JMIR Publications

ABSTRACT

Background:

Adequate sleep supports mood regulation and injury prevention in high school female athletes; however, insufficient sleep is common in this population. Wearable devices enable continuous assessment of objective sleep metrics, but the impact of personalized feedback based on daily sleep data remains underexplored.

Objective:

This study evaluated the feasibility of an mHealth intervention that delivered personalized sleep feedback to high school female soccer players. A large language model generated feedback from objective Fitbit sleep metrics and delivered it via LINE on days with valid prior-night data. Sleep efficiency was the primary preliminary efficacy outcome. We explored additional sleep outcomes and examined mood state and sports injury severity as exploratory outcomes.

Methods:

We conducted a pilot randomized controlled trial as a closed-user-group study among 25 high school female soccer players from a single team. We completed recruitment, consent, and Fitbit setup in person. Weeks 1 to 4 constituted the baseline period. After a week-5 questionnaire, we stratified participants by median baseline total sleep time and the Japanese version of the Pittsburgh Sleep Quality Index score and randomly allocated them (1:1) to intervention (n=12) or control (n=13). During weeks 6 to 9, the intervention group received GPT-4o–generated feedback based on Fitbit metrics and individual baseline averages on days with valid prior-night data, whereas the control group received daily nonpersonalized sleep information. Feasibility outcomes included message delivery success rate, valid Fitbit wear nights, questionnaire completion rate, technical problems, and adverse events. Secondary and exploratory outcomes included objective and subjective sleep measures, mood state, and sports injury severity assessed using a modified Japanese version of the Oslo Sports Trauma Research Center Questionnaire on Health Problems.

Results:

All 25 participants completed the trial. Message delivery success was 100% in both groups. Median valid Fitbit wear nights were 26 (IQR 24-28) in the intervention group and 22 (IQR 15-26) in the control group. Questionnaire completion was 100% at both assessment points. No major technical problems, adverse events, or privacy breaches occurred. Change in sleep efficiency favored the intervention group (Hodges–Lehmann between-group difference 2.00%, 95% CI 0.50-4.00; P=.046; r=0.35). Nominal between-group differences also favored the intervention for rapid eye movement sleep duration, PSQI-J total score, Sleep Difficulty Score, and Total Mood Disturbance. We observed no between-group differences in total sleep time, wake after sleep onset, deep or light sleep duration, daytime sleepiness, or sports injury severity.

Conclusions:

The mHealth intervention was feasible in a high school female soccer team. Although the short-term intervention did not increase total sleep time, preliminary findings suggest improvements in sleep efficiency, subjective sleep quality, sleep difficulty, and mood state. Larger and longer trials are needed to evaluate effectiveness and injury-related outcomes. Clinical Trial: UMIN Clinical Trials Registry UMIN000061189; https://center6.umin.ac.jp/cgi-open-bin/ctr_e/ctr_view.cgi?recptno=R000070024