Differences Between Sleep Logs and Actigraphy Combined With Electroencephalography in Adults With Sleep Disturbances

Abstract

Background:

Actigraphy plus single-channel electroencephalography (EEG) has a high agreement with polysomnography and is considered a simple method for monitoring sleep at home. No studies, however, have evaluated its agreement with sleep logs.

Purpose:

A sleep log was compared to a single-channel EEG in adults with sleep disturbances. The determinants of the agreement between the two recording methods were explored.

Method:

This prospective comparison substudy included 71 participants aged 22–67 years sampled from a previous study of nonpharmacological therapy for sleep disturbances. The participants concurrently used a sleep log and a single-channel EEG machine for 6 days to measure their sleep at home. Total sleep time (TST) and sleep onset latency (SOL) were compared.

Results:

Correlations between measurements were .43 for TST (p < .001) and .30 (p < .01) for SOL. Bland–Altman analysis revealed that the mean bias between methods was 81.54 min (standard deviation [SD] = 66.72 min) for TST and 5.58 min (SD = 25.38 min) for SOL. SDs for both TST and SOL were higher for the sleep log than for EEG. A multivariate linear regression model demonstrated that age (standardized coefficient β = .42, p = .001) was a significant predictor of the mean difference in SOL measurements, accounting for 13.7% of the total variance in the mean difference.

Conclusion:

Sleep log and single-channel EEG sleep measures are distinct from one another, and age predicts the mean level of disagreement between the two measures in adults with sleep disturbances.