0999 Can Electroencephalogram-modulated Music Facilitate Falling Asleep?

Introduction The electroencephalogram (EEG) reflects the falling asleep process through a progressive reduction in the beta power (15-30 Hz) and a corresponding increase in the theta power (4-8 Hz) band. The log-ratio “ρ=10×log(β/θ)” monotonically decreases by an order of magnitude as sleep initiates. Methods We modulated the volume of audio using the average “ρ” of 115 sleep EEG recordings from a previous study and tested the effect of the modulated audio on sleep latency and ρ of nine participants (6M/3F; 38.8 ± 11.9 years old) in this study.Subjects were randomly assigned to two groups: music (N=5) and white-noise (N=4). Within each group, three conditions were tested: baseline, audio (music or noise), and modulated-audio (gradually decreasing volume). Per participant, two sleep recordings per audio condition and 5 baseline nights were collected at home using a sleep wearable EEG system. The audio was played using earbuds connected to an MP3 player (in baseline no audio was played). Subjects were instructed to start the EEG recording and audio rendering at their sleep intent time. The volume of the audio was proportional to “ρ” such that (Vol= 15 dB for ρ = −4). Both audio streams: music or white noise lasted for 10 minutes. In the music condition, subjects could select among three songs known to promote relaxation. Results Across all conditions, latency to the first sleep epoch was shortest in the modulated-music (6.88 ± 4.08 min) condition but was not statistically significantly different from baseline (13.72 ± 13.65 min). Latency to stable N2 (first two-minute N2 bout) was shortest in the modulated-music condition (9.17 ± 15.5 min) with a medium effect size (Cohen’s d = 0.4) compared to baseline (23.7±20.9 min). The value of “ρ” ten minutes after the 1st sleep epoch was the lowest in the modulated-music condition with a statistically trending difference compared to baseline (p=0.057). Conclusion Modulated-music applied at sleep-intent time was associated with the shortest sleep latency and fastest ρ decay. This promising effect needs to be validated in a larger sample size to assess statistical significance. Support (If Any) NA