Personalized Machine Learning Intervention to Improve Sleep Quality Using Wearable Technology in Healthy Middle-Aged Adults From Mexico City: Protocol for a Pilot Randomized Controlled Trial.

Background: In 2019, global sleep surveys reported that 80% of adults want to improve their sleep quality, and in 2021, 45% were reported to be dissatisfied with their sleep. In 2025, among American adults, 37% reported sleep dissatisfaction and 38% reported not feeling energized after sleep. These findings are consistent with data from the 2016 nationally representative survey of Mexican adults (aged ≥18 years), in which 37% reported sleep problems. Objective: This protocol describes a pilot randomized controlled trial (RCT) testing whether a single personalized sleep intervention driven by machine learning (ML) using consumer wearable data can improve sleep scores compared with generic sleep hygiene education in healthy middle-aged adults from Mexico City. Methods: This pilot RCT plans to enroll 32 participants (16 per arm; stratified by sex) in Mexico City. All participants wear Samsung Galaxy Watch 4 devices for 60 days. During days 1-30 (baseline), objective sleep data (10 variables related to duration, efficiency, sleep stages, movements, cycles, and recovery metrics) are collected. The control group receives generic sleep hygiene education. The experimental group receives personalized recommendations on day 30 based on the top predictive sleep parameters identified by ML models using Shapley Additive Explanations analysis and recursive feature elimination. The primary outcome is the sleep score (scale 1-100; composite device metric) during days 31-60, which is analyzed using analysis of covariance, with the baseline sleep score as a single covariate. The secondary outcome is the Pittsburgh Sleep Quality Index (PSQI) global score (scale 0-21; subjective validation), which is assessed at baseline, day 30, and day 60. Results: The study is currently in progress. Recruitment started in August 2024 and ended in July 2025. Data collection is expected to be completed by December 2025. The study will compare 960 nights from the control group with 960 nights from the experimental group to explore whether ML interventions can improve sleep scores using wearable technology, generate a dataset from objective data for iterative model training and analysis, correlate objective and subjective sleep quality metrics, and establish whether a feasible framework for proactive sleep quality approaches can be developed. The results will be available by March 2026, and the findings will be submitted for publication within 6 months of study completion. Conclusions: This pilot study establishes the feasibility and preliminary effect size for ML-personalized sleep interventions using consumer wearables within a manufacturer-independent personalization framework. The approach combines objective device monitoring with a subjective measure (PSQI) to test whether precision targeting of individual sleep parameters outperforms generic recommendations. If validated, this methodology could advance sleep interventions from universal protocols toward individualized behavioral targeting. The resulting dataset will enable model refinement and provide preliminary evidence for scaling personalized sleep health interventions in healthy populations.