Assessing the Level and Determinants of Active Aging in China: An LDA-Based Topic Modeling Approach

While the concept of active aging has been extensively studied in high-income countries, China faces distinct demographic challenges, including a rapidly growing elderly population, accelerated aging, and aging prior to widespread economic prosperity. These trends highlight the urgent need for a context-specific conceptual and evaluative framework to measure active aging, tailored to China's socio-cultural and economic realities.

This study employs the Latent Dirichlet Allocation (LDA) topic model to construct a multidimensional indicator system for measuring active aging among older adults in China. Drawing on five waves of nationally representative panel data from the China Health and Retirement Longitudinal Study (CHARLS), spanning 2011 to 2020, we evaluate individual-level active aging scores using a quantitatively derived framework.

The measurement system consists of six core dimensions and 21 indicators: (1) physical health and functional capacity, (2) psychological well-being and life satisfaction, (3) family caregiving and social security, (4) economic security and intergenerational support, (5) social participation and enabling environments, and (6) lifelong learning and self-management. All scale-based measures demonstrated acceptable internal consistency (Cronbach's alpha ≥ 0.70). The average active aging score among the full sample was 0.4912±0.0907.

Active aging levels in China have shown consistent improvements over the observation period, with the most pronounced gains in the eastern region. The central region has seen a narrowing of differences, while the eastern, northeastern, and western parts of the country have seen a widening of differences. Key positive correlates of active aging include educational attainment, urban residence, male gender, alcohol consumption, and being married. Negative associations were found for older age, geographic region, presence of chronic conditions, number of surviving children, and smoking. Among these, education attainment, urban-rural status, age and gender emerged as the most influential factors.