Abstract
Aging is associated with changes in sleep, and improving sleep may have important consequences for the health, cognition, and quality of life of older adults. Many prescription sleep aids increase the risk of nighttime falls, have adverse effects on next-day cognition, and are associated with increased mortality. Melatonin, a hormone secreted at night, increases sleep duration in young adults but only when administered during the day when endogenous levels are low. In a month-long cross-over study, we randomized 24 healthy older (age >55, mean 64.2 ± 6.3 years) participants to receive two weeks of placebo and two weeks of either a low (0.3 mg) or high (5.o mg) dose of melatonin 30 minutes before sleep. Sleep was polysomnographically recorded and was scheduled during both the biological day and night using a forced desynchrony design. While 0.3 mg melatonin had a trend towards increasing sleep efficiency overall, this was due to it’s effects on sleep during the biological day. In contrast, 5 mg melatonin significantly increased sleep efficiency during both biological day and night, mainly by increasing the duration of stage 2 nonREM sleep and slightly shortening awakenings. Melatonin should be further explored as a sleep aid for older adults.
Keywords: biological clock, sleep, circadian rhythm, aging, melatonin, hypnotic effect
INTRODUCTION
Older people have a high prevalence of sleep deficiency, including fragmented sleep, insomnia, and early morning awakenings 1–3. Even healthy older adults without chronic medical conditions or sleep disorders have shorter sleep with more awakenings than young adults 4,5. Recently, both cross sectional and prospective epidemiological studies 6,7 have revealed that older adults who sleep less than 5 hours per night have more than double the risk of Alzheimer’s disease and mortality over a 4-5 year follow-up interval, possibly due to the accumulation of amyloid beta when sleep is disrupted and/or shortened 8,9. Unfortunately, epidemiologic studies have also revealed that older adults who chronically take prescription hypnotics to help their sleep also have higher levels of incident dementia 7,10,11 as well as higher mortality 7,12, although the direction of causality is not known. Prescription hypnotics have also been associated with greater nighttime fall risk 13 and greater risk of hip fracture 14 in older adults.
Exogenous supplementation of the pineal hormone melatonin, levels of which are often lower in older people 15–17, has been hypothesized to improve sleep duration in older people. Melatonin, via MT1 receptors, acts to suppress the firing rate of neurons in the suprachiasmatic nucleus (SCN) 18–20, the central circadian pacemaker, thereby signaling the start of the biological night 21,22. The human SCN, unlike most of the brain, has a high concentration of melatonin receptors 20. Studies in young adults have indicated that melatonin acts as a chronohypnotic, improving sleep only when endogenous melatonin is not present and having little effect during the biological night when endogenous melatonin levels are high 23.
There is evidence that the aged SCN is less sensitive to melatonin 24. The number and density of MT1-expressing neurons in the SCN are decreased with normal aging, and decreased to an even greater extent in individuals the later stages of Alzheimer’s disease than in age-matched controls 25. That latter finding may explain the sleep and circadian disruptions associated with Alzheimer’s disease 26–30. There is also evidence of reduced day-night amplitude in melatonin receptor mRNA expression in the SCN and decreased receptor binding in the SCN with age 31. Consolidation of the nocturnal sleep episode may require that the brain, and particularly the SCN, is exposed to sufficiently high levels of melatonin to activate a critical number of MT1 receptors necessary to quiet neuronal firing in the SCN 32,33.
Given the evidence that both the number and the affinity of melatonin receptors may decline with age, and that the amount of nocturnal melatonin secretion may decline in some older people, particularly those individuals whose pineal glands become calcified with age 34, we evaluated the sleep-promoting effects of a high dose as well as a low dose of exogenous melatonin in older adults. In addition, the findings from young adults that exogenous melatonin only improves sleep when endogenous levels are low 23 led us to test the sleep-promoting effects of melatonin before sleep scheduled across all circadian phases. We therefore conducted a study in healthy older adults without sleep complaints to test whether melatonin can improve sleep when it is scheduled during the biological nighttime or during the biological daytime, and if so, whether the sleep-improving effects are dose-dependent.
MATERIALS AND METHODS
Participants.
Participants were recruited for the study via newspaper advertisements and notices to community organizations. All were medically and psychologically healthy, as assessed during a screening process prior to study. Medical screening included routine clinical tests on blood and urine, an electrocardiogram, a chest radiograph, a physical examination, and a medical history to rule out acute or chronic illnesses and medication use. Psychological screening included the MMPI-2 35 and the Geriatric Depression Scale 36, the Folstein Mini-Mental State Exam 37, and a structured interview with a clinical psychologist to rule out current or past psychopathology. None were under the care of a physician for any chronic medical condition and none were regularly taking medications. They were instructed to abstain from caffeine, nicotine, alcohol, and all prescription and over-the-counter medications during the three weeks prior to study; compliance with this aspect of the study was verified upon admission to the laboratory by comprehensive toxicological analysis of their urine.
All participants were without major sleep complaint by history and were evaluated for sleep disorders by all-night polysomnography prior to admission. Average AHI of the participants empaneled into the studies was 11.99 ± 7.3 (range: 0 - 29) and average PLMI was 4.52 ± 9.1 (range: 0 - 35). All participants were instructed to maintain a regular (± 30 minutes) sleep-wake schedule with 8 hours in bed at their habitual times for the 3 weeks prior to study. During the week immediately prior to the study, compliance with this regular schedule was verified with a wrist activity monitor (Actiwatch-L, Philips Respironics, Murrysville, PA).
Each gave written informed consent prior to study; the study was reviewed and approved by the Partners HealthCare Human Subjects Committee and was conducted in accordance with the principles outlined in the Declaration of Helsinki.
A total of 25 participants began the inpatient portion of the study. One withdrew consent on the 5th day of the study and was not included in any of the analyses reported here. One participant was withdrawn from the study by the investigators shortly before the end of condition 2 (see below), but her data were included in the analyses. The 24 participants (13 women, 11 men) ranged in age from 55 to 78 years (mean ± SD: 64.2 ± 6.3 years), and their habitual bedtimes were, on average, 22:54 ± 0:44.
Experimental Protocol.
Each study began with three baseline (BL) adaptation days, consisting of 16 hours of wakefulness and an 8-hour sleep episode scheduled according to each participant’s habitual bedtime and wake time. This was followed by a forced desynchrony protocol 38–40, during which the participants were scheduled to a rest-activity cycle length of 20 hours, with sleep episodes scheduled to begin 4 hours earlier each day and continue for 1/3 of each cycle (6 hours 40 minutes; see Figure 1). During the remainder of each 20-hour “day” (13 hours 20 minutes), participants were awake and ambulatory within their study room. This was continued for 30 cycles (equivalent to 24 calendar days), followed by three 24-hour re-adaptation days which were scheduled at the same times as the baseline days.
Figure 1.
Double raster plot of study protocol. Relative time of day is across the 48-hour x-axis; days of the experiment are plotted both to the right of and beneath the previous day. Boxes represent scheduled sleep episodes. Open boxes on nights 1-3 represent 8-hour scheduled baseline sleep episodes during which all participants received placebo on all three baseline nights. From night 4 through night 33, participants were scheduled to live on a 20-hour “day”, with bedtimes four hours earlier each night than the previous night. Scheduled sleep episodes were 6:40 and wake episodes were 13:20. Hashed boxes represent melatonin/placebo nights, open boxes represent single-blind placebo nights. Participants were randomized to receive placebo on each of the 12 nights during one of the two randomization blocks; following 6 nights of washout in which all participants received placebo, they were then randomized to the opposite condition (melatonin/placebo) on each of the 12 nights in the second randomization block.
Experimental Conditions.
During the entire study, ambient light intensity during the scheduled waking episodes was dim [approximately 0.0087 W/m2 (~3.3 lux) at 137 cm from the floor facing towards the walls with a maximum of 0.048 W/m2 (15 lux) at 187 cm from the floor facing towards the ceiling], and during scheduled sleep episodes all lights were turned off (complete darkness).
Throughout their study, each participant lived in a private study room that was free of external time cues. The study room had no windows, clocks, or other indication of time of day, and the participants were not permitted to watch television or listen to the radio. Staff members were trained to avoid discussion of time-of-day information and did not wear watches. Participants were administered short tests of subjective mood and alertness approximately every 30 minutes, and approximately every two hours were administered a performance test battery lasting approximately 20 minutes 41. During their free time between tests, participants were allowed to pursue sedentary activities in their study room, which typically included reading, listening to music, watching videos, or pursuing hobbies.
Drug Condition and Randomization.
Identically-appearing melatonin and placebo capsules were manufactured by the Investigational Drug Service (IDS) of the Brigham and Women’s Hospital Pharmacy. Capsules were tested at manufacture, and then again annually to ensure that they contained the specified dosage (Chemir Polytech). Upon each participant’s enrollment in the inpatient portion of the study, the IDS randomized the participant (see below) and packaged their 36 melatonin capsules individually, labeling them with participant number and order of administration.
During the baseline nights (sleep episodes 1-3), the re-adaptation nights (sleep episodes 34-36) and sleep episodes 16-21, all participants were scheduled to receive a placebo capsule. The remaining nights of each study were divided into two conditions, for the first 12 (condition 1, sleep episodes 4-15) and final 12 (condition 2, sleep episodes 22-33) nights of the forced desynchrony portion of the study. Each participant was randomly assigned to receive placebo in one of the conditions and melatonin in the other condition. Two doses of melatonin were tested, 0.3 mg (low dose) and 5 mg (high dose), with each participant receiving the same dose throughout the 12 nights of their melatonin condition. A total of four randomization groups thus resulted [placebo followed by 0.3 mg; placebo followed by 5.0 mg; 0.3 mg followed by placebo; 5.0 mg followed by placebo]. Pills were administered thirty minutes prior to each bedtime by a CCI nurse. Study staff involved in participant interactions, data collection, data processing, and preliminary data analysis remained blind to the condition of each participant until all had completed the study and all sleep data had been scored.
Data Collection.
Core body temperature was collected continuously throughout each study using a rectal thermistor. Throughout most of the study, blood samples were collected hourly via an indwelling intravenous catheter connected to a 12-foot tubing so that samples could be collected from outside the room during scheduled sleep episodes. Blood samples were kept on ice for up to 1 hour prior to being centrifuged. The resulting plasma was frozen until the samples were assayed for melatonin. Plasma samples were assayed at one of two assay laboratories. The initial 10 participants’ samples were assayed by the Core Laboratory at the BWH CCI; because this facility was unable to accommodate the large volume of samples from this study, the remaining 14 participants’ samples were assayed at Pharmasan Labs (Osceola, WI).
The EEG was recorded during all sleep episodes using a standard montage (C3, C4, O1, O2), referenced to contralateral mastoids (A1, A2). In addition to the EEG, two electrooculograms (left outer canthus, right outer canthus), one submental electromyogram, and a 2-lead electrocardiogram were recorded. All signals were acquired using a digital ambulatory sleep recording system (Vitaport-2 or 3, Temec Instruments, Kerkrade, B.V., The Netherlands). The EEG signals were high-pass filtered at a time constant of one second and low-pass filtered at 70 Hz (Bessel fourth-order antialiasing; > 80 dB). Finally, the signals were digitized with a resolution of 12 bit (range 500 μV; sampling rate 256 Hz, storage rate 128 Hz), stored on a Flash RAM card, and downloaded offline after wake time. All sleep episodes were scored visually according to standard criteria 42 by trained scorers who were blind to the study conditions.
We excluded sleep episode 1 from analysis for all participants due to the “first night” effect. We also excluded from our analysis any sleep episode that was missing more than 5% of the scheduled epochs due to equipment or sensor malfunction 23. Sleep episodes 2 and 3 were averaged for each participant to determine baseline night information.
We defined sleep latency (SL) as the time from lights off until the occurrence of any stage of sleep. Awakenings were defined as the number of epochs or series of epochs within a sleep episode scored as wake; the average duration of all awakenings for each sleep episode was also calculated. For sleep-stage data, NREM sleep (stages 2-4), rapid-eye movement (REM) sleep, slow wave sleep (SWS; stage 3 and 4), and stages 1-4 were calculated in minutes. Wakefulness during scheduled sleep and sleep efficiency (SE) were calculated as the percentage of total time in bed between lights off and lights on.
Each sleep episode was assigned a circadian phase of bedtime (see below). We then assigned each sleep episode during the FD segment (sleep episodes 4-33) to biological day (90-240°) or biological night (240-90°).
Data Analysis.
The core body temperature (CBT) data from each condition of the study was assessed for intrinsic circadian period using non-orthogonal spectral analysis (NOSA) 39,40. This analysis takes into account the 20-hour periodicity in the data resulting from the imposed rest-activity schedule, and then simultaneously searches for a periodicity in the circadian range (search range = 15 to 30 hours). Using the period and the projection of the CBT minimum on the first day of each condition (assigned circadian phase 0°), we then assigned a circadian phase from 0° to 359° to each minute of the FD segment of the study and used this to assign each sleep episode a circadian phase at bedtime.
Descriptive statistics are presented as percentage or mean (± standard deviation) unless otherwise noted. Linear mixed-effects models for repeated measures were used to study the effect of DRUG (melatonin, placebo) on sleep efficiency, sleep latency, and sleep time in various sleep stages. Generalized linear mixed-effects models with Poisson distribution were used to study the effect of DRUG (melatonin, placebo) on number of awakenings. While DRUG was the main effect, SEX (female, male), ORDER (placebo-melatonin, melatonin-placebo), and TIME OF DRUG (biological day, biological night) were also tested in the models. SUBJECT was treated as a random effect in the models to account for individual variability. Proportional hazards regression models were used to study the effect of DRUG (melatonin or placebo) on duration of awakenings. Robust sandwich covariance estimates were used to account for intra-individual correlations. Residual plots and model assumptions were examined to assess model fit. For all statistical tests, the critical significance level is defined as α = 0.05. All reported degrees of freedom and two-sided p-values are from the final statistical model for each measure. Final statistical models include DRUG as the primary variable of interest, all significant main effects, and all significant interactions, but exclude non-significant interactions. Note that because significant interactions vary between measures, degrees of freedom are not uniform. The analyses were performed with SAS 9.4.
Results in figures are presented as mean ± standard deviation, with all observations first averaged within, and then across participants.
RESULTS
A total of 564 polysomnographically-recorded sleep episodes from the 24 participants were included in the analysis. Due to missing data from equipment or sensor malfunctions 23, one baseline sleep episode from each of two participants was excluded, and nine nights from the forced desynchrony segment of the study (one night each from five participants, and two nights from two) were excluded. For one participant, post hoc analysis of the plasma melatonin data suggested that he inadvertently received melatonin on two of the nights he was supposed to have received placebo; these two sleep episodes were excluded from analysis.
There was no difference in age, sex distribution, habitual sleep-wake timing, or morningness-eveningness score between the participants in the low melatonin dose groups vs. those in the high melatonin dose groups (see Table 1). On the baseline nights, we did not find any difference in sleep onset latency between the low and high dose groups, there were no significant differences in the duration of any sleep stage, and there were no differences in the number or average duration of awakenings (see Table 1).
Table 1.
Baseline characteristics of the Low Dose (0.3 mg) Melatonin and High Dose (5.0 mg) Melatonin groups.
| Low Dose n=12 mean (std dev) | High Dose n=12 mean (std dev) | T value or Z valuea | P-value | |
|---|---|---|---|---|
| Age (years) | 64.67 (6.15) | 63.75 (6.68) | −0.35 | 0.73 |
| Sex | 6M, 6F | 5M, 7F | 0.168 | 0.682 |
| Habitual Bedtime | 23.12 (0.54) | 22.69 (0.86) | −0.146 | 0.158 |
| Habitual Waketime | 7.13 (0.52) | 6.83 (0.78) | −0.347 | 0.729 |
| Morningness Eveningness Score | 59.86 (6.15) | 63.11 (9.83) | 0.678 | 0.498 |
| Baseline Sleep Efficiency (%) | 80.83 (5.38) | 78.59 (8.24) | −0.791 | 0.437 |
| Stage 1 (min) | 53.46 (31.64) | 51.81 (23.77) | −0.144 | 0.887 |
| Stage 2 (min) | 202.94 (27.59) | 176.23 (41.70) | −1.850 | 0.078 |
| SWS (min) | 42.83 (36.39) | 59.79 (33.88) | 1.181 | 0.250 |
| REM (min) | 86.81 (26.02) | 88.10 (19.33) | 0.138 | 0.892 |
| TST (min) | 386.04 (24.98) | 375.94 (39.66) | −0.747 | 0.463 |
| Wake (min) | 91.94 (25.59) | 102.54 (39.68) | 0.778 | 0.445 |
| SoL (min) | 18.39 (17.37) | 19.58 (19.54) | −0.123 | 0.902 |
| Awakenings (n) | 28.42 (9.86) | 28.79 (17.43) | −0.549 | 0.583 |
| Awakenings (average duration) | 3.78 (2.31) | 4.09 (1.91) | 0.779 | 0.436 |
std dev=standard deviation; min=minutes, SWS=slow wave sleep, REM=rapid eye movement sleep; TST= total sleep time, SoL=sleep onset latency.
Wilcoxon Rank Sum test
We did not find a significant sex difference in any outcome, and there was no significant order effect (placebo first, melatonin first) for either the low dose or the high dose for any sleep outcome examined. We therefore combined the placebo-first/melatonin-second and melatonin-first/placebo-second groups for the low dose and for the high dose for all subsequent analyses.
In the low-dose group, while there was no significant difference in sleep efficiency in the melatonin condition compared with the placebo condition, there was a trend for greater sleep efficiency and TST overall (see Table 2 and Figure 2, lower panel) as well as during the biological daytime (see Table 3). Looking at all the sleep episodes, when the participants were in the low-dose melatonin condition they had significantly more stage 1 and stage 2 sleep, significantly less slow-wave sleep, and no difference in REM sleep than in the placebo condition. Sleep latency was not significantly different between the low-dose melatonin and placebo conditions, and while the number and average duration of awakenings were not different between the two conditions, there was a trend for less wakefulness in the low-dose melatonin condition (see Table 2).
Table 2.
Overall sleep stage durations by dosage group. Mean (and standard deviation) for each sleep stage are presented in minutes, along with sleep efficiency and number of awakenings; model-fit group difference estimates (and standard error) are presented, as are T values and p values for the dosage group comparisons. Left: Placebo vs. Low Dose (0.3 mg) Melatonin; Right: Placebo vs. High Dose (5.0 mg) Melatonin.
| Overall | Overall | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Placebo mean (std dev) | Low Dose mean (std dev) | Estimate (std err) | T value (df=266) | P-value | Placebo | High Dose | Estimate (std err) | T value (df=272) | P-value | |
| Sleep Efficiency (%) | 71.57 (7.15) | 74.65 (6.31) | 3.05 (1.81) | 1.69 | 0.0927 | 69.42 (9.04) | 75.68 (5.68) | 6.24 (1.81) | 3.44 | 0.0007 |
| Stage 1 (min) | 33.94 (16.52) | 38.59 (23.76) | 4.61 (1.59) | 2.9 | 0.0041 | 36.75 (21.31) | 40.22 (19.41) | 3.37 (1.49) | 2.27 | 0.0242 |
| Stage 2 (min) | 138.23 (18.22) | 152.20 (21.17) | 14.01 (4.68) | 3 | 0.003 | 119.81 (24.96) | 146.25 (21.05) | 26.50 (4.21) | 6.29 | <0.0001 |
| Stage SWS (min) | 47.04 (23.75) | 41.50 (26.85) | −5.56 (1.78) | −3.13 | 0.002 | 54.28 (22.60) | 44.59 (21.99) | −9.69 (1.98) | −4.9 | <0.0001 |
| Stage NREM (min) | 219.21 (23.48) | 232.29 (27.39) | 13.02 (5.36) | 2.43 | 0.0159 | 210.84 (25.56) | 231.05 (16.82) | 20.17 (5.14) | 3.92 | 0.0001 |
| Stage REM (min) | 66.37 (14.03) | 65.76 (14.29) | −0.61 (2.83) | −0.22 | 0.829 | 66.47 (14.93) | 71.22 (11.59) | 4.73 (2.93) | 1.61 | 0.1078 |
| TST (min) | 285.58 (28.41) | 298.06 (25.10) | 12.37 (7.22) | 1.71 | 0.0878 | 277.30 (36.10) | 302.27 (22.77) | 24.91 (7.25) | 3.44 | 0.0007 |
| Stage Wake (min) | 113.50 (28.75) | 100.58 (25.26) | −12.84 (7.16) | −1.79 | 0.074 | 122.12 (36.10) | 97.15 (22.76) | −24.91 (7.26) | −3.43 | 0.0007 |
| SoL (min) | 20.53 (12.15) | 17.37 (9.37) | −3.20 (3.93) | −0.81 | 0.4172 | 17.84 (9.51) | 16.03 (9.58) | −1.78 (2.31) | −0.77 | 0.4419 |
| Number of awakenings | 23.81 (9.04) | 24.53 (10.49) | 0.03 (0.02) | 1.18 | 0.2385 | 22.43 (9.67) | 24.45 (13.30) | 0.09 (0.02) | 3.49 | 0.0006 |
| Average duration of awakenings | 6.03 (2.68) | 5.19 (2.09) | 0.17 (0.10) | 2.76a | 0.0968 | 7.00 (3.38) | 4.93 (2.97) | 0.33 (0.10) | 10.18a | 0.0014 |
std dev=standard deviation; std err=standard error; df=degrees of freedom; min=minutes, SWS=slow wave sleep, NREM=non REM sleep, REM=rapid eye movement sleep, TST=total sleep time, SoL=sleep onset latency.
=chi-square statistic, df=1
Figure 2.
Overall sleep efficiency by condition and dosage group. Mean (± standard deviation) sleep efficiency averaged across all 12 nights in the placebo condition (white bar, left) and melatonin condition (grey bar, right). Upper Panel: Placebo vs. High Dose (5.0 mg) Melatonin; Lower Panel: Placebo vs. Low Dose (0.3 mg) Melatonin.
Table 3.
Sleep stage durations by dosage group and biological time of scheduled sleep opportunity. Mean (and standard deviation) for each sleep stage are presented in minutes, along with sleep efficiency and number of awakenings; model-fit group difference estimates (and standard error) are presented, as are T values and p values for the dosage group comparisons. Biological Night: sleep opportunities beginning between 240 and 90 degrees; Biological Day: sleep opportunities beginning between 90 and 240 degrees. Top Panel: Placebo vs. Low Dose (0.3 mg) Melatonin; Bottom Panel: Placebo vs. High Dose (5.0 mg) Melatonin.
| Biological Night | Biological Day | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Placebo mean (std dev) | Low Dose mean (std dev) | Estimate (std err) | T value (df=144) | P-value | Placebo mean (std dev) | Low Dose mean (std dev) | Estimate (std err) | T value (df=109) | P-value | |
| Sleep Efficiency(%) | 78.54 (6) | 80.37 (5.54) | 1.44 (1.88) | 0.77 | 0.4449 | 62.3 (10.03) | 66.69 (11.12) | 4.55 (2.58) | 1.76 | 0.0804 |
| Stage 1 (min) | 35.4 (17.62) | 39.48 (23.0) | 4.36 (2.06) | 2.12 | 0.036 | 32.85 (16.24) | 37.54 (25.37) | 4.60 (2.51) | 1.83 | 0.0703 |
| Stage 2 (min) | 149.29 (21.03) | 160.2 (21.03) | 10.89 (5.67) | 1.92 | 0.0569 | 122.84 (26.57) | 140.04 (29.81) | 16.62 (6.88) | 2.42 | 0.0174 |
| Stage SWS (min) | 47.72 (26.64) | 43.82 (27.8) | −3.89 (2.40) | −1.62 | 0.1075 | 45.58 (21.65) | 38.91 (25.94) | −6.68 (2.43) | −2.75 | 0.007 |
| Stage NREM (min) | 232.41 (18.82) | 243.5 (19.85) | 10.44 (6.04) | 1.73 | 0.086 | 201.28 (34.49) | 216.48 (43.63) | 14.95 (8.44) | 1.77 | 0.0794 |
| Stage REM (min) | 81.07 (15.32) | 77.47 (16.52) | −3.83 (2.98) | −1.29 | 0.1997 | 47.16 (11.72) | 49.69 (14.18) | 3.30 (3.04) | 1.09 | 0.2803 |
| TST (min) | 313.48 (23.89) | 320.97 (22.16) | 5.89 (7.48) | 0.79 | 0.4323 | 248.44 (39.78) | 266.17 (44.17) | 18.37 (10.27) | 1.79 | 0.0765 |
| Stage Wake (min) | 85.68 (24.01) | 78.46 (21.9) | −5.65 (7.49) | −0.76 | 0.4515 | 150.51 (40.25) | 131.62 (44.70) | −19.64 (10.16) | −1.93 | 0.0558 |
| Latency (min) | 19.32 (19.02) | 12.69 (5.4) | −7.52 (6.37) | −1.18 | 0.2397 | 20.94 (12.17) | 23.33 (15.57) | 2.29 (3.62) | 0.63 | 0.5283 |
| Number of awakenings | 23.59 (8.27) | 25.38 (10.69) | 0.07 (0.03) | 2.13 | 0.0349 | 23.93 (10.53) | 23.28 (10.48) | −0.03 (0.04) | −0.68 | 0.497 |
| Duration of awakenings (min) | 4.46 (2.15) | 3.65 (168) | 0.23 (0.14) | 2.66a | 0.1026 | 8.34 (4.75) | 7.38 (3.99) | 0.13 (0.16) | 0.65a | 0.4185 |
| Placebo mean (std dev) | High Dose mean (std dev) | Estimate (std err) | T value (df=148) | P-value | Placebo mean (std dev) | High Dose mean (std dev) | Estimate (std err) | T value (df=111) | P-value | |
| Sleep Efficiency (%) | 76.68 (7.62) | 80.67 (4.52) | 3.89 (185) | 2.1 | 0.0374 | 60.47 (12.06) | 69.7 (9.02) | 8.87 (2.84) | 3.12 | 0.0023 |
| Stage 1 (min) | 38.68 (23.50) | 42.22 (20.07) | 2.89 (2.05) | 1.41 | 0.1604 | 34.33 (20.21) | 37.54 (19.52) | 3.85 (2.02) | 1.91 | 0.0591 |
| Stage 2 (min) | 130.68 (27.30) | 151.51 (21.39) | 20.74 (4.84) | 4.28 | <0.0001 | 105.93 (28.53) | 141.69 (27.28) | 34.01 (6.82) | 4.98 | <0.0001 |
| Stage SWS (min) | 58.18 (26.46) | 46.72 (22.77) | −11.14 (2.57) | −4.33 | <0.0001 | 50.25 (19.94) | 41.64 (22.18) | −8.03 (2.93) | −2.74 | 0.0071 |
| Stage NREM (min) | 227.54 (19.61) | 240.45 (15.95) | 12.53 (5.42) | 2.31 | 0.0224 | 190.51 (36.57) | 220.86 (26.80) | 29.78 (8.46) | 3.52 | 0.0006 |
| Stage REM (min) | 78.78 (17.27) | 81.78 (13.30) | 2.95 (3.32) | 0.89 | 0.3754 | 51.03 (15.12) | 57.5 (11.12) | 5.57 (3.84) | 1.45 | 0.1502 |
| TST (min) | 306.32 (30.37) | 322.24 (18.09) | 15.56 (7.40) | 2.1 | 0.0372 | 241.54 (48.24) | 278.37 (36.09) | 35.38 (11.35) | 3.12 | 0.0023 |
| Stage Wake (min) | 93.09 (30.22) | 77.11 (17.72) | −15.64 (7.43) | −2.1 | 0.037 | 157.89 (48.36) | 121.13 (36.17) | −35,28 (11.36) | −3.11 | 0.0024 |
| Latency (min) | 15.78 (10.53) | 11.52 (8.70) | −4.11 (2.16) | −1.9 | 0.0598 | 19.95 (11.11) | 23.25 (19.84) | 2.41 (4.25) | 0.57 | 0.5714 |
| Number of awakenings | 24.24 (10.78) | 24.02 (11.33) | −0.01 (0.03) | −0.17 | 0.8633 | 19.9 (9.27) | 25.22 (16.28) | 0.22 (0.04) | 5.72 | <0.0001 |
| Duration of awakenings (min) | 4.52 (3.01) | 3.37 (133) | 0.33 (0.19) | 3.09a | 0.0789 | 10.26 (5.08) | 6.63 (5.04) | 0.37 (0.20) | 3.40a | 0.0653 |
std dev=standard deviation; std err=standard error; df=degrees of freedom; min=minutes, SWS=slow wave sleep, NREM=non REM sleep, REM=rapid eye movement sleep;TST=total sleep time, SoL=sleep onset latency.
=chi-square statistic, df=1
In the high-dose group, overall there was significantly greater sleep efficiency and TST in the melatonin condition compared with the placebo condition (see Figure 2, upper panel). When the participants were in the high-dose melatonin condition they had significantly more stage 1 and stage 2 sleep, significantly less slow-wave sleep, and no difference in REM sleep than in the placebo condition. Sleep latency was not significantly different between the high-dose melatonin and placebo conditions. There were significantly more, but shorter awakenings in the high-dose melatonin condition than in the placebo condition, and overall the participants had significantly less wakefulness in the high-dose melatonin condition compared with the placebo condition (see Table 2).
In the low-dose group there was no significant difference in sleep efficiency or TST in either the biological day or the biological night between the low-dose melatonin or placebo conditions, although there was a trend towards greater sleep efficiency and TST in the low-dose melatonin condition during the biological day (see Figure 3, lower panel). Although there were trends for differences in sleep stages, the only significant differences were a significant increase in Stage 1 sleep and an increase in the number of awakenings during the biological night in the low-dose melatonin condition compared with the placebo condition (see Table 3). During the biological day, there was significantly more stage 2 sleep and significantly less slow wave sleep in the low-dose melatonin condition than in the placebo condition (see Table 3).
Figure 3.
Sleep efficiency during the biological day and biological night by condition and dosage group. Biological Day: sleep opportunities beginning between 90 and 240 degrees; Biological Night: sleep opportunities beginning between 240 and 90 degrees. Upper Panel: Mean (± standard deviation) sleep efficiency averaged across all Biological Day (left) or Biological Night (right) sleep episodes in the Placebo condition (white bars) and High Dose (5.0 mg) Melatonin condition (grey bars). Lower Panel: Mean (± standard deviation) sleep efficiency averaged across all Biological Day (left) or Biological Night (right) sleep episodes in the Placebo condition (white bars) and Low Dose (0.3 mg) Melatonin condition.
The high-dose melatonin group showed significantly greater sleep efficiency and TST when sleep was scheduled during either the biological day or the biological night compared with the placebo condition (see Figure 3, upper panel). The high-dose melatonin group showed significantly more stage 2 sleep and nonREM sleep, and significantly less slow wave sleep and wakefulness when sleep was scheduled during either the biological day or the biological night compared with the placebo condition (see Table 3). During both the biological day and biological night there was a trend for the duration of awakenings to be shorter in the high-dose melatonin condition compared with the placebo condition, and while the number of awakenings was not different between the high-dose melatonin condition and placebo condition during the biological night, there were significantly more awakenings in the high-dose melatonin condition compared with placebo when sleep occurred during the biological day (see Table 3).
DISCUSSION
In a randomized, placebo-controlled trial of 24 healthy older adults scheduled to sleep at all different circadian phases, we found that pre-sleep administration of a 5 mg dose of melatonin increased total sleep time by 25 minutes compared with placebo. When only considering sleep scheduled during the biological daytime when endogenous melatonin was absent, pre-sleep administration of 5 mg melatonin as compared to placebo increased sleep duration by more than half an hour. When 5 mg melatonin was administered before sleep scheduled to occur during the biological nighttime when endogenous melatonin was present, it was still able to increase sleep duration significantly, by more than 15 minutes. In contrast, while pre-sleep administration of low-dose (0.3 mg) melatonin increased stage 1, stage 2, and nonREM sleep duration, neither the 12-minute change in total sleep time overall or the 18-minute increase in TST during biological day sleep was statistically significant. These data reveal that a higher dose of melatonin may be required to reliably increase total sleep time in healthy older adults, consistent with evidence from animal studies that MT1 receptor affinity decreases with aging.
These findings in healthy older adults contrast with findings from a similar study we carried out in young adults 23. In that study, we used a between-participant design to compare the impact of a low dose (0.3mg), high dose (5.0mg), or placebo on sleep scheduled on a 3-week, 20-hour forced-desynchrony protocol. We found no significant change in sleep efficiency during nighttime sleep for either the low dose or high dose melatonin groups compared with the placebo group, likely because average nocturnal sleep efficiency was already quite high in those young participants, at 88%. When we examined sleep scheduled to occur during the biological daytime when endogenous melatonin was not present, both low dose and high dose melatonin significantly improved sleep efficiency, with no difference between the doses in those young adults.
Our current results also contrast with those from several prior studies in which melatonin was tested in older adults. In general, those studies did not find any effect of melatonin on the duration of sleep in older adults, whether they had sleep complaints or not [see 43 for a summary]. However, those prior studies used lower doses of melatonin (typically 2 mg) than the 5 mg dose we found to be effective. Additional studies are needed to determine whether doses between 2 and 5 mg are effective for sleep promotion in older adults.
In summary, we demonstrated that, in healthy older adults without major sleep complaints, a 5 mg dose of melatonin can improve sleep in both the biological daytime and the biological nighttime. While these findings remain to be replicated in larger trials, they suggest that melatonin holds potential for improving sleep in older adults. A recent report found that over the past two decades there has been a three-fold increase in the number of US adults age 65 and older who report using melatonin in the past month 44, despite the lack of existing evidence for its efficacy and despite recommendations from the American Academy of Sleep Medicine that physicians should not recommend its use to older patients 43. Future studies should be carried out to investigate whether higher (between 2 and 5 mg) doses of melatonin improve sleep in older adults with insomnia or other sleep complaints, and to investigate the mechanism(s) underlying the age-related difference in response to melatonin that we have observed. With that information, personalized, age-related dosing regimen for melatonin to improve sleep can be developed.
ACKNOWLEDGMENTS
We wish to thank the study participants; D. McCarthy and C. O’Brien for participant recruitment; H. Lam of the BWH IDS for melatonin capsule production and participant randomization; K.B. McKay, M.J. Duverne, R. Webb, V. Hobbes, V. Sparkes, and R.F. Dimanche of the DSCD Chronobiology Core for assisting with the data collection and processing; the BWH CCI staff; B.J. Lockyer, E. Riel, and G. Renchkovsky of the DSCD Sleep Core for scoring the sleep recordings; and J.K. Wyatt, R.J Hughes, and D.J. Dijk for thoughtful discussions regarding a similar study conducted in young adults. The studies were supported by NIH grant P01 AG09975 and were conducted in the Brigham and Women’s Hospital General Clinical Research Center supported by M01 RR02635; additional support for data analysis was provided by NIH grants AG06072 and AG044416; JFD received support from the Brigham and Women’s Hospital BRI Fund to Sustain Research Excellence.
POTENTIAL CONFLICTS OF INTEREST
Drs. Duffy and Mr. Ronda declare that they have no potential conflicts of interest. Dr. Wang reports that she is a consultant to the National Sleep Foundation. Dr. Czeisler reports grants/contracts to BWH from Dayzz Live Well, Ltd., Delta Airlines, FAA, Jazz Pharmaceuticals, NHLBI, NIA, NIOSH, NASA, Puget Sound Pilots, Regeneron Pharmaceuticals/Sanofi and DOD; grants/gifts to Monash University from CDC Foundation, with funding from BNY Mellon, and WHOOP; is/was a paid consultant or received lecture fees from Emory University, Inselspital Bern, UCLA, Institute of Digital Media and Child Development, Klarman Family Foundation, National Council for Mental Wellbeing, National Sleep Foundation, Physician’s Seal, SRS Foundation, Tencent Holdings, Teva Pharma Australia, With Deep, & Vanda Pharmaceuticals Inc, in which Dr. Czeisler holds an equity interest; received travel support from Aspen Brain Institute, Bloomage International Investment Group, UK Biotechnology and Biological Sciences Research Council, Bouley Botanicals, Dr. Stanley Ho Medical Development Foundation, EBRS, German National Academy of Sciences (Leopoldina), National Safety Council, National Sleep Foundation, Stanford Medical School, Tencent Holdings, and Vanda Pharmaceuticals; receives research/education support through BWH from Arbor Pharmaceuticals, Avadel Pharmaceuticals, Beijing Zhaode Healthcare Management Consulting Co., Bryte, Alexandra Drane, Eisai, Harmony Biosciences, Jazz Pharmaceuticals, Johnson & Johnson, Mary Ann & Stanley Snider via Combined Jewish Philanthropies, NeuroCare, Inc., Optum, Philips Respironics, Regeneron Pharmaceuticals, Regional Home Care, ResMed, San Francisco Bar Pilots, Sanofi, Schneider, Simmons, Sleep Cycle. Sleep Number, Sysco, Teva Pharmaceuticals Industries, Vanda Pharmaceuticals; is/was an expert witness in legal cases, including those involving Advanced Power Technologies, Aegis Chemical Solutions LLC, Amtrak; Casper Sleep Inc, C&J Energy Services, Catapult Energy Services Group, Covenant Testing Technologies, Dallas Police Association, Enterprise Rent-A-Car, Espinal Trucking/Eagle Transport Group/Steel Warehouse Inc, FedEx, Greyhound Lines Inc/Motor Coach Industries/FirstGroup America, PAR Electrical Contractors Inc, Product & Logistics Services LLC/Schlumberger Technology Corp/Gelco Fleet Trust, Puckett Emergency Medical Services LLC, Puget Sound Pilots, Union Pacific Railroad, United Parcel Service, and Vanda Pharmaceuticals; serves as the incumbent of an endowed professorship provided to Harvard University by Cephalon, Inc.; and receives royalties from McGraw Hill, and Philips Respironics for the Actiwatch-2 and Actiwatch Spectrum devices. Dr. Czeisler’s interests were reviewed and are managed by the BWH and MGB in accordance with their conflict of interest policies.
References
- 1.Foley DJ, Monjan AA, Brown SL, Simonsick EM, Wallace RB, Blazer DG. Sleep complaints among elderly persons: An epidemiologic study of three communities. Sleep 1995;18(6):425–432. [DOI] [PubMed] [Google Scholar]
- 2.Feinberg I, Koresko RL, Heller N. EEG sleep patterns as a function of normal and pathological aging in man. J Psychiatr Res 1967;5:107–144. [DOI] [PubMed] [Google Scholar]
- 3.Bliwise DL. Sleep in normal aging and dementia. Sleep 1993;16(1):40–81. [DOI] [PubMed] [Google Scholar]
- 4.Bliwise DL. Normal Aging. In: Kryger MH, Roth T, Dement WC, eds. Principles and Practice of Sleep Medicine. 4th Ed. Philadelphia: Elsevier Saunders; 2005:24–38. [Google Scholar]
- 5.Dijk DJ, Duffy JF, Czeisler CA. Age-related increase in awakenings: Impaired consolidation of nonREM sleep at all circadian phases. Sleep 2001;24(5):565–577. [DOI] [PubMed] [Google Scholar]
- 6.Robbins R, Quan SF, Weaver MD, Bormes G, Barger LK, Czeisler CA. Examining sleep deficiency and disturbance and their risk for incident dementia and all-cause mortality in older adults across 5 years in the United States. Aging (Albany NY) 2021;13(3):3254–3268. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Robbins R, Weaver MD, Barger LK, Wang W, Quan SF, Czeisler CA. Sleep difficulties, incident dementia and all-cause mortality among older adults across 8 years: Findings from the National Health and Aging Trends Study. J Sleep Res 2021;30(6):e13395. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ju YE, McLeland JS, Toedebusch CD, et al. Sleep quality and preclinical Alzheimer disease. JAMA Neurol 2013;70(5):587–593. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Ju Y-ES, Lucey BP, Holtzman DM. Sleep and Alzheimer disease pathology—A bidirectional relationship. Nat Rev Neurol 2014;10(2):115–119. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Robbins R, DiClemente RJ, Troxel AB, et al. Sleep medication use and incident dementia in a nationally representative sample of older adults in the US. Sleep Med 2021;79:183–189. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Billioti de Gage S, Moride Y, Ducruet T, et al. Benzodiazepine use and risk of Alzheimer’s disease: Case-control study. BMJ 2014;349:g5205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kripke DF, Langer RD, Kline LE. Hypnotics’ association with mortality or cancer: A matched cohort study. BMJ Open 2012;2(1):e000850. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Herzig SJ, Rothberg MB, Moss CR, et al. Risk of in-hospital falls among medications commonly used for insomnia in hospitalized patients. Sleep 2021;44(9):zsab064. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Cumming RG, Le Couteur DG. Benzodiazepines and risk of hip fractures in older people: a review of the evidence. CNS drugs 2003;17(11):825–837. [DOI] [PubMed] [Google Scholar]
- 15.Sharma M, Palacios-Bois J, Schwartz G, et al. Circadian rhythms of melatonin and cortisol in aging. Biol Psychiatry 1989;25:305–319. [DOI] [PubMed] [Google Scholar]
- 16.Liu R-Y, Zhou J-N, van Heerikhuize J, Hofman MA, Swaab DF. Decreased melatonin levels in postmortem cerebrospinal fluid in relation to aging, Alzheimer’s disease, and apolipoprotein E-e4/4 genotype. J Clin Endocrinol Metab 1999;84:323–327. [DOI] [PubMed] [Google Scholar]
- 17.Youngstedt SD, Kripke DF, Elliot JA, Klauber MR. Circadian abnormalities in older adults. J Pineal Res 2001;31:264–272. [DOI] [PubMed] [Google Scholar]
- 18.Dubocovich ML, Rivera-Bermudez MA, Gerdin MJ, Masana MI. Molecular pharmacology, regulation and function of mammalian melatonin receptors. Front Biosci 2003;8:d1093–1108. [DOI] [PubMed] [Google Scholar]
- 19.Dubocovich ML. Melatonin receptors: Role on sleep and circadian rhythm regulation. Sleep Med 2007;8 Suppl 3:34–42. [DOI] [PubMed] [Google Scholar]
- 20.Liu J, Clough SJ, Hutchinson AJ, Adamah-Biassi EB, Popovska-Gorevski M, Dubocovich ML. MT1 and MT2 melatonin receptors: A therapeutic perspective. Annu Rev Pharmacol Toxicol 2016;56:361–383. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.McArthur AJ, Hunt AE, Gillette MU. Melatonin action and signal transduction in the rat suprachiasmatic circadian clock: Activation of protein kinase C at dusk and dawn. Endocrinol 1997;138(2):627–634. [DOI] [PubMed] [Google Scholar]
- 22.Cagnacci A. Melatonin in relation to physiology in adult humans. J Pineal Res 1996;21:200–213. [DOI] [PubMed] [Google Scholar]
- 23.Wyatt JK, Dijk DJ, Ritz-De Cecco A, Ronda JM, Czeisler CA. Sleep facilitating effect of exogenous melatonin in healthy young men and women is circadian-phase dependent. Sleep 2006;29:609–618. [DOI] [PubMed] [Google Scholar]
- 24.von Gall C, Weaver DR. Loss of responsiveness to melatonin in the aging mouse suprachiasmatic nucleus. Neurobiol Aging 2008;29(3):464–470. [DOI] [PubMed] [Google Scholar]
- 25.Wu YH, Zhou JN, Van Heerikhuize J, Jockers R, Swaab DF. Decreased MT1 melatonin receptor expression in the suprachiasmatic nucleus in aging and Alzheimer’s disease. Neurobiol Aging 2007;28(8):1239–1247. [DOI] [PubMed] [Google Scholar]
- 26.Harper DG, Volicer L, Stopa EG, McKee AC, Nitta M, Satlin A. Disturbance of endogenous circadian rhythm in aging and Alzheimer disease. Am J Geriatr Psychiatry 2005;13(5):359–368. [DOI] [PubMed] [Google Scholar]
- 27.Wang JL, Lim AS, Chiang WY, et al. Suprachiasmatic neuron numbers and rest-activity circadian rhythms in older humans. Ann Neurol 2015;78(2):317–322. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Hood S, Amir S. The aging clock: circadian rhythms and later life. J Clin Invest 2017;127(2):437–446. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Khachiyants N, Trinkle D, Son SJ, Kim KY. Sundown syndrome in persons with dementia: An update. Psychiatry Investig 2011;8(4):275–287. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Wu YH, Swaab DF. The human pineal gland and melatonin in aging and Alzheimer’s disease. J Pineal Res 2005;38(3):145–152. [DOI] [PubMed] [Google Scholar]
- 31.Benloucif S, Masana MI, Dubocovich ML. Responsiveness to melatonin and its receptor expression in the aging circadian clock of mice. Am J Physiol 1997;273(6):R1855–1860. [DOI] [PubMed] [Google Scholar]
- 32.Gillette MU, McArthur AJ. Circadian actions of melatonin at the suprachiasmatic nucleus. Behav Brain Res 1996;73:135–139. [DOI] [PubMed] [Google Scholar]
- 33.Barinaga M. How jet-lag hormone does double duty in the brain. Science 1997;277(5325(25)):480. [DOI] [PubMed] [Google Scholar]
- 34.Mahlberg R, Kienast T, Hadel S, Heidenreich JO, Schmitz S, Kunz D. Degree of pineal calcification (DOC) is associated with polysomnographic sleep measures in primary insomnia patients. Sleep Med 2009;10(4):439–45. [DOI] [PubMed] [Google Scholar]
- 35.Butcher JN, Dahlstrom WG, Graham JR, Tellegen A, Kaemmer B. The Minnesota Multiphasic Personality Inventory-2 (MMPI-2): Manual for administration and scoring. Minneapolis, MN: University of Minnesota Press; 1989. [Google Scholar]
- 36.Yesavage JA, Brink TL, Rose TL, et al. Development and validation of a geriatric depression screening scale: A preliminary report. J Psychiatr Res 1982;17(1):37–49. [DOI] [PubMed] [Google Scholar]
- 37.Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12(3):189–198. [DOI] [PubMed] [Google Scholar]
- 38.Dijk DJ, Czeisler CA. 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 1995;15(5):3526–3538. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Czeisler CA, Duffy JF, Shanahan TL, et al. Stability, precision, and near-24-hour period of the human circadian pacemaker. Science 1999;284:2177–2181. [DOI] [PubMed] [Google Scholar]
- 40.Duffy JF, Cain SW, Chang AM, et al. Sex difference in the near-24-hour intrinsic period of the human circadian timing system. Proc Nat Acad Sci 2011;108(suppl. 3):15602–15608. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Silva EJ, Wang W, Ronda JM, Wyatt JK, Duffy JF. Circadian and wake-dependent influences on subjective sleepiness, cognitive throughput, and reaction time performance in older and young adults. Sleep 2010;33(4):481–490. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Rechtschaffen A, Kales A. A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. Washington, D.C.: U.S. Government Printing Office; 1968. [DOI] [PubMed] [Google Scholar]
- 43.Sateia MJ, Buysse DJ, Krystal AD, Neubauer DN, Heald JL. Clinical practice guideline for the pharmacologic treatment of chronic insomnia in adults: An American Academy of Sleep Medicine clinical practice guideline. J Clin Sleep Med 2017;13(2):307–349. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Li J, Somers VK, Xu H, Lopez-Jimenez F, Covassin N. Trends in use of melatonin supplements among US adults, 1999-2018. JAMA 2022;327(5):483–485. [DOI] [PMC free article] [PubMed] [Google Scholar]