Experimental Gerontology, Vol. 26, pp. 233--243, 1991
0531-5565/91 $3.00 + .00 Copyright © 1991 Pergamon Press plc
Printed in the USA. All rights reserved.
SLEEP AND CIRCADIAN RHYTHMS
TtMOTaY H. MONK Sleep Evaluation Center, Western Psychialric Institute and Clinic, University of Pittsburgh School of Medicine, 3811 O'Hara Street, Pittsburgh, Pennsylvania 15213
Abstract -- The human biological clock or "circadian system" serves the function of preparing the body and mind for restful sleep at some times of day and active wakefulness at others. The observed circadian rhythms result from three interacting processes: endogenous rhythm generation mechanisms, entrainment mechanisms to keep these rhythms "on track," and exogenous masking processes stemming from changes in environment and behavior. In both advanced age and extraterrestrial travel these processes, particularly the latter two, can be dramatically effected with a consequent disruption in sleep and daytime functioning. We are currently studying the age-related effects on these processes in healthy 80 year olds using two different protocols. The fast involved 36 h of constant wakeful bedrest which "unmasked" the endogenous component of circadian rhythms in various physiological and psycholngical functions. This experiment revealed little difference between nine old men and four young men in the endogenous body temperature rhythm, but revealed quite marked differences between the age groups in subjective activation and objective performance rhythms. A similar pattern occurred in the second experiment, which was a phase shift study comparing five old women with eight middle-aged men in their response to an acute 6 h phase advance in routine. Despite rather short-lived age-related differences in circadian rhythms, there were dramatic differences between the age groups in measures of sleep, mood, activation, and performance efficiency lasting for a week or more after the phase shift. Key Words: Sleep, drt+suliAnrhythms, bnmnn, age, phase shift
INTRODUCTION HOMO SAPIENS' development on a planet with a 24 h rotation has endowed the species with a biological clock or circadian system. This timekeeping system is endogenous and selfsustaining, serving to set the stage for sleep at some times of day and active wakefulness at others (Aschoff, 1981). This is done using "circadian rhythms" which comprise gradual fluctuations over the 24 h period in a broad range of physiological and psychological measures. There are three aspects to the human circadian system which are important from both aging and space perspectives, namely the rhythm generation processes, masking processes, and entrainment processes. These aspects will be discussed in turn, followed by a discussion of some recent studies of our own.
Correspondence to: Timothy H. Monk. 233
r H. MONK
The core of the circadian system is the endogenous circadian pacemaker IECPI which i~ a biological timekeeping system sending out a rhythmic 24 h, or approximate, signal (Conroy & Mills, 1970). This process is self-sustaining and continues to generate circadian rhythms even when activity and conditions are utterly constant (Froberg, 19771. In the human we cannot observe the workings of the ECP directly, but have to infer its status from the circadian rhythms we observe. Unfortunately, such rhythms provide an inexact estimate of the true ~,tatus of the circadian system as they comprise a mixture of both endogenous processes (from the circadian pacemaker) and exogenous ones (masking effects) which result simply from changes in the activity, posture and environment of the individual (Minors & Waterhouse, 19891. Thus. fi)r example, the fall in body temperature due to sleep and/or inactivity will be superimposed upon temperature changes occurring from differences in time of day. We, therefore, need protocols in which we can study the endogenous rhythms in situations where the masking effects are held constant and the rhythms observed all stem from internal processes. Such protocols will be described later. With advanced age, as with space travel, there can be a breakdown in the temporal coherence of the masking processes affecting an individual's circadian system. Although underlying reasons may differ between the two groups (e.g., illness, weakness, or boredom in the old, operational considerations in space) there are often intrusions of wakefulness during the "night" and sleep during the "'day" which can mask circadian rhythms in an apparently chaotic way. Thus it is important for both gerontological and space investigations that circadian masking processes are better understood. Also important are entrainment mechanisms. The human circadian system is precise, but inherently inaccurate, often having a natural period that is rather slower (usually by about an hour) than the 24 h days of the earth and society. We can see this using "'free-running'" experiments in which volunteers live tbr several weeks at a time under temporal isolation, with no external time cues whatsoever impinging on them. When allowed to choose their own bedtimes and waketimes, a "'day length" of approximately 25 h is often selected, that is, tbr healthy young men (Wever. 1979). Interestingly, this "natural" free-running period tends to decline with advancing age, with periods much closer to 24 h appearing in the few 70 and 80 year olds observed in such studies (Weitzman et al. 1982: Monk. 1989a). Such effects might at least partially explain some of the early evening sleepiness and early morning wakefulness of the elderly. Because of the non-24 h periodicity of the circadian system, there has to be an active process keeping the circadian system "'on track" at both the correct period (24 h in normal society) and at the correct orientation relative to the timing of sleep and wakefulness. We shall call this the entrainment mechanism. One should emphasize that the entrainment mechanism is always in use keeping the circadian system on track. Without it, unacceptable drift would occur, creating a mismatch in timing between the circadian system and the world around us. We particularly notice the impact of the entrainment mechanism with sudden changes in routine as occasioned by shiftwork and transmeridianal flight. After these acute shifts in routine the entrainment mechanism is challenged to get the circadian system back on track (Aschoff et al., 1975). Because it was only designed to handle minor adjustments of 60 min or so, this process is a slow one, taking several weeks to adjust completely to a continuous night work routine, for example (Knauth et al., 1978). The agents used by the entrainment mechanism in order to decide where to put "'day" and where to put " n i g h t " are the various time cues or zeitgebers that impinge on the individuaL.
SLEEP AND CIRCADIAN RHYTHMS
Some zeitgebers, such as daylight, are more potent than others, such as social contacts, activity, and meal patterns; but all have the potential for affecting the timing of the circadian system (Wever, 1988). In the area of circadian entrainment there are also important parallelisms between aging and space travel. Recent studies have shown that many housebound or institutionalized elderly people experience little or no outdoor illumination levels, denying them access to the physical zeitgebers of daylight and darkness which are so important in keeping the human circadian system "on-track." Such zeitgebers are also denied, of course, to the space traveler, and both groups can be likened to those experiencing prolonged undersea voyages in nuclear submarines, for whom circadian dysfunction has been well-documented (Colquhoun et al., 1978). In both groups normal social zeitgebers are also diminished, and an improved understanding of the role of both physical and social circadian zeitgebers would undoubtedly benefit both the treatment and care of the elderly and the successful operation of prolonged space missions. The remainder of this paper will discuss some of our research funded presently by the NIA (P01 AG06836) and previously by NASA (NCC2-253) which is concerned with rhythm generation and entrainment mechanisms and their effect on sleep and daytime performance. RHYTHM GENERATION We are using an unmasking protocol which was developed by Dr. Czeisler (Czeisler et al., 1985) from an original protocol of the late Dr. John Mills (Mills et al., 1978). This protocol involves 36 h of continuous wakeful bedrest. We are studying the effects of advanced age in the very healthy, that is, aged but robust, in order to address the simple question of whether the human clock runs down after 80 years (i.e., after about 29,000 circadian cycles) producing a weaker (lower amplitude) signal. We have compared nine men aged 80 years plus (~ = 82) with four young men aged 20-30 years of age (~ = 26). Currently we are only in the second year of funding, so results are preliminary. When we studied the "normal" circadian temperature rhythms of these subjects we saw that there did indeed appear to be a reduced amplitude in the old compared to the young (Fig. 1). On an individual basis, there appeared to be an amplitude reduction of about 35%. However, as discussed earlier, one needs to unmask circadian rhythms in order to determine true rhythm generation differences between old and young. The 36 h unmasking study was performed in complete temporal isolation, with the subject remaining wakeful but in bed, and meals replaced by hourly dietary supplements. This report will concentrate on body (rectal) temperature (sampled every minute), subjective vigilance and global affect (sampled every hour), and objective performance efficiency (sampled every two hours). From Fig. 2 it is clear that the endogenous rectal temperature rhythms of old and young were very similar in both timing and amplitude, indicating that in this robust elderly sample there appeared to be no age-related deterioration in the circadian rhythm generation processes. On an individual basis the mean rhythm amplitude was about equal in the two groups (0.28°C old vs. 0.27°C young). A different pattern, though, emerged for unmasked psychological rhythms of mood, activation and performance. There was much more evidence of age-related differences. In subjective vigilance (Fig. 3, top panel) which was assessed using a visual analogue scale technique (Monk, 1989b); the two age groups started off about equal, but the young group suffered much more of a deterioration during the night hours than did the old, a finding confirmed by technicians' observations. The young did, however, exhibit more of a recovery
372 36.g -~ OLO MALES
~ 36.6~h i 36.5.,<
Y O U N G MAI I:'$
~9 zl :3 TIME OF" DAY
Fig. 1. Circadian temperature rhythms from nine old males (solid line) and four young males (broken linel using ambulatoD' monitoring under their normal home routine.
after the night, a pattern which was particularly clear in objective vigilance [a 30-minute version of the Mackworth (1948) Clock Vigilance Task] where there was a more marked recovery, that is, circadian rhythm, in the young than in the old (Fig. 3, middle panel). In manual dexterity (Fig. 3. bottom panel) there was a slowing due to age (hence different scales for the two groups), but an almost linear decline for the elderly contrasted with a distinct circadian rhythmicity for the young. So downstream from the endogenous circadian pacemaker there are age related changes in circadian performance rhythms. These performance effects are important for both the elderly and the space traveler and must be further explored. ENTRAINMENT MECHANISMS Another aspect of our NIA funded research (and one that uses NASA-funded research as a control condition) is concerned with entrainment mechanisms. We have developed a 15-daytime isolation protocol involving a single acute 6 h phase advance in routine. Subjects live in special apartments isolated from all time cues for 15 days at a time. No clocks, windows, or radios are allowed but a strict routine is enforced (Monk et al., 1985). For the first 5 days that routine is tailored to be equivalent to the subject's home routine as measured by a sleep diary kept for two weeks prior to the study. On the sixth night the subject retires to bed as usual, but
SLEEP AND CIRCADIAN RHYTHMS
AVERAGETEMPERATUREDURING 3 6 H
~ .... ff '..;.: '. :"--"":' ~' ' /".,
,< n¢ laJ
I.d n=) ¢y i,i n ]E
: : , .,',. ,:'. ii : ~' ' ",
1~t1, ) 0.
01 O0 . 12
24 2'o,', ' HOURS INTO STUDY
Fig. 2. Average rectal temperature plotted as a function of time into a 36 h unmasking experiment (see text) for nine old males (solid line) and four young males (broken line). Time of day is also indicated.
is woken (without comment) 6 h earlier than normal, that is, after about 1 or 2 h of sleep. For the remainder of the study all mealtimes and bedtimes are on the new routine, that is, 6 h earlier than before (Monk et al., 1988). The phase shift is unheralded, and about half the subjects do not realize what has happened. A range of measures are taken including continuous rectal temperature, sleep recording, and mood and performance assessments. In previous NASA funded research (Monk et al., 1988) we studied eight middle-aged men (x = age 47). In our NIA funded work (and, again, we are only in year 2, so this is preliminary) we have so far studied five women in their 80's (x = age 82). Looking first at the circadian temperature rhythm we can see that this single acute phase shift had a quite dramatic effect on both groups (Fig. 4). One can quantify both amplitude and phase on an individual basis (i.e., not looking at group curves). This was done by fitting a sinusoid to each subject's data on a daily basis using a least squares technique (Monk & Fort, 1983). Averaging across the subjects then yielded an estimate of amplitude (mid level to peak of sinusoid) and acrophase (timing of sinusoid peak) for each day of the study. The analysis revealed initial age related differences, with old females showing different effects than the controls on the first two or three post shift days (Fig. 5). Phase adjustment, in particular was initially much faster for the older group. After a few days, however, the adjustment patterns were remarkably similar.
~,UBJEt]TIV[ vIGILANCE DURING
,1 ,i ee
HOURS NTO STUDY
OIECTIVE VIGILANCEDURING.36H UNMASKING
PI~[~iS PER MINUTE DtJ~ING 3~I UNMASKINO
OUD~ .P T9 28
18 27 1726
i t6 25
1S 24 MJ~S
o;,', 13' 15' , ' , ' '
21 2 3 Ot TtME OF 0AY
' , ' , , ' , ' ' 1 7 1i9
Fig. 3. Average subjective vigilance (top panel), objective vigilance (middle panel), and manual dexterity (bottom panel) from nine old males (solid line) and four young males (broken line) plotted as a function of time into a 36 h unmasking experiment (see text). Note separate scales for young and old in the bottom panel.
SLEEP A N D C I R C A D I A N R H Y T H M S
AVERAGE TEIWPERATURE -
t DAYS INTO STUDY
AVERAGE IIDvlPERATURE -
361 • 56,7 •
t DAYS INTO STUDY
Fig. 4. Averagerectal temperature plotted as a functionof days into study for five old females (upper panel) and eight middle-aged male controls (lower panel). The phase shift occurred on Day 5. The major age differences were in sleep and performance (again emphasizing the importance of downstream effects). This report will concentrate on measures of sleep efficiency (percent of the " n i g h t " spent asleep), subjective vigilance, and manual dexterity. In sleep efficiency (Fig.
TEMPERATURE RHYTHM ACROPHASE
0 14.5 (~)
OLD !rEua.LES 95 ^ I
DAY O F S T U D y
TEMPERATURE RHYTHM AMPLITUDE 0,ll 0 .~0 0.~9. o 3B.
• • ,"
0 36 0.55 0.34 • O33 0.~2 O.$T
0 2& 0.25-
0.2~ 0 23 0.2.2 0.21. 0 20 0.19 0.18
/ OLD reid AL£S
DAY O f STUDY
Fig. 5. Average least squares estimates of acrophase (upper panel) and amplitude (lower panel) of the circadian temperature rhythm (see text) from five old females (solid lines) and eight middle-aged male controls (broken line).
6, top panel) there was a very different pattern between the two age groups, with post shift night two being the worst postshift night for the middle aged male controls, but the best postshift night for the old females (indeed one of them commented that it was the best night of sleep she had for 30 years!). Thereafter, things got better (albeit in a zig-zag fashion) for the controls, but
SLEEP AND CIRCADIAN ~
O f ~nJoY
SUBJE~I~ V~l.~0[: OLD~
~ . CONllq0t3
m ~ u
0AV 0F STU~
MANUAL DF.XllSNlFY: 0t.D FBdALE8 VS. ~
Om OONm0~ 30 24p ~x23P 3,122.
31 10I! -
4 I I
I I I O ? I
I I Wl1
I I mlIN
Fig. 6. Average sleep, efficiency (top panel), subjective vigilance (middle panel), and manual dexterity (bottom panel) from five old females (solid line) and eight middle-aged male controls (broken line). The phase shift is marked A. Note different scales for the two age groups on the bottom panel.
T H. MONK
worse for the old females. Differences also appeared in subjective vigilance, where much more of a shift-related deterioration was observed in the first three post-shift days for the controls than for the old females (Fig. 6, middle panel). In manual dexterity performance speed (Fig. 6, bottom panel), there was a pronounced practice effect over the 15 days of the study in the controls with a slight decrement for about three days following the phase shift. For the old females, performance was uniformly slower, and although performance speed showed the same three day decrement after the phase shift, it failed to improve much over baseline levels thereafter. Clearly, age was dramatically affecting the medium-term response to an acute phase shift, despite the comparatively short-lived nature of the age-related effects observed in circadian rhythm parameters. Should these phase shift results hold up with an increased sample size they have important ramifications for our understanding of older people's entrainment mechanisms. Although circadian rhythm differences might be rather minor after the first two or three post-shift days, fairly dramatic differences can appear in sleep, mood, and activation for a week or more after the phase shift. Such differences could have major functional significance.
CONCLUSIONS In both extraterrestrial environments and advanced age, there are disruptions of zeitgebers and a change in masking effects. These can lead to circadian dysfunction, resulting in sleep disruption and impairments of waking function and mood. Cross fertilization of experiments and findings between the areas of space and aging are likely to be most fruitful in sleep and circadian rhythms research. Acknowledgment~ -- Grateful thank,', are due to my professional colleagues Drs. Daniel J. Buysse and David B. Jarrett, to Mr, Jeff Dettling and Mr. Bart Billy for subject recruitment, screening, training, and stud)' management; to our technicians for actually running the studies and to Ms. Jamie Nageley for analyzing the data and preparing the graphs. This work was supported by NIA grant AG06836 and a previous NASA co-operative agreement (NCC2-253) to Dr. Monk while he was at Cornell University Medical College.
REFERENCES ASCHOFF, J. Handbook ~Behavioral Neurobiology, Volume 4, New York: Plenum. 1981. ASCHOFF, J., HOFFMAN, K., POHL, H.. and WEVER. R. Re-entrainment of circadian rhythms alter phase-shifts of the zeitgeber. Chronobiologia 2, 23-78, 1975. COLQUHOUN, W.P.. PAINE, M.W.P.H., and FORT, A. Circadian rhythm of body temperature during prolonged undersea voyages. Av., Sp. and Environ. Med. 49, 671-678. 1978. CONROY, R.T.W.L., and MILLS, J.N. Human Circadian Rhythms, Churchill London, 1970. CZEISLER, C.A.. BROWN. E.N., RONDA, J.M., KRONAUER, R.E,, RICHARDSON, G.S., and FREITAG, W.D. A clinical method to assess the endogenous circadian phase ~ECP) of the deep circadian oscillator in man. Sleep Res. 14, 295, 1985. FROBERG, J.E. Twenty-tour-hour patterns in human performance, subjective and physiological variables and differences between morning and evening active subjects. Biol. Psychol. 5~ 119-134, 1977. KNAUTH. P.. RUTENFRANZ, J., HERMANN, G., and POPPL, S.J. Re-entrainment of physiological functions in experimental shiftwork studies. Ergonomics 21, 775-783, 1978. MACKWORTH, N.H. The breakdown of vigilance during prolonged visual search. O.J. L~-p. Psychol. 1, 6-21, 1948. MILLS, J.N., MINORS, D.S., and WATERHOUSE, J.M. The effects of sleep upon human circadian rhythms. Chronobiologia 5, 14-27, 1978. MINORS, D.S. and WATERHOUSE, J.M. (Eds.). Masking and biological rhythms. Chronobiol. Int. 6, I-110, 1989. MONK. T.H. Circadian rhythm. Clin. Geriat. Med. 5, 331-346, 1989a.
MONK, T.H. A visual analogue scale technique to measure Global Vigor and Affect (GVA). Psychiatry Res. 27, 89-99, 1989b. MONK, T.H., FOOKSON, J.E., KREAM, J., MOLINE. M.L., POLLAK, C.P., and WEITZMAN, M.B. Circadian factors during sustained performance: Background and methodology. Behav. Res. Methods Instrum. Comput. 17, 19-26, 1985. MONK, T.H. and FORT, A COSINA--A cosine curve fitting program suitable for small computers. Int. J. Chronobiol. 8, 193-222, 1983. MONK, T.H., MOLINE, M.L., and GRAEBER, R.C. Inducing jet lag in the laboratory: Patterns of adjustment to an acute shift in routine. Aviat. Space Environ. &led. 59, 703-710. 1988. WEITZMAN, E.D., MOLINE, M.L., CZEISLER, C.A., and ZIMMERMAN, J.C. Chronobiology of aging: Temperature, sleep-wake rhythms and entrainment. Neurobiol. Aging 3, 299-309, 1982. WEVER, R. In: The Circadian System of Man: Results of E.tperiments Under Temporal Isolation, Springer-Verlag. New York, 1979. WEVER, R.A. Order and disorder in human circadian rhythmicity: Possible relations to mental disorders. In: Biological Rhythms and Mental Disorders, Kupfer, D.J., Monk, T.H. and Barchas, J.D. (Editors), pp. 238-324, Guilford Press, New York, 1988.