al., 2004; Ayas et al., 2006; Barger et al., 2006). Performance can be affected whether sleep loss is in the form of a night of substantially reduced sleep, a night of total sleep deprivation, or a series of less drastic, but more chronic, restricted sleep hours. A 1997 study by Dinges et al. revealed that when sleep is restricted to the level that is commonly experienced by astronauts (i.e., 4 to 6 hours per day), a "sleep debt" accrues and, in less than 1 week, performance deficits during waking hours reach levels of serious impairment.
Chronic reduction of sleep can impact performance in a manner that is similar to that of total sleep deprivation. A study by Van Dongen et al. (2003), which used 48 subjects, evaluated the specific performance effects of chronic sleep restriction in comparison to the effects of 3 nights of total sleep deprivation. Sleep restriction conditions included 14 consecutive nights of 8, 6, or 4 hours of sleep opportunity, with actual sleep quantity validated by polysomnography recordings. Subjects who were subjected to sleep restriction conditions underwent neurobehavioral assessments every 2 hours during their scheduled wakefulness, while subjects who were subjected to the sleep deprivation condition were tested every 2 hours throughout their total 88 hours of sleep deprivation.
The neurobehavioral assessment battery that was used in the Van Dongen et al. (2003) study included the psychomotor vigilance task (PVT). The PVT— which determines alertness and the effects of fatigue on cognitive performance (as determined by lapses in response time and accuracy of responses) by measuring the speed with which subjects respond to a visual or an auditory stimulus (by pressing a response button) � has become a standard laboratory tool for the assessment of sustained performance in a variety of experimental conditions (Dorrian et al., 2005). The PVT detects changes in basic neurobehavioral performance that involve vigilant attention, response speed, and impulsivity; and it has been extensively validated in ground-based laboratory studies to detect cognitive deficits that are caused by a variety of factors (e.g., restricted sleep, sleep/ wake shifts, motion sickness, residual sedation from sleep medications) (Dinges and Powell (1985), Van Dongen et al. (2003), Drummond et al. (2005)). The PVT is an optimal tool for repeated use, in contrast to some other cognitive measures, as studies have shown no minimal learning effects and aptitude differences when using the PVT (Van Dongen et al., 2003; Balkin et al., 2004; Dorian et al., 2003).
Results from these laboratory studies indicate that multiple consecutive sleep episodes of 4 or 6 hours significantly erode performance on the PVT and on measures of working memory, and that performance under these two conditions (i.e., 4 or 6 hours) was comparable to the performance that is found under conditions of 1 to 2 days of total sleep deprivation. Surprisingly, by the end of the 14 days of sleep restriction, subjects in the 4- and 6-hour sleep period conditions reported feeling only slightly sleepy. As these reports were taken when performance was at its lowest level, this indicates that the subjects may no longer have been aware of their performance deficits because of inadequate recovery sleep (Van Dongen et al., 2003) (figure 3-2).
Subjects who spent 4 hours in bed reached levels of impairment at 6 days and of severe impairment at 11 days. Subjects who spent 6 hours in bed reached levels of impairment at 7 days. Interestingly, it appears that subjects who spent 8 hours in bed approached levels of impairment. Figure 3-3, which is from Belenky et al. (2003), however, demonstrates that subjects who spent 9 hours in bed did not approach these levels of impairment, indicating that 9 hours in bed may be needed to alleviate the risk of performance errors.
Similar performance effects resulting from chronically restricted sleep can also be seen in the Category I study by Belenky et al. (2003) and in figure 3-3. This study involved 66 subjects who were observed in four conditions (i.e., 3, 5, 7, and 9 hours in bed) for 7 days. PVT testing showed severe impairments in reaction time under the 3-hour condition, with lapses in responses increasing steadily across the 7 days of sleep re-
Risk of Performance Errors Due to Sleep Loss, Circadian Desynchronization, Fatigue, and Work Overload
91
