Qiu Junjun, Zhang Yuge, Jia Binbin, Li Danyang
Center for Aging and Health Sciences, The College of Sports Science and Technology of Wuhan Sports University, Wuhan, China.
Institute of Sports Industry, Kookmin University, Seoul, Republic of Korea.
Front Public Health. 2025 Jul 29;13:1630009. doi: 10.3389/fpubh.2025.1630009. eCollection 2025.
Population aging in China exhibits pronounced spatial heterogeneity, driven by complex interactions among demographic dynamics, economic development, healthcare infrastructure, environmental conditions, and lifestyle factors. Understanding which determinants exert the strongest-and most geographically variable-effects is critical for designing targeted healthy-aging policies. This study investigates the multiscale influences on the city-level aging rate in Hubei Province, comparing patterns in 2010 and 2020.
We applied Multiscale Geographically Weighted Regression (MGWR) to annual data for 17 cities in Hubei. Explanatory variables encompassed demographic indicators (birth rate, mortality rate), economic affluence (per-capita GDP), healthcare infrastructure indicators (quantity of health institutions and service enterprises), environmental measures (per capita urban park green space, centralized treatment rate of sewage treatment plants), and physical activity prevalence. MGWR's adaptive bandwidth selection enabled each predictor to operate at its optimal spatial scale, while model fit was assessed via AICc, adjusted R, and residual diagnostics.
In 2010, spatial variability in aging was dominated by economic (SD≈0.36) and healthcare disparities (SD≈0.31). By 2020, these disparities had largely converged, and demographic divergence-particularly heterogeneous birth-rate effects (SD≈0.42)-became the primary driver. Crucially, physical activity emerged as the most potent local accelerator of aging in 2020 (mean β≈-0.60, SD≈0.25), statistically significant in over half of cities, and operating at a fine spatial scale.
The temporal shift from structural inequality to demographic and lifestyle determinants underscores the evolving landscape of population aging. MGWR's multi-bandwidth approach revealed that physical-activity interventions must be tailored at the city level, while fertility and economic policies warrant regional coordination. These findings demonstrate MGWR's advantage over global or single-bandwidth models in capturing layered spatial processes. Future research should employ finer spatial units, longitudinal designs, and integrate psychosocial variables to further elucidate healthy-aging pathways.
中国的人口老龄化呈现出显著的空间异质性，这是由人口动态、经济发展、医疗基础设施、环境条件和生活方式因素之间的复杂相互作用驱动的。了解哪些决定因素产生最强且地理差异最大的影响，对于制定有针对性的健康老龄化政策至关重要。本研究调查了对湖北省城市层面老龄化率的多尺度影响，并比较了2010年和2020年的模式。
我们将多尺度地理加权回归（MGWR）应用于湖北省17个城市的年度数据。解释变量包括人口指标（出生率、死亡率）、经济富裕程度（人均GDP）、医疗基础设施指标（卫生机构和服务企业数量）、环境指标（人均城市公园绿地、污水处理厂集中处理率）以及体育活动普及率。MGWR的自适应带宽选择使每个预测变量能够在其最佳空间尺度上运行，同时通过AICc、调整后的R和残差诊断来评估模型拟合度。
2010年，老龄化的空间差异主要由经济（标准差≈0.36）和医疗差距（标准差≈0.31）主导。到2020年，这些差距在很大程度上趋于一致，而人口差异，特别是出生人口率的异质性影响（标准差≈0.42）成为主要驱动因素。至关重要的是，体育活动在2020年成为老龄化最有力的局部加速因素（平均β≈ -0.60，标准差≈0.25），在超过一半的城市中具有统计学意义，并在精细的空间尺度上发挥作用。
从结构不平等向人口和生活方式决定因素的时间转变凸显了人口老龄化格局的演变。MGWR的多带宽方法表明，体育活动干预必须在城市层面进行量身定制，而生育和经济政策需要区域协调。这些发现证明了MGWR在捕捉分层空间过程方面优于全局或单带宽模型。未来的研究应采用更精细的空间单元、纵向设计，并整合社会心理变量，以进一步阐明健康老龄化途径。
