Network and Bayesian DAG Analysis of Healthy Lifestyle Beliefs, Educational Stress, and Family Relationships Among Adolescents.

BACKGROUND: Healthy lifestyle beliefs are important for adolescent health promotion, yet they are embedded within broader psychosocial contexts involving family relationships and educational stress. Although these domains are likely to interact, little is known about how their specific components are organized within a broader psychosocial system. This study examined the interrelationships among healthy lifestyle beliefs, educational stress, and family relationship quality among Bangladeshi adolescents using network analysis and directed acyclic graph (DAG) modeling. METHODS: Data from 501 college-attending adolescents in Bangladesh were analyzed using the Healthy Lifestyle Beliefs Scale, the Educational Stress Scale for Adolescents, and the Brief Family Relationship Scale. An undirected Gaussian graphical model was estimated to examine network structure, centrality, bridge centrality, predictability, and stability. Gender differences were assessed using the Network Comparison Test. Directed acyclic graph analyses were conducted at the subscale level to explore statistical dependency structures and edge orientations among domains. RESULTS: The strongest positive edge in the network was observed between family cohesion and family expressiveness (0.520), whereas the strongest negative edge was observed between family cohesion and family conflict (-0.232). Confidence in handling problems effectively emerged as the most central node (strength = 1.325). The strongest bridge node was belief in family support for achieving goals (0.522), followed by belief in achieving personal goals and family expressiveness. Network stability was good (CS-coefficient = 0.75). Gender comparison indicated a significant difference in network structure (p = 0.003), but not in global strength (p = 0.511). DAG analyses identified dependency structures linking family relationships, educational stress, and healthy lifestyle beliefs; edge orientations were interpreted as exploratory rather than as evidence of temporal or causal ordering. CONCLUSIONS: Healthy lifestyle beliefs, educational stress, and family relationship quality were interconnected within a broader psychosocial system among adolescents. Coping-related healthy lifestyle beliefs appear central within this system, while family relational processes and academic stress showed meaningful connections with health-related beliefs. These findings may inform future longitudinal and preventive research examining coping-related beliefs alongside family communication, support, and academic stress.