Background: Just-in-time adaptive interventions (JITAIs) use real-time monitoring to deliver personalized support at optimal moments, demonstrating potential for improving lifestyle behaviors in weight management. Objective: This study provides an overview of how JITAIs have been used or developed for weight management in adults with excess body weight. Methods: This scoping review followed Arksey and O’Malley’s 5-step framework and the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) checklist to ensure methodological rigor. Eight electronic databases (PubMed, Cochrane Library, Embase, CINAHL, PsycINFO, IEEE Xplore, Scopus, and Web of Science) were searched from journal inception to November 13, 2024, along with gray literature and hand-searched references. Two independent reviewers conducted data extraction for all included studies. Descriptive statistics were used to summarize study characteristics, followed by a nonlinear, inductive qualitative content analysis of the extracted data to identify and synthesize recurring concepts and characteristics of JITAI-based weight management interventions. Results: Thirty-five studies on JITAIs for weight management were included, focusing on dietary behavior (25/35, 71.4%), physical activity (20/35, 57.1%), and self-weighing (17/35, 48.6%). Types of support included prompts (n=33), feedback (n=24), recommendations of coping strategies (n=7), and educational information (n=5). A total of 31.4% of studies used machine learning for decision-making, while the rest used rule-based algorithms. Retention rates varied from 74% to 100%, and compliance from 15.1% to 94.6%. Greater user engagement was associated with improved weight loss outcomes. Across interventions, significant improvements were observed in weight, waist circumference, BMI, and blood pressure, alongside increased physical activity, healthier dietary behaviors, and reductions in sedentary time. Conclusions: While JITAIs show potential for improving lifestyle habits by providing the right intervention at the right time and in the right setting, most studies lacked theoretical grounding and were not conceptualized as JITAIs. Furthermore, terminology and reporting were inconsistent, which hindered evaluation and comparison across studies. Nevertheless, most studies incorporated varied distal and proximal outcomes, behavioral theories, intervention delivery methods, and data acquisition methods, and demonstrated positive outcomes in weight, physical activity, and dietary behaviors. This review demonstrates JITAIs’ potential in weight management but highlights the field’s early stage of development. Future research should focus on improving reporting standards, optimizing JITAI components such as the integration of behavioral theories and machine learning, and enhancing user engagement and long-term effectiveness by incorporating passive sensing, personalization, and adaptive feedback mechanisms. Trial Registration:
Learning Evolving Latent Strategies for Multi-Agent Language Systems without Model Fine-Tuning
arXiv:2512.20629v1 Announce Type: cross Abstract: This study proposes a multi-agent language framework that enables continual strategy evolution without fine-tuning the language model’s parameters. The core




