Assessment of Fitbit Charge 4 for sleep stage and heart rate monitoring against polysomnography and during home monitoring in Huntington’s disease
Fitbit Charge 4 accuracy for tracking sleep stages and heart rate compared to standard tests at home in Huntington’s disease
AI simplified
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.
AI simplified