Nightcap Measurement of Sleep Quality in Self-Described Good and Poor Sleepers

Transcription

1 Sleep, 17(8): American Sleep Disorders Association and Sleep Research Society Sleep Quality, Behavioral Scales and Sleep Scoring Nightcap Measurement of Sleep Quality in Self-Described Good and Poor Sleepers Edward F. Pace-Schott, lunko Kaji, Robert Stickgold and 1. Allan Hobson Laboratory of Neurophysiology, Harvard Medical School, Boston, Massachusetts, U.S.A. Summary: The Nightcap is a home-based sleep monitoring device that reliably differentiates rapid eye movement sleep, nonrapid eye movement sleep and wake states using eyelid and body movement measurements. This study documents its capacity to measure differences in sleep latency and sleep efficiency between self-described good and poor sleepers drawn from a normal population. Ten self-described "good" sleepers and II self-described "poor" sleepers were selected from a pool of college students. These groups differed significantly on selection parameters and on subjective estimates of sleep quality obtained each morning during the study. Each subject wore the Nightcap at home for nights. Statistically significant differences in Nightcap-measured sleep latency and sleep efficiency were obtained between groups using individual subject means. In individual subjects, Nightcap measurements of sleep latency were correlated with subjective estimates of sleep latency. Poor sleepers were less accurate in estimating their sleep onset latency than were good sleepers. The demonstrated sensitivity of the Nightcap to good and poor sleep in these normal subjects augurs well for its application in a clinical setting. Key Words: Nightcap-Sleep latency - Sleep efficiency - Insomnia. Insomnia and other sleep disorders constitute a serious public health concern (1,2) and there is a need for an economical means of assessing sleep in the home and prescreening potential candidates for laboratorybased or ambulatory polysomnography (PSG). Although PSG studies provide accurate details of sleep architecture, most simple ambulatory sleep monitors only identify periods of sleep and wakefulness (3-7). The Nightcap, a recently developed home-based sleep monitoring device, offers an intermediate level of information, identifying periods of wake, rapid eye movement (REM) and nonrapid eye movement (NREM) sleep (8-11). Nightcap data analyzed by a computerized algorithm have been shown to predict sleep state in normal subjects with 87% accuracy as compared with simultaneous PSG (9). Objective monitoring of sleep parameters is an important component in the clinical evaluation of poor sleep. In many studies, sleep latency and sleep efficiency have been found to differ between insomniac and normal sleepers (12-20). These two parameters have also been shown to differentiate good and poor sleepers (21). Sleep duration (12,16,20-23) and num- Accepted for publication July Address correspondence and reprint requests to Edward F. Pace Schott, Laboratory of Neurophysiology, Harvard Medical School, 74 Fenwood Road, Boston, MA 02115, U.S.A. 688 ber of awakenings (16,21) also may differ between such groups. In addition, insomniacs are found to characteristically overestimate sleep latency and underestimate number of awakenings (2,24). In the current study, we test the Nightcap's sensitivity to good and poor sleep in a normal population as a prelude to clinical studies. The current study determines the ability of the Nightcap to detect objective differences between self-described good and poor sleepers. Based upon results of our pilot studies (10,11), a PSG study of good and poor sleepers (21) and PSG studies of insomniacs (12-20), we predicted longer sleep latency and poorer sleep efficiency in the poor sleeper group. METHODS Eighteen to 24-year-old college students who responded to advertisements gave estimates of habitual sleep latency, number of nighttime awakenings and degree of morning fatigue. Those near the high and low extremes of these self-report parameters were further screened and excluded if they slept with a partner, took sleep-affecting medication, or had chronic physical or mental conditions affecting sleep. Remaining subjects with the most severe sleep complaints or, alternatively, with self-reports of especially good sleep were assigned to "poor" and "good" sleeper groups,

2 NIGHTCAP STUDY OF GOOD AND POOR SLEEPERS head movement sensor 2. eyelid movement sensor mount 3. eyelid sensor lead 4. eyelid sensor with adhesive backing 5. bandanna (worn "pirate style") 6. wires from sensors to Nightcap unit 7. Nightcap recording unit B. J fl lilt I ~.IL It I' Eye I 12,00 1 :00 2,00 3:00 4,00 S.OO fo,j If! I " I ill I I,,, 12,00 LOo 2, ,00 5,00 Movements -w,,", ] - RIM Polygraph -NRFM Heal Movementc,; ] Nightcap Fig. 1. A. Photo-based drawing of the Nightcap and its mode of attachment using a bandanna. B. Sample output and analysis by the NC Analyzer program. Top trace: Histogram plot of Nightcap-detected eyelid movements; Second trace: Hypnogram representing the manually scored PSG record using Rechtschaffen and Kales' criteria (31); Third trace: Hypnogram of computer-scored Nightcap data; Fourth trace: Histogram plotting Nightcap-detected head movements. On all hypnograms, the top level represents wake, the second level represents REM, and the third level is NREM. Movement time is not shown. The lower axis indicates the time of night. [Illustration from Ajilore et al. (9)] respectively. Five female and five male good sleepers and five female and six male poor sleepers were selected. Subjects completed a prestudy sleep-quality questionnaire, gave informed consent and were paid $100 each to participate in the study. During Days 1 to 7 of the study, subjects completed a daily questionnaire before and after each night. Beginning on Day 8, subjects wore the Nightcap each night and continued to complete the daily questionnaire. Subjects completed at least 12 "artifact-free recording nights" (defined as nights free of Nightcap malfunction, sensor breakage, gross environmental disturbance or clear noncompliance). Nightcap data were periodically downloaded to a Macintosh computer. The Nightcap used in this study is a modification of devices described by Mamelak and Hobson (8) and Ajilore et al. (9). The major differences between the current device and that described by Ajilore et al. (9) included 1) improved signal-to-noise characteristics of the piezo eyelid movement sensor and 2) improvements in the supporting software and scoring algorithm (NC Analyzer 3.6b, Laboratory of Neurophysiology, Harvard Medical School, Boston, MA). Figure 1 illustrates the Nightcap device, its mode of attachment and a sample computer printout with hand-scoring of a simultaneous polysomnogram. Sleep latency, sleep efficiency, sleep duration and number of awakenings for each subject night were computed using Nightcap data. Parameter means were then determined for each subject and used in comparisons between good and poor sleeper groups. Only data from artifact-free nights were used in computing parameter means for each subject. RESULTS The expected subjective differences between the two groups selected were confirmed both with regard to their perception of habitual sleep (p < 0.001, onetailed t tests for each of three prestudy screening questions) and with regard to their night-by-night estimates of sleep quality (p < 0.01, multivariate analysis of variance, nine sleep quality-related daily questionnaire items). In comparison with self-described good sleepers, selfdescribed poor sleepers showed significantly longer Nightcap-measured sleep latency (25.8 vs minutes) and significantly lower Nightcap-measured sleep efficiency (0.91 vs. 0.95). Both one-tailed t tests and two-tailed Mann-Whitney Utests comparing individual subject means detected statistically significant differences in these two parameters between these two groups (Table 1). In contrast, self-described good and poor sleeper groups did not differ with respect to Nightcap-measured sleep duration or number of awakenings (Table 1). Self-described good sleepers more accurately estimated their Nightcap-measured sleep latency than did self-described poor sleepers. The mean subjective error for poor sleepers was significantly greater than that of good sleepers (18.5 vs. 6.3 minutes) when compared by a two-tailed t test and by a two-tailed Mann-Whitney U test (Table 1). Espie et al. (25) suggested that longer time periods Sleep. Vol. 17. No.8, 1994

3 690 E. F. PACE-SCHOTT ET AL. TABLE 1. Comparison of Nightcap- measured sleep parameters between self-described good and poor sleeper groups Mean latency Mean subjective error as % Sleep Sleep latency sleep latency group mean Sleep duration No. of n (minutes) error (minutes)a latency" efficiency (minutes) awakenings Good sleepers (6.1) 6.3 (2.4) 61.3 (23.3) 0.95 (0.02) (43.7) 0.86 (0.46) Poor sleepers lib 25.8 (21.7) 18.5 (12.6) 71.7 (48.9) 0.91 (0.07) (53.1) 1.21 (0.77) t test P < 0.05 P < om ns p < 0.05 ns ns (I-tailed) (two-tailed) (I-tailed) U test (two-tailed) p < om p < ns p < 0.05 ns ns Good and poor sleeper groups' sleep parameters were compared using t tests and Mann-Whitney U tests on individual subject means. Parameters were computed using artifact-free records obtained from each subject (for recording quality criteria, see Methods). Standard deviations are shown in parentheses next to parameter means. a Each subject's mean subjective sleep latency error was computed by averaging the absolute values of the nightly differences between Nightcap measurements and morning estimates of sleep latency. Each subject's mean latency error as percent of group mean latency was determined by dividing each night's subjective sleep latency error by his or her group mean sleep latency (10.2 and 25.8 minutes for good and poor sleeper groups, respectively). b Data from one poor sleeper who frequently neglected to estimate sleep latency were excluded from all the objective/subjective comparisons thereby reducing the number of poor sleepers to 10 for these comparisons. may be inherently more difficult to estimate than shorter ones. Noting the two-fold group difference in objective sleep latency in addition to the three-fold group difference in mean subjective sleep latency error (Table I), we performed a second percentage-based comparison to see if the group difference in accuracy of estimating sleep latency might reflect such overall poorer estimation of longer periods. When nightly subjective errors were expressed as a percentage of the average Nightcap-measured sleep latency for a subject's group, the mean subjective error of the poor sleeper group was no longer significantly greater than that of the good sleeper group (Table 1). Despite overestimations, variation in a subject's perceived sleep latency did reflect his or her nightly variation in Nightcap-measured sleep latency. Spearman rank correlations relating nightly estimates to Nightcap-measured sleep latencies were positive in 17 subjects (p < 0.01, binomial test), statistically significant (p < 0.05) in nine subjects, and the average Spearman correlation coefficient (rs = 0.40) was significantly greater than zero (p < 0.001, two-tailed t test). DISCUSSION Our results show that the Nightcap and its software can demonstrate objective differences in sleep latency and sleep efficiency between groups of self-described good and poor sleepers. In polysomnographic studies, these are the two objective sleep parameters that most consistently differentiate insomniac versus normal sleepers (12-20). The magnitude of difference in mean sleep latency between good and poor sleeper groups can be expressed as a percentage ofthe good sleeper group's mean sleep latency. When expressed in this way, the difference in mean sleep latency between good and poor sleeper groups in this study (153% of the mean sleep latency for the good sleeper group) is higher than the PSGmeasured sleep latency difference between good and poor sleepers (105%), which we calculated in the same manner from mean data reported by Monroe (21). This difference also lies within the lower range of differences calculated from mean values reported in seven PSG studies (12-16,18,20) comparing insomniacs and normals (range = %, mean = 221%, SD = 84%). When expressed in the same manner, the difference in mean sleep efficiency between good and poor sleeper groups in this study (4.2% of the mean sleep efficiency for the good sleeper group), although statistically significant, is lower than the difference calculated from Monroe's PSG data (9.3%) as well as differences calculated from mean values reported in eight PSG studies (12-16,18,20,26) comparing insomniacs and normals (range = %, mean = 14.5%, SD = 5.4%). In this study, therefore, the Nightcap is probably detecting the small objective differences in sleep quality that exist between good and poor sleepers in an essentially normal population, which does not include clinically diagnosed chronic insomniacs. Contributing to the low magnitude of sleep latency and sleep efficiency group differences in the current study (as well as the failure to detect group differences in sleep duration and number of awakenings) are the following three factors. 1) Normals and insomniacs differ more in subjective estimates of habitual, overall sleep quality than in either nightly sleep diary reports or PSG measurements (27). Because subjects were selected on the basis of self-reports of habitual sleep patterns, an overestimation of actual differences apparently occurred during subject selection. 2) The Nightcap is not designed to detect the brief «1-2-minute) nighttime arousals that have been re- Sleep, Vol. 17, No.8, 1994

4 NIGHTCAP STUDY OF GOOD AND POOR SLEEPERS 691 ported in some PSG studies (e.g. reference 15). If such brief awakenings contributed to the disturbed sleep reported by our poor sleepers, group differences in the measured number of awakenings might have been correspondingly reduced. For the same reason, sleep efficiency in poor sleepers might have been more inflated than sleep efficiency in good sleepers, thereby decreasing the magnitude of sleep efficiency differences between the groups. 3) Sleep patterns of many college students are characterized by chronic voluntary sleep deprivation during the week followed by rebound sleep on weekends (28). Such patterns may increase group variance in sleep duration among both good and poor sleepers, thereby reducing group differences. More greatly inflated sleep latency estimates in the poorer sleepers agree with previous studies which found that insomniacs overestimate sleep latency (24,29). Poor sleepers are often ascribed a greater tendency to overestimate sleep onset latency than good sleepers (2). However, poor sleepers are also attempting to estimate a longer objective sleep latency and, in the current study, the difference between good and poor sleepers' subjective error is not statistically significant if expressed as a percentage of their group's average actual latency. This finding suggests that poor sleepers' greater overestimation of sleep latency may be influenced by an inherently greater difficulty in estimating longer versus shorter time periods as suggested by Espie et al. (25). We consider the sensitivity of the Nightcap to the relatively small differences between our normal good and poor sleepers to be encouraging of future work in a clinical setting. Combined PSG and Nightcap studies on a variety of sleep-disordered populations, including a sample of diagnosed insomniacs, are currently underway. Chronic insomniacs and persons suffering from other sleep disorders may display longer nocturnal arousals (30), which are more readily detected by the Nightcap. In addition, we are conducting further development of the sleep-state scoring algorithm, which may result in improved detection of microarousals. Acknowledgements: This project was funded by grants from the MacArthur Foundation and NIH (MH-48, 832). The authors wish to thank Healthdyne Technologies, Inc., for providing Nightcaps and related accessories and the Biostatistics Consulting Clinic at the Harvard School of Public Health and Cynthia D. Rittenhouse for statistical aid and helpful discussion. REFERENCES I. Ford DE, Kamerow DB. Epidemiologic study of sleep disturbances and psychiatric disorders. An opportunity for prevention? JAm Med Assoc 1989;262: Morin CM. Insomnia, psychological assessment and management. New York: Guilford Press, Alihanka J, Vaahtoranta K. A static charge sensitive bed: a new method for recording body movements during sleep. Electroencephalogr Clin NeurophysioI1979;46: Franklin J. The measurement of sleep onset latency in insomnia. BehavRes Ther 1981;19: Lichstein KL, Nickel R, Hoelscher n, Kelly JE. Clinical validation of a sleep assessment device. Behav Res Ther 1982;20: Oglivie RD, Wilkinson RT. Behavioral versus EEG-based monitoring of all-night sleep/wake patterns. Sleep 1988;11: Cole RJ, Kripke DF, Gruen W, Mullaney DJ, Gillin JC. Automatic sleep/wake identification from wrist activity. Sleep 1992; 15: Mamelak A, Hobson JA. Nightcap: a home-based sleep monitoring system. Sleep 1989; 12: Ajilore OA, Stickgold R, Rittenhouse C, Hobson JA. Nightcap: home-based evaluation of a portable sleep monitor. PsychophYSiology, 1994: Ajilore OA. Digitizing dreams: validation of a home-based sleep monitor and its use in sleep quality assessment. Senior thesis. Harvard College, Cambridge, MA, II. Ajilore OA, Stickgold R, Rittenhouse C, Hobson JA. Assessment of good and poor sleepers using the Nightcap home-based sleep monitor. Sleep Res 1993;22: Beutler LE, Thornby 11, Karacan I. Psychological variables in the diagnosis of insomnia. In: Williams RL, Karacan I, eds. Sleep disorders: diagnosis and treatment. New York: John Wiley & Sons, 1978: Bonnet MH, Arand DL, McNulty TB. 24-Hour metabolic rate in matched normals and insomniacs. Sleep Res 1993;22: Coates n, George J, Killen JD, Marchini E, Hamilton S, Thorensen C. Discriminating good sleepers from insomniacs using all-night polysomnograms conducted at home. J Nerv Mental Dis 1982; 170: Frankel BL, Coursey RD, Buchbinder R, Snyder F. Recorded and reported sleep in primary insomnia. Arch Gen Psychiatry 1976;33: Gaillard JM. Chronic primary insomnia: possible physiopathological involvement of slow wave sleep deficiency. Sleep 1978; 1: Gillin JC, Duncan W, Pettigrew KD, Frankel BL, Snyder F. Successful separation of depressed normal and insomniac subjects by EEG sleep data. Arch Gen Psychiatry 1979;36: Hauri P, Fisher J. Persistent psychophysiologic (learned) insomnia. Sleep 1986;9: Kuderian RH, Oglivie RD, McDonnell G, Simons IA. Behavioral response home monitoring of good and insomniac sleepers. Canad J PsychoI1991;45: Schneider-Helmert D. Twenty-four-hour sleep-wake function and personality patterns in chronic insomniacs and healthy controls. Sleep 1987; 10: Monroe LJ. Psychological and physiological differences between good and poor sleepers. J Abnormal PsychoI1967;72: Salin-Pascual RJ, Roehrs TA, Merlotti LA, Zorick F, Roth T. Long-term study of the sleep of insomnia patients with sleep state misperception and other insomnia patients. Am J Psychiatry 1992; 149: Stepanski E, Zorick F, Roehrs T, Young D, Roth T. Daytime alertness in patients with chronic insomnia compared with asymptomatic control subjects. Sleep 1988; II : Carskadon MA, Dement WC, Mitler MM, Guilleminault C, Zarcone VP, Spiegel R. Self-reports versus sleep laboratory findings in 122 drug-free subjects with complaints of chronic insomnia. Am J Psychiatry 1976;133: Espie CA, Lindsay WR, Espie LC. Use of the sleep assessment device (Kelley and Lichstein, 1980) to validate insomniacs' selfreport of sleep pattern. J Psychopathol Behav Assessment 1989; 1: Mendelson WB, Garnett D, Gillin JC, Weingartner WB. The experience of insomnia and daytime and nighttime functioning. Psychiatry Res 1984; 12: Sleep. Vol. 17, No.8, 1994

5 692 E. F. PACE-SCHOTT ET AL. 27. Mendelson WB, James SP, Garnett D, Sack DA, Rosenthal NE. A psychophysiological study of insomnia. Psychiatry Res 1986; 19: Lack LC. Delayed sleep and sleep loss in university students. J Am College Health 1986;35: Knab B, Engel RR. Perception of waking and sleeping: possible implications for the evaluation of insomnia. Sleep 1988; II : Kales A, Bixler EO, Vela-Bueno A, Cadieux RJ, Soldatos CR, Kales JD. Biopsychobehavioral correlates of insomnia, III: polygraphic findings of sleep difficulty and their relationship to psychopathology. Int J Neurosci 1984;23: Rechtschaffen A, Kales A, eds. A manual of standardized terminology. techniques and scoring system for sleep stages in human subjects. Los Angeles: Brain Information Service/Brain Research Institute, University of California at Los Angeles, Sleep. Vol. 17. No