PubMed1986
Ten insomniacs with age- and sex-matched controls had studies of baseline sleep, relation of polygraphically defined sleep to retrospective reports, and arousal thresholds to electronic tones or to a recording of a voice calling out the subject's name. The two groups differed in 10 out of 13 questions about habitual sleep and daytime feelings. In contrast, polygraphic measures of baseline sleep indicated only that insomniacs tended to have slightly less total sleep and had a small but significant increase in early morning awakening time. Unlike the descriptions of habitual sleep, the subjects' retrospective reports of the previous night's sleep differed significantly only for the variable of total sleep time, and there were virtually no differences in the description of their status at a given moment. Auditory arousal thresholds were similar in the two groups, and both went back to sleep and stayed asleep with equal facility. These findings suggest that subjectively poor sleep is not necessarily "light" sleep. For both groups, arousal thresholds differed across the sleep stages, and thresholds to hearing the subject's name were lower than those in response to electronic tones. Although insomniacs had as much polygraphically defined sleep as controls between the forced awakenings of the arousal threshold studies, they perceived their sleep to be only approximately half as long. Insomniacs described themselves as having been awake more frequently than controls in 8 out of 10 forced awakening situations. In one case, insomniacs also overestimated the time between awakenings. In both groups, there was little relationship between reported habitual aspects of sleep and baseline polygraphically defined sleep variables. On questionnaires the following mornings, however, in both groups there was a positive correlation of subjective quality of sleep on the baseline nights with percentage of rapid eye movement sleep, and a negative correlation to various aspects of slow-wave sleep.