Fitbit Charge 4 accuracy for tracking sleep stages and heart rate compared to standard tests at home in Huntington’s disease
Updated
Abstract
Fitbit Charge 4 estimated total sleep and wake times in participants with Huntington's disease with intraclass correlations ranging from 0.79 to 0.96 compared to polysomnography.
- Fitbit detected waking with 68.6 ± 15.5% sensitivity and 93.7 ± 2.5% specificity, indicating reliable identification of wake periods.
- Rapid eye movement sleep was detected with a sensitivity of 78.7 ± 31.9% and specificity of 95.6 ± 2.3%, suggesting good accuracy.
- Deep sleep detection had lower sensitivity at 56.4 ± 28.8% but maintained high specificity of 95.0 ± 4.8%.
- Heart rate measurements from Fitbit were strongly correlated with those from polysomnography, with a mean absolute error of 1.16 ± 0.42 beats/min.
- Longer sleep and shorter wake times were observed during home monitoring compared to hospital data, while percentage sleep stage times remained consistent.
Simplified
Funding
Competing interests
All authors have seen and approved this manuscript. Work for this study was performed at University Hospital Zurich and University College Dublin. Research was supported by the EU Joint Programme for Neurodegenerative Disease Research with funding from Alzheimer’s Society, Secretary of State for Health and Social Care, Health and Care Research Wales, Public Health Agency Northern Ireland, Jacques and Gloria Gossweiler Foundation, Bundesministerium für Bildung und Forschung, Narodowe Centrum Badań i Rozwoju, Swiss National Science Foundation (32ND30_185548), and Health Research Board (JPND-HSC-2018-003). The authors report no conflicts of interest.
PubMed