Guangzhou’s Elder Care Map Reveals a City Split Between Connected Cores and Stranded Peripheries

When a megacity ages, the question of who can reach elder care stops being a matter of how many nursing homes a district has and becomes a matter of how districts connect to one another. A new study of Guangzhou, China, published in Frontiers in Public Health, maps the city’s pension care services not as a collection of isolated administrative units but as a living network, and the picture it reveals is strikingly unequal. Three central districts form a tightly woven core, while eight peripheral districts hang loosely off the edges, dependent on that core for nearly everything.

The research team, led by Lianhua Liu and colleagues at Guangzhou Huashang College and Guangzhou International Economics College, chose Guangzhou for good reason. As of 2024, the city had more than 2.1 million registered residents aged 60 and older, nearly 20 percent of its registered population. But those older residents are spread unevenly across 11 districts: Yuexiu District counts 351,130 people over 60, while Nansha District has fewer than 95,000. In Liwan District, more than 31 percent of residents are over 60, compared with just over 13 percent in Huangpu, an 18-point gap that means neighboring districts face fundamentally different care demands.

The resources meant to meet those demands are just as unevenly distributed. Of the city’s 57,452 nursing home beds, Baiyun District holds 23.13 percent while Conghua District holds a mere 1.35 percent. To capture this imbalance rigorously, the researchers built a multidimensional evaluation framework covering three dimensions: institutional care capacity, home- and community-based care capacity, and medical support capacity. All indicators were standardized per 1,000 older residents so that districts with large elderly populations would not be unfairly penalized, and the entropy weighting method was used to assign objective weights to each indicator based on how much information it actually carried.

The results exposed a gap far wider than most planners would guess. Baiyun District scored 0.9772 on the comprehensive pension care development index, essentially a near-perfect combined score across all three dimensions, while Haizhu District scored just 0.0774, a range of nearly 0.90. Baiyun emerged as a comprehensive supply center, strong in institutions, community services, and medical support alike. Tianhe District ranked second with the city’s best medical support capacity at 0.8972, but its institutional care lagged, marking it as a medical-resource-driven district. Liwan excelled in institutional beds but showed weak medical backing, while Nansha and Conghua built impressive grassroots networks of community canteens and health stations despite thin institutional and medical resources.

But the study’s real innovation lies in what came next. The researchers recognized that in a megacity, care does not stop at district lines: older adults cross boundaries for institutional admission, medical referrals, and family visits, and resources flow through transportation networks. So they constructed an interdistrict transport impedance index using actual public transportation travel times drawn from Gaode Map, measuring the shortest combined bus, metro, and walking routes between the 11 district government seats. A modified gravity model then combined each district’s service capacity score with these travel-time costs to estimate the potential strength of care linkages between every pair of districts.

The transport data alone told a story of two cities. Tianhe District enjoys the best connectivity, with an average public transport time of just 65.9 minutes to other districts and the lowest impedance index at 8.1633. At the other extreme, Conghua and Zengcheng face average travel times of 136.6 and 141 minutes respectively, impedance values above 12.5 that translate into real barriers: longer trips for professional care, harder referrals, and fewer family visits. Accessibility, the study confirms, decays steadily from the center toward the southern, northern, and eastern periphery.

Feeding the linkage estimates into social network analysis revealed the architecture underneath. Of 110 possible directed connections between districts, 61 exceeded the strength threshold, yielding a network density of 0.5545. Baiyun District connected directly with all 10 other districts and posted a betweenness centrality of 31.167, meaning it sits on the shortest paths linking most other pairs of districts. Tianhe followed with a betweenness score of 13.167. Together these two districts account for over 90 percent of the network’s total bridging capacity, an extraordinary concentration that makes the entire system’s resilience hinge on just two nodes.

The core-periphery analysis made the hierarchy explicit. Liwan, Tianhe, and Baiyun form the network’s core, with an internal connection density of 1.000, meaning every core district links strongly to every other. Peripheral districts, by contrast, show an internal density of only 0.250, meaning they barely connect to each other at all and instead channel their ties toward the core. Haizhu District is the most isolated case, with just one strong connection to the rest of the network and centrality scores that leave it almost entirely dependent on core nodes. Notably, the study found that a district’s local resources and its network position do not always match: Yuexiu holds strong medical capacity but sits near the network’s edge, while resource-rich Baiyun dominates connectivity.

These findings carry a pointed message for urban policy: building more facilities in a district does not guarantee that its older residents can actually reach care. Spatial equity in aging megacities, the authors argue, is relational as much as territorial. It depends not only on where resources sit but on whether districts can connect to and use resources located elsewhere. Their prescriptions follow from the network structure itself: establish a standardized cross-district coordination mechanism covering bed availability, referrals, and follow-up, supported by a citywide digital platform; develop secondary service centers in Panyu, Huangpu, Conghua, and Nansha that play to each district’s strengths rather than replicating the core; and improve transport links and horizontal cooperation among peripheral districts, from shared beds to community meal services and emergency assistance.

The authors are candid about the study’s limits. The gravity model estimates potential connections based on capacity and travel times, not observed flows such as actual cross-district admissions or institutional collaborations, and the analysis captures a single moment rather than the network’s evolution over time. Future work, they suggest, could incorporate multi-year data and operational records from care institutions to track how these networks change and whether coordination mechanisms actually work. For now, the study offers something megacity planners rarely have: a map showing exactly where the threads of an aging city’s care system are woven tight, and where they are coming apart.

Subject of Research: Spatial network structure and equity of pension care services across districts in Guangzhou, China

Article Title: Structural differentiation and collaborative optimization of the spatial network of pension care service in megacity behemoths: a case study of Guangzhou, China

Article References: Liu, L., Lan, L., Li, Y., & Wang, Z. (2026). Structural differentiation and collaborative optimization of the spatial network of pension care service in megacity behemoths: a case study of Guangzhou, China. Frontiers in Public Health, 14, Article 1960926. https://doi.org/10.3389/fpubh.2026.1960926

Image Credits: AI Generated

DOI: 10.3389/fpubh.2026.1960926

Keywords: elder care, pension care services, Guangzhou, spatial network analysis, population aging, urban equity, gravity model, transport accessibility, core-periphery structure, public health policy, social network analysis, megacity governance