Adequate sleep is crucial for mood regulation and injury prevention, particularly among high school female athletes. Despite its importance, insufficient sleep is prevalent in this demographic. This study explores the feasibility and preliminary effects of an AI-supported mobile health (mHealth) intervention designed to provide personalized sleep feedback based on daily sleep data.
Study Overview
The pilot randomized controlled trial involved 25 high school female soccer players from a single team. The intervention aimed to evaluate the technical and operational feasibility of delivering personalized sleep feedback generated by GPT-4o through a messaging platform. Participants were randomly assigned to either an intervention group, which received personalized feedback, or a control group, which received nonpersonalized sleep information.
The study was conducted over several weeks, with the baseline period spanning four weeks followed by a nine-week intervention phase. During this time, the intervention group received feedback based on their Fitbit sleep data, while the control group received general sleep information.
Results
All participants completed the trial, demonstrating a 100% success rate in message delivery and questionnaire completion. The intervention group showed a median of 26 valid Fitbit wear nights, compared to 22 in the control group. Importantly, the primary outcome of sleep efficiency favored the intervention group, with a statistically significant difference observed (Hodges-Lehmann between-group difference of 2%, 95% CI 0.50%-4%; P=.046; r=0.35).
Additional nominal differences were noted in secondary outcomes, including the Pittsburgh Sleep Quality Index (PSQI-J) total score and sleep difficulty score. Exploratory outcomes such as REM sleep duration also indicated potential benefits from the intervention.
What this study does not show
While the study indicates feasibility and preliminary efficacy, it does not establish long-term effects or health outcomes directly resulting from the AI-generated feedback.
Evidence scorecard
- Analytic validity: Established
- Clinical validity: Established
- Clinical utility: Partial
This evidence is based on a randomized controlled trial involving human participants.