Sleep traits, screen use, and emotional and behavioral problems: A short-term longitudinal cross-lagged panel network...

The longitudinal associations between multidimensional sleep traits, distinct types of screen time, and emotional and behavioral problems (EBPs) in adolescents remain inadequately understood, with traditional analytical methods struggling to capture their directional, symptom-level dynamics.

This longitudinal study included 4,800 Chinese adolescents (57.2% female; mean [SD] age, 15.84 [0.57] years) who completed baseline and 6-month follow-up surveys. Sleep traits, screen use, and EBPs were assessed using validated scales. A cross-lagged panel network (CLPN) model was employed to map directional predictive pathways, with centrality analyses (out-expected influence [OEI], in-expected influence [IEI], and bridge expected influence [BEI]) identifying core and bridge symptoms. Gender-specific subgroup networks were also estimated and compared.

EBPs demonstrated high temporal stability and dense internal connectivity, with "hyperactivity" (HA) and "emotional symptoms" (ES) serving as central nodes. Prosocial behaviors (PB) negatively predicted subsequent HA. Bridge pathways indicated that daytime dysfunction (DDF) and sleep disturbance (SDT) predicted subsequent ES and HA. Bidirectional associations were observed between HA and playing video games. Centrality analyses showed that ES, HA, DDF, and SDT were influential nodes with high OEIs and BEIs. Gender-specific networks were highly correlated (Pearson's r = 0.733).

ES and HA were central in the short-term longitudinal network linking sleep traits, screen use, and EBPs. Sleep-related problems, particularly DDF and SDT, emerged as key bridging symptoms connecting sleep and EBPs and may precede maladaptive screen use. These findings highlight potential targets for strategies for early identification and intervention to disrupt self-reinforcing cycles of adolescent EBPs.