Subjective sleep quality and spectral analysis of electroencephalography during nocturnal sleep

INTRODUCTION: The relation between subjective sleep quality and the stages of polysomnography is unclear. Computerized sleep analysis provides a reliable tool for the study of sleep dynamics. OBJECTIVE: To study the relation between subjective sleep quality and spectral EEG parameters during nocturnal sleep. MATERIAL AND METHODS: Nocturnal sleep were recorded ambulatory in 26 subjects aged from 8 to 63. The following variables were extracted from EEG signal (Fz-Pz) quantified through the Fast Fourier Transformation (FFT): delta maximum, spectral peak, delta sleep time, efficiency and desynchronized sleep. A sleep diary was administered in the morning to assess the subjective sleep quality. Multiple regression and Anova were used for statistical analysis. RESULTS: Efficiency and delta maximum became significant predictors accounting for 54% of subjective sleep variance. The item 'how was your sleep?' was answered in a different way according to the delta maximum. CONCLUSIONS: Sleep depth, measured by spectral analysis, is an important predictor of the subjective sleep quality. The sleep classification in discrete stages ignores this feature.