Chongqing Municipality (29°33′49″N, 106°32′48″E), one of four megacities directly governed by the Chinese government, provides a unique setting to study urbanization, aging, and the intersection of health, well-being, and sustainability. Covering 82,400 km²—the size of Austria—it combines densely populated urban cores with extensive rural areas. According to the 2020 Census, Chongqing had 32.09 million residents, with 69.5% (22.3 million) living in urban areas and 30.5% (9.8 million) living in rural regions (6). Of this population, 17.8% (5.7 million) were aged 60 or above, with rural districts such as Qijiang and Wanzhou exceeding 25% (33). The old-age dependency ratio reached 22.1% in 2023, reflecting mounting social pressures, particularly in rural communities where youth outmigration has left more than 40% of households with at least one older adult(s) resident (34). Chongqing’s mountainous terrain, fragmented urban growth along the Yangtze River, and dense urban morphology create distinctive mobility challenges for older adults. At the same time, the municipality has been designated a national pilot zone for senior care reform, implementing initiatives such as community-based elder-care centers, age-friendly neighborhood retrofits (e.g., barrier-free facilities, safe crossings), and integrated health–social care programs. These characteristics make Chongqing a highly policy-relevant case for examining environmental supports for healthy aging. However, because this study employed a cross-sectional design, we did not measure direct exposure to these reforms. Our findings could therefore be interpreted as context-specific associations between walkability attributes and PF, rather than as representative of all Chinese municipalities.

The research was conducted in 2023, utilizing data collected in the field by research teams from both Yibin University and Sichuan University of Science & Engineering in China. A stratified random sampling approach was adopted at the community level to ensure variation in age, gender, and residential environments. The study focused on analyzing social and built-environment factors influencing health outcomes for older people in Chongqing. The study involved 2,900 randomly selected participants, aged 60–85. We selected one senior citizen per household from the registration of all community-dwelling older adults in Chongqing. Ultimately, 982 individuals (33.8%) volunteered for primary research. Participants (346 people; 35.2%) agreed to an on-site examination held at various community centers throughout Chongqing City from April 3 to June 30, 2023. During these sessions, participants completed a self-administered, paper-based questionnaire to provide sociodemographic information and report depression symptoms. Additionally, trained research personnel assessed cognitive performance. Health evaluations, including physical functionality and body composition, were also conducted. Each on-site evaluation lasted approximately 2 h, and participants received a book voucher upon completion. All participants signed an informed consent form, and no minors were involved in this study. This study was approved by the Yibin University Institutional Ethics Committee (November 17, 2022). This research examined 346 individuals (197 women and 149 men) who completed a comprehensive on-site evaluation.

The differences in population characteristics were examined using an independent t-test or X² test. Multilinear regression was used to investigate the relationship between environmental walkability attributes and objectively measured physical performance. Several covariates were controlled for in this analysis, including age, living status, duration of residence, educational attainment, employment status, lower limb functional performance, depression, and cognitive function. These covariates were classified according to gender to avoid multicollinearity with four PF outcomes, an a priori decision based on well-established biological and social differences in PF and mobility between older men and women. Gender stratification is widely applied in gerontology and environmental health research to uncover heterogeneity obscured by pooled models. Since PA appears to mediate the relationship between built-environmental characteristics and physical performance, individuals incapable of PA were excluded from the study. Additionally, each environmental factor was independently analyzed within each model (without mutual adjustment due to collinearity) to determine its overall effect. Primary inferences were based on the locally constructed Composite Walkability Index. Walk Score, a third-party proxy, was considered contextual to compare directionality but not emphasized due to conceptual overlap and limited calibration in dense Chinese settings. Stata 18.0 (Stata Corporation, College Station, TX, United States) was used to perform the analysis, and the probability threshold of p < 0.05 was used to determine significance.

In this study, women constituted (56.94%) more participants than men (43.06%) (Table 1). Female participants had a higher average age (78.70 ± 5.94; mean ± SD) than their male counterparts (76.20 ± 5.50). Table displays a statistically significant difference (***p < 0.001). In terms of educational attainment, higher (%) of men held ≥ tertiary degrees (44.16%), while most women had below-tertiary education (74.50%), indicating a statistically significant difference (***p < 0.001). Lower educational attainment among women may reflect cohort differences in schooling access, potentially influencing health literacy and resource navigation. The distribution of living status revealed that most participants, regardless of gender (88.32 and 84.56% women and men, respectively), lived with others, whereas a smaller proportion (~10%) lived alone. Concerning employment status, over 75% of men were either employed or receiving a pension, while the largest % of women relied on pensions rather than employment. Regarding lower extremity functional pain, women reported pain more frequently, with nearly 50% of both genders indicating no pain; however, mild pain was more prevalent among women (41.62%). Lower-limb pain was recorded on a 1–4 scale (1 = none, 2 = mild, 3 = moderate, 4 = severe). The average length of residence was similar for both men and women, measuring from approximately 10 to 50 years with considerable variability. Longer residence duration may indicate greater environmental familiarity; it is adjusted for residence years to reduce bias from cumulative exposure differences. Depression scores showed similar mean values for both genders, with scores ranging from approximately 0 to 7. Cognitive function demonstrated comparable results for men and women, with mean scores ranging from approximately 25 to 30. Because depressive symptoms and cognitive performance may constrain outdoor mobility and balance tasks in older adults, we included both as covariates.

Figure 1 shows that walkability attributes significantly influence physical performance in older men, particularly concerning balance and mobility. Higher population density within a 1,000-m radius was positively associated with their balance, evidenced by a beta coefficient of 3.42 (95% CI: 0.43 to 6.41). There is evidence that denser residential areas promote postural stability. Similarly, availability of destinations within 1,000 m significantly enhanced one-legged stance performance, with a beta of 5.20 (95% CI: 2.19 to 8.22). This finding implies that proximity to essential locations benefits both mobility and stability. Intersection density within 500-m buffers showed one of the strongest effects on balance, yielding a beta value of 3.73 (95% CI: 0.35 to 7.12). Higher populations and intersection densities may entail frequent micro-adjustments during short trips (curbs, crowds, and frequent crossings), stimulating sensorimotor control and balance. This finding reinforces the notion that well-connected street networks are vital for improving lower-body function. Conversely, hand-grip strength was only moderately influenced by public transport accessibility, with a beta of 0.92 (95% CI: 0.14 to 2.00). There is some evidence to suggest that closer transit access may encourage more movement, contributing to better muscle function. However, gait speed and performance on the timed up-and-go test were not significantly associated with walkability measures in this sample, suggesting that these outcomes may depend on additional factors not captured in our study. Non-significant associations for gait speed and timed up-and-go activities need to be interpreted cautiously; our models did not include PA behavior, which could confound or mediate these outcomes. Moreover, the Walk Score—a composite measure of neighborhood walkability—was positively associated with balance, showing a beta coefficient of 3.37 (95% CI: −0.02 to 6.77). This result implies that residing in highly walkable areas supports functional stability. Compared with men’s stronger balance associations with density/connectivity, women’s estimates (below) show comparatively large associations for destination access and the composite index. Overall, these findings indicate that dense, well-connected environments with accessible destinations were associated with better balance and mobility outcomes among older men. While these associations are consistent with environments that may support fall prevention, our study did not measure falls directly, and causal inferences cannot be made. However, additional interventions, such as structured exercise programs, may be necessary to improve walking speed and body strength.

Figure 2 displays that walkability attributes significantly influence physical performance in older women, particularly concerning mobility, balance, and body strength. Higher population density within a 1,000-m radius was positively associated with hand-grip strength, exhibiting a beta value of 0.32 (95% CI: 0.06 to 0.58). This result suggests that dense residential areas encourage PA, enhancing body function. The availability of destinations within 500 m demonstrated a negative association with timed up-and-go performance, with an empirical value of −0.33 (95% CI: 0.00 to −0.66). For timed up-and-go performance, negative coefficients indicate faster (better) performance. This result indicates that having nearby destinations improves mobility and reduces walking difficulty. Similarly, availability of destinations within 1,000 m boosted balance, reflected in a beta coefficient of −0.51 (95% CI: 0.09 to −1.1).

It appears that providing easy access to amenities promotes movement, which enhances postural stability. One of the strongest positive effects on balance was found to be the density of intersections within 500 m, with a beta of 1.94 (95% CI: 1.32 to 2.55), which supports the notion that well-connected street networks enhance the dynamic stability of older women. The 1,000-m intersection density buffer also displayed a beta value of 1.54 (95% CI: 1.10 to 1.98), further reinforcing the idea that an increase in pedestrian-friendly intersections contributes to improved balance. Access to public transport had a moderate impact on hand-grip strength, with a beta of 0.37 (95% CI: 0.04 to 0.70). This result implies that older women living closer to transit stations may engage in more PA, enhancing their body function. Both hand-grip strength and balance improved in response to the Walk Score, a measure of neighborhood walkability, with beta values of 0.46 (95% CI: 0.13 to 0.79) and 0.47 (95% CI: 0.11 to 0.84), respectively. This evidence suggests that more walkable neighborhoods support muscular endurance and postural stability in older women. While men’s balance related more to density/connectivity, women’s PF, particularly timed up-and-go and balance, was more responsive to access to nearby destinations and the composite index. This suggests that neighborhood design prioritizes short-distance, essential service access, which may be especially beneficial for older women. Overall, these findings emphasize that highly connected, destination-rich, and walkable environments significantly enhance mobility and balance in older women, reducing fall risk and supporting independence. However, improvements in gait speed may necessitate additional targeted interventions, such as structured PA programs.

We acknowledge several limitations to our research. First, its cross-sectional design prevents causal inference, so the observed associations between walkability and PF could be interpreted as correlational rather than causal. Second, the study sample was relatively healthy, and individuals unable to engage in PA were excluded; this likely produced conservative estimates, as those with the greatest functional decline—who might benefit most from supportive environments—were underrepresented. Third, our analyses assumed linear associations between walkability and PF, yet threshold or nonlinear effects are plausible. Future research may apply more flexible approaches, such as spline regression or generalized additive models, to capture these potential patterns and identify optimal environmental thresholds for intervention. Fourth, several relevant pathways were not directly measured. In particular, PA—a widely recognized mediator of walkability-function relationships—was unavailable in our dataset, and other contextual factors such as greenery, perceived traffic safety, and incidental social engagement may also play significant roles. Lastly, although the use of 500 m and 1,000 m network buffers strengthened ecological validity by better approximating daily activity spaces, these may still not fully capture the complexity of older adults’ mobility environments. Despite these limitations, the use of objectively measured environmental exposures and PF outcomes adds robustness to our findings.

These results point to the need for gender-sensitive neighborhood designs. In dense urban settings, close access to essential services may support older women’s PF. In contrast, safe and well-connected pedestrian networks may be more relevant to men’s mobility and balance. Cities around the world are increasingly adopting intelligent technology and policy innovations to address these challenges. Several countries, including Denmark and Canada, are utilizing artificial intelligence (AI) and sensor-based monitoring to enhance pedestrian safety, while the Chinese government is experimenting with age-friendly urban design initiatives in selected areas. For example, Copenhagen has integrated pedestrian monitoring systems in older-dense districts to optimize signal timings for safe crossings. Integrating telehealth services into walkable environments can foster independent aging, thereby alleviating pressure on hospitals and eldercare facilities. Telehealth and ambient-assisted living technologies, when co-located within walkable neighborhoods, can synergistically support aging in place, consistent with WHO age-friendly city principles. The sustainable urban models of cities like Vienna and Copenhagen exemplify how the integration of green infrastructure, healthcare, and transportation can create resilient urban systems that promote both public health and environmental sustainability.