How Do Sense of Place and Perceived Restorativeness Affect Psychological Benefits from Urban Green Spaces for Older Adults? A Cross-Sectional Study

Abstract

With the intensifying trend of population aging, the positive effects of Urban Green Space (UGS) on Psychological Well-being (PW) among older adults have garnered increasing attention. Previous studies examined the unidirectional pathways through which objective UGS exposure indicators influenced PW via Sense of Place (SOP) or Perceived Restorativeness (PR). However, little empirical work has addressed how UGS exposure affects PW in older adults through a dual mediation pathway encompassing both SOP and PR. To address this gap, this study investigated representative urban parks in Fuzhou, China. Using the Normalized Difference Vegetation Index (NDVI) and Normalized Difference Water Index (NDWI) as objective exposure indicators, along with survey data capturing older adults’ perceived characteristics (PC) of UGS, SOP, and PR, we developed a multilevel structural equation model to explore how NDVI, NDWI, and PC influence PW through the dual mediators of SOP and PR. Results indicated that (1) NDVI and PC influence PW either directly or indirectly via the mediators SOP and PR, with PC exhibiting a significantly stronger effect than NDVI; (2) NDWI has no significant effect on SOP and PW, but it indirectly influences PW through PR; (3) PR moderated the link between SOP and PW; (4) under the dual-pathway mechanism, PR contributes more substantially to PW than SOP. This study enriches the understanding of how UGS contributes to PW and advocates for integrating exposure characteristics, place qualities, and restorative elements into the framework of age-friendly city planning to guide targeted health interventions for the elderly. Moreover, SOP–PR insights from an East Asian city inform global aging urbanism.

1. Introduction

With accelerating urbanization and rapid population aging, psychological well-being (PW) issues among older adults (such as depression and anxiety) are becoming increasingly prominent worldwide [1]. Urban Green Space (UGS), which integrates natural contact, emotional restoration, and opportunities for social interaction, is widely regarded as a vital natural resource for enhancing public well-being and alleviating PW problems [2]. Various international initiatives (e.g., the World Health Organization’s “Healthy Cities” campaign) and national strategies (e.g., China’s “Age-Friendly Cities” policy) emphasize the importance of high-quality UGS development to promote emotional restoration and social connection, thereby enhancing older adults’ subjective well-being and PW, and improving urban livability and inclusiveness [3].

How UGS influences PW has become a focal point in healthy city research. In previous studies, UGS exposure was often quantified by objective physical indicators such as the Normalized Difference Vegetation Index (NDVI), which measured the amount of greenness within a spatial area [4]. The exposure of blue spaces (BS) is commonly measured using the Normalized Difference Water Index (NDWI), which serves as a key indicator for quantifying the characteristics and information of water bodies [5]. However, an increasing number of studies emphasize that “green space exposure” encompasses not only the physical presence of greenery but also individuals’ subjective experiences of contact and perception with the green environment [6]. Due to limitations in physical condition and mobility, older adults tend to rely more on their psychological perception of nearby accessible green spaces than on spatial indicators alone [7]. Relying solely on objective measures may lead to underestimation or overestimation of the effects of UGS on their PW [8]. Integrating objective indicators with subjective perceptions has become a crucial means of examining the relationship between UGS exposure and PW in older adults. Scholars have pointed out that subjective perception is a psychological process in which individuals, through their sensory systems, receive external stimuli and actively integrate their personal experiences, such as prior memories, emotional tendencies, and cognitive preferences, to process and interpret these signals, ultimately forming personalized cognition and experiences of objects or environments [9,10,11,12]. The Perceived Sensory Dimensions (PSD) scale is primarily used to assess people’s subjective perceptions of UGS quality [12]. It encompasses perceptual dimensions such as facilities, culture, social interaction, biodiversity, safety, and accessibility, which collectively capture older adults’ subjective evaluations of UGS quality.

Current studies have explored the mechanisms through which UGS affects health from two primary directions: one focuses on Perceived Restorativeness (PR). PR refers to an individual’s subjective perception of the psychological and attentional restorative potential of a given environment. It is commonly measured using the Perceived Restorativeness Scale (PRS), which assesses the restorative characteristics of environments, as explained by Kaplan’s Attention Restoration Theory (ART) and Ulrich’s Stress Reduction Theory (SRT). ART posited that natural environments help restore directed attention that was prone to fatigue, thereby enhancing cognitive functioning [13]; SRT suggested that individuals’ proximity to green spaces or natural environments induced positive psychophysiological responses, which can rapidly alleviate feelings of tension, anxiety, and anger, promote a sense of calmness and relaxation, and thereby contribute to psychological well-being [14]. The other emphasizes the positive role of Sense of Place (SOP) in promoting emotional well-being and life satisfaction. SOP primarily referred to the relationship between people and specific locations. The enhancement of SOP (or place attachment) can facilitate psychological restoration and strengthen the capacity for perceived well-being [15]. Although both mechanisms have received extensive empirical support [15,16], studies that integrate the dual mediating roles of SOP and PR in influencing PW remain relatively scarce, and the existing literature predominantly focuses on younger or general populations, with limited attention to the elderly. In addition, issues such as inconsistent green space quality and significant disparities in accessibility for the elderly in second-tier Chinese cities exacerbate the vulnerability of older adults’ PW and reinforce social inequality [17].

The current studies have several limitations. Firstly, due to age-related declines in physical functioning, sensory perception, and mobility, relying solely on objective exposure indicators fails to accurately capture older adults’ actual environmental experiences. Therefore, it is necessary to incorporate subjective perception measures to more comprehensively assess their psychological health outcomes. Secondly, existing studies tended to conceptualize SOP and PR as parallel or independent mediators of UGS effects, without systematically examining their directional relationship or interactive mechanisms, and Structural Equation Modeling (SEM) provides an optimal solution and is one of the most commonly used tools for analyzing complex relationships and pathways among multiple variables. These approaches often overlook individuals’ subjective perceptions of the environment and their deep emotional connections with SOP, which may serve as a prerequisite for perceived restoration [16]. This is particularly true for older adults who have developed a stronger sense of belonging through long-term exposure to UGS. Thirdly, numerous studies have pointed out that Fuzhou City in Fujian Province is characterized by high-density land use and an aging population at greater health risk [18,19,20]. The high-density urban environment further exacerbates the health burdens of vulnerable groups such as older adults, leading to problems such as insufficient physical activity, metabolic risks, obesity, cardiovascular diseases, and coronary heart disease [21,22]. Researching UGS in Fuzhou can therefore help reduce health disparities, improve public health, and contribute to the development of a resilient and inclusive city, while also filling the existing research gap that focuses on older populations. Therefore, exploring the unique roles of SOP and PR in linking objective and subjective environmental perceptions with older adults’ PW through a multilevel SEM is of critical importance.

This study conducted a field survey among older adults in four urban parks in Fuzhou City. We established a theoretical framework linking “dependent on or related to UGS exposure characteristics (objective and subjective)—psychological mediators (SOP + PR)—health outcomes,” aiming to investigate the following key questions: (1) Do UGS quantity indicators (NDVI and NDWI) and subjective Perceived Characteristics (PC) exert differential impacts through the distinct mediating pathways of SOP and PR? (2) Is there a chained mediating relationship between SOP and PR? Our study not only expands and refines existing theoretical models but also clarifies how PW can be enhanced in aging and densely populated urban environments.

2. Theoretical Review and Research Hypotheses

2.1. Association Between UGS Exposure and SOP Among Older Adults

2.1.1. The Impact of UGS Exposure Characteristics on Older Adults’ Health

In recent years, increasing attention has been paid to the psychological benefits of green space exposure for vulnerable groups such as older adults. The literature typically emphasized several dimensions of green space utility, including usefulness (area-based spatial exposure measures such as green coverage ratio or NDVI), visibility (green view index, i.e., the proportion of greenery visible within the streetscape), accessibility (distance-based spatial exposure), and attractiveness (quality-based evaluations) [23]. Studies showed that higher NDVI values were significantly associated with lower risks of mortality, psychological stress, anxiety, and depression [24,25,26]. Increasing vegetation coverage effectively regulates ambient temperature and improves air quality [27], thereby enhancing overall health outcomes [28,29]. A higher NDWI value indicates more pronounced characteristics of open water bodies [30], which are associated with restorative environmental features and better psychological restoration [31,32]. This study investigated four urban parks that exhibited different levels of NDVI and NDWI, meaning that the attributes or characteristics of the parks (NDVI and NDWI) were treated as influential variables, with the assumption that the NDVI and NDWI values were determined by each specific park under investigation.

As an essential component of natural ecosystems, blue spaces (BS) within parks, various forms of surface water adjacent to green areas, have often been regarded as integral parts of green environments. These spaces similarly possessed substantial psychological and physiological restorative potential, improving older adults’ mental health through pathways such as stress reduction, promotion of physical activity, and facilitation of social interaction [33]. However, for older adults, BS provided distinctive sensory experiences and emotional healing values that differed from those of green spaces, yet these benefits had long been underestimated [33]. The Normalized Difference Water Index (NDWI) was one of the most widely used objective indicators for examining the relationships between blue spaces, mental health, and well-being [34]. Previous studies demonstrated associations between NDWI and depression as well as overall mental health; however, research focusing specifically on older adult populations remained insufficient.

The Health Exposure Theory and the Health Affordances Theory both emphasize that the impact of natural environments on health does not arise directly from objective exposure but is realized through individuals’ subjective perception and interactive experiences [35,36]. The former noted that contact with nature could promote health by alleviating stress, encouraging physical activity and social cohesion, and enhancing psychological restoration. Meanwhile, the latter focused on the interactive potential between the individual and the environment, specifically, the perceived supportive opportunities the environment provides for health-related behaviors. Previous research further indicated that the health benefits of an environment depend on its ability to meet the usage needs of specific populations [37]. For example, uneven gravel paths or dense vegetation may offer explorative and attractive qualities for children, yet pose accessibility barriers for older adults with mobility limitations. For older adults, functional features such as flat walkways, continuous pathways, sufficient resting points, and socially supportive settings were key prerequisites for their perception, engagement, and subsequent psychological benefits. Previous studies have demonstrated that positive perceptions of green space exerted a promotive effect on multiple dimensions of health in older adults [38]. This suggests that the health value of green spaces depended not only on the quantity of exposure but more crucially on older adults’ subjective perception and evaluation of environmental quality. Accordingly, we propose the following hypotheses:

H1a. 

UGS (NDWI, NDVI, PC) positively affects PW.

2.1.2. The Role of Green Space Exposure Characteristics in Shaping Older Adults’ SOP

SOP refers to the relationship between individuals and specific locations or environments, encompassing perceived value, meaning, and symbolic significance [39]. According to [40], SOP can be conceptualized as a three-dimensional construct comprising place attachment, place identity, and place dependence. Place attachment denotes a strong emotional bond between individuals or groups and specific places, often characterized by irreplaceability and emotional resonance [41,42,43]. Place identity reflects how people integrate particular places into their self-concept, emphasizing symbolic and psychological identification with those places [44]. Place dependence refers to the extent to which individuals functionally rely on a location to fulfill specific needs [45]. Among older adults, SOP is often expressed as a profound “residential rootedness.” Following retirement, older individuals tend to engage in prolonged reflection on personal place-related experiences, memories, and community participation processes through which SOP typically reaches its peak [46]. This rootedness deepens familiarity and reliance on the surrounding environment, strengthening emotional bonds and fostering a heightened SOP belonging. However, the mechanisms through which UGS exposure contributes to SOP in older adults remain poorly understood [47].

Higher NDVI levels enhanced older adults’ ability to perceive greenness [48] and encouraged opportunities for nature exposure. For older adults, positive nature experiences fostered social cohesion, environmental familiarity, and dependency, which in turn stimulated positive emotional responses, environmental identity, and belonging through recollection and cognition, thereby reinforcing their sense of place [49]. Exposure to tree-shaded green spaces facilitated psychological satisfaction among older adults, which positively influenced place attachment [50,51].

Interactions with BS not only provided aesthetic pleasure and emotional regulation but also played a vital role in forming a sense of place attachment and belonging among older adults [34]. This process was primarily sustained by fostering self-awareness, which reinforced individuals’ sense of attachment and restorative feelings toward BS and in turn promoted psychological restoration and well-being [52]. In addition, BS contributed to the shaping of our place identity within the natural environment [53].

PC represented older adults’ subjective cognition and evaluation of UGS and demonstrated potential in shaping their SOP. Grounded in the Theory of Place Qualities and Social Support Theory [54,55], high-quality green spaces not only fulfilled older adults’ daily needs in physical aesthetics, functional requirements, facilities, culture, safety, species, and accessibility but also fostered social connections and emotional exchange [56,57,58,59,60]. Such open, safe, and inclusive interactive spaces enabled older adults to acquire emotional support, instrumental support, informational support, and appraisal support through social interactions. This social support strengthened individuals’ self-worth and emotional identification, thereby deepening their place attachment and place identity [55,61]. Concurrently, park environments that incorporated familiar landscape features, cultural symbols, and memory cues facilitated a sense of belonging among older adults by guiding the recollection and sharing of meaningful past experiences, thus promoting the formation of SOP [62,63,64]. Accordingly, the following hypotheses are proposed:

H2a. 

UGS (NDWI, NDVI, PC) has a positive effect on SOP.

2.2. Association Between Green Space Exposure Characteristics and PR

ART and SRT provide theoretical frameworks to explain how natural environments facilitate psychological restoration. According to SRT, humans possess an innate biological predisposition to engage with natural environments, which helps reduce stress and anxiety, regulate mood, and enhance PW [14,65]. Natural environments reduce physiological arousal by inhibiting sympathetic nervous activity and enhancing parasympathetic responses, thus lowering stress-related indicators such as heart rate and blood pressure and promoting psychological recovery [66]. Compared to urban built environments, natural settings are more likely to elicit positive affective responses and yield superior restorative benefits [14]. For example, BS in parks promoted emotional restoration and PW through their distinctive visual and auditory characteristics (such as water bodies and the sound of flowing water), while their open vistas and tranquil atmosphere further enhanced the psychological restoration effects [31]. ART emphasizes that exposure to nature facilitates the recovery of directed attention, consequently improving cognitive functioning [67]. Mental fatigue and attentional depletion, frequently triggered by both internal and external distractions, can be effectively mitigated through exposure to green spaces [13,68]. Currently, PR is most commonly measured using the PRS, which comprises four dimensions: being away, extent, fascination, and compatibility [68]. The specific connotations of these dimensions are as follows: (1) being away: psychological or physical distancing from daily stressors to facilitate attentional recovery; (2) extent: a coherent and immersive environment that supports exploratory engagement; (3) fascination: effortless attention elicited by inherently interesting stimuli in nature; and (4) compatibility: the degree to which the environment supports one’s personal goals and activities [68]. Accordingly, the following hypotheses are proposed:

H3a. 

UGS (NDWI, NDVI, PC) has a positive effect on PR.

2.3. The Linkages Between SOP, PR, and PW

SOP reflects the long-term psychological attachment individuals form through their subjective perception of the environment and plays a critical role in sustaining PW among older adults. Prior research suggests that human-place bonds are often forged through psychological, rather than purely physical, interactions, with specific spatial attributes eliciting strong emotional responses [69,70]. Neuroscientific findings indicate that emotional responses to place are not purely subjective phenomena but are underpinned by specific brain structures, particularly the hippocampus and amygdala, which encode spatial memory and emotional value, thereby enhancing SOP’s role in shaping individual well-being [71]. An empirical study conducted in Hong Kong confirms a significant positive correlation between SOP and subjective well-being [72].

PR is defined as the perceived restorative potential that individuals experience in restorative environments, associated with attention recovery, stress reduction, and restorative experiences [13]. Many studies have linked the key dimensions of PR (being away, fascination, extent, and compatibility) to enhanced emotional states and increased well-being. Landscape components and sensory environments (visual and auditory) within UGS exert both direct, indirect, and moderating influences on PR, emotional states, and individual preferences [73]. PR can preferentially capture people’s aesthetic preferences and perceived dimensions in the environment, benefiting enjoyment and well-being [48]. A recent survey across four typologies of urban parks revealed a strong positive association between PR and subjective vitality, pleasure, excitement, and happiness [57].

In addition, some studies have examined the interconnection between SOP and PR, suggesting that place attachment significantly enhances individuals’ perceived restorative experiences [15,74]. Interacting with familiar places often evokes nostalgic memories and a sense of recognition, which strengthens emotional bonding and enhances PR [15]. Although previous studies acknowledged that SOP had restorative potential and that SOP and PR could serve as mediators in the relationship between green space and health, few studies further analyzed how subjective and objective exposure characteristics of UGS differed and interacted via the SOP and PR pathways. Specifically, Liu et al. used synthesized photographs of urban parks and conducted questionnaire surveys to explore the unidirectional relationships among local landscape characteristics, place attachment, and PR [15]. Van Dinter et al. carried out an online survey to examine the direct relationships between individual and park characteristics, SOP, and subjective well-being, without including the PR variable [16]. In contrast, the present study conducted a field survey in four urban parks in Fuzhou City, where each respondent was physically present and actively engaged with the park environment. This approach enabled a more authentic and accurate assessment of older adults’ PC, SOP, and PR. Furthermore, by integrating objective environmental exposure and subjective perceptions into a multilevel SEM framework, while simultaneously incorporating both SOP and PR, our model provided a more comprehensive and empirically grounded approach, substantially improving upon previous studies that relied on synthetic image evaluations or online survey data. Therefore, the following hypotheses were proposed, and the conceptual framework is illustrated in Figure 1.

H1b. 

SOP has a positive effect on PR.

H2b. 

SOP has a positive effect on PW.

H3. 

PR has a positive effect on PW.

H4. 

SOP and PR play a chained mediating role between UGS (NDWI, NDVI, PC) and PW.

3. Methods

3.1. Study Area and Urban Parks

Fuzhou, situated along the southeastern coast of China (25°15′–26°39′ N, 118°08′–120°31′ E), features a terrain that descends from mountainous west to low-lying coastal east. Over 70% of the city consists of hilly and mountainous landscapes. The region has a subtropical monsoon climate, with a permanent population of approximately 8.47 million and a high degree of urbanization. The city boasts an integrated ecological network of mountains, rivers, forests, grasslands, urban areas, farmlands, and coastal systems, with rich biodiversity. Its urban green coverage rate exceeds 43.1%.

Park selection was guided by the following criteria: (1) provision of diverse cultural ecosystem services; (2) variation in park size; (3) high visitation frequency among older adults; and (4) geographic distribution across different areas of Fuzhou. Based on preliminary field investigations, four parks were selected as study sites: Xihu park, Jinan park, Qinting park, and Fuzhou national forest park. Xihu park, situated in the densely populated and aging core of Gulou District, is a representative example of traditional classical garden design in Fuzhou. Jinan park, located in the central Jin’an District adjacent to Gulou, integrates Niugangshan Park, Helin Ecological Park, and Jinan lake park. As a newly developed ecological park, it features comprehensive facilities and efficient management. Qinting park and Fuzhou national forest park are located in the urban-rural fringe of Jin’an District. While Qinting Lake Park primarily serves surrounding residential communities, Fuzhou National Forest Park attracts a wider demographic (Figure 2 and Figure 3).

3.2. Questionnaire Survey

Drawing on prior studies [46], we targeted elderly residents aged 60 and above who had lived in Fuzhou for no less than 15 years for participation in the questionnaire survey. It is important to highlight that all respondents were assessed on-site and had firsthand experience with all four parks under study, which was crucial for the validity of our findings. The questionnaire comprised five sections: PC, PR, SOP, PW, and demographic information (see Appendix A,

Table A1. Perceived characteristics scale (Likert scale ranging from 1 to 7 agree).

Facilities
Convenient entrances and exits
Asphalt, concrete, or gravel walkways
Accessible facilities
Seats and benches
Trash bins
Directional and informational signage
Picnic tables and chairs
Drinking fountains
Restrooms
Rain shelters or shaded areas
Snack kiosks
Convenient parking lots
Fitness equipment
Culture
Sculptures
Exotic plants
Scenic lakes, creeks, ponds, and other water features
Public art installations
Historical buildings or monuments of famous individuals
Social
Usable children’s playgrounds
Usable open spaces (e.g., for dancing, practicing Tai Chi, singing, or playing chess)
Social interactions with others
Fitness or sports activities
Walking
Science education programs
Enjoying the scenery
Relaxing and resting
Species
One can detect several animals, like birds, insects, etc.
The urban park or urban open space consists of natural plant and animal populations
There are many native plants to study
Safety
Good lighting
Ground visibility
CCTV surveillance
Calm, quiet, and peaceful atmosphere
Seeing many people actively using the park
Parks are clean
Free from traffic noise
Feeling safe within the park
No perception of air pollutants
Free from toxic plants
Effective mosquito and pest control
Accessibility
Do you find the distance from your residence to the park convenient?
How would you assess the convenience of transportation around the park?
Do you perceive significant traffic congestion around the park?

,

Table A2. Sense of place Scale (Likert scale ranging from 1 disagree to 7 agree).

Place Attachment
This park is one of my favorite urban parks.
I really miss this park when I am away for long.
Place identity
This park reflects the type of person I am.
This park makes me feel like I can be myself.
This park says very much about me.
Place dependency
This park meets my needs better than any other park.
This park is a good place to do the things I enjoy most.
There are better parks than this park.

,

Table A3. Perceived Restorativeness Scale (Likert scale ranging from 1 disagree to 7 agree).

Fascination
Places like that are fascinating.
In places like this, my attention is drawn to many interesting things.
In places like this, it is hard to be bored.
Being away
Places like that are a refuge from nuisances.
To get away from things that usually demand my attention, I like to go to places like this.
To stop thinking about the things that I must get done, I like to go to places like this.
Coherence
There is a clear order in the physical arrangement of places like this.
In places like this, it is easy to see how things are organized.
In places like this, everything seems to have its proper place.
Scope
That place is large enough to allow exploration in many directions.
In places like this, there are few boundaries to limit my possibility for moving about.

,

Table A4. WHO-5 Scale (Likert scale ranging from 1 disagree to 7 agree).

WHO-5
I have felt cheerful in good spirits
I have felt calm and relaxed
I have felt active and vigorous
I woke up feeling fresh and rested
My daily life has been filled with things that interest me

and

Table A5. Socio-demographic Characteristics Scale.

Gender	Male	Female
Age	60–65	66–70	71–75	76–80	81 and above
Educational Attainment	Elementary school and below	Junior high school	Senior high school	Senior high school and undergraduate	Graduate student and above
Income	Within 3000 RMB	3000~5000	5000~8000	8000~10,000	10,000 RMB and above
Name of the park currently visited	Xihu park	Jinan park	Qinting park	Fuzhou national forest park

: Questionnaires). All questionnaire items were measured using a 7-point Likert scale, ranging from 1 “strongly disagree” to 7 “strongly agree” [75].

(1)

The PC scale was adapted from the Perceived Sensory Dimensions (PSD) framework [12], characteristics of age-friendly parks [48], and park quality guidelines [76]. It covers six perceptual dimensions with a total of 65 items: facilities (13), cultural features (5), social aspects (8), biodiversity (3), safety (11), and perceived accessibility (3).

(2)

The mediating variables included SOP and PR. SOP comprised eight items across three dimensions: place attachment, place identity, and place dependency [16,40]. PR was assessed using the PRS-11 scale, which includes four components: fascination, being away, coherence, and scope [77].

(3)

PW was measured using the World Health Organization’s five well-being index (WHO-5), a widely validated self-reported scale for assessing overall PW [23]. The scale includes five items, such as “I have felt cheerful and in good spirits” and “I have felt calm and relaxed.”

(4)

Four demographic variables were included: gender (binary), age (coded from 1 to 5 from youngest to oldest), education level (coded 1–5 from lowest to highest attainment), and monthly income (coded 1–5 from below 3000 to above 10,000 RMB).

A pilot study involving 100 samples was conducted before the formal survey. Reliability and validity testing, along with exploratory factor analysis, confirmed the questionnaire’s sound design and strong internal consistency. The formal survey was conducted in June and July 2024 using balanced sampling methods during morning hours (avoiding extreme weather) to ensure data accuracy. During the survey, staff introduced the study and obtained informed consent from the elderly participants before they completed the questionnaire.

3.3. Demographic Characteristics

Table 1. Descriptive Statistics of Demographic Variables.

Demographic Variables	Number/Percentage
Gender
Male	403 (48.03%)
Female	436 (51.97%)
Age
60–65	315 (37.54%)
66–70	214 (25.51%)
70–75	152 (18.12%)
76–80	105 (12.51%)
80–	53 (6.32%)
Educational qualifications
Primary school or below	393 (46.84%)
Secondary school	242 (28.84%)
High School	101 (12.04%)
College or above	62 (7.39%)
Postgraduate and Above	41 (4.89%)
Gross monthly income
3000 or below	368 (43.86%)
3000–5000	242 (28.84%)
5000–8000	107 (12.75%)
8000–10,000	62 (07.39%)
10,000 or higher	43 (5.13%)

Note: N = 839.

presents the descriptive statistics for the participants’ demographic characteristics. A total of 856 questionnaires were collected during the formal survey, of which 839 were deemed valid, resulting in a high response rate of 97.98%. In general, the gender distribution was relatively balanced (48.03% male, 51.97% female), and the largest age group was 60–65 years (37.54%). According to the Seventh National Population Census Bulletin of Fuzhou City (data up to 2020), the gender and age distribution of the participants in this study was consistent with the overall population structure of Fuzhou. The majority of participants had a high school education or below (87.72%), and 43.86% of respondents reported a monthly income of less than 3000 yuan.

3.4. Acquisition of Objective UGS Exposure Indicators

Objective UGS exposure indicators (NDVI and NDWI) were calculated using ArcGIS Pro 3.1.6 and ENVI 5.3 software. Given the potential limitations of satellite imagery, such as cloud cover and seasonal variability, we carefully examined and selected imagery that best represented the peak vegetation conditions under clear and cloud-free weather. Consequently, a high-resolution Sentinel-2A image acquired on 6 August 2023, was chosen for analysis to minimize these influences. The imagery was obtained from the European Space Agency (https://dataspace.copernicus.eu/ (accessed on 29 July 2024) and had a spatial resolution of 10 m. According to [36], negative NDVI values may be considered when analyzing the combined effects of green and blue spaces; Given that the elderly respondents completed the questionnaires during on-site visits (rather than through online recall-based surveys) and considering their limited mobility range, the use of traditional buffer scales might not accurately reflect their true exposure levels. To avoid restoration effect bias resulting from this discrepancy, this study measured UGS exposure based on the actual park boundaries within which individual activities occurred. The reflectance values of the near-infrared and red bands are represented by NIR and R, respectively (Equation (1)). The green band was represented by “Green,” and the NDWI was calculated as shown in Equation (2). Figure 4 and Figure 5 showed the results.

$$\mathrm{N}\mathrm{D}\mathrm{V}\mathrm{I}={\displaystyle \frac{\mathrm{N}\mathrm{I}\mathrm{R}-\mathrm{R}}{\mathrm{N}\mathrm{I}\mathrm{R}+\mathrm{R}}}$$

(1)

$$\mathrm{N}\mathrm{D}\mathrm{W}\mathrm{I}={\displaystyle \frac{\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}-\mathrm{N}\mathrm{I}\mathrm{R}}{\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}+\mathrm{N}\mathrm{I}\mathrm{R}}}$$

(2)

4. Results

4.1. Statistical Results of Variables

Table 2. Statistical results of NDVI, NDWI, PC, SOP, PR, and PW variables.

Variables	Mean (Standard Deviation)
Independent variable
NDVI
Fuzhou National Forest Park	0.461 (0.189)
Qinting Park	0.172 (0.201)
Xihu Park	0.235 (0.245)
Jinan Park	0.261 (0.231)
NDWI
Fuzhou National Forest Park	−0.409 (0.194)
Qinting Park	−0.144 (0.209)
Xihu Park	−0.168 (0.257)
Jinan Park	−0.208 (0.236)
PC	4.795 (1.094)
Facilities (FAC)	4.840 (1.375)
Culture (CU)	4.822 (1.478)
Social (SO)	4.794 (1.449)
Species (SP)	4.768 (1.511)
Safety (SE)	4.745 (1.425)
Accessibility (AC)	4.767 (1.560)
Mediators
SOP	4.891 (1.154)
Place Attachment (PA)	5.002 (1.518)
Place Identity (PI)	4.880 (1.484)
Place Dependence (PD)	4.827 (1.465)
PR	4.810 (1.152)
Fascination (FAS)	4.753 (1.537)
Being Away (BA)	4.799 (1.554)
Coherence (CO)	4.871 (1.502)
Scope (SC)	4.819 (1.566)
Dependent variables
PW	4.747 (1.474)

presents the means and standard deviations of the UGS exposure indicators (NDVI and NDWI) and the latent variables measured by the questionnaire across the four parks. Regarding objective exposure to green space, the Fuzhou national forest park exhibited markedly higher NDVI (Mean = 0.461) values than the other three urban parks, reflecting its dense vegetation cover and strong ecological function, However, the NDWI is lower. In contrast, Qinting park reported the lowest NDVI (0.172) values, likely because the lake occupies a large proportion of the park area (NDWI Mean = −0.144), leaving less space for green vegetation. Notably, Xihu Park had the highest standard deviations for both NDVI (0.245), suggesting a more fragmented and spatially heterogeneous distribution of green space, which may be attributed to its traditional garden-style layout.

Regarding subjective perception variables, older respondents generally expressed positive attitudes toward the park environments. All mean scores for questionnaire dimensions exceeded the midpoint value of 4, with PC = 4.795, SOP = 4.891, PR = 4.810, and PW = 4.747, indicating that elderly participants had relatively positive experiences with the current UGS. Specifically, the highest sub-dimension scores were observed in facility quality within PC (4.840), place attachment within SOP (5.002), and coherence within PR (4.871).

4.2. Confirmatory Factor Analysis

This study employed Amos 28.0 to conduct confirmatory factor analysis (CFA) to examine the correlations between hypothesized factors and latent variables [48].

Table 3. Goodness-of-fit metrics of the measurement model.

Model Fit Index	X2/dF (Chi-Square)	Comparative Fit Index (CFI)	Normed Fit Index (NFI)	Root Mean Square Error of Approximation (RMSEA)
Recommended values	1–3	>0.95	>0.90	<0.08
Results	1.130	0.993	0.940	0.012

presented the model fit indices for the measurement model. According to [78], model fit is considered ideal when CFI > 0.95, NFI > 0.90, and RMSEA < 0.08. Therefore, the model in this study demonstrated a good fit.

Regarding convergent validity, all standardized factor loadings ranged from 0.738 to 0.865 and were significant at p < 0.001, meeting the criteria proposed by [79] for acceptable convergent validity. The composite reliability (CR) values ranged from 0.804 to 0.956, and the average variance extracted (AVE) values ranged from 0.596 to 0.690. According to the thresholds of CR > 0.7 and AVE > 0.5 [80,81], these results indicate that all constructs showed good internal consistency and convergent validity. Furthermore, the square roots of the AVE values exceeded the inter-construct correlations, confirming adequate discriminant validity among the measurement dimensions [80].

4.3. Multilevel SEM

SEM is a statistical approach used to construct, estimate, and test causal relationship models, allowing for the measurement of both the strength and direction of causal pathways, and enabling the simultaneous examination of direct, indirect, and total effects among multiple variables [4]. This study constructed a multilevel SEM in which NDWI, NDVI, and PC were specified as independent variables; SOP and PR as mediating variables; and PW as the dependent variable, with demographic characteristics included as control variables. The multilevel SEM model showed satisfactory fit, with X²/dF = 1.143, CFI = 0.991, NFI = 0.929, and RMSEA = 0.013, aligning with the acceptable criteria proposed by [78].

Table 4. Path coefficients and multiple parameters in multilevel SEM.

Model Pathways	R2	Estimate	Std Estimate	S.E.	C.R.	p
NDVI→SOP	0.410	2.593	0.285	0.574	4.514	***
NDWI→SOP		0.704	0.065	0.611	1.152	0.249
PC→SOP		0.528	0.518	0.058	9.149	***
NDVI→PR	0.570	3.485	0.394	0.538	6.475	***
NDWI→PR		2.575	0.246	0.548	4.699	***
PC→PR		0.320	0.322	0.058	5.532	***
SOP→PR		0.299	0.307	0.064	4.648	***
NDVI→PW	0.490	1.982	0.155	0.689	2.875	0.004
NDWI→PW		0.739	0.049	0.703	1.050	0.293
PC→PW		0.456	0.317	0.076	6.013	***
SOP→PW		0.252	0.179	0.084	2.996	0.003
PR→PW		0.310	0.214	0.086	3.610	***
Sex→PW		−0.006	−0.006	0.032	−0.201	0.840
Age→PW		−0.038	−0.013	0.081	−0.465	0.642
Edu.→PW		0.084	0.068	0.057	1.464	0.143
Inc.→PW		0.015	0.012	0.056	0.264	0.792

Note: ***: p < 0.001.

presented the path coefficients and parameter estimates for the multilevel SEM, and Figure 6 showed the standardized estimates of the multilevel SEM path analysis. Both NDVI and PC exerted significant positive effects on SOP, PR, and PW, indicating that hypotheses H1a, H2a, and H3a were partially supported. This indicated that high vegetation coverage and positive environmental perceptions not only directly promoted older adults’ PW but also enhanced the potential of SOP and PR in influencing PW. This further suggests that older adults’ emotional attachment to UGS and psychological restoration experiences served as key psychological mechanisms through which environmental characteristics were translated into health benefits. Additionally, these findings underscored the importance of both the quantity and quality of green spaces in constructing age-friendly environments. Notably, in terms of influencing PW, PC (β = 0.317, p < 0.001) demonstrated a significantly stronger effect than NDVI (β = 0.155, p = 0.004), indicating that the health benefits derived by older adults were more dependent on their perception of the environment than on objective greenness measures. Within the PC dimension, facilities (β = 0.75) and accessibility (β = 0.73) were the most influential subcomponents, whereas the influence of species diversity was comparatively minimal (β = 0.69). This finding demonstrated that well-maintained facilities and accessibility were the strongest environmental predictors of PW in the older adults.

SOP exerted significant positive effects on both PR (β = 0.307, p < 0.001) and PW (β = 0.179, p = 0.003), supporting hypothesis H1b, H2b. This indicates that long-term emotional attachment to a place can translate into health benefits through the psychological restoration process. In terms of influencing SOP, PC (β = 0.518, p < 0.001) demonstrated a significantly stronger effect than NDVI (β = 0.285, p < 0.001).

PR exerted a significant positive effect on PW (β = 0.214, p < 0.001). In terms of influencing PR, both NDVI (β = 0.394, p < 0.001) and PC (β = 0.322, p < 0.001) demonstrated the strongest predictive power, while NDWI (β = 0.246, p < 0.001) showed the smallest effect among these predictors.

An interesting finding was that NDWI showed no significant effect on SOP (β = 0.065, p = 0.249) or PW (β = 0.049, p = 0.293), but it indirectly influenced older adults’ PW through PR. This indicates that although BS did not directly enhance older adults’ SOP or PW, the ambiance and sensory attributes of BS were more important than the mere physical water area. The mechanism through which BS affected older adults emphasized emotional regulation and psychological restoration rather than the long-term construction of emotional attachment.

A CI is a range on either side of the estimate, within which we can be 95% confident that the true value lies [82].

Table 5. The direct, indirect, and total effects of NDVI, NDWI, and PC on WHO-5 scores.

Model Pathways	Direct Effect Coef.	Indirect Effect Coef.	Total Effect Coef.	LLCI	ULCI	p	Effect Ratio
NDVI→PW	1.982		3.953	0.506	3.618	0.013	50.14%
NDVI→SOP→PW		0.652		0.166	1.462	0.006	16.49%
NDVI→PR→PW		1.079		0.387	2.013	0.002	27.30%
NDVI→SOP→PR→PW		0.24		0.064	0.620	0.001	6.07%
NDWI→PW	0.739		1.778	−0.755	2.262	0.308	41.54%
NDWI→SOP→PW		0.177		−0.095	0.713	0.184	9.95%
NDWI→PR→PW		0.797		0.268	1.612	0.002	44.80%
NDWI→SOP→PR→PW		0.065		−0.039	0.294	0.204	3.65%
PC→PW	0.456		0.737	0.284	0.670	0.000	61.96%
PC→SOP→PW		0.133		0.032	0.250	0.010	18.07%
PC→PR→PW		0.099		0.031	0.202	0.002	13.45%
PC→SOP→PR→PW		0.049		0.018	0.100	0.002	6.66%

Note: LLCI is the lower limit of the 95% range of estimates, and ULCI is the upper limit of the 95% range of estimates.

summarizes the direct, indirect (comprising simple mediation, chain mediation, and total mediation), and total effects across all specified paths. NDVI and PC both exerted significant indirect effects on PW through SOP and PR, as the 95% confidence intervals did not include zero. Specifically, the total effect of NDVI on PW was 3.953, with a direct effect of 1.982 and an indirect effect of 1.971 (p < 0.001). However, since the confidence intervals for the pathways NDWI→SOP→PW and NDWI→SOP→PR→PW included zero, these mediating effects were not significant. In contrast, the NDWI→PR→PW pathway’s confidence interval excluded zero, confirming its significant mediating effect. Therefore, hypothesis H4 was partially supported. For PC, the total effect was 0.737 (direct effect = 0.449; indirect effect = 0.456, p < 0.001). These findings further indicate that in the pathways through which NDVI and PC influence PW, SOP not only plays an independent mediating role but also functions jointly with PR in a chain mediation role.

5. Discussions

This study systematically investigated the complex relationships between objective UGS exposure (NDVI and NDWI), subjective PC, and elderly PW, with a particular focus on the mediating roles of SOP and PR. The findings revealed the associations and path relationships among these variables, highlighting the importance of the proposed theoretical framework “UGS exposure characteristics (objective and subjective)—psychological mediators (SOP + PR)—health effects” in guiding age-friendly UGS development. Furthermore, demographic variables had no significant effect on PW in this study, which may be due to the long-term life experiences that foster stable emotional ties and identity with UGS, thereby reducing the interference of individual background factors.

In response to research question one, our findings reveal that among older adults, objective exposure to UGS (NDVI and NDWI) and subjective PC exert differentiated effects on PW through the mediating pathways of SOP and PR. We found that subjective perceptions were stronger predictors of PW in the older adults than objective exposure. This finding was consistent with the study by [83], which emphasized that vulnerable groups were more influenced by the quality of GS than by other GS exposures. Across the pathways influencing SOP, PR, and PW, PC emerged as nearly the most explanatory factor. This may be attributed to the fact that environmental perception engages complex multisensory processes and involves cognitive-affective dimensions [84]. Compared to passive exposure, active perception exerts a more substantial impact on mental well-being, particularly in emotionally resonant and restorative settings [85]. Second, PC exerted a greater influence on SOP than on PR. While SOP captures long-term emotional ties emerging from sustained human–environment interactions, and is shaped by comprehensive perceived attributes (e.g., facilities, cultural value, sociability, safety, and accessibility), PR emphasizes immediate psychological restoration, relying more on environmental features such as tranquility, openness, and spatial extent, making its scope narrower. Therefore, UGS attributes are more effective in strengthening place attachment and place identity [74,86], rather than directly enhancing restorative potential. Lastly, PR exhibited a stronger impact on PW than SOP. Theoretically, PR addresses short-term, rapid attentional recovery in restorative environments, whereas SOP reflects long-term emotional ties, identity, and sense of belonging. Accordingly, PR may exert more direct and immediate effects on emotional and cognitive restoration in older adults. This finding was consistent with the ART [68].

Interestingly, NDWI did not directly affect SOP or PW but exerted an indirect effect on PW through PR. This suggested that the pathway through which BS influences health tends to follow a “restorative-cognitive mechanism” rather than an “emotional attachment mechanism,” which aligns with the ART [13]. Visual stimuli undergo a stimulus-organism-response process, where open water bodies influence environmental perception and psychological restoration. BS with expansive and continuous water features facilitates emotional and attentional recovery more effectively in older adults. However, the mere physical presence of water features is insufficient to foster deep emotional attachment and identity toward a place, as the sense of place is a complex construct developed over the long term. This also reveals the differentiated psychological mechanisms through which natural elements promote health among older adults.

Diverging from prior research, ref. [48] highlighted fascination as the most powerful dimension of PR. In contrast, our findings indicate that “extent” and “coherence” are the most influential PR dimensions. While extent reflects the spatial expansiveness of the environment, coherence pertains to the organizational clarity and logical structure of environmental settings. Open environments provide visual comfort that strengthens orientation, promotes perceived safety, and facilitates psychological restoration [12,87]. Spaces that are well-organized, serene, and secure reduce environmental distractions and cognitive burden, offering opportunities for rest and contemplation, which in turn help to alleviate psychological stress [12,66,68]. Accumulated evidence from landscape ecology similarly suggested that UGS with higher connectivity and continuity were more likely to enhance visual perception and facilitate restorative outcomes [88]. Consistent with [89], our results highlight that readability, coherence, and accessibility should be prioritized in the design and management of UGS.

For research question two, our study demonstrated that both subjective and objective features of UGS exposure influence older adults’ PW via a sequential mediation pathway involving SOP and PR, indicating a chained mediation mechanism. Several studies lend indirect support to our findings. Van Dinter et al. [16] revealed that SOP mediates the association between park features and individuals’ subjective well-being, with park attributes directly linked to SOP and SOP exerting a positive and significant effect on subjective well-being. Parks are among the most preferred places for older adults to visit and relax. Familiar environmental cues in the local context can elicit place attachment, strengthen the continuity of SOP and place identity [59], and contribute to enhanced restorative experiences [90]. Even exposure to virtual representations of parks can foster place attachment and promote feelings of relaxation [89]. Previous studies showed that place attachment positively contributed to individuals’ psychological restoration [15,90,91]. When people stay in places with positive emotional attachment, those places have restorative potential [74]. The addition of distinctive landscape features can enrich perceptual engagement and more effectively foster place attachment and restorative experiences [15]. The integration of high-quality UGS characteristics with locally distinctive landscape features can maximize individuals’ restorative outcomes [4,15,91]. Compared to other demographic groups, older adults are more readily evoked into positive emotional states through UGS exposure, which in turn fosters emotional bonding and environmental interaction, leading to increased restorative potential. Consistent with previous research, our results demonstrated that vegetation coverage, well-maintained facilities, and accessibility were among the most critical environmental characteristics in age-friendly parks.

In response to the final research question, this study provides three empirically grounded recommendations for the planning and design of age-friendly UG (

Table 6. Empirically grounded recommendations for planning and design of age-friendly UGS.

Variable	Strategy	Implementation Measures
Sense of Place	Urban planning decisions	Optimizing urban design parameters, distinctive urban landmarks, urban features, and city branding strategies
Perceived Restorativeness	Promoting the extent and coherence of spatial design	Focusing on coherent landscape structures (e.g., green spaces and water bodies) and tranquil areas
Perceived Characteristics	Improving the Quality and Quantity of Urban Green Space	Enhancing facility quality and improving accessibility

). First, we advocate for urban planning decisions that enhance the SOP by prioritizing design urban parameters [92], distinctive urban landmarks, urban features [93,94], and city branding strategies [95]. And actively shaping place-based qualities, spatial forms, landscape elements, and diversity [15,57]. Second, although our findings suggest that SOP contributes to older adults’ PW by fostering emotional bonds and identity, UGS environments with restorative attributes tend to deliver superior health benefits compared to SOP alone. This underscores the need to prioritize the inclusion of restorative green environments when constructing or retrofitting age-friendly urban areas, especially the green space with extent and coherence. Finally, to improve older adults’ PW and overall well-being, greater emphasis should be placed on perceptible green space qualities (e.g., good facilities, nearby accessibility) than on the amount of exposure. We recommend implementing these three strategies and formulating interventions that are informed by the empirical findings of this research.

While the contributions of this study are substantial, several limitations should be acknowledged, alongside directions for future research: First, the study was conducted only in Fuzhou, a subtropical coastal second-tier city in Fujian Province. Future research should include comparative studies across cities with differing geographic, climatic, and socio-economic contexts to test the generalizability of the proposed theoretical framework. Second, the present study did not differentiate elderly subgroups with chronic conditions such as cardiovascular disease, diabetes, respiratory illnesses, or cognitive impairments. Given that these populations often exhibit distinct attitudes toward and needs for UGS [96], their PC, SOP, and PR may vary accordingly. Future research should adopt a stratified approach to examine how diverse health profiles affect the applicability of this framework. Finally, as this study primarily relied on self-reported measures to assess SOP, PR, and PW, future research should consider incorporating wearable physiological sensing technologies, such as electroencephalography, heart rate variability, and skin conductance to more objectively and accurately capture the mental health benefits of green space exposure [97].

6. Conclusions

6.1. Key Findings

The key findings are summarized as follows: (1) There were significant differences in the pathways from objective UGS exposure (NDVI) and subjective PC to PW through SOP and PR. (2) Green space affected the elderly PW via SOP and PR, existing in the “SOP-PR” link of the mediated relationship, and PR moderated the link between SOP and PW. (3) NDWI showed no significant direct effect on SOP or PW but exerted an indirect impact on PW through PR. (4) Among older adults, facility quality and accessibility emerged as the most salient features shaping PC; within SOP, place attachment showed the strongest explanatory power, whereas coherence and scope were the most influential components within the PR pathway. (5) UGS exposure characteristics, place qualities, and restorative elements are critical for guiding age-friendly urban planning. All hypothesis testing results are presented in

Table 7. Hypothesis testing results.

Hypothesis	Path Tested	Result
H1a	NDWI→PW	Not supported
	NDVI→PW	Supported
	PC→PW	Supported
H2a	NDWI→SOP	Not supported
	NDVI→SOP	Supported
	PC→SOP	Supported
H3a	NDWI→PR	Supported
	NDVI→PR	Supported
	PC→PR	Supported
H1b	SOP→PR	Supported
H2b	SOP→PW	Supported
H3	PR→PW	Supported
H4	NDWI→SOP→PR→PW	Not supported
	NDVI→SOP→PR→PW	Supported
	PC→SOP→PR→PW	Supported

.

6.2. Research Contributions

Based on field investigations in Xihu Park, Jin’an Park, Qinting Park, and Fuzhou National Forest Park in Fuzhou, China, this study used a multilevel SEM to construct a theoretical framework of “physical environmental features—psychological mediators (SOP + PR)—health outcomes,” and systematically explored the complex impacts of objective UGS exposure (NDVI, NDWI) and subjective PC on elderly PW through SOP and PR. The results demonstrated that NDVI, PC, SOP, and PR are important variables influencing PW.

This study advances the existing “UGS exposure–health” theoretical framework by empirically demonstrating the feasibility of enhancing PW among older adults through strengthening their SOP. Accordingly, future research should incorporate the dimension of SOP to gain a more comprehensive understanding of the relationship between UGS exposure and health outcomes. These insights are of considerable relevance to public health, epidemiology, environmental psychology, and urban planning, and call for broader comparative and empirical investigations by researchers across diverse global contexts.