Abstract
The effect of steady noise on sleep was examined experimentally for 24 nights by sleep polygraphy of a twenty-eight-year-old man. The parameters used for sleep polygraphy were sleep stage, sleep latency, frequency of sleep stage shift per hour, average sleep depth, the gradient and intercept of a regression line of sleep depth against time, and subjective sleep. The sleep stage was judged and estimated based on the criteria of Rechtschaffen & Kales by the authors' original system using a microcomputer. For the calculation of sleep depth at a given time, stages W (waking), 1, REM, 2, 3 and 4 were weighted to be 0, 1, 1.5, 2, 3 and 4, respectively. Sleep parameters were compared between the data of non-noise and noise-exposed nights. The noise exposure was overnight stationary pink noise of 40, 50 and 60 dB (A).
General findings were a decreased REM stage, an increased stage 2, and deepened average sleep depth of the 60 dB (A)-night sleep compared with the control and 40 dB (A) exposure nights. When the control nights were excluded and parameters were compared between the three levels of stationary noise exposure, stage 2, the intercept of the regression line, and average sleep depth were increased almost in proportion to the increased noise exposure. In contrast, sleep latency, frequencies of sleep stage shift, stages W, 1, 3, MT, and REM, as well as the gradient of the regression line were decreased with increased noise exposure.
Although the decrease of stage 3 and REM at 60 dB (A) indicated the possibility of sleep disturbance, the subject continued to sleep well as indicated by the shortened sleep latency and deepened average sleep depth even under the environment of 60 dB (A) stationary noise exposure.
