REAL-WORLD SLEEP TRACKING: COMPARING OBJECTIVE AND SELF-REPORTED SLEEP DATA

Sleep problems constitute a growing public health concern, tied to adverse physical and mental health consequences. Home-based sleep tracking can collect longitudinal, low‑cost data, yet the quality of data in real-life conditions remains insufficiently explored. This exploratory study evaluated the feasibility and validity of combining sleep protocol data with two distinct trackers: a wearable (Fitbit Charge 6, FBC) and a contactless sensor (Withings Sleep Analyzer, WSA). Over the span of 300 days, 18 healthy adults (M = 33.6 years, female = 16) provided data on sleep duration (SD), wake after sleep onset (WASO), sleep efficiency (SE), sleep onset, and wake-up times. Feasibility was assessed through protocol adherence. Agreement of self-report and trackers was evaluated by Spearman correlations and Bland-Altman plots. Protocol adherence was high (86.5%), whereas data completeness was lower for trackers (FBC: 74.9%; WSA: 66.1%). Both devices correlated strongly with self-reports for SD, sleep onset, and wake-up time (all r > .735), but less so for WASO and SE (all r < .281). Compared to self-reports, the wearable showed shorter (Mean Diff. = 6.91 min) and the contactless sensor longer SDs (Mean Diff. = –9.21 min). Both recorded sleep onset and wake‑up times more than 8 min earlier than self‑reports. These results indicate that sleep-tracking modalities are differentially suited to specific research aims, with self-reports enhancing ecological validity and informing methodological guidelines for scalable, low-cost home-based sleep phenotyping in longitudinal neurobehavioral and population-level studies.