AI‐Driven Personalized Recommendations and Real‐Time Monitoring for the Prevention and Management of Obesity: Clinical Trial Protocol for the Multicentre BETTER4All Personalized Intervention
Eva Karaglani, Christina Patmiou, Paris Kantaras, Ioanna Panagiota Kalafati, Maria Kafyra, Vasiliki Vavouraki, Marta Gaspar, Ana Rito, Louise Seconda, Anestis Dougkas, Julie‐Anne Nazare, Anna Ek, Ioannis Ioakeimidis, Alexis Kyriacou, Panagiota Veloudi, Constantinos Deltas, Vanessa Bullon‐Vela, Maira Bes‐Rastrollo, Paulina Krzywicka, Aleksandra Luszczynska, Dimitrios Aletras, Chrysa Episkopou, Anastasios Delopoulos, Gianna Karanasiou, Vera Stavroulaki, Rooholla Poursoleymani, Ioannis Papathanail, Stavroula Georgia Mougiakakou, Ioannis Roumeliotis, Aristotelis Ballas, Christos Diou, George Dedoussis, Yannis Manios,ABSTRACT
Background
Obesity arises from complex interactions among genetic, biological, behavioral, and environmental factors. Traditional “one‐size‐fits‐all” interventions often yield modest, short‐term outcomes, failing to account inter‐individual variability. Advances in artificial intelligence (AI), polygenic risk scores (PRS), and digital health technologies now enable personalized and adaptive lifestyle interventions.
Objective
This paper presents the design and evaluation framework of the BETTER4All personalized intervention, a multicentre randomized controlled trial within the EU‐funded BETTER4U project. The trial aims to integrate biological and behavioral data into tailored obesity prevention and management strategies.
Methods
The BETTER4All trial will be conducted across seven European countries, targeting 1022 adults with body mass index (BMI) ≥ 25 kg/m 2 . Participants will be randomized to an intervention group receiving AI and PRS‐based lifestyle recommendations via a smartphone application, wearable device, and structured counseling, or to a control group receiving standard hypocaloric diet and lifestyle advice. The primary outcome is change in BMI at 6 months. Secondary outcomes include weight‐loss maintenance at 1 year, metabolomic and gut microbiota biomarkers, lifestyle behaviors, and psychosocial parameters. Process evaluation incorporates measures of fidelity, acceptability, feasibility, and appropriateness, alongside adherence metrics from the digital tools.
Conclusion
The BETTER4All trial will provide evidence on the effectiveness and scalability of AI‐driven, biologically informed personalization in obesity care, offering a methodological framework for integrating multimodal digital data, from omics and biomarkers to behavior and lifestyle, contributing to the advancement of precision public health strategies.
Trial Registration: