Stable co-occurrence and model-dependent within-person prospective coupling between problematic smartphone use and depressive symptoms: A four-wave study among college students.

BACKGROUND: Problematic smartphone use (PSU) involves addiction-like impaired control and functional interference, but its longitudinal association with depressive symptoms may reflect stable between-person co-occurrence or dynamic within-person coupling. OBJECTIVE: Guided by I-PACE, we tested whether PSU and depressive symptoms were linked through stable individual differences, prospective within-person deviations, or both. METHODS: Chinese college students were surveyed in September and December 2025 and March and June 2026. Analyses retained 851 participants with at least two observed waves; 189 eligible students were first observed at T2. RI-CLPM, LCM-SR, missingness, and closed-cohort sensitivity analyses were conducted. RESULTS: The RI-CLPM attributed approximately 49 %-53 % of PSU variance and 57 %-61 % of depressive-symptom variance to stable between-person differences. The random intercepts were strongly correlated (r = 0.693, 95 % CI [.619, 0.767], p < 0.001). Stationarity constraints were not rejected, robust Δχ2(8) = 7.92, p = 0.441. The stationary PSU-to-depressive-symptom path was B = 0.045, SE = 0.018, p = 0.010; the mean standardized effect was β = 0.126, SE = 0.051, 95 % CI [.027, 0.225], p = 0.013. The reverse mean effect was β = 0.038, p = 0.457, and the standardized directional contrast was nonsignificant, Δβ = 0.088, 95 % CI [-0.024, 0.201], p = 0.125. The LCM-SR cross-lag was nonsignificant (B = 0.036, SE = 0.039, 95 % CI [-0.039, 0.112], p = 0.344). CONCLUSIONS: PSU and depressive symptoms showed strong stable co-occurrence. The modest forward association was model-dependent and did not establish a mechanism, causal effect, or statistically reliable directional difference.