Enhancing older adults' perceived convenience of internet-based health management and facilitating the shift from cognition to actual behavior are important pathways to promote the health of older adults and actively respond to population aging in the digital and intelligent era. However, little research has explored whether and how smart devices empower older adults in internet-based health management. This study develops an integrated cognition-behavior analytical framework to analyze the empowering effect of smart devices on internet-based health management among older adults and examines the mediating roles of internet usage ability and internet usage motivation in the relationship between smart devices and internet-based health management. Based on data drawn from the China Longitudinal Aging Social Survey (CLASS) conducted in 2018 and 2020, this study employs logistic regression analysis and structural equation modeling to test the research hypotheses proposed for the research questions, and further conducts a systematic analysis on the robustness of the results with Propensity Score Matching (PSM), the two-stage residual inclusion (2SRI) method, the Heckman two-step procedure and the replacement of the core explanatory variable.

Results indicate that smart devices significantly improve older adults' perceived convenience of internet-based health management and increase their likelihood of engaging in such behaviors. This conclusion remains robust across a series of rigorous tests. Based on cognitive and behavioral dimensions, older adults are classified into four types: active participants, passive participants, potential participants, and bystanders, with the results confirming clear transition pathways among these types. Specifically, smart devices enhance older adults' perceived convenience of internet-based health management by improving their information search ability and internal health motivation, thereby facilitating the conversion of bystanders into potential participants. Furthermore, smart devices promote the transition from potential participants to active participants by boosting their sense of social adaptation. However, software usage ability does not show a significant mediating effect, suggesting that some older adults with perceived convenience still fail to act due to their insufficient internet operational skills. In addition, substantial heterogeneity is found in the effects of different smart devices on older adults' internet-based health management, and the same device exerts distinctly different effects on their cognition versus actual behavior. For instance, smart wristbands and smart watches tend to encourage older adults to become potential participants, whereas all-in-one smart devices and smart cameras are more likely to facilitate their transition into active participants. These findings highlight the need for tailored interventions that facilitate older adults' adoption of digital technologies in line with their types of internet-based health management. Both community-based digital skills training and intergenerational digital support within families may help enhance older adults' software usage ability.

This study contributes to the literature by advancing the analytical focus from behavior alone to an integrated cognition-behavior perspective. Therefore, this shift clarifies the heterogeneous empowerment pathways of smart devices for older adults' internet-based health management. The findings provide robust empirical evidence for optimizing the supply of internet-based health services for older adults and promoting the development of age-friendly digital health systems.