Y, Qin L, Gu R, Feng J, et al. Association of residential air pollution and green space with all-cause and cause-specific mortality in individuals with diabetes: an 11-year prospective cohort study. EBioMedicine. 2024; 108:105376.\u003c/li\u003e\n\u003cli\u003eSangkham S, Phairuang W, Sherchan SP, Pansakun N, Munkong N, Sarndhong K, et al. An update on adverse health effects from exposure to PM2.5. Environmental Advances. 2024; 18:100603.\u003c/li\u003e\n\u003cli\u003eYang B-Y, Zhao T, Hu L-X, Browning MHEM, Heinrich J, Dharmage SC, et al. Greenspace and human health: An umbrella review. The Innovation. 2021; 2:100164.\u003c/li\u003e\n\u003cli\u003eBailie CR, Ghosh JKC, Kirk MD, Sullivan SG. Effect of ambient PM(2.5) on healthcare utilisation for acute respiratory illness, Melbourne, Victoria, Australia, 2014-2019. J Air Waste Manag Assoc. 2023; 73:120\u0026ndash;32.\u003c/li\u003e\n\u003cli\u003eYang J, Zhang B. Air pollution and healthcare expenditure: Implication for the benefit of air pollution control in China. Environment International. 2018; 120:443\u0026ndash;55.\u003c/li\u003e\n\u003cli\u003eHajat A, Hsia C, O\u0026apos;Neill MS. Socioeconomic Disparities and Air Pollution Exposure: a Global Review. Curr Environ Health Rep. 2015; 2:440\u0026ndash;50.\u003c/li\u003e\n\u003cli\u003eLiu H, Yin P, Qi J, Zhou M. Burden of non-communicable diseases in China and its provinces, 1990-2021: Results from the Global Burden of Disease Study 2021. Chin Med J (Engl). 2024; 137:2325\u0026ndash;33.\u003c/li\u003e\n\u003cli\u003eQiu T, Shi L, Zhou W, Deng J, Sun G. From individual burden to social risk pooling: drivers of the declining out-of-pocket health expenditure share in China (2009-2024). Int J Equity Health. 2026; 25:33.\u003c/li\u003e\n\u003cli\u003eZhang Y, Chen S, Liu D. A measurement study of the environmental quality and medical expenditures of elderly individuals: causal inference based on machine learning. Archives of Public Health. 2024; 82:195.\u003c/li\u003e\n\u003cli\u003eJin B, Li Z. Air pollution, healthcare use, and inequality: Evidence from China. Economic Modelling. 2024; 141:106905.\u003c/li\u003e\n\u003cli\u003eWang E, Zhu M, Lin Y, Xi X. Multilevel medical insurance mitigate health cost inequality due to air pollution: Evidence from China. International Journal for Equity in Health. 2024; 23:153.\u003c/li\u003e\n\u003cli\u003eZhou L, Zhong Q, Yang J. Air Pollution and Household Medical Expenses: Evidence From China. Frontiers in Public Health. 2022; Volume 9 - 2021.\u003c/li\u003e\n\u003cli\u003eSeyler BC, Luo H, Wang X, Zuoqiu S, Xie Y, Wang Y. Assessing the impact of urban greenspace on physical health: An empirical study from Southwest China. Front Public Health. 2023; 11:1148582.\u003c/li\u003e\n\u003cli\u003eZhang L, Wu Y. Negative Associations between Quality of Urban Green Spaces and Health Expenditures in Downtown Shanghai. Land [serial on the Internet]. 2022; 11(8).\u003c/li\u003e\n\u003cli\u003eWei Y, Wang Y, Di Q, Choirat C, Wang Y, Koutrakis P, et al. Short term exposure to fine particulate matter and hospital admission risks and costs in the Medicare population: time stratified, case crossover study. Bmj. 2019; 367:l6258.\u003c/li\u003e\n\u003cli\u003eZhang C, Zhang L. The relationship between toxic air pollution, health expenditure, and economic growth in the European Union: fresh evidence from the PMG-ARDL model. Environmental Science and Pollution Research. 2024; 31:21107\u0026ndash;23.\u003c/li\u003e\n\u003cli\u003eJaafar H, Razi NA, Azzeri A, Isahak M, Dahlui M. A systematic review of financial implications of air pollution on health in Asia. Environ Sci Pollut Res Int. 2018; 25:30009\u0026ndash;20.\u003c/li\u003e\n\u003cli\u003eSong Y, Chen B, Ho HC, Kwan M-P, Liu D, Wang F, et al. Observed inequality in urban greenspace exposure in China. Environment International. 2021; 156:106778.\u003c/li\u003e\n\u003cli\u003eCao Y, Li G. Extensive inequality of residential greenspace exposure within urban areas in China. Science of The Total Environment. 2024; 948:174625.\u003c/li\u003e\n\u003cli\u003eShandong Provincial Government. Shandong Province ecological environment status report.\u003c/li\u003e\n\u003cli\u003eShandong Provincial Bureau of Statistics. Statistical Yearbook. Available from: http://tjj.shandong.gov.cn/col/col6279/index.html.\u003c/li\u003e\n\u003cli\u003eYu Y, Liu D, Hu S, Shi X, Tang J. Spatiotemporal Heterogeneity of Vegetation Cover Dynamics and Its Drivers in Coastal Regions: Evidence from a Typical Coastal Province in China. Remote Sensing [serial on the Internet]. 2025; 17(5).\u003c/li\u003e\n\u003cli\u003eXiong C, Ma H, Liang S, He T, Zhang Y, Zhang G, et al. Improved global 250\u0026thinsp;m 8-day NDVI and EVI products from 2000-2021 using the LSTM model. Sci Data. 2023; 10:800.\u003c/li\u003e\n\u003cli\u003eGascon M, Cirach M, Mart\u0026iacute;nez D, Dadvand P, Valent\u0026iacute;n A, Plas\u0026egrave;ncia A, et al. Normalized difference vegetation index (NDVI) as a marker of surrounding greenness in epidemiological studies: The case of Barcelona city. Urban Forestry \u0026amp; Urban Greening. 2016; 19:88\u0026ndash;94.\u003c/li\u003e\n\u003cli\u003eWei J, Li Z, Lyapustin A, Wang J, Dubovik O, Schwartz J, et al. First close insight into global daily gapless 1\u0026thinsp;km PM2.5 pollution, variability, and health impact. Nature Communications. 2023; 14:8349.\u003c/li\u003e\n\u003cli\u003eLi Q, Dong L, Zhang L. Have pensions reduced the relative poverty? ----- empirical analysis from China CHARLS data. Heliyon. 2023; 9:e22711.\u003c/li\u003e\n\u003cli\u003eErreygers G, Van Ourti T. Measuring socioeconomic inequality in health, health care and health financing by means of rank-dependent indices: a recipe for good practice. J Health Econ. 2011; 30:685\u0026ndash;94.\u003c/li\u003e\n\u003cli\u003eGravelle H. Measuring income related inequality in health: standardisation and the partial concentration index. Health Econ. 2003; 12:803\u0026ndash;19.\u003c/li\u003e\n\u003cli\u003eTwohig-Bennett C, Jones A. The health benefits of the great outdoors: A systematic review and meta-analysis of greenspace exposure and health outcomes. Environmental Research. 2018; 166:628\u0026ndash;37.\u003c/li\u003e\n\u003cli\u003eWang S, Sun J, Xu Z, Di Tanna GL, Chen M, Downey LE, et al. Residential greenspace and multiple chronic health conditions in China: a cross-sectional study. J Glob Health. 2025; 15:04218.\u003c/li\u003e\n\u003cli\u003eDominski FH, Lorenzetti Branco JH, Buonanno G, Stabile L, Gameiro da Silva M, Andrade A. Effects of air pollution on health: A mapping review of systematic reviews and meta-analyses. Environmental Research. 2021; 201:111487.\u003c/li\u003e\n\u003cli\u003eWerder E, Lawrence K, Deng X, Braxton Jackson W, Christenbury K, Buller I, et al. Residential air pollution, greenspace, and adverse mental health outcomes in the U.S. Gulf Long-term Follow-up Study. Science of The Total Environment. 2024; 946:174434.\u003c/li\u003e\n\u003cli\u003eZhao Y, Yang S, Zhang X. What does poor air quality lead to? - The influence of air pollution on elderly medical expenditures in China. Cities. 2026; 169:106543.\u003c/li\u003e\n\u003cli\u003ePi T, Wu H, Li X. Does Air Pollution Affect Health and Medical Insurance Cost in the Elderly: An Empirical Evidence from China. Sustainability [serial on the Internet]. 2019; 11(6).\u003c/li\u003e\n\u003cli\u003eHou Z, Kim MJ, Zhang N. Air pollution, defensive behaviors, and medical expenditures. Journal of Asian Economics. 2025; 98:101931.\u003c/li\u003e\n\u003cli\u003eBi Y, Wang Y, Yang D, Mao J, Wei Q. Urban green spaces and resident health: an empirical analysis from data across 30 provinces in China. Frontiers in Public Health. 2024; Volume 12 - 2024.\u003c/li\u003e\n\u003cli\u003eSadeh M, Brauer M, Dankner R, Fulman N, Chudnovsky A. Remote sensing metrics to assess exposure to residential greenness in epidemiological studies: A population case study from the Eastern Mediterranean. Environment International. 2021; 146:106270.\u003c/li\u003e\n\u003cli\u003eWeng Z, Tong D, Wu S, Xie Y. Improved air quality from China\u0026apos;s clean air actions alleviates health expenditure inequality. Environ Int. 2023; 173:107831.\u003c/li\u003e\n\u003cli\u003eZHANG J, JIANG H, ZHANG W, MA G, WANG Y, LU Y, et al. Cost-benefit analysis of China\u0026apos;s Action Plan for Air Pollution Prevention and Control. ENGINEERING Management. 2019; 6:524\u0026ndash;37.\u003c/li\u003e\n\u003cli\u003eLi S, Wu D, Liu L, Yang L, Wang Y, Cao S, et al. Is it worth implementing the Blue Sky Defense Battle initiative? A cost-benefit analysis of the Chengdu case. Integrated Environmental Assessment and Management. 2025; 21:425\u0026ndash;41.\u003c/li\u003e\n\u003cli\u003eZhou Q, Chen J, van den Bosch CCK, Zhang W, Zhu L, Vera YMR, et al. Constructing an Aims-Indicators-Methods framework for Green Space System Planning in China. Urban Forestry \u0026amp; Urban Greening. 2022; 67:127437.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"international-journal-for-equity-in-health","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"ijeh","sideBox":"Learn more about [International Journal for Equity in Health](http://equityhealthj.biomedcentral.com)","snPcode":"12939","submissionUrl":"https://submission.nature.com/new-submission/12939/3","title":"International Journal for Equity in Health","twitterHandle":"@equityhealthj","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Air pollution, Greenspace, Household healthcare expenditure, Health inequality","lastPublishedDoi":"10.21203/rs.3.rs-9378849/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9378849/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eEnvironmental factors such as air pollution and access to greenspace are increasingly recognised as important determinants of population health and healthcare expenditure. This study examined the relationship between ambient PM2.5, urban greenspace, and household healthcare expenditure in China, and assessed how healthcare spending was concentrated across these gradients of environmental exposure.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eWe conducted a cross-sectional analysis using data from the 6th Health Services Survey in Shandong Province, linking household healthcare expenditure to residential greenspace (NDVI) and ambient PM2.5 at the village level. We used Generalised Linear Mixed Models with village-level random effects to estimate the relationship between household healthcare expenditure and environmental exposures. To examine inequalities in healthcare expenditure, we calculated concentration indices (CIs), quantifying the distribution of healthcare expenditure across levels of greenspace and air pollution.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eA total of 27,603 individuals were included in the analysis. Higher NDVI exposure was associated with lower household medical expenditure (Q2: β = \u0026minus;0.21, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01; Q3: β = \u0026minus;0.21, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01; Q4: β = \u0026minus;0.26, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01), whereas higher ambient PM2.5 concentrations were linked to increased expenditure (Q2: β\u0026thinsp;=\u0026thinsp;0.13, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01; Q3: β\u0026thinsp;=\u0026thinsp;0.17, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01; Q4: β\u0026thinsp;=\u0026thinsp;0.30, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01). In both models, older age (45\u0026ndash;65 and \u0026ge;\u0026thinsp;65 years) and underweight status were associated with higher costs (NDVI: β\u0026thinsp;=\u0026thinsp;0.13\u0026ndash;0.32, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01; PM2.5: β\u0026thinsp;=\u0026thinsp;0.15\u0026ndash;0.35, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01), as were households earning above 60% of median income (NDVI: β\u0026thinsp;=\u0026thinsp;0.10, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01; PM2.5: β\u0026thinsp;=\u0026thinsp;0.11, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01). Health insurance was linked to lower expenditure in the NDVI model (β = \u0026minus;0.10, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) but not in the PM2.5 model. Inequality analyses indicated that household healthcare expenditure was disproportionately concentrated among residents with lower greenspace (NDVI: PCI\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.04, 95% CI: \u0026minus;0.05 to \u0026minus;\u0026thinsp;0.03) and higher PM2.5 exposure (PCI\u0026thinsp;=\u0026thinsp;0.06, 95% CI: 0.05 to 0.07).\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eOur findings highlight the importance of integrating environmental equity into public health and policy interventions to reduce healthcare costs and inequalities.\u003c/p\u003e","manuscriptTitle":"Ambient PM2.5, residential greenspace, and household healthcare expenditure in Shandong, China","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-03 09:56:53","doi":"10.21203/rs.3.rs-9378849/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-05-07T23:08:28+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-06T04:19:42+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"160094844937246051058035319225439041644","date":"2026-04-22T07:31:20+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-22T05:41:02+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-15T17:44:43+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-15T17:44:40+00:00","index":"","fulltext":""},{"type":"submitted","content":"International Journal for Equity in Health","date":"2026-04-10T11:17:47+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"international-journal-for-equity-in-health","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"ijeh","sideBox":"Learn more about [International Journal for Equity in Health](http://equityhealthj.biomedcentral.com)","snPcode":"12939","submissionUrl":"https://submission.nature.com/new-submission/12939/3","title":"International Journal for Equity in Health","twitterHandle":"@equityhealthj","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"aa7f4e27-e85b-4697-b4e3-fe010397c51c","owner":[],"postedDate":"May 3rd, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Revision requested","date":"2026-05-07T23:08:28+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-06T04:19:42+00:00","index":22,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[],"tags":[],"updatedAt":"2026-05-07T23:23:28+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-03 09:56:53","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9378849","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9378849","identity":"rs-9378849","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}