Multi-Modal Digital Interventions for Adolescent Weight Management

Adolescent obesity remains a global public health challenge with far-reaching implications for physical and psychological well-being. In Taiwan, the prevalence of overweight and obesity among adolescents aged 13-18 has exceeded 30%, mirroring trends observed worldwide. The COVID-19 pandemic has further exacerbated this issue, intensifying weight gain among youth due to disrupted routines, reduced physical activity, poor sleep hygiene, and increased consumption of energy-dense foods. Despite growing awareness and reported intentions to manage weight among adolescents, there remains a substantial gap between intention and sustained healthy behavior.

This study seeks to bridge this intention-behavior gap through the development and evaluation of a multi-modal digital intervention, grounded in the Theory of Planned Behavior (TPB). TPB posits that intention-the most proximal predictor of behavior-is influenced by three constructs: attitude, subjective norm, and perceived behavioral control. Each construct is shaped by a distinct set of beliefs, including behavioral, normative, and control beliefs. However, limited research has explored how these constructs interact to promote healthy weight control behaviors or inadvertently contribute to unhealthy practices such as meal skipping, purging, or misuse of diet pills.

Building on the PI's previous qualitative and quantitative studies validating TPB in Taiwanese adolescents with obesity, this project designs an intervention that targets these belief systems directly. The intervention integrates artificial intelligence-driven virtual coaching, virtual reality (VR)-based simulation scenarios, gamified applications, and real-time feedback and monitoring through the Virti digital platform. Each intervention module is specifically aligned with a TPB construct:

Module A (Attitude): Uses persuasive VR messages to emphasize benefits of healthy weight control (e.g., appearance, health improvement).

Module B (Subjective Norm): Reinforces perceived social support through peer modeling and family involvement.

Module C (Perceived Behavioral Control): Builds self-efficacy by offering coping strategies to overcome common barriers (e.g., dealing with food sharing, time constraints).

The intervention will be delivered biweekly over three months, with each session lasting approximately two hours. Participants will interact with the virtual nurse avatar, respond to scenario-based questions, and receive personalized feedback. Data collection will occur at baseline, one month, and three months post-intervention.

The randomized controlled trial will enroll adolescents aged 12-18 years with a BMI at or above the 85th percentile, recruited from junior and senior high schools in Taipei. Stratified randomization will ensure balanced group assignment across districts. The intervention group will receive the full TPB-based digital program, while the control group will receive standard health education.

To evaluate intervention efficacy, the study will utilize:

TPB-based questionnaires measuring belief constructs and intention,

A weight control behavior scale distinguishing healthy vs. unhealthy practices,

Standardized tools to assess body image (MBSRQ) and mental health (PHQ-A),

Anthropometric and body composition measurements.

Quantitative data will be analyzed using mixed-design ANOVA and structural equation modeling (SEM) to assess changes over time and the mediating role of psychological factors. Thematic analysis of qualitative data (collected via platform interaction and interviews) will further elucidate adolescents' experiences and the intervention's impact.

This research is expected to make a significant contribution by:

Unpacking the cognitive mechanisms that translate weight control intentions into specific behaviors,

Demonstrating how digital interventions can be effectively tailored using TPB constructs,

Offering an innovative, scalable, and theory-driven model for adolescent obesity intervention that prioritizes psychological safety, personalization, and user engagement.

Ultimately, this study aims to inform policy and practice by providing empirical evidence for the use of AI-enhanced, VR-integrated interventions to promote healthy lifestyle behaviors among adolescents-a critical population for early preventive efforts.