Hongwei Li, Xiaopan Xu, Cheng Mao, Qingqi Liu
The global popularity of short-form video applications such as TikTok has increased rapidly. However, longitudinal evidence on how distinct usage patterns affect mental health outcomes, particularly depression, remains limited. The present study investigated whether active versus passive TikTok use differentially predicts depression through the mediating roles of autistic traits and self-esteem. A three-wave longitudinal survey was conducted in China at six-month intervals. A total of 590 university students (aged 17-24 years) completed questionnaires assessing active/passive TikTok use and depression at Time 1 (T1), autistic traits and self-esteem at Time 2 (T2), and depression again at Time 3 (T3). Mediation analyses were performed using PROCESS macro (Model 6), controlling for gender, age, daily TikTok usage time, and T1 depression. Active TikTok use indirectly predicted lower depression through increased self-esteem, but not through autistic traits. In contrast, passive TikTok use predicted higher depression via three distinct pathways: through elevated autistic traits, through reduced self-esteem, and through the serial mediation of autistic traits leading to reduced self-esteem. Active and passive TikTok use exhibit divergent longitudinal associations with depression, mediated by distinct socio-cognitive mechanisms. These findings highlight the importance of distinguishing between usage modes when evaluating the mental health implications of short-video platforms.