| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Circadian, The contribution of the circadian timing
system to the age-related advance of sleep-wake timing was investigated
in two experiments. In a constant routine protocol, we found that the
average wake time and endogenous circadian phase of 44 older subjects
were earlier than that of 101 young men. However, the earlier circadian phase of the older subjects actually occurred later relative to their
habitual wake time than it did in young men. These results indicate
that an age-related advance of circadian phase cannot fully account for
the high prevalence of early morning awakening in healthy older people.
In a second study, 13 older subjects and 10 young men were scheduled to
a 28-h day, such that they were scheduled to sleep at many circadian
phases. Self-reported awakening from scheduled sleep episodes and
cognitive throughput during the second half of the wake episode varied
markedly as a function of circadian phase in both groups. The rising
phase of both rhythms was advanced in the older subjects, suggesting an
age-related change in the circadian regulation of sleep-wake propensity. We hypothesize that under entrained conditions, these age-related changes in the relationship between circadian phase and
wake time are likely associated with self-selected light exposure at an
earlier circadian phase. This earlier exposure to light could account
for the earlier clock hour to which the endogenous circadian pacemaker
is entrained in older people and thereby further increase their
propensity to awaken at an even earlier time.
aging; circadian rhythm; sleep; insomnia; core body temperature; forced desynchrony
A COMMON FEATURE OF AGING is the advance of habitual
bedtimes and wake times to earlier hours (1, 27, 31, 35).
The complaint of early morning awakening, rare among young adults but
reported by 15-20% of the older population (18, 25, 26), probably
represents an extreme of this advance in sleep-wake timing. These
changes in sleep-wake timing have been hypothesized to be related to
age-related changes in the endogenous circadian pacemaker (7, 32, 42).
A number of studies have reported that the timing of human circadian
rhythms changes with age (24, 37-39, 41), finding that the daily
rhythm of body temperature, a commonly used marker of the circadian
pacemaker's phase and amplitude, advances to an earlier hour in older
persons (30, 42). However, many of those studies measured body
temperature in subjects on a sleep-wake routine, and, because the
endogenous circadian rhythm of body temperature is masked by changes in
posture, activity, and sleep-wake state, the finding of an earlier
timing of the temperature rhythm in older subjects studied under
ambulatory conditions could have been a direct result of
thermoregulatory responses evoked by the earlier timing of sleep and
wakefulness in those same older subjects.
Therefore, several years ago, we carried out a study comparing the
endogenous component of the circadian temperature rhythm in young and
older subjects using the constant routine (CR) procedure (9). This
procedure minimizes the influence of factors known to mask the
endogenous component of the circadian body temperature rhythm, such as
activity, postural changes, and changes in sleep-wake state. We
observed age-related changes in the endogenous component of the core
body temperature cycle and found that the endogenous temperature nadir
of the older subjects occurred at an earlier clock hour than that of
the young adults. We also observed a correlation between habitual wake
time and endogenous circadian temperature phase. Those findings raised
the question of causality. Was the earlier timing of the temperature
minimum an indirect result of age-related differences in the timing of
light exposure (and hence circadian entrainment) due to the earlier
habitual sleep and wake times of the older subjects, or were the
earlier habitual sleep and wake times of the older subjects a
consequence of the earlier circadian phase of entrainment?
Since that time, we have carried out two new studies in an attempt to
investigate further the role of the endogenous circadian pacemaker in
the regulation of early morning awakening in older people. The first
study represents an extension of our earlier report, carried out on a
larger and more highly screened group of young and older subjects, who
maintained a regular but self-selected sleep-wake schedule for 3 wk
before study. We used the CR technique to unmask the endogenous
component of the core body temperature cycle to assess the relationship
between body temperature, a marker of the circadian pacemaker, and the
timing of the habitual sleep-wake schedule. The second study was
carried out in an attempt to address the question of causality. This
second study used the forced desynchrony technique, in which subjects
were scheduled to a 28-h sleep-wake cycle, resulting in sleep episodes
distributed uniformly across the circadian cycle, with each commencing
after the same duration of prior wakefulness. By assessing
self-reported awakening after each sleep episode while experimentally
manipulating the phase relationship between the sleep-wake cycle and
circadian rhythmicity, this study allowed us to compare self-assessed
awakening time between young and older subjects across the entire range
of circadian phases. Additionally, a neurobehavioral test of cognitive
throughput was administered during waking episodes in the same study to
enable us to compare the circadian rhythm of waking performance between young and older subjects.
Subjects
ABSTRACT
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References
INTRODUCTION
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References
MATERIALS AND METHODS
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References
All subjects were healthy, as assessed during a complete physical examination before study. Medical screening also included clinical biochemical tests on blood and urine, electrocardiogram, chest radiograph (older subjects only), psychological screening tests (Minnesota Multiphasic Personality Inventory and Beck Inventory), and a screening interview with a clinical psychologist for subjects in experiment 2. Subjects were instructed to abstain from caffeine, nicotine, alcohol, and all medications during the 3 wk before study; on admission to the laboratory, comprehensive toxicological analysis of their urine was performed to verify that they were drug free at the time of study.
All subjects were without sleep complaint by history. Older subjects
were evaluated for sleep disorders by polysomnography before admission,
and volunteers with evidence of sleep apnea [apnea/hypopnea index
(AHI) 
10] or periodic leg movements associated with arousals
[periodic leg movement index (PLMI) 20] were not empaneled. Average AHI of the older subjects empaneled into the studies
was 3.34 ± 2.43 (range 0.2-8.51), and average PLMI was 4.91 ± 4.36 (range 0.4-19.6).
No subject reported a history of rotating shift work or regular night work in the 3 yr before study or travel across more than two time zones in the 3 mo before study. Subjects were instructed to keep a regular but self-selected sleep-wake schedule and maintain an 8-h time in bed each night for at least 3 wk before entering the laboratory for study and to record these self-selected sleep and wake times each day. They wore a wrist activity monitor for 1 wk before admission to the laboratory, and on admission to the laboratory the activity data were checked against their self-reported sleep and wake times to verify compliance with this regular schedule. Only those volunteers whose activity data indicated that they had achieved their target bedtimes and wake times (±0.5 h) on at least 12 of the 14 bedtimes and wake times during the week before study were empaneled for participation in the protocol. Data from three young men who were empaneled into the study were not included in the present analysis because they failed to meet the prestudy screening criteria due to missing or incomplete prestudy sleep-wake logs.
All protocols were reviewed and approved by the Human Research Committee of the Brigham and Women's Hospital. Before beginning his or her study, each subject gave written informed consent.
Experimental Protocols
After three baseline adaptation days, consisting of 16 h of wakefulness and an 8-h sleep episode scheduled according to each subject's habitual bedtime and wake time, a 26- to 53-h CR was carried out to assess the endogenous circadian phase of the core body temperature rhythm (see Fig. 1 and days 1-4 of Fig. 2). The CR, which is described in detail elsewhere (2, 3, 10), is a refinement of a procedure first proposed by Mills et al. (28) and is designed to minimize or distribute evenly across the circadian cycle factors known to mask the endogenous component of the core body temperature rhythm. Thus, during the CR, subjects were restricted to wakeful bed rest in a semirecumbent posture, in dim indoor light (~10-15 lx), with food intake distributed in hourly snacks. This initial baseline and CR are referred to as experiment 1.
|
|
A subset of 10 young subjects and 13 older subjects (9 men, 4 women) participated in a forced desynchrony protocol (13) after their CR, during which time subjective sleep quality data were obtained. This protocol (referred to as experiment 2) used an imposed sleep-wake cycle length of 28 h, which is beyond the range of entrainment of the human circadian pacemaker. This resulted in sleep episodes evenly distributed across circadian phases after a common length of prior wakefulness and allowed us to examine self-reported early awakening at many different circadian phases. During the 18-23 cycles of forced desynchrony, subjects were scheduled to a 28-h "day," with sleep episodes scheduled to begin 4 h later each day and continue for one-third of each cycle (9 h 20 min; see Fig. 2). During the remainder of each 28-h day (18 h 40 min), subjects were awake and ambulatory within their study room.
Experimental Conditions
During the three baseline days, ambient light intensity was ~150 lx (ordinary indoor room light), during the CRs and during waking hours on the 28-h forced desynchrony days it was ~10-15 lx (dim indoor light), and during scheduled sleep episodes it was <0.03 lx (darkness).Throughout his or her study, each subject remained in a study room that was free of external time cues. The study rooms had no windows, clocks, or other indication of time of day, and the subjects were not permitted to watch television or listen to the radio. Technicians were trained to avoid discussion of time-of-day information and did not wear watches.
Data Collection
Sleep-wake logs were maintained by each subject for at least 3 wk before entering the laboratory. Subjects were instructed to record the time of retiring, the estimated time to fall asleep, the time of awakening, and the time of arising from bed. The self-reported times of retiring and awakening from the week immediately before entering the laboratory were averaged for each subject, and the averaged times were used as the habitual bed and wake times in the present analysis. On admission to the laboratory, the average bed and wake times of each subject were adjusted slightly, if necessary, to equal 8.0 h and were used as the scheduled baseline bed and wake times.Core body temperature was collected continuously throughout each study using a rectal thermistor (Yellow Springs Instruments, Yellow Springs, OH).
After each scheduled wake time in the laboratory, subjects were asked to complete a postsleep questionnaire (PSQ) which contained a series of questions about the previous sleep episode. The PSQs from the 23 subjects in experiment 2 were analyzed; results from only one question ("Did your final awakening in the morning occur when the laboratory personnel came in to awaken you? If not, how much earlier?") are presented here.
During scheduled waking episodes, an assessment of neurobehavioral performance was carried out by administering a test of cognitive throughput approximately one time per hour. The pencil and paper test consisted of a series of pairs of two-digit numbers (22) and tested the speed of mental calculation of sums, a test of cognitive throughput. Subjects were allowed 4 min to sum as many of the pairs as possible, and tests were scored as number attempted per 4-min test. The cognitive throughput data from the 20 subjects in experiment 2 for whom the data were available are presented here.
Data Analysis
Endogenous circadian phase of the core temperature data collected during the CRs (experiment 1) was assessed by the maximum likelihood fit of a two-harmonic regression model with first-order autoregressive noise (3). The first 5 h of data and the final 30 min of data from the CR were excluded from analysis due to the masking effects of waking and changing posture at the beginning and end of the CR. The minimum of the fundamental and the first harmonic components of the model fit to the data were averaged, and this average was used as a phase reference marker, referred to as core body temperature minimum (CBTmin). Average waveforms of the CBT data for the third scheduled day and the CR were compiled by first taking each subject's minute-by-minute temperature data series and calculating the average value per hour, beginning with scheduled wake time on the third scheduled day. These hourly average values per subject were then averaged for the young and older groups so that average curves for each of the groups could be made.The variances of the young and older groups of subjects were compared using an F test. The means of the young and older groups of subjects were compared using an unpaired t-test after verification of variance homogeneity; in cases in which the variance was unequal, an approximate t was calculated. All statistical analyses were two tailed. Correlation analysis was performed using Pearson's correlation coefficient. Linear regression analysis was performed by minimizing the variance of the perpendicular distance to the line (21), a method that makes no assumption about which variable is dependent or independent.
In the subset of subjects who participated in experiment 2, the core temperature data from the entire study (excluding the 3 baseline days) was assessed for intrinsic circadian period using nonorthogonal spectral analysis. This analysis takes into account the 28-h periodicity in the data resulting from the sleep-wake schedule and then simultaneously searches for a periodicity in the circadian range (search range 20-30 h). From this estimate of intrinsic circadian period and the phase of the fitted minimum, a circadian phase (from 0 to 359°) was assigned to each minute of the study, with 0° corresponding to the minimum of the waveform fit to the entire temperature data series.
The intrinsic circadian period of each subject in experiment 2 was used to calculate the circadian phase of every scheduled wake time. Each PSQ was then assigned a circadian phase corresponding to the scheduled time of awakening on the day it was administered. Data from each subject were binned in 45° bins (3 circadian hours), and a mean early awakening value per bin was calculated for each subject. Mean data for the two groups (young and older) were then calculated per bin.
Cognitive throughput was assessed for each of the 10 older and 10 young subjects in experiment 2 for whom data were available. The data were first normalized with respect to each individual's mean and SD (mean of 0 and SD of 1) from the forced desynchrony portion of the study. Each test of cognitive throughput was assigned a circadian phase corresponding to the time the test began, as described for the PSQ data above, and a time since awakening based on elapsed time since scheduled wake time. Data from the first and second one-half of each waking day were analyzed separately to examine the effect of increased amounts of wakefulness on task performance in the two age groups. Data from each subject were then binned in 45° bins (3 circadian hours), and a mean value per bin was calculated for each subject. Mean data for the young and older groups were then calculated per bin.
Responses to the PSQ question were analyzed using a two-factor ANOVA to assess the difference between the two age groups (factor group) and the difference between circadian phases within each age group (factor phase), whereas the cognitive throughput data were analyzed using a three-factor ANOVA to assess the effect of length of time awake (factor one-half of waking episode). The SAS program (SAS Institute, Cary, NC) was used for statistical analysis.
| |
RESULTS |
|---|
|
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References
|
|---|
Comparison of Older Men and Women
The ages, bedtimes, wake times, and sleep lengths were not found to be significantly different between older men and women, although given the variances and the sample size, we only had the power (
= 0.10) to
detect as significant a difference of 0.75 h between the bed and wake
times. We found no significant difference in the timing of
CBTmin between the older men and
women, although we only had the power ( = 0.10) to detect as
significant a difference of 1.5 h. Although the interval between
CBTmin and reported wake time was
shorter in the older men than in the older women, the difference was
not statistically significant, although we only had the power ( = 0.10) to detect as significant a difference of 1.3 h between the
groups.
For these reasons, results from the older men and women were pooled for all subsequent comparisons with results from the young subjects.
Experiment 1
Comparisons between the young and older subjects revealed that the average reported wake time from the week before entering the laboratory of the older group was more than an hour earlier than that of the young group, with correspondingly earlier bedtimes (see Table 1). Although all subjects were instructed to maintain an 8-h sleep episode each night during the 3 wk before entering the laboratory, the reported sleep length of the older subjects was on average 15 min less than for the young subjects (P < 0.01).
|
When the core body temperature data from the scheduled day and night before the CR were averaged with respect to clock hour and the average waveforms for the young and older groups were compared using t-tests on the hourly average values, the core body temperature data did not differ significantly during most of the daytime hours (from 0600 to 2100, only hours 0800 and 0900 were significantly different; see Fig. 3). However, during the late evening and nighttime hours, the temperature data of the older subjects were significantly higher than that of the young men (all hours from 2200 to 1000, P < 0.05), reaching a nadir earlier, probably reflecting the earlier sleeping times of the older subjects.
|
When the average waveforms of the core body temperature data were aligned with respect to habitual wake time rather than actual clock hour and the young and older groups were again compared, the curves did not differ significantly during the daytime hours on the scheduled day (see Fig. 4A). However, within 2 h of the beginning of the scheduled sleep episode, the average temperature data for the older subjects began to differ significantly from that of the young subjects and continued to be significantly higher throughout the rest of the scheduled sleep episode (all hours from bedtime + 3 h to bedtime + 8 h, P < 0.02). From these masked data, it appears that the older group of subjects reached a nadir of temperature earlier in the sleep episode than did the young group. Under the unmasking conditions of the CR (see Fig. 4B), the average temperature data for the older subjects are similar to those of the young subjects during the regular waking hours but again begin to differ from those of the young subjects during the hours corresponding to the usual sleep episode (time period from 2 h after usual bedtime to 7 h after usual bedtime), despite the fact that the subjects did not change posture, remained in constant dim light, and were kept awake by a technician. It is only under the constant conditions of the CR procedure that it becomes apparent that the nadir of the temperature rhythm in the older subjects actually occurs later within the scheduled sleep episode than it does for the young subjects, leading to the wake time of the older subjects occurring at an earlier circadian phase.
|
When a two-harmonic regression model was fit to the CR temperature data for each subject, the fitted minimum of the model (CBTmin) occurred on average at an earlier clock hour in the older subjects compared with the young subjects (P < 0.04) (see Table 1). Both reported sleep length and CBTmin showed significantly greater variability in the older group than in the young subjects (sleep length, F43,100 = 2.32, P < 0.001; CBTmin, F43,100 = 1.85, P < 0.02).
There was a significant positive correlation between habitual wake time and CBTmin when young and older subjects were combined (r = 0.61, P < 0.0001; see Fig. 5); when analyzed separately, this relationship continued to be significant (young, r = 0.64, P < 0.0001; older, r = 0.57, P < 0.0001). When a linear regression minimizing the variance of the perpendicular distance to the line was fit to the data, the slopes fit to the young and older groups were significantly different (young, 0.471 ± 0.05; older, 0.266 ± 0.06; P < 0.001), indicating that the relationship between usual wake time and the timing of the circadian rhythm of core body temperature is not the same in the two groups.
|
The interval between CBTmin and habitual wake time was 0.6 h shorter for older subjects than for the young subjects (P < 0.05), indicating that the older subjects awaken sooner after their CBTmin than do the young subjects (see Fig. 6 and Table 1). Furthermore, there was a significantly higher variability in this measure in the older subjects (F43,100 = 2.21, P < 0.002, see Fig. 6).
|
Experiment 2
PSQs, baseline nights. Responses on the PSQ question from the baseline nights were evaluated for the overall baseline period (nights 1 + 2 + 3) as well as for each baseline night separately. There was no significant difference detected between the two age groups on any of the baseline nights (average of all 3 baseline nights, mean ± SD, young, 11.4 ± 20.9 min; older, 16.6 ± 18.2 min and range, young, 0-55 min; older, 0-60 min).
PSQs, forced desynchrony. There was a main effect of age in the level of self-assessed wake before scheduled wake time at nearly all circadian phases, with older subjects reporting longer overall durations of self-assessed early awakening when compared with young subjects (F15,1 = 14.50, P < 0.001; see Fig. 7).
|
Cognitive throughput during waking, forced desynchrony. There was both a circadian- and a wake-dependent variation in cognitive throughput in both groups of subjects, although the scores of the older subjects were lower than those of the young before normalization (mean ± SD, older subjects, 78.9 ± 25.9; young subjects, 107.1 ± 36.9; P < 0.07). Division of the 18.67-h waking day in half and examination of the circadian variation of cognitive throughput revealed a significant effect of both circadian phase (ANOVA for repeated measures, F7,126 = 8.2, P < 0.001) and one-half of waking day (F1,18 = 83.39, P < 0.0001). Repeated-measures ANOVA performed on the latter one-half of the waking day revealed a significant interaction between age and circadian phase (F7,126 = 2.13, P < 0.05), indicating that the effects of circadian phase are different in the two age groups in the latter one-half of the waking day. Post hoc t-tests between the two age groups during the latter one-half of the waking day revealed significant differences in the bins centered 3 and 6 h after the temperature minimum, at 45 and 90° (P < 0.01, see Fig. 8). The circadian waveforms of cognitive throughput during the latter one-half of the waking day followed similar courses in the two age groups from ~180-0° (corresponding to late afternoon to early morning under entrained conditions). Thereafter, the waveform of the older subjects reached a minimum in the bins centered just before and at the temperature minimum and then began to rise, whereas the waveform of the young men reached a minimum 3-6 h later, in the bin centered 3 h after the temperature minimum, and did not begin to rise until 6 h after the temperature minimum.
|
| |
DISCUSSION |
|---|
|
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References
|
|---|
Our analysis of reported sleep patterns and endogenous circadian phase in experiment 1 revealed that whereas the self-reported wake time of older subjects occurs at an earlier clock hour and at an earlier endogenous circadian phase than does that of young subjects, this endogenous circadian phase actually occurs later with respect to the habitual sleep episode in older subjects. Thus the relationship between the habitual sleep-wake cycle and entrained circadian phase is significantly different in older compared with young adults, as illustrated in Fig. 5. Experiment 1 represents one of the largest studies of endogenous circadian phase in healthy young and older subjects reported to date. Furthermore, we have used data collected in a CR procedure to account for differences in the timing of activity between the groups so that these findings are not confounded by exogenous masking effects (15). The importance of using unmasked data such as those collected in a CR can be seen in Figs. 3 and 4, in which the true phase and the phase relationship between the core body temperature rhythm and the timing of the habitual sleep episode are obscured when subjects follow a scheduled sleep-wake routine (Figs. 3 and 4A) but are revealed under the constant behavioral conditions of the CR (Fig. 4B). Furthermore, the unmasking conditions of the CR revealed information about the reported age-related amplitude reduction of the circadian rhythm of core body temperature (9, 40, 42). Although the temperature data collected during the usual waking hours were quite similar in the young and older subjects (see Fig. 4), the unmasked CR temperature data of the older subjects did not fall as low during the usual nighttime hours as did that of the young men, even in the absence of sleep and postural changes. This suggests that the age-related reduction of core body temperature amplitude is not only independent of age-related changes in sleep but is due to changes in only the portion of the circadian cycle corresponding to habitual nighttime.
One explanation that has been suggested for the earlier endogenous circadian phase and usual wake time in older subjects would be an age-related shortening of the intrinsic period of the circadian pacemaker. Entrainment theory predicts that the phase angle of entrainment is dependent on intrinsic period (34), and an age-related shortening of intrinsic period has been reported for some rodents (32, 33). However, analysis of the intrinsic period of the core body temperature, plasma melatonin, and plasma cortisol data from the young and older subjects who participated in experiment 2 revealed that in these healthy subjects there is no age-related shortening of intrinsic circadian period (8). Furthermore, although we did observe a significant correlation between intrinsic period and endogenous circadian temperature phase in the young subjects who participated in experiment 2, that relationship was not found to be significant in the older subjects in that experiment. On the basis of these results, we conclude that variations in intrinsic period between age groups, at least among healthy individuals, is probably not the primary cause of the observed difference in phase.
An age-related change in the phase-response curve (PRC) to light could also explain the earlier circadian phase in older individuals. A reduced phase-delay and/or an increased phase-advance portion of the PRC in older people could produce the age-related advance of circadian phase observed in experiment 1. However, in a series of experiments carried out in our laboratory, we observed no significant difference in phase delays and smaller phase advances in response to three cycles of 5-h exposures to bright indoor light in older subjects (23). Those preliminary findings do not support an age-related change in the shape of the PRC as a explanation for the earlier entrained CBTmin observed in experiment 1. However, it is unclear whether there are age-related differences in the response to single pulses of light or to light of lower intensity that may be more relevant to entrainment under normal conditions.
Results obtained from experiment 2 extend the findings from experiment 1 and provide some insight as to mechanisms underlying the age-related changes in entrained phase and the timing of awakening. The forced desynchrony technique, in which sleep and wake episodes are scheduled to occur at all circadian phases, revealed that the circadian drive for wakefulness is strongest in the evening hours in both young and older subjects and that the overall shape of the circadian rhythms of early morning awakening and cognitive throughput are similar in both groups. However, the early awakening data from experiment 2 also revealed that older subjects report more difficulty maintaining sleep when it is scheduled at abnormal circadian phases than do young subjects. Furthermore, those data demonstrated that there is a narrower window within the circadian cycle during which older subjects report the ability to maintain sleep than there is for young subjects. Analysis of objective (polysomnographically recorded) sleep data from the same forced desynchrony studies is consistent with the narrowing of the circadian phase window, during which the sleep episode can be maintained in older subjects, found in the analysis of these subjective reports (14).
What is perhaps most remarkable about these findings is that older subjects report being unable to sustain sleep just after the temperature nadir, at the circadian phase of peak sleep propensity in young subjects (6, 13). This finding cannot be explained by factors that would affect the phase of entrainment, because the comparison of awakening time between young and older subjects was done with reference to each subject's own circadian phase.
Within the framework of our current understanding of sleep timing (11, 13, 17), there are several possible explanations as to why these healthy older subjects report being unable to remain asleep after the minimum of the core body temperature rhythm has occurred. One possibility is an age-related change in the homeostatic process regulating sleep. This could occur by a variety of means, including a lower sleep pressure at sleep onset (such as a slower buildup of sleep pressure during the waking day coupled with a decreased threshold to initiate sleep) or by a more rapid decline of homeostatic sleep drive during sleep, resulting in spontaneous awakening after a shorter duration of sleep. However, these explanations alone cannot account for why the difference in wakefulness at the end of the scheduled sleep episode between the young and older subjects is quite pronounced at some circadian phases and is absent at others. Furthermore, the similar age-related change observed in the circadian waveform of the neurobehavioral performance data collected during wake episodes indicates that age-related changes in sleep dynamics are not likely to fully account for why older subjects report being unable to remain asleep after the minimum of the core body temperature rhythm.
A second possibility is that the observed increase in wakefulness during scheduled sleep in older people, which is particularly prominent in the early morning hours under entrained conditions, is related to changes in circadian signals that contribute to sleep consolidation. Because the circadian drive for sleep has been shown to be most pronounced in the early morning hours (13), an age-related reduction in the circadian drive for sleep could explain why differences between young and older people are especially large at some phases and not at others. Furthermore, such a mechanism could account for both the sleep findings and the findings from the neurobehavioral performance data collected during wakefulness. Although the amplitudes of the sleep latency and sleep propensity rhythms have been shown to vary little with age in healthy subjects (19, 36), the increased level of core body temperature during the usual nighttime hours observed in the older subjects in the present study may be correlated with an age-related decrease in the circadian drive for sleep during that part of the circadian cycle. Furthermore, in both the subjective sleep data and the cognitive throughput data, the overall circadian waveform of the older subjects showed a narrower trough due to changes in the ascending portion of the curve. These data suggest that there is a change in the relationship between the circadian rhythm of core body temperature and the underlying circadian rhythm of wake propensity that is manifested not only during sleep but also during waking when cognitive throughput is assessed.
It has been suggested that the primary role of the circadian pacemaker, the suprachiasmatic nuclei (SCN), in sleep regulation is to promote wakefulness at particular circadian phases (17). This model suggests that at phases when the SCN are not actively promoting wakefulness, sleep will occur if sufficient homeostatic drive has built up. Our current findings in older subjects, together with our previous findings in young adult subjects (12, 13, 16), suggest that the circadian system (perhaps via the SCN) may also promote sleep, in particular in the early morning hours, i.e., close to the CBT nadir, and that this circadian drive for sleep may change with age.
Another possibility is that there is an age-related change in the interaction between the homeostatic and circadian processes. This is supported by preliminary analyses of the objective sleep data from the forced desynchrony studies in experiment 2 (14). We found that whereas sleep latency and sleep propensity during the first one-half of the scheduled sleep episode did not differ significantly between young and older subjects, there were significant differences in the latter part of the scheduled sleep episode. The older subjects slept for a shorter duration, with the reduction in total sleep time mainly occurring in the final 2 h of the scheduled sleep episode and resulting in a much lower sleep efficiency. In the present study, the differences in the neurobehavioral performance rhythm between the young and older subjects were observed only in the latter one-half of the waking day, supporting the notion that there is an age-related change in the interaction between the homeostatic and circadian processes. Although a change in the interaction between the sleep and circadian processes is supported by our data, the nature of this potential change remains unknown.
Our present findings are consistent with data from two recent studies in middle-aged subjects. In a study of simulated jet lag, Moline and colleagues (29) reported that although the temperature rhythm of middle-aged subjects adapted as quickly as that of young adults, the subjective and objective sleep complaints of the middle-aged group were higher. In another study of simulated shift work, Campbell (4) reported that although bright light was effective in shifting the timing of the core body temperature rhythm of middle-aged subjects, their sleep and waking performance was still impaired, leading him to speculate that middle-aged subjects might be less "phase tolerant" than young subjects. Our data indicate that the basis of such reported age-related insomnia at adverse circadian phases is likely to be the narrower circadian phase window within which older people can maintain consolidated sleep.
The results of this study have important consequences for understanding the phase of circadian entrainment of older subjects. Our present findings, together with other recent findings from our laboratory (2, 20), suggest a possible mechanism for the earlier clock hour of awakening observed in older subjects. If older subjects find it more difficult to remain asleep after the nadir of the temperature cycle than do young subjects and awaken at an earlier circadian phase, as our data indicate, they will have more exposure to light during the portion of the circadian cycle when humans are most sensitive to phase-advance shifting (10). Even if they remain indoors, such indoor light intensity levels are now known to exert significant phase-shifting effects (2) and could result in the resetting of the circadian pacemaker to an earlier clock hour, even in the absence of an age-related difference in the PRC to light, intrinsic period, or exposure to outdoor levels of light.
Our findings indicate that the sleep of older people can be maintained at high efficiency only within a narrower window of a phase advance-shifted circadian cycle. This phase relationship is such that the modest age-related phase advance, together with the age-related change in the timing of sleep with respect to circadian phase can have a large impact on the ability to consolidate sleep in the morning hours. Given the significantly greater variability in the phase relationship between sleep and circadian rhythms in older subjects under baseline conditions, the impact on disrupted sleep in the latter part of the sleep episode is further compounded. Our present results imply that a treatment regimen that seeks to establish and maintain a stable and optimal clock hour of entrainment and internal phase relationship between sleep and circadian rhythms in older people should lead to reduced early morning awakening and a longer sleep episode (5).
| |
ACKNOWLEDGEMENTS |
|---|
We thank the subject volunteers for their participation in the studies, the subject recruitment staff of our laboratory (M. Martens, G. Spelman, A. J. Chiasera, J. J. Daly, J. Stromsten, D. Margolis, J. Kao, R. McCarley, and B. Rich), the supervisory staff of the General Clinical Research Center Environmental Scheduling Facility (ESF) and Intensive Physiological Monitoring (IPM) Unit (T. L. Shanahan, E. B. Martin, Jr., A. E. Ward, D. W. Rimmer, G. Jayne, and S. Driscoll), the technical staff of the ESF and IPM for subject monitoring and data collection, J. M. Zeitzer, D. B. Boivin, and D. W. Rimmer, who carried out some of the studies included in the present analysis, M. E. Jewett and K. P. Wright for comments on the manuscript, J. M. Ronda, J. F. Mitchell, and A. McCollom for assistance with computer software and hardware, and E. F. Hall and E. J. Silva, who assisted with data processing and analysis.
| |
FOOTNOTES |
|---|
This study was supported in part by Grants R01 AG-06072 and P01 AG-09975 from the National Institute on Aging; Grant R01 MH-45130 from the National Institute of Mental Health; and Grants NAG9-524 and NAGW-4033 from the National Aeronautics and Space Administration. These studies were carried out in a General Clinical Research Center supported by Grant M01 RR-02635 from the National Center for Research Resources.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Address for reprint requests: J. F. Duffy, Circadian, Neuroendocrine and Sleep Disorders Section, Brigham and Women's Hospital, Harvard Medical School, 221 Longwood Ave., 438c, Boston, MA 02115-5817.
Received 11 March 1998; accepted in final form 21 July 1998.
| |
REFERENCES |
|---|
|
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References
|
|---|
1.
Bliwise, D. L.
Review: sleep in normal aging and dementia.
Sleep
16:
40-81,
1993[Medline].
2.
Boivin, D. B.,
J. F. Duffy,
R. E. Kronauer,
and
C. A. Czeisler.
Dose-response relationships for resetting of human circadian clock by light.
Nature
379:
540-542,
1996[Medline].
3.
Brown, E. N.,
and
C. A. Czeisler.
The statistical analysis of circadian phase and amplitude in constant-routine core-temperature data.
J. Biol. Rhythms
7:
177-202,
1992
4.
Campbell, S. S.
Effects of timed bright-light exposure on shift-work adaptation in middle-aged subjects.
Sleep
18:
408-416,
1995[Medline].
5.
Campbell, S. S.,
D. Dawson,
and
M. W. Anderson.
Alleviation of sleep maintenance insomnia with timed exposure to bright light.
J. Am. Geriatr. Soc.
41:
829-836,
1993[Medline].
6.
Carskadon, M. A.,
and
W. C. Dement.
Distribution of REM sleep on a 90 minute sleep-wake schedule.
Sleep
2:
309-317,
1980[Medline].
7.
Czeisler, C. A.,
J. S. Allan,
S. H. Strogatz,
J. M. Ronda,
R. Sánchez,
C. D. Ríos,
W. O. Freitag,
G. S. Richardson,
and
R. E. Kronauer.
Bright light resets the human circadian pacemaker independent of the timing of the sleep-wake cycle.
Science
233:
667-671,
1986
8.
Czeisler, C. A.,
J. F. Duffy,
T. L. Shanahan,
E. N. Brown,
J. F. Mitchell,
D.-J. Dijk,
D. W. Rimmer,
J. M. Ronda,
J. S. Allan,
J. S. Emens,
and
R. E. Kronauer.
Reassessment of the intrinsic period (
) of the human circadian pacemaker in young and older subjects (Abstract).
Sleep Res.
24A:
505,
1995.
9.
Czeisler, C. A.,
M. Dumont,
J. F. Duffy,
J. D. Steinberg,
G. S. Richardson,
E. N. Brown,
R. Sánchez,
C. D. Ríos,
and
J. M. Ronda.
Association of sleep-wake habits in older people with changes in output of circadian pacemaker.
Lancet
340:
933-936,
1992[Medline].
10.
Czeisler, C. A.,
R. E. Kronauer,
J. S. Allan,
J. F. Duffy,
M. E. Jewett,
E. N. Brown,
and
J. M. Ronda.
Bright light induction of strong (Type 0) resetting of the human circadian pacemaker.
Science
244:
1328-1333,
1989
11.
Daan, S.,
D. G. M. Beersma,
and
A. A. Borbély.
Timing of human sleep: recovery process gated by a circadian pacemaker.
Am. J. Physiol.
246 (Regulatory Integrative Comp. Physiol. 15):
R161-R178,
1984
12.
Dijk, D.-J.,
and
C. A. Czeisler.
Paradoxical timing of the circadian rhythm of sleep propensity serves to consolidate sleep and wakefulness in humans.
Neurosci. Lett.
166:
63-68,
1994[Medline].
13.
Dijk, D.-J.,
and
C. A. Czeisler.
Contribution of the circadian pacemaker and the sleep homeostat to sleep propensity, sleep structure, electroencephalographic slow waves, and sleep spindle activity in humans.
J. Neurosci.
15:
3526-3538,
1995[Abstract].
14.
Dijk, D.-J.,
J. F. Duffy,
E. Riel,
and
C. A. Czeisler.
Altered interaction of circadian and homeostatic aspects of sleep propensity results in awakening at an earlier circadian phase in older people (Abstract).
Sleep Res.
26:
710,
1997.
15.
Dijk, D.-J.,
T. L. Shanahan,
and
C. A. Czeisler.
Individual differences and non-additive effect of rest-activity cycle and circadian melatonin phase on core body temperature. (Abstract)
Sleep Res.
26:
709,
1997.
16.
Dijk, D.-J.,
T. L. Shanahan,
J. F. Duffy,
J. M. Ronda,
and
C. A. Czeisler.
Variation of electroencephalographic activity during nonrapid eye movement and rapid eye movement sleep with phase of circadian melatonin rhythm in humans.
J. Physiol. (Lond.)
505:
851-858,
1997[Medline].
17.
Edgar, D. M.,
W. C. Dement,
and
C. A. Fuller.
Effect of SCN lesions on sleep in squirrel monkeys: evidence for opponent processes in sleep-wake regulation.
J. Neurosci.
13:
1065-1079,
1993[Abstract].
18.
Foley, D. J.,
A. A. Monjan,
S. L. Brown,
E. M. Simonsick,
R. B. Wallace,
and
D. G. Blazer.
Sleep complaints among elderly persons: an epidemiologic study of three communities.
Sleep
18:
425-432,
1995[Medline].
19.
Haimov, I.,
and
P. Lavie.
Circadian characteristics of sleep propensity function in healthy elderly: a comparison with young adults.
Sleep
20:
294-300,
1997[Medline].
20.
Jewett, M. E.,
D. W. Rimmer,
J. F. Duffy,
E. B. Klerman,
R. E. Kronauer,
and
C. A. Czeisler.
The human circadian pacemaker is sensitive to light throughout subjective day without evidence of transients.
Am. J. Physiol.
273 (Regulatory Integrative Comp. Physiol. 42):
R1800-R1809,
1997
21.
Keeping, E. S.
Introduction to Statistical Inference. Princeton, NJ: Van Nostrand, 1962.
22.
Klein, K. E.,
H.-M. Wegmann,
G. Athanassenas,
H. Hohlweck,
and
P. Kuklinski.
Air operations and circadian performance rhythms.
Aviat. Space Environ. Med.
47:
221-230,
1976[Medline].
23.
Klerman, E. B.,
J. F. Duffy,
R. E. Kronauer,
and
C. A. Czeisler.
Altered phase resetting characteristics in the elderly (Abstract).
Sleep Res.
24A:
518,
1995.
24.
Lieberman, H. R.,
J. J. Wurtman,
and
M. H. Teicher.
Circadian rhythms of activity in healthy young and elderly humans.
Neurobiol. Aging
10:
259-265,
1989[Medline].
25.
Mant, A.,
and
E. A. Eyland.
Sleep patterns and problems in elderly general practice attenders: an Australian survey.
Community Health Stud.
12:
192-199,
1988[Medline].
26.
McGhie, A.,
and
S. M. Russell.
The subjective assessment of normal sleep patterns.
J. Ment. Sci.
108:
642-654,
1962.
27.
Miles, L. E.,
and
W. C. Dement.
Sleep and aging.
Sleep
3:
119-220,
1980.
28.
Mills, J. N.,
D. S. Minors,
and
J. M. Waterhouse.
Adaptation to abrupt time shifts of the oscillator(s) controlling human circadian rhythms.
J. Physiol. (Lond.)
285:
455-470,
1978
29.
Moline, M. L.,
C. P. Pollak,
T. H. Monk,
L. S. Lester,
D. R. Wagner,
S. M. Zendell,
R. C. Graeber,
C. A. Salter,
and
E. Hirsch.
Age-related differences in recovery from simulated jet lag.
Sleep
15:
28-40,
1992[Medline].
30.
Monk, T. H.,
D. J. Buysse,
C. F. Reynolds III,
D. J. Kupfer,
and
P. R. Houck.
Circadian temperature rhythms of older people.
Exp. Gerontol.
30:
455-474,
1995[Medline].
31.
Myers, B. L.,
and
P. Badia.
Changes in circadian rhythms and sleep quality with aging: mechanisms and interventions.
Neurosci. Biobehav. Rev.
19:
553-571,
1995[Medline].
32.
Pittendrigh, C. S.,
and
S. Daan.
Circadian oscillations in rodents: a systematic increase of their frequency with age.
Science
186:
548-550,
1974
33.
Pittendrigh, C. S.,
and
S. Daan.
A functional analysis of circadian pacemakers in nocturnal rodents. I. The stability and lability of spontaneous frequency.
J. Comp. Physiol. [A]
106:
223-252,
1976.
34.
Pittendrigh, C. S.,
and
S. Daan.
A functional analysis of circadian pacemakers in nocturnal rodents. IV. Entrainment: pacemaker as clock.
J. Comp. Physiol. [A]
106:
291-331,
1976.
35.
Prinz, P. N.,
M. V. Vitiello,
M. A. Raskind,
and
M. J. Thorpy.
Geriatrics: sleep disorders and aging.
N. Engl. J. Med.
323:
520-526,
1990[Medline].
36.
Richardson, G. S., M. A. Carskadon, E. J. Orav, and W. C. Dement. Circadian variation of
sleep tendency in elderly and young adult subjects. Sleep 5, Suppl. 2: S82-S94, 1982.
37.
Sharma, M.,
J. Palacios-Bois,
G. Schwartz,
H. Iskandar,
M. Thakur,
R. Quirion,
and
N. P. V. Nair.
Circadian rhythms of melatonin and cortisol in aging.
Biol. Psychiatry
25:
305-319,
1989[Medline].
38.
Sherman, B.,
C. Wysham,
and
B. Pfohl.
Age-related changes in the circadian rhythm of plasma cortisol in man.
J. Clin. Endocrinol. Metab.
61:
439-443,
1985[Abstract].
39.
Touitou, Y.,
J. Sulon,
A. Bogdan,
C. Touitou,
A. Reinberg,
H. Beck,
J.-C. Sodoyez,
E. Demey-Ponsart,
and
H. Van Cauwenberge.
Adrenal circadian system in young and elderly human subjects: a comparative study.
J. Endocrinol.
93:
201-210,
1982[Abstract].
40.
Vitiello, M. V.,
R. G. Smallwood,
D. H. Avery,
R. A. Pascualy,
D. C. Martin,
and
P. N. Prinz.
Circadian temperature rhythms in young adult and aged men.
Neurobiol. Aging
7:
97-100,
1986[Medline].
41.
Webb, W. B.
The measurement and characteristics of sleep in older persons.
Neurobiol. Aging
3:
311-319,
1982[Medline].
42.
Weitzman, E. D.,
M. L. Moline,
C. A. Czeisler,
and
J. C. Zimmerman.
Chronobiology of aging: temperature, sleep-wake rhythms and entrainment.
Neurobiol. Aging
3:
299-309,
1982[Medline].
43.
Zeitzer, J. M.,
R. E. Kronauer,
and
C. A. Czeisler.
Photopic transduction implicated in human circadian entrainment.
Neurosci. Lett.
232:
135-138,
1997[Medline].
This article has been cited by other articles:
|
|
 |
|
  C. E. Kline, J. L. Durstine, J. M. Davis, T. A. Moore, T. M. Devlin, M. R. Zielinski, and S. D. Youngstedt Circadian variation in swim performance J Appl Physiol, February 1, 2007; 102(2): 641 - 649. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Gronfier, K. P. Wright Jr., R. E. Kronauer, M. E. Jewett, and C. A. Czeisler Efficacy of a single sequence of intermittent bright light pulses for delaying circadian phase in humans Am J Physiol Endocrinol Metab, July 1, 2004; 287(1): E174 - E181. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. K. Baehr, C. I. Eastman, W. Revelle, S. H. L. Olson, L. F. Wolfe, and P. C. Zee Circadian phase-shifting effects of nocturnal exercise in older compared with young adults Am J Physiol Regulatory Integrative Comp Physiol, June 1, 2003; 284(6): R1542 - R1550. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D.-J. Dijk and S. W. Lockley Functional Genomics of Sleep and Circadian Rhythm: Invited Review: Integration of human sleep-wake regulation and circadian rhythmicity J Appl Physiol, February 1, 2002; 92(2): 852 - 862. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. F. Duffy, J. M. Zeitzer, D. W. Rimmer, E. B. Klerman, D.-J. Dijk, and C. A. Czeisler Peak of circadian melatonin rhythm occurs later within the sleep of older subjects Am J Physiol Endocrinol Metab, February 1, 2002; 282(2): E297 - E303. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. A. Czeisler, J. F. Duffy, T. L. Shanahan, E. N. Brown, J. F. Mitchell, D. W. Rimmer, J. M. Ronda, E. J. Silva, J. S. Allan, J. S. Emens, et al. Stability, Precision, and Near-24-Hour Period of the Human Circadian Pacemaker Science, June 25, 1999; 284(5423): 2177 - 2181. [Abstract] [Full Text]
|
|
| ||||||||||||||
, Older subjects
(n = 43); 
, young subjects
(n = 97); solid bar, usual sleep
episode of older subjects (mean ± 1 SD); open bar, usual sleep
episode of young subjects (mean ± 1 SD). Shown are data from
scheduled day
3, the 24-h period before the CR. Data
are plotted with respect to actual time of day. Temperature data for
the 24-h scheduled day before CR were first averaged in hourly bins for
each subject, beginning at wake time on that scheduled day; data for
all subjects in each group were then averaged per hour, and mean ± SE is shown.
0.27 ± 0.05 and older subjects =