Report / Paper metadata
A double-edged sword: unpacking the functional heterogeneity of mobile social media use and its differential associations with exercise identity and social-media-specific sedentary time among older adults
A survey of older social-media users in Shandong found different associations for different platform functions. The study concerns self-reported exercise identity and sitting during social-media use, not measured exercise changes or proof that an algorithm causes inactivity.
Who participated?
The analysis included 452 adults aged at least 60 who reported using mobile social media daily for at least 30 minutes over the previous three months. Recruitment combined purposive community-based selection and referrals through participants' networks. Non-users, nursing-home residents and hospitalized patients were excluded. This is a selected group of active users, not a probability sample of older adults across China.
What outcomes were measured?
Questionnaires measured platform functions, perceived value, immersive experience, exercise identity and social-media-specific sedentary time. Exercise identity concerns how participants see themselves and how they believe others recognize them as exercisers. It is not accelerometer-measured activity. The sitting measure concerns time associated with social-media use, not all daily sedentary behavior.
What patterns did the model report?
The overall direct associations of the combined function measure with either outcome were not statistically significant. Some separate functions differed: information retrieval, online guidance and social comparison were associated with stronger exercise identity. The combined video-and-personalization function was associated with weaker exercise identity and more reported sitting during social-media use. These function-level associations should not be turned into a claim that all social media helps or harms activity.
Do modeled pathways establish a mechanism?
The statistical model linked attention/interaction and recording to exercise identity through perceived value. The overall indirect pathway through immersive experience to sedentary time was not significant, although the video-and-personalization pathway was reported at a corrected significance boundary. A modeled indirect association in a one-time survey does not show that one event occurred first and caused another. The paper's stronger language about effects needs to be read alongside its explicit causal limits.
Could the direction run the other way?
People who already identify strongly as exercisers may select fitness guidance or information. People who already spend longer sitting with social media may encounter more video recommendations. Unmeasured health conditions, physical limitations and personality could influence both use and reported outcomes. Adjustment for demographic variables does not remove all such explanations.
What can be concluded?
The results suggest that lumping platform functions together can hide different reported patterns. They do not establish an effective intervention, actual exercise improvement or a causal risk from personalized video. All measures were self-reported, and the regional non-probability sample limits transfer to other populations. This original explanation uses the full paper and preserves the difference between cross-sectional modeling and a tested behavioral intervention.
- Kind
- Paper metadata
- Identifier
- PMC13642746