Can an in-ear sensor measure how quickly healthy adults fall asleep?

Details

Detecting sleep latency—the time it takes to fall asleep—during the Multiple Sleep Latency Test (MSLT) with scalp electroencephalography (scalp EEG) is time-consuming. The study evaluated a novel in-ear EEG sensor against scalp EEG during MSLT in healthy adults, with and without sleep restriction.

The researchers recruited 25 healthy adults aged 28.5±5.3 years who underwent two MSLTs with simultaneous scalp and in-ear EEG recording. Each test followed a randomly assigned sleep restriction of ≤5 hours or a usual night's sleep of ≥7 hours. Reaction time and the Stroop test assessed the functional impact of sleep restriction. EEGs were scored using American Academy of Sleep Medicine 2012 criteria, blinded to the measurement mode and study conditions. Agreement between scalp and in-ear EEG was assessed with Bland–Altman analysis.

Technically acceptable data were obtained from 23 adults for 69 of 92 naps under sleep restriction and from 25 adults for 85 of 100 naps after usual sleep. An increase in reaction time confirmed meaningful sleep restriction: mean ± SD was 238±30 ms versus 228±27 ms; P=0.045. Under sleep restriction, in-ear EEG had sensitivity of 0.93 and specificity of 0.80 for detecting sleep latency, with substantial agreement (κ=0.71). After usual sleep, sensitivity was 0.91 and specificity was 0.89, again with substantial agreement (κ=0.79).

The in-ear sensor detected reduced sleep latency after sleep restriction, which was sufficient to impair both reaction time and cognitive function. Substantial agreement was observed between scalp and in-ear EEG for sleep latency. The authors concluded that the new in-ear EEG technology has significant value as a convenient measure of sleep latency. The abstract's keywords include electroencephalography, in-ear EEG, MSLT, excessive daytime sleepiness, sleep onset and sleep restriction.