Digital health solutions for chronic disease physical activity management: wearable devices, artificial intelligence, and public health implementation

Physical inactivity and sedentary behavior are major modifiable risk factors in chronic disease management, yet conventional clinic-based follow-up is poorly suited to capture real-world behavioral dynamics or deliver timely personalized support. Wearable devices provide longitudinal data on steps, activity intensity, sedentary time, sleep, heart rate, and selected physiological signals, while artificial intelligence (AI) may support feedback personalization, risk prediction, dynamic goal setting, and remote coordination. This review synthesizes evidence on wearable- and AI-supported physical activity management across diabetes, obesity, cardiovascular disease, chronic respiratory disease, cancer survivorship, and older-adult multimorbidity. Current evidence most consistently supports improvements in behavioral outcomes, including steps, physical activity levels, self-monitoring, and in some cases sedentary behavior. Evidence for functional and intermediate clinical outcomes is promising but heterogeneous, while evidence for long-term clinical endpoints, cost-effectiveness, and health-system integration remains less definitive. AI-specific evidence remains comparatively early, heterogeneous, and often feasibility-oriented, so claims about AI-enabled benefit require cautious interpretation. The review argues that wearable devices and AI should not replace clinical care, but should be understood as components of digital public health closed loops that connect continuous sensing, personalized behavioral support, clinical actionability, governance, and equity. Future research should move beyond isolated devices or apps toward validating explainable, actionable, equitable, and sustainable digital intervention pathways in real chronic disease populations and health systems.