Aging-in-Place Attachment Among Older Adults in Macau’s High-Density Community Spaces: A Multi-Dimensional Empirical Study

Abstract

This study explores key factors influencing Aging-in-Place Attachment (AiPA) among older adults in Macau’s high-density community spaces, emphasizing interactions between the built environment, behavior, and psychology. A multidimensional framework evaluates environmental, behavioral, human-factor, and psychological contributions. A mixed-methods, multisource approach was employed. This study measured spatial characteristics of nine public spaces, conducted systematic behavioral observations, and collected questionnaire data on place attachment and aging intentions. Eye-tracking and galvanic skin response (GSR) captured visual attention and emotional arousal. Hierarchical regression analysis tested the explanatory power of each variable group, supplemented by semi-structured interviews for qualitative depth. The results showed that the physical environment had a limited direct impact but served as a critical foundation. Behavioral variables increased explanatory power (~15%), emphasizing community engagement. Human-factor data added ~4%, indicating that sensory and habitual interactions strengthen bonds. Psychological factors contributed most (~59%), confirming AiPA as a multidimensional construct shaped primarily by emotional and social connections, supported by physical and behavioral contexts. In Macau’s dense urban context, older adults’ desire to age in place is mainly driven by emotional connection and social participation, with spatial design serving as an enabler. Effective age-friendly strategies must extend beyond infrastructure upgrades to cultivate belonging and interaction. This study advances environmental gerontology and architecture theory by explaining the mechanisms of attachment in later life. Future work should explore how physical spaces foster psychological well-being and examine emerging factors such as digital and intergenerational engagement.

1. Introduction

As the global trend of population aging intensifies, especially in high-density Asian cities, the concept of “aging in place”, the ability of older adults to continue living safely and comfortably in familiar environments, has become a pressing urban and public health concern [1,2]. This issue is particularly salient in Macau, a compact, multicultural city where the aging population is rising rapidly and community-based care systems are still evolving [3]. In this context, older adults’ emotional attachment to public spaces is emerging as a key factor for their psychological resilience, social participation, and long-term well-being [4,5]. A growing body of literature underscores that the success of aging in place is deeply rooted in place attachment—the emotional bond individuals form with specific environments [6,7]. For older adults, such bonds transcend private dwellings and increasingly encompass community public spaces where everyday activities and social interactions unfold [1,8,9].

In high-density urban contexts such as Macau, community public spaces—including local parks, small plazas, temple forecourts, and neighborhood rest areas—function as extensions of older adults’ everyday living environments [10,11]. Often regarded as communal “living rooms,” these spaces host routine activities such as morning exercises, casual conversations, board games, sunbathing, and neighborhood interactions [12]. Through sustained, repetitive use, they become repositories of memory and emotion, woven into the fabric of older adults’ lived experiences [13]. As such, these spaces transcend their physical utility and evolve into emotionally and socially supportive environments [14,15]. Recent studies have shown that continued engagement in such spaces significantly strengthens older adults’ sense of community identity and emotional attachment [5]. Such community spaces fulfill essential psychological needs by offering companionship, routine, and a connection to local culture, thereby supporting aging-in-place through emotional bonds to the environment [16].

Place attachment, commonly defined as the emotional bond between people and specific physical environments, is a foundational concept in environmental psychology and gerontology [7,17,18]. Existing studies identify a wide range of factors influencing place attachment in later life, including physical environmental qualities, social opportunities, and psychological needs [19,20,21]. However, these factors are rarely examined in an integrated manner. Meanwhile, increasing evidence suggests that such influences are not exerted independently, but rather through complex mediating pathways, such as environmental satisfaction, social participation, or place meaning [22,23,24]. Despite this, empirical research remains limited, particularly in relation to aging in place within the context of public spaces in high-density Asian cities. Moreover, the relationship between place attachment and aging in place remains insufficiently understood. Existing traditional place attachment theories have yet to effectively capture the long-term residential intentions and emotional bonds of older adults within the specific context of later-life stages, leaving a significant gap in current research.

To address these research gaps, this study introduces and operationalizes a specialized form of attachment, referred to as “aging-in-place attachment (AiPA)”, which reflects the aging experience within community environments and is conceptualized as an emotionally grounded commitment by older adults to age within the public spaces of their community. Consistent with Scannell & Gifford’s Person–Process–Place framework [6], AiPA leaves the Person and Place poles intact but extends the Process pole by foregrounding three late-life indicators. Building upon place attachment theory and incorporating the physical, psychological, and behavioral specificities of aging, this study argues that AiPA is particularly salient in high-density urban settings such as Macau, where spatial limitations, cultural hybridity, and demographic pressures converge. To empirically examine this concept, the study proposes a comprehensive analytical framework that investigates the formation mechanisms of AiPA among older adults in Macau. It draws a multisource methodology combining spatial, behavioral, psychological, and physiological data. The study investigates four principal dimensions of influence: physical environmental variables [25,26], behavioral patterns [27,28,29,30,31,32,33], human-factor metrics [34,35], and psychological factors [16,25,27,28,36,37,38,39,40,41,42,43]. The outcome variable, aging-in-place attachment, is conceptualized as an extension of place attachment theory and is measured using three indicators: willingness to remain in the current location, satisfaction with the environment, and perceived dependence on the local community.

The central research question guiding this study is as follows: In the context of high-density Asian cities such as Macau, to what extent do physical environmental features, spatial behavior patterns, human-factor data, and psychological variables explain older adults’ attachment to aging in place, and how do these variables contribute, individually and collectively, to the explanatory power of the proposed empirical model? Regarding the analysis method, while Structural Equation Modeling (SEM) has been widely adopted in this field for its capacity to model latent constructs and complex mediation pathways, hierarchical regression analysis offers a complementary approach that allows for a clearer examination of the incremental explanatory power of grouped variables, particularly useful when evaluating nested or context-specific factors in aging-related studies. This stepwise approach not only complements SEM but also better isolates context-sensitive variable clusters in empirical aging studies. By applying hierarchical regression analysis, this study sequentially assesses how these variables explain AiPA, allowing us to test the relative and layered importance of physical, behavioral, physiological, and psychological factors.

Ultimately, this study delivers three interlocking contributions. First, it introduces AiPA—a construct that fuses classic place-attachment theory with the explicit intention to remain in one’s community, thereby translating an abstract psychological bond into a policy-relevant indicator for age-friendly planning. Second, by integrating five data streams—site-scale spatial audits, behavioral observations, eye-tracking + GSR metrics, matched questionnaires, and in-depth interviews—this study assemble a multi-layer evidence base that has been employed in only a handful of related studies, allowing the relative weights of environmental, behavioral, sensory, and psychological drivers to be compared within one unified model. Third, the focus on Macau’s hyper-dense, culturally hybrid public spaces fills a geographical gap in the literature and yields design guidance for other Asian cities where land scarcity and rapid population aging collide. Taken together, these advances offer both a theoretical bridge between environmental psychology and gerontology and an evidence-based toolkit for architects, planners, and policymakers seeking to create emotionally resonant, socially vibrant, and spatially feasible environments for older adults. Future research can extend this framework by testing intergenerational dynamics, digital engagement, or institutional supports to refine AiPA across diverse urban contexts.

2. Literature Review

2.1. Understanding Place Attachment

Place attachment is commonly defined as a positive emotional bond between individuals and specific places [44]. Scannell and Gifford (2010) conceptualize it as a multidimensional construct involving Person, Process, and Place dimensions–essentially, who is attached, how this attachment is expressed (affectively, cognitively, and behaviorally), and what place characteristics form the focus of attachment [6]. In gerontological contexts, this emotional connection to place is especially salient. Older adults often have longer residence tenures and narrower activity ranges, which can deepen their sense of attachment to home and community environments [7]. For example, Wiles et al. (2012) found that “aging in place” is perceived by older adults as advantageous largely due to feelings of security, familiarity, and identity derived from staying in one’s home and neighborhood [1]. Over time, older people accumulate rich life memories and experiences in their surroundings, which strengthens emotional commitment to those places [45]. Research in rural UK settings noted that long-term older adults developed multifaceted place attachment (physical, social, temporal, and psychological), with an additional historical dimension reflecting the depth of accumulated memories [25]. These findings underscore that aging amplifies the importance of place as a source of continuity and meaning in life. Accordingly, we treat AiPA as a contextual refinement—not an additional dimension—of the classical model, with its three indicators mapping directly onto the Process component that captures how attachment is enacted.

Notably, the mechanisms of place attachment can differ by life stage or residential history. Younger newcomers in a community often establish a sense of belonging relatively quickly through new social networks, whereas older adults–especially those who relocate later in life–tend to rely more on environmental familiarity and past place-based experiences to re-form attachments [6,7,46]. In contrast, older adults tend to build attachment through long-term environmental familiarity and emotionally significant memories. Studies show that older adults are more invested in their immediate home and neighborhood environments, whereas younger people often express attachment at larger spatial scales, such as the city or region [20,47]. Moreover, older people’s attachments are more vulnerable to environmental disruptions; even minor changes to a long-familiar milieu can threaten their sense of stability and well-being (often described as a “rootedness” or place dependence effect) [45]. In short, late-life place attachment has unique characteristics, rooted in longevity of place experience, sensitivity to change, and profound ties to personal identity, that require specialized theoretical and empirical attention. While ‘place attachment’ broadly refers to emotional bonding with place, this study emphasizes AiPA as its context-specific extension that reflects sustained affective commitment to aging within one’s familiar community environment. This study aims to connect environmental psychology with age-sensitive public space design.

2.2. Environmental and Social Foundations of Place Attachment Among Older Adults

Recent scholarship has proposed several conceptual frameworks to understand the mechanisms by which older adults form emotional bonds with place. For instance, Aliakbarzadeh Arani et al. conducted a systematic scoping review and identified five core dimensions of place attachment in later life [21]. Building upon this integrative view, this study focuses on three primary clusters: (a) physical environmental attributes, (b) social interaction and behavioral patterns, and (c) psychological-emotional factors, to capture the multidimensional pathways through which older adults develop aging-in-place attachment.

2.2.1. Physical Environment

The physical environment is a critical factor influencing the development of place attachment among older adults, particularly through three interrelated dimensions: physical familiarity, aesthetic appeal, and elder-friendly accessibility. Drawing on Burholt’s. An empirical study of over 900 rural elders in England and Wales, aesthetic attachment, relating to visual and sensory qualities, reflecting the emotional impact of pleasing environments on older individuals [25]. Expanding on these findings, Pereira et al. [48] and Lies et al. [26] found that design features such as natural landscaping, walking paths, and green gathering areas foster natural attachment by creating restorative and serene experiences that support socialization and independence among older adults. Additionally, Wiles et al. [31,43] argue that dwellings and surrounding green spaces can evoke physical familiarity, reinforcing emotional bonds through daily use and sensory reassurance. Fu et al. (2025) demonstrated that wide spatial openness, elevated sky views, and high green cover not only contributed to visual comfort but also increased older adults’ environmental satisfaction and emotional attachment, particularly in compact urban settings [49]. In urbanized and high-density settings such as Macau, the dimension of elder-friendly accessibility is equally critical. For instance, a recent study in Macau found that neighborhoods with higher visible greenery and a greater density of resting places were positively associated with better functional independence in older residents [12]. Pedestrian-oriented designs, continuous smooth sidewalks, ramps, clear signage, and rest areas encourage older adults to venture out and engage with the community, thereby strengthening their emotional bond to the locale [50]. Conversely, environmental barriers such as steep slopes, poor lighting, lack of seating, or slippery surfaces create mobility difficulties and reduce elders’ sense of belonging [50]. However, from an architectural perspective, it is essential to emphasize that the perceived influence of the physical environment may often be underreported in survey-based research. As previous scholars have noted, self-reported data, especially among older adults, may not fully capture the subconscious or indirect effects of spatial configurations [51,52]. This could be due to limited environmental awareness, cultural communication patterns, or the implicit nature of spatial experience.

2.2.2. Social Interaction and Behavior

Social interaction and behavioral routines play a pivotal role in shaping place attachment among older adults, especially in the context of aging in place. Empirical studies show that older adults who engage more frequently in social, recreational, and cultural interactions report stronger place attachment and enhanced well-being [53,54]. For example, Hwang and Sim [54] demonstrated that older adults living alone in South Korea who participated in more social activities had significantly better self-rated health and happiness. Likewise, community social support networks are crucial: having nearby family or supportive neighbors makes elders feel cared for and rooted in place [23,27,28,29,30,31]. Older adults’ emotional attachment to their communities is largely shaped by social interactions and neighborhood ties [24]. Community public spaces that facilitate both peer-to-peer and intergenerational engagement, such as conversations, group exercise, or recreational activities, can significantly fulfill older adults’ emotional needs and strengthen their sense of belonging and place identity [24]. Moreover, residential longevity has been consistently associated with higher levels of place attachment. The longer older individuals reside in a familiar environment, the more accumulated memories and life experiences they associate with that place, resulting in deeper emotional bonds [20]. This also explains why relocation from familiar neighborhoods often leads to emotional distress among older adults. Therefore, social behaviors are not merely reflections of the physical or social environment; they are constitutive of place attachment processes. Particularly in aging populations, such behavioral routines function as both indicators and reinforcers of emotional bonds to place, making them a critical dimension of AiPA. This study situates social interaction and spatial behavior as core components in the multidimensional analysis of AiPA.

2.2.3. Psychological and Emotional Factors

Underlying psychological needs and perceptions ultimately drive an older person’s attachment to place; emotional security, autonomy, identity, and continuity of life story are all internal motives that link a person to their environment [16,25,27,28,36,37,38,39,40,41,42,43]. As physical capacities decline with age, familiar places serve as a crucial source of comfort and control—providing a sense of mastery over one’s daily life and freedom from anxiety about the unknown [1]. Butcher and Breheny (2016) describe how older Māori in New Zealand drew on a “comforting and comfortable dependence” on their land and family ties to achieve a feeling of autonomy in later life [45]. This illustrates a broader point: familiar home and community environments can function as an extension of self, enabling older adults to maintain identity and dignity [31,39,40,42,55,56,57]. Long-term residents often attach deep place meaning to their surroundings, with memories and personal or cultural symbols embedded in the landscape [24]. Such places become repositories of life stories—evoking nostalgia and affirming one’s sense of continuity over time. When an older person has strong affective ties and memories connected to where they live, they are typically reluctant to leave that place voluntarily [24]. In summary, psychological attachments, the feelings of belonging, historical identity, and the emotional relief that places provide are powerful determinants of whether an elder truly feels “at home” and willing to age in place.

2.3. Multi-Path Pathways and Mediating Mechanisms

It is worth noting that these environmental, social, and psychological factors do not operate in isolation; rather, they may influence older adults’ place attachment through complex, interacting pathways. Recent research emphasizes that certain mediators, such as environmental satisfaction, social participation, or place meaning, often explain how various factors translate into stronger attachment or well-being [23,58]. Clark notes that satisfaction with one’s home and neighborhood is tightly linked to older adults’ desire to remain living there, reiterating its role as a key component of place attachment [58]. Hwang and Sim’s SEM model (2021) showed that beyond its direct benefits, frequent social interaction contributed to greater neighborhood satisfaction, which subsequently enhanced the subjective well-being of older adults [54]. Another study in Iran by Lak et al. (2023) used a structural equation approach to demonstrate that older people’s perceptions of public space attributes (e.g., security, walkability, aesthetics), together with latent factors like place attachment and life satisfaction, collectively influenced their physical and mental health outcomes [23]. Other empirical studies have confirmed that place attachment mediates the relationship between attitudes toward aging and subjective well-being, thereby emphasizing its pivotal role in the aging process [22]. These multiple routes illustrate that place attachment is not a single monolithic influence, but rather a bundle of mechanisms through which aging in place becomes impactful to older adults’ quality of life. These intertwined pathways call for empirical models that not only assess direct effects but also account for mediating mechanisms and cumulative layering of influences. The current study aims to respond to this call by structuring its analytical framework accordingly.

2.4. Theoretical Positioning and Contributions of the Present Study

This study builds upon existing research on place attachment in later life, responding to calls for a more integrative understanding of how environmental, behavioral, and psychological factors interact to shape aging-in-place outcomes. While prior studies have examined these domains separately, few have proposed an empirical framework that captures their dynamic interplay—particularly in the high-density, culturally complex urban contexts of Asia.

To address this gap, this study introduces the concept of Aging-in-Place Attachment (AiPA) as a context-specific extension of place attachment. This study advocates comprehensive frameworks that examine these layered influences together rather than in silos. By accounting for the combined effects of environmental design, social context, and individual psychology, and the mediating variables among them, scholars and urban practitioners can better identify what truly fortifies older adults’ desire and ability to age in place. This study builds on a design-sensitive model of place attachment, aiming to empirically validate the multidimensional pathways proposed in the literature and to inform age-friendly urban design with an evidence-based understanding of late-life place attachment.

3. Materials and Methods

Overview of the mixed-methods design. To guide the reader through the study’s multi-layered methodology, the authors summarize the full sequence here before presenting each module in detail. Five data-collection modules were executed in the order listed: (1) a physical audit of nine community spaces (N = 9 sites); (2) systematic behavioral observation of older-adult users at those sites (N ≈ 7800 observational snapshots); (3) a laboratory-based eye-tracking + galvanic-skin-response experiment with 63 elders drawn from the same neighborhoods; (4) a matched questionnaire survey of those 63 participants covering place attachment, behavioral intention and demographics; and (5) semi-structured interviews with 13 of the surveyed elders to elicit narrative depth. All quantitative streams were integrated through hierarchical regression to assess the incremental explanatory power of each variable block, while the interview material was thematically analyzed to contextualize and challenge the statistical results. The subsequent Section 3.1, Section 3.2, Section 3.3, Section 3.4, Section 3.5, Section 3.6 and Section 3.7 describe each module in full.

3.1. Study Area

This study identified 9 community public spaces in Macau through a purposive sampling strategy, guided by criteria of representativeness, cultural diversity, and research feasibility. Selection considerations included the spatial scale of the site, the functional variety of facilities, and, most importantly, the frequency of older adult usage. Preference was given to locations situated within neighborhoods with a high density of older adults, as these spaces are deeply embedded in daily routines and social practices. To reflect Macau’s unique socio-cultural landscape, place meaning, social configuration, environmental atmosphere, and levels of local integration, the selected sites were categorized into three typologies: (1) traditional Chinese-style spaces, (2) Portuguese-influenced spaces, and (3) hybrid Sino-Portuguese spaces. Although the spatial sizes varied, all nine spaces share characteristics of medium-scale open areas typical of Macau’s urban plazas, which have historically supported community interaction and daily living.

Figure 1 illustrates the spatial distribution and cultural classification of the selected sites, mapping their location across various urban districts. It also includes comparative data on the proportion of older adult users observed during weekday and weekend periods, providing an empirical foundation for assessing patterns of use and engagement. Data was collected through systematic field observations conducted specifically for this study. A preliminary version of these observational data and site classifications was previously published in the context of instrumental activities of daily living (IADL) research [12]. The current study extends this earlier work by applying the data within a new conceptual and analytical framework focused on AiPA.

3.2. Physical Features of Community Public Spaces

The assessment of Macau’s public spaces was grounded in the conceptual framework of aging in place [58], with particular attention to how the built environment can support autonomy, well-being, and sustained social interaction in later life. Drawing from an interdisciplinary scoping review of place attachment and environmental gerontology [21], three key dimensions of physical features were identified: spatial characteristics, natural elements, and elder-friendly and accessible design attributes [12].

For spatial characteristics, metrics included site area, spatial enclosure, paved surface ratio, and visual brightness uniformity. Natural elements were evaluated through metrics such as visible greenery ratio, plant species diversity, and green area proportion. The elder-friendly and accessible design dimension included facility diversity, the density of resting and leisure amenities, and entrance accessibility. All data were collected through systematic field observation, supported by local geospatial sources where applicable.

Table 1. Operational definitions and measurement methods of physical environmental features in Macau’s public spaces.

Environmental Dimension	Indicator	Measurement Description	Data Collection Method	Source
Spatial Characteristics	Area	Total surface area (m2) of the designated public space, measured based on ground boundaries.	On-site survey and spatial mapping	[59]
	Percentage of Ground Pavement	Proportion of ground covered by impervious materials (e.g., concrete and pavers) relative to total area.	On-site visual classification and area computation	[60]
	Spatial Enclosure	Ratio between the length of enclosing physical structures and the total perimeter. Reflects perceived enclosure.	GIS and On-site measurement	[61]
	Spatial Visual Brightness	Luminance readings collected at uniform grid points within the space; the standard deviation is used to express perceived brightness contrast.	Light meter sampling and spatial averaging	[62]
Natural Characteristics	Green Coverage Ratio	Ratio of green elements (lawns, planters, and trees) to the total ground surface area.	Field observation and area calculation	[63]
	Visible greenery ratio	Proportion of visible greenery within field of view based on horizontal photos taken at eye level and analyzed through sampling.	Image-based sampling and visual coding	[64]
	Plant Diversity	Number of distinct plant species adjusted for site area using the modified Patrick index: R = S/log A.	Botanical field survey	[65]
	Green Space Ratio	Area dedicated to ecological or recreational planting (e.g., tree beds, lawns, and flower zones) as a percentage of total site area.	Land use sketch and calculation	[66]
Elder-friendly and Accessibility Features	Functional Diversity	Number of distinct facility types (e.g., seating, toilets, and ramps) per site, standardized as F = N/log A.	Inventory mapping	[67]
	Recreational Facilities	Total number of grouped recreational installations (e.g., benches, play zones, and fitness sets).	Direct field count	[68]
	Density of Resting Facilities	Number of seating/resting elements per 100 m2 of space.	Unit-based calculation	[69]
	Pathway Connectivity	Calculated using C = E/log P, where E = number of entrances and P = perimeter length, representing spatial accessibility.	Field perimeter measurement and entry point count	[70]

(adapted from [12]) succinctly lists every metric, the computation formula, data source, and reference. For example, spatial enclosure was derived through a combined GIS-and-field procedure. Building-footprint and parcel data (scale 1:1000) were obtained from the Macau Online Map and imported into ArcGIS Pro 3.1. For each public-space polygon, the Boundary-Facing Length—the cumulative length of built or vegetated edges situated ≤ 3 m from the space boundary—was extracted with the Near analysis tool. These lengths were then verified in situ with a Leica Disto D510 laser range-finder (mean absolute error = 0.07 m). Spatial enclosure was measured by the ratio of the enclosing boundary length to the perimeter of the space, producing the following values: 44.44%, 37.22%, 7.96% for the three Chinese-style squares; 59.10%, 84.01%, 27.80% for the Portuguese-style squares; and 32.32%, 66.44%, 78.36% for the Sino-Portuguese hybrids. The nine-site range (7.96–84.01%) and overall mean (48.6%) confirm a wide spectrum of enclosure conditions across Macau’s community spaces. These SE percentages, together with site area, paved-surface ratio, and visual-brightness uniformity, form the spatial-features block later entered into the hierarchical regression models.

3.3. Behavioral Observation

This study used systematic non-participant observation at all nine sites to document older-adult engagement. The protocol was informed by the SOPARC system (System for Observing Play and Recreation in Communities) developed by McKenzie [71], and further structured according to Gehl’s [72] recommendations for multi-day, multi-period behavioral scans in urban settings.

Preliminary counts in two test plazas revealed five elder-activity peaks—early-morning exercise (08:30–10:30), late-morning errands (10:30–12:30), post-lunch respite (14:30–16:30), pre-dinner gathering (16:30–18:30) and evening strolls (19:30–21:30). Each site was therefore scanned in these five periods on eight non-consecutive days (four weekdays, four weekend days) randomly distributed across one month, producing 800 min of observation per site. In every period, the observer selected the first 20 min (e.g., 08:30–08:50) and performed a slow left-to-right sweep; an individual was recorded only when the focal action remained clearly visible for at least 30 s.

Actions were first classified as necessary, optional, or social. Necessary behaviors (e.g., hurried shortcutting) were logged for completeness but, following Gehl, omitted from later statistics because they are weakly design-sensitive. Optional behaviors were subdivided into static (resting, reading, listening to music, viewing scenery) and dynamic (walking, fitness-walking, tai-chi, use of exercise equipment, gardening, dog-walking) sub-types. Social behaviors comprised any interaction involving two or more persons—conversation, child-minding, chess or card games, group opera, community volunteering, family gatherings, and the like. When multiple actions occurred, the behavior occupying the longest share of the scan was logged to avoid double-counting. Two trained coders double-scored 12% of the video-assisted scans (Cohen’s κ = 0.89); one coder then completed the remaining files.

The number of older adults, their proportion relative to total users, and the frequency of each activity type were systematically recorded. Contextual factors, such as weather conditions or special events, were also documented to control environmental influences. Observations conducted under poor weather conditions were excluded and rescheduled.

3.4. Human-Factor Data Collection

To explore how older adults respond visually and emotionally to community public spaces, this study employed an integrated approach using eye-tracking and GSR data for 63 older adults. Visual stimuli consisted of nine high-resolution images depicting typical public spaces in Macau. These images were selected for their cultural representativeness, environmental complexity, and relevance to the daily experiences of adults aged 65 and over. Photos were captured at natural eye-height (1.55 m) under consistent daylight conditions (12–14 klx), ensuring standardization across all scenes. Each image measured 4500 × 2532 dpi (16:9 aspect ratio) and was georeferenced to reflect actual locations. Static images were chosen as visual stimuli due to their high experimental controllability and proven validity in simulating real-world visual experiences. Henderson [73] showed that static scenes elicit naturalistic gaze patterns, while Berto [74] confirmed their effectiveness in triggering environmental responses. Compared to interactive 3D models, images reduce older adults’ variability and physical exertion and allow precise control over visual input.

To ensure representativeness and cognitive resonance, the nine stimuli were drawn from three spatial types—open plazas, linear street-courts, and pocket gardens—within each of Macau’s major cultural morphologies: Chinese, Portuguese, and hybrid. Although the selected spaces vary in physical size, their inclusion was guided by the medium-scale characteristics typical of Macau’s largos, which are generally recognized as appropriate for the everyday use and mobility patterns of older adults. Candidate images were first assessed by three architectural experts using a 5-point representativeness scale, and only those reaching a content validity ratio (CVR) of ≥0.78 were retained. This selection was further validated by six local older adults who rated their familiarity with each scene (mean ≥4.0/5). To minimize saliency bias and control visual complexity, we used Vision Complexity v1.4 to compute edge density (target range: 2.4–3.0%) and Shannon entropy (target range: 6.1–6.4 bits); only images falling within these thresholds were included.

Areas of interest (AOIs) were defined within each image to categorize visual attention based on three physical dimensions of place attachment: (1) spatial attachment (e.g., building façades, sculptures), (2) nature attachment (e.g., vegetation, water features), and (3) age-friendly/accessibility features (e.g., seating, pathways, signage). AOIs were defined within each image to categorize visual attention based on three physical dimensions of place attachment: (1) spatial attachment (e.g., building façades, sculptures), (2) nature attachment (e.g., vegetation, water features), and (3) age-friendly/accessibility features (e.g., seating, pathways, signage). AOIs were manually delineated using Tobii Pro Lab software (Version 24.21) and were designed to cover approximately 80–90% of each image frame, ensuring visual balance across the three categories. The remaining 10–20% of each image was left as a buffer zone to preserve natural gaze behavior. Representative thumbnails overlaid with AOI boundaries are presented in Figure 2.

The experiment was conducted using the Tobii Pro Nano eye-tracker (Tobii, Stockholm, Sweden, 60 Hz sampling rate, 0.3° accuracy) and the Shimmer3 GSR+ sensor (Shimmer wearable Sensor Technology, Dublin, Ireland) to record physiological responses. Data collection occurred in familiar elder activity centers to maximize comfort and ecological relevance. Participants were seated 60–65 cm from a 19-inch monitor, with ambient luminance maintained at 400–500 lux (measured using a digital lux-meter) and A-weighted background noise kept below 40 dB to minimize sensory distractions. Prior to data collection, all the participants provided written informed consent and completed a 9-point calibration procedure using Tobii Pro Lab. Calibration was accepted when the root-mean-square gaze error was below 0.4°, and it was repeated whenever significant head movement occurred during the session. Each participant viewed the same sequence of nine visual stimuli, randomized using a Latin square design. Each stimulus block consisted of a 5 s title screen, a 30 s stimulus image, and a 60 s mid-gray wash-out screen to mitigate carry-over effects.

Synchronization between the Tobii Pro Nano and the Shimmer3 GSR+ was achieved through Tobii Pro Lab software, which ensured precise temporal alignment of eye-tracking and skin conductance data without requiring external hardware triggering. Gaze and GSR signals were recorded simultaneously throughout the experiment. A complete experimental session, including calibration and stimulus presentation, lasted approximately 30 min. The procedural workflow is illustrated in Figure 3. This dual-modality approach provided empirical insight into the multisensory mechanisms of elder place attachment, capturing both attentional engagement and physiological arousal in response to visual features of community public spaces.

3.5. Questionnaire Survey

The questionnaire captured perceptions and intentions to analyze psychological mechanisms of AiPA. Human-factor data collection and questionnaire administration shared the same 63 elders who had just completed the eye-tracking + GSR protocol, thereby ensuring a one-to-one correspondence between physiological and self-report measures. This consistent participant design enhances the complementarity and comparability of the data, allowing for a multidimensional understanding from both psychophysiological and cognitive–affective perspectives.

As for sampling and recruitment, community-center rosters drawn from the ten census tracts with the highest proportion of residents aged ≥65 years were first stratified by population density; within each stratum, sex (~50% women/men) and age-band quotas (65–74, 75–84, ≥85 years) were imposed. Face-to-face contact with 80 elders produced 68 individuals who satisfied the inclusion criteria (≥65 yrs; ≥5 yrs continuous local residency; MoCA-C ≥ 18; normal/-corrected vision). Sixty-three completed both the laboratory session and the questionnaire, giving an effective response rate of 78.8% (63/80). Demographic profile (analytic n = 63). The final sample comprised 33 women and 30 men (mean ± SD age = 72.6 ± 6.1 yrs). Educational attainment was primary or below (61.9%), secondary (31.7%), and tertiary (6.3%); monthly personal income fell into four brackets: <MOP 4000 (15.9%), 4000–9999 (42.9%), 10,000–19,999 (38.1%), and ≥20,000 (3.1%). Median local residency was 28 years, 62% were currently married, and 93.7% self-identified as ethnically Chinese (4.8% Macanese; 1.5% other).

The present questionnaire’s purpose serves as the primary source for evaluating subjective perceptions and behavioral expectations. It also underwent a pilot test and ethical review to ensure appropriateness and robust protection of participants’ personal data. The questionnaire consists of two major modules. The first module focuses on individual background and cognitive–affective responses, including demographic information, patterns of space use, cultural orientation, and perceptions related to aging in place. Notably, this module includes the study’s core dependent variable, Aging-in-Place Attachment, a sub-concept developed to enrich the classical theory of place attachment. The second module assessed place attachment and environmental intention toward nine space scenarios. Together, these tools generate subjective perceptual data that corresponds with the physiological responses captured in the experimental setting. The Place Attachment Scale is based on the Abbreviated Place Attachment Scale (APAS) proposed by Boley [75], which divides place attachment into two dimensions:

Place identity and place dependence, each measured by three concise items to ensure reliability while maintaining feasibility for older adults. The Environmental Behavioral Intention component is designed with reference to key theories in environmental psychology, such as Kaplan’s restorative theory and Ajzen’s theory of planned behavior, assessing participants’ perceived comfort, relaxation, social engagement, and intention to linger in the given spatial scenarios. All attitudinal statements employed 5-point Likert scales (1 = strongly disagree, 5 = strongly agree). The interviewers were formally trained and followed an assisted-completion protocol to minimize literacy barriers; any questionnaire with > 10% missing items was excluded (none within the final 63 cases). The complete English-language instrument has been moved to Appendix A. The questionnaire functions not merely as a supplementary dataset but as a pivotal and independent tool for capturing sociopsychological variables. Together with the physiological data, it establishes a multidimensional, integrative framework for analysis.

3.6. Semi-Structured Interviews

To illuminate the psychological mechanisms that underpin the quantitative indicators of aging-in-Place Attachment, a program of semi-structured interviews was carried out with 13 elders—approximately 20% of the analytic cohort (n = 63)—recruited through maximum-variation purposive sampling so that sex, age band, educational attainment, and cultural orientation were all represented. Each conversation was conducted in Cantonese in a quiet room of the respondent’s community center immediately after completion of the laboratory and survey sessions and lasted 30–40 min. The interviews were audio-recorded with informed consent, transcribed verbatim, translated into English, and analyzed in NVivo 12 by following Braun and Clarke’s thematic procedure [76]; inter-coder agreement reached κ = 0.82, and thematic saturation was evident after the twelfth transcript.

The interview guide comprised four thematic modules and thirteen open-ended prompts (

Table 2. Semi-structured interview guide.

Module	Code	Interview Prompt
Personal background & aging-in-place status	a	Length of residence and daily activity areas
	b	Intention to remain in Macao; sense of belonging
Spatial–emotional ties	c	Frequently used outdoor public spaces and reasons
	d	Emotional attachment to specific space types or cultural styles
	e	Attractive spatial attributes and fit with personal preferences
	f	Relative attachment to Chinese-style, Portuguese-style, or hybrid settings
Everyday behavioral patterns	g	Typical activities and their frequency
	h	Influence of design/features on activity choice
	i	Nature of social interactions and their emotional benefits
	j	Design features that could better support activities and attachment
Future expectations and recommendations	k	Satisfaction with current public-space provision
	l	Desired improvements to support daily life and emotional well-being
	m	Vision for age-friendly public spaces in Macao

). This structure ensured the coverage of personal background, spatial–emotional ties, everyday behavioral patterns, and future expectations while still allowing respondents to elaborate freely. Although only about 20% of the questionnaire participants took part in the interviews, and their responses are not broadly representative, their narratives offer valuable subjective insights. These perspectives provide potential explanations for selected quantitative findings and contribute to the theoretical understanding of aging in place.

3.7. Data Analysis

This study employed a multi-stage data analysis strategy to examine the explanatory power of various factors on older adults’ AiPA in the Macau public spaces. Data sources included structured spatial measurements, behavioral observations, questionnaire responses, and physiological data collected through eye-tracking and electrodermal activity (EDA). Prior to analysis, all datasets were cleaned, screened for outliers, and normalized using min–max scaling when appropriate. The questionnaire data underwent reliability and validity checks to ensure the robustness of the instrument.

To address the central research question—namely, how much additional variance each block of variables contributes to AiPA—we conducted hierarchical regression analysis in IBM SPSS Statistics (version 26; IBM Corp., Armonk, NY, USA). Variables were entered stepwise in four blocks to evaluate their incremental contribution to explaining AiPA. This approach allowed the assessment of how each factor group improved model fit (ΔR²) and explanatory power. Significance was set at p  <  0.05.

4. Results

4.1. Questionnaire Reliability and Validity Analysis

To ensure the robustness of key constructs—particularly place attachment and AiPA—the questionnaire underwent comprehensive reliability and validity assessments.

Reliability analysis was conducted using Cronbach’s alpha to evaluate the internal consistency of each multi-item scale. The results indicated that all core constructs exceeded the commonly accepted threshold of 0.70, demonstrating satisfactory reliability. Specifically, the aging-in-place attachment scale (α = 0.754), composed of three items reflecting residential continuity, environmental satisfaction, and community dependence, exhibited acceptable internal coherence. Similarly, high levels of reliability were observed for place identity (α = 0.915), place dependence (α = 0.908), place attachment (α = 0.934), and place behavioral intention (α = 0.921). Item-total correlations confirmed the contributions of individual items to overall scale reliability, with selective deletion analysis reinforcing the decision to retain theoretically meaningful items.

Validity analysis included content and construct validation. Content validity was established using well-established scales adapted to the local context. Construct validity was tested via Principal Component Analysis (PCA) with Varimax rotation. For the aging-in-place attachment scale, the Kaiser–Meyer–Olkin (KMO) value was 0.677 and Bartlett’s test was significant (p < 0.001), indicating sampling adequacy. A single factor was extracted (eigenvalue = 2.046), accounting for 68.19% of total variance, with all factor loadings exceeding 0.78—supporting a unidimensional structure. Similar procedures confirmed the factorial validity of the other latent constructs.

These results confirm that the questionnaire exhibits strong psychometric properties, thereby ensuring the reliability and construct validity required for subsequent statistical modeling.

4.2. Hierarchical Regression Analysis

Based on the preceding discussion and dataset structure, this method allowed for the stepwise entry of predictor blocks to assess the incremental explanatory power of each category. Specifically, physical environmental attributes were entered as Block 1 in the SPSS regression model, followed by spatial behavioral variables (Block 2), human-factor data (Block 3), and finally, place-related psychological variables (Block 4).

As shown in

Table 3. Summary of Hierarchical Regression Models for Aging-in-Place Attachment.

Model	R	R Square	Adjusted R Square	Std. Error of the Estimate	ΔR2	F Change	df1	df2	Sig. F Change
1	0.231 a	0.054	0.048	5.734	0.054	10.61	3	563	0
2	0.451 b	0.204	0.195	5.273	0.15	35.209	3	560	0
3	0.494 c	0.244	0.225	5.176	0.04	3.652	8	552	0
4	0.914 d	0.836	0.831	2.418	0.592	989.76	2	550	0

^a. Constant, age-friendliness and accessibility, spatial attachment, and nature attachment. ^b. Model 1 + social behavior, dynamic behavior, and Static Behavior. ^c. Model 2 + Avg. skin conductance, Avg. First Fixation Time (AOI), Avg. Visit Duration, Conductance Frequency, Total Fixation Duration (AOI), Avg. Visit Count, Avg. Fixation Count (AOI), and Avg. Total Visit Time. ^d. Model 3 + place attachment, behavioral intention.

, the progressive inclusion of predictor blocks significantly improved the model’s explanatory power. Model 1, which included only physical environment indicators, yielded an R² of 0.054, suggesting that environmental factors alone explained only 5.4% of the variance in AiPA—a limited contribution. When behavioral variables were added in Model 2, R² increased markedly to 0.204 (ΔR² = 0.150), indicating the substantial explanatory power of behavioral engagement in public spaces. Model 3, which incorporated human-factor data, achieved a modest R² improvement to 0.244 (ΔR² = 0.040), suggesting that while biometric and behavioral metrics provided added value, their contribution was relatively smaller. Finally, Model 4 introduced place-psychological variables, resulting in a substantial R² increase to 0.836 (ΔR² = 0.592).

Further examination of standardized coefficients (see

Table 4. Coefficients of the final hierarchical regression model (Model 4).

Variable	Unstandardized Coeff. (B)	Std. Error	Standardized Coeff. (Beta)	t	Sig.	Tolerance	VIF
Constant	2.896	0.907	—	3.192	0.001	—	—
Spatial Attachment	0.424	0.142	0.069	2.995	0.003	0.563	1.776
Nature Attachment	−1.927	0.205	−0.454	−9.403	0	0.128	7.798
Age-Friendliness and Accessibility	−0.309	0.143	−0.063	−2.166	0.031	0.349	2.868
Static Behavior	−2.753	0.297	−0.41	−9.284	0	0.153	6.528
Dynamic Behavior	−1.674	0.239	−0.269	−6.999	0	0.203	4.934
Social Behavior	5.219	0.239	0.814	21.866	0	0.216	4.638
Total Fixation Duration (AOI)	0.164	0.043	0.096	3.852	0	0.479	2.086
Avg. Fixation Count (AOI)	−0.004	0.004	−0.03	−0.993	0.321	0.333	3.007
Avg. First Fixation Time (AOI)	−0.02	0.044	−0.008	−0.443	0.658	0.919	1.089
Avg. Total Visit Time	−0.673	0.226	−0.111	−2.983	0.003	0.217	4.601
Avg. Visit Duration	7.302	2.161	0.091	3.379	0.001	0.41	2.437
Avg. Visit Count	0.042	0.013	0.084	3.119	0.002	0.409	2.446
Avg. Skin Conductance	−0.057	0.069	−0.017	−0.824	0.41	0.701	1.427
Conductance Frequency	0	0.061	0	−0.007	0.994	0.821	1.218
Behavioral Intention	1.076	0.185	0.177	5.824	0	0.324	3.088
Place Attachment	0.723	0.032	0.67	22.604	0	0.34	2.938

) in Model 4 reveals nuanced insights. Among physical environment variables, spatial attachment (β = 0.069, p = 0.003) exhibited a weak but statistically significant positive effect, suggesting that familiarity and a sense of belonging may modestly enhance AiPA. In contrast, natural attachment showed a significant negative effect (β = −0.454, p < 0.001), implying that those with a stronger affinity toward natural environments may be less inclined to remain in urbanized public spaces. The accessibility and age-friendliness index also showed a small negative association (β = −0.063, p = 0.031), which may reflect a limited impact of infrastructural improvements on deeper emotional attachment.

With respect to behavioral variables, both static (β = −0.410, p < 0.001) and dynamic behavior indicators (β = −0.269, p < 0.001) were negatively associated with AiPA. This suggests that passive occupation of space or overemphasis on exercise-related activities may not contribute positively to emotional bonding with place, possibly due to reduced opportunities for interaction or the functional, transient nature of such engagements. In contrast, social behaviors (β = 0.814, p < 0.001) had the strongest positive effect, reaffirming the pivotal role of interpersonal interaction in the formation of place attachment in later life.

Human factors produced mixed effects. Total gaze duration within areas of interest (β = 0.096, p < 0.001), single visit duration (β = 0.091, p = 0.001), and frequency of visits (β = 0.084, p = 0.002) all showed significant positive associations with AiPA, indicating that visual attention and habitual presence reinforce emotional attachment. However, total dwell time (β = −0.111, p = 0.003) exhibited a negative effect, suggesting that mere physical presence without meaningful engagement may not foster attachment. Other biometric indicators, such as electrodermal activity and fixation counts, did not reach statistical significance (p > 0.05), pointing to limited direct predictive power in this context.

Most notably, place psychological variables had the strongest predictive effect. Place attachment (β = 0.670, p < 0.001) emerged as the most influential factor in the final model, reinforcing the centrality of emotional bonding in aging-in-place decisions. Behavioral intention toward continued engagement with the place (β = 0.177, p < 0.001) was also a significant predictor, suggesting that intentionality mediates the relationship between place experience and long-term settlement preferences. Together, these findings substantiate the theoretical model proposed in earlier chapters, affirming that AiPA in high-density urban environments is shaped by a layered interplay of physical, behavioral, biometric, and psychological factors, with place-based emotion serving as the most critical determinant.

The ANOVA results (

Table 5. ANOVA ^a results of hierarchical regression models.

Model		Sum of Squares	df	Mean Square	F	Sig.
1	Regression	1046.524	3	348.841	10.61	0.000 b
	Residual	18,510.1	563	32.878
	Total	19,556.63	566
2	Regression	3983.88	6	663.98	23.877	0.000 c
	Residual	15,572.75	560	27.808
	Total	19,556.63	566
3	Regression	4766.675	14	340.477	12.707	0.000 d
	Residual	14,789.95	552	26.793
	Total	19,556.63	566
4	Regression	16,340.81	16	1021.301	174.673	0.000 e
	Residual	3215.818	550	5.847
	Total	19,556.63	566

^a. Dependent variable: attachment to aging in place; ^b. Predictors: constant, age-friendliness and accessibility, spatial attachment, and Nature Affinity; ^c. Predictors: Model 1 + social behavior, dynamic behavior, and Static Behavior; ^d. Predictors: Model 2 + human-factor metrics; ^e. Predictors: Model 3 + place attachment and behavioral intention.

), generated using IBM SPSS Statistics Version 26.0, confirm the overall statistical significance of all four regression models (p < 0.001), with Model 4 achieving the highest F-value, further validating the robustness of the full model. These results underscore the necessity of integrating both objective and subjective dimensions in urban aging research and offer an empirical basis for designing emotionally resonant, socially vibrant, and behaviorally inclusive elder-friendly public spaces.

4.3. Conceptual Definition and Weighted Model of AiPA

To synthesize both the quantitative findings and qualitative insights, this study defines AiPA as a multi-dimensional construct shaped by four key domains—environment (E), behavior (B), human-factor (H), and psychology (P)—corresponding to four regression blocks that together explained 83% of its variance. This study represents AiPA using the following conceptual formula:

AiPA = α·E + β·B + γ·H + δ·P + ε(Q)

(1)

where

• E = physical environmental features;
• B = observed behavioral patterns;
• H = human-factor metrics;
• P = psychological and emotional variables;
• α, β, γ, δ = empirically derived weights from the model;
• Q = Qualitative narratives from semi-structured interviews;
• ε(Q) = A qualitative interpretive adjustment term reflecting lived experiences and symbolic meaning.

Based on the final hierarchical regression results(ΔR²): α = 0.054, β = 0.150, γ = 0.040, δ = 0.592; and the cumulative adjusted R² = 0.836. Thus, the normalized weighted model becomes as follows:

AiPA = 6.5%·E + 18.0%·B + 4.8%·H + 70.7%·P + ε(Q)

(2)

The normalized weights show psychological factors dominate (≈70.7%), highlighting the central role of emotional bonding in aging-in-place decisions. Behavioral engagement (B) accounts for a substantial secondary role (18%), while physical attributes (E) and human-factor cues (H) serve as foundational yet relatively modest contributors. Here, Q represents insights derived from semi-structured interviews, serving as a qualitative explanatory adjustment term that does not directly affect the statistical weights but enriches the model by revealing lived experiences, cultural meanings, and affective narratives, and supplements internal logic. This mixed-method formulation of AiPA provides a comprehensive framework to assess aging-in-place dynamics. It offers an empirically grounded, transdisciplinary conceptual tool to guide future evaluation and design of age-supportive urban environments. Having established the relative weights of each domain, we next interpret their theoretical and design implications.

5. Discussion

This study addresses three gaps raised in the introduction—(i) the lack of a metric that couples classic place-attachment theory with the explicit intention to age in place, (ii) the scarcity of multisource evidence that weighs environmental, behavioral, sensory and psychological drivers within a single model, and (iii) the near-absence of data from hyper-dense Asian streetscapes. By operationalizing the AiPA construct and analyzing five convergent data streams collected in Macau, we show that late-life attachment is dominated by psychological bonding (≈71% of explained variance), followed by social behavior (18%), human-factor engagement (5%), and the built environment (7%). These findings from Macau’s ultra-dense streets thus offer design guidance for other space-constrained cities seeking to foster successful aging in place.

5.1. Emotional and Social Dimensions as Core Drivers of AiPA

The results confirm that emotional bonds and social behaviors are key drivers of AiPA, aligning with place attachment theory [6,21]. In the hierarchical model, the inclusion of place attachment and related psychological variables yielded the largest increase in explanatory power, underscoring that an older person’s sense of belonging and identity with a locale is paramount in the decision to age in place. This aligns with longstanding observations that, in later life, familiar places become intertwined with one’s self-concept and life history [7]. Likewise, social engagement emerged as a critical ingredient: the frequency of sociable behaviors (e.g., conversing, group activities) had a highly positive association with AiPA (β ≈ 0.81, p < 0.001). This echoes prior studies, which show that strong neighborhood ties and community participation anchor older adults to their environment and enhance well-being [7,54]. In contrast, solitary or purely utilitarian use of space (e.g., sitting alone or exercising without social interaction) did not contribute to attachment; such static or non-social activities were negatively associated with AiPA. This observation reinforces theories that place attachment in later life is co-constructed through shared experiences and interactions rather than through passive habitation of the environment [54,77]. From an architectural perspective, these results highlight that the design and programming of public spaces should prioritize opportunities for social interaction and community-building among seniors. Seemingly ordinary community spaces, such as a small plaza, a park corner, or a temple forecourt, can serve as vital emotional refuges and “third places” for older adults. In Macau’s compact neighborhoods, such spaces function as extensions of the home environment, offering accessible venues for daily routines, face-to-face contact, and leisure. Over years of repeated use, these settings become imbued with personal memories and shared meaning. Consistent with the notion of public space as a community asset [78]. For practitioners in architecture and urban design, this finding underlines the importance of creating age-friendly communal spaces that are not only physically comfortable but also socially vibrant–places where older people can form friendships, sustain rituals, and reinforce their sense of identity and belonging.

5.2. The Facilitating Role of the Physical Environment

While emotional bonds and social life dominate AiPA, the physical environment remains a crucial facilitator in the background. Although environmental features alone explained just 5% of variance, they play an indirect yet enabling role in fostering attachment. A well-designed, familiar, and safe environment provides the stage upon which social and psychological processes play out. As our results suggest, spatial features had a modest positive effect on attachment, implying that legible and elder-friendly design can help older adults feel at ease and “at home.” This corresponds with the idea that environmental familiarity and comfort enable the confidence to engage socially, thereby nurturing attachment over time [7,50]. Participants who valued greenery reported lower AiPA, suggesting a mismatch between their preferences and Macau’s dense, built environment. This finding highlights a potential design gap: incorporating more nature or biophilic elements into urban community spaces could help satisfy the preferences of nature-oriented seniors and enhance their emotional connection to otherwise hardscaped settings. Macau’s dense neighborhoods offer limited greenery, so improving even small aspects might significantly improve emotional fulfillment for users who yearn for natural environments. This counterintuitive result may indicate that when basic amenities (ramps, seating, lighting, restrooms) are present and meet a minimum threshold, further enhancements in infrastructure do not automatically translate into stronger place bonds. Older people appreciate such features, but their emotional attachment depends more on what they do in space than on the checklists of physical provisions. However, it is important to note that without a baseline of good design–safety, comfort, and accessibility–many older adults would likely not use the space enough to develop any attachment at all. For instance, if a plaza has unsafe crossings or no seating, seniors may avoid it, forfeiting the chance to socialize and form memories there [50]. In this way, the physical environment plays a supportive, enabling role: it can encourage or discourage engagement. The findings suggest that in high-density cities, architects and planners should view environmental design as a facilitator of social interaction and psychological comfort. In summary, the built environment alone may not create attachment, but it sets the conditions for attachment to flourish by making spaces welcoming and usable for older people.

5.3. Sensory Engagement and Human-Factor Insights

By integrating human-factor metrics (eye-tracking and GSR data) into the analysis, this study offers a novel perspective on how older adults physically and emotionally engage with public space, an area rarely examined in traditional architecture and gerontology research. Although adding these sensory–cognitive indicators produced a moderate increase in explained variance (about 4%), the results are positive.

Longer gaze duration on salient features, higher visit frequency, and longer per-visit fixation all showed positive associations with AiPA, implying that active visual sampling, not merely being on-site, feeds place bonding. This echoes Kaplan’s restorative-environment theory that “soft fascination” fosters cognitive ties to the setting [79] and aligns with recent neuro-ergonomic work demonstrating that repeated attentional loops consolidate spatial memory [80]. The negative link between total dwell time (β = −0.11, p = 0.003) and AiPA becomes intelligible when read through the environmental press–competence model. Lawton & Nahemow posited that when environmental challenges exceed an older person’s adaptive capacity, prolonged immobility may signal passive coping rather than attachment; subsequent field studies of mobility limitation in later life support that view [81]. Webber likewise found that elders who “stay put” for long stretches in public plazas often do so because they feel unable—not unwilling—to circulate [82]. Our data suggests the same: if time on a bench is not punctuated by scanning, conversation, or micro-movement, it contributes little to emotional anchoring.

The physiological arousal measure (GSR) did not show a meaningful relationship with attachment. This aligns with findings that older adults often exhibit attenuated sympathetic arousal in familiar, low-threat settings [83]. Low GSR in our sample thus likely reflects a comfort–security state rather than disengagement, echoing socio-emotional-selectivity research showing that seniors prioritize calm, meaningful environments over novelty [84]. This insight cautions researchers that not all physiological signals of engagement are straightforward in the context of place bonding. Taken together, these results suggest a dual mechanism: focused visual attention cements memory traces, while low-arousal safety feelings consolidate attachment.

From a design standpoint, these human-factor findings highlight the importance of creating environments that can capture attention and invite repeated use. Features that draw the eye, such as natural elements, public art, or active street life, and that sustain interest over time, can encourage older visitors to mentally engage with space. Likewise, providing environments that reward exploration (e.g., varied pathways, points of interest, comfortable micro-spaces to pause) may lead to more frequent and longer visits, strengthening attachment. As architects and urban designers consider age-friendly spaces, multi-sensory design becomes crucial. Future research could build on this by combining physiological data with neurocognitive measures and in-depth qualitative interviews, examining how specific design elements evoke emotional responses and meaning for older users. This incorporation of human-factor data reinforces how older adults perceive and interact with the physical setting in real time is an integral piece of the attachment puzzle, one that merits further exploration in both research and practice.

5.4. Multidimensional Pathways and the Life-Course Perspective

AiPA emerges from the interaction of physical, social, and psychological factors rather than from any single domain. Instead, the results support an integrative view long suggested by environmental gerontologists and place theorists [21,58]. For example, while the physical environment provides opportunities or barriers, those translate into attachment largely through their influence on behavior and satisfaction. That the model explains over 80% of the variance supports the view that AiPA reflects a lifetime of accumulated place experiences: the built environment they inhabit, the social relationships they maintain there, and their internal reflections and meanings all feed into a holistic sense of place identity. This perspective is consonant with Rowles’ classic observations of place attachment as a layered phenomenon (physical, social, autobiographical) built over time, and with recent calls to examine aging in place through multiple lenses simultaneously [9]. One implication is that mediating mechanisms likely play a significant role. Prior studies using SEM have indeed found that factors like neighborhood satisfaction or perceived safety mediate the relationship between environmental qualities and well-being in older populations [23,54]. This study sees indirect evidence of such pathways: higher social activity levels coincided with greater attachment, and one can infer that an age-friendly environment facilitates those activities by making elders feel safe and comfortable to participate. Similarly, our human-factor data suggest that engagement and attention mediate the environment-attachment link; features that captured gaze and curiosity likely made the environment more enjoyable, fostering attachment beyond the physical amenities themselves. Adopting a life-course perspective, it is also important to recognize how personal history interweaves with these factors. Many of our participants had decades-long familiarity with their neighborhood, meaning that their attachment is also the product of temporal depth. The longer one lives in a place, the more opportunities there have been for the physical setting to serve as a backdrop for memories (children raised, festivals celebrated, hardships endured) and for the social network to solidify. Longer residence deepens attachment not only through behavior [7,20], but through the life narratives that build place identity over time. For urban scholars and designers, this multifaceted understanding means that promoting aging-in-place is not just a matter of modifying bricks-and-mortar; it requires supporting the social fabric and acknowledging the personal histories tied to a place. Interventions should be synergistic: physical improvements must go hand in hand with community programs and services that engage older residents, and both should resonate with the cultural and historical context that seniors value. The findings encourage moving beyond siloed approaches to aging-in-place research and practice, instead embracing models that link environment, behavior, and psychology into one continuum of experience.

5.5. Theoretical Contributions and Methodological Insights

By conceptualizing and empirically testing “Aging-in-Place Attachment,” this study contributes to theory in both place attachment research and environmental gerontology, particularly from an architectural vantage point. Classic place attachment theory has long recognized that people form deep bonds with places, but it has not explicitly focused on the intention to age in place as a distinct facet. This study extends the concept by highlighting how, in late life, attachment involves not just loving a place, but being committed to staying put in that place as one grows older. This adds a new dimension to place attachment that is closely tied to autonomy, security, and identity in old age. The results support theoretical arguments that familiar places provide a sense of continuity and control that is critical for “successful aging” [77]. This emphasis on continuity echoes observations by Wiles [1] and others that aging in place is valued because it allows elders to maintain a sense of self and normalcy through stable surroundings. Additionally, our findings nuance the concept of rootedness, the long-term embeddedness in a place. While scholars like Lewicka [7] have noted that length of residence often strengthens attachment, our analysis suggests that such rootedness is not a passive result of time alone. It is actively reinforced by ongoing engagement, the social ties one nurtures, and the personal meanings one continually ascribes to the environment. In other words, time in place contributes to attachment insofar as it is filled with meaningful relationships and experiences. This highlights a theoretical point: place attachment in late life is a dynamic, evolving process, not merely a residual of the past. The study provides empirical evidence in an urban public-space context for these ideas. The authors observed that when the environment supports older adults’ needs, it bolsters their sense of autonomy and attachment. Conversely, if an environment is inconvenient or hostile, it can undermine an elder’s confidence and desire to remain there, a phenomenon consistent with the environmental stress model [85]. By showing that supportive physical settings amplify competence yet must be complemented by informal, people-centered infrastructures to secure lasting bonds, AiPA provides a trans-disciplinary bridge linking environmental-press theory with contemporary scholarship on social infrastructure. As such, it furnishes planners with a diagnostic lens that targets emotional as well as physical outcomes of age-friendly interventions.

To complement the theoretical discussion, we offer several methodological reflections. Hierarchical regression was strategically chosen for its capacity to quantify the incremental explanatory power of grouped variables, revealing the distinct contributions of physical, behavioral, physiological, and psychological domains to AiPA. To verify the robustness of our hierarchical model, we performed several diagnostic tests, including multicollinearity checks (all VIF values < 3.1), residual normality assessments, and sensitivity analyses using bootstrap methods (1000 samples), confirming model stability and the reliability of findings. In addition, the block-wise hierarchy observed in our regression (physical ≈ 5% < human-factor ≈ 4% < behavior ≈ 15% < psychology ≈ 60%) should not be read as evidence that the first two domains are negligible. Rather, their comparatively low direct contributions imply that their influence is channeled through higher-order mechanisms. Conceptually, the built environment furnishes affordances that scaffold social interaction; those interactions, in turn, consolidate emotional bonds, while sensory engagement provides the perceptual glue that links experience to meaning. Put differently, Environment→Behavior→Perception→Psychology→AiPA may constitute a layered causal chain, with each tier amplifying or dampening the effect inherited from the previous one.

While hierarchical regression effectively demonstrates each domain’s relative contribution, it remains limited in its ability to clarify complex interrelationships among variables. Structural Equation Modeling (SEM), as a complementary analytical method, is particularly suited to addressing this limitation, as it simultaneously estimates multiple mediating and indirect pathways among latent constructs. In a preliminary SEM analysis from related ongoing research, we found initial evidence suggesting that physical environmental attributes influence AiPA primarily through behavioral and psychological mediators, rather than through direct effects alone. Specifically, behavioral engagement appears to play a nuanced mediating role, while psychological factors, especially place attachment, significantly mediate and strengthen these relationships. Although these SEM results are preliminary and are therefore not tabulated in the present article, they lend empirical support to the mediation logic implied by our regression hierarchy. Given the exploratory nature of these initial findings, although a full SEM analysis exceeds the current scope of this manuscript, future research employing comprehensive SEM frameworks is strongly recommended to rigorously test these mediation pathways and refine the theoretical understanding of AiPA in high-density urban environments.

5.6. Reflections from Qualitative Insights

Although the in-depth interviews in this study involved 12 older adults (approximately 20% of the survey sample), their narratives offer valuable qualitative insights that complement the quantitative findings. These subjective accounts highlight the emotional underpinnings of place attachment and aging-in-place experiences in Macau’s high-density public spaces. Respondents commonly described emotional bonds rooted in long-term residence, intergenerational ties, and stable social networks, indicating that attachment is shaped not only by the physical environment but also by accumulated life experiences and social embeddedness.

At the broadest level, the regression shows that psychological factors and social behavior dominate the variance in AiPA (β_place attachment = 0.67; β_social behavior = 0.81). Interviewees echoed this primacy: they repeatedly framed neighborhood plazas as “second living rooms” where long-standing acquaintances provide emotional security, and where “seeing familiar faces every morning” is perceived as a guarantee of continued autonomy. Thus, the qualitative testimonies reinforce the quantitative finding that social embed-ded-ness is the principal engine of AiPA. A second point of convergence concerns sensory engagement. Longer gaze duration within AOIs and higher visit counts were modest but significant predictors of attachment (β ≈ 0.09). Participants independently narrated that visual cues such as “hand-painted tiles” or “ancient banyans” anchor memories and invite repeat visitation. These statements give phenomenological substance to what the eye-tracking metrics register only in the abstract.

Not all linkages are confirmatory, however. They also supplied a counterview to one statistical result: while the model assigns a negative weight to nature-attachment (β = −0.45), interviewees repeatedly stressed that Macau and existing pocket parks are “too small, leaving their longing for everyday activities and greenery unsatisfied rather than absent, hence the negative coefficient reflects spatial scarcity, not a lack of ecological affinity. A similar nuance arises for age-friendly accessibility. Quantitatively, the index shows a small negative association with AiPA (β = −0.06), yet elders rarely complained about ramps or handrails in conversation. Instead, they lamented construction noise, crowding, and the loss of “resting niches.” The implication is that once basic accessibility thresholds are met, incremental hardware upgrades matter less than the experiential quality of the space—an insight that refines what the purely numerical coefficient can tell.

Taken together, these points illustrate a complementary relationship between methods: quantitative analysis identifies the structural weight of each domain, while qualitative narratives clarify the meanings, contingencies, and blind spots behind those weights. Where the two strands concur, confidence in the findings is strengthened; where they diverge, fresh hypotheses and design cues emerge. In this sense, the mixed-methods synthesis not only validates but also extends the explanatory reach of the study, offering a better understanding of how Macau’s elders come to— or decline to—age in place.

5.7. Limitations

The authors note that this cross-sectional, hierarchical regression approach only reveals direct associations. The model identifies key predictors of AiPA, but cannot test indirect paths, such as environmental influence via neighborhood satisfaction. From an architectural standpoint, it is important to note that apparent low effects in data may reflect methodological limitations or underreporting by participants, rather than a true lack of environmental influence [51,52]. Future research should therefore employ SEM or path analysis, as performed in other aging-in-place studies [54], to disentangle these chains of influence. The Macau-specific context limits generalizability, but future work could explore how digital features shape place attachment. Similarly, intergenerational dynamics and institutional support were outside the scope of our analysis but are likely important for holistic aging-friendly communities. Nevertheless, the mixed-methods, multisource data offer a rich, convergent picture of AiPA processes, paving the way for broader validation in diverse urban aging settings.

6. Conclusions

In high-density urban environments like Macau, the ability of older adults to age in place comfortably and meaningfully is less about any single physical feature and more about the synergy between people and their environment. This research introduced the Aging-in-Place Attachment to understand that synergy. By examining physical space, observed behaviors, sensory responses, and psychological bonds in concert, this study demonstrated that older adults’ attachment to place–and their commitment to continue living in their community–arises primarily from the emotional and social fulfillment they derive there. Place attachment and behavioral intention proved to be decisive factors in our empirical model, reflecting the deep human need for belonging and purpose. Social interactions and active engagement in community life were shown to greatly amplify these attachments. In contrast, physical environmental attributes played a more supportive role: they were necessary to enable engagement but, by themselves, accounted for only a small portion of the attachment motivation. Even human-factor cues mattered insofar as they facilitated those positive experiences and feelings. Together, these findings confirm that AiPA is a multilayered construct–one that is built on the foundation of a welcoming environment, but ultimately crowned by the edifice of social connection and emotional significance.

From an architectural and urban planning standpoint, the study provides evidence-based guidance for creating age-friendly cities. While an accessible, safe, and aesthetically pleasing physical environment is important, it is only the beginning. Equally crucial is designing and managing public spaces in ways that encourage social interaction, ongoing participation, and a sense of ownership among older residents. This could mean incorporating flexible spaces for community events, comfortable seating arrangements that facilitate conversation, elements of local culture and history that instill pride and identity, and green or tranquil spots that offer psychological relief in a dense cityscape. By nurturing these conditions, architects and planners can help forge the emotional bonds that truly anchor seniors in their communities. The concept of AiPA introduced here offers a practical framework for evaluating such interventions: success can be measured not just by usage statistics, but by whether older people feel more attached and committed to aging in their place as a result.

This study extended traditional place attachment theory into the gerontological realm, demonstrating that aging-in-place attachment can be empirically measured and modeled. In doing so, this study bridged disciplines, linking environmental psychology’s insights on emotional bonds, sociology’s emphasis on community interaction, and architecture’s focus on space and form. The high explanatory power of the integrated model (over 80%) suggests that comprehensive, interdisciplinary approaches are indeed capable of capturing the complex reality of aging in place better than siloed analyses. The findings reinforce the idea that successful aging in an urban context is a collaborative achievement between people and their environments. For policymakers, this means that efforts to help seniors age in place should be multifaceted, combining physical improvements with social programs and psychological support.

In conclusion, aging-in-place attachment encapsulates the idea that home for older adults extends beyond the four walls of their dwelling to the wider community milieu that gives them joy, security, and a sense of belonging. This study in Macau’s public spaces has empirically validated this concept and highlighted the predominance of human-centered factors in cultivating attachment. Moving forward, this study envisions that city designers and researchers will take up this mantle, exploring new ways to strengthen the emotional connection between older adults and their neighborhoods. As cities worldwide grapple with rapid population aging, adopting an AiPA lens can help ensure that urban environments are not just livable, but lovable for those in their later years. By investing in both the social hearts and the physical bones of communities, we can empower more elders to not only remain in place, but also to truly thrive in place for the rest of their lives.