Author(s): YANG, P., CHIU, Y.H., WANG, H.C., LIU, C.M., CHEN, S.L., CHUANG, Y.T., Institution: CHANG JUNG CHRISTIAN UNIVERSITY, Country: TAIWAN, Abstract-ID: 155

INTRODUCTION:
Sedentary behavior is prevalent among university faculty and staff, leading to adverse health outcomes. However, time constraints often hinder regular participation in exercise. This study evaluates the feasibility and effectiveness of a time-efficient, technology-integrated "8-Week Turn Up" intervention, highlighting its potential to address common barriers to health and wellness. The program utilizes an active-passive hybrid training device combined with Artificial Intelligence (AI) assistance to improve body composition in a workplace setting.
METHODS:
The study recruited university staff members to participate in an 8-week intervention (September to November 2025). The protocol required participants to complete two 15-minute sessions per week. Technological integration included: (1) AI-assisted computer vision models to extract skeleton node coordinates for motion trajectory verification while preserving privacy; and (2) smartwatches to record physical activity data. Anthropometric measurements, including skeletal muscle mass (SMM), fat mass (FM), and waist circumference (WC), were collected at baseline and at the post-intervention time point.
RESULTS:
The intervention demonstrated high feasibility with a 92.3% completion rate (24 out of 26 participants met the protocol criteria). Data from these 24 completers were analyzed. Although statistical significance for fat mass reduction was not reached (p = 0.08), a clear trend was observed, with a mean reduction of 0.62 kg and an effect size approaching medium (Cohen's d = -0.37). Participants also achieved a mean reduction in waist circumference of 0.73 cm, demonstrating clinical relevance given the minimal time commitment. Notably, skeletal muscle mass was well-preserved (p = 0.13), indicating that the AI-assisted training stimulus was sufficient to counteract the potential muscle loss associated with weight reduction.
CONCLUSION:
The integration of AI assistance with short-duration active-passive training offers a viable solution for workplace health promotion, evidenced by the high adherence rate. The findings suggest that this time-efficient intervention can effectively induce trends in fat loss and waist reduction while preserving muscle mass. The use of AI for trajectory tracking provides an objective method for verifying exercise adherence in unsupervised or semi-supervised settings.