M-health

To investigate the effects of simple dietary management and diet combined with exercise capacity-targeted intelligent lifestyle management on improving metabolic indicators in patients with overweight/obesity and metabolic syndrome.

The general goal is to design, develop and evaluate a personalized, self-guided and trans-diagnostic internet-based intervention to prevent anxiety and depression, based on predictive risk algorithms and decision support systems (DSS), in Spanish and Chilean adult population. Methods: We will conduct a three-arm parallel randomized controlled trial with one year of follow-up. A total of 2,595 depression- or/and anxiety-free participants (865 per group), aged 18-65, will be recruited and randomly assigned to one of two intervention groups or to the usual-care group (in a 1:1:1 ratio). Both interventions, Pandora-1 and Pandora-2, will be implemented via a smartphone application, the Pandora App. Pandora-1 is a self-guided and transdiagnostic intervention that includes 4 interactive intervention modules (move more, sleep better, improve relationships and emotional well-being), as well as predictive risk algorithms, decision support systems, and monitoring and feedback to implement personalized plans for the prevention of anxiety and depression. Pandora-2 is a psycho-educative intervention with predictive risk algorithms, minimally interactive and without personalization. The primary outcome is the combined rate of the onset of anxiety or depression (DSM-V diagnoses as measured by the CIDI interview) at 6 and 12 months. The secondary outcomes are the reduction of depressive (PHQ-9) and anxious (GAD-7) symptoms, risk of depression and anxiety (predictD and predictA risk algorithms), and the improvement of mental \& physical quality of life (SF-12), as well as acceptation and satisfaction with Pandora apps (u-MARS) and adverse effects (ad hoc questionnaire), which will be assessed at 1, 6 and 12 months. As mediators will be measured social support (Duke-UNC-2 items), physical activity (BPAQ-2), sleep (AIS-5), and repetitive negative thinking (PTQ-9) evaluated at 1, 6 and 12 months. We will use ActiGraph-GT3X accelerometers to assess physical activity and sleep at 1 and 6 months, in a subsample of 404 sedentary study participants.

In Taiwan, an estimated 2.3 million individuals have diabetes, with a 44% increase observed among young adults and adolescents. Poor dietary habits and sedentary lifestyles are major risk factors for type 2 diabetes. The widespread use of smartphones has facilitated the development of digital health technologies, including digital food photography and artificial intelligence (AI), which show promise for personalized nutrition care and health promotion. While such technologies have demonstrated short-term success in diabetes management, their long-term effectiveness remains uncertain. This study aims to evaluate the effectiveness of a digital eHealth care intervention for individuals with diabetes. Participants will be recruited from the Diabetes Shared Care Network and community care centers in Taiwan and followed for 12 months. Eligible participants will be randomly assigned by computer to either a control or an eHealth care group. • eHealth Group: Receives a 10-minute digital nutrition education session using the lab-developed "3D/AR MetaFood food portion education platform" (https://sketchfab.com/susanlab108/collections) and is required to submit weekly dietary records through food images using the "Formosa FoodAPP." Participants will receive immediate dietary feedback from nutritionists, followed by AI-generated personalized feedback on the glycemic index (GI) and glycemic load (GL) of their meals. They will also be provided with educational videos on healthy eating, physical activity, and selecting low-GI/GL foods. Anthropometric measurements and baseline questionnaires will be collected at enrollment. Blood biochemistry, including HbA1c, will be measured at baseline, and at 3, 6, 9, and 12 months. Collected food image data will be used to train AI systems for real-time dietary feedback and to explore the relationship between nutrient intake and long-term glycemic control.