Analysing Digital Engagement Patterns: A Machine Learning Investigation into Social Anxiety Among Adolescents with ADHD

Abstract

Background: The relationship between screen time and mental health among adolescents, particularly those identified as “digital natives”, has become increasingly pertinent, especially in the context of heightened digital engagement during the COVID-19 pandemic. This study aims to examine the associations between digital engagement patterns, ADHD severity, and anxiety symptoms in adolescents aged 11–18. Methods: A cross-sectional observational study was conducted with a sample of 97 adolescents diagnosed with ADHD. Participants completed validated questionnaires assessing daily digital activities and anxiety symptoms while caregivers provided additional contextual information. Data were analyzed using Random Forest regression to identify relationships between screen time, ADHD severity, and anxiety symptoms. Results: The findings indicate an average screen time of 6.6 h on weekdays and 8.1 h on weekends, with social media emerging as the primary activity, particularly among female participants. Notable gender differences were identified, with family dynamics influencing digital engagement; family disputes had a more significant impact on girls than boys. Furthermore, specific anxiety traits, i.e., physical symptoms and harm avoidance, significantly predicted time spent on social media across all genders considered in the study, while others, namely social anxiety and separation anxiety, were less valuable. Conclusions: This study highlights the complex interplay between screen time, ADHD, and anxiety symptoms in adolescents. These findings underscore the importance of tailored interventions that address the nuanced relationships between digital engagement and mental health in this population, warranting further research to develop effective management strategies.