The aging population presents significant challenges globally, particularly in China, where the proportion of older adults is projected to reach 33.9% by 2050. A critical concern is ensuring access to nutritious meals for this growing demographic. Community-based senior dining halls (CSDHs) offer a potential solution, providing affordable and accessible meals within local communities. However, many CSDHs in China face financial difficulties due to low dining volume. This study aims to understand the relationship between built environment characteristics (both external and internal) and CSDH dining volume to inform strategies for improving their financial sustainability and enhancing the well-being of older adults. The research questions focus on whether a convenient external environment increases the likelihood of older adults traveling to CSDHs and whether an attractive internal environment, conducive to social interaction, encourages their usage. The study addresses a gap in the literature by directly examining the link between built environment and CSDH usage, utilizing objective dining volume data instead of relying on potentially biased self-reported questionnaires.

Existing research highlights various factors influencing dining volume, including food quality and diversity, and socioeconomic factors like family income. The built environment's role, however, remains under-explored regarding CSDH usage. Previous studies have examined the impact of built environment on older adults' travel behaviors to other destinations like parks and restaurants, using the 3Ds framework (Density, Diversity, and Design). While these studies suggest a potential link, direct evidence regarding CSDHs is lacking. Previous studies on restaurant usage, community center visits, and park usage have also shown the impact of built environment on older adult's willingness to visit. However, studies often rely on self-reported data which could introduce bias.

This study focused on 50 CSDHs in Yuexiu District, Guangzhou, China, an area with a high concentration of older adults. Dining volume data for the second quarter of 2023 was obtained from the Yuexiu District Civil Affairs Bureau, providing objective measures of CSDH usage. External built environment factors were assessed using a Geographic Information System (GIS) with three buffer zones (400m, 600m, 800m) around each CSDH to account for older adults' walking distances. Factors included floor area ratio (FAR) as a measure of density, street density, number of bus stops, distance to the nearest subway station, and normalized difference vegetation index (NDVI) as design factors, and catering point of interest (POI) diversity. Internal built environment factors, assessed through on-site audits, included the availability of kitchens, dine-in seating, healthcare centers, number of entrances, floor level, accessibility of entrances, barrier-free handrails, number of tables and seats, availability of wash basins, water dispensers, clocks, TVs, and other amenities. Covariates such as older population density and food taste (based on public voting data) were also considered. Multiple linear regression models were used to analyze the association between dining volume (log-transformed) and the built environment factors, with separate models for each buffer zone. Multicollinearity was checked using variance inflation factor (VIF), and variables with VIF >4 were removed.

The adjusted R² values for the regression models ranged from 0.476 to 0.494, indicating a substantial proportion of variance in dining volume explained by the built environment factors. For external environment, lower FAR and shorter distances to subway stations were significantly and negatively associated with higher dining volume across buffer zones. For internal environment, the availability of dine-in seating, lower floor levels, and the presence of water dispensers were positively associated with higher dining volume. The number of seats also showed a significant positive association, with a noticeable increase in dining volume for CSDHs with 21-30 and more than 30 seats compared to those with less than 10. Further analysis showed a sharp increase in dining volume when FAR was below 6.00 and a significant drop when the distance to subway stations exceeded 500 meters. Similarly, CSDHs with over 30 seats showed a significant increase in dining volume (over 4000 meals on average).

The findings support the hypotheses that both external and internal built environments significantly influence CSDH usage. Lower FAR likely reflects older residential areas with a higher concentration of older adults, better walking conditions, and potentially better cognitive health among residents, all contributing to increased CSDH visits. Proximity to subway stations enhances accessibility, particularly crucial for older adults with mobility limitations. The positive association between internal factors (dine-in seating, number of seats, water dispensers) indicates the importance of creating a comfortable, welcoming, and socially interactive dining environment. The negative association of higher floor levels underscores the need for accessibility in CSDH location choices. These findings have significant implications for urban planning and CSDH design to improve accessibility and usage, ultimately promoting community-based senior care and addressing the 'difficulty in dining' issue.

This study demonstrates the significant impact of both external and internal built environment factors on CSDH dining volume in Guangzhou. Optimizing these factors can improve CSDH efficiency, reduce reliance on public funding, and enhance social interaction among older adults. Future research should consider incorporating individual factors of older adults, ambient environmental factors (lighting, noise), service quality, and marketing strategies to provide a more comprehensive understanding. Qualitative studies, involving surveys with older adults, would further validate the findings and provide valuable insights into their preferences and perceptions.

This study has some limitations. Individual factors of older adults (socioeconomic status, health conditions, etc.) were not considered due to data unavailability. Ambient environmental factors (lighting, noise) were also not included due to time and resource constraints. Specific CSDH factors such as service quality and marketing strategies were not assessed. Future research should address these limitations for a more complete understanding. The current study focused on objective measures to provide clear implications for urban planning, however future studies should include subjective perceptions from older adults.