The Evolution of Ecological Well-Being Performance and Its Effects on Population Longevity: A County-Level Spatiotemporal Analysis of Hubei Province, China

Abstract

Building an ecological civilization and promoting national health are crucial for high-quality development. These goals are linked to ecological well-being performance (EWP). This study aimed to evaluate EWP based on county-scale input–output relationships, analyze its spatiotemporal evolution, and explore how EWP changes affect longevity through spatial spillover and interaction mechanisms. We first used the super-SBM model to assess county-level EWP from 2000 to 2020. Then, spatial econometric models and geographical detectors were applied to analyze the impact of EWP on longevity. The results show a persistent uptrend in overall EWP, indicating that Hubei Province has enhanced its sustainable development capacity. Regions with high EWP values have distinct characteristics. There is polarization in the east, expansive connectivity in the west, and fragmentation in the center, forming a clear “core–edge” structure. The improvement in EWP directly promotes male, female, and overall population longevity and has spatial spillover effects. EWP also interacts with the natural environment and socioeconomic development, serving as a key factor promoting population longevity within Hubei Province. These findings provide a reference for regions in China or other developing countries to understand the relationships between the extension of population lifespan and regional sustainable development.

1. Introduction

In the past few decades, countries worldwide have made efforts to coordinate the relationship between human beings and the natural ecological environment in order to increase the level of human well-being. China has experienced rapid socioeconomic development since the reform and opening up, yet the extensive mode of economic growth has caused serious environmental pollution and resource constraints [1], posing a great threat to public health [2]. After the 18th National Congress of the Communist Party of China, a series of goals and policies concerning the construction of an ecological civilization and green development were proposed, which strongly supported high-quality development in China. As in most developed countries, the proportion of elderly people aged 80 years and older in China has been gradually increasing, and the total number of centenarians in 2020 was already 6.65 times greater than that 20 years ago. The China Association of Gerontology and Geriatrics (CAGG) posits that China has entered an era of population longevity [3]. The relationship between human health and the environment is regarded as the most fundamental human–land relationship [4], and harmonious and sustainable human–land relationships positively affect human beings. Hence, clarifying regional differences in sustainability and how these differences affect human lifespan is highly important for coordinated regional development in the era of population longevity. This study addresses these aims by analyzing regional ecological well performance and investigating the role of its change in contributing to human longevity.

Ecological well-being performance (EWP) is an indicator of the efficiency with which natural ecological resource expenditure is transformed into human well-being [5]. EWP takes into account economic, environmental, and social dimensions comprehensively, illustrating the extent of coordination between advancements in human well-being and ecological resource utilization [6]. In comparison to other indicators, EWP matches the requirement for the synergistic integration of the “environment–economy–welfare” triad, which is in harmony with the developmental objectives of human society [7]. Therefore, EWP has been widely used as an effective tool for measuring the degree of regional sustainable development [8]. Regional differences in EWP and population longevity and the causes of their formation have been widely studied by researchers [9,10,11,12]. To our knowledge, however, few previous studies have focused on the regional effects of changes in EWP [13], and no study has investigated what types of regional health effects are produced by changes in EWP. Moreover, macro scales, such as countries, provinces, and cities, are the main research units in existing studies [14,15], and less attention has been given to spatial interactions.

To fill these gaps, this study attempted to conduct an analysis with Hubei Province in China as its study area. Hubei Province has a well-known human longevity phenomenon and plays an important role in the Rise of Central China strategy. In recent years, it has undergone economic transformation and upgrading, moving forward toward high-quality development. The COVID-19 pandemic raised alarms about economic and social vulnerabilities. Hubei Province endured significant population health impacts during the crisis. Exploring the relationship between EWP and population lifespan in Hubei is crucial for addressing health risks in the post-pandemic era. This study aimed to evaluate EWP on the basis of the input–output relationship at the county scale and analyze its spatiotemporal evolution, on the one hand, and to examine and summarize the mechanism of influence of EWP changes on population longevity through the lens of spatial spillover and interaction on the other hand.

With these aims, the contributions of this research can be summarized as follows. First, we construct a county-level evaluation index system of EWP, which is beneficial for identifying the characteristics of regional disparity and evolution of EWP at a relatively fine spatial scale. To do that, we systematically reviewed studies that combined the use of geographic information data. Following the measurement of inputs related to resource consumption and environmental pollution and outputs related to economic development, social security, and environmental optimization, we conducted a county-level spatiotemporal analysis of EWP in Hubei Province. These insights allow policymakers to take targeted measures through a precise view of the regional imbalance of EWP.

Second, the study complements the literature on the health effects of EWP changes, especially the spatial spillover effect. Specifically, improving EWP can not only benefit local human longevity directly but also have a positive effect on neighboring human longevity indirectly. This finding expands the existing research that has focused on the feedback of EWP changes on the economy and environment, and it also restates the significance of interregional equity, indicating that to achieve sustainable human–land development, attention should be given to the coordination among regions.

Third, this study also enriches the geographical analysis of population longevity. We summarized the impact of ecological and environmental changes on population longevity from the perspectives of spatial effects and the moderating role of improvements in human–land relations. We constructed a mechanism framework for the impact of EWP improvement on population longevity. The experience of Hubei Province can provide an important reference for other provinces in China and even for the Global South, such as India and African countries, to coordinate human–land relations and extend human lifespans.

The research framework is shown in Figure 1, and the structure of this article is as follows. Section 2 is the literature review. The research materials and methods are presented in Section 3, which is followed by the research results in Section 4. Then, we discuss the findings in Section 5. Finally, Section 6 gives the conclusions.

2. Literature Review

The origin of EWP can be traced back to a concept in steady-state economics introduced by Daly [16]. The concepts of ecological footprint and the human development index were subsequently developed [17]. A variety of approaches are used to measure EWP, such as the ratio between the level of human well-being and ecological consumption [18,19] and mathematical modeling methods, including the stochastic frontier analysis (SFA) model [20] and the data envelopment analysis (DEA) model [21,22].

The differences across regions and the evolution of EWP have received increasing attention from scholars [23,24,25]. In China, EWP tends to decrease from east to west and south to north [26]. Scholars specializing in urban agglomerations in China have reported that EWP presents a multicore network radiation pattern in the Beijing–Tianjin–Hebei urban agglomeration [14] and an obvious core–edge structure distribution in the urban agglomeration in the middle reaches of the Yangtze River [27]. The imbalanced spatial distribution of EWP across geographical scales is similar to the differences in economic development. Therefore, many researchers have analyzed the impact of economic development on EWP and found that it is complex and variable [28,29]. In addition, other factors influencing EWP spatial patterns, such as urbanization [30], industrial agglomeration [31], environmental regulation, and green transformation [26,32], have been detected.

The promotion of EWP entails a decrease in the costs associated with obtaining the same level of benefits, which generates feedback for the various components of the sustainable development system. Xiao and Zhang [33] proposed that EWP enhancement creates feedback for green innovation efficiency via economic growth, social tolerance, and environmental modification. Moreover, a higher EWP positively contributes to the efficiency of intensive land use through the demand for spatial optimization and the sustainable utilization of idle land [34]. Zhu et al. [13] indicated that EWP enhancement shifts from restraining economic growth to promoting it as the development mode of the economy shifts from extensive to intensive, and it has a continuous and intensified impact on carbon emission reduction in general. Although researchers have begun to consider the effects of changes in EWP, few in-depth studies have been conducted, especially on the effects on population health.

An increase in human lifespan is an important manifestation of population health. Population longevity is the result of a combination of subjective (genes, behavioral habits, etc.) and objective (health care, habitat, etc.) conditions [35,36], and the latter matters more [37]. Population longevity levels are often evaluated by the proportion of elderly people, the proportion of centenarians, and average life expectancy at birth [10,38,39]. Studies have explored the spatiotemporal patterns of population longevity levels in China. Many of the areas with high longevity in China are located east of the Huhuanyong Line, indicating that the suitability of climate and terrain [40,41] and the degree of socioeconomic development significantly affect population longevity [42]. Air-pollutant exposure and poor green space coverage are barriers to population longevity [43]. Environmental elements are often integrated to influence the lifespan of a population, resulting in comprehensive effects [44]. However, the impacts of such a change in the degree of coordination between ecological, economic, and social environments, i.e., the degree of sustainable development, have been less explored.

Reviewing previous studies, the following research gaps still need to be addressed. First, most existing EWP studies focus on national, provincial, or prefecture city scales. However, it is necessary to analyze EWP at the finer county scale in China. County-level analysis can more accurately reveal spatial patterns; moreover, the county is an important carrier in the new urbanization construction of China [45]. Second, the health effects of EWP changes need to be further explored, especially through the lens of spatial spillovers. In addition, because of the gender heterogeneity in the spatial distribution of longevity [10], the effects of EWP on population longevity should be explored by gender. With the release of the results of China’s fifth, sixth, and seventh population censuses, analysis can be conducted on the basis of county-scale data to explore the effects of EWP changes in terms of population longevity.

3. Materials and Methods

3.1. Study Area

Hubei Province is located in central China, in the middle reaches of the Yangtze River, and has a provincial area of 185,900 km² (Figure 2). The average annual temperature in Hubei Province is 15 °C to 17 °C, and the average annual precipitation is between 800 and 1600 mm. Its internal terrain is diverse, supporting the division of the province into various areas: the northeastern Hubei hilly area, southeastern Hubei hilly area, northern Hubei hillock plain area, Jianghan plain, northwestern Hubei mountainous area, and southwestern Hubei mountainous area [46].

In the early 2000s, with the pursuit of rapid economic growth, environmental pollution and ecological damage were extremely severe in Hubei Province. Following the construction of the Yangtze River Economic Belt and the Yangtze River Protection Project, the transformation and upgrading of the economy, combined with ecological restoration, have begun to achieve positive effects, creating a significant change in the level of sustainable development within the province of Hubei.

Currently, there are 13 prefecture-level cities and 103 county-level administrative units in Hubei Province. To scientifically and accurately analyze the evolution of EWP spatial distribution, the 103 counties under the 2020 administrative division were used as the research units in each year covered by the study.

3.2. Evaluation Indicators

3.2.1. Ecological Well-Being Performance

To measure EWP, we established an index system based on various indicators from the perspective of the input–output relationship. According to this perspective, EWP consists of input factors, which refer mainly to ecological resource consumption, and output factors, which ultimately refer to human well-being.

For resource consumption, owing to differences in research scales, there are variations in the selection of resource consumption indicators. In general, it is more common and necessary to select indicators to measure the consumption of energy and land resources than to measure the consumption of water resources [47]. Moreover, it is currently difficult to find unified and comprehensive indicators and data for measuring water resource consumption at the county scale in China. Therefore, according to the research of Ma and Zhang [48], we mainly considered land resource consumption, which is measured by the developed area and cropland area of a county, and energy consumption, which is measured by the energy index. The energy index is calculated according to the process of Liu et al. [49]. Related studies have shown a significant linear relationship between nighttime light data and energy consumption. Energy consumption in daily life and production can be simulated with nighttime light data. We obtained the energy index by summing the raster values of nighttime lights in each county. According to Bian et al. [50], Guo and Ou [51], and other related studies, even though the labor force is not a natural resource, it serves as important human capital in the whole process, from ecological inputs to well-being outputs. Therefore, the labor force was considered a non-resource input and was measured as the population aged 15–64. Environmental pollution is usually manifested as an undesirable output, but considering the meaning of EWP, environmental pollution can be regarded as a type of cost of environmental damage that affects the output of human well-being [12,14]. Therefore, the environmental cost was included in the input side of EWP and was measured by the concentration of PM_2.5.

For human well-being outputs, the three aspects of economic, social, and environmental well-being were considered, with reference to the human development index (HDI) and existing research. Economic well-being was measured by GDP per capita [31], which indicates economic development. Social well-being was measured by universal education, health care, and overall quality of life, according to the indicators of the average years of schooling [22], the number of beds in health institutions per 1000 persons [52], and average life expectancy at birth [53]. It is difficult to obtain uniform environmental indicators for the measurement of environmental well-being at the county level. Therefore, the normalized difference vegetation index (NDVI), which reflects the degree of green space coverage and spatial greening level [54,55], was selected to indicate the spatial afforested level in a county as a measure of environmental well-being. The evaluation indices of county EWP are shown in

Table 1. Evaluation index of county EWP.

Dimension	Criteria	Two-Level Index	Indicators	Unit	Reference
Input indicators	Land resource input	Construction land consumption	Developed area	km2	Ma et al.(2019) [47]; Ma & Zhang (2021) [48]
		Cropland consumption	Cropland area	km2	Ma et al.(2019) [47]
	Energy input	Energy consumption	Energy consumption index	(index)	Ma & Zhang (2021) [48]
	Ecological environment destruction	Environmental quality	PM2.5 concentration	µg/m3	Zhu et al. (2022) [12]; Xia & Li (2022) [14]
	Nonecological resource input	Labor force	Population aged 15 to 64	104 person	Bian et al. (2020) [50]; Guo & Ou (2023) [51]
Output indicators	Economic well-being	Level of economic development	GDP per capita	CNY	Liu et al. (2024) [31]
	Social well-being	Universal education	Average years of schooling	year	Han et al. (2025) [22]
		Health care	Number of beds in health institutions per 1000 persons	bed/1000 persons	Hu et al. (2021) [52]
		Overall quality of life	Average life expectancy at birth	year	Deng at al. (2021) [53]
	Environmental well-being	Level of favorable environment	NDVI	(index)	Leslie et al. (2010) [54]; Wilker et al. (2014) [55]

.

3.2.2. Human Longevity Level

The longevity index is used to indicate the longevity of the age structure of a population. The benchmark for measuring old age is usually age 60 or 65 globally. Longevity can be regarded as the further aging of the elderly population. In general, 80 years of age is considered the threshold for longevity, and scholars typically use the proportion of the population aged 80 to 90 years old and older to reflect longevity progression [56]. Although many studies use life expectancy to measure population lifespan, this indicator, based on the prospective estimation of current age-specific mortality rates, struggles to reflect the actual survival status and spatial distribution characteristics of the elderly long-lived population. While the number or proportion of centenarians is also used to measure extreme longevity, the rarity and spatially discrete distribution of centenarians make related indicators prone to sample sparsity issues in county-level studies, thus failing to effectively capture the collective longevity level and spatial associations. With reference to existing studies [10], the longevity index was constructed as the ratio of the population aged 85 and older to the population aged 65 and older. The longevity index was used as an indicator of the general longevity level of a county.

3.3. Models

3.3.1. Super-SBM Model

In this study, the super-SBM model was selected for the measurement of EWP. Traditional data envelopment analysis (DEA) suffers from the problem of slack input and output variables, which makes the results inaccurate. The super-SBM model, introduced by Tone [57], represents an improvement over the traditional DEA model and better supports efficiency sorting among multiple simultaneously effective DMUs, which helped us to analyze the differences in EWP more accurately [58]. The specific formulas are as follows:

$$\rho ={\displaystyle \frac{1-\frac{1}{M}{\displaystyle \sum _{m=1}^M\frac{s_m^{-}}{x_{mk}}}}{1+\frac{1}{N}{\displaystyle \sum _{n=1}^N\frac{s_n^{+}}{y_{nk}}}}}$$

(1)

$$\mathrm{Subject} \mathrm{to}\left\{\begin{array}{l}{x}_k={\displaystyle \sum _{i=1}^I{\lambda }_i{x}_{mi}+s_m^{-},m=1,2, \dots , M}\\ y_k={\displaystyle \sum _{i=1}^I{\lambda }_i{y}_{ni}+s_n^{+}, \mathrm{n}=1,2, \dots , N}\\ {\displaystyle \sum _{i=1}^I{\lambda }_i=1, i=1,2, \dots , I}\\ {\lambda }_i\ge 0, s_m^{-}\ge 0, s_n^{+}\ge 0 \end{array}\right.$$

(2)

In Equations (1) and (2), ρ represents the EWP of the research unit; M and N refer to the number of input and output variables, respectively; I represents the number of DMUs; $s_m^{-}$ and $s_n^{+}$ represent the slacks of the inputs and outputs, respectively; $x_m$ and $y_n$ represent the input and output vectors, respectively; $x_{mk}$ and $y_{nk}$ represent the input and output, respectively; ${\lambda }_i$ represents the weight coefficients; and i represents DMU_i, i.e., county i.

3.3.2. Spatial Autocorrelation Model

Spatial autocorrelation analysis can measure the spatial agglomeration characteristics of the study objects on the basis of the similarity of their spatial locations and attributes [59], so we conducted it to determine the spatial clustering characteristics of EWP and population longevity levels. The global spatial autocorrelation results reported by Moran’s I are also important for determining whether variables can be used for further spatial regression analyses. Moran’s I is calculated as follows:

$$Moran’s I={\displaystyle \frac{n{\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j\ne 1}^n{W}_{ij}(x_i-\overline{x})(x_j-\overline{x})}}}{{\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j\ne 1}^n{W}_{ij}{\displaystyle \sum _{i=1}^n{\left(x_i-\overline{x}\right)}^2}}}}}$$

(3)

In Equation (3), n is the number of counties; x_i and x_j denote the values of the attributes explored in counties i and j, respectively; and $\overline{x}$ represents the mean values of the attributes explored. W_ij refers to the spatial weight matrix. We used a spatial inverse distance weight matrix calculated from the inverse of the distance between counties. Moran’s I values ranged between −1 and 1. Moran’s I > 0 indicates the existence of spatial clustering, and vice versa. When Moran’s I is 0, the data are randomly distributed.

3.3.3. Spatial Econometric Model

Spatial econometric models were used to explore the direct effects of EWP variation on population longevity and the spatial spillover effects. The major spatial econometric models are the spatial lag model (SLM), spatial error model (SEM) and spatial Durbin model (SDM). The general formula is as follows:

$$\left\{\begin{array}{l}{Y}_{it}=\rho W{Y}_{it}+\phi W{X}_{it}+\beta X_{it}+{\epsilon }_{it}\\ {\epsilon }_{it}=\lambda W{\epsilon }_{it}+{\nu }_{it}\end{array}\right.$$

(4)

In Equation (4), i refers to county i, t is the year, Y_it is the dependent variable, X_it refers to the independent variable, W is the spatial weight matrix, ${\epsilon }_{\mathit{it}} \mathrm{and} {\upsilon }_{\mathit{it}}$ denote the random error, β refers to the regression coefficients, ρ is the spatial lag coefficient of the dependent variable, and $\phi$ is the spatial lag coefficient of the independent variable. When $\lambda$ = 0, the above equation represents the spatial Durbin model; when $\lambda$ = 0 and $\phi$ = 0, the above equation represents the spatial lag model; and when $\rho$ = 0 and $\phi$ = 0, the above equation represents the spatial error model.

The scope of spatial spillover effects at the county level is relatively limited, with spatial impacts likely confined to proximate regions. Traditional inverse distance spatial weight matrices incorporate all spatial units into the consideration of spillover effects. In analyses of spatial spillovers at the county scale, such matrices—by accounting for all study units—may introduce noise from distant, weakly correlated spatial units, potentially distorting the results of the models. To address this issue, a K-nearest neighbor (KNN) spatial weight matrix offers a superior alternative, as it restricts spatial spillover effects to the closest K units for localized consideration [60]. However, conventional KNN matrices assign an equal weight of 1 to all neighbors, neglecting the distance decay of spatial effects across different proximities.

Thus, this study integrated the inverse distance and KNN approaches, employing a distance-weighted KNN spatial weight matrix in modeling. The distance-weighted KNN matrix thus combined the principles of geographic contiguity and distance decay to evaluate spatial spillovers at the county scale. This matrix assigned different spatial weights to the K nearest neighbors based on their distances to the focal unit, enabling a more precise measurement of how local EWP impacts population longevity in adjacent regions. To select the optimal K value, this study compared the Akaike information criterion (AIC) and Bayesian information criterion (BIC) of models under different K values to identify the model with the best performance [61]. Through the statistics of the number of adjacent counties in each county of Hubei Province, it was found that the number of counties with 5 adjacent counties was the largest, followed by those with 6 adjacent counties. The number of adjacent counties mainly ranged from 3 to 7. Therefore, in this study, distance-weighted KNN spatial weight matrices were constructed with K ranging from 3 to 7 for comparison.

3.3.4. Geographical Detector

We used a geodetector to analyze the effects of EWP interactions with environmental elements on population longevity levels. Geodetectors allow the examination of the relationships between variables on the basis of the detection of spatially stratified heterogeneity. The basic assumption of a geodetector is that if a variable has a significant effect on another variable, the spatial distributions of the two variables should be similar. Geodetectors describe the associations between variables via the q statistic (Equation (5) below). In hierarchical comparisons, the stratification between variables can be in terms of geographical space or attributes, so geodetectors can be used flexibly [62].

$$q=1-{\displaystyle \frac{{\displaystyle \sum _{\mathrm{s}=1}^L{N}_s{\sigma }_s^2}}{N{\sigma }^2}}$$

(5)

In Equation (5), s = 1, …, L is the stratification of variable Y or X; N_s and N are the numbers of strata and all samples, respectively; ${\sigma }_{\mathrm{s}}^2$ and ${\sigma }^2$ are the variances of the Y values of strata s and all samples, respectively; and q takes a value of [0, 1]. The larger the q statistic is, the more similar the spatial distributions of variables X and Y are, and the greater the effect of variable X on variable Y is. By calculating the q value when stratifications of different variables X1 and X2 are superimposed, the geodetector can detect various kinds of interactions among different X variables without being limited to the multiplicative interactions required by econometrics. According to the change in the effect q value of different X factors after superpositioning relative to the original X, the interaction effects can be classified into 5 types, as shown in

Table 2. Interaction types of geographical detectors.

q Value	Interaction Type
q (X1 ∩ X2) < min [q (X1), q (X2)]	Non-linear weakening
min [q (X1), q (X2)] < q (X1 ∩ X2) < max [q (X1), q (X2)]	Single-factor weakening
q (X1 ∩ X2) = q (X1) + q (X2)	Independent
q (X1 ∩ X2) > max [q (X1), q (X2)]	Bivariable enhancement
q (X1 ∩ X2) > q (X1) + q (X2)	Non-linear enhancement

.

3.4. Data Sources

The vector maps used in the study were drawn from the 2020 Administrative District Map of Hubei Province obtained from the Department of Natural Resources of Hubei Province. The land area data were obtained from China’s National Land Use and Cover Change (CNLUCC) datasets. Nighttime light data, which measure energy consumption, were obtained from China DMSP-OLS-like data with a 1 km spatial resolution. The PM_2.5 particulate matter concentration data, which measure environmental pollution, come from the 1 km resolution annual PM_2.5 datasets for the Yangtze River Economic Belt. The demographic data required for the calculation of the labor force, average schooling years, life expectancy, and longevity levels were collected from the 5th, 6th, and 7th Hubei Provincial Population Censuses published by the Hubei Provincial Statistics Bureau. The NDVI data were calculated from the MODIS dataset released by NASA and integrated by month on the basis of MOD13Q1 image extraction. Other physical geography data used in the study were obtained from the National Earth System Science Data Center. Socioeconomic data were obtained from the statistical yearbooks or statistical bulletins of national economic and social development from cities, states, and counties in Hubei Province. The GDP per capita was deflated to eliminate price changes, taking 2000 as the base period. Analyses were conducted using MATLAB 2021, ArcGIS Pro 3.1, Stata 17, and R 4.4.1.

4. Results

4.1. Spatial–Temporal Evolution of EWP in Hubei Province

4.1.1. Temporal Changes and Regional Differences in EWP in Hubei Province

To assess EWP in Hubei Province, we first aimed to determine how EWP evolved at the county level from 2000 to 2020 by measuring the established indicators via the super-SBM model, and then visualized the results of the average level of EWP in Hubei Province and its subregions (Figure 3). Overall, in Hubei Province, EWP presented a continuously increasing trend, from 0.187 in 2000 to 0.243 in 2010 and then to 0.413 in 2020. This result reflects the overall increase in the capacity for sustainable development in the province, with unit resource consumption and environmental cost gradually transforming into greater human well-being. At the subregional level, western Hubei consistently led in terms of EWP. The EWP of the southeastern Hubei hilly area and Jianghan plain steadily increased, and the EWP of the northeastern Hubei hilly area, northern Hubei hillock plain area, and northwestern Hubei mountainous area increased swiftly over the past ten years, whereas the EWP of the southwestern Hubei mountainous area first decreased but then increased. Since the 18th National Congress of the CPC in 2012 promoted the construction of an ecological civilization as a national strategy, Hubei Province has actively promoted the protection of the eco-environment and the adjustment of industrial development in cities and towns, eliminating outdated production facilities and carrying out targeted pollution control. The level of green and low-carbon development in Hubei Province has increased steadily overall, improving the quality of life and the general well-being of society.

4.1.2. Spatial Pattern Evolution of Hubei Province’s EWP

The spatial distribution of EWP was visualized in ArcGIS 10.7 with grouping by the natural breaks method (Figure 4). Almost all of the areas with high EWP in 2000 were located in mountainous areas in western Hubei Province, whereas eastern Hubei Province began to present high EWP values in later years. In 2010 and 2020, several of the main urban districts of Wuhan (Jianghan, Wuchang, Hongshan, Dongxihu) gradually emerged as the main high-EWP zones in eastern Hubei Province as counterparts to the western part of the province. However, the level of EWP at the outskirts of this high-value zone decreased dramatically. This is mainly because the economic and social development of the main urban districts brought about the most favorable well-being conditions. Moreover, in the last decade, various urban problems were first addressed at the core of the cities; thus, more attention has been given to coordination among the population, economy, society, and environment in these areas. The high-value areas in the western region are connected mainly by high-value areas and sub-high-value areas, with a stable distribution in the border area of southwestern Hubei. The scope of the high-value area in the southwestern Hubei mountainous area decreased in 2010 but then rebounded in 2020, as the average EWP value first decreased and then increased. The EWP changes in the southwestern mountainous area mirror the transformation of the economic development mode within it. The EWP tends to remain low or decline during early periods, accelerating extensive economic growth and later increasing.

We conducted a global spatial autocorrelation analysis of EWP in Hubei Province by year (

Table 3. Moran’s I of EWP in Hubei Province.

Year	Moran’s I	Z	p
2000	0.126	2.32	0.02
2010	0.296	5.1	<0.01
2020	0.391	6.655	<0.01

). The results show that Moran’s I increased gradually over the three measurement years, with all values passing the test at the 5% significance level. This illustrates the significant spatial clustering of EWP in the counties of Hubei Province, which gradually increased. The top 10 counties in terms of EWP across the three years were compared, and more than 85% of them were municipal districts. The average EWP value of all the municipal districts at the three time points is 0.418, which is higher than the average value of all the counties in Hubei Province (0.281) and the average value of the nonmunicipal district counties (0.2). Therefore, the EWP in Hubei Province shows a very obvious “core–edge” structure during the study period. This indicates that the capacity for regional sustainable development in Hubei Province is still unbalanced. With the advancement of the Great Protection of the Yangtze River and ecological civilization construction, the extensive production modes in urban areas have been transformed. However, urban core areas, endowed with superior development conditions, often exhibit a “siphon effect” on their peripheral zones. The limited radiation effect of core areas is particularly evident in the county-level regions of Hubei Province [45]. Consequently, the regional polarization effect of EWP is highly pronounced. A similar EWP core–edge structure has also been observed in previous studies [12,27].

4.2. Spatial Effects of EWP on Human Longevity

4.2.1. Model Processing

We further explored the health effects of changes in EWP on population longevity levels after conducting spatial and temporal pattern analysis. We took the population longevity level as the dependent variable and EWP as the core independent variable. We conducted a global spatial autocorrelation analysis of overall longevity, male longevity, and female longevity (

Table 4. Moran’s I of longevity levels of different population groups in Hubei Province.

Year	Overall Longevity			Male Longevity			Female Longevity
	Moran’s I	Z	p	Moran’s I	Z	p	Moran’s I	Z	p
2000	0.361	6.153	<0.01	0.451	7.641	<0.01	0.295	5.043	<0.01
2010	0.304	5.203	<0.01	0.386	6.6	<0.01	0.215	3.732	<0.01
2020	0.414	7.018	<0.01	0.443	7.517	<0.01	0.319	5.443	<0.01

). Since there were significant spatial correlations between each of the dependent variables and the core independent variables, we used spatial econometric models to analyze the direct impacts and spatial spillover effects of EWP on population longevity levels. Considering gender differences, we conducted regressions separately for the total, male, and female populations.

Moreover, referring to existing studies and considering the human–Earth relationship system, we included several environmental factors as control variables in the model to improve the strength of the statistical test and to better summarize the impact of EWP changes on the longevity of the population. Among the control variables, the impact of the natural geographical environment was measured by annual average temperature (TEM), annual precipitation (PRE), and altitude (ALT), and the impact of the human geo-environment, i.e., economic and social development, was measured by per capita GDP (PGDP) and the population urbanization rate (URB). The description of variables is showed in appendix (

Table A1. Descriptive variables.

Variables	Symbol	Unit	Observations	SD	Mean	Min	Max
Total longevity level	TLI	%	309	1.76	5.4	1.82	10.22
Male longevity level	MLI	%	309	1.73	4.11	0.87	9.47
Female longevity level	FLI	%	309	1.89	6.58	2.43	11.04
Ecological well-being performance	EWP	index	309	0.37	0.28	0.01	1.71
GDP per capita	PGDP	CNY	309	26,556	14,951	120.3	401,886
Urbanization rate	URB	ratio	309	0.26	0.52	0.1	1
Annual average temperature	TEM	°C	309	1.62	16.48	9.25	18.33
Annual precipitation	PRE	mm	309	200.7	1402	909.5	1926
Altitude	ALT	m	309	373.1	287.8	22.38	1676

).

To better align the variables with a normal distribution and address the model’s heteroskedasticity, natural logarithmic transformations were applied to average annual temperature, annual precipitation, average altitude, GDP per capita, and the population urbanization rate prior to conducting the regression analysis. Following the OLS benchmark regression estimation, the variance inflation factor (VIF) values for all the explanatory variables were less than 5, indicating the absence of multicollinearity. A core focus of this study was examining the spatial spillover effects of EWP changes on population longevity, thus requiring spatial econometric models to estimate the spatial spillovers of the independent variables. Although significant results were obtained from the LM tests and robust LM tests for spatial econometric model selection, to ensure robustness, we further constructed the SDM, SLM, and SEM based on different distance-weighted KNN spatial weight matrices and selected the most appropriate model.

4.2.2. Regression Results

The results of the regression models are shown in

Table 5. Regression results.

Var.		OLS	SDM					SLM	SEM
			K = 3	K = 4	K = 5	K = 6	K = 7	K = 6	K = 6
Coef.	EWP	0.16 ** (2.56)	0.12 *** (2.80)	0.11 *** (2.78)	0.11 *** (2.71)	0.10 ** (2.52)	0.10 ** (2.45)	0.11 *** (2.60)	0.11 *** (2.61)
	lnPGDP	0.60 *** (9.33)	0.16 ** (2.51)	0.14 ** (2.27)	0.14 ** (2.28)	0.14 ** (2.13)	0.14 ** (2.10)	0.23 *** (4.22)	0.25 *** (3.42)
	lnURB	0.16 * (1.92)	0.12 ** (1.98)	0.11 * (1.84)	0.09 (1.59)	0.09 (1.51)	0.08 (1.36)	0.10 * (1.68)	0.16 *** (2.74)
	lnTEM	1.64 *** (4.11)	0.75 (1.58)	0.96 ** (2.01)	1.01 ** (2.18)	1.06 ** (2.30)	1.05 ** (2.24)	0.52 * (1.83)	2.02 *** (4.90)
	lnPRE	0.02 (0.47)	0.17 *** (3.28)	0.18 *** (3.45)	0.17 *** (3.28)	0.17 *** (3.43)	0.17 *** (3.44)	0.04 (1.26)	0.19 *** (3.76)
	lnALT	2.17 ** (2.49)	1.91 *** (3.06)	2.07 *** (3.32)	1.96 *** (3.17)	1.97 *** (3.17)	1.96 *** (3.16)	1.52 *** (2.58)	2.71 *** (4.64)
	W × EWP		0.07 (1.07)	0.09 (1.18)	0.09 (1.04)	0.13 (1.42)	0.14 (1.43)
	W × lnPGDP		0.21 *** (2.74)	0.27 *** (3.32)	0.25 *** (2.84)	0.28 *** (2.96)	0.29 *** (2.80)
	W × lnURB		−0.04 (−0.42)	−0.10 (−0.93)	−0.07 (−0.55)	−0.08 (−0.58)	−0.08 (−0.50)
	W × lnTEM		−0.02 (−0.03)	−0.39 (−0.69)	−0.59 (−1.03)	−0.75 (−1.27)	−0.72 (−1.15)
	W × lnPRE		−0.24 *** (−3.67)	−0.25 *** (−3.56)	−0.22 *** (−3.12)	−0.23 *** (−3.13)	−0.24 *** (−3.14)
	W × lnALT		−0.89 (−1.06)	−0.62 (−0.66)	−1.42 (−1.40)	−1.77 (−1.59)	−1.96 (−1.62)
	ρ		0.46 *** (8.32)	0.48 *** (8.10)	0.51 *** (8.11)	0.49 *** (7.41)	0.49 *** (6.99)	0.59 *** (11.23)
	λ								0.71 *** (11.64)
	sigma2		0.12 *** (12.17)	0.12 *** (12.21)	0.12 *** (12.21)	0.12 *** (12.25)	0.12 *** (12.27)	0.12 *** (12.22)	0.13 *** (11.83)
Direct effect	EWP		0.14 *** (2.93)	0.13 *** (2.95)	0.13 *** (2.84)	0.12 *** (2.74)	0.12 *** (2.66)	0.12 ** (2.57)
	lnPGDP		0.20 *** (3.37)	0.18 *** (3.12)	0.18 *** (3.03)	0.17 *** (2.82)	0.17 *** (2.72)	0.25 *** (4.51)
	lnURB		0.13 ** (2.18)	0.11 * (1.88)	0.10 * (1.69)	0.09 (1.58)	0.09 (1.43)	0.11 * (1.88)
	lnTEM		0.79 * (1.82)	0.96 ** (2.17)	1.00 ** (2.31)	1.03 ** (2.40)	1.03 ** (2.35)	0.55 * (1.89)
	lnPRE		0.14 *** (3.06)	0.16 *** (3.29)	0.15 *** (3.17)	0.16 *** (3.37)	0.16 *** (3.38)	0.05 (1.35)
	lnALT		1.95 *** (2.97)	2.16 *** (3.30)	1.95 *** (2.99)	1.92 *** (2.98)	1.91 *** (2.98)	1.68 *** (2.69)
Indirect effect	EWP		0.22 * (1.80)	0.26 * (1.86)	0.27 * (1.70)	0.34 ** (1.96)	0.36 * (1.92)	0.15 ** (2.21)
	lnPGDP		0.48 *** (5.23)	0.60 *** (5.79)	0.60 *** (5.08)	0.65 *** (4.93)	0.65 *** (4.53)	0.31 *** (4.67)
	lnURB		0.04 (0.24)	−0.07 (−0.39)	−0.01 (−0.06)	−0.04 (−0.17)	−0.04 (−0.15)	0.14 * (1.76)
	lnTEM		0.56 (0.81)	0.13 (0.17)	−0.14 (−0.17)	−0.43 (−0.50)	−0.37 (−0.41)	0.68 * (1.90)
	lnPRE		−0.28 *** (−3.2)	−0.30 *** (−3.06)	−0.26 ** (−2.47)	−0.27 ** (−2.47)	−0.29 ** (−2.53)	0.06 (1.27)
	lnALT		0.11 (0.07)	0.79 (0.43)	−0.64 (−0.3)	−1.33 (−0.58)	−1.69 (−0.68)	2.14 ** (2.33)
Total effect	EWP		0.36 ** (2.38)	0.40 ** (2.39)	0.40 ** (2.16)	0.47 ** (2.36)	0.48 ** (2.29)	0.27 ** (2.42)
	lnPGDP		0.68 *** (6.85)	0.79 *** (7.09)	0.78 *** (6.20)	0.82 *** (6.00)	0.82 *** (5.50)	0.56 *** (5.12)
	lnURB		0.17 (0.92)	0.04 (0.18)	0.09 (0.35)	0.05 (0.20)	0.04 (0.14)	0.25 * (1.85)
	lnTEM		1.35 ** (1.98)	1.09 (1.51)	0.86 (1.07)	0.60 (0.72)	0.65 (0.77)	1.24 * (1.94)
	lnPRE		−0.13 (−1.51)	−0.14 (−1.47)	−0.11 (−1.06)	−0.11 (−1.05)	−0.13 (−1.19)	0.11 (1.32)
	lnALT		2.06 (1.04)	2.95 (1.33)	1.30 (0.52)	0.59 (0.22)	0.22 (0.08)	3.82 ** (2.56)
N		309	309	309	309	309	309	309	309
R2		0.78	0.82	0.83	0.83	0.83	0.84	0.80	0.74
AIC		356.87	285.79	280.04	279.66	279.35	279.91	292.53	288.73
BIC		383.01	382.85	377.11	376.73	376.42	376.97	367.20	318.59
log-likelihood		−171.44	−116.89	−114.02	−113.83	−113.68	−113.95	−126.27	−136.36

Note: ***, **, and * indicate significance levels of 1%, 5%, and 10%, respectively; the numbers in parentheses are the t values.

. By comparing the AIC, BIC, and log-likelihood values, the spatial Durbin model constructed using a distance-weighted KNN matrix with K = 6 can be identified as the most suitable model for analysis. With a mean of 5.17 neighboring counties, the value of K = 6 reflects the typical spatial connectivity intensity among Hubei’s counties. In addition to the SDM, this study constructs the SLM and the SEM using the same K = 6-based matrix for comparison. Although the SLM and SEM exhibit better BIC values than the SDM does, the model’s AIC, log-likelihood, and R² values indicate that the SDM remains the most appropriate model. For model comparison, all the dependent and independent variables in every model were standardized via Z-score normalization, and the constants of all the models, which were always zero, were not reported in the results.

The results of the regressions indicate that improvement in EWP has had a significant positive effect on population longevity in Hubei Province. The efficient transformation from resource consumption to human well-being has resulted in less ecological destruction, higher levels of well-being, and a better degree of regional sustainability. As a result, the quality of life of residents has increased significantly, which is reflected in the higher level of longevity among the elderly population. Most control variables also have significant effects, reflecting the underlying influence of the physical and human geographical environment. Additionally, the significant coefficients of the spatial error term further validate that the factors influencing population longevity are intricate, with unaccounted for spatial spillover effects from factors that are not fully captured by the regression analysis.

This study separately constructed the SDM for overall population longevity and gender-specific longevity. All the coefficients of the spatial lag term pass the test at the 1% significance level, which further indicates the simultaneous existence of the spatial lag effect of population longevity. The existence and intensity of spatial spillover effects for the explanatory variables cannot be determined by their spatial lag terms (W × X) alone. Instead, interpretation must be conducted through effect decomposition of the SDM [63]. This study performs effect decomposition using a partial differentiation approach (

Table 6. Decomposition of spatial effects on population longevity by gender.

Var.		Population Longevity
		Overall	Male	Female
Direct effect	EWP	0.12 ***(2.74)	0.12 ***(2.92)	0.11 **(2.2)
	lnPGDP	0.17 ***(2.82)	0.13 **(2.24)	0.21 ***(3.07)
	lnURB	0.09(1.58)	−0.08(−1.36)	0.24 ***(3.52)
	lnTEM	1.03 **(2.4)	1.26 ***(3.12)	0.74(1.5)
	lnPRE	0.16 ***(3.37)	0.12 ***(2.81)	0.18 ***(3.4)
	lnALT	1.92 ***(2.98)	1.67 ***(2.77)	2.05 ***(2.82)
Indirect effect	EWP	0.34 **(1.96)	0.35 **(2.21)	0.32 *(1.78)
	lnPGDP	0.65 ***(4.93)	0.67 ***(5.48)	0.61 ***(4.4)
	lnURB	−0.04(−0.17)	0.08(0.35)	−0.15(−0.58)
	lnTEM	−0.43(−0.5)	−0.27(−0.34)	−0.52(−0.57)
	lnPRE	−0.27 **(−2.47)	−0.19 *(−1.89)	−0.34 ***(−2.93)
	lnALT	−1.33(−0.58)	−1.29(−0.61)	−1.14(−0.49)
Total effect	EWP	0.47 **(2.36)	0.48 ***(2.63)	0.44 **(2.13)
	lnPGDP	0.82 ***(6)	0.79 ***(6.31)	0.82 ***(5.83)
	lnURB	0.05(0.2)	0(0.01)	0.09(0.32)
	lnTEM	0.6(0.72)	0.99(1.29)	0.22(0.26)
	lnPRE	−0.11(−1.05)	−0.07(−0.68)	−0.15(−1.4)
	lnALT	0.59(0.22)	0.38(0.16)	0.91(0.34)
ρ		0.49 ***(7.41)	0.48 ***(7.30)	0.43 ***(6.05)
N		309	309	309
R2		0.83	0.86	0.78
AIC		279.35	240.54	362.81
BIC		376.42	337.6	459.88
log-likelihood		−113.68	−94.27	−155.41

Note: ***, **, and * indicate significance levels of 1%, 5%, and 10%, respectively; numbers in parentheses are t values.

).

The findings confirm that changes in EWP not only exert a significant positive effect on local population longevity levels but also generate significant positive spatial spillover effects on adjacent counties. The direct and spatial spillover effects of improved EWP on longevity are stronger for men than for women.

4.3. Interactive Influence of EWP with Environmental and Socioeconomic Factors

We used geodetectors to explore the interactive effects of EWP with other geo-environmental factors among the entire population and its subgroups during the study period. The interaction detector was used after the optimal data classification method was followed, with the q value maximized. The values on the diagonal in each matrix represent the effects of the factors themselves, whereas the first volumes represent the interaction effects (Figure 5). Generally, there are significant interactions between factors, and the effects are stronger when factors interact with EWP than when they are considered individually.

An examination of the interactions between factors reveals non-linear enhancement among most of the factors. The combined effects of the factors are greater than the sum of their individual effects. Additionally, some factors exhibit bivariable enhancement, with the combined effect surpassing the maximum effect of the two individual factors. Compared with the effects of a single factor, the interactions of EWP with PGDP, URB, and TEM are generally more prominent. Over time, the interaction between EWP, GDP per capita, and the urbanization rate gradually took the absolute lead according to the order of intensity, regardless of whether the entire population or its subgroups were considered. This illustrates that, under sustainability conditions, increased efficiency in converting resource consumption into well-being can regulate and promote the generation of greater health effects from economic and social development. In terms of gender heterogeneity, we found that the average increment of the interaction effects was greater for men (0.214) than for women (0.206), which is similar to the spatial econometric results mentioned above, indicating that male longevity is more sensitive to changes in EWP.

4.4. Influencing Mechanism of EWP on Longevity in Hubei Province

To summarize the effect analysis mentioned above, we present the mechanism of EWP enhancement on the longevity level of the population in Hubei Province (Figure 6). On the one hand, improvements in population longevity cannot occur without the role of the natural geographic environment and socioeconomic development. In the case of Hubei Province, its warm and humid climate conditions and certain altitude and relief conditions form a suitable natural geographic environment. Moreover, economic growth and social progress have led to improved health care, health technology, and social security. Moreover, improved EWP plays an important facilitating role in the human–land relationship system. This means more human well-being gains in exchange for fewer eco-environmental costs, representing a stronger capacity for regional sustainable development. As the population longevity level is driven by the natural geographic environment and socioeconomic development, there are interactions between EWP and elements of the natural ecosystem and the human system. A more harmonious human–land relationship amplifies the effects of environmental influences. The combination of these influences leads to a synergistic increase in population longevity across the region through direct enhancement and spatial spillover.

5. Discussion

China, the world’s largest developing country, is currently in a critical period of economic transformation and development. Hubei Province, known for the phenomenon of population longevity, is an important province in terms of the Rise of Central China strategy and the development of the Yangtze River Economic Belt. Over the past two decades, economic development and transformation in Hubei Province have brought about a series of chain reactions. The variation in the relationship between resource consumption and social welfare output is reflected in the spatial distribution and evolution of the EWP. In contrast to existing studies, which have focused mostly on provincial- or prefecture-level cities [14,15], we analyzed the EWP and population longevity at a finer county scale via geographic information data [47,48], which enabled the representation of a more detailed spatial and temporal pattern.

The evolution of the spatial distribution of EWP in Hubei Province generally corresponds to the process of economic development in the province. The mean EWP showed continuous growth over the study period, especially in the last decade, demonstrating that the overall sustainable development capacity of Hubei Province increased as its ecology was optimized and that people’s livelihoods improved due to the construction of an ecological civilization and economic transformation. These findings match those of existing studies. Xuan’en, Hefeng, and other counties in southwestern Hubei have long been economically underdeveloped regions within Hubei Province [45]. Compared with the eastern and central parts of Hubei Province, the southwestern part of Hubei Province is relatively weak in terms of human activities and regional development intensity because of the mountainous barrier. The relatively high values of EWP occurring in these regions may imply low resource consumption and low well-being outputs, as mentioned in the study by Zhu et al. [12].

According to the environmental Kuznets curve (EKC) hypothesis, the degree of pollution and damage to the regional environment first increases and then decreases with economic growth. Thus, during periods of extensive economic development that are structurally dominated by high-energy-consumption industries, EWP tends to remain at a low level or decline, and it increases after the economic development method is transformed. Dietz et al. [28] argued that the relationship between EWP and economic development does not follow the EKC. In our study, however, the EWP decreased and then increased in some of the counties in the southwestern Hubei mountainous area, such as Changyang and Dianjun, indicating that the EKC is reasonably well supported. Changyang and Dianjun are counties within Yichang city, which achieved rapid economic growth with the development of the heavy chemical industry in approximately 2010 owing to its mineral resources. However, Enshi Prefecture was undergoing industrial restructuring to promote the development of secondary industry during the same period. As a result, the eco-environmental costs accompanying these processes caused the EWP within the southwestern Hubei mountainous area to decrease rather than increase. With the construction of the West Hubei Ecological and Cultural Tourism Circle, the protection of the Yangtze River, and the promotion of the green and high-quality development of industries in coastal areas, the quality of human settlements [64] and peoples’ livelihoods and well-being improved by 2020, while high-quality development was pursued economically.

The spatial distribution pattern reflects the imbalance of EWP enhancement among regions in Hubei Province. The advanced and high-level regions have not yet exerted a radiating driving force on surrounding areas, which is reflected in the “core–edge” structure of the spatial distribution of EWP and the strengthening of spatial agglomeration. Related studies also present this spatial distribution of EWP [27], but our study presents it at the more detailed county level, clarifying the differentiation between municipal districts and suburban counties within prefecture-level cities.

This study reveals a set of mechanisms for promoting population longevity through EWP enhancement in Hubei Province. Wang et al. [65] qualitatively concluded that longevity is affected by the harmonious development of regional ecology, regional society, and the regional economy. We used spatial econometric models and geodetectors to empirically investigate how harmonious regional human–land relationships can contribute to population longevity. We focused on the synergistic relationships among ecology, economy, and society, considering the effect of the regional sustainable development level on population longevity. Harmonious human–land relationships act as “amplifiers”, meaning that higher EWP interactions with other environmental factors can reinforce the effects of those factors on population longevity. As mentioned by Han et al. [22], the low-carbon city pilot policy in China has local neighbor impacts on EWP. The effect of EWP enhancement can also radiate to neighboring areas. Existing studies using spatial econometric models of analysis have found that population lifespan may be affected by neighboring regions, and these effects may exceed the influence of local factors, indicating the presence of spatial dependence [38,66]. The results of our study also support this view. The harmonious human–Earth relationship in a certain county not only promotes the longevity of the population within the county but also has spatial spillover effects on neighboring counties.

We also revealed gender differences in the impact of EWP elevation on population longevity. During the study period, economic growth in most areas of Hubei Province was driven by secondary industry, and the working environments of high-pollution and high-energy-consumption industries tended to negatively impact the health of a large group of male workers. Along with EWP improvement, green development and the elimination of outdated industry led to better working environments. Therefore, male longevity is more sensitive to changes in EWP. As for the results of the other control variables in the study, the contributions of Hubei Province’s warm, humid, and moderate climate, altitude, and level of economic development to the increase in longevity are similar to those estimated in previous studies [40,41]. We suggest that the negative effect of urbanization rates on male longevity levels may be further linked to gender heterogeneity in lifestyle. In Hubei Province, even throughout China, the current social division of labor remains such that men are more engaged in occupations, whereas women focus on household chores [67,68]. Men face more occupational pressure and are less attentive to personal care than women are. Men are also prone to smoking and drinking addictions, whereas women place greater emphasis on personal care and quality of life. With increasing urbanization, the likelihood of males experiencing work pressure and adopting bad behavioral habits increases to some extent. Moreover, women may enhance their health via medical care and other means, aided by modern technology and the benefits of urban living.

In response to the above findings, we propose the following policy recommendations. Firstly, governments should pay attention to the counties with lower levels of EWP and slower rates of enhancement, summarize the bottlenecks of EWP enhancement in light of the actual situation of these counties, and improve the well-being of society in a targeted manner. Then, spatial interactions need to be further emphasized in the synergistic enhancement of EWP and population longevity. High-level areas need to further enhance their ability to radiate positive effects and drive positive changes in surrounding areas, fostering sustainable intercounty relationships. Last but not least, the gender heterogeneity in the effects of EWP on population longevity levels requires specific attention, and targeted actions should be taken.

6. Conclusions

This study aimed to assess ecological well-being performance (EWP) from the perspective of county-scale input–output dynamics. It explores the spatiotemporal patterns of EWP evolution and analyzes how changes in EWP influence population longevity, with a particular focus on the roles of spatial spillover effects and interactive mechanisms. The findings effectively respond to these objectives as follows.

First, the persistent uptrend in overall EWP in Hubei Province not only indicates an increased capacity for sustainable development within the province but also has wider applicability. The identified spatial patterns of EWP reveal polarization in the east, connection into a surface in the west, and a collapse in the middle, along with a “core–edge” structure.

Second, EWP has a very significant direct contribution and spatial spillover effect on male, female, and overall longevity, highlighting the crucial role of EWP in population health. This finding calls for the environmental science, public health, and social policy fields to focus on EWP in promoting population longevity without ignoring spatial transmission between regions.

Finally, we summarize the mechanisms revealed after effects analysis by spatial econometrics and geographical detectors. A higher EWP represents an enhanced capacity for regional sustainable development, which can reinforce the positive impact of natural and socioeconomic environments on population longevity.

Based on this research’s findings, we put forward relevant policy recommendations, such as emphasizing the spatial interactions in the synergistic enhancement of EWP and population longevity. This study also has several limitations due to the limited availability of county-level data. On the one hand, the comprehensiveness of the EWP evaluation index system is limited to some extent. In addition, we could only consider the traditional and major factors of the physical geographical environment and the level of economic and social development affecting population longevity in addition to EWP. Future research could focus on collecting more detailed and extensive data to improve the accuracy of measuring EWP and better account for all the relevant factors influencing population longevity.