Nightcap: a reliable system for determining sleep onset latency

Previous studies have demonstrated that the Nightcap home-based sleep-monitoring system can differentiate waking, NREM and REM sleep based on eyelid movements (ELMs) and head movement behavior. The present study aims at determining the reliability of the Nightcap in determining the human sleep onset latency (SOL) as revealed by standard polysomnography (PSG). Four naps were recorded in each of ten normal subjects using both PSG and the Nightcap simultaneously. The Nightcap algorithm scored sleep onset as the first of 4 consecutive 30-sec epochs with less than 5 ELMs. The mean percentage of agreement between the Nightcap and PSG was 93% (k = 0.79), and the average absolute difference was 45 sec (13.3% of SOL(PSG)). SOL(NC) differed by less than 1 min in 85% of onsets. Recordings of EEG activity from 90 sec before and after PSG-identified sleep onsets were subjected to spectral analysis. Changes in spectral power in the theta (4-7 Hz) and alpha (8-12 Hz) frequency bands during the transition into light sleep correlated well with eyelid behavior. However, changes in ELM density predicted sleep onset better than did changes in theta and alpha spectral power. These results suggest that the Nightcap may be a potential alternative to the PSG technique in the assessment of SOL in normal subjects.