The BETTER4U (Preventing obesity through Biologically and bEhaviorally Tailored inTERventions for you) project, funded by the European Union, aims to address obesity through biologically and behaviourally tailored interventions. Obesity is a major public health issue influenced by genetic, metabolic, and lifestyle factors. Despite current weight management interventions, many individuals face challenges due to these varied influences. The BETTER4U project seeks to improve weight management by incorporating artificial intelligence (AI) and polygenic risk scores (PRS) to personalize interventions. The goal is to test the effectiveness of these personalized interventions in improving weight loss compared to standard care, using advanced monitoring tools and AI models.

The BETTER4ALL personalized intervention is a multicentre, open-label, parallel-group randomized controlled trial (RCT) involving seven study sites across Cyprus, France, Greece, Poland, Portugal, Spain, and Sweden. A total of 1,022 participants with overweight or obesity (BMI ≥ 25 kg/m²), aged 18-65 years, will be enrolled. Participants will be randomized into two groups: an intervention group receiving personalized lifestyle recommendations based on AI and PRS, and a control group receiving standard care recommendations. The intervention will last six months, followed by a six-month follow-up assessment.

The intervention's key aspects include wearable devices and a mobile application to monitor participants' behaviour, including physical activity, sleep, and eating habits. The intervention also integrates genetic, metabolic, and environmental data to provide tailored recommendations for weight loss. Participants' outcomes will be assessed regarding BMI, weight loss maintenance, changes in clinical biomarkers, body composition, and other lifestyle parameters.

This RCT will provide valuable insights into the effectiveness of personalized weight management strategies. AI-driven personalized recommendations and real-time monitoring represent a significant shift from traditional, one-size-fits-all approaches. The results of this study could offer a more effective and sustainable model for obesity management, particularly by accounting for individual genetic predispositions and lifestyle factors. Furthermore, by evaluating the impact of the intervention on a wide range of health outcomes, including biomarkers and psychosocial factors, the study will provide a comprehensive understanding of how personalized interventions can improve overall health and weight management.

In addition to contributing to the scientific understanding of obesity and its management, this project has the potential to influence public health strategies, offering a more personalized, data-driven approach to obesity prevention and treatment. By integrating genetic, environmental, and lifestyle factors, the BETTER4U intervention could pave the way for future innovations in digital health and obesity management.

Obesity is a major global public health issue with significant physical, psychosocial, and financial implications, especially in vulnerable populations. It results from complex interactions between genetic, metabolic, lifestyle (e.g., diet, physical activity, sleep), psychological, and sociodemographic factors. Individual responses to weight management interventions vary considerably, necessitating personalized strategies to prevent and treat obesity. Data from clinical trials have demonstrated that response to weight loss treatments is considerably characterized by an inter-individual variation, resulting from a combination of multiple genetic and phenotypic factors, interacting in a non-linear manner. Thus, a tailor-made, evidence-based weight loss intervention is considered to be the optimal solution for obesity management. Modern technology and the use of artificial intelligence (AI) models provide the potential to individualize weight management by including more information about an individual before delivering a treatment plan, in a way that has not been possible before.

Recently, the European Commission (EC) funded the large-scale BETTER4U project (Preventing obesity through Biologically and bEhaviorally Tailored inTERventions for you) to improve weight management through a tailor-made intervention using AI. The project consortium consists of 28 partners across Europe, Israel, and Australia, including health and technology scientists, legal and communication experts, and representatives from the European Association for the Study of obesity, EASO.

The BETTER4U project started in November 2023 and will run until October 2027, having the potential to produce new knowledge on weight management in an adult population. In addition, the envisioned privacy-respecting BETTER4U platform will be capable for detailed behavioural monitoring, will provide an infrastructure for continued and widespread use of real-world data beyond the end of the project, using real-world data for understanding the contextual circumstances and health status, as a valuable clinical research and patient management tool.

The BETTER4U work plan is broken down into 9 work packages (WPs) and its methodology is orchestrated in 4 phases, namely: Phase 1: identification of all weight-gain related determinants; Phase 2: creation of an AI-based causal model; Phase 3: pilot and randomized controlled trials (RCT) for the creation of the BETTER4U integrated methodology; and Phase 4: communication and dissemination of the methodology.

The study will be implemented in seven intervention sites (Cyprus, France, Greece, Poland, Portugal, Spain, Sweden). The overall aim of this study is to assess the effectiveness of the personalized intervention on weight loss over six months, as well as evaluate its impact on other lifestyle parameters, clinical indices, and biomarkers, while also assessing the maintenance of weight loss over a follow-up period of another six months.

The intervention will be mainly delivered by dietitians/nutritionists and potentially other health professionals experienced in delivering interventions that target nutrition, physical activity, and other lifestyle parameters modification (called "implementers" hereinafter). In all cases, a certified dietitian/nutritionist should always supervise the intervention delivery. A centralized training will be provided by HUA (WP7 leader) in English to train study sites' representatives/research team on how to deliver the intervention (both for the intervention and the control groups). Then, using the train-the-trainer approach, each site's responsible representatives/research team will train in either English or the local language all the site implementers before initiating the recruitment of study participants. AUTH, UBERN, WINGS, and HUA technical partners will support the training on the use of the intervention tools. These include the BETTER4U App and Intervention Platform (WP6 tools), incorporating the causal AI model developed in WP5 and the monitoring tools that will allow the successful calculation of the BETTER4U Core Behaviour Indicators and Living Environment Indicators (BCBIs and LEIs) developed within WP6.

Overall, during the BETTER4All Personalized Intervention, there will be three types of data collected:

In-App data: these data will be the ones collected regularly via the BETTER4U App by the participants, either passively and unobtrusively (i.e. the BCBIs and LEIs) or actively by taking meal photos (e.g. via the Go Food functionality) and answering related questions on meal type (breakfast/ lunch/ dinner), food/ drink type/ content (fish, meat, pasta, alcohol etc.), as well as perception of appetite at the time of meal consumption. Details to be provided in SOP to describe use of the app and what the Go Food is, how to take photos, the questions accompanying this process, etc.

Online data reported through RedCap: these data will be the questionnaires answered by the participants online via RedCap at each visit, either on site during the visit or online for a period of ± one week from the visit.

Anthropometric and blood pressure measurements and biological samples collection: these data include measurements of body weight, waist circumference, body composition and blood pressure that will be conducted during the in-person visits, i.e. V2, V4, V5, as well as the blood and stool samples collection that will be repeated at V4. For all measurements, the same procedures as in V0 will be followed based on the relevant SOPs. Anthropometric and blood pressure measurements will be recorded on the Participant Evaluation Form.

Participants will also receive a structured counselling session, based on the evidence for psychosocial behavior change techniques, which may be characteristic of effective weight change interventions. The choice of the effective behavior change techniques was also based on Group Lifestyle Balance approach.

After the completion of the 6-month intervention period, the end-of-study visit (V4) will be scheduled. During this visit, participants will receive the final counselling session and will be asked to complete the end-of-study measurements and questionnaires. They will also be reminded that an additional follow-up visit after another 6 months will take place (i.e., V5 - 6-month follow-up visit). In the meantime, they will be advised to maintain the lifestyle changes that have been achieved, taking advantage of the counselling they received in the previous period, as well as to continue using the digital tools (wearable, App) and monitoring their lifestyle behaviours.

During the 6-month follow-up visit (V5), participants will return the wearables to the research team and will be asked to repeat the completion of questionnaires and measurements to evaluate the maintenance of potential improved behaviours and weight loss achieved during the intervention period.

After the end of the total 6-month intervention period for the last participant, each implementer will be asked to complete three questionnaires assessing the acceptability, feasibility, and appropriateness of the intervention. This will allow a first exploratory assessment of the three key implementation outcome measures, assessing stakeholders' perceptions of the intervention and the implementation strategies. The three questionnaires are the Acceptability of Intervention Measure (AIM), Intervention Appropriateness Measure (IAM), and Feasibility of Intervention Measure (FIM).

The personal smartphone Android devices of each participant will be used, and participants will be provided with access to the deployed mobile App. More specifically, the mobile device will have the following functions:

It will act as the equipment hub for the connection of the BETTER4U smartwatches, locally collecting the generated data through Bluetooth technology, and, whenever possible, locally calculating relevant behavioural indicators. The device will act as a data collection tool itself, used for the collection of IMU-sensor data (i.e., accelerometer) and geolocation data (e.g., used for urban and socioeconomic context analytics, assessment of users' mobility and physical activity). It will act as the means of self-reported data collection from the participant, with repeated, episodic questions being communicated and answered through the smartphone device.

It will act as the means of photograph taking of meal occasions (including foods, drinks, etc.).

Through the app, it will serve as a snapshot of the users' collected data, showcasing some of their key indicators that can be computed directly on the device.

The wearable devices (smartwatches) to be used are going to be bought off-the-shelf devices, which will act as a passive monitoring tool and collect data from the following data sources:

Raw Data: Accelerometer Data (3-channel, 25Hz sampling rate) Readily Available Data: Stress Data (1 sample/minute), Heart Rate Variability (1 sample/minute), Steps Count (1 sample/minute), Sleep Structure (sleep duration, sleep stage duration), Calories Burned (1 sample/minute), SpO2 (1 sample/minute).

The smartwatch device data will be synced frequently with the user's smartphone, utilizing the Bluetooth Low Energy (BLE) protocol. Then they will be transferred to the BETTER4U servers, where the derived data calculation will take place. Extensive signal processing and AI/ML algorithms are used to derive meaningful information from the acquired signals, such as:

• Analysing GPS trajectories (from smartphone devices) and/or accelerometer signals (from smartwatch devices) to detect periods of movements and periods of staying at a place.
• Analysing the accelerometer of smartwatch devices to detect food intake gestures.
• Analysing images of food to recognize food categories, estimate volume and nutritional information, leading to Derived Data calculation. Derived Data will be used to calculate BCBIs and LEIs.

The BETTER4U Intervention Platform will act as an interface that facilitates the interactions between implementers and the AI model, by operating conjointly with the back-end services developed in WP5, to produce individualized guidelines for the management and prevention of weight gain. Furthermore, the web application will act as a tool that enables implementers to have a closer look at the behaviour of study participants by providing a detailed breakdown of the extracted indicators and how they evolve over time; thus, allowing for an on-the-fly assessment of the suggested individualized intervention scheme and a measurement of compliance.

Inclusion Criteria:

• Age: Participant aged 18-65 years
• Body Mass Index (BMI): BMI ≥ 25 kg/m² (overweight or obesity)
• Willingness and ability to use wearable devices and an Android mobile application for the duration of the study
• Owning a smartphone device with Android operating system
• Eating utensil technique: dominant hand gestures that correspond to handling of food via a fork or a spoon
• Language proficiency: participant has the ability to read and understand the language in which the mobile app and study materials are provided
• Consent: participant is willing to provide informed consent to participate in the study
• Residency: participant is resident of the country
• Availability: participant is able to participate for the full three-week duration of the study and comply with the study protocol

Exclusion Criteria:

• Pregnant or breastfeeding or intending to get pregnant in the short-term
• Comorbidities which might affect inflammation levels (i.e. type 1 diabetes, uncontrolled type 2 diabetes, unstable cardiovascular disease, eating disorders, gastrointestinal disorders)
• Mental illness affecting cognitive and communication skills, lifestyle and/or dietary habits
• History of cancer within 5 years prior to intervention onset
• Physical Limitations: any physical or mental condition that would prevent the participant from using the wearable device or mobile application as intended
• Any other criterion which would deem the participant unsuitable, according to the investigator's impression
• Technical Incompatibility: Individual does not own a compatible smartphone or are unable to use the provided wearable devices for technical reasons
• Documented and/or self-reported rapid changes in body weight in the six months preceding intervention onset, attributed to a diagnosed medical condition
• Interventional drug treatment in a clinical trial in the period within six months prior to intervention onset, which might affect intervention impact
• Inability to adhere to the proposed diet regimens, due to medical (i.e. gluten intolerance, allergies, swallowing problems) or religious reasons
• Medication promoting weight loss
• Bariatric surgery in the 12 months preceding intervention onset
• Use of supplements promoting weight loss or sleep quality, mood enhancing supplements, and/or consumption of probiotics and medications that influence the microbiome in the last three months preceding intervention onset Participation in a different clinical trial protocol or participation in different programs aiming at weight loss during the six months preceding intervention onset