Mixed

Consumer wearable devices (specifically Apple Watch) can accurately predict sleep stages (Wake, NREM, REM) and sleep-wake status by combining raw acceleration, photoplethysmography (PPG) heart rate, and a circadian 'clock proxy' feature using neural network classifiers.

If you use a smartwatch (like an Apple Watch) to track sleep, you can trust its sleep-wake detection more than previously thought, especially if it uses heart rate data. However, do not rely on it for diagnosing sleep disorders. For research or personal optimization, combining movement with heart rate variability and circadian timing models provides a robust estimate of sleep stages, bridging the gap between consumer tech and clinical science.