y. Sleep Res. (1999) 8, 237-245 REVIEW Does sleep fragmentation impact recuperation? A review and reanalysis NANCY JO WESENSTEN, THOMAS J. BALKIN atid GREGORY BELENKY Department of Neurobiology and Behavior, Division of Neuropsychiatry, Walter Reed Army Institute of Research, Washington DC, 20307-5100, USA Accepted in revised form 24 April 1999; received 13 October 1998 SUMMARY Studies have shown that next-day performance and alertness are impaired by sleep fragmentation procedures even when total sleep time (TST) is unaffected. Based on these studies it has been hypothesized that both the duration and continuity of sleep determine its recuperative value. This review of the literature suggests that when sleep fragmentation procedures increase the relative amount of stage 1 sleep, next-day performance and alertness are impaired. Other studies suggest that stage 1 sleep has little or no recuperative value. Total sleep time, however, is typically defined as the sum of time spent in sleep stages 1, 2, 3, 4, and REM. In the present paper it is shown that when stage 1 sleep is excluded from TST, a stronger relationship between TST and subsequent alertness and performance emerges - and the need to invoke 'sleep continuity' as a variable that contributes independently to recuperative sleep processes is obviated. In the same way that partial or total sleep deprivation impairs alertness and performance, it is proposed that sleep disruption also impairs alertness and performance by reducing true recuperative sleep time. KEYWORDS alertness, arousals, disruption, fragmentation, performance, sleep stage INTRODUCTION The physiological mechanisms by which sleep restores and sustains alertness and cognitive performance are not yet known, and the functional significance of many sleep architecture parameters remains obsctire. However, one sleep parameter - steep dtiration - clearly impacts the recuperative value of sleep (Kleitman 1963). Although the relationship between sleep duration and recuperation may not be linear (Lumley et al. 1986), it is generally true that longer sleep durations result in enhanced performance and alertness during subsequent wakefulness (see Dinges et al. 1997; Wilkinson et al. 1966). Apparent exceptions to the principle that sleep duration Correspondence: Nancy Jo Wesensten PhD., Department of Neurobiology and Behavior, Division of Neuropsychiatry, Walter Reed Army Institute of Research, Washington DC, 20307-5100, USA. Tel: (301) 295 7826; Fax: (301) 295 7445 determines recuperation have been reported. In some studies it is suggested that daytime alertness can be impaired in both normal subjects and subjects with sleep disorders, despite nocturnal sleep periods of nortnal duration (e.g. Carskadon et al. 1982; Rosenthal et al. 1984; Stepanski et al. 1984; Bonnet 1985). Based on such findings, it was hypothesized that in addition to sleep duration 'sleep continuity' must independently mediate recuperative sleep process(es) (Bonnet 1985). That is, it has been proposed that disruptions of sleep, whether 'external' (e.g. noises, tones) or internal (e.g. apneas, periodic leg movements), impair recuperation even though TST may be unaffected; however, the mechanism(s) by which fragmentation reduces the recuperative value of sleep are not specified. In fact, recent reviews of the efTects of sleep length, architecture, and continuity on daytime alertness fail to provide operational (and therefore testable) definitions of 'sleep continuity' and fragmentation (Gillberg 1995; Chugh et al. 1996; Kimoff 1996). 1999 European Sleep Research Society 237