Physical Activity
Conditions
Keywords
exercise, large language model, generative AI, artificial intelligence, physical activity, chatbot
Brief summary
This is a pilot RCT examining physical activity outcomes at 3 months across 3 conditions: use of large language model (LLM) fully self-directed; use of LLM with guidance; and use of online tools to help with sleep.
Detailed description
To prepare for a future fully powered RCT, the investigators are conducting a pilot RCT. The pilot RCT will enroll n=39 insufficiently active primary care patients, and randomize them to one of 3 conditions: LLM alone condition: participants are recommended to use a publicly available, general-purpose LLM for physical activity promotion over the next 3 months (n=13); LLM enhanced condition: participants are recommended to use a publicly available, general-purpose LLM for physical activity promotion over the next 3 months and additionally attend an initial session providing guidance on LLM use and weekly text message with suggestions on using the LLM (n=13); or Control condition: participants are recommended to use publicly available tools to help with sleep (websites with tips for sleep and guided meditation videos) over the next 3 months (n=13). At study onset, all participants will be given basic information about exercise safety. At 3 months, moderate-to-vigorous physical activity will be measured with a validated self-report measure. Feasibility and acceptability outcomes such as recruitment and retention--metrics essential for advancing to a subsequent large-scale trial-will be measured.
Interventions
BEHAVIORALLLM alone
Participants are recommended to use a publicly available, general-purpose LLM for physical activity promotion
BEHAVIORALLLM enhanced
Participants are recommended to use a publicly available, general-purpose LLM for physical activity promotion over the next 3 months and additionally attend an initial session providing guidance on LLM use and weekly text message with suggestions on using the LLM
BEHAVIORALSleep promotion
Participants are recommended to use publicly available tools to help with sleep (websites with tips for sleep and guided meditation videos)
Sponsors
University of Florida
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
PREVENTION
Masking
NONE
Eligibility
Sex/Gender
ALL
Age
18 Years to 75 Years
Healthy volunteers
No
Inclusion criteria
* Age 18-75 * Reports a desire to increase physical activity level * Not meeting minimum aerobic PA as indicated by L-CAT endorsement indicating level of activity below moderate (moderate defined as \>= 5x/week activity or higher) * Reports willingness to use LLMs to help with exercise goals. * Access to internet through phone data plan or home Wi-Fi.
Exclusion criteria
* Conditions warranting more extensive supervision of physical activity, including: * Any history of MI or stroke, arrhythmias, or heart failure * Uncontrolled hypertension * Type 1 or 2 diabetes * COPD or asthma * Recent orthopedic surgery or fracture * Advanced Parkinson's * Pregnant or planning to become pregnant in next 9 months * Active cancer undergoing treatment * Recent major surgery (\<12 weeks) * Use of assistive devices for ambulation * History of eating disorder diagnosis * Dementia diagnosis * Unable to read on own. * Hospitalization for mental health reasons in past year * Has used LLM to support physical activity goals in prior 3 months. * Currently enrolled in a structured program to promote physical activity or weight loss.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Recruitment feasibility | 6 months | Ability to meet recruitment goal of 39 participants within an acceptable timeframe |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Retention | 3 month | Ability to retain 80% of participants at 3 month assessment |
| Physical activity | 3 months | MET-minutes per week based on international physical activity questionnaire (IPAQ)-short form |
Contacts
CONTACTMegan A McVay, PhD
Outcome results
None listed