With the widespread use of smartphones among adolescents, smartphone addiction has become a growing mental health concern. Adolescents' limited self-regulation makes them particularly vulnerable to using smartphones to escape real-life stress, heightening addiction risk. However, the heterogeneity of addictive behaviors and the dynamic role of experiential avoidance have been underexplored. This 6-month longitudinal study surveyed 547 Chinese primary and secondary students using the Smartphone Addiction Scale (SAS) and the Acceptance and Action Questionnaire-II (AAQ-II). Latent profile analysis (LPA) and latent transition analysis (LTA) were applied to identify subgroups and examine transitions between these subgroups. Cross-lagged panel network analysis (CLPN) revealed key symptom interactions between experiential avoidance and addiction. The study identified two addiction subgroups: a stable "low-risk group" (84.9 percent) and a "high-risk group," 51.4 percent of whom transitioned to low risk over time. Logistic regression showed that experiential avoidance significantly predicted high-risk membership (odds ratios [OR] = 1.083-1.102) and deterioration within the low-risk group (OR = 1.036). The CLPN identified "online intimacy" (SPA-3) and "hesitation and overcautious" (EA-7) as driver nodes, with "withdrawal symptoms" (SPA-2) serving as a central node. These findings emphasize the crucial role of experiential avoidance in adolescent smartphone addiction and suggest symptom-level targets for early intervention. The results support acceptance and commitment therapy (ACT) as a promising approach for reducing smartphone addiction among youth.
Keywords: adolescents; cross-lagged panel network; experiential avoidance; latent profile analysis; latent transition analysis; smartphone addiction.