KMAN Publication Inc. (KMANPUB) Journal of Adolescent and Youth Psychological Studies (JAYPS) 2981-2526 7 Serial Number 45 2026 05 01 Bayesian Network Modeling of Youth Substance Use Risk Based on Trauma Exposure, Reward Sensitivity, Peer Pressure, Sleep Instability, and Psychological Inflexibility Bayesian Network Modeling of Youth Substance Use Risk Based on Trauma Exposure, Reward Sensitivity, Peer Pressure, Sleep Instability, and Psychological Inflexibility 1 14 10.61838/kman.jayps.5472 EN Journal Article 2025 09 10 <table> <tbody> <tr> <td> <p><strong>Objective:</strong> The present study aimed to develop and evaluate a Bayesian network model for predicting youth substance use risk based on trauma exposure, reward sensitivity, peer pressure, sleep instability, and psychological inflexibility among adolescents and emerging adults in the United States.</p> <p><strong>Methods and Materials: </strong>This study employed a cross-sectional predictive modeling design using Bayesian probabilistic network analysis. The statistical population consisted of adolescents and emerging adults aged 15 to 22 years from educational institutions in the United States, from whom 1,248 participants were selected through multistage stratified sampling. Data were collected using the Substance Use Risk Profile Scale, Childhood Trauma Questionnaire-Short Form, Behavioral Activation System Scale, Peer Pressure Inventory, Sleep Condition Indicator, and Acceptance and Action Questionnaire-II. Data analysis was conducted using SPSS version 29 and Python 3.11. Bayesian network structure learning was implemented using the Hill-Climbing optimization algorithm with Bayesian Information Criterion scoring. Predictive performance was evaluated using classification accuracy, precision, recall, F1-score, sensitivity, specificity, and area under the receiver operating characteristic curve.</p> <p><strong>Findings: </strong>The findings demonstrated significant positive relationships among trauma exposure, reward sensitivity, peer pressure, sleep instability, psychological inflexibility, and substance use risk. Psychological inflexibility showed the strongest correlation with substance use risk (r = .65, p &lt; .01), followed by trauma exposure (r = .61, p &lt; .01). The Bayesian network model demonstrated high predictive performance with a training accuracy of 89.41%, testing accuracy of 86.93%, and area under the curve value of 0.91. Sensitivity analysis indicated that psychological inflexibility represented the strongest predictor within the network structure, followed by trauma exposure, peer pressure, reward sensitivity, and sleep instability. The probabilistic network further revealed that trauma exposure functioned as a central upstream variable influencing psychological inflexibility, peer pressure susceptibility, and sleep instability pathways associated with elevated substance use vulnerability.</p> <p><strong>Conclusion: </strong>The findings suggest that youth substance use vulnerability is shaped by a multidimensional network of emotional, cognitive, behavioral, and social risk mechanisms. Bayesian network modeling provided a powerful computational framework for identifying conditional dependency pathways among trauma exposure, reward sensitivity, peer pressure, sleep instability, psychological inflexibility, and substance use risk. The results highlight the importance of trauma-informed and multidimensional prevention strategies targeting emotional regulation, sleep health, peer influence processes, and psychological flexibility in adolescents and emerging adults.</p> </td> </tr> </tbody> </table> https://journals.kmanpub.com/index.php/jayps/article/download/5472/9881