| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
RESEARCH ARTICLE |
a The Intervention Research Center for Late Life Mood Disorders and the Clinical Neuroscience Research Center, Department of Psychiatry, University of Pittsburgh School of Medicine, Pennsylvania
Charles F. Reynolds III, Department of Psychiatry, University of Pittsburgh School of Medicine, 3811 O\|[apos ]\|Hara Street, Room 1135-E, Pittsburgh, PA 15213 E-mail: reynoldscf{at}msx.upmc.edu.
Decision Editor: Toni C. Antonucci, PhD
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
MOST elderly persons spend excessive time in bed and experience a loss in the consolidation and depth of sleep, particularly after the age of 75, and an increase in daytime sleepiness. These changes in sleep consolidation, depth, and daytime sleepiness have a large impact on quality of life, level of functioning, and ability to remain independent in the old-old (Ford and Kamerow 1989
; Morin and Gramling 1989; Prinz, Vitiello, Raskind, and Thorpy 1990; Wauquier, van Sweden, Lagaay, Kemp, and Kamphuisen 1992). Data from our longitudinal observational studies indicate that decrements in sleep quality, affect, and cognition among the healthy old-old are associated with poor sleep efficiency, and that "inefficient" or poorly consolidated sleep is correlated with subsequent declines in indicators of mental and physical health status (Dew et al. 1994; Hoch et al. 1994; Reynolds, Frank, Perel, Imber, and Kupfer 1994). These relationships appear to be bi-directional because increasing medical illness is also usually associated with increasing disturbance of sleep in later life, both loss of continuity and depth of sleep (for review, see Bliwise 1993). Data such as these underscore the need to devise safe and effective behavioral and educational interventions for preserving and enhancing sleep quality and for preventing age-related declines in sleep continuity and depth among the elderly population. As part of our long-standing investigation of the role of sleep in successful aging, we have carried out a pilot intervention study to explore the ability of a behavioral versus a psychoeducational intervention to have an impact on measures of sleep consolidation and depth over a 1-year period. The hypothesis that we eventually aim to test is that maintaining good sleep consolidation and depth (i.e., preventing loss of sleep quality) will contribute to preventing declines in functional status and thereby diminish functional morbidity in the old-old. As a preliminary step to testing that hypothesis, we examined the feasibility of carrying out behavioral and psychoeducational interventions in the old-old, determining the magnitude of their impact on polysomnographically recorded sleep. It is these data which we present here.
To this end, our pilot study tested two interventions in a controlled trial for improving the "inefficient" sleep of old-old participants (i.e., sleep that is fragmented by frequent wakefulness and rendered less efficient and deep): (a) a modification of sleep-restriction therapy and (b) sleep-hygiene education. We selected these two approaches based upon our previous work (Hoch et al. 1987) and their effectiveness in treating elderly participants with chronic insomnia (Friedman, Bliwise, Yesavage, and Salom 1991; Spielman, Saskin, and Thorpy 1987). In our previous work, we have observed that modest self-imposed reduction of time in bed of 30–60 minutes nightly and a highly entrained lifestyle were associated with the maintenance of superior sleep in a study of healthy elderly nuns (Hoch et al. 1987). The nuns fell asleep more quickly and had less early morning awakening than age-matched female controls. As a pilot project, therefore, our aim was to determine whether two sleep health interventions produce measurable benefits to sleep consolidation and depth, functional status, and quality of well-being in noncomplaining elders in good mental and physical health, while inducing minimal negative effects (e.g., daytime sleepiness). In order to provide a benchmark for the further interpretation of our data, we constructed an archival control group using data from similar participants (healthy elderly volunteers without sleep disorders) followed longitudinally without any kind of intervention.
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
). Participants were also required to have a score of less than seven on the Hamilton Depression Rating Scale (HDRS, Hamilton 1960) and of 28 or greater on the Folstein Mini-Mental Status Examination (MMSE; Folstein, Folstein, and McHugh 1975). Participants also had a physical examination, electrocardiogram, complete blood count, thyroid function tests, and chemistry screen to detect serious or uncontrolled physical health problems as well as medication use that could affect sleep or mood. Participants under a physician's care for stable medical illness (e.g., heart disease, hypertension, arthritis, diabetes, thyroid disease) and with health conditions that posed no major limitation to activities of daily living were admitted to the study. With respect to exclusion criteria, participants who had an apnea-hypopnea index of 20 or greater or a sleepiness index of 50 or greater (i.e., a mean sleep latency of 
10 min) on the multiple sleep latency test were excluded from the study. No sex, race, or ethnic group restrictions were applied. Participants gave written informed consent before entering the study. We matched each of the 21 participants on age and gender with a control participant in our archives who had met the same inclusion/exclusion criteria and who had participated in our longitudinal, observational study of sleep in successful aging, without entering an intervention trial. As shown in Table 1 , archival control participants did not differ from respondents participating in the trial on key demographic and clinical measures.
|
) to the bed-restriction condition (which also included sleep-hygiene education) or to the sleep-hygiene condition (without prescribed restriction of time in bed). Participants were stratified before group assignment so that each intervention group had an equal proportion of "inefficient sleepers" (i.e., EEG sleep efficiency [ratio of time spent asleep to time in bed] of 80% or lower) and "good sleepers" (i.e., above 80% sleep efficiency). Sleep efficiency was determined at baseline (T1) sleep evaluation. There were four participants in the bed-restriction condition with sleep efficiencies greater than 80% and five in the sleep-hygiene condition at greater than 80%. The minimum baseline sleep efficiency in the sleep-hygiene cell was 73.5%; there were four participants in bed restriction with baseline sleep efficiency below 67%.
Time in bed restriction.
We used an 8-week period for the initial active intervention, assigning participants to a 30-min restriction of their available time in bed. Participants were instructed to delay bedtime by 30 minutes a day (for example, going to bed at 10:30 p.m., rather than 10:00 p.m.). Participants were also allowed to take a 30-minute nap between 2:00 p.m. and 4:00 p.m. daily, as needed. The intent of this intervention was to provide a slow and conservative approach to accumulating a modest sleep debt in order to improve sleep consolidation (efficiency) and depth. The method of intervention varied somewhat from the approach used for insomnia sufferers. Typically when restriction of time in bed is used as a treatment for chronic insomnia, the amount that time in bed is reduced depends upon the patient's initial average total sleep time and average time in bed. We chose to reduce time in bed by a standard amount (30 minutes) because our participants were not patients with insomnia. For patients with insomnia, the reduction is tailored more to the participant's individual needs. During the initial phase, participants met weekly for 1 hour with the project coordinator (F.H.) and principal investigator (C.C.H.). The first contact focused on education about the principles of sleep restriction and included instructions in delaying of bedtime by 30 minutes, keeping regular bedtimes and wake-up times, making use of the opportunity to take 30 minute naps, and keeping daily logs of their sleep-wake schedule. Subsequent weekly meetings reviewed daily sleep logs, assessed daytime sleepiness, reinforced bed restriction and good sleep hygiene (described below), completed study measures, facilitated compliance, and answered questions. Time 2 (T2) sleep studies occurred at the end of the active acute intervention phase (Week 8).
During continuation treatment, Weeks 9–24, participants met every other week for 30 minutes with the project coordinator to encourage and assess compliance, assess daytime sleepiness, complete study measures, reinforce bed-restriction and sleep-hygiene education, and to troubleshoot any incident problems with sleep. Time 3 (T3) sleep studies occurred at the end of the continuation phase (Week 24).
Maintenance treatment continued for 6 months beyond the end of continuation. Participants met with the project coordinator for 30 minutes monthly to foster continued compliance, answer questions, and complete study assessment procedures. The final (T4) set of sleep studies occurred at the end of maintenance treatment (Week 52).
Sleep-hygiene education intervention.
Participants randomly assigned to this condition received sleep-hygiene education from the project coordinator and principal investigator, followed the same meeting schedule, and completed the same study measures as did participants in the bed-restriction condition. Initial contacts focused on education about the principles of sleep hygiene, including the effects of caffeine, tobacco, alcohol, and medications; the benefits of moderate exercise and dietary practices as they pertain to sleep; and attention to room temperature, noise, lighting, and pre-bedtime routines. Participants were specifically not instructed about the amount of time spent in bed but did receive instruction about keeping regular bedtimes and wake-up times and about regularizing naps. Subsequent meetings reinforced sleep-hygiene principles, fostered compliance, answered questions, and provided support related to sleep.
Archival control group.
Participants in this group had received no sleep-related intervention but were participants in our study of "Sleep and Sleep Quality in Successful Aging" (NIMH [NIH] Grant MH52266). These participants had the same baseline (T1) and 1-year (T4) assessments as participants in this intervention study.
Assessment procedures and outcome measures.
A structured interview for sleep disorders based upon the work of Buysse and colleagues 1994 was administered at baseline to evaluate participants for DSM-IV sleep disorder diagnoses. The Pittsburgh Sleep Diary (Monk et al. 1994) was completed daily by all participants in order to ascertain information about bedtime, wake-up time, napping, subjective sleep quality, mood, and vigor. In-laboratory total recording time was set based upon diary information. At each of the four assessment points, participants had three consecutive nights of sleep recording during their usual sleep time in private bedrooms of the Clinical Neuroscience Research Center (University of Pittsburgh School of Medicine). Electroencephalographic, electro-oculographic, and bipolar submental electromyographic activity were recorded on Grass model 78B polygraphs with high pass filter settings (EEG) of 0.3 and low pass filter settings of 30. Airflow was measured via oral and nasal thermistors and respiratory effort with pizo-electric crystal movement detectors. Spectral analysis of the EEG signal was performed as described elsewhere (Vasko et al. 1997). Oxygen saturation was measured via finger probe with the Ohmeda 3700 Biox Oximeter (Ohmeda, Louisa, Coloradao). Each participant also completed a multiple sleep latency test (MSLT) following the third night of each set of sleep studies to document objectively the level of daytime sleepiness (Carskadon, Dement, and Mitler 1986). The MSLT consists of five opportunities for participants to nap, with the first nap beginning 3 hours after the participant awakened from nocturnal sleep. The four remaining naps were at 2-hour intervals throughout the day. Each nap recording lasted up to 20 minutes; the time it took participants to fall asleep was measured and used to calculate the mean sleep latency.
The major laboratory-based outcome measures dealt with aspects of sleep continuity (sleep efficiency), depth of sleep (percent delta sleep and delta EEG power), sleep-disordered breathing (apnea-hypopnea index), and daytime sleepiness (mean sleep latency on the MSLT). These measures were selected because they reflect age-dependent changes in sleep and/or because we wished to document any adverse effects of restricting time in bed (e.g., Apnea Hypopnea Index, AHI, daytime sleepiness). Clinical measures dealt with self-reported sleep quality (Pittsburgh Sleep Quality Index score; Buysse, Reynolds, Monk, Berman, and Kupfer 1989), the diary-based measures of self-reported alertness and mood on awakening (Monk et al. 1994), self-reported symptoms of depression (Beck Depression Inventory; Beck, Ward, Mendelson, Mock, and Erbaugh 1961), anxiety (Brief Symptom Inventory; Derogatis and Melisaratos 1983), well-being (Campbell Well-Being Scale; Campbell, Converse, and Rodgers 1976), and quality of life (MOS-SF36, Ware and Sherbourne 1992; and Global Assessment Scale, Endicott, Spitzer, Fleiss, and Cohen 1976).
Costs and payments to participants
Respondents participating in the study incurred no costs beyond the time and effort involved. Each participant was paid $150 for completion of each 3-day series of nocturnal sleep studies, $35 for each MSLT, and $15 for each of 22 office visits. Thus, each participant completing the entire year of the pilot study in either the sleep-hygiene or bed-restriction condition was paid $555 for sleep studies and MSLTs and $330 for office visits, for a total of $885. Participants in the non-intervention control condition were paid $185 for each set of sleep studies at the start and finish of the year, for a total of $370.
Analyses.
One night of baseline sleep was recorded to allow for first-night effects. Visually-scored sleep variables used the means of Nights 2 and 3 in each sleep series. EEG power data were based upon Night 2, and apnea-hypopnea index was based upon the means of Nights 1 and 2. There was a reduced n (reduced sample size) at the end of continuation (T3) and maintenance treatment (T4). The number of observations available at each time point are indicated on the figure for each variable (see Results section). Attrition was minimal and did not differ significantly between groups. We performed a Kruskal-Wallis test on the 8-week, 6-month, and 1-year change scores in order to determine whether the two intervention conditions had different effects over time. A Kruskal-Wallis test was also used to evaluate differential changes between participants in the two intervention groups and those in the archival (nonintervention) control group from baseline to 1-year follow-up.
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Inspecting the data for total recording period (TRP: Fig. 1 and Table 2 ), it appears that as a group the bed-restriction participants were compliant, because we set TRP to diary and because their TRPs were proportionately lower. Reducing time in bed did not affect time spent asleep, but we did detect significant change scores for sleep efficiency, percent delta sleep, and EEG spectral power (Table 3 ). Participants in the bed-restriction condition showed differentially greater improvement in sleep efficiency at 1 year, greater percentage of slow wave sleep at 6 months, and greater delta spectral power at each time point. The participants in the sleep-hygiene condition showed a trend toward improved mood on morning awakenings in the first 8 weeks. (See Fig. 2.) We also detected a trend toward shorter sleep latencies (MSLT) at 24 weeks in the bed-restricted groups (though not to a clinically significant degree), but no differential impact on subjective sleep quality (PSQI) or on apnea-hypopnea index. There were three participants in both groups who had an increase in the number of diary-reported naps at each of the follow-up time point from baseline.
|
|
|
|
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
). Otherwise, increases in longevity are likely to be associated primarily with prolongation of dependency and unsatisfactory quality of life. Any understanding of how to attenuate the sleep changes and impairment that contribute to functional morbidity and decreased quality of later life would have an impact not only on elderly adults themselves, but also on society, by reducing psychiatric morbidity, burden to families, and rates of institutionalization. Impaired sleep in elderly care recipients contributes to caregiver strain (Rabins, Mace, and Lucas 1982; Sanford 1975) and increases the probability of placement in long-term care (Pollak, Perlick, Linsner, Wenston, and Hsieh 1990). The data from the current pilot study show the feasibility of implementing a yearlong program of sleep-hygiene education and modest restriction of time in bed, in elderly participants who are good sleepers and continuing to do well but who can also be expected to become increasingly vulnerable to sleep disruption and functional impairment as they advance into the final years of life. Our data suggest that both approaches have something positive to offer elderly participants. Thus, sleep-hygiene education was associated with initial improvements in a sense of well-being on awakening, whereas modest sleep restriction led to sustained improvements in sleep continuity and depth—those aspects of sleep most vulnerable to disruption in later life. These data suggest that age-related declines in sleep continuity and depth can be slowed and, perhaps, reversed. Because of lower baseline sleep efficiencies in some of the bed-restriction participants, we cannot exclude the possibility of a regression to the mean.
The data also suggest a favorable risk/benefit ratio. Elderly persons who undertake sleep-restriction therapy for insomnia sometimes report excessive daytime sleepiness (Friedman et al. 1991), and that was occasionally observed in our participants and reflected in the MSLT data, which showed a trend to shorter sleep latencies in the bed-restriction condition (not clinically significant). The increase was not judged to be clinically significant because the mean sleep latency continued to be well above 10 minutes. The restriction level of time in bed used here (30 minutes nightly) was less than that typically employed in sleep-restriction therapy of chronic insomnia (2–3 hours nightly).
We believe that these data show the feasibility of testing a much more interesting and important hypothesis, namely, that maintaining good sleep consolidation and depth will ultimately contribute to preventing declines in functional status and thereby to the compression of functional morbidity among the old-old, the fastest growing segment of the population. The results also make sense and are in the direction that would invite a controlled test of this hypothesis. Finally, our findings are broadly consistent with those of Morin and colleagues, who have recently shown the efficacy of cognitive behavioral interventions in improving sleep quality in elderly insomniac patients over a 1-year period (Morin, Colecchi, Stone, Sood, and Brink 1999).
Limitations of the current study include the relatively small sample size and the absence of a true nonintervention control group that the archival control only partially addresses. A more definitive randomized clinical trial will require both a larger sample size and a longer period of treatment to determine the preventive value of either intervention (good sleep practices alone vs good sleep practices plus restriction of time in bed) in maintaining good mental and physical health in the final years of life. There is always a danger in extrapolating from results of specific treatments with clinical patients and applying the same treatment approach to a nonclinical, nonpatient population. This danger is perhaps even greater when we do not know exactly what sleep performance should be considered optimal in the age group studied. The pilot data presented in this report suggest that a fixed "dose" of bed restriction may have value in slowing or even reversing age-dependent declines in sleep quality and daytime well-being. Further controlled assessment appears warranted by these data.
|
Acknowledgments |
|---|
Received for publication December 7, 1999. Accepted for publication May 25, 2000.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. F. van den Berg, H. M. E. Miedema, J. H. M. Tulen, A. K. Neven, A. Hofman, J. C. M. Witteman, and H. Tiemeier Long Sleep Duration is Associated With Serum Cholesterol in the Elderly: The Rotterdam Study Psychosom Med, November 1, 2008; 70(9): 1005 - 1011. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
