WEARABLE DEVICES AND AI-DRIVEN PHYSICAL ACTIVITY MONITORING IN PUBLIC HEALTH: IMPLICATIONS FOR CHRONIC DISEASE PREVENTION AND HEALTH EQUITY – A NARRATIVE REVIEW

Background: The escalating global prevalence of non-communicable diseases necessitates proactive epidemiological strategies. Physical inactivity is a primary, modifiable risk factor in the pathogenesis of these chronic conditions. Consequently, digital epidemiology has attracted profound interest, particularly through wearable devices integrated with artificial intelligence (AI), which facilitate the continuous, real-time monitoring of physiological markers. Aim: This study aims to synthesize current evidence on AI-driven wearable devices in public health, evaluating their biological and behavioral mechanisms in chronic disease prevention, technical modalities, clinical applications, and implications for global health equity. Methods: A narrative literature review was conducted utilizing PubMed, Scopus, and Web of Science, focusing on peer-reviewed studies published between 2010 and 2026. Eligible publications included observational studies, algorithm validation research, and randomized controlled trials evaluating machine learning and physical activity monitoring. Results: AI-driven wearables effectively promote disease prevention via early anomaly detection and personalized behavioral interventions. Major modalities - including accelerometry, photoplethysmography (PPG), and continuous glucose monitoring (CGM) - powered by deep learning, demonstrate substantial clinical benefits in managing cardiovascular and metabolic disorders. However, critical limitations remain, including algorithmic bias, data privacy concerns, and the exacerbation of the digital divide. Conclusions: Wearable devices and AI analytics are highly promising adjuncts in public health surveillance. To realize their full preventative potential, future research and policy must prioritize protocol standardization, algorithmic fairness, and inclusive design to ensure equitable accessibility across diverse demographic groups.