Despite accumulating research on the associations between TikTok use and mental health, few studies have differentiated among specific types of use or examined the directionality of these relationships. To address these gaps, we explored the longitudinal relationships between four types of TikTok use and key mental health indicators. A total of 715 adult TikTok users completed two surveys over a two-month interval, assessing passive viewing, participatory use, contributory use, and problematic use of TikTok, alongside depression, social anxiety, and life satisfaction. Prospective associations were analyzed using cross-lagged panel network analysis. Results revealed that influence predominantly flowed from mental health indicators to TikTok use behaviors. Specifically, problematic use was primarily driven by depressive symptoms and social anxiety. Participatory use was predicted by higher life satisfaction and negative self-view, but reduced by social anxiety. Life satisfaction also predicted contributory use. In contrast, passive viewing showed minimal unique links to mental health once other effects were considered. Findings were largely consistent whether the four TikTok use dimensions were modeled using composite scores or individual items, indicating the robustness of the results. Overall, TikTok use may reflect users' psychological states rather than directly causing mental health changes. Interventions aimed at promoting healthier engagement should focus not only on usage restrictions, but also on increasing users’ awareness of their psychological states while using the platform.