This study employed a cross-lagged panel network model to examine the longitudinal relationships between problems of sleep, internalizing and externalizing problems in adolescents.

This study gathered data at four different time points (T1, T2, T3, and T4) for students enrolled in Grades 7 and 8, with an interval of approximately six months between each time point. The present sample comprised 1,281 Chinese adolescents, including 636 girls, with a mean age of 12.73 years (SD = 0.68) at baseline. Cross-lagged panel network modeling was used to estimate longitudinal relationships between symptoms at adjacent time points. Network replicability was assessed by comparing the T1→T2 network with the T2→T3 network and the T2→T3 network with the T3→T4 network.

The anxious/depressed symptom emerged as the most predictive of other symptoms and were also the most prospectively influenced by other symptoms. Cross-cluster edges predominantly flowed from internalizing and externalizing symptoms to sleep problems. Additionally, externalizing symptoms exhibited distinct patterns: aggression predicted more sleep and internalizing symptoms, whereas delinquent behavior predicted fewer of these issues.

These findings suggest that mental health problems contribute to later sleep disturbances, with internalizing symptoms playing a central role in adolescent psychopathology.

Psychotic-like experiences (PLEs), especially for persistent PLEs, are highly predictive of subsequent mental health problems. Hence, it is crucial to explore the psychopathological associations underlying the occurrence and persistence of PLEs. This study aimed to explore the above issues through a longitudinal dynamic network approach among PLEs and psychological and psychosocial factors.

A total of 3,358 college students completed two waves of online survey (from Oct 2021 to Oct 2022). Socio-demographic information was collected at baseline, and PLEs, depressive and anxiety symptoms, and adverse life events were assessed in both waves. Cross-lagged panel network analyses were used to establish networks among individuals with baseline PLEs as well as those without.

At baseline, 455(13.5%) students were screened positive for PLEs. Distinct dynamic network structures were revealed among participants with baseline PLEs and those without. While ‘psychomotor disturbance’ had the strongest connection with PLEs in participants with baseline PLEs, ‘suicide/self-harm’ was most associated with PLEs in those without. Among all three subtypes of PLEs, bizarre experiences and persecutory ideation were the most affected nodes by other constructs in participants with baseline PLEs and those without, respectively. Additionally, wide interconnections within the PLEs construct existed only among participants without baseline PLEs.

The study provides time-variant associations between PLEs and depressive symptoms, anxiety symptoms, and adverse life events using network structures. These findings help to reveal the crucial markers of the occurrence and persistence of PLEs, and shed high light on future intervention aimed to prevent and relieve PLEs.