Artificial Intelligence Across the Obesity Continuum: From Mechanistic Insights to Global Precision Prevention and Therapy

Obesity is a complex, rapidly escalating global health challenge that demands innovation across biology, clinical care, and public health. This review synthesizes evidence on artificial intelligence (AI) revolutionizing obesity research and management. In mechanistic discovery, AI techniques like deep neural networks and graph architectures integrate multi-omics, microbiome, and wearable-sensor data to elucidate metabolic signatures and gene-environment interactions. Clinically, AI enables predictive modeling of treatment response and supports adaptive trial designs. For pediatric obesity, machine learning facilitates early risk detection and personalized digital therapeutics, enhanced by privacy-preserving methods like federated learning. At the population level, spatial analytics and multi-omics modeling uncover environmental drivers, informing precision public health initiatives. The trustworthy deployment of these technologies hinges on cross-cutting imperatives: explainability, fairness, and data-quality assurance. The review compares key AI methodologies-from classical machine learning to large language models and causal inference frameworks-while addressing associated ethical and infrastructural challenges. It proposes a phased road map for equitable integration, positioning AI as a unifying framework that bridges molecular insights, individualized interventions, and population-wide strategies for more effective and scalable obesity prevention and care.