Short-video engagement, health-related behaviors, and mental well-being: Evidence from a longitudinal study of Chinese youth

The rapid growth of short-form video platforms presents both potential public health considerations and new avenues for social engagement, yet longitudinal evidence on how different patterns of short video use are linked to youth mental health remains limited. Using two waves of nationally representative data from the China Family Panel Studies, this study applies complementary analytical approaches: two-way fixed effects models (FE) to assess within-person associations, and cross-lagged structural equation modeling (SEM) to exploratorily examine temporal ordering and indirect pathways. FE estimates suggest that moderate engagement with short videos is linked to lower subsequent levels of depressive symptoms, whereas habitual daily use shows no direct within-person association. Exploratory SEM analyses further indicate patterns consistent with indirect pathways: moderate watching is correlated with higher physical activity, which in turn is linked to lower depression; meanwhile, daily watching is correlated with later sleep timing, which is associated with higher depressive symptoms. Smoking did not show a significant mediating role. These indirect effect analyses are constrained by the two-wave design and reliance on frequency-based measures, and should be interpreted as identifying plausible associations rather than establishing causal mediation. The findings underscore the importance of distinguishing patterns of use and the need for more nuanced measurement in research on youth digital engagement and mental health.