Overweight , Obesity
Conditions
Keywords
Artificial Intelligence, Just-in-Time Adaptive Intervention, Weight Management, Health Behavior Change
Brief summary
This pilot study evaluates the needs, feasibility, and preliminary effects of an artificial intelligence (AI) health coach-based just-in-time adaptive intervention for weight management in adults with overweight or obesity. Participants receive a wearable device and use a WeChat-based platform during the intervention period. The system collects wearable data and self-reported information, and provides timely behavior-change support related to physical activity, sedentary behavior, sleep, diet self-monitoring, and weight-management self-regulation. The AI health coach provides conversational support and personalized suggestions based on predefined intervention rules and participant inputs. The main purpose of this study is to assess whether this AI-supported intervention is feasible and acceptable for adults with overweight or obesity. The study also explores changes in weight-related outcomes, health behaviors, self-efficacy, sleep, and quality of life before and after the intervention.
Detailed description
Overweight and obesity are common chronic health problems that require sustained support for daily behavior change. Digital health interventions may help extend weight-management support into everyday life, but many existing programs rely on generic education, retrospective feedback, or burdensome manual self-monitoring. This study evaluates an AI health coach-based just-in-time adaptive intervention designed to provide timely and individualized support for weight-management behaviors. This is a single-arm pilot study conducted among adults with overweight or obesity. Participants use a wearable device and a WeChat-based intervention platform during the study period. The intervention combines passive wearable sensing, participant-reported dietary self-monitoring, rule-based just-in-time intervention triggers, and an AI conversational health coach. Intervention content focuses on physical activity, sedentary behavior, sleep-related routines, dietary self-monitoring, and self-regulation for weight management. The AI health coach provides conversational guidance, encouragement, and behavior-change suggestions. Intervention messages are generated or selected based on participant data and predefined rules, with the goal of delivering support at moments when participants may benefit from timely prompts or feedback. The study evaluates feasibility and acceptability indicators, including wearable use, participant engagement, dietary self-monitoring, and interaction with the AI health coach. Preliminary intervention effects are explored by comparing baseline and post-intervention measures, including weight-related outcomes, body composition, physical activity, sleep, eating-related self-efficacy, and quality of life. The findings will inform the refinement of AI-supported just-in-time adaptive interventions for future controlled trials in weight management.
Interventions
BEHAVIORALAI Health Coach-based Just-in-Time Adaptive Intervention
The intervention provided timely behavior-change support for weight management through an AI health coach delivered via a WeChat-based platform. The system used wearable device data and participant inputs to support physical activity, sedentary behavior reduction, sleep-related routines, dietary self-monitoring, and self-regulation. Participants received conversational guidance, encouragement, and personalized suggestions based on predefined intervention rules and participant data.
Sponsors
The Fourth Affiliated Hospital of Zhejiang University School of Medicine
Study design
Allocation
NA
Intervention model
SINGLE_GROUP
Primary purpose
TREATMENT
Masking
NONE
Intervention model description
Single-arm pilot study in which all participants received an AI health coach-based just-in-time adaptive intervention for weight management.
Eligibility
Sex/Gender
ALL
Age
18 Years to 60 Years
Healthy volunteers
No
Inclusion criteria
* Aged 18 to 60 years. * Body mass index of 24 kg/m2 or higher, meeting the Chinese criteria for overweight or obesity. * Able to use a smartphone, WeChat-based platform, and wearable device during the study period. * Willing to participate in the AI health coach-based just-in-time adaptive intervention and complete study assessments. * Provided written informed consent.
Exclusion criteria
* Pregnancy, lactation, or planning pregnancy during the study period. * Severe cardiovascular, cerebrovascular, hepatic, renal, endocrine, psychiatric, or other major diseases that may affect study participation or safety. * Medical conditions or medications that may substantially affect body weight or body composition. * Contraindications to physical activity or inability to complete the intervention procedures. * Current participation in another weight-management intervention or clinical study. * Inability to understand the study procedures or complete the required assessments.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Participant retention rate | From enrollment to post-intervention assessment, approximately 31 days | Retention rate was defined as the proportion of enrolled participants who completed the post-intervention assessment. |
| Valid wearable use rate | During the 31-day intervention period | Valid wearable use rate was defined as the proportion of intervention days with valid wearable data. A valid wearable day was defined as a day with at least 10 hours of wear time or at least 180 minutes of main sleep data. |
| Active engagement rate | During the 31-day intervention period | Active engagement rate was defined as the proportion of intervention days on which participants had at least one interaction with the AI health coach or at least one dietary self-monitoring record. |
| Acceptability of the AI health coach-based intervention | Post-intervention assessment, approximately 31 days | Acceptability was assessed using a post-intervention questionnaire evaluating participants' perceived usefulness, satisfaction, and willingness to continue using the AI health coach-based intervention. Higher scores indicate greater acceptability. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Change in body weight | Baseline and post-intervention assessment, approximately 31 days | Body weight was measured at baseline and post-intervention. The outcome was the change in body weight from baseline to post-intervention. |
| Change in body mass index | Baseline and post-intervention assessment, approximately 31 days | Body mass index was calculated from measured body weight and height. The outcome was the change in body mass index from baseline to post-intervention. |
| Change in body fat percentage | Baseline and post-intervention assessment, approximately 31 days | Body fat percentage was measured at baseline and post-intervention. The outcome was the change in body fat percentage from baseline to post-intervention. |
| Change in waist circumference | Baseline and post-intervention assessment, approximately 31 days | Waist circumference was measured at baseline and post-intervention. The outcome was the change in waist circumference from baseline to post-intervention. |
| Change in visceral fat level | Baseline and post-intervention assessment, approximately 31 days | Visceral fat level was measured at baseline and post-intervention. The outcome was the change in visceral fat level from baseline to post-intervention. |
| Change in physical activity | Baseline and post-intervention assessment, approximately 31 days | Physical activity was assessed using the International Physical Activity Questionnaire-Short Form. The outcome was the change in physical activity from baseline to post-intervention. |
| Change in sleep quality | Baseline and post-intervention assessment, approximately 31 days | Sleep quality was assessed using the Chinese version of the Pittsburgh Sleep Quality Index. The outcome was the change in sleep quality score from baseline to post-intervention. |
| Change in eating self-efficacy | Baseline and post-intervention assessment, approximately 31 days | Eating self-efficacy was assessed using the Chinese version of the Weight Efficacy Lifestyle Questionnaire-Short Form. The outcome was the change in eating self-efficacy score from baseline to post-intervention. |
| Change in health-related quality of life | Baseline and post-intervention assessment, approximately 31 days | Health-related quality of life was assessed using the EQ-5D-5L. The outcome was the change in health-related quality of life from baseline to post-intervention. |
Countries
China
Contacts
PRINCIPAL_INVESTIGATORJie Yu, Master
4th Affiliated Hospital, School of Medicine, Zhejiang University, China
Outcome results
None listed