manipulation of the human circadian system with bright light and melatonin
TRANSCRIPT
MANIPULATION OF THE HUMAN CIRCADIAN SYSTEM WITH BRIGHT LIGHT AND MELATONIN
Stephen John Deacon, BSc Hons
December 1994
A thesis submitted in accordance with the requirements of the University of Surrey for the degree of Doctor of Philosophy
Endocrinology and Metabolism Group School of Biological Sciences University of Surrey GUILDFORD Surrey, GU2 SXH
This thesis is dedicated to my mother and late father,
Anne and Peter, and my brother Phil.
ACKNOWLEDGEMENTS
Firstly, I would like to thank Lara Blann for her unlimited help,
encouragement and support, not to mention her considerable patience throughout this
work. I have thoroughly enjoyed being involved in a relatively new and rapidly
expanding field and I am very grateful to Prof. Josephine Arendt for her enormous
support, expert supervision and contagious enthusiasm for the field. In addition, I
am very grateful to her for encouraging me to attend conferences in many a foreign
land; a Cornishman has never seen so much of the world before !
I am indebted to Judie English for her unfailing support and advice in every
aspect of my work. I would also like to thank Benita Middleton and everyone else
in the Chronobiology laboratory for their help and for making my stay so enjoyable.
I am most grateful to Dr. J. Wright for clinical supervision of all trials and
to the following people for their technical help, advice and medical assistance: Dr.
J. Waterhouse, Dr. D. S. Minors (both at the University of Manchester), Dr. P.
Totterdell (University of Sheffield) and Prof. S. Folkard (University of Wales), Dr.
Sunil Sethi and Dr. Julie Lovegrove (both at the University of Surrey).
I would like to thank the Medical Research Council, the School of Biological
Sciences at the University of Surrey and especially Stockgrand Ltd. for funding my
studies and supporting my attendance at conferences and workshops. Stockgrand
Ltd. (University of Surrey) under the management of Prof. J. Arendt and Judie
English has provided enormous financial support without which much of this clinical
research could not be performed. Finally, but by no means least, a huge thank you to all subjects who took part
in these studies. Good subjects are worth their weight in gold, and I would like to
thank in particular those who were involved in the most arduous and `probing' of
studies, Lara Blann, Sandra and Francis Hill, Nicola Wootton, David Hall, Simon
Gamble, Sean and Julie Lovegrove, Alison Weaver, Mark Ruutel and Tim Marczylo.
3
SUMMARY
The primary aim of this thesis was to develop an experimental model, as close
to real life as possible, to simulate circadian rhythm disturbances associated with jet
lag and shift work. This would then serve to investigate treatment strategies using
exogenous melatonin and bright light. A second aim was to study potential
mechanisms through which melatonin exerts its effects. However, before developing
such models and treatment strategies, it was important to validate a circadian rhythm
marker which gave a reliable assessment of the behaviour of the endogenous body
clock. Many hormones and other blood constituents have been shown to be
influenced by posture through the influence of gravity. Melatonin in plasma and
saliva was found to be concentrated by moving from a supine to a standing position,
and diluted after an opposite change in posture. The endogenous body temperature rhythm is more susceptible to masking
influences than melatonin. A purification method was employed to reduce masking
effects on the temperature rhythm after a simulated time-zone transition (in
collaboration with Drs. Waterhouse and Minors, University of Manchester, UK).
Masking led to overestimates in the rate of adaptation of the body temperature
rhythm. Such methods may provide a more accurate estimation of the behaviour of
the endogenous clock in a normal environment. A number of non-invasive techniques are employed to measure the melatonin
rhythm in saliva, and its metabolite, 6-sulphatoxymelatonin (aMT6s), in urine. These
various methods have not all been compared under phase-shifted conditions. Close
relationships between the calculated peak times of melatonin in plasma and saliva and
of aMT6s in urine, together with the plasma melatonin onset time were established.
In a partially controlled environment, appropriately timed bright light (- 1200
lux) and darkness (< 1 lux) in combination with a shift in the sleep/wake cycle facilitated the adaptation of the endogenous clock and behavioural functions to a
gradual 9h advance and delay phase-shift in external time cues within 3-5 days.
Subjects then resumed their baseline sleep/wake schedule in a normal environment,
thus experiencing a rapid 9h advance or delay phase-shift of local time cues.
4
Adaptation of circadian rhythms and behavioural functions took longer to adapt to
advance phase-shifts than to delay shifts, underlining the preferred direction of
reentrainment of the biological clock. The administration of melatonin (5mg) near bedtime for 3 days after a rapid
9h advance phase-shift significantly improved sleep, alertness and performance
efficiency, even in the presence of bright light (2000 lux) timed to counteract the
phase-shifting effects of melatonin. Improvements using melatonin treatment were
observed within one day and before significant adaptation of the biological clock
occurred. In addition to a phase-shifting effect on the endogenous biological clock,
melatonin can induce acute temperature suppression and sleepiness. A significant
relationship was determined between melatonin-induced temperature suppression and
the subsequent phase-shift in the biological clock after administration of 5mg at 1700h. This relationship was further found to be dose-related, suggesting that acute
changes in body temperature by melatonin may be a primary event in phase-shifting
mechanisms. Significant sleepiness was only associated with pharmacological plasma
melatonin levels. With pharmacological doses, melatonin may exert its beneficial
effects on circadian rhythm disturbances through a combination of transient sedation,
temperature suppression and phase-shifting mechanisms.
5
CONTENTS
Acknowledgements
Summary
List of Figures
List of Tables
List of Abbreviations
CHAPTER 1- INTRODUCTION
3
4
15
21
22
1.1 Consequences of biological rhythmicity 25
1.2 Control of biological rhythmicity 28
1.2.1 Location of the body clock 29
1.2.2 Neural projections between the SCN and the eye 30
1.2.3 The pineal gland 30
1.2.4 Neural projections from the SCN to the pineal gland 32
1.2.5 Cellular mechanisms of melatonin synthesis 32
1.2.6 Synthesis of melatonin 34
1.2.7 Other control mechanisms of melatonin synthesis 35
1.2.8 Melatonin secretion 37
1.2.9 Melatonin metabolism 38
1.2.10 Target sites and potential circadian effects of
melatonin action 40
1.2.11 Summary - potential physiological roles for melatonin 43
1.3 Factors affecting melatonin secretion 44
1.3.1 Light suppression of melatonin secretion in humans 44 1.3.1.1 Intensity of light exposure 1.3., 1.2 Timing of light exposure 1.3.1.3 Duration of light exposure 1.3.1.4 Wavelength of light exposure 1.3.2 Non-visible radiation 1.3.3 Changing day-length (photoperiod)
44
45
46
46
48
48
6
1.3.4 Puberty
1.3.5 Advancing age 1.3.6 Normal menstrual cycle
1.3.7 Body size 1.3.8 Exercise/stress
1.3.9 Alcohol ingestion
51
52
54
54
55
56
1.3.10 Depression and seasonal affective disorder 57
1.3.11 Circadian rhythm disorders 58
1.4 Phase-shifting response of the human circadian system to bright
light and melatonin 59
1.4.1 Range of entrainment 59
1.4.2 Entrainment of the human circadian system to the
light/dark cycle 60
1.4.3 Phase-shifting effects of light 60
1.4.4 Phase-shifting effects of melatonin 64
1.5 Abrupt phase-shifts in local time cues 66
1.5.1 Jet lag 66
1.5.2 Shift-work 68
1.5.3 Susceptibility to jet lag and maladaptation to shift-work 69
1.5.3.1 Age 69
1.5.3.2 Circadian rhythm characteristics 69
1.5.3.3 Personality 70
1.6 Bright light and melatonin treatment for jet lag and shift-work 71
1.6.1 Bright light treatment for jet lag 71
1.6.2 Bright light treatment for shift-work 73
1.6.3 Melatonin treatment for jet lag 74
1.6.4 Melatonin treatment for shift-work 75
1.7 Bright light and melatonin treatment for SAD 76
1.8 Effects of exogenous melatonin on other sleep/wake disorders 77
1.8.1 Blindness 77
1.8.2 Sleep phase syndrome disorders 77
1.8.3 Insomnia 78
7
1.9 Adverse effects of melatonin 79
1.9.1 Endocrinological effects 80
1.10 Acute effects of melatonin and bright light 81
1.10.1 Sedative effects of melatonin 81
1.10.2 Arousal/stimulating effects of bright light 82
1.10.3 Stimulating effect of bright light on body temperature 83
1.10.4 Acute effects of exogenous melatonin on temperature 84
1.11 Potential pharmacologic and behavioural treatment strategies for circadian rhythm disturbances 85
1.11.1 Manipulation of the sleep/wake cycle 85
1.11.2 Melatonin agonists and antagonists 85
1.11.3 Physical activity and hypnotics 87
1.12 Research proposal and aims of thesis 89
CHAPTER 2- CLINICAL CONSIDERATIONS PRIOR TO STUDIES.
PHYSIOLOGICAL AND BEHAVIOURAL MEASUREMENTS AND ANALYSIS
2.1 Clinical trials: general considerations 93
2.1.1 Subject criteria 93
2.1.2 Details of the melatonin preparation 94
2.1.3 Health and safety effects of light 95
2.2 Collection and storage of samples 96
2.3 Radioimmunoassay of melatonin and aMT6s 97
2.3.1 Basic principle and summary of procedure 97
2.3.2 RIA of melatonin in human plasma 98
2.3.3 RIA of melatonin in human saliva 102
2.3.4 RIA of 6-sulphatoxymelatonin in human urine 105
2.4 Light intensity measurement 108
2.5 Core body temperature recording 108
2.6 Activity diaries 108
2.7 Behavioural measurements 109
2.7.1 Mood and sleep quality ratings 110
8
2.7.2 Sleep logs 110
2.7.3 Performance measurements: Search and memory (SAM) test 111
2.8 Cosinor analysis 112
CHAPTER 3- INFLUENCE OF POSTURE ON MELATONIN
CONCENTRATIONS IN PLASMA AND SALIVA
3.1 Introduction 117
3.2 Subjects and methods 118
3.2.1 Subjects 118
3.2.2 Design 119
3.2.3 Sampling 119
3.2.4 Statistical analysis 120
3.3 Results 121
3.4 Discussion 127
CHAPTER 4- INDUCING CIRCADIAN RHYTHM DISTURBANCES IN
PARTIAL ENVIRONMENTAL ISOLATION: AN INITIAL STUDY.
COMPARISON OF THE MELATONIN RHYTHM IN PLASMA, URINE AND
SALIVA BEFORE AND AFTER A FORCED PHASE SHIFT
4.1 Introduction 131
4.2 Subjects and methods 133
4.2.1 Subjects 133
4.2.2 Study design 134
4.2.3 Sampling 135
4.2.4 Behavioural measurements 137
4.2.5 Hormone assays 137
4.3 Statistical analysis 138
4.3.1 Hormonal and alertness rhythms 138
4.3.2 Sleep data and daily means for performance and alertness 138
4.4 Results 139
9
4.4.1 Melatonin 139
4.4.2 Alertness 147
4.4.3 Performance 147
4.4.4 Sleep 148
4.5 Discussion 153
CHAPTER 5- INDUCING CIRCADIAN DISTURBANCES IN PARTIAL
ENVIRONMENTAL ISOLATION: A RAPID 9h ADVANCE PHASE-SHIFT
AFTER A GRADUAL 9h DELAY SHIFT
5.1 Introduction 160
5.2 Subjects and methods 161
5.2.1 Subjects 161
5.2.2 Study design 161
5.2.3 Sampling 165
5.2.4 Behavioural measurements 165
5.2.5 aMT6s assays 165
5.3 Statistical analysis 166
5.3.1 Temperature and aMT6s rhythms 166
5.3.2 Mood and performance rhythms 166
5.3.3 Sleep data and daily means for mood and performance 167
5.4 Results 167
5.4.1 Ambient light monitoring 168
5.4.2 Baseline 168
5.4.3 Gradual forced 9h delay phase-shift 172
5.4.4 Adaptation to a 9h rapid advance phase-shift (placebo treatment) 181
5.5 Discussion 188
5.5.1 Gradual 9h delay phase-shift 188
5.5.2 Rapid 9h advance phase-shift 190
10
CHAPTER 6- INFLUENCE OF MELATONIN ON ADAPTATION TO A
LARGE RAPID ADVANCE PHASE-SHIFT IN THE PRESENCE OF
CONFLICTING BRIGHT LIGHT
6.1 Introduction 195
6.2 Subjects and methods 197
6.2.1 Subjects 197
6.2.2 Study design 197
6.2.3 Sampling 199
6.2.4 Behavioural measurements 199
6.2.5 Hormone assays 199
6.3 Statistical analysis 200
6.3.1 aMT6s and core body temperature profiles 200
6.3.2 Mood and performance daily means and sleep data 200
6.3.3 Mood and performance rhythms 200
6.4 Results 201
6.4.1 Ambient light monitoring 201
6.4.2 aMT6s 202
6.4.3 Temperature 202
6.4.4 Mood 202
6.4.5 Performance 207
6.4.6 Sleep 207
6.5 Discussion 216
CHAPTER 7- PURIFICATION OF THE CORE BODY TEMPERATURE
RHYTHM AFTER SIMULATED TIME ZONE TRAVEL
7.1 Introduction 222
7.2 Methods 224
7.2.1 Purification procedure 225
7.3 Statistical analysis 227
7.3.1 Baseline (D-2 to DO) 227
11
7.3.2 Gradual 9h delay phase-shift (D 1 to D5) 227
7.3.3 Rapid 9h advance phase-shift (D6 to D11) 228
7.4 Results 229
7.4.1 Normal baseline conditions 229
7.4.2 Adaptation to a gradual 9h delay phase-shift 229
7.4.3 Adaptation to a rapid 9h advance shift 233
7.5 Discussion 239
7.5.1 Normal baseline conditions 239
7.5.2 Adaptation to a gradual 9h delay phase-shift 240
7.5.3 Adaptation to a rapid 9h advance phase-shift 241
CHAPTER 8- INDUCING CIRCADIAN RHYTHM DISTURBANCES IN
PARTIAL ENVIRONMENTAL ISOLATION: ADAPTATION TO A 9h DELAY
PHASE-SHIFT AFTER A GRADUAL 9h ADVANCE SHIFT
8.1 Introduction 247
8.2 Subjects and methods 248
8.2.1 Subjects 248
8.2.2 Study design 248
8.2.3 Sampling 251
8.2.4 Behavioural measurements 252
8.2.5 aMT6s 252
8.3 Statistical analysis 252
8.3.1 Temperature and aMT6s rhythms 252
8.3.2 Mood and performance rhythm 253 8.3.3 Sleep data and daily means for mood and performance 253
8.4 Results 254
8.4.1 Baseline 254
8.4.2 Adaptation to a gradual 9h advance phase-shift 258
8.4.3 Adaptation to a rapid 9h delay phase-shift 263
8.5 Discussion 272
8.5.1 Adaptation to a gradual 9h advance phase-shift 272
12
8.5.2 Adaptation to a rapid 9h delay phase-shift 273
CHAPTER 9- ACUTE EFFECTS OF A SINGLE MELATONIN TREATMENT
ON CIRCADIAN PHASE, BODY TEMPERATURE AND BEHAVIOUR
9.1 Introduction
9.2 Subjects and methods 9.2.1 Subjects
9.2.2 Study design
9.2.3 Sampling and measurements 9.2.4 Hormone assay
9.3 Statistical analysis 9.4 Results
277
279
279
279
280
280
281
281
9.4.1 Acute effects of melatonin treatment (D3) 281
9.4.2 Delayed effects of melatonin (one day after treatment D4) 284
9.4.3 Relationship between temperature suppression, exogenous
melatonin levels and degree of shift 288
9.5 Discussion 290
CHAPTER 10 - DOSE-RESPONSE EFFECTS OF MELATONIN ON
CIRCADIAN PHASE, BEHAVIOUR AND TEMPERATURE SUPPRESSION
10.1 Introduction
10.2 Subjects and methods 10.2.1 Subjects
10.2.2 Preparation of melatonin/placebo 10.2.3 Study design
10.2.4 Sampling 10.2.5 Behavioural measurements 10.2.6 Melatonin assay
10.3 Statistical analysis 10.4 Results
293
294
294
294
295
296
297
298
298
300
13
10.4.1 Acute effects of melatonin treatment (D3) 300
10.4.2 Delayed effects of melatonin treatment (one day after
treatment, D4) 308
10.4.3 Relationship between individual degree of phase-shift
and magnitude of temperature suppression 315
10.4.4 Relationship between individual exogenous plasma melatonin levels, temperature suppression and degree of phase-shift 315
10.5 Discussion 317
CHAPTER 11 - GENERAL DISCUSSION
11.1 Determining the behaviour of the endogenous body clock 323
11.1.1 Masking 323
11.2.2 Different markers of melatonin secretion 324
11.2 An experimental model to study adaptation strategies 326
11.2.1 Adaptation to gradual phase-shifts 326
11.2.2 Adaptation to rapid shifts - simulation of jet lag and
maladaptation to shift work 328
11.2.3 Preferred direction of adaptation 328
11.3 Effects of bright light and melatonin during adaptation to rapid
phase-shifts 329
11.4 Repeated melatonin administration 330
11.5 Mechanisms of action of melatonin 331
11.5.1 Phase-shifting effect on the biological clock 331
11.5.2 `Hypnotic' effects 332
11.6 Masking or cue for entrainment? 334
11.7 Relationship between the acute effects and phase-shifting
properties of melatonin and bright light 335
11.8 Future work 337
PUBLICATIONS AND PRESENTATIONS REFERENCES
341 344
14
LIST OF FIGURES
CHAPTER 1
1.1 Schematic representation of the neural pathway between light
perception in the eye and the pineal gland 31
1.2 Schematic diagram showing neural pathway from light/darkness
input to melatonin 33
1.3 Synthesis of melatonin 36
1.4 Major melatonin metabolism pathways in the liver 39
1.5 Human phase-response curves for bright light and melatonin 63
CHAPTER 2
2.1 Parameters of a circadian rhythm
CHAPTER 3
113
3.1 Immediate effect of posture on plasma and salivary melatonin levels in alternate postural positions 123
3.2 Long-term effects of posture on plasma melatonin levels 125
3.3 Long-term effects of posture on salivary melatonin levels 126
CHAPTER 4
4.1 Diagrammatic design of the study 134
4.2 Circadian plasma melatonin profiles before and after a forced
phase-shift 142
4.3 Urinary aMT6s circadian profiles before and after a forced phase-shift 143
4.4 Circadian melatonin profiles in saliva before and after a forced
phase-shift 144
4.5 Relationship between acrophase estimates for melatonin, aMT6s,
15
alertness and melatonin onset time 145
4.6 Alertness daily means and acrophase estimates before, during
and after a forced phase-shift 149
4.7 Subjective alertness profiles before and after a forced phase-shift 150
4.8 Performance measures before, during and after a forced phase-shift 151
4.9 Subjective sleep parameters before, during and after a forced
phase-shift 152
CHAPTER 5
5.1 Diagrammatic design of the study 164
5.2 aMT6s profiles before and during a gradual 9h delay phase-shift 169
5.3 Temperature profiles before and during a gradual 9h delay phase-shift 170
5.4 Alertness profiles before and during a gradual 9h delay phase-shift 171
5.5 Acrophases of the aMT6s and temperature rhythms during a gradual 9h delay phase-shift 173
5.6 Adaptation of the alertness rhythm during a gradual delay
phase-shift 174
5.7 Adaptation of the SAM-1 response time performance rhythm during
a gradual delay phase-shift 175
5.8 Adaptation of the SAM-5 response time performance rhythm during
a gradual delay phase-shift 176
5.9 Daily means for mood before and during a gradual 9h delay
phase-shift 177
5.10 Subjective sleep measures during a gradual forced 9h delay
phase-shift 178
5.11 Daily means for SAM-1 performance measures before and during
a gradual forced 9h delay phase-shift 179
5.12 Daily means for SAM-5 performance measures before and during
a gradual forced 9h delay phase-shift 180
5.13 Adaptation of sleep to a rapid 9h advance phase-shift 182
5.14 Adaptation of mood to a rapid 9h advance phase-shift 183
16
5.15 Adaptation to a rapid 9h advance phase-shift: daily mean
performance 184
5.16 Adaptation of aMT6s and temperature rhythms to a rapid 9h advance
phase-shift 185
5.17 Adaptation of the temperature to a rapid 9h advance phase-shift 186
5.18 Direction of re-entrainment of individual aMT6s rhythms after a rapid 9h advance phase-shift 187
CHAPTER 6
6.1 Diagrammatic design of the study 198
6.2 Effect of melatonin and bright light on aMT6s profiles during adaptation
to a rapid 9h advance phase-shift 203
6.3 Effect of melatonin and bright light on the adaptation of the aMT6s
acrophase to a rapid 9h advance phase-shift 204
6.4 Influence of melatonin and bright light on the adaptation of the
temperature rhythm to a rapid 9h advance phase-shift 205
6.5 Effect of melatonin and bright-light on the adaptation of the temperature
cosinor rhythm to a rapid 9h advance phase-shift 206
6.6 Influence of melatonin and bright-light on the adaptation of mood
to a rapid 9h advance phase-shift 208
6.7 Influence of melatonin and bright-light on the adaptation of the
alertness rhythm to a rapid 9h advance phase-shift 209
6.8 Influence of melatonin and bright-light on the adaptation of the
alertness cosinor rhythm to a rapid 9h advance phase-shift 210
6.9 Influence of melatonin and bright-light on the adaptation of SAM-1
performance to a rapid 9h advance phase-shift 211
6.10 Influence of melatonin and bright-light on the adaptation of SAM-5
performance to a rapid 9h advance phase-shift 212
6.11 Influence of melatonin and bright-light on the adaptation of SAM-1
performance cosinor rhythm to a rapid 9h advance phase-shift 213
6.12 Influence of melatonin and bright-light on the adaptation of SAM-5
17
performance cosinor rhythm to a rapid 9h advance phase-shift 214
6.13 Influence of melatonin and bright-light on the adaptation of sleep
to a rapid 9h advance phase-shift 215
CHAPTER 7
7.1 Purification of the core body temperature rhythm 230
7.2 Comparison of the rate of adjustment between the aMT6s and
temperature (raw and purified) rhythms to a gradual 9h delay shift 231
7.3 Relationship between the rate of adaptation of purified and raw
temperature rhythms and the aMT6s rhythm to a gradual 9h delay
phase-shift 232
7.4 Effect of purification on the temperature rhythm amplitude during a
gradual 9h delay phase-shift 234
7.5 Effect of purification on the rate of adjustment of the temperature
rhythm to a rapid 9h advance phase-shift 235
7.6 Effect of purification on the rate of adjustment of the temperature
rhythm amplitude to a rapid 9h advance phase-shift 237
7.7 Effect of melatonin and bright light on the adaptation of the purified
body temperature rhythm 238
CHAPTER 8
8.1 Diagrammatic design of the study 250
8.2 Temperature profiles before and during a gradual 9h advance
phase-shift 255
8.3 Adaptation of the urinary aMT6s rhythm to a gradual 9h advance
phase-shift 256
8.4 Alertness profiles before and during a gradual 9h advance phase-shift 257
8.5 Acrophases of aMT6s and temperature rhythms during a gradual 9h
advance phase-shift 259
8.6 Adaptation of mood to a gradual 9h advance phase-shift 260
18
8.7 Adaptation to a gradual 9h advance phase-shift: daily mean
performance 261
8.8 Adaptation of sleep to a gradual 9h advance phase-shift 262
8.9 Alertness profiles during adaptation to a rapid 9h delay phase-shift 264
8.10 Adaptation of mood to a rapid 9h delay phase-shift 265
8.11 Adaptation to a rapid 9h delay phase-shift: daily mean performance 266
8.12 Adaptation of sleep to a rapid 9h delay phase-shift 267
8.13 Adaptation of the temperature rhythm to a rapid 9h delay phase-shift 268
8.14 Adaptation of aMT6s and temperature rhythm to a rapid 9h delay
phase-shift 269
8.15 Adaptation of the urinary aMT6s rhythm five days after a rapid 9h
delay phase-shift 270
8.16 Direction of reentrainment of individual masked temperature rhythms
after a rapid 9h delay phase-shift 271
CHAPTER 9
9.1 Saliva concentrations of melatonin after ingestion of 5mg
of melatonin taken orally in gelatin/lactose capsules 282
9.2 Acute effects of 5mg melatonin on body temperature and alertness 283
9.3 Acute phase-shifting effect of 5mg melatonin treatment on the
endogenous melatonin rhythm one day after treatment 285
9.4 Influence of melatonin on the body temperature nadir during the
second night after late afternoon treatment 286
9.5 Effect of late afternoon melatonin treatment on sleep the
following night 287
9.6 Relationship between individual degree of phase-shift and
temperature suppression 289
19
CHAPTER 10
10.1 Plasma concentrations of melatonin after ingestion of low doses of
melatonin taken orally in corn oil 302
10.2 Acute dose-dependent suppression of body temperature by melatonin 303
10.3 Dose-dependent decrease in subjective alertness immediately after
melatonin treatment 304
10.4 Acute dose-dependent effects of melatonin on performance accuracy 306
10.5 Acute dose-dependent effects of melatonin on performance efficiency 307
10.6 Dose-response relationship between oral melatonin and degree of
phase-shift in melatonin rhythm one day after treatment 310
10.7 Dose-dependent phase advance in the temperature nadir during the
second night after melatonin treatment 311
10.8 Delayed effects of low oral doses of melatonin on the alertness rhythm
one day after treatment 312
10.9 Dose-dependent effects on sleep on the first night following melatonin
treatment. 314
10.10 Relationship between individual degree of phase-shift and temperature
suppression 316
20
LIST OF TABLES
Table 2.1 Percent side-effects reported more than once after taking
5mg melatonin to alleviate jet lag (from 1984 to Nov. 1994) 95
Table 3.1 Mean ± SD difference and range in % change in melatonin
levels between standing and supine positions. 124
Table 4.1 Correlation and linear regression analysis of acrophase
estimates between plasma melatonin, salivary melatonin
and urinary aMT6s, together with plasma melatonin onset 146
Table 7.1 Individual differences in purified temperature acrophase
estimates (in hours) from mean baseline (D-2 to DO) during
adaptation to a 9h advance phase-shift. 236
Table 10.1 Acute effects of melatonin treatment 305
Table 10.2 Delayed effects of melatonin treatment 313
21
LIST OF ABBREVIATIONS
aMT6s 6-sulphatoxymelatonin
ANOVA Analysis of variance AUC Area under the curve
cAMP cyclic 3', 5'-adenosine monophosphate
D Day of study DGDW Double glass distilled water
EEG Electroencephalogram
F Female
FDA Food and drug administration
GABA Gamma-aminobutyric acid
h Hours
5-HT 5-Hydroxytrypamine
HIOMT Hydroxy-O-methyltransferase
ICER Inducible cAMP early repressor
IGL Intergeniculate leaflet
LSN Lateral septal nucleus
LH Luteinizing hormone
M Male
MT Melatonin
N Night of study
NA Noradrenaline
NAS N-acetylserotonin
NAT N-acetyltransferase
NREM Non-rapid eye movement
ns Non significant
PPO 2,5-diphenyloxazole
POPOP 1,4-bis-[2-(5-phenyloxazolyl)]benzene
PRC Phase-response curve
PVN Paraventricular nucleus
REM Rapid eye movement
22
RHT Retinohypothalamic tract
RIA Radioimmunoassay
SAD Seasonal affective disorder
SAM Search and memory test
SCG Superior cervical ganglion SCN Suprachiasmatic nuclei
SD Standard deviation
SEM Standard error of the mean UV Ultraviolet
VAS Visual analogue scale
23
Tempora mutantur, et nos mutwnur in illis
Times change, and we change with them LYLY F. uphues 1.1578
CHAý. ''rER 1 INTRODUCTION
1.1 CONSEQUENCES OF BIOLOGICAL RHYTHMICITY
Most organisms are able to predict and adapt to changes in the environment,
phenomena that are essential to their survival and well-being in an ever changing and
often unfriendly environment. From the unicellular flagellate to the higher vertebrate
this anticipation and response to timed events in the environment is important in
preparing the organism for a variety of events such as waking and activity throughout
the day, `winding down' in preparation for sleep at night, reproductive function and
hibernation at certain times of year.
These preparations are brought about by various rhythms in physiological and
behavioural functions. The phase (time) relationship between various rhythms and
the environment needs to be coordinated so that the correct biological response occurs
at the right time of day or year. For example, for animals that breed only once a
year, it is vitally important that heightened reproductive function coincides in both
males and females so that the offspring may be born at the optimum time of year for
survival.
In the absence of all environmental influences, many biological rhythms still
persist but `free-run' with slightly different periodicities. It is because of this
endogenous nature of biological rhythms that organisms are considered to have an
internal rhythm-generating system or a biological clock. Possession of an inherent
body clock allows the measurement of environmental time so that the body can
predict regular external changes instead of just reacting to them, allowing an optimum
response. The most prominent endogenous rhythms are those related to the daily 24
hour light-dark cycle. In humans, the endogenous biological timing system free-runs
25
with a period of close to 25 hours, hence the term "circadian", from the latin circa -
about and dies - day (Aschoff, 1965; Wever, 1979,1983,1986; Lewy and
Newsome, 1983; Kennaway and Van Dorp, 1991). A 25h internal clock is obviously
an inaccurate forecaster of daily (24h) changes in the environment. Hence, stable
synchronization of the endogenous biological clock to the environment depends on
external synchronising factors or time cues, such as temperature, humidity, social
cues and the light/dark cycle. Many of these rhythms are entrained to 24 hours
mainly by the daily light/dark cycles. This ability of the circadian system to be
influenced by external zeitgebers also allows the body to adapt to changes or phase-
shifts in environmental time.
Human beings are unique in that they consciously and repeatedly alter their
daily schedules to suit their own particular lifestyles, overriding the natural
mechanisms which influence the rhythmicity of the circadian timing system. The
natural tendency to fall asleep at times of great propensity, which so many organisms
are powerless against, can be stalled until either the urge to sleep becomes
overwhelming and sleep becomes inevitable, or the disruption of the sleep pattern
becomes permanent through continual change and the individual may begin to suffer
from a sleep disorder. This conscious decision to change the timing of retiring to bed
and waking up is invaluable to many individuals especially in the modern world
where they may have to work night shifts or fly to a new time zone. However, there
is a price to pay.
Abrupt shifts in local time cues or zeitgebers can result in various rhythmic
functions becoming out of phase with each other and the external environment. The
natural tendency of the internal body clock to adapt to such changes is gradual and
26
slow. It may take days or even weeks for some circadian rhythms to resynchronize
to a large phase-shift. As a consequence, the predictive ability of the body clock
becomes transiently affected with deleterious effects on human physiology and
behaviour. For example, humans can become sluggish, cold and sleepy during
waking hours when starting on a new night shift because the circadian timing system
has prepared the body for rest and sleep. Subsequent decrements in performance and
alertness can increase the risk of accidents or errors which may reduce productivity
and safety. Since shift workers include air-traffic controllers, nuclear submarine and
power plant operators, emergency doctors and intensive care nurses, it is important
to understand circadian rhythms in the work place not just from an ergonomic point
of view but also a safety one. In addition, disturbances in the circadian system have
been identified in a number of disorders. Treatment strategies to facilitate the
adaptation or resetting of biological rhythms are therefore of enormous value.
Light (and darkness) together with the hormone melatonin are closely involved
with the circadian timing system and have been proposed as potential treatments for
circadian rhythm disorders.
This chapter will review current knowledge on the structure, organisation and
regulation of the circadian timing system, with particular emphasis on the melatonin
circadian rhythm, and its influences on human physiology and behaviour.
Disturbances in this pacemaker system will then be discussed, with a review of
potential treatment strategies concentrating on the use of light and melatonin. There
are a number of other more immediate effects of light and melatonin on human
physiology and behaviour, which may or may not be related to their ability to
influence the circadian timing system, which will also be discussed. Finally, the
27
melatonin secretory rhythm is used as a circadian rhythm marker of the internal
biological clock and exogenous factors affecting this rhythm will be described.
1.2 CONTROL OF BIOLOGICAL RHYTHMS
There is accumulating evidence for a multioscillatory structure of the circadian
system in rodents and humans coupled through the suprachiasmatic nuclei of the
hypothalamus (SCN). In everyday conditions all circadian rhythms such as the
sleep/wake cycle, body temperature and melatonin are synchronized or entrained with
each other and environmental time. In conditions free from environmental time cues
and where the sleep/wake cycle is altered, the temperature and sleep/wake rhythms
may dissociate from one another with different free-running periods of approximately
24.5h and 33h, respectively (Aschoff, 1965; Wever, 1979). Other physiological
rhythms can assume a periodicity associated with either the temperature or sleep/wake
rhythm (Moore-Ede et al, 1983).
Evidence for the involvement of two subsidiary oscillators controlling the
circadian timing system has been used by Pittendrigh and Daan (1976) and Illnerovä
and Vanecek (1982), to describe the response of animals to photoperiodic changes in
the environments. One oscillator, the morning or "dawn" oscillator is thought to
regulate the transition between night and day and control the offset of melatonin
secretion. The other oscillator, the evening or "dusk" oscillator, may trigger the
onset of melatonin secretion by controlling the transition between day and night, see
section 1.3.3. In humans there is, however, an indication of a third pacemaker which
drives endogenous memory performance rhythms with an inherent period of 21 hours
28
(Folkard et al, 1983), thus underlining the complex nature of the system. However,
the available evidence at present indicates that all circadian rhythms, with the
exception of visual sensitivity (Terman and Terman, 1985) and activity anticipatory
to feeding (Krieger et al, 1977; Stephan et al, 1979; Boulos et al, 1980; Clarke and
Coleman, 1986), are abolished by SCN lesions.
1.2.1 Location of the body clock
The SCN, located at the base of the hypothalamus in close proximity to the
optic chiasm, are thus thought to be the site of the major circadian pacemaker.
Surgical isolation of SCN cells from other parts of the brain leaves a persistent
circadian rhythm of neural activity in the SCN which is no longer found in other parts
of the brain (Inouye and Kawamura, 1979). In vitro, SCN cells generate a 24 hour
rhythm in neural activity (Green and Gillette, 1982; Groos and Hendriks, 1982).
Taken together, these studies indicate that if there are no biochemical, neural or
thermal effects influencing the isolated SCN, then the SCN must have a self-contained
oscillator (at least in rodents).
Neural transplantation procedures have provided further support that the
circadian timing system originates from within the SCN. Circadian rhythmicity
reappears after foetal SCN tissue is transplanted into rodents with their SCN
destroyed (Lehman et al, 1987; Sawaki et al, 1984; Drucker-Colin et al, 1984). In
addition, transplanting SCN tissue from hamster mutants with an endogenous
circadian period of 20 or 22 hours (instead of the typical 24 hours), into hamsters
whose own 24 hour clock has been destroyed, will restore rhythmicity that reflects
29
the donor's circadian period (Ralph et al, 1990).
1.2.2 Neural Projections between the SCN and the eye
The SCN receive time of day information from the light-dark cycle through
monosynaptic optical pathways from the retina mainly via the retinohypothalamic tract
(RHT), but also from the intergeniculate leaflet (IGL) of the lateral geniculate body
(the geniculohypothalamic tract) (Moore, 1973; Pickard et al, 1987). If these
connections are destroyed the circadian rhythms lose their entrainment and free-run
(Rusak, 1977). Potential neurotransmitters thought to be involved in conveying
light/dark signals via the RHT include glutamate and possibly acetylcholine (Earnest
and Turek, 1985; Keefe et al, 1987; Meijer et al, 1988). Connections from the IGL
to SCN consist of mainly gamma-aminobutyric acid fibres and projections which
contain the neurotransmitter neuropeptide Y (Moore and Speh, 1993). Other
neurotransmitters such as 5-hydroxytryptamine (5-HT), vasoactive intestinal peptide
and vasopressin have also been suggested as mediators of photic information
(Reghunandanan et al, 1993).
1.2.3 The Pineal Gland
The end organ of this visual system of photoperiodic time measurement is the
pineal gland which releases the hormone melatonin (N-acetyl-5-methoxytryptamine),
during darkness in an entrained individual (Wurtman et al, 1963). There are other
secretory products of the pineal gland which include other indoles and peptides which
30
Figure 1.1 Schematic Representation of the Neural Pathway between Light Perception in the Eye and the
Pineal Gland.
Paraventricular Nucleus
Suprachiasmatic nucleus
Eye
Chiasm
Brainstem
Pineal Gland
31
may have biological activity (Pevet, 1983a, 1983b), but their significance is as yet
unclear.
1.2.4 Neural projections from the SCN to the pineal gland
The neural connections which link the SCN to the pineal gland was first
demonstrated by Berk and Finklestein (1981) and Stephan et al (1981) who identified
projections from the SCN to the paraventricular nucleus of the hypothalamus (PVN)
with further projections leading into the sympathetic chain of the spinal cord and
synapsing in the superior cervical ganglion (SCG), Figure 1.1. Post-ganglionic
sympathetic nerves from the SCG release noradrenaline (NA) which acts through
adrenergic receptors on the pineal gland cell (pinealocyte) membrane to stimulate the
production melatonin. Destruction and stimulation of this pathway has confirmed its
neural role in melatonin production (Klein et al, 1983; Yanovski et al, 1987).
1.2.5 Cellular mechanisms of melatonin synthesis
During the day, light stimulates the activity of the SCN which ultimately
inhibits the release of noradrenaline (NA) from sympathetic neurons in the pineal
gland. However, at night this inhibitory influence ceases and NA is released (Wolfe
et al, 1962; Brownstein and Axelrod, 1974) and interacts with specific adrenergic
receptors on the pinealocytes to stimulate the production of melatonin, Figure 1.2
(Parfitt et al, 1976).
On interaction with the 81-adrenergic receptor, adenylate cyclase is stimulated
32
Figure 1.2 Schematic diagram showing neural pathway from light/darkness input to melatonin secretion
Light/ Eye darkness
y
Pineal gland
Spinal Cord
Tryptophan
'Serotonin I N-Uc'c tyl
trujr. 5fýrusý. ý
N-acetyl serotonin
I Hyclro-gillclcyle-o-
ý ýnýiel! 1 ýI ýl/raii.
iý /erase ýf.
Hypothalamus
N-acetvl Beta I transferase A<-r-m adreriergic synthesis receptor
L Alpha x adrelleTgic receptor
Superior cervical ganglion
`1 Noradrenalin release from
nerve terminal
*0)
via a guanine nucleotide-binding regulatory protein, resulting in an increase in the
intracellular second messenger cyclic 3', 5'-adenosine monophosphate (CAMP,
Axelrod and Zatz, 1977). Elevated levels of cAMP stimulate the production of N-
acetyltransferase (NAT) which is usually the rate limiting step in the production of
melatonin (Klein et at, 1970). There is also activation of an a, -adrenergic receptor
by NA which increases intracellular Cat+, stimulating protein kinase C. This in turn
amplifies the activation of cAMP and so augments the production of NAT (Sugden
et at, 1985a, 1985b; Sugden et at, 1987; Vanacek et at, 1985) through a mechanism
involving gene transcription and translation. A third adrenergic receptor (a2) has
been identified recently in the mammalian pineal gland but its involvement is
uncertain (Simonneaux et at, 1991; Schaad and Klein, 1992).
Recently, rhythmic circadian control of gene expression has been demonstrated
in the pineal gland. Stehle et at (1993) have identified a potent regulator of cAMP-
induced gene transcription known as ICER (inducible cAMP early repressor) which
shows high levels during the night and low levels in the day. It is suggested that
ICER may regulate NAT activity by repressing NAT gene expression (Takahashi,
1993).
1.2.6 Synthesis of melatonin
Melatonin is synthesized primarily in the pineal gland although synthesis also
occurs in the retina (Pang et al, 1977; Vakkuri et al, 1987), the Harderian gland
(Pang et al, 1977; Menendez-Pelaez et al, 1987) and the gut (Raikhlin et al, 1975;
Huether et al, 1992; Yaga et al, 1993). However, in mammals the amount of
34
melatonin produced from these non-pineal sources is considered to be small and
probably used locally since in most studies removal or destruction of the pineal results
in plasma levels or levels of its metabolite in urine which are undetectable by
radioimmunoassay (Arendt et at, 1980; Lincoln et at, 1981; Bittman et at, 1983a, b)
or gas chromatography - mass spectrometry (Markey and Buell, 1982; Tetsuo et at,
1982; Neuwelt and Lewy, 1983).
Melatonin is synthesized from its precursor 5-HT which is formed by
hydroxylation and decarboxylation of the amino acid, tryptophan (Axelrod, 1974),
Figure 1.3. The rise in NAT activity during the night through sympathetic
innervation from the SCN then converts serotonin to N-acetylserotonin (NAS,
Weissbach et at, 1961); NAT catalyses the rate limiting step in the production of
melatonin (Klein and Weller, 1970; Klein, 1979). Elevated levels of a second
enzyme, hydroxy-O-methyltransferase (HIOMT), during the night catalyses the
conversion of NAS to melatonin (Axelrod and Weissbach, 1960).
Practically every aspect of melatonin's synthesis and release is rhythmic
(Klein, 1979), even regulation of gene expression of the rate limiting enzyme (Stehle
et at, 1993).
1.2.7 Other control mechanisms of melatonin synthesis
A number of receptor sites have been identified on the pineal gland for
glutamate, gamma-amino-n-butyric acid (GABA), 5-HT, benzodiazepines, dopamine,
steroid hormones and prostaglandins but their function remains unknown (Cardinali
and Vacas, 1978; Cardinali and Ritta, 1983; Ebadi and Govitrapong, 1986). In vivo,
35
Figure 1.3 Synthesis of Melatonin
CH2CH(NH2)0O0H
TRYPTOPHAN
H I 7iyptophan hydroxylase
i HO CH2CH(NH2)COOH
5-HYDROXYTRYPTOPHAN
H I
Aromatic amino acid decarboxylase
HO + CH2CH2NH2
SEROTONIN (5-hydroxyttyptamine)
H
N-acetyltransferase i
HO _I CH2CH2NHCOCH3
N-ACETYLSEROTONIN (5-hydroxy-N-acetylthyptamine)
H I
Hydroxyindole-O-methyltransferase
CH3o CH2CH2NHCOCH3
MELATONIN (5-metho)cy-N-acetylhyptamine)
H
36
the hormonal modulation of melatonin synthesis appears to have a negligible effect
(Champney et al, 1985,1986).
1.2.8 Melatonin secretion
The melatonin profile in peripheral blood is a good reflection of pineal
melatonin synthesis at almost the same time (Wilkinson et al, 1977; Illnerovä et al,
1978), since the pineal gland releases the indole directly and primarily into the blood
stream rather than the cerebral spinal fluid, with minimal storage (Rollag et al, 1978;
Pang, 1985). Highly lipophilic in nature, melatonin is readily diffusible and rapidly
distributed throughout all tissues (Kopin et al, 1961).
The secretory rhythm of melatonin is characterised by an onset of secretion
in humans around 2100-2200h, rising to a peak around 0200-0300h and falling to
daytime levels around 0800-0900h in the morning. The melatonin secretory rhythm
is a strongly inherent rhythm, persisting in the absence of all environmental time cues
with endogenous periodicity of close to 25h (Lewy and Newsome, 1983; Arendt et
al, 1985a; Wever, 1986; Kennaway and Van Dorp, 1991). The timing of the free-
running rhythm is very precise and the start of its production can sometimes be
predicted in advance with an accuracy of within 30 minutes (Lewy and Newsome,
1983). The shape, period and amplitude of the plasma melatonin circadian rhythm
is highly reproducible in healthy individuals entrained to 24h (Broadway et al, 1987),
and may therefore provide a reliable indication of the behaviour of the internal
biological clock.
A measurable melatonin rhythm may not however be essential for the well-
37
being of an individual. Some people exhibit very low levels of melatonin (< 10
pg/ml) in the plasma without any apparent detrimental effects on the circadian system
or health (Arendt et al, 1985). Since probable melatonin receptors have been
identified in the human hypothalamus pineal melatonin may act locally in much lower
concentrations in these individuals.
1.2.9 Melatonin metabolism
In humans, the metabolism of melatonin occurs predominantly in the liver to
form 6-hydroxymelatonin and N-acetylserotonin, which are then excreted in the urine
as conjugates of the glucuronide or sulphate (Jones et al, 1969; Young et al, 1985),
Figure 1.4. Vakkuri et al (1987) identified a further metabolite, a cyclic isomer of
2-hydroxymelatonin, which accounted for 5% of the total urinary melatonin
metabolites after ingestion of 100mg exogenous melatonin. In rabbits, a different
pathway exists in the central nervous system whereby cleavage of the indole ring
results in the formation of kynuranamines (Hirata et al, 1974).
The end products of melatonin metabolism found in urine provide a means of
measuring a related substance which reflects melatonin output from the pineal in a
non-invasive and practical way. aMT6s is a good index of plasma melatonin
secretion in both timing and amplitude (Bojkowski et al, 1987b).
38
Figure 1.4 Major Melatonin Metabolism Pathways in the Liver
HO CH2CH2NH0O0H3 CH3O
H
5-Hydroxy-N-acetyltryptamine (N-acetylserotonin)
H MELATONIN
COCH3 Cyclic 2-hydroxymelatonin
CH2CH2NHCOCH3
6-Hydroxymelatonin
CH2CH2NHCOCH3 6-Sulphatoxymelatonin
(59-79%)
6-Glucuronide Melatonin (13-29%)
39
1.2.10 Target sites and potential circadian effects of melatonin action
The SCN has been proposed as a site for the actions of melatonin to feedback
onto the pacemaker and regulate the circadian timing system (e. g. Cassone, 1991).
In mammals, high affinity melatonin binding sites have been characterised most
frequently in the SCN of the hypothalamus, the pars tuberalis of the pituitary gland
and the retina (Dubocovich, 1988; Carlson et al, 1989; Duncan et al, 1989; Morgan
et al, 1989; Weaver et al, 1989; Cassone, 1991). Human adult and foetal SCN tissue
are also reported to contain such putative melatonin receptors (Reppert et al, 1988;
Weaver et al, 1993). The high affinity melatonin binding site is associated with at
least one signal transduction mechanism involving regulatory G-proteins and therefore
provides a functional role for this binding site (Carlson et al, 1989; Morgan et al,
1989a, b; Rivkees et al, 1989; Laitinen and Saavedra, 1990; Stankov et al, 1992).
High affinity, specific melatonin binding sites comparable to those found in the
chicken retina have been proposed as ML-1 type receptors (Dubocovich, 1988).
There are lower affinity sites which are pharmacologically distinct from the ML-1
type receptor and Dubocovich (1988) has proposed that these sites may be referred
to as ML-2 sites. However, there is no evidence of separate physiological roles for
these sites as yet. A daily rhythm for high affinity melatonin receptors in the SCN
has also been described (Laitinen et al, 1989).
A possible feedback response of melatonin on the biological clock can be seen
in vitro. McArthur et al (1991) and later Mason et al (1993) demonstrated that
exogenous melatonin acts directly within the SCN to phase-shift the circadian
rhythmicity of neural activity. This feedback system on the SCN may allow
40
melatonin to modulate the phase relationship between the circadian system and the
environmental light/dark cycle which can alter the duration (and timing) of the
melatonin signal on a daily basis (see section 1.3). There are also efferent neural
connections from the SCN into other areas of the brain which provide a means by
which melatonin can indirectly influence behavioural and physiological functions
(Watts, 1991). The densest projections from the SCN lead into the adjacent
paraventricular nucleus (PVN) with a smaller projection to the lateral septal nucleus
(LSN, Watts et at, 1987). The PVN efferents connect the hippocampus, amygdala
and nucleus accumbens with the LSN projecting to the hypothalamus. These in turn
may influence the activity of the medial forebrain bundle which is crucial for many
aspects of behaviour such as arousal and feeding (Swanson, 1987).
Exogenous melatonin will entrain the free-running activity rhythms of rats
(Redman et at, 1983; Cassone et at, 1986). Suitably timed melatonin administration
to rats after a large shift in the light dark cycle can alter the rate and direction of
reentrainment of both the activity rhythms and N-acetyltranferase activity (Armstrong
and Redman, 1985; Redman and Armstrong, 1988; Illnerovä et at, 1989).
Strong evidence for melatonin's role in reproduction comes from seasonal
breeders. The duration of the melatonin secretory rhythm provides a direct reflection
of the length of night in photoperiodic animals (Arendt, 1986), and is able to transmit
this photoperiodic information to seasonal breeders to time reproductive behaviour
and function (Tamarkin et at, 1985). Infusion of physiological levels of melatonin
in animals timed to mimic the secretion pattern in different seasons is able to induce
the appropriate endocrinological and behavioural response (Bartness et at, 1993).
However, anatomical evidence demonstrating a relationship between the SCN and the
41
neuroendocrine parts of the hypothalamus is limited.
There do not seem to be any direct connections from the SCN to the median
eminence where gonadotrophin-releasing hormone is released to influence the
secretion of luteinizing hormone from the pituitary gland. However, since high
affinity melatonin binding sites have been found in the pituitary gland (Morgan and
Williams, 1989) and projections from some parts of the PVN lead into the preoptic
area of the hypothalamus where nuclei are closely associated with numerous
neuroendocrine functions, there is potential for melatonin to influence many aspects
of neuroendocrinology through indirect and direct mechanisms (see Watts, 1991 for
review).
Many structures within the preoptic region of the hypothalamus are intimately
involved in many aspects of circadian hypothalamic functions such as fluid balance,
sleep and thermoregulation (Swanson, 1987) and minor projections do lead into this
area from the SCN (Berk and Finklestein, 1981; Stephan et al, 1981). There are
however, major pathways which indirectly connect the two sites. For example, the
primary nucleus for body temperature regulation is the medial preoptic area of the
hypothalamus (Hellon, 1972; Hensel, 1973; Satinoff, 1978) which is directly
connected to an area just ventral to the PVN known as the subparaventricular zone
(sPVHz). This site in turn is directly supplied by dense projections from the SCN
(Watts et al, 1987). Thus morphological studies support the broad spectrum of SCN-
dependent effects.
42
1.2.11 Summary - potential physiological Roles for Melatonin
The presence of high affinity melatonin receptors within the SCN and the
ability of exogenous melatonin to alter the circadian rhythms of SCN neural activity
and rodent behaviour, suggest that pineal melatonin plays some role in the circadian
organisation in mammals.
Melatonin has a rapid and widespread distribution from the circulatory system
into all body tissues and fluids indicating that it may have a wide ranging modulatory
role over a number of different mechanisms within the body. Melatonin easily passes
through cellular membranes and into the nucleus (Joss, 1981; Menendez-Pelaez and
Reiter, 1993). The possibility that melatonin could modify the function of many
tissues and organs even at the nuclear level is therefore reasonable. Recent evidence
that melatonin is highly effective in neutralising the destructive activity of free-
radicals (oxidizing agents) suggests a role which does not require a membrane bound
receptor (Poeggeler et al, 1993; Reiter et al, 1993; Tan et al, 1993).
In addition, it has been shown that melatonin can modulate the response of the
immune system in some conditions (Blask et al, 1993; Maestroni, 1993). Circadian
rhythmicity occurs in number of immune mechanisms (Fernandes et al, 1976,1979;
Abo et at, 1981; Kawate et al, 1981) and to some extent tumour growth in
experimental animals can be influenced by photoperiod and melatonin administration
at different times of day (Bartsch and Bartsch, 1981; Stanberry et al, 1983). Thus
the pineal gland through its rhythmic circadian release of melatonin can influence the
immune system and possibly malignant disease (Blask, 1984).
43
1.3 FACTORS AFFECTING MELATONIN SECRETION
Melatonin is considered to be a reliable circadian rhythm marker and is used
as such in various physiological and pathological studies. Changes in melatonin
secretion may indicate a role for melatonin and the circadian timing system. It is
however, important to exclude disturbing (masking) influences unrelated to the object
of the study.
1.3.1 Light suppression of melatonin secretion in humans
Although the melatonin rhythm is remarkably stable in healthy humans, one
environmental factor known to greatly influence its production, via suppression, is the
light/dark cycle. Whether this suppressant effect is related to the phase-shifting or
entraining property of light is a matter of great interest.
1.3.1.1 Intensity of light exposure
Acute exposure to light at night shuts down the synthesis of melatonin rapidly.
The degree of suppression of nighttime melatonin depends upon the brightness of the
light. In humans, Lewy et al (1980) originally found suppression of melatonin
secretion in response to light of 1500 and 2500 lux. Light of 500 lux initially showed
no suppression although in a later study by Lewy et al (1985) some melatonin
suppression was found with 500 lux. Further studies have shown that the melatonin
synthesis is affected by lower intensities of light. Bojkowski et al (1987a) reported
44
partial but significant suppression at 300 lux (30 minutes at 0030h) and McIntyre et
al (1989a) indicated some suppression with 200 lux (lh at midnight). In all cases
melatonin levels returned to normal values within 60 minutes after cessation of
treatment. Waking, and maintaining subjects awake in <1 lux has no effect on
plasma melatonin levels (Lewy et al, 1980; Bojkowski et al, 1987a). The threshold
intensity of full spectrum light needed to suppress human melatonin levels has yet to
be determined but may reveal effective suppression at even lower levels. Using
monochromatic (green) light, Brainard et al (1988), have observed melatonin
suppression as low as 5.5 lux in some individuals. Pupil dilation may also be
important in determining the degree of illuminance falling on the retina and the level
of light-induced melatonin suppression (Gaddy et al, 1993). Some individuals with
manic-depressive illness have been found to be supersensitive to light and it has been
suggested that this may provide an appropriate marker for manic-depressive-illness
(Lewy et al, 1985).
1.3.1.2 Timing of Light Exposure
The response of the pineal gland to light, at least in rats, may be dependent
on the timing of the light treatment, where NAT activity does not recover to the same
extent after treatment in the latter half of the night as in the first half (Illnerovä and
Vanecek, 1984). In man, variations in the ability to detect low light intensities have
been found over the course of the night suggesting that light-induced suppression of
melatonin secretion may alter with time (Terman and Terman, 1985; Bassi and
Powers, 1986). This has been confirmed by Owen and Arendt (1992), who
45
demonstrated in humans living in Antarctica, that suppression by both dim light (m
300 lux) and bright light (-2200 lux) was more effective in the second half of the
night and was dependent upon the season (suppression less effective in summer).
However, in contrast, McIntyre et al (1989b) could find no time-dependent
differences in the ability of light to suppress melatonin production using 1000 and
3000 lux. In addition, they and others (Bojkowski et al, 1987a; Lewy et al, 1980)
found no rebound effect in melatonin levels after light exposure during the early dark
phase (around the melatonin secretion onset time) unlike Beck-Friis et al (1985) using
2500 lux treatment.
1.3.1.3 Duration of light exposure
A short duration (15 minutes) light pulse of 1500 lux at either 2200h or 2300h
or even a5 minute pulse of 2500 lux at 0200h can reduce nocturnal melatonin levels
(Byerley et al, 1989; Petterborg et al, 1991a). All-night exposure to bright light
(3000 lux) can completely abolish the nighttime melatonin profile (Strassman et al,
1987), suggesting that this model could be used for studies needing "functional
pinealectomy" in humans.
1.3.1.4 Wavelength of light exposure
The environment in which humans live consists almost exclusively of broad
spectrum light whether it is natural light or artificial fluorescent and incandescent
light. In mammals, including humans, the blue/green wavelength of the light
46
spectrum seems to be maximally effective at suppressing melatonin synthesis
(Brainard et al, 1984,1985). The visual system in humans is most sensitive to green
light (Riggs, 1965), however, there may be a change in sensitivity to lower
wavelengths of light during the course of the night (Honma et al, 1992).
Unfortunately, most of these studies differ in the timing, intensity and duration
of their light exposure. Further experiments are required to see whether humans have
the same inverse relationship between intensity and duration as has been found for rat
NAT suppression (Vanecek and Illnerovä, 1982). In rats, a short pulse of bright light
has a similar suppressant effect on NAT activity as dim light of longer duration
(Vanecek and Illnerovä, 1982). Previous lighting history may also prove to be
important in the response of melatonin suppression in terms of sensitivity to light.
Many of these studies were conducted under different latitudes with different
photoperiodic environments of varying natural illuminance. Lynch et al (1981)
demonstrated that rats on alternating artificial bright light/dim light cycles respond as
if dim light was natural darkness, and rats on alternating cycles of dim light and
darkness recognise dim light as natural light. In addition, laboratory-raised and wild-
captured ground squirrels, exposed to different intensities of light, responded
differently to light exposure at night (Reiter et al, 1983).
In summary, the suppressive effects of light on melatonin secretion are
complex depending on the intensity, timing, duration and wavelength of the light
treatment together with the photoperiodic history and sensitivity of the individual.
47
1.3.2 Non-visible radiation
Bright full spectrum light is not the only type of radiation known to affect
melatonin production. Non-visible radiation in the region of the electromagnetic
spectrum is reported to influence the production of melatonin (Reiter, 1992a). In
rats, low frequency (50-60 Hz) or static magnetic field exposure can suppress
melatonin levels at night (Olcese et al, 1988; Lerchl et al 1991). The pathway
through which the pineal gland perceives this magnetic information is suggested to
be via the retina in much the same way as light is recognised, although the exact
mechanism remains unclear and the function unknown (Olcese et al 1985,1990; Kato
et al, 1994).
1.3.3 Changing Daylength (Photoperiod)
The circadian pacemaker via melatonin secretion provides information to the
body about the length of day which changes with the time of year. This information
is used in many species to control seasonal rhythms in physiology and behaviour,
such as reproductive function and coat growth (Pittendrigh and Daan, 1976).
In all non-human mammalian species studied, the duration of nocturnal
melatonin secretion and NAT activity is extended under a shorter "winter" day
(longer night) and reduced under a longer "summer" day (shorter night), (Review:
Illnerovä, 1988). It is not known how the body interprets the melatonin signal
providing information about environmental time and signalling the appropriate
endocrine and behavioural adjustments to be made. It may be the timing at which the
48
light and/or darkness phases hit the melatonin rhythm or the timing between the
melatonin rhythm and a rhythm of sensitivity of the receiving organ. A combination
of the above are likely, depending on the species, together with the direction in which
the melatonin rhythm alters (Reiter, 1987; Illerovä, 1988).
In humans studies, the response to seasonal changes in natural photoperiod
have been variable. Some groups have shown a seasonal variation in melatonin
plasma concentrations (Arendt et al, 1977,1979; Wirz-Justice and Richter, 1979;
Touitou et al, 1984), although highest levels were not consistently associated with a
short photoperiod. Others have found no variation in total aMT6s production with
the time of year (Beck-Friis et al, 1984; Griffiths et al, 1986; Sack et al, 1986;
Bojkowski et al, 1988). However, Beck-Friis et al (1985) have found that in winter,
melatonin levels were raised later in the morning and subsequent studies have shown
that the melatonin secretion is phase-advanced by up to 2h in the winter compared to
summer (Illernovä et al, 1985, Arendt and Broadway, 1986; Griffiths et al, 1986;
Kennaway and Royles, 1986; Bojkowski and Arendt 1988). These studies suggest
that the changing season affects the timing of the melatonin rhythm more than the
amplitude.
More recently it has been demonstrated that humans respond both
physiologically and behaviourally to changes in artificial photoperiods similar to
animals (Buresovä et al, 1992; Wehr, 1991; Wehr et al, 1993). The nocturnal
duration of melatonin secretion along with sleep-phase, cortisol, prolactin and body
temperature lengthened on exposure to a short "winter" photoperiod (16h darkness:
8h light) from a longer "summer" one (14h light: 10h darkness). In the winter,
natural darkness falls in the early evening hours and, in the absence of artificial light,
49
the production of melatonin occurs earlier (phase-advancing). Darkness lifts in the
later hours of the morning in winter delaying the offset of melatonin secretion until
dawn appears. Wehr (1991) suggests that the mechanism for this phenomenon could
employ a two-oscillator system; one which controls the onset of melatonin secretion
and sleep, and one for the offset. The dual oscillator model has been previously
demonstrated by Pittendrigh and Daan (1976) using ̀ dawn' and ̀ dusk' oscillators and
later by Illernovä and Vanecek (1982) with their studies on rodents. Illernovä and
Vanecek (1982) simulated a dawn and dusk signal using brief 1 minute pulses of
bright light, giving a so-called ̀skeleton' photoperiod.
Further evidence for a dawn and dusk time cue in humans comes from
Broadway et al (1987) who simulated a 12.5h spring ̀ skeleton' photoperiod well into
the Antarctic winter, when men live entirely under artificial light (< 500 lux). They
applied two bright light pulses (of 1 hour duration at 2500 lux), one in the morning
and one in the evening, and produced a melatonin profile similar to the pattern in
summer. Dim light was not effective (< 500 lux). After the photoperiodic time cues
are removed, this response to a change in the light/dark cycle lasts for at least 24h
(Wehr et al, 1993), or longer if the sleep/wake cycle is fixed (Illnerovä et al, 1993),
suggesting that the photoperiodic response might be conserved within the human
circadian timing system.
These studies indicate that humans can respond to seasonal photoperiod
variations. However, the intensity of light appears to be important (Broadway et al,
1987) and may explain why a durational response in nocturnal melatonin was not
found in some studies since humans in industrialised societies live in artificial lighting
environments, whether in the home and at work, for most of the year and may not
50
receive bright enough light, or suitably timed photoperiodic signals, for a seasonal
response to be seen (Illnerovä et al, 1985).
1.3.4 Puberty
Puberty is associated with elevated activity of the hypothalamic-pituitary-
gonadal axis, bringing about sexual maturation. Evidence of a relationship between
melatonin and puberty has been conflicting, mainly due to methodological differences
such as day- and night-time sampling, limited number of subjects and single nighttime
samples (Review: Cavallo, 1993). In extensive studies with frequent nighttime
sampling in normal healthy children, plasma melatonin levels show little difference
between sexes and show a fall in the nighttime peak of melatonin secretion with age
and sexual maturation (Attanasio et al, 1985; Waldhauser et al, 1988; Cavallo, 1992).
In contrast, some reports show no change in total excretion of the urinary metabolite
of melatonin aMT6s from childhood to adult (Tetsuo et al, 1982). These results have
led to the suggestion that any increase in melatonin with age is due to increased body
mass (Waldhauser et al, 1988), see section 1.3.7. However, Brun et al (1992) have
found that urinary excretion of 6-sulphatoxymelatonin increases from early to mid
childhood taking into account body mass.
Pineal tumours in humans are associated with disturbances in pubertal
development (Kitay, 1954). However, other factors may be responsible for changes
in the hypothalamic-pituitary-gonadal axis such as tumour tissue spreading into the
hypothalamus and elevated secretion of chorionic gonadotropin (Wass et al, 1982;
Neuwelt and Lewy, 1983). The relationship between the pineal gland and puberty
51
has yet to be fully established.
1.3.5 Advancing age
In many species the nocturnal rise in melatonin secretion undergoes a gradual
decline with age (review: Reiter, 1992b). The fall in amplitude of the nighttime rise
in melatonin levels has also been shown with age in humans (Iguchi et al, 1982,
Touitou et al, 1981), together with a shorter nighttime profile (Nair et al, 1986).
aMT6s excretion follows a similar fall with advancing age (Sack et al, 1986;
Bojkowski et al, 1990), although Young et al (1988) found no fall in aMT6s
excretion with age. However, Young et al (1988) grouped subjects into two wide age
populations (36.5-79.9 years compared to 18-36.49 years) which allows only gross
comparisons to be made between two variable age groups. Significant changes may
only occur in later years perhaps only after 60 years as demonstrated by Bojkowski
et al (1990).
One theory for the reduction in melatonin secretion involved pineal
calcification which increases with age (Welsh, 1985). However, the activity of the
pineal enzymes involved in the synthesis of melatonin is not affected by calcium
deposits in the pineal gland (Wurtman et al, 1964) and more recent studies have
confirmed that pineal calcification does not affect melatonin production in young to
middle-aged adults (Commentz et al, 1986; Bojkowski and Arendt, 1990).
Another possible mechanism to account for the attenuation of the melatonin
rhythm with age is the concomitant decline of available B-adrenergic receptor numbers
on the pinealocyte membrane (Greenberg and Weiss, 1978; Henden et al, 1992) and
52
a, -adrenergic receptors in the SCN (Weiland and Wise, 1990). Food restriction in
rats can preserve the age-associated fall in these receptor numbers and nighttime
activity of NAT (Stokkan et al, 1991; Henden et al, 1992). A reduction in the
neuronal activity of the SCN with age may also contribute to a dampening of the
melatonin rhythm (Satinoff et al, 1993).
Administration of exogenous nighttime melatonin or engraftment of pineal
glands from young to old mice has been found to promote longevity in rats (Pierpaola
and Regelson, 1994). In addition to changes in melatonin output with age, a
concomitant fall in the endogenous temperature rhythm amplitude is observed
(Czeisler et al, 1992), which can be prevented with the melatonin agonist S-20242 in
rats (Koster-van Hoffen et al, 1993). These studies suggest that melatonin and the
pineal gland may act as an internal clock mechanism involved in the aging process.
The alterations in the circadian system with age may affect the individuals'
ability to be entrained in a normal environment or adapt to changes in a new time
zone. Recently sleep disorders, which are common in the elderly (Feinberg, 1974)
have been correlated with disturbances in the melatonin rhythm (Haimov et al, 1994).
The elderly account for the majority of the population who are diagnosed with
advance sleep phase syndrome (Moldofsky et al, 1986).
In aging hamsters the circadian clock either becomes less responsive or
changes its response to the phase-shifting and synchronising effects of
benzodiazepines and light/darkness (Zee et al, 1992; Van Reeth et al, 1993). In
humans, age-related differences in adaptation to shiftwork or a new time zone have
been attributed to a reduced ability of the internal clock to adjust (Akerstedt and
Torsvall, 1981; Graeber, 1982).
53
1.3.6 Normal menstrual cycle
Wetterberg et al (1976) have indicated that melatonin levels showed apparent
cycling in line with the menstrual cycle, peaking at the time of menstrual bleeding
and reaching a minimum at ovulation, and suggested a role for melatonin in normal
menstrual cyclicity. Correlational evidence shows that this menstrual melatonin
profile is inversely related to the changes in temperature throughout the cycle
(Nakayama et al, 1992). Amplitude changes and phase-shifts in the luteal phase have
also been reported by others (Hariharasubramanian and Nair, 1982; Webley and
Leidenberger, 1986). However, Brzezinski et al (1988) and Fellenberg et al (1982)
have failed to confirm these findings. Melatonin has been shown to be present in
human preovulatory follicles (Brzezinski et al, 1987) and nocturnal melatonin levels
can stimulate the production of progesterone in human granulosa cells (Webley and
Leidenberger, 1986). The circadian role of melatonin in human ovarian function
remains to be elucidated although there is evidence for melatonin's role in the
reproductive function of many mammals (review: Reiter, 1980).
1.3.7 Body size
There is contradictory evidence for a relationship between melatonin levels and
body size. As body weight increases in both men and women, plasma melatonin
levels have been shown to increase (Arendt et al, 1982; Ferrier et al, 1982).
However, no correlation has been found for aMT6s (Nair et al, 1986; Sack et al,
1986; Bojkowski and Arendt, 1990).
54
Height may also influence melatonin levels. Beck-Friis et al (1984) found a
negative correlation between melatonin and height. However, a similar finding by
Sack et al (1986) was no longer significant when age was taken into account and
other studies could not confirm the relationship (Nair et al, 1986; Bojkowski and
Arendt, 1990).
Differences in melatonin levels have been observed in the 2 sexes and obesity
by some workers. Sack et al (1986) found low aMT6s levels in urine with increasing
weight in women and the opposite for men. Beck-Friis et al (1984) found a negative
association between obesity and melatonin levels in a mixed population and Arendt
et al (1982) demonstrated a similar negative correlation in women. However, in
obese children, Tamarkin et al (1982) have found normal plasma melatonin levels.
The relationship between melatonin levels and measurements of body size remains
unclear, but huge individual differences in melatonin secretion (Arendt, 1988) may
disguise a true effect.
These studies do suggest that consideration of weight, height and obesity index
is warranted when measuring melatonin and emphasises the need to use individuals
as their own controls.
1.3.8 Exercise/stress
Stressful stimuli can stimulate the sympathetic nervous system by increasing
circulating levels of NA (Axelrod and Reisine, 1984). It is therefore conceivable that
stress could increase the production of melatonin through increased sympathetic
innervation of the pineal or through direct stimulation of pinealocytes by systemic
55
NA.
The effects of exercise on melatonin concentrations in humans is conflicting.
At night a reduction in melatonin production has been observed after physical exercise
in the dark (Monteleone et at, 1990). However, physical exercise during light-
induced suppression of melatonin causes no effect (Monteleone et at, 1992). During
the day Vaughan et at (1979) demonstrated no changes in melatonin levels after a 100
yard dash. However, more strenuous exercise has been shown to transiently elevate
melatonin concentrations (Carr et at, 1981; Ronkainen et at, 1986). Similar findings
have been demonstrated in rats whilst swimming at night (Troiani et at, 1987).
Stress-related hormones and catecholamines are not involved in this response, since
removing the adrenal glands and severing sympathetic innervation has no effect on
this elevation in the rat blood (Wu et at, 1987; Troiani et at, 1988). The active NA
uptake mechanism of sympathetic nerve endings in the pineal may also provide a
protective response against acute stress-induced stimulation of the pineal (Parfitt and
Klein, 1976). No corresponding changes in rat NAT or HIOMT activity are evident
with the increase in blood melatonin (Troiani et at, 1987). The mechanism of action
therefore remains unknown.
1.3.9 Alcohol ingestion
Alcohol at doses usually ingested socially (0.05-1g/kg wt) can suppress
melatonin secretion when administered before the melatonin onset time (Ekman et al,
1993; Röjdmark et al, 1993). The mechanism of action remains unknown but may
involve inhibition of adrenergic activation of the pinealocytes (Chik et al, 1987,
56
1992).
In summary, whilst the melatonin circadian rhythm is remarkably stable within
individuals, there are a number of factors which may disturb the endogenous secretion
pattern of melatonin leading to inappropriate estimates of circadian parameters.
Hence, in any studies involving melatonin measurement as a marker of circadian
time, alcohol and exercise should be restricted and subjects/patients should be drug-
free wherever possible and matched by age, weight, height and reproductive status.
1.3.10 Depression and seasonal affective disorder
Disturbances in the circadian timing system have been associated with
depression particularly when they are linked to periodic manic episodes. A phase-
advance theory has been proposed for depressed patients where sleep, temperature
and cortisol rhythms have been shown to be phase-advanced compared to controls
(review: Kripke, 1983). The amplitude of nocturnal melatonin secretion may be
reduced although the evidence is conflicting (e. g. Wetterberg et al, 1978; Claustrat
et al, 1984; Thompson et al, 1988).
A number of antidepressant drugs which affect noradrenergic and serotonergic
function have been shown to stimulate melatonin production in humans. For
example, desipramine can advance the onset and increase the duration of melatonin
in humans (Franey et al, 1986; Cowen et al, 1985; Skene et al, 1994), presumably
by inhibiting the reuptake of NA at the nerve synapses in the pineal gland (Checkley
et al, 1986). 5-HT uptake inhibitors such as fluvoxamine and melatonin synthesis
57
precursors can elevate melatonin levels (Namboodiri et al, 1983; Demisch et al,
1986; Skene et al, 1994). In addition, the phosphodiesterase inhibitor rolipram, has
been shown to increase melatonin (Checkley et al, 1987). All have antidepressant
activity. These studies underline the importance of noradrenergic, serotonergic and
cAMP mechanisms in melatonin production but there is little evidence so far to
suggest a causal effect of melatonin in depression.
One depression that reappears every autumn and winter, is known as seasonal
affective disorder (SAD; Rosenthal et al, 1984), and has been linked to a seasonal
disturbance in the circadian timing system. A delay in the rhythm of melatonin (and
aMT6s), cortisol and temperature has been observed in SAD patients in various
studies (Lewy et al, 1987; Terman et al, 1988; Avery et al, 1989; Wirz-Justice et al,
1990; Wirz-Justice et al, 1993), although this has not been confirmed by others
(Rosenthal et al, 1990; Checkley et al, 1993; Eastman et al, 1993).
1.3.11 Circadian rhythm disorders
There are a number of other situations or pathological states where the timing
of the endogenous melatonin rhythm is out of phase with environmental zeitgebers
and apparently with other circadian rhythms. These include jet lag (Arendt et al,
1987), maladaptation to shift work (Waldhauser et al, 1986; Touitou et al, 1990;
Sack et al, 1992; Shanahan et al, 1994) and blindness (Lewy and Newsome, 1983)
and are described in detail in sections 1.5,1.6 and 1.8.
58
1.4 PHASE-SHIFTING RESPONSE OF THE HUMAN CIRCADIAN
SYSTEM TO BRIGHT LIGHT AND MELATONIN
1.4.1 Range of entrainment
There is a limited `length of the day' to which humans can adapt without
causing circadian rhythms to become misaligned from the environment ("external
desynchronization"), or from one another ("internal desynchronization"), and continue
oscillating with their own free-running periods (Wever, 1983). Human physiological
rhythms can normally entrain to daylengths between 23 and 27h (Wever, 1983,
1986). Outside this range of entrainment rhythms such as temperature, cortisol and
melatonin (Lewy and Newsome, 1983) will naturally free-run with a periodicity close
to 25h (Wever, 1974,1986,1989; Kennaway and Van Dorp, 1991). This free-
running periodicity is considered to reflect the internal period of the circadian
pacemaker. However, the range of entrainment also depends upon the strength of the
zeitgeber (Wever, 1983). Bright light of 3000 lux during the light phase (0.1 lux
during darkness) has been found to increase the range of entrainment of the
temperature rhythm (Wever, 1983,1986,1989) by as much as 6h a day from its free-
running period, i. e. 19-31h. Shorter durations of bright light (2h at 2000-4000 lux)
exposure administered just before bedtime can also facilitate the entrainment of the
circadian system to a 26h day (Eastman and Miescke, 1990).
59
1.4.2 Entrainment of the human circadian system to the light/dark cycle
Originally, the light/dark cycle was not thought to be as effective as social
time cues in synchronising the human circadian system (Wever, 1979). However, in
these early studies light was freely available during the dark phase and social cues
(audible gongs) signalled light/dark transitions. With stricter control of the light/dark
cycle and social cues, the light/dark cycle was able to entrain the temperature and
activity rhythms (Czeisler et al, 1981). Many blind subjects do not receive zeitgebers
of adequate strength to entrain their circadian system indicating that the light/dark
cycle is an important zeitgeber in man (Orth and Island, 1969; Miles et al, 1977;
Sack et al, 1987; Arendt et al, 1988; Klein et al, 1993). Brighter intensity light
(> 3000 lux) throughout the light phase of the light/dark cycle effectively entrains the
temperature rhythm to a 24h day (Wever et al, 1983; Honma et al, 1987).
1.4.3 Phase-shifting effects of light
The ability of circadian timing system to entrain to environmental time cues
relates to the ability of external zeitgebers to shift the circadian pacemaker in time.
Suitably timed bright light of sufficient intensity and duration will induce phase-shifts
in the rhythms of various physiological variables such as core body temperature,
cortisol and melatonin secretions (Broadway et al, 1987; Shanahan and Czeisler,
1991). Repeated daily exposure to bright light (2500 lux) in the morning can phase-
advance the onset of melatonin secretion, whereas evening light treatment can induce
phase-delays, even when the sleep/wake cycle is held constant (Lewy et al, 1987).
60
Evening exposure to much stronger intensity bright light (> 7000 lux) for one week,
can phase-delay the endogenous temperature and cortisol rhythms by as much as 6h
in some individuals, whilst maintaining a constant sleep/wake cycle and regular social
interactions (Czeisler et al, 1986). In addition, Czeisler et al (1989) demonstrated
that bright light pulses (> 7000 lux) given on three consecutive days can shift the
temperature, cortisol and urine output rhythms, advancing them with morning light
exposure or delaying them with evening light treatment.
Bright light can also acutely phase-shift the circadian pacemaker after a single
exposure. Buresovä et al (1991) reported an a 0.6-2.2h advance in the melatonin
rhythm one day after exposure to bright light (3000 lux) between 0300-0900h. A
single 3h bright light pulse (5000 lux) administered up to 5h before the temperature
minimum phase-delayed the melatonin rhythm by an average of 74 minutes (van
Cauter et al, 1994). Average phase-advances of 28 minutes were evident between 1-
4h after the temperature minimum.
In constant routine ('unmasked') conditions, the magnitude and direction of
shift of the plasma melatonin, plasma thyrotropin, core body temperature, plasma
cortisol and urine production rhythms are similar (Czeisler et al, 1989; Shanahan and
Czeisler, 1991; van Cauter et al, 1994). Taken together these studies strongly
suggest that light has an unique ability to phase-shift the master human circadian
pacemaker governing a number of physiological circadian rhythms.
These findings have led to the establishment of a phase response curve (PRC)
for bright light which predicts the magnitude and direction of shift of the circadian
pacemaker, in response to light pulses at various times of day (Minors et al, 1991;
van Cauter et al, 1994), Figure 1.5. A classical PRC using single pulses of bright
61
light has been described by Minors et al (1991) using the core body temperature
rhythm as a circadian phase marker of the biological clock. Bright light administered
before the temperature minimum (0400-0500h clock time) caused a phase-delay and
light given after the temperature minimum phase-advanced the rhythm. Phase-
responses were obtained under 3 different conditions where 1) subjects were initially
entrained to a 24h day before being exposed to a 3h pulse of 5000 lux bright light,
2) subjects were free running before exposure to 3h of bright light at 9000 lux and
3) subjects were free-running before exposure to 3h of 9000 lux on three consecutive
days. Larger phase-shifts were obtained with subjects free-running and even more
so with repeated daily light exposure. A similar phase response effect of bright light
(5000 lux) has been demonstrated by van Cauter et al (1994) using plasma melatonin
and thyrotropin as circadian rhythm markers, and Honma and Honma (1988) using
the free-running sleep/wake rhythm as a circadian phase marker.
Czeisler et al (1989) have generated a light phase response curve which can
induce much larger phase-shifts in temporal isolation and under constant routine
conditions. This is based on repeated daily exposure of bright light in combination
with darkness. Using 5h of bright light (7000-12000 lux) on 3 consecutive days and
maintaining regular 8h darkness (sleep) episodes, phase-shifts of up to 12 hours in the
temperature rhythm were achieved after 3 days. The greatest shifts were obtained
when light pulses were centred over the temperature minimum with subjects in
darkness (and sleep) during the day. The time of the temperature minimum is
considered to be the most sensitive phase of the circadian system where the largest
phase-shifts can be induced (Czeisler et al, 1989). These results reflect a combination
of imposed light/dark and sleep/wake cycles. in achieving the shift.
62
Figure 1.5 Human phase-response curves for bright light and melatonin
y V
pU ed ý
-12
Time relative to
-8 -4 0 +4 +8 +12 temperature minimum (h)
1600 2000 2400 0400 0800 1200 1600 Clock time (h)
Phase-response curves (PRC) for bright light and melatonin showing the magnitude and direction of phase-shift in the endogenous biological clock after bright light or melatonin treatment at different times of the day. Two light PRCs for single 3h bright light pulses are shown and are plotted using the midpoint of light exposure. One used the core body temperature rhythm as the circadian phase marker (, light intensity = 5000/9000 lux, ----: no estimate made). The second light PRC
used the endogenous melatonin rhythm as a phase marker (- , light intensity =
5000 lux). Two melatonin PRCs are shown and both have used the endogenous melatonin rhythm as the circadian phase marker. One shows a cumulative phase- shift reponse after repeated melatonin treatment (0.5mg) for 4 days (-). The
second shows the response to single 3h melatonin infusions at physiological doses
only at four different times of day (A), and is plotted using the start of each infusion. Adapted from Minors et al, 1991, Lewy et al, 1992, van Cauter et al, 1994 and Zaidan et al, 1994.
63
Response curves allow the prediction of the effects of light treatment at
different times of day and therefore could be used therapeutically in conditions where
there are underlying circadian rhythm disturbances, e. g. jet lag. However, it must
be noted that the generation of such phase response curves in a free-running state or
in temporal isolation may not apply to a "normal" environment where the circadian
system will be influenced by daily zeitgebers which alter rhythm parameters. For
example, natural ambient bright light exposure can prevent the desired phase-shift
with artificial bright light if appropriately timed to oppose the desired shift (Eastman
and Miescke, 1990). Control of ambient light intensity in the normal environment
is therefore important when treating circadian rhythm disturbance.
1.4.4 Phase-shifting effects of melatonin
Under normal conditions, repeated melatonin administration (2mg and 8mg)
at 1700h or 2200h will phase-advance its own endogenous rhythm (Arendt et al,
1985a; Wright et al, 1986; Mallo et al, 1988). The cumulative phase-shifting effect
of repeated melatonin administration has been extended to cover responses at several
times of day and has led to the establishment of a PRC for melatonin (Lewy et al,
1992) with almost a reverse phase relationship to the light PRC (Minors et at, 1991;
van Cauter et al, 1994), Figure 1.5. Using the onset of melatonin secretion in plasma
as a circadian phase marker, the PRC was determined with reference to the
individuals melatonin onset time in plasma, which for normal sighted individuals
occurs at 2100-2200h clocktime (Arendt et al, 1977; Lewy et al, 1992; Deacon and
Arendt, 1994). Using 2200h as a guide, phase-advances were apparent between
64
1200-2000h and phase-delays were evident between 0400-1000h. A zone of little or
no response occurred between 2000-0200h, Figure 1.5.
Recently, the acute phase-shifting effects of single melatonin infusions have
been described in terms of a phase-response curve (Zaidan et al, 1994). Intravenous
infusion of physiological levels of melatonin starting at 1600h and 2000h advanced
the endogenous melatonin rhythm and when administered between 0400-0700h and
1200-1500h the melatonin profile was delayed, Figure 1.5. In contrast to the PRC
of Lewy et al, Zaidan and workers reported a larger phase-shift later in the day.
Such work suggests that melatonin has phase-shifting properties of its own and
provides a means of resetting the biological clock similar to the ability of bright light,
where disturbances in biological rhythms exist, for example abnormal sleep phase
syndromes and shiftwork.
65
1.5 ABRUPT PHASE-SHIFTS IN LOCAL TIME CUES
Two aspects of modern society where the local time cues are quickly altered,
time-zone travel and shift work, result in behavioural and physiological problems
which can interfere with the individuals ability to perform everyday tasks and make
them more susceptible to having accidents, compromising productivity and safety.
1.5.1 Jet lag
Tiredness during the day, difficulty in sleeping at night, poor performance and
alertness together with digestive problems are common complaints after rapid
transmeridian flight over many time zones (Siegel et al 1969; Winget et al, 1984;
Comperatore and Krueger, 1990). Travel weariness and sleep loss are associated
with time-zone travel, however, the symptoms of jet lag tend to persist longer than
the time taken to recover from these effects. Underlying circadian rhythm
disturbances are also associated with the jet lag state (D6sir et al, 1981; F6vre-
Montange et al, 1981). On arrival at their new destination, jet travellers are exposed
to a range of social and environmental zeitgebers and imposed routines that are
shifted in time, although their body clocks are not yet adapted. The jet lag state tends
to persist whilst the circadian system slowly adapts to new local time cues at a rate
of 1-1.5h a day (Aschoff et at, 1975; Klein and Wegmann, 1980). The length of
time for different circadian rhythms to readjust to a new time zone or schedule is
variable resulting in a temporary internal dissociation between circadian rhythms
(Elliot et al, 1972; Aschoff, 1978; Klein and Wegmann, 1980; D6sir et at, 1981).
66
Even rhythm parameters (e. g. minimum and maximum) may adapt differently,
suggesting that different control mechanisms regulate different aspects of rhythmic
functions (Ddsir et al, 1981). However, masking of endogenous circadian rhythms
by external factors such as activity and sleep may influence the apparent rate of
adjustment and should be considered (Folkard et al, 1991; Minors and Waterhouse,
1992a, 1992b).
The speed of resynchronization of circadian rhythms depends upon the number
and direction of time zones crossed together with the strength of the zeitgeber
(Aschoff et al, 1975; Klein and Wegmann, 1980). Resynchronization takes longer
after eastward rather than westward flights. To adjust the circadian rhythms to an
eastward time zone, the circadian timing system needs to be advanced, i. e. moved
forward in time. In contrast, adjustment to a westward time zone requires the
circadian timing system to be delayed, i. e. moved backwards in time. This is
achieved by either temporarily shortening or lengthening the period/cycle of the
internal clock. For example, in normal everyday life the period of the endogenous
body clock m 24h. Adopting a longer period, say 25h, gradually delays the timing
of the circadian system by lh a day until the desired shift is achieved. Then
environmental and behavioural time cues will reinforce and entrain the biological
clock to the new 24h cycle.
67
1.5.2 Shift work
The major industrial disasters Chernobyl and Bhopal occurred during the night
shift when sleepiness related accidents are common (Folkard et al, 1978; Akerstedt,
1988; Mitler et al, 1988). Often the adaptation of circadian rhythms to the newly
imposed work schedules is never complete and individuals sleep and work at
abnormal circadian times even after many years of night shift work (Colquhoun et al,
1969; Folkard et al, 1978; Akerstedt and Gillberg, 1981; Czeisler et al, 1990; Sack
et al, 1992). Although night shift workers are deprived of sleep at night they may
still suffer from disturbed sleep during the day (Akerstedt and Gillberg, 1981;
Czeisler et al, 1982). Continual adjustment of the circadian system to new work
schedules is associated with health problems that include sleep disorders, digestive
problems and a higher risk of cardiovascular disease (Czeisler et al, 1983; Akerstedt
et al, 1984).
Adaptation of the circadian system to newly imposed changes in the
sleep/wake cycle may be hindered, firstly by individuals not maintaining the same
sleep/wake schedule on days off and secondly, insufficient time to adapt during
rotating shift schedules. Shift workers are also often exposed to other zeitgebers such
as the natural light/dark cycle and social contacts which prevent adaptation to new
schedules (Eastman, 1987). This problem is often compounded by the fact that night
shift workers are typically exposed to light intensities at work which are much less
than natural outdoor light, and hence the zeitgeber strength is attenuated.
68
1.5.3 Susceptibility to jet lag and maladaptation to shift work
Susceptibility to jet lag and problems adjusting to shift work varies with
different individuals.
1.5.3.1 Age
One of the factors affecting the ability to adapt to new time zones and shift
work is age. A reduced ability to adapt has been associated with increased difficulty
in sleeping at inappropriate circadian phases (Graeber, 1982; Moline et al, 1992).
In a simulated jet lag paradigm in which the subjects' day/night routine was advanced
rapidly by 6h, Moline et al (1992) found that middle-aged men showed no difference
in the rate of readjustment of the core body temperature rhythm compared to younger
controls but experienced more disrupted sleep at night and suffered increased
sleepiness and fatigue during the day. This study indicates that the disruption of the
pattern or timing mechanism of sleep is more likely to occur after abrupt phase-shifts
in middle-aged men than in younger individuals. More fatigue, greater disruption of
the sleep pattern and slower adaptation of the temperature rhythm has also been
observed in middle-aged men compared younger controls following transmeridian
flight (Hauty and Adams, 1965; Evans, 1970; Evans et al, 1972).
1.5.3.2 Circadian rhythm characteristics
The stability and amplitude of circadian rhythms may also influence the rate
69
of adaptation in individuals (Reinberg et al, 1978; Klein and Wegmann, 1980;
Graeber, 1982; Winget et al, 1984; Comperatore and Krueger, 1990). Some workers
suggest that individuals with larger circadian rhythm amplitudes (e. g. body
temperature) find it harder to adapt to phase-shifts (Reinberg et al, 1978). The
stability of circadian rhythms is assessed by the standard deviation (SD) over 12
consecutive days (Colquhoun and Folkard, 1978). A low SD indicates a more stable
rhythm and is associated with a lower rate of shift. The circadian phase of the
individual may also predict the rates of adaptation to abrupt phase-shifts. For
example, after an 8h eastward (advance) time-zone shift, Colquhoun (1979) found that
individuals with earlier temperature maximums entrained faster.
1.5.3.3 Personality
The circadian phase may also be related to certain personality types. A
number of studies suggest that evening-type individuals adapt to shift work easier than
morning-types (e. g. Folkard and Monk, 1981; Hildebrandt and Stratmann, 1979).
Morning-type people are more alert in the morning and show an earlier peak time of
the body temperature rhythm (Home and Östberg, 1976). Thus when the new time
zone or shift schedule requires a delay of the circadian system, it is logical that the
earlier the circadian phase of the individual the longer it will take for them to adapt.
This is confirmed by field studies showing that the speed of circadian adjustment to
night work of morning-type people is reduced on delayed shifts (Costa et al, 1989).
70
1.6 BRIGHT LIGHT AND MELATONIN TREATMENT FOR JET LAG
AND SHIN WORK
The circadian system cannot adapt fast enough to abrupt changes in local time
cues because it has a limited range of entrainment. However, by increasing the
strength of the zeitgeber the circadian system may resychronise itself with the external
environment faster.
1.6.1 Bright light treatment for jet lag
The rate of adaptation may depend upon the time-zone travellers activities (and
environment) after the flight. In a study by Klein and Wegmann (1974), the rate of
reentrainment of performance rhythms after a 6h phase-shift were much slower in one
group of subjects kept indoors compared to another group who enjoyed the freedom
of outdoor activities. Differences in social zeitgebers were considered to be
responsible for this effect. However, exposure to brighter intensity light outdoors
may also have contributed to the faster adaptation through mechanisms of entrainment
and phase-shifting (Czeisler et al, 1981,1986,1989; Wever et al, 1983; Honma and
Honma, 1988; Minors et al, 1991; van Cauter et al, 1994). These effects of light
have led to its use in facilitating the readaptation of circadian rhythms to the new
local time and alleviating the associated behavioural problems.
In 1984, Daan and Lewy suggested a light treatment strategy for jet lag using
exposure to natural daylight, based on animal light phase response curve theory (see
section 1.4.3). After two subjects travelled east over 9 time zones (9h phase-
71
advance), they exposed themselves to natural daylight for one week either between
0700-1000h local time, intended to phase-delay, or between 1000-1300h local time,
intended to phase-advance. The temperature rhythm of the latter reentrained faster
by one week compared to the former. However, this observation conflicts with
human light phase response curve theory (Czeisler et al, 1989; Honma and Honma,
1988; Minors et al, 1991; van Cauter et al, 1994) since light exposure between 1000-
1300h should have delayed the temperature rhythm, although individual circadian
phase, differences in daylight intensity and other competing zeitgebers may account
for the discrepancies.
The actual intensity of light can alter the rate of reentrainment after an abrupt
phase-shift. In a laboratory setting, Wever (1985a) observed that the rate of body
temperature adaptation after a phase-delay of the light dark cycle was faster by
0.4h/day with indoor light intensities of 2000-5000 lux compared to < 1500 lux, over
the first few days post-shift, suggesting that the strength of the zeitgeber is important.
There is also evidence for delayed effects of bright light. In one study, no
overall differences were observed in the rate of body temperature adaptation
comparing light treatments of 2500 lux and 300-500 lux administered for 4h on
waking after a 6h phase-advance (Moline et al, 1989). Both light groups advanced
at the same rate throughout the treatment period of 4 days, but after the fifth post-
shift day the bright light group delayed for 2 days. No differences in objective sleep
measurements were identified. In a similar phase-shift paradigm, Samel et al (1992)
treated individuals with bright light exposure (3500 lux) at either 0400h or 1300h
(mid-points of light exposure) for two days. Only on the seventh day after the phase-
shift did the morning bright light group show an earlier peak time in the temperature
72
rhythm. Another study has shown improvements in polysomnographic sleep
patterns with morning exposure of bright light (>3000 lux) compared to dim light
(< 500 lux) for 3 days after an 8h advance time zone shift (Sasaki et al, 1989). Cole
and Kripke (1989) analysed subjective measurements of sleep after time zone advance
shifts of 6.5-10h, followed by light treatment upon wakening for 3 days consisting of
bright white light (2000 lux) or dim red light (< 100 lux). No group mean
differences between the two light treatments were apparent although on at least one
of the light treatment days early morning bright light treatment appeared to
consolidate sleep into a single period at night.
To date the effects of bright light treatment in alleviating jet lag are
inconclusive despite successful treatment for maladaptation to shift work.
1.6.2 Bright fight treatment for shift work
In laboratory simulations of night shift work, normal adaptation can be
achieved within a week (Rutenfranz et al, 1981). However, this contrasts markedly
to field based studies, underlining the importance of the strength and timing of natural
environmental zeitgebers, in particular the natural light/dark cycle (Akerstedt et al,
1977; Folkard et al, 1978; Reinberg et al, 1984; Monk, 1986).
Bright light (<2800 lux) just prior to sleep and light control the following
morning can facilitate the adaptation to a 26h day (Eastman and Miescke, 1990). By
maintaining this schedule over a number of days, it provides a method for adjusting
the circadian system gradually during slowly-rotating shift schedules (i. e. a week or
more between different shifts, Eastman, 1987).
73
By increasing the strength of the light zeitgeber and adjusting the timing of
exposure, a few studies have achieved the desired rapid adjustment, without treatment
prior to the new shift. Controlled exposure to bright light (> 7000 lux for 4 nights)
during the whole night, and darkness during the day, can induce complete adaptation
to simulated night shift work, despite the morning commute home in natural light
(Czeisler et al, 1990). Shorter durations of bright light exposure (> 6000 lux, 2400-
0400h) during the night shift are also sufficient to significantly improve daytime sleep
and nocturnal performance (Dawson and Campbell, 1991; Bougrine et al, 1993).
In addition to an effect on the circadian system, bright light may have direct
arousing or stimulating effects which can benefit mood and performance during the
night shift, see section 1.10.2.
Most of the studies described above have involved young people (< 30 years).
Although bright light therapy may be effective in facilitating adaptation to shiftwork
in these people, older individuals have been found to be less responsive to the same
light treatments (Campbell et al, 1994). This underlines the importance of adapting
bright light treatment to suit particular individuals.
1.6.3 Melatonin treatment for jet lag
Exogenous melatonin (5mg) has been shown to alleviate jet lag symptoms,
facilitate the reentrainment of the endogenous melatonin and cortisol rhythms and
resynchronize the melatonin excretion rhythm after simulated or actual time zone
shifts (Arendt et al, 1987; Samel et al, 1991). Other studies have shown melatonin
to successfully alleviate subjective feelings of jet lag in field studies (Arendt and
74
Aidhous, 1988; Nickelsen et al, 1991b; Claustrat et al, 1992; Petrie et al, 1989,
1993).
In all these studies, melatonin was administered near bedtime after the flight.
In a few cases, pre-flight treatment of melatonin was used to initiate the phase-shift
in the circadian system in order to accelerate the resynchronization process (Arendt
et al, 1987; Petrie et al, 1989; Samel et al, 1991). Melatonin was administered at
a time when it would be naturally released in the new time zone. In a recent study,
Petrie et al (1993) compared the efficacy of melatonin in alleviating jet lag in aircrew
with and without pre-flight treatment for 2 days. Post-flight treatment alone
demonstrated faster alleviation of jet lag symptoms, suggesting that pre-flight
treatment was not necessary. Pre-flight melatonin treatment was timed assuming that
every individual was entrained to local time before the flight. However,
administering melatonin treatment to aircrews at the correct circadian time pre-flight
for every individual is complicated by their frequent time-zone travel.
1.6.4 Melatonin treatment for shift work
During rotating shift work, Armstrong et al (1986) showed that melatonin
administration (5mg) favoured the reentrainment of the body temperature rhythm to
the new schedules. During night shift work, melatonin administered at the desired
bedtime improved the quality and duration of sleep (Folkard et al, 1993). One
performance parameter showed deleterious effects but others were unchanged or
slightly improved (Folkard et al, 1993).
75
1.7 Bright light and melatonin treatment for SAD
In early studies, the effectiveness of morning bright light exposure as opposed
to evening therapy in treating this disorder was thought to be a result of the circadian
phase-shifting effects of light (Lewy et al, 1987). The phase-shift theory suggests
that circadian rhythms are delayed with respect to the external clock time and
subjects' sleep times. Thus advancing rhythms with morning light treatment, resets
the circadian system. However, the circadian timing of light therapy is no longer
thought to be critical since improvements are now seen with light treatment no matter
what time of day treatment is given (James et al, 1985; Wehr et al, 1986; Jacobsen
et al, 1987; Wirz-Justice and Anderson, 1990; Wirz-Justice et al, 1993), and
improvements in SAD patients have been noted whether they were initially phase-
delayed or not (Wirz-Justice et al, 1993).
Another theory to explain the antidepressant effects of phototherapy is the
melatonin hypothesis. Here the reduction of the duration of melatonin secretion in
SAD patients by simulating a summer photoperiod, was thought to alleviate the
depression (Rosenthal et al, 1984). However, since light treatment in the middle of
the day improves the symptoms without affecting the melatonin rhythm, other factors
are thought to be involved (Wehr et al, 1986). Use of the B-adrenergic antagonist
atenolol, which causes suppression of melatonin secretion, shows no remission in the
disorder, and therefore disproves the hypothesis (e. g. Rosenthal et al, 1988).
Light therapy remains the "choice" treatment for SAD, but these
biological/psychological effects of light do not appear necessarily to involve the
melatonin rhythm and the circadian timing system.
76
1.8 EFFECTS OF EXOGENOUS MELATONIN ON OTHER SLEEP/WAKE
DISORDERS
1.8.1 Blindness
Many blind people keep regular schedules and report few sleep problems.
However, others suffer from periodic insomnia and daytime fatigue which relates to
the inappropriate phase position of endogenous rhythms as a function the desired sleep
times (Miles et al, 1977; Lewy and Newsome, 1983; Klein et al, 1993).
Exogenous melatonin (0.5-5mg) administered close to the desired bedtime can
synchronise the sleep/wake cycle in blind subjects to a 24h day and improve the
quality of life (Arendt et al, 1988; Folkard et al, 1990; Sack et al, 1991; Jan et al,
1994). Similar findings were described after administering 0.5mg earlier in the
evening at 1800h (Palm et al, 1991). However, the timing of administration seems
to be more important for some blind subjects. Tzischinsky et al (1992b) reported
successful treatment of sleep disturbances with melatonin administration at 2000h but
not between 2200-2300h.
1.8.2 Sleep phase syndrome disorders
Difficulties in initiating sleep at night and problems with waking in the
mornings are characteristic features of a type of insomnia known as delayed sleep
phase syndrome. Despite their "sleep onset insomnia" patients generally have no
difficulty in maintaining sleep once initiated with the onset and offset of sleep
77
occurring nearly at the same clock times each day (Weitzman et al, 1981).
Individuals often have difficulties in maintaining a normal working life since morning
starts will inevitably curtail sleep and result in daytime fatigue (Alvarez et al, 1992).
There is therefore an inappropriate relationship between the circadian sleep/wake
rhythm and environmental time cues.
Entrainment to the environment can be facilitated with exogenous melatonin.
Late evening melatonin treatment has been shown to advance sleep/wake cycles
towards normal time with an overall improvement in the quality of life (Dahlitz et al,
1991; Alvarez et al, 1992; Tzischinsky et al, 1993).
1.8.3 Insomnia
Supraphysiological levels of melatonin (75mg oral dose) administered in the
late evening improved subjective measurements of total sleep time and daytime
alertness in chronic insomniacs (MacFarlane et al, 1991), although changes in sleep
structure with polygrahic recordings were less clear in a similar group of patients
after 1mg and 5mg of melatonin at 2100h (James et al, 1987). In artificially induced
insomnia, Waldhauser et al (1990) demonstrated significant improvements in sleep
disruption and onset after 80mg melatonin given at 2100h and suggested that these
changes were similar to the effects of hypnotics. Beneficial effects of far lower doses
of melatonin (0.5-5mg) have been observed on disturbed sleep/wake patterns and
overall behaviour in children with chronic neurological disabilities and behavioural
problems, where other sleep management techniques including sedatives, strict
scheduling of the sleep/wake cycle and psychological procedures, had proven
78
unsuccessful (Jan et al, 1994).
In conditions where the nighttime melatonin profile has been abolished after
destruction of the pineal gland, the lack of endogenous melatonin may prove to be
detrimental. Petterborg et al (1991b) reported on a subject whose pineal gland had
been destroyed through radiation therapy for a pineal astrocytoma and as a
consequence suffered from insomnia and daytime sleepiness. Daily melatonin
treatment during the late evening improved subjective sleep and daily mood and led
to the synchronisation of growth hormone and prolactin secretory rhythms (melatonin
and temperature were not measured).
1.9 ADVERSE EFFECTS OF MELATONIN
The successful alleviation of behavioural problems by melatonin, with the
realignment of at least the sleep/wake rhythm with the local time cues in a number
of these circadian rhythm disorders, indicates a wide therapeutic spectrum for this
compound. However, the effects of exogenous melatonin on other physiological and
behavioural parameters under a range of doses and under long term treatment must
first be established. Melatonin, even in very high doses has extremely low toxicity
in rodents (Sugden, 1983). In humans, no serious adverse reactions have been
reported to date with short-term or long term use of low doses. Somnolence may
occur, but this may be an important physiological effect of the hormone and not a
pharmacological side-effect, see section 1.10.1.
79
1.9.1 Endocrinological effects
Low milligram doses of melatonin have little acute or delayed effects on the
levels of a number of endocrine variables including growth hormone, luteinizing
hormone (LH), testosterone, cortisol, follicular stimulating hormone and thyroxine
even after repeated daily administration for up to one month (Lissoni et al, 1986;
Wright et al, 1986; Waldhauser et al, 1987; Mallo et al, 1988; Petterborg et al,
1991; Terzolo et al, 1993), although in one report melatonin acutely amplified LH
pulses in women when given in the morning (Cagnacci et al, 1991). Slight changes
in prolactin levels may occur but only at certain times of day (Lissoni et al, 1986;
Wright et al, 1986; Waldhauser et al, 1987; Mallo et al, 1988; Terzolo et al,
1993). In much higher doses over longer periods (300mg/day for 4 months),
melatonin has been shown to reduce levels of luteinizing hormone in women. This
preparation in combination with a progestin is undergoing clinical trials as a
contraceptive (Voordouw et al, 1992).
Cortisol and melatonin secretion are thought to be inversely correlated
(Wetterberg et al, 1982). The masked rhythms are, however, readily dissociated
under phase-shifted conditions (Fevre-Montange et al, 1981). Although melatonin
administration to humans may facilitate the reentrainment of cortisol rhythms after a
time zone change (Arendt et al, 1987), in other situations exogenous melatonin will
shift the phase of its own rhythm but not the cortisol rhythm (Wright et al, 1986;
Mallo et al, 1988; Zaidan et al, 1994). It is not surprising that circadian rhythms
show a close correlation to each other under entrained conditions. However, a causal
effect is not supported.
80
1.10 ACUTE EFFECTS OF MELATONIN AND BRIGHT LIGHT
1.10.1 Sedative effects of melatonin
Reports of the `hypnotic' effects of melatonin in humans have been
inconsistent. Daytime administration of melatonin given orally, intravenously or
intranasally in doses of 0.3mg to 240mg has been shown to increase drowsiness,
reduce performance and advance sleep onset time (Cramer et al, 1974; Vollrath et al,
1981; Arendt et al, 1984; Lieberman et al, 1984; Waldhauser et al, 1990; Dollins et
al, 1993,1994). Using objective measurements of sleep (polysomnography), Anton-
Tay (1974) reported significant changes in sleep structure with increased REM density
and stage 2 sleep after morning administration of lg melatonin for 6 days. Nickelsen
et al (1989) reported increased drowsiness in the morning (0900h) but not in the
evening (1900h), after an oral dose of 50mg, whereas Arendt et al (1984,1985a)
demonstrated that subjective evening fatigue only increased significantly after 4 days
with 2mg melatonin daily at 1700h. James et al (1987) after late evening (2300h)
treatment of lmg and 5mg melatonin found an increased REM latency with 5mg only,
but the sleep onset and duration were not affected by either treatment. No changes
in sleep structure were identified after 50mg of intravenous melatonin administered
just before bedtime even though psychomotor activity was reduced and latency to
sleep shortened (Cramer et al, 1974).
In summary, there are no clear effects of supra-physiological levels of
melatonin on human sleep architecture when measured objectively. It is only in
conditions of circadian rhythm disturbance that melatonin seems to consistently
81
improve sleep quality and adjust the sleep/wake cycle when administered close to the
desired bedtime, see sections 1.7 and 1.8. Whether this improvement in sleep by
melatonin is via sleep inducing/initiating processes and/or circadian mechanisms is
still unclear. It is interesting to note that giving melatonin in doses sufficient to
entrain the motor activity rhythms in the rat, no changes in sleep EEG measurements
have been observed (Tobler et al, 1994).
1.10.2 Arousal/stimulating effects of bright light
In addition to an effect on the circadian system, bright light (z 1000 lux) can
directly and acutely enhance alertness and performance during the night (Campbell
and Dawson, 1990; Czeisler et at, 1990; Badia et at, 1991; Dawson and Campbell,
1991; Hannon et at, 1992) and in the early morning (Clodore et at, 1990), although
it must be noted that in one study, Dollins et at (1993) reported no stimulatory effect
on alertness and performance with light exposure at night. True physiological arousal
with bright light exposure at night has been associated with changes in waking EEG,
but only during the treatment. Badia et at (1991) demonstrated increased beta activity
associated with increased light intensity which was considered to be a reflection of
the stimulating effects of bright light.
The effects of bright light exposure on normal nocturnal sleep are less clear.
In some cases, bright light had no effect on sleep latency (Drennan et at, 1989;
Bunnell et at, 1992), whereas one study demonstrated increased difficulties in
initiating sleep after bright light (Dijk et at, 1991). In the latter study, longer sleep
latencies were associated with elevated body temperature rather than alterations in
82
brain electrical activity.
The mechanism of action of the acute effects of bright light remains unknown.
However, it has been suggested that the arousal effects may be mediated via
melatonin and its thermoregulatory effect on temperature (Badia et al, 1993).
1.10.3 Stimulating effect of bright light on body temperature
Further evidence of true physiological arousal with bright light exposure has
been demonstrated with increases in body temperature. It has been shown that, under
sleep deprivation conditions, nighttime administration of bright light of z 2500 lux
versus dim light of <_ 50 lux increased tympanic temperature (Badia et al, 1990a,
1990b, 1991; Murphy et al, 1991; Myers and Badia, 1991); administration of bright
light versus dim light during the day showed no difference (Murphy et al, 1991;
French et al, 1990). In addition, Dijk et al (1991) showed that after a single 3h
exposure of bright and dim light (- 2500 lux) before bedtime, core body temperature
was elevated and remained raised for the initial 4h of sleep after the bright light
treatment. A dose-response curve for the effects of light intensity on tympanic
temperature has been established by Myers and Badia (1993). Body temperature
increases were associated with increasing light intensity but this effect was only
apparent after 2100h, which is the approximate time of the onset of melatonin
secretion in plasma (Lewy et al, 1989).
These acute effects of bright light on temperature are only clear during the
nighttime or darkness hours when bright light has a suppressive effect on melatonin,
suggesting that melatonin may mediate this temperature elevating effect.
83
1.10.4 Acute effects of exogenous melatonin on temperature
There is a consistent inverse relationship between the endogenous melatonin
and core body temperature rhythms under normal (Cagnacci et al, 1992), free-running
and non-24h environments (Wever, 1989) which is independent of the sleep/wake
cycle (Akerstedt et al, 1979, Cagnacci et al, 1992).
Melatonin's acute hypothermic properties have been reported in a number of
studies (Carman et al, 1976; Cagnacci et al, 1992; Dawson and Encel, 1993; Dollins
et al, 1994). Suppression of endogenous melatonin levels with the B-adrenergic
antagonist atenolol, attenuated the night time fall in body temperature and reduced the
amplitude by about 40%, suggesting that endogenous melatonin is directly responsible
for at least 40% of the body temperature rhythm (Cagnacci et al, 1992). In addition,
a divided dose of 1.75mg oral melatonin given in the evening completely inhibited
the stimulatory effect of late evening light treatment on body temperature (Cagnacci
et al, 1993).
Strassman et al (1991) have confirmed melatonin's role in the bright light
stimulation of body temperature using a melatonin infusion paradigm (Strassman et
al, 1987) during bright light suppression of endogenous melatonin. Exogenous
melatonin, infused to reproduce an approximate physiological profile, reversed the
bright light-induced elevation of body temperature.
84
1.11 POTENTIAL PHARMACOLOGIC AND BEHAVIOURAL TREATMENT
STRATEGIES FOR CIRCADIAN RHYTHM DISTURBANCES
1.11.1 Manipulation of the sleep/wake cycle
One way of speeding up the readaptation of the circadian system to large
phase-shifts could simply involve manipulating the sleep/wake cycle. The principal
is to split the phase-shift over a number of days (Samel et al, 1994). For example,
to achieve a 9h advance phase-shift, the sleep/wake schedule can be advanced by 5
hours one day and then a further 4 hours the following day. This system prevents the
"antidromic" effect where subjects adapt to the new local time cues by delaying
instead of advancing, thereby shortening the time taken to adjust completely (Samel
et al, 1994).
1.11.2 Melatonin agonists and antagonists
New melatonin analogues have recently been developed. Hopefully these will
help to elucidate the function and mechanisms through which melatonin exerts its
effects on body physiology and in particular the circadian system.
In vitro studies have shown that Luzindole (2-benzyl-N-acetyltryptamine)
inhibits the activation of melatonin receptors and may provide a appropriate
pharmacological melatonin antagonist for studying the physiological role of melatonin
(Dubocovich, 1988). Kennaway et al (1989) demonstrated that the melatonin agonist,
6-chloromelatonin, promoted the synchronization of the urinary 6-sulphatoxymelatonin
85
rhythm after an 8h shift of the light/dark cycle. Recently the amino acid sequence
of the melatonin receptor has been determined thus enabling the development of
melatonin agonists and antagonists with specific structural relationship to the
melatonin receptor (Ebisawa et al, 1994). Development of newer analogues has
resulted in the formation of napthalene bioisosteres with high affinity and specific
binding to melatonin receptor sites in the pars tuberalis (Yous et al, 1992; Howell et
al, 1991).
One such compound, S-20098 (N [2-(7-methoxy-l-naphthyl) ethyl] amide), has
been shown to facilitate the reentrainment of rat locomotor activity rhythms when
free-running or after an 8h advance shift of the light/dark cycle (Bonnefond et al,
1993; Redman and Guardiola-Lemaitre, 1993). In a rat model for delayed sleep-
phase syndrome, S20098 proved effective in advancing delayed activity rhythms,
resetting to them to darkness onset (Armstrong et al, 1993). In vivo studies have
shown S-20098 to mimic the effects of melatonin in phase-shifting the rhythm of
neural activity of the rat SCN (Mason et al, 1993). Another related melatonin agonist
S-20242 can prevent the age-associated reduction in temperature rhythm amplitude
and stability in rats (Koster-van Hoffen et al, 1993).
In humans, doses of up to 800mg of S-20098 have been well tolerated with
one subject experiencing tachycardia and slight hypotension at 1200 mg (Van Reeth
et al, 1994). Morning administration of S-20098 has shown similar hypothermic
effects to melatonin but at higher doses (Van Reeth et al, 1994).
86
1.11.3 Physical activity and hypnotics
In the normal environment, multiple zeitgebers probably act to entrain the
circadian system and when suitably timed may prove useful in resetting the circadian
system after an abrupt shift, or in clinical disorders. In animals, arousal and activity
act as strong zeitgebers. Periods of physical activity and arousal have been shown
to entrain and phase-shift rodent activity rhythms (Mrosovsky and Salmon, 1987;
Reebs and Mrosovsky, 1989; Wickland and Turek, 1991; Hastings et al, 1992).
Recently, the acute phase-shifting effects of single timed pulses of physical activity
at night on the human circadian system have been reported (Van Reeth et al, 1994).
A 3h pulse of exercise prior to the temperature minimum phase-delayed the melatonin
rhythm by 1-2h one day after exercise exposure. Smaller delays were induced after
the temperature minimum.
Benzodiazepines used in the treatment of insomnia, anxiety and other stress-
related problems can induce phase-shifts in rodent activity rhythms (Turek and Losee-
Olsen, 1986). However, this ability is associated with an increase in activity after
administration since restraining hamsters can prevent triazolam-induced phase-shifts
(Van Reeth and Turek, 1989). How increased arousal transducer information into
phase-shifting signals is unknown but is thought to involve the intergeniculate leaflet
of the hypothalamus (Johnson et al, 1988). Connections from here to the SCN
contain the neurotransmitter neuropeptide Y (Moore and Speh, 1993) which has been
shown to phase-shift the neuronal activity in the rat SCN (Medanic and Gillette,
1992). There is also serotonergic input from the raphe nuclei to the IGL and the
SCN which may modulate the phase-shifting response of non-photic zeitgebers
87
(Prosser et al, 1990; Smale et al, 1990). 5-HT agonists can phase-shift the clock in
vitro and in vivo similar to those shown by triazolam and other non-photic stimuli
(Prosser et al, 1990; Tominaga et al, 1992).
In humans, benzodiazepines facilitate sleep adaptation after transmeridian
flight, but not the rate of adjustment of the endogenous body clock (Nicholson et al,
1985; Donaldson and Kennaway, 1991). Hypnotics are commonly used to facilitate
sleep and in shift work studies they facilitate sleep during the day and may improve
alertness and performance during the night shift in some people (Bonnet et al, 1988;
Walsh et al, 1991). They may indirectly affect the circadian pacemaker by imposing
a more regimented sleep/wake and therefore light/dark cycle.
88
1.12 RESEARCH PROPOSAL AND AIMS OF THESIS
Field studies on jet lag are expensive and difficult to control with the
collection of biochemical data an obvious problem. Laboratory simulations provide
a more controlled setting with which to study these disorders in detail. However,
previous simulations have involved complete isolation from the subjects normal
environment. The primary aim of this work was to develop an experimental model
to simulate circadian rhythm disorders such as jet lag and maladaptation to shift work
without complete environmental isolation and to study melatonin and bright light as
potential treatment strategies.
However, in developing such models, it is important to use a physiological or
behavioural rhythm marker that gives an accurate reflection of the behaviour of the
endogenous biological clock. Overt (measured) circadian rhythms are subject to
behavioural and environmental masking influences such as light, sleep, activity, stress
and meals, and can result in the unreliable estimation of the circadian time of the
internal biological clock. Compared to other circadian rhythm markers melatonin is
less susceptible to exogenous masking influences other than light, and is extensively
used to assess circadian rhythm status. Thus any factors influencing its concentration
other than endogenous rhythmicity must be assessed. Changing posture influences
the levels of a number components in the blood. If melatonin appears to be
influenced by a change in posture then it must be controlled for when designing
studies. The effects of changing posture on melatonin concentrations in plasma and
saliva were therefore assessed.
One method of separating these masking effects from the endogenous profile
89
of the temperature rhythm is a mathetical "purification" procedure developed by
Minors and Waterhouse (1988,1992). This procedure was used to "purify" the
temperature data after a simulated rapid time zone (in collaboration with Drs.
Waterhouse and Minors, University of Manchester, UK) to give a more reliable
estimation of the endogenous behaviour of the body clock. The problem of masking
and the usefulness and assumptions of such a procedure are discussed with reference
to the experiment.
Due to the difficulties in measuring the melatonin secretory rhythm in the
blood, non-invasive techniques have been employed to measure the melatonin rhythm
in saliva and its urinary metabolite rhythm in the urine. However, the relationship
between melatonin in the blood, melatonin in saliva and its urinary metabolite under
different conditions must be firmly established. Since most of the following
experiments involve a phase-shift, the relationship between melatonin rhythm
parameters in different mediums were evaluated under normal and phase-shifted
conditions.
Light and melatonin therapy to treat jet lag can be timed to induce phase-
advances or phase-delays, and their respective PRCs mirror each other to some
extent, i. e. they act in opposite directions. Combined treatments have not yet been
studied. In real life, natural light exposure may occur at inappropriate times for the
most expeditious reentrainment of circadian rhythms and oppose the effects of
melatonin treatment. So, after an abrupt phase-shift in local time cues, the beneficial
effects of suitably timed melatonin treatment were studied in the presence of bright
light exposure timed to oppose the phase-shifting effects of melatonin.
Finally, a group of studies were carried out to elucidate the mechanisms
90
through which melatonin may exert its beneficial effects on jet lag and other circadian
rhythm disorders.
Experimental treatment for jet lag has commonly involved repeated daily
administration of melatonin at the same time of day to facilitate the reentrainment of
circadian rhythms, but little is known about the phase-shifting effects of single
melatonin treatments. In addition to these circadian effects, melatonin also causes
immediate effects on body temperature and behaviour after acute administration. It
is possible that these effects of melatonin are related. In an attempt to study any
relationship between these phenomena, the effectiveness of a single 5mg melatonin
treatment -a dose commonly used to treat circadian rhythm disorders - to phase-shift
the endogenous clock was compared to its acute effects on temperature and behaviour.
It has been demonstrated that zeitgeber strength - at least with regards to light
- is important in entraining and phase-shifting circadian rhythms. A human dose-
response phase-shifting effect of melatonin has not been demonstrated but would help
to determine optimum doses of melatonin needed to treat jet lag and other circadian
rhythm disorders. The phase-shifting effects of single, low oral doses of melatonin
were studied and compared to their immediate effects on body temperature and
behaviour.
91
CHAPTER 2
CLINICAL CONSIDERATIONS PRIOR TO STUDIES.
PHYSIOLOGICAL AND BEHAVIOURAL MEASUREMENTS
AND ANALYSIS
Clinical considerations prior to studies.
Physiological and behavioural measurements and analysis
2.1 CLINICAL TRIALS: general conditions
2.1.1 Subject criteria
The subjects for these trials were healthy volunteers under no medication
except for minor analgesics and oral contraceptives, except for the studies described
in Chapters 9 and 10. In the latter studies, aspirin-based analgesics and females
taking oral contraceptives were excluded because of their possible influence on body
temperature. Ingestion of any drug throughout the trials was noted.
Further exclusion criteria were as follows:
1) Subjects who had either undertaken a transmeridian flight or night shift work
within one month of the commencement of the trial.
2) Subjects who normally consumed more than two units of alcohol per day.
3) Knowledge or suspicion of pregnancy.
4) Familial or personal history of psychiatric disorders, migraine, epilepsy, sleep
disorders or food allergies.
5) Heavy smokers (> 10 per day).
6) Subjects who had donated blood within the preceding 3 months.
All subjects gave informed consent and the study protocols were approved by
the Advisory Committee on Ethics of the University of Surrey. In addition, in studies
93
where blood sampling was necessary, subjects read and signed the University of
Surrey's consent form for the donation of blood for teaching or research.
2.1.2 Details of the melatonin preparation
The melatonin used in all studies is licensed by the FDA for experimental use
in normal human volunteers (M5250 FD). A clinical trials exemption certificate is
held by Prof. J. Arendt (University of Surrey, UK) under who's supervision these
trials were conducted. It is chemically synthesised and supplied by Sigma Chemicals
Co. Ltd. (Poole, Dorset, UK., product no. W 5318-3). It is used either as a gelatine-
lactose capsule, prepared by Penn Pharmaceuticals Ltd. (Gwent, Wales), with quality
control at the University of Surrey, or as specified in the relevant chapters.
Possible side-effects of melatonin
Melatonin has been administered orally (2-10mg) in a number of studies with
minimal side-effects (Arendt et al, 1987; Arendt and Aldhous, 1988, Wright et al,
1986, Samel et al, 1991; Jan et al, 1994). Table 2.1 shows the percentage number
of people who suffered from a particular side-effect as a result of taking melatonin
(5mg) for the alleviation of jet lag in field studies carried out by Prof. J. Arendt
(University of Surrey, UK).
94
Table 2.1 Percent side effects reported more than once after taking 5mg melatonin
to alleviate jet lag (from 1984 to Nov. 1994)
Side-effect Placebo (n = 112) Melatonin (n=484)
Sleepiness 1.78 8.3
Headache 2.7 1.65
Nausea 0.9 0.8
Dizziness 0 1.4
2.1.3 Health and safety effects of light
Moderately bright, full spectrum light up to a maximum of 2000 lux at eye
level was used in these experiments. The source of light consisted of a light box with
6-8 light tubes made by True-Lite, Duro-Test Corporation, Fairfield, NJ, USA,
obtained through Full Spectrum Lighting, High Wycombe, Buckinghamshire, UK.
The intensities of light used were much less than that experienced on a bright sunny
day (up to 100 000 lux, Thorington, 1985). Normal artificial room intensity lighting
at eye level ranges from 50-600 lux. Full spectrum light (2500 lux for up to 6 hours)
is used clinically to treat seasonal affective disorder (Terman et al, 1990; Terman and A
Terman, 1990). To date, no problems have been reported even'the use of relatively
high intensity light (10 000 lux) full spectrum light (Terman and Terman, 1990;
95
Goodwin, 1992).
The levels of UV light from conventional fluorescent lamps and the light boxes
used are extremely low and not thought to be detrimental to health (Whillock et al,
1988). The trend towards the use of brighter high efficiency sources suggests that
phototherapy lamps are unlikely to have any adverse effects on health and may even
be advantageous.
2.2 COLLECTION AND STORAGE OF SAMPLES
Plasma: Blood samples were collected into lithium heparin tubes and centrifuged
for 15 minutes at 3000g. The plasma was separated carefully within 25 minutes of
collection, and stored at -20°C until analysis.
Saliva: Saliva samples were collected into plastic vials. In some studies saliva
was collected with stimulation by chewing for 2 minutes on parafilm. The method
of collection was constant within each study. There is a possibility that stimulation
may affect salivary melatonin levels through altering salivary flow rates. However,
no significant differences in the ratio between melatonin levels in plasma and saliva
have been observed with and without stimulation of flow rate by chewing gum or
parafilm (McIntyre et al, 1987; Miles, 1989).
Caffeine and smoking were excluded during saliva sampling due to
interference with the assay. All samples were checked for the presence of blood and
food which may have contributed to high salivary melatonin levels in other studies
(McIntyre et at, 1987; Miles, 1989). Samples were stored at -20°C until analysis.
96
After thawing, if any samples contained visible particulate matter they were
centrifuged before assay.
Urine: Urine volumes were recorded on collection and a 5m1 aliquot placed
in a plastic vial and stored at -20°C until analysis. aMT6s has been found to be very
stable in urine for up to 5 days at room temperature or 4°C and for up to 2 years at -
20°C either with or without preservative (Bojkowski, 1988).
2.3 RADIOIMMUNOASSAY OF MELATONIN AND aMT6s
2.3.1 Basic principle and summary of procedure
The sample is incubated with a fixed amount of antiserum specific to the
melatonin/aMT6s (antigen) together with trace amounts of radioactive
melatonin/aMT6s (labelled antigen) which competes with the non-radioactive antigen
for the limited number of antibody binding sites. Since the number of antibody
binding sites and the amount of radioactive melatonin/aMT6s is constant, the amount
of radioactivity associated with the antibody is dependent upon the concentration of
melatonin/aMT6s (antigen) in the sample. When equilibrium is established, the free
and antibody-bound fractions of melatonin/aMT6s are separated either by charcoal
adsorption of free antigen or with a second precipitating antibody system.
The melatonin/aMT6s concentration in unknown samples can be determined
by comparing the amount of radioactivity measured in either the bound or the free
fraction to that obtained with a set standards of known concentration.
97
2.3.2 RIA of melatonin in human plasma
Plasma melatonin levels were determined by direct RIA using the method
described by Fraser et al (1983) which is used routinely in our laboratory.
Materials: Antiserum (product no. AB/S/02 or AB/S/021) and melatonin-free
plasma (product no. LMHP1), were obtained from Stockgrand Ltd., University of
Surrey, UK. 3H-melatonin (product no. TRK-798) was purchased from Amersham
International (Amersham, Buckinghamshire, UK). 2,5-diphenyloxazole (PPO) and
1,4-bis-[2-(5-phenyloxazolyl)]benzene (POPOP, low sulphur grade) were obtained
from Fisons (UK). All other chemicals including melatonin (product no. M-5250)
were obtained from Sigma Chemical Co. Ltd. (Poole, Dorset, UK).
Reagents
Buffer. All dilutions were made in buffer made freshly every week. This
consisted of tricine (0.1 mol/1), 0.9% NaCl and 0.1% gelatin made up in double-
glass-distilled water (DGDW) to give a stock solution at pH 5.5. In order to dissolve
the gelatin, the buffer was heated to 50°C for 30 minutes.
Antiserum: The antiserum was supplied freeze-dried and was reconstituted with
2m1 of DGDW to provide an intermediate dilution of 1: 10. It was then divided into
50µ1 aliquots and stored frozen at -20°C. Diluting one 501A1 aliquot to 20m1 with
tricine buffer provides enough antiserum for 100 tubes with a dilution of 1: 4000.
98
Dextran-coated charcoal: Norit activated charcoal (2 %) was suspended in tricine
buffer for 5 minutes. After centrifugation at 4°C at 1000rpm, the supernatant and any
fines around the sides of the vessel were discarded. The charcoal was resuspended
in the original volume of tricine buffer and then stirred with 0.02 % Dextran T70 for
at least lh at 4°C, before storing at 4°C. The suspension was freshly made each
week.
Radiolabel: An intermediate dilution of the stock 3H-melatonin was prepared by
diluting 20µl to 2m1 in absolute ethanol and storing at -20°C. A working solution was
freshly prepared by further diluting with tricine buffer to give approximately 5000cpm
in 1O0µ1.
Scintillation fluid: The toluene-based scintillant was prepared by adding 12.5g PPO
and 0.75g POPOP to 2.5 litres of toluene.
Melatonin standards: A stock melatonin standard was prepared every 3 months
by dissolving 10mg melatonin in 0.5m1 absolute ethanol. This was made up to lOml
with DGDW to give a concentration of 1 mg/ml and stored at 4°C. The working
standard was freshly made for each assay from this stock solution as follows:
100µ1 of 1mg/ml solution to 100ml in DGDW to give 1µg/ml
500141 of 1µg/ml solution to 50m1 in DGDW to give lOng/ml
125µl of lOng/ml solution to 2.5m1 in melatonin-free plasma to give 0.5ng/ml.
99
Standards covering the range 2.5 to 250 pg/ml for the standard curve were prepared
with further dilutions in melatonin-free plasma as follows:
Melatonin standard Melatonin-free Melatonin conc. 0.5ng/ml (µ1) plasma (µl) pg/0.5m1
500 0
5 495 2.5
10 490 5.0
25 475 12.5
50 450 25.0
100 400 50.0
200 300 100
500 - 250
Duplicate tubes were prepared for all samples, standards, non-specific binding tubes,
quality control samples. Total counts tubes were prepared in quadruplicate.
Procedure
Plasma samples, melatonin standards and quality control samples in 500µ1
aliquots were mixed with 200µ1 antiserum then incubated for 30 minutes at room
temperature. 100µl 3H-melatonin were then added, mixed again and the assay
incubated for 18h at 4°C. The antibody-bound melatonin was separated from the free
fraction by mixing with 5001A1 of dextran-coated charcoal then incubating for 15
minutes at 4°C. The charcoal was continuously stirred for 30 minutes prior to
100
addition to make sure it was suspended properly in the buffer. Rapid addition of the
charcoal to all assay tubes was necessary to reduce intra-assay variation. After 15
minutes, the assay tubes were centrifuged at 1500g for 15 minutes at 4°C. 7O0µ1 of
the supernatant were transferred into vials containing 4m1 of scintillation fluid. The
vials were then shaken for lh in order to extract all the 3H-melatonin into the organic
phase to provide an appropriate medium for radioactivity detection without quenching.
Radioactivity in the tubes were counted in a suitable B-counter and the sample
concentrations calculated using a'RIAcalc' program.
The RIAcalc program automatically determines the percentage of total counts
bound or free, and then plots them as a function of the known concentrations of
melatonin standards. A curve is fitted through the standard points based on the
smoothing spline calculation method. This fits a continuous curve to each pair of
standard points using the conventional method of least squares, but in addition, it
calculates the best-fitting curve with the minimum number of turning points and is as
smooth as possible. The concentration of the unknown samples are then determined
from this curve.
101
2.3.3 RIA of melatonin in human saliva
Salivary melatonin levels were determined using a modification (English et al,
1993) of the direct RIA using the method described by Fraser et al (1983), and used
routinely in our laboratory.
Materials: Antiserum (product no. AB/R/03) was obtained from Stockgrand Ltd.,
University of Surrey, UK. The separation system SAC-CEL (a solid-phase, second
antibody [donkey anti-rabbit] coated cellulose suspension) was purchased from IDS,
Bolden, Tyne & Wear, UK). 125I-2-iodomelatonin was purchased from Amersham
International (Amersham, Buckinghamshire, UK). All other chemicals including
melatonin (product no. M-5250) were obtained from Sigma Chemical Co. Ltd.
(Poole, Dorset, UK).
Reagents
Bq er: as described in 2.1.1 except that 0.2 % gelatine was used and the buffer was
adjusted to pH 8 with 1M NaOH.
Antiserum: The antiserum was supplied freeze-dried (133µl of a 1: 10 dilution) and
was reconstituted with 2m1 DGDW to provide an intermediate dilution of 1: 150.5O1&1
aliquots were stored at -20°C. Diluting one 501A1 aliquot to lOml with tricine buffer
provided enough antiserum for 100 tubes with a dilution of 1: 30 000.
102
Separation system: A solid phase second antibody coated cellulose suspension,
SAC-CEL, was used as supplied.
Radiolabel: "I-2-iodomelatonin was diluted in buffer to give approximately 10 000
cpm in l00µ1.
Melatonin standards: The stock solution was prepared as described in section 2.1.2.
The working solution was freshly prepared as follows:
10011 of lmg/ml melatonin solution to 100ml in DGDW to give 1µg/ml
5O01A1 of 1µg/ml melatonin solution to 50m1 in DGDW to give lOng/ml
5O1A1 of long/ml melatonin solution to 2.5m1 in buffer to give 0.2ng/ml
Standards covering the range 2 to 200 pg/ml for the standard curve were prepared
with further dilutions in buffer as follows:
Melatonin standard Assay buffer Melatonin conc. 0.2ng/ml (µl) (µl) pg/ml
500 0 5 495 2 10 490 4 25 475 10 50 450 20 125 375 50 250 250 100
500 - 200
103
Duplicate tubes were prepared for all samples, standards, non-specific binding
tubes, quality control samples. Total counts tubes were prepared in quadruplicate.
Procedure
Saliva samples, quality controls and standards in 500141 aliquots were incubated
with 100141 antiserum at room temperature for 30 minutes. 100µl of '25 I-2-
iodomelatonin were added and the assay was incubated at room temperature for 4h.
The antibody-bound melatonin and free fractions were separated using 100/41 of SAC-
CEL. After addition, the assay was intermittently mixed for lh at room temperature,
centrifuged at 1500g for 10 minutes at 20°C and the suspension discarded.
Radioactivity of the remaining pellet was measured in a gamma counter and the
sample concentrations read from the standard curve using a `RIACa1c' program.
104
2.3.4 RIA of 6-sulphatoxymelatonin in human urine
6-sulphatoxymelatonin (aMT6s) levels were determined using the method
described by Aldhous and Arendt et al (1988), and used routinely in our laboratory.
Materials
Antiserum (product no. AB/S/04), 'I-aMT6s, charcoal-stripped (aMT6s-free)
urine and aMT6s standard were obtained from Stockgrand Ltd., University of Surrey,
UK. All other chemicals were obtained from Sigma Chemical Co. Ltd. (Poole,
Dorset, UK).
Reagents
Buffer. - as described in 2.1.1 .
Dextran-coated charcoal: as described in 2.1.1
Antiserum: The antiserum was supplied freeze-dried and was reconstituted with
1ml DGDW and 9m1 of buffer added to provide an intermediate dilution of 1: 100.
100µl aliquots were stored at -20°C. Diluting one 50µl aliquot to 20m1 with tricine
buffer provided enough antiserum for 100 tubes with a dilution of 1: 20 000.
Radiolabel: `I-aMT6s was diluted in buffer to give approximately 8000 to 10 000
cpm per l00µ1.
105
Charcoal-stripped urine: This was supplied freeze-dried (1O01A1 per vial) and was
reconstituted with 25m1 buffer to give a dilution of 1: 250.
aMT6s standards: The stock was supplied as 200pg/ml diluted in 1: 250 with charcoal-
stripped plasma. Standards covering the range 2 to 50 ng/ml for the standard curve
were prepared with further dilutions in buffer as follows:
aMT6s standard aMT6s-free urine aMT6s conc. 200pg/ml(µ1) 1: 250 dilution (µl) ng/ml
500 0
5 495 0.5
10 490 1
20 480 2
40 460 4
70 430 7
100 400 10
200 300 20
500 - 50
Duplicate tubes were prepared for all samples, standards, non-specific binding
tubes, quality control samples. Total counts tubes were prepared in quadruplicate.
106
Procedure
Urine samples were diluted 1: 250 with buffer prior to the assay. Urine
samples, quality controls and standards in 500141 aliquots were incubated with 200µ1
antiserum at room temperature for 30 minutes. 100µ1 of ' I-aMT6s were then added,
mixed again and the assay incubated for 15-18h at 4°C. The antibody-bound aMT6s
was separated from the free fraction by mixing with l00141 of dextran-coated charcoal
then incubating for 15 minutes at 4°C. The charcoal was continuously stirred for 30
minutes prior to addition to make sure it was suspended properly in the buffer. Rapid
addition of the charcoal to all assay tubes was necessary to reduce intra-assay
variation. After 15 minutes, the assay tubes were centrifuged at 3500g for 15
minutes at 4°C and the supernatant immediately discarded. Radioactivity of the non-
antibody-bound fraction in the charcoal pellet was counted in a suitable gamma
counter and the sample concentrations calculated using a' RlAcalc' program.
107
2.4 LIGHT INTENSITY MEASUREMENT
Light intensity was measured using a handheld digital light meter obtained
from Full Spectrum Lighting, High Wycombe, Buckinghamshire, UK. Measurements
were taken at eye level directly in front of the eyes towards the light source and other
fields of view likely to be used by the subjects (i. e. towards computer screen,
television, etc. ).
2.5 CORE BODY TEMPERATURE
Core body temperature was recorded every 6 minutes using rectal probes and
Squirrel Minilogger recording devices (Grant Instruments, Cambridge, UK). Subjects
wore rectal probes and recording devices continuously except during
defecation/showering/bathing when necessary. Perception of the probe decreased
after one day.
2.6 ACTIVITY DIARIES
Subjects were required to record their activity in a diary during each day of
the study (certain experiments only). The day was divided into half-hourly intervals
and subjects ticked the appropriate box if they:
1) consumed any food/drink (identifying a meal as opposed to a snack)
108
2) exercised (anything more intensive than walking)
3) relaxed (rested)
4) were stressed
5) any drug ingestion and alcohol
6) removed the temperature probe
7) were sedentary whilst working
8) were showering
within the half-hour.
These diaries were necessary to determine reasons for missing or erroneous
values in the raw data and to assess subject conformity. In addition, they enabled
exogenous influences (e. g. exercise and stress) which affect body temperature to be
mathematically removed to give a more accurate profile of the endogenous
temperature rhythm, see Chapter 7.
2.7 BEHAVIOURAL MEASUREMENTS
All behavioural assessments were performed using handheld computers, Psion
Organisers (Psion UK, Greenford, UK). Programs for the Psions were written and
kindly provided by Dr P. Totterdell (University of Sheffield, UK) and Professor S.
Folkard (University of Swansea, UK) who have used this method of recording
behavioural parameters extensively (Totterdell and Folkard, 1992).
109
2.7.1 Mood and sleep quality ratings
Subjective alertness was rated on a visual analogue scale as follows:
Very drowsy Very alert
The extreme right end of the line represented the most alert a subject ever felt, and
the extreme left end of the scale the most drowsy that they ever felt. In a similar
fashion subjective cheerfulness (Very sad ý--- Very happy), calmness (Very calm
ý----º Very tense) and sleep quality (Very poor ý--ý Very good) were rated.
Since these ratings were purely subjective and depended to a great degree on
how individuals interpreted these scales, subjects were used as their own controls.
2.7.2. Sleep logs
Subjects completed sleep logs daily on waking and recorded:
1) time of retiring to bed, 2) time of trying to sleep, 3) how long it took to fall
asleep, 4) number of night awakenings, 5) the total duration of night awakenings,
6) time of waking and 7) time of getting up. Sleep quality was rated on a VAS as
described above. The sleep logs and sleep quality scales used were originally
developed by Prof. A. Borb6ly (University of Zurich, Switzerland).
110
2.7.3 Performance measurements: Search and memory (SAM) tests
Subjects were required to memorise either one (SAM-1) or five (SAM-5)
letters of the alphabet after which they were shown two lines of alphabetic characters.
Every time a memorised letter was recognised they pressed a key on the computer.
Subjects performed these tasks as quickly as possible whilst keeping their performance
as accurate as possible. For example, when required to identify the letters RKTL
Z in the following sequence of letters:
GHjJVSYPRVBMFGBHJM
FGDWZXCFWGLVQWQZQWYJH
the response key would be pressed 8 times (the target letters have been highlighted
and underlined as an example).
The performance parameters of accuracy, response time and efficiency were
determined from the data and were defined as the percentage of correct responses,
time taken to complete the task and accuracy/square root response time. Since
performance variables are highly dependent upon the individual, for example some
people are naturally slow but accurate, others are naturally quick but inaccurate,
subjects were used as their own controls. Where grouped comparisons were
necessary all values were taken as a percentage of the individuals own baseline before
grouping.
111
2.8 COSINOR ANALYSIS
Circadian rhythms are not only characterized by the period (i. e. the rhythm
repeats itself about every 24h) but also the shape, amplitude and mean level. The
studies described in this thesis were concerned with the phase (time) relationship
between endogenous circadian rhythms and the environmental zeitgebers after induced
phase-shifts together with the amplitude of the rhythm. The rhythmic parameters can
be determined using various curve-fitting programs or just through visual inspection
of the data. However, interpreting the phase or amplitude of the rhythm from a set
of data can often lead to unreliable estimates, especially where there are difficulties
in the collection of data. For example, comparing the maximum/minimum values
from infrequent data points in a 24h profile to identify phase differences from day to
day can lead to problems associated with using a single possible aberrant value and
which may not be the actual maximum or minimum point of the rhythm. The
frequency of sampling is therefore of considerable importance when trying to identify
the true parameters. In some cases, e. g. urine collection, frequent sampling is
impractical and integrated samples are obtained over a period of time after
accumulation in the bladder. This obviously leads to a loss of definition of the
circadian profile since each sample represents a average of values over the whole
collection period. However, urine sampling is convenient and practical and provides
accurate estimates of total 24h excretion.
One method of estimating the parameters of a rhythm based on all the
collected sample points is to fit a curve which "best fits" the data. This can provide
a powerful means of calculating rhythmic parameters with irregular and infrequent
112
sampling points.
One curve-fitting method known as "cosinor analysis" provides measures of
all parameters except for the shape of the rhythm which it assumes to be sinusoidal
in shape (Nelson et al, 1979; Minors and Waterhouse, 1988). The best-fitting cosine
curve is fitted to the data by the conventional method of least squares. Estimates of
the amplitude (variation in concentration over time), acrophase (peak time) and mesor
(mean concentration over time) of the rhythm are provided from the "best-fitting"
cosine curve. Figure 2.1 illustrates the terms used to describe these circadian rhythm
parameters.
Acrophase (time of peak)
Nadir (time of minimum)
Figure 2.1 Parameters of a circadian rhythm
113
A major assumption of cosinor analysis is that the rhythm fits a cosine curve
even though many rhythmic circadian functions clearly do not. There may also be
instances where the shape of the rhythm changes such as after time-zone travel or
workers on alternating night/day shifts. In order to check whether the cosinor-
derived parameters are valid, two estimates of the "goodness of fit" are determined.
Firstly, cosinor analysis identifies whether the data are more likely to fit a straight
line as opposed to a cosine curve. A "significant fit" is set at p<0.05 such that the
probability that the data will fit a straight line is !. 5%. Secondly, the percentage
variability in the data accounted for by the cosine curve (% rhythm) is given. A
100% rhythm would indicate a perfect fit.
In order to use this method of cosinor analysis, the period of the rhythm must
be known or assumed. Under entrained conditions where the subject is exposed to
the normal 24h zeitgebers, the period of the rhythm is expected to be 24h. However,
by stating a range of periods for the cosinor analysis program, it is possible to
identify the probable period of the rhythm by identifying the period with the best %
rhythm and lowest probability of a straight line fit.
In the following studies, cosinor analysis was used to identify whether a
particular 24h data set reliably fitted a 24h cosine curve during the baseline or control
days under normal entrained conditions. Repeated measures ANOVA was used to
confirm a significant time of day (rhythmic) effect. If significant rhythmicity was
established by these two measures, all subsequent comparisons were made with
respect to the control values. An estimate of reentrainment (and reestablishment of
24h cosinor rhythmicity) were provided by the two estimates of "goodness of fit"
described above. Comparisons with raw data profiles and ANOVA performed on the
114
raw data were used to confirm the results from cosinor analysis.
Since for some performance and mood parameters, the number of data points
collected in one day was small and only obtained during wakefulness, several days'
data were combined to achieve a more reliable estimate of rhythmic parameters using
cosinor analysis (Arbogast et al, 1983).
There are other methods of determining rhythm parameters (see Klemfuss and
Clompton, 1993 and Refinetti, 1993 for comparison of methods), but few allow the
phase of the rhythm to be estimated. Cosinor analysis provides an appropriate
method of determining the phase and amplitude of circadian rhythms (Minors and
Waterhouse, 1988) and was therefore employed in the studies described in this thesis.
However, comparison of the cosinor-derived values with raw data were used to
confirm the results. The cosinor analysis program used was written and kindly
provided by Dr D. S. Minors (University of Manchester, UK).
115
CHAPTER 3
INFLUENCE OF POSTURE ON MELATONIN
CONCENTRATIONS IN PLASMA AND SALIVA
Influence of Posture on Melatonin Concentrations in Plasma and Saliva
3.1 INTRODUCTION
The force of gravity causes significant changes in the human circulatory
system and with the continual changing of posture in everyday life the influence of
gravity on various blood parameters becomes apparent. In healthy subjects, when an
upright position is adopted after lying recumbent (supine), the plasma volume is
reduced and the concentration of various plasma components, especially that of the
proteins and of the constituents bound to them, increases. This rise in plasma
constituent concentration is in the order of 10-20% and the effect is reversible on
reversing the postural position (Stoker and Wynn, 1966; Husdan et al, 1973; Tan et
al, 1973; Hagan et al, 1978).
The pineal hormone melatonin is secreted into the blood in a highly stable and
reproducible pattern within individuals (Broadway et al, 1987). An important feature
of its secretion is that, with the exception of light (Lewy et al, 1980; Bojkowski et
al, 1987a; McIntyre et al, 1989a), it is less susceptible to behavioural and
environmental masking (disturbing) influences, than other overt (measured) circadian
rhythms such as core body temperature and cortisol secretion. These latter rhythms
are masked by factors including sleep, waking, meals and stress (Czeilser et al, 1976;
English et al, 1982; Chapter 7) and can lead to inaccurate estimates of the circadian
time of the internal biological clock. For this reason the melatonin circadian rhythm
is thought to provide an appropriate and reliable marker of the endogenous biological
horologue under normal conditions, as well as in constant routine and in other
117
isolated environments.
Although melatonin, a lipophilic molecule, easily filters through capillary
membranes, up to 70% of melatonin may be weakly bound to albumin (Cardinali et
al, 1972) which is not readily diffusible, and albumin-bound hormones do not pass
into the surrounding tissues (Riad-Fahmy et al, 1982); the ratio of albumin in plasma
to saliva is 10,000: 1 (Vining and McGinley, 1984). It is therefore possible that a
shift in posture may cause a change in plasma melatonin levels.
Salivary melatonin is thought to reflect closely the changes in plasma
melatonin levels (Vakkuri, 1985; Miles et al, 1987; McIntyre et al, 1987; Nowak et
al, 1987), and since saliva is becoming a routine method of determining melatonin
secretion it is important to consider the effects of posture on salivary melatonin levels
as well. It is the purpose of this investigation to study the possible masking influence
of posture on melatonin levels.
3.2 SUBJECTS and METHODS
3.2.1 Subjects
7 healthy volunteers (3M: 4F), aged 24-35 years (26.3 ±4 years, mean f
SD), one of whom was a smoker, refrained from heavy exercise and drinking
alcoholic beverages or those containing caffeine for 12 hours prior to and during each
stage of the study. On each day of the study, the subjects moderated their fluid
intake to no more than 1.5L daily.
118
3.2.2 Design
The 4-day study was divided into two periods (stages 1 and 2) of 2 days, 1
week apart, and was performed in October during the rising phase of the melatonin
rhythm, between 2100h and 0130h. On all 4 days subjects remained in a sitting
position from 2100-2400h and under domestic lighting (150-300 lux) from 2100-
2300h. From 2300-0130h, subjects remained awake under dim light (< 10 lux).
After this baseline period the treatments were applied within each stage as a
randomised crossover design. From 2400h the subjects assumed a different postural
position as follows:
Stage la: 2400h - standing, 0030h - supine, 0100h - standing,
Stage lb: 2400h -supine, 0030h - standing, 0100h - supine,
Stage 2a: 2400-0130h - supine,
Stage 2b: 2400-0130h - standing.
Brief, minimal and slow movement was allowed throughout each stage.
3.2.3 Sampling
Blood samples (5 mis, collected by dim red torch light, < 10 lux, after
2300h) and saliva samples (after chewing on parafilm) were taken hourly via an
indwelling forearm cannula, from 2100-2400h and then every 10 minutes from 2400h
until 0130h. Only limited water was allowed immediately after saliva sampling, to
reduce any possible interference with salivary melatonin estimation. Urine was
collected to determine any differences in aMT6s excretion due to changes in posture.
119
After completely voiding the bladder at 2400h, urine was collected at the end of 90
minutes in the supine (stage 2a) and standing (stage 2b) positions.
Samples were collected, stored and analysed as described in section 2.2. The
interassay coefficients of variation for the plasma assay were 12.4%, 9% and 7.3%
at 19.4,36.8 and 191 pg/ml (n =10). The interassay coefficients of variation for the
saliva assay were 16%, 10%, 9% and 6% at 8.7,20,36 and 58 pg/ml (n=8). For
aMT6s, the interassay coefficients of variation were 16.8%, 11.6% and 5.8% at 2.4,
14.7 and 31 ng/ml. The minimum level of detection for each assay was 3.8 ± 0.24
pg/ml, 1.9 ± 0.15 pg/ml and 0.3 ± 0.05 ng/ml (mean ± SEM), respectively.
3.2.4 Statistical analysis
All plasma and salivary melatonin values were calculated as a percentage
change in melatonin levels from the initial point of assuming a different postural
position. Each subject was used as their own control, in view of the large inter-
individual differences in melatonin levels (see section 1.3). Since each stage was of
a random crossover design, any differences between standing and supine positions
were compared between stage la and b and between stage 2a and b. Area under the
curve (AUC) analysis followed by Student's paired t-tests (two-tailed) were used to
identify significant differences between postures. Paired comparisons were also made
at each 10 minute sampling point between standing and supine using Student's paired
t-tests (two-tailed). Correlations between plasma and salivary melatonin levels were
carried out using Pearson's correlation. All values lower than the detection limit of
the assay were excluded from the calculations. This corresponded to 5% of the
120
plasma melatonin data and 6% of the salivary melatonin data. All values are quoted
as mean ± SD unless stated otherwise.
3.3 RESULTS
All subjects completed stage 1. Four subjects completed stage 2 and one
further subject completed stage 2a and 60 minutes of stage 2b. No subject reported
any oral lesion and no blood contamination was observed in the saliva samples.
However, salivary melatonin was unmeasurable in the one subject who smoked,
through contamination or interference in the assay, and was discarded from the
analysis of salivary melatonin.
Figure 3.1 shows the percentage change in melatonin levels within 30 minute
periods in alternate postural positions from 2400h. Changing from standing to supine
and back to standing (stage la), and vice versa when the positions were reversed
(stage lb), caused marked differences in percentage change of both plasma and
salivary melatonin levels. AUC analysis identified significant differences between
standing and supine positions with greater increases associated with the standing
position (Figure 3.1). Significant changes between postures were observed within 10
minutes in all three time periods for plasma melatonin, Figure 3.1 A. Differences
in salivary melatonin levels between standing and supine positions were less
significant than with plasma, Figure 3.1 B. Table 3.1 shows the mean difference in
% change between the two postures.
Figure 3.2 shows the individual profiles for the percentage change in plasma
121
melatonin levels over a 90 minute period between stages 2a and b. Differences
between standing and supine positions were significant for plasma melatonin only
(p<0.01) with AUC analysis. The changes in salivary melatonin levels were more
variable and proved not to be significant, Figure 3.3.
Good correlation between plasma and salivary melatonin concentrations was
obtained during hourly sampling (r=0.83, p<0.0001, y= 0.59 [± 0.01]x -2 [±
1.2]) from 2100-2400h. Although still highly significant, the correlation between the
two melatonin mediums was not as good during short term sampling (r=0.555,
p<0.0001, y= 0.23 [± 0.02]x + 5.7 [± 0.9]). Salivary melatonin levels were
estimated to represent 46.2 ± 18.1 % and 40.4 ± 18.7% during hourly and 10 minute
sampling, respectively, of the corresponding plasma levels.
There were no significant differences in aMT6s levels between standing and
supine positions. After 90 minutes in the standing position, aMT6s levels were
calculated to be 103 ± 17% of the levels found after an equivalent period lying
recumbent.
122
Figure 3.1 Immediate effect of posture on plasma and salivary melatonin levels in alternate postural positions
A. 120r-
100
-20 '
-60J
" Standing 0 Supine
T
" Supine 1" Standing jO Standing j0 Supine
i
riiiII ýý
2400/0 10 20 30/0 10 20 30/0 10 20 30
Time (h/mins. )
B.
2400/0 10 20 30/0 10 20 30/0 10 20 30
Time (h/mins. )
Percentage change in A. plasma melatonin and B. salivary melatonin levels within 30 minutes in alternate postural positions from 2400h. Mean +/- SD, plasma: n=7, saliva: n=6. * p<0.05, ** p<0.01, *** p<0.001: significant differences between standing and supine positions (Stage la (S) versus stage lb (p), paired t-test, two-tailed).
123
Table 3.1 Mean (± SD) difference and range in % change in melatonin levels
between standing and supine positions.
Plasma
Stage 1
2400-0030h
0030-0100h
0100-0130h
35.2 ± 21.3 (n=7)'
range -7.5 to 78
33.5 ± 16.2 (n=7)"
range 5.6 to 85
36.7 ± 17.5 (n=7)"
range 7.8 to 71
Saliva
42 ± 27 (n=6)**
range -16.9-81
30.9 ± 34.4 (n=6)
range -23 to 113
31.9 ± 44.8 (n=6)
range -21 to 88
Stage 2
2400-0130h
2400-0100h
18.6 ± 14.8 (n=4)"
range -13 to 59
34.1 ± 41.7 (n=5)
range -13 to 132
25.7 ± 46.3 (n=4)
range -41.8 to 93.1
28.6 ± 43.9 (n=5)
range -41 to 93
'p<0.05, "p<0.01: significance determined from AUC analysis followed by paired
t-test.
124
Figure 3.2 Long term effects of posture on plasma melatonin levels
150
100
50
Upright , Standing/Supine
Subject 1
Subject 5
50
-100 A
300
200
100
0 -100- 200-
150-
100-
50-
A
-50-1 2200 0 10 20 30 40 50 60 70 80 90 2100 2300
Upright i Standing/Supine
250
150
-100 2200 0 1020 30 40 50 60 70 80 90 2100 2300
Time (h/mins. )
Time (h/mins. )
Individual profiles showing the percentage change in plasma melatonin levels after adopting a supine (0, stage 2a) or a standing (0, stage 2b) position for 90 minutes from 2400h. Significant difference (p<0.01) over 90 minutes between supine and standing positions (n=4, excluding subject five, AUC analysis, paired t-test). The changes in plasma melatonin levels reached significance (p<0.05, paired t-test, n=4) at 20,40,50,60 and 70 minutes between stages 2a and 2b.
125
Figure 3.3 Long term effects of posture on salivary melatonin levels
t
15
10
5
-5 "9 .9
5/-10
e 01 . - -5
-10
20/-15
15
10
5
0
-5
Upright ; Standing/Supine Subject 2
2200 010 20 30 40 506070 8090 2100 2300
Time (h/mins. )
Upright ; Standing/Supine
2200 010 20 30 40 50 6070 80 90 2100 2300
Time (h/mins. )
Individual profiles showing percentage change in salivary melatonin levels after adopting a supine ( , stage 2a) or a standing (0, stage 2b)
position from 2400h for 90 minutes. No significant differences between postures were seen.
126
3.4 DISCUSSION
When moving from a supine position to a standing position, blood volume
becomes redistributed from the central pool within the chest, to the lower limbs
where small molecules pass through the capillary membranes into the interstitial space
by diffusion due to the increased fluid pressure in the plasma. The plasma volume
is reduced, thus increasing the concentration of nonfilterable blood constituents such
as red cells, and proteins together with their bound components (Stoker and Wynn,
1966; Husdan et al, 1973; Tan et al, 1973; Hagan et al, 1978), including albumin-
bound hormones (Riad-Fahmy et al, 1982).
The results from this study indicate that changing from the standing to the
supine position caused a marked fall in plasma and salivary melatonin levels which
was reversed on returning to the standing position once again. When these positions
were reversed over the same period of time an opposite effect of similar magnitude
was seen. These differences were significant for plasma melatonin within 10 minutes
as calculated between stages la and lb and within 20 minutes comparing stages 2a
and 2b. These changes showed no reliable (p > 0.1) progressive change with time
indicating that the changes in plasma melatonin levels associated with the effects of
posture were rapid and remained constant after approximately 20 minutes. This
agrees well with other posture-related studies affecting various blood constituents
where little change occurred after 20 minutes (Hagan et al, 1978; Tan et ao.
The effects on salivary melatonin levels were more variable. This may be
associated with frequent and forced sampling (every 10 minutes) which probably
changes salivary composition and secretory mechanisms of the gland (Dawes, 1972),
127
modifying the true melatonin concentration in saliva. A high degree of correlation
between plasma and salivary melatonin levels was observed throughout hourly
sampling similar to that shown previously by other studies (Vakkuri, 1985; McIntyre
et al, 1987; Nowak et al, 1987; Miles et al, 1985,1987). Although the overall
proportion of melatonin levels in saliva to plasma, throughout hourly and 10 minute
sampling, remained comparable, the correlation fell considerably throughout 10
minute sampling. Only the linear relationship throughout hourly sampling
corresponded well with previous studies. For example, McIntyre et al (1987)
demonstrated a correlation of r=0.90 and the regression relationship as:
y (salivary melatonin) = 0.55 [± 0.05] x (plasma melatonin) - 2.6 [± 2.1]
(f SEM).
Differences in urinary aMT6s output associated with alterations in posture
might be expected to be delayed due to liver metabolism and renal excretion
processes. Therefore changes were likely to become apparent after 90 minutes.
However, there were no differences in urinary aMT6s output between standing and
supine positions, suggesting that the changes in melatonin concentrations observed in
plasma and saliva were likely to be fluid distribution effects rather than a change in
total melatonin output.
Abalan et al (1992) recently noted a difference of 32.6% ± 34.3 (mean f
SD, range: -20% to 91 %) in total plasma cortisol - another circadian phase marker -
on standing for 40 minutes compared to remaining supine, during the rising phase
of the cortisol rhythm. A similar difference in plasma and salivary melatonin levels
was observed in this study even though sampling was more frequent and over
different periods of time.
128
Although considerable interindividual variation was evident, these results
indicate that posture has a masking effect on plasma and salivary melatonin levels.
Changes due to posture may prove to be important when interpreting data from
subjects in different postural positions, for example, comparing bedridden patients and
healthy controls, and may have implications in situations where the body is exposed
to other environments where body fluid shifts are also experienced, such as
spaceflights, water immersion and exercise. Changes in melatonin and cortisol levels
due to posture may alter circadian parameters such as daily means and amplitudes and
influence acrophase values. Samel et al (1993) studied the effects on the circadian
system of 7 days of bed rest with a constant sleep/wake cycle and a regular
photoperiod. Although neither melatonin or cortisol were measured, small changes
in rhythm acrophases were evident for noradrenaline, adrenalin, potassium, heart rate
and urine volume and significant changes in rhythm amplitudes were observed. These
changes may be attributable to inactivity and posture.
This study indicates the necessity for standardising the timing of blood
sampling and indeed saliva sampling with prior knowledge of the postural position
and the time in that position. Care should also be taken in the interpretation of
results from high frequency saliva sampling.
129
CHAPTER 4
INDUCING CIRCADIAN RHYTHM DISTURBANCES IN
PARTIAL ENVIRONMENTAL ISOLATION: AN INITIAL
STUDY.
COMPARISON OF THE MELATONIN RHYTHM IN
PLASMA, URINE AND SALIVA BEFORE AND AFTER A
FORCED PHASE SHIFT
Inducing circadian rhythm disturbances in partial environmental isolation: an initial study
- Comparison of the melatonin rhythm in plasma, urine and saliva before and after a forced phase-shift
4.1 INTRODUCTION
Rapid time-zone change and shift work can lead to conflict between the new
external time cues (zeitgebers) and the internal biological horologue with consequent
physiological disturbances and deleterious effects on, for example, sleep and
performance (Seigel et al, 1969; Fevre-Montange et al, 1981; Winget et al, 1984;
Touitou et al, 1990). One treatment strategy for facilitating the adaptation of circadian
rhythms to rapid shifts in external time cues and alleviating the associated behavioural
problems, is the use of bright light (Daan and Lewy, 1984; Czeisler et al, 1990;
Eastman, 1992).
When suitably timed, bright light of sufficient intensity and duration will
induce phase-shifts in physiological variables such as core body temperature, cortisol
and melatonin secretion (Czeisler et al, 1986; Broadway et al, 1986,1987; Dijk et
al, 1987,1989; Lewy et al, 1987; Minors et al, 1991; Shanahan and Czeisler, 1991;
van Cauter et al, 1994), even when other zeitgebers, such as the sleep/wake cycle and
social cues, are kept constant (Czeisler et al, 1986; Lewy et al, 1987). When a light
pulse is used in combination with a shift in the sleep/wake schedule, far greater
phase-shifts can be achieved (Czeisler et al, 1989; Shanahan and Czeisler, 1991).
The magnitude and direction of the shift in melatonin and core body temperature
rhythms are similar when measured under constant routine (`unmasked') conditions
131
suggesting a common control mechanism (Shanahan and Czeisler, 1991).
The establishment of a phase-response curve for light has allowed predictions
of the magnitude and direction of the shift based upon the timing of the light pulse.
Exposure to bright light in the evening can delay circadian rhythms and morning light
exposure will produce advance shifts, with the greatest shifts achieved with light
exposure close to the temperature minimum (Minors et al, 1991, Czeisler et al, 1989,
section 1.4). Such a phase-response curve for light may be applied to treat circadian
rhythm disorders such as jet lag and maladaptation to shift work (Daan and Lewy,
1984; Czeisler et al 1990; Eastman, 1992). It may also provide a means of
simulating such disorders in a laboratory.
Field studies of shift workers and jet travellers are expensive and difficult to
control. One primary aim of this study was to devise a simple, rapid and inexpensive
method to study the adaptation of circadian rhythms, sleep, mood and performance,
during and after a phase-shift without complete environmental isolation. In this initial
study a combination of moderately bright light and subsequent imposed darkness (and
sleep) was used to induce a gradual and slow phase-shift of the circadian system in
the laboratory. Subjects then resumed their baseline sleep/wake cycle, in their normal
environment, inducing a rapid shift of external time cues similar to that experienced
by jet travellers or shift workers. The subsequent readaptation of the subjects in their
normal environment was then studied.
Under normal environmental conditions melatonin is less susceptible to
masking (disturbing) influences by external factors, other than light (Lewy et al,
1980), compared to other circadian rhythms such as core body temperature and
cortisol secretion. Melatonin was therefore employed as an appropriate tool for
132
measuring the phase of the internal body clock. The usefulness of self-rated alertness
as a behavioural circadian rhythm marker was assessed.
Melatonin has been variously measured and compared in plasma, saliva, and
its metabolite aMT6s in urine under normal environmental conditions (Arendt et al,
1985b; Markey et al, 1985; Vakkuri, 1985; Bojkowski et al, 1987b; McIntyre et al,
1987; Nowak et al, 1987). The onset of its secretion in dim light has been
extensively used by Lewy and co-workers as a circadian phase marker (for example,
Lewy and Sack, 1989). A second aim of this study was to assess the relative validity
of these various approaches in determining the circadian phase position under normal
and phase-shifted conditions.
4.2 SUBJECTS and METHODS
4.2.1 Subjects
6 healthy volunteers (3M: 3F), aged 22-26, under no medication and all non-
smokers, refrained from heavy exercise and caffeine-containing drinks for 12 hours
before sampling of body fluids and restricted their alcohol consumption (maximum
2 units per day) throughout the experiment. Using the Home and Osterberg
Morningness/Eveningness questionnaire (1976), four subjects were classified as
"moderately morning" types, one as an "intermediate" type and one as a "moderate
evening" type.
133
4.2.2 Study design
The study was performed in early December 1991. Figure 4.1 shows a
diagrammatic representation of the study.
DAYS (D)
-7 -6 -5 -4 -3 -2 -1 01234567
Baseline
Regular sleep/wake cycle and normal environment
Slow, forced Post-treatment: 6h delay Readaptation to a normal
phase-shift environment with baseline sleep/wake cycle.
Rapid 6h advance phase-shift
Figure 4.1 Diagrammatic design of the study.
Baseline: For one week prior to light treatment (days (D) -7 to 0), the subjects
were required to maintain a regular sleep/wake schedule, retiring to bed at 2330h and
waking at 0800h, remaining in the dark (< 1 lux) between these times.
Light/Dark Treatment: Immediately following the baseline week, the subjects
were exposed to a 6h pulse of moderately bright, full spectrum light (average 1200
lux at eye level) at progressively later times over 3 consecutive days. The light
treatment occurred during the following periods:
DI: 2000-0200h, D2: 2200-0400h, D3: 2400-0600h .
134
Each period of light treatment was followed by 8 hours of imposed darkness
(< 1 lux). During the remaining hours of each day strict light control was not
imposed until dusk (1630h), but the subjects were asked to refrain from periods of
exposure to bright sunlight. From dusk until the light treatment period, subjects
remained in domestic intensity light (150-300 lux at eye level). Throughout the light
treatment (D1-3) meal-times were shifted progressively by 2 hours each day in line
with the imposed light/dark and sleep/wake cycle. Subjects were allowed to choose
their activity (e. g. reading, writing, watching videos) as long as they glanced
periodically (at least every 5 minutes) at the light source. They were continuously
supervised throughout these periods.
Post-Light/Dark Treatment: Immediately following the last light/dark treatment (i. e
after 1400h on D4), subjects were required to assume the baseline sleep/wake
schedule, i. e. retiring to bed at 2330h and waking at 0800h for 4 days (D4-8). The
subsequent readaptation of the subjects in their normal environment was then
observed. Subjects were only allowed to sleep during the periods of darkness
throughout this study. Naps were not allowed. In addition, subjects were asked to
avoid bright sunlight.
4.2.3 Sampling
Blood, urine and saliva samples were taken on the last baseline day (DO), the
first post-light treatment day (D4) and the fourth post-treatment day (D7). All blood
samples obtained throughout the dark-phase were taken in the recumbent position and
135
daytime blood and saliva samples after sitting for at least 10 minutes. On all
sampling days subjects remained in the clinical investigation unit and during waking
hours the light intensity ranged from 100 to 300 lux.
Blood: An indwelling cannula was inserted into the forearm of each subject,
under local anaesthetic, between 1630h and 1700h. Blood samples (5 ml) were
obtained hourly from 1700h for 24 hours on each sampling day. Sampling during
darkness was performed in dim red light (< 1 lux). A total of 396 ml of blood was
removed from each person.
Saliva: Samples (2-3ml) were obtained hourly from 1700h for 24 hours
(following 2 minutes of chewing on parafilm), except during sleep, on the sampling
days. Meals and drinks were taken just after saliva sampling and a thorough
mouthwash taken afterwards to reduce any possible interference with salivary
melatonin estimation.
Urine: Sequential 4-hourly urine samples (8-hourly when asleep) were
collected from 1600h throughout 24 hours on DO, 4 and 7. Each subject voided at
1600h on each sampling day. All urine was then collected throughout each four hour
period and the bladder completely emptied at the end of that collection period. The
volume of each sample was noted and a5 ml aliquot from each sample collected.
For information on preparation and storage of the samples, see section 2.2.
136
4.2.4 Behavioural Measurements
During waking hours throughout the entire experimental period, subjective
alertness was rated and performance tasks were carried out every 2 hours. Subjective
alertness was rated on a 10cm visual analogue scale (VAS, " Drowsy a-. Alert ") and
high and low memory capacity performance tasks were used (1- and 5-letter search
and memory [SAM] tasks, respectively). Subjects were required to practice the tests
for at least three days before the study. In addition, daily sleep logs were kept.
Section 2.7 describes these behavioural measurements in detail.
4.2.5 Hormone assays
Methods of measuring melatonin and aMT6s are described in section 2.3. The
interassay coefficients of variation for the various assays were as follows:
Plasma melatonin: 10.4%, 15.2% and 9% at 22,40 and 210 pg/ml (n=10).
Salivary melatonin: 15.3 %, 11.7 % and 5.5 % at 24,40 and 67 pg/ml (n =6).
Urinary aMT6s: 16.8%, 11.6% and 5.8% at 2.4,14.7 and 31 ng/ml (n=6).
The minimum level of detection for each assay was 6 pg/ml, 3.2 pg/ml and 0.3
ng/ml, respectively.
137
4.3 STATISTICAL ANALYSIS
4.3.1 Hormonal and alertness rhythms
The raw hormonal and alertness data were analysed by two-factor ANOVA
for repeated measures (day of study and time as factors). Subsequently, a cosinor
analysis program was used to derive the rhythm characteristics of both the alertness
and hormonal data (see section 2.8). ANOVA followed by Duncan's Multiple Range
Test, were applied to the cosinor-derived parameters of acrophase and amplitude. All
data below the limit of detection of the assay were excluded from the analysis, except
for the purposes of the cosinor program where values lower than this limit were
assigned a value of half the minimum detection limit of the assay.
The onset and offset of plasma melatonin secretion were defined as twice the
minimum detection limit of the assay when two consecutive values either exceeded
or were below this amount, respectively. The relationships between concentrations
of melatonin in saliva and plasma, and aMT6s in urine, and between their respective
acrophases and plasma melatonin onset time were determined by linear regression
analysis and Pearson's Correlation.
4.3.2 Sleep data, and daily means for performance and alertness
Daily means for alertness and performance (accuracy, response time and
efficiency) were determined for each individual and calculated as a percentage of
individuals' own mean baseline. One factor ANOVA (day of study) followed by
Duncan's multiple range test were applied to the sleep data and the daily means for
alertness and performance (accuracy, response time and efficiency). Performance
138
accuracy, response time and efficiency were defined as described in section 2.7.3.
Values quoted are mean ± SEM unless stated otherwise.
4.4 RESULTS
No subjects reported any oral lesion or discomfort throughout the study and
no saliva sample appeared to be contaminated with blood. On 4 of the baseline days
(D-6 to -3) one subject suffered from flu-like symptoms which subsequently resulted
in reduced alertness and performance, and disrupted his sleep pattern. The data
obtained on these days were excluded from the analysis.
4.4.1 Melatonin
The profiles of plasma melatonin, salivary melatonin and urinary aMT6s are
shown in Figures 4.2 to 4.4, respectively.
Inspection of the aMT6s data indicated a marked phase delay on the first post-
treatment day (D4) in 5 out of 6 subjects. One subject showed a 3-fold increase in
the amplitude of aMT6s on D4 with no phase change and, in contrast to all the other
subjects, a poor correlation with plasma and salivary melatonin. This subject was
consequently excluded from the analysis of aMT6s.
A significant time-dependent effect of treatment (ANOVA: phase-delay) was
shown by aMT6s (p < 0.01) and plasma melatonin (p < 0.05) only, on the first post-
treatment day (D4), which was no longer present on the fourth post-treatment day
(D7) (p>0.1).
139
Cosinor analysis gave most significant fits for plasma melatonin (17 out of 18
profiles at p<0.001), followed by salivary melatonin (14 out of 18 at p<0.05) and
urinary aMT6s (10 out of 15 at p<0.1) reflecting the missing data during sleep for
both salivary melatonin and aMT6s. However, the percentage of the variance
accounted for by a cosine curve was greater than 81 % in the aMT6s profiles.
Using the acrophase data derived from the cosinor programme, a significant
phase delay (ANOVA) was present in all measures of melatonin secretion on the first
post-treatment day. The mean peak time was shifted by 2.67 ± 0.3h (n=5) for
aMT6s (p < 0.001), 2.35 ± 0.29h (n = 6) for plasma melatonin (p < 0.001) and 1.97
± 0.32h (n=6) for salivary melatonin (p<0.01), 1 day post-treatment, Figure 4.5.
Plasma melatonin onset also showed a significant delay on D4 (3.12 ± 0.74h, n =6,
p<0.001). These shifts did not significantly differ from each other (p>0.1). The
acrophase estimates and melatonin onset time for the fourth post-treatment day (D7)
were not significantly different from baseline (p>0.1, ANOVA), but were
significantly different from D4 (p < 0.05, ANOVA), indicating reentrainment by D7.
The plasma melatonin offset time shifted to a lesser degree (1.33 ± 0.35h, n=6)
compared to its onset time.
Table 4.1 shows the correlation of acrophases between plasma and salivary
melatonin and urinary aMT6s, together with plasma melatonin onset times.
There was a direct correlation between plasma and salivary melatonin in
samples taken simultaneously (r=0.758, p<0.001) before the light treatment,
immediately after the phase-shift (r=0.792, p<0.001) and on D7 (r=0.576,
p<0.01). The difference between plasma and salivary melatonin levels remained
constant throughout the study. Salivary melatonin levels were estimated to represent
140
a mean ± SD of 40 ± 6.9 % of the corresponding plasma levels throughout the
baseline day (DO), 44.1 ± 4.9% over the first post-treatment day and 37.4 ± 4.6%
on the fourth post-treatment day, to give an overall mean ± SD value of 40.4 f
5.3%.
The amplitude of the individual urinary aMT6s rhythms correlated well with
the amplitude of the plasma melatonin rhythm (r=0.73, p<0.001). In addition, the
total output of urinary aMT6s in 24h significantly correlated with the total 24h
melatonin concentration in plasma, as assessed by area under the curve analysis
(r=0.71, p<0.001).
The degree of shift for aMT6s on D4 was inversely correlated (P<0.05) to
the pre-treatment amplitude of the aMT6s rhythm, such that the greater the amplitude
the smaller the phase-shift. No such correlations were found to be statistically
reliable with plasma or salivary melatonin rhythms. No significant change in the
rhythm amplitude was apparent after the phase-shift with any of the measures of
melatonin secretion.
141
Figure 4.2 Circadian plasma melatonin profiles before and after a forced phase-shift
IIIIIIIIIIIIIIIIIIIIII 1800 2200 0200 0600 1000 1400
Time (hour of day)
I
Circadian profiles for plasma melatonin showing (a) phase-delay on the first post-treatment day (D4, ) compared to baseline (DO, O) and (b) readaptation by the fourth post-treatment day (D7, A ); no significant difference from baseline (DO). Darkness from 2330 to 0800h. Phase-delay indicated by significant time-dependent treatment effect (* p<0.05; two-way ANOVA for repeated measures). Values represent mean +/- SEM (n=6).
142
Figure 4.3 Urinary aMT6s circadian profile before and after a forced phase-shift
1600- 2000
2000- 2330- 0800- 1200- 2330 0800 1200 1600
Time (hour of day)
Circadian profiles for urinary aMT6s showing (a) phase-delay on the first post-treatment day (D4, ) compared to baseline (DO, Q) and (b) readaptation by the fourth post-treatment day (D7,0) - no difference from baseline (DO). Darkness from 2330 to 0800h. Phase-delay indicated by significant time-dependent treatment effect (** p<0.01; two-factor ANOVA for repeated measures). Values represent mean +/- SEM (n=5).
143
Figure 4.4 Circadian melatonin profile in saliva before and after a forced phase-shift
a) 20ý
15ý
1o
5.
OJ
0 1800 2200 0200 0600 1000 1400
Time (hour of day)
Circadian profiles for salivary melatonin, without night-time sampling, showing (a) phase-delay in the offset of secretion on the first post-treatment day (D4, ) compared to baseline (DO, D) and (b) readaptation by the fourth post-treatment day (D7, f). Darkness from 2330 to 0800h. Values represent mean +/- SEM (n=6).
144
Figure 4.5 Relationship between acrophase estimates for melatonin, aMT6s, alertness and melatonin onset time
I T
4 T
0 Day of study
2000
1800 ý
1600 ý
1400
0800
0600
0400
***
7
Acrophase estimates for urinary aMT6s (0), plasma MT (o), salivary MT (A), plasma MT onset ( ) and alertness (4) on baseline (day 0), one day post-treatment (day 4) and four days post-treatment (day 7). Mean +/- SEM (n=5 for aMT6s, n=6 for other variables). ** p<0.01, *** p<0.001: significant difference from baseline (ANOVA). D7 was significantly different from D4 (p<0.05) but not from DO for all variables (Duncan's multiple range).
0200 ý
2400 ý
2200 ý
145
Table 4.1 Correlation and linear regression analysis of acrophase estimates between
plasma melatonin, salivary melatonin and urinary aMT6s, together with the time of
plasma melatonin onset. y= mean (± SEM) x+ mean (± SEM).
Urinary aMT6s acrophase = 0.82 (± 0.18) plasma melatonin acrophase + 1.45 (± 0.93) r=0.754, p<0.001, n=15
Plasma melatonin acrophase = 0.60 (t 0.2) salivary melatonin acrophase + 2.35 (t 0.93)
r=0.6, p<0.01, n=18
Urinary aMT6s acrophase = 0.78 (± 0.19) salivary melatonin acrophase + 2.09 (± 0.78) r=0.718, p<0.001, n=18
Plasma melatonin acrophase = 0.62 (f 0.1) plasma melatonin onset time + 5.06 (t 0.23)
r=0.849, p<0.001, n=18
Urinary aMT6s acrophase = 0.7 (± 0.14) plasma melatonin onset time + 5.6 (± 0.23)
r=0.69, p<0.01, n=15
Salivary melatonin acrophase = 0.39 (f 0.15) plasma melatonin onset time + 4.55 (f 0.24)
r=0.53, p<0.05, n=18
146
4.4.2 Alertness
Using the cosinor-derived acrophase parameters (83% significant fits at
p<0.05), no significant difference in the timing of the alertness acrophase was
observed throughout the baseline days (mean acrophase 14.84 ± 0.28h), Figure 4.6.
A maximum delay phase-shift of 6.33 ± 1.3h (p<0.01) was evident on the last
treatment day (D3). By D4, there was still a clear phase delay in the rhythm (shift:
3.08 ± 0.67h, p<0.01, n=6, ANOVA; 1 day post-treatment compared to DO). This
effect is shown in Figure 4.6 and 4.7. A phase delay was also indicated with
ANOVA on the raw data (p<0.01), figure 4.7. The alertness rhythm reentrained by
the second post-treatment day (D5, Figure 4.6). No reliable difference in amplitude
was apparent although the daily mean alertness fell significantly during the first day
of the light treatment (p<0.01, Duncan's Multiple Range). The degree of shift in
the alertness rhythm on D4 (compared to DO) was not significantly different from the
shift in urinary aMT6s, plasma melatonin, salivary melatonin and the plasma
melatonin onset time (p>0.1), Figure 4.5.
4.4.3 Performance
No change in performance was observed over the week prior to the treatment
(p > 0.1), but a deterioration was observed during and after treatment, Figure 4.8.
SAM-5 accuracy scores fell earlier than SAM-1 (first [D1] versus third [D3]
treatment days, respectively) and returned to baseline values earlier than SAM-1 (D6
versus D7, respectively, after returning to the normal environment). Response times
were not significantly affected although a slight decrease was apparent over the
treatment period. Efficiency only suffered significantly for both tasks on the first day
147
after the rapid advance of external time cues (D5; p<0.01).
4.4.4 Sleep
No significant changes in any sleep parameters were apparent over the week
prior to treatment, Figure 4.9. The quality of sleep deteriorated on the first night (by
30%, p<0.05) and third night (by - 40%, p<0.01) of the treatment period.
Sleep duration decreased (by : lh, p<0.05), sleep latency increased (by - 50
minutes, p<0.05) and sleep quality fell significantly (by = 60%, p<0.05) on the
first post-treatment night (N4, i. e. first night after the rapid 6h advance shift of
external time cues), Figure 4.9.
148
Figure 4.6 Alertness daily means and acrophase estimates before, during and after a forced phase-shift
24: 00-
22: 00-
20: 00-
18: 00-
16: 00-
14: 00 ý
12: 00
a)
T---ý---l i 1 I I I -6 -4 -2 0246
Day
4 4 f Alertness (a) acrophase estimates and (b) daily means throughout an 8-day baseline (D-7 to DO), a 3-day delay phase-shift period (D1-3) and after returning to normal baseline sleep/wake cycle and environment on D4. A
significant delay phase-shift was apparent on D3 and D4 with reentrainment by D5. Daily mean alertness was only significantly affected on D1. * p<0.05, **
p<0.01: significant difference from baseline values (Duncan's multiple range test). Values represent mean +/- SEM
, n=6 (D-6 to D-3, n=5). 4 Blood
sampling days.
149
Figure 4.7 Subjective alertness profiles before and after a forced phase-shift
130-
120-
110-
100-
90-
80-
70-
60-
50-
a)
24 4 t""'`T- T-T ý
8 12 16 20 24
50 24 48 12 16 20 24
130-
120- b)
110- 100-
90-
80-
70-
60-
Time (hour of day)
Subjective alertness profiles showing (a) phase-delay on the first post-treatment day (D4, ) compared to baseline (DO, Q) and (b) readaptation by the fourth post- treatment day (D7, A); no significant difference from baseline. Darkness from 2330h to 0800h on DO and D7 (M) and 0600 tol400h and 2330 to 0800h on D4 (EM). Phase-delay indicated by significant time-dependent treatment effect (* * p<0.01, two-way ANOVA for repeated measures). Values represent mean
+/- SEM (n=6).
150
Figure 4.8 Performance measures before, during and after a forced phase-shift
110-
105 ý
loo-
95 -
90-
85-ý
80
110-
loo-
90 -
80 -
70
*
-6 -4 -2 0246
-6 -4 -2 024 I '1
6 1I1
Day of study
Daily mean performance measures for (a) accuracy and (b) efficiency showing high- and low-memory load capacity tasks (SAM-5, , and SAM-1, ", respectively). No significant variation over the 8-day pre-treatment period (D- 7 to DO), significant fall in SAM-5 accuracy over the slow delay phase-shift (D1-3), and significant decrements in all performance parameters on the first day after returning to normal baseline sleep/wake schedule (D5). * p<0.05: significant difference from baseline (Duncan's multiple range test).
+ Blood
sampling days. Values represent mean +/- SEM, n=6 (D-6 to D-3, n=5).
151
Figure 4.9 Subjective sleep parameters before, during and after a forced phase-shift
130 H
110
90 `..
70 ** v,
50 **
ý 30
-8 -6 -4 -2 0246 100]
b) y
80
60
a)
40 -1
20 ý
Llý _8
ý J, J- ý 0
9.5 H
8.5 H
7.5 ý
6.5 -1
5.5
C)
0 ýrl lll -4 -2 02
III III
46
-8 -6 -4 -2 0246
Night of study
Subjective measures of sleep for (a) sleep quality (b) sleep latency and (c) sleep duration showing no significant variation over the 9-night baseline period (N-8 to NO), fall in sleep quality over a slow delay forced phase-shift (N 1-3), and significant deterioration on the first night (N4) after the rapid advance phase-shift on D4 for all parameters. * p<0.05, ** p<0.01: significant difference from baseline (Duncan's multiple range test). 4 Blood sampling nights. Values represent mean +/- SEM, n=6 (D-6 to D-3, n=5).
152
4.5 DISCUSSION
One objective of this study was to assess the usefulness of non-invasive
methods of determining melatonin secretion in the form of salivary melatonin and
urinary aMT6s before and after a phase-shift. The close relationship between plasma
melatonin onset as a phase marker, and the calculated peak times of melatonin in
plasma and saliva, and aMT6s in urine, suggests that all these approaches give valid
information. A high degree of correlation between the onset time and acrophase for
melatonin in plasma has also been determined in a recent study using shift workers
and blind subjects (Blood et al, 1994). During the light treatment it was not possible
to assess correctly the phase position of circadian rhythms due to the direct, acute
effects ("masking") of light. For example, with sufficiently bright light melatonin is
suppressed and core body temperature is increased (Lewy et al, 1980; Badia et al,
1993; Myers and Badia, 1993). For this reason, the immediate post-light treatment
period provides the nearest estimate to the actual phase-shift induced. During this
period (D4), during the baseline (DO), and after readaptation (D7) the melatonin onset
and mean peak values occurred in conditions of dim light or darkness. In such
conditions, it is assumed that the measured phase position of melatonin represented
the actual phase position on those days, and that the magnitude of the actual induced
shift was greater than or equal to the measured shift on D4.
In order to minimise sleep disturbance which may have compromised
behavioural measurements, saliva and urine sampling were avoided during the sleep-
phase. This resulted in the loss of definition in the circadian profiles, particularly for
salivary melatonin. Certainly the estimates of the melatonin rhythm amplitude in
153
saliva would not be reliable. In spite of this problem a comparable phase-shift in the
salivary melatonin rhythm was obtained. Clearly, however, the saliva data would
greatly have been enhanced by night-time sampling. The salivary melatonin onset
time was not a useful index of the rhythm in this study since a high proportion (67%)
of onset times occurred during the sleep phase after the phase-shift.
In this present study the ratio of salivary melatonin levels to plasma melatonin
levels remained constant (approximately 40%) and the correlation remained highly
significant even immediately after the rapid phase-shift. Even though the number of
samples with detectable melatonin levels in saliva were reduced through no night-time
sampling, the ratio agrees well with other reports where samples were obtained
throughout 24h in a normal environment. In these studies, the ratio of melatonin in
saliva to plasma was found to represent 32% (Nowak et al, 1987), 30-60% (Vakkuri,
1985), 27-32% (Miles et al, 1987), and an estimated 55% (McIntyre et al, 1987).
McIntyre et al (1987) also showed a good correlation (r=0.92, p<0.005) when
plasma melatonin and salivary melatonin levels were suppressed with one hour of
bright light between 2400h and 0100h. Nickelson et al (1991a) suggested that the
relationship between salivary melatonin and urinary aMT6s concentrations may
change following a9 hour forced phase-shift. There is no evidence for this
phenomenon in this present study. However, the induced shift was much smaller,
and more frequent saliva collections were made, at least during waking hours.
A high degree of correlation (r=0.73) was observed between the rhythm
amplitudes for plasma melatonin and urinary aMT6s and agrees well with Bojkowski
et al (1987b) who demonstrated a correlation of r=0.70. In addition, a significant
relationship between the individual total 24h plasma melatonin and urinary aMT6s
154
concentrations was observed (r=0.71), which has been established by previous studies
under normal conditions (r=0.76: Markey et al, 1985; r=0.75: Bojkowski et al,
1987b; r=0.72: Nowak et al, 1987), suggesting that urinary aMT6s is a reliable
index of plasma melatonin levels even after an imposed shift in the circadian system.
The practical advantage of urinary aMT6s is that it provides an integrated
measure for melatonin production over the night-time period and allows repeated and
longitudinal studies in the same individual to be performed. However, it is not
possible to measure subtle changes in the rhythms due to limited frequency of
sampling. The delayed aMT6s acrophase compared to that of plasma melatonin is
presumably due to the delay in hepatic metabolism of melatonin and the time taken
for the passage and diffusion of the metabolite through the kidney before appearing
in the urine.
A clear phase-delay in the alertness rhythm was evident, although alertness
was clearly influenced by the sleep/wake cycle, since it was only rated during
wakefulness. However, it is clear from the alertness data on D4 that the fall of the
rhythm occurred significantly later in the evening compared to baseline suggesting a
delay in the endogenous alertness rhythm. In addition, the degree of shift was
comparable to that of melatonin which is not affected by the sleep/wake state (Folkard
et al, 1984; Morris et al, 1990). This may imply a causal relationship as suitably
timed melatonin administration can modify alertness (Arendt et al, 1987, see chapters
9 and 10) and advance the timing of the self-rated fatigue and the closely associated
sleep propensity rhythm (Arendt er al, 1984; Tzischinsky et al, 1992a). However,
in this study it is more likely that the treatment induced a shift in central rhythm
generating mechanisms common to both variables.
155
A significant decline in daily mean alertness was apparent only on the first
treatment day. It is likely that this was due to blood sampling the previous night and
during the day. Subjective sleep quality and SAM-5 accuracy did fall during this
night and day but not significantly. However, disturbances in sleep, alertness and
performance were not associated with all blood sampling periods, hence it is difficult
to attribute changes in behaviour to blood sampling. Perhaps the novelty of
cannulation and sleep in a new environment compromised behaviour on this day.
Disturbances in sleep quality over the treatment period (during the slow delay
phase-shift) were observed. There were also important and significant decrements in
performance accuracy during this period with time-dependent differences in high and
low memory capacity tasks. These observations suggest that even a brief shift in the
sleep/wake cycle may compromise sleep and performance.
The light/dark treatment schedule was chosen to enhance the natural tendency
of circadian rhythms in the majority of people to delay. Since humans have a free-
running period of about 25h, the natural tendency is to delay by approximately
lh/day. Wever et al (1983), Wever (1986) and Aschoff and Wever (1981) have
demonstrated that physiological rhythms can entrain to a 26h day even without bright
light treatment. However, bright light can enhance the entrainment to longer days
(Wever et al, 1983; Wever, 1989). In the present study, using a 26h sleep/wake
schedule over 3 days, with bedtime and wake time 2h later each day, the evening
light treatment was aimed to hit the delay portion of the phase-response curve each
day. It is possible that the light treatment was not of sufficient intensity or duration
or the light suitably timed to promote the adaptation of sleep and performance to the
new sleep/wake cycle. However, sleep quality and performance accuracy were the
156
only sleep and performance parameters to be adversely affected. Eastman and
Miescke (1990) found that in a similar 26h sleep/wake schedule, bright light (1770-
2800 lux) administered for 2h before bedtime facilitated the entrainment of the
temperature rhythm in 74% of the subjects and promoted better sleep. However,
shielding from natural bright light for 6h after sleep was necessary for this effect.
Only 17% of the subjects entrained with natural light exposure at any time of day.
Strict light control was not imposed until dusk in the present study and may help to
explain the disturbances in sleep and performance over the imposed 26h sleep/wake
schedule due to the effect of inappropriate light exposure on the circadian system.
These behavioural disturbances may also have been confounded by the effects of
blood sampling and the novel sleep environment.
Individual differences may also account for the behavioural disturbances
following adaptation to an extended day. The majority of subjects in this study were
moderately "morning types". There is a suggestion that "evening type" individuals
adapt quicker to imposed shifts in the sleep/wake cycle (Folkard and Monk, 1981;
Hildebrandt and Stratmann, 1979).
The major and consistent deleterious effects on sleep and performance
occurred when the subjects were rapidly returned to their normal baseline sleep/wake
cycle and environment on D4. A transient increase in latency, fall in quality and
duration were observed on the first night after the rapid advance phase-shift, together
with decrements in performance accuracy and efficiency the following day. The
disturbances in sleep and performance (possibly as a consequence of disrupted sleep)
can be attributed to the inappropriate phase relationship between the endogenous sleep
propensity rhythm and the imposed sleep/wake cycle (Dijk and Czeisler, 1994), i. e.
157
the subjects were trying to sleep at a time when the circadian pacemaker was
promoting wakefulness. As a result sleep deprivation probably made it even more
difficult to perform well the following day (Carskadon and Dement, 1979,1981).
In this initial study, the use of a three-cycle stimulus of moderate intensity
bright light (1200 lux), suitably timed and coupled with periods of imposed darkness
(< 1 lux) and sleep, clearly induces significant phase-shifts in a major endogenous
hormonal circadian rhythm (melatonin) and a major behavioural rhythm (self-rated
alertness). The subsequent readaptation of the subjects in their normal environment
produced transient but major behavioural disturbances. It should therefore be possible
to simulate circadian rhythm disorders and related problems such as jet lag and
shiftwork, similar to those seen in field studies, in an inexpensive and controlled
environment where the subjects are left in their normal environment for most of the
time. Clearly since the induced phase-shift was small and the behavioural problems
transient on readaptation, larger phase-shifts are necessary in order to study the
efficacy of treatment strategies in treating circadian rhythm disorders.
158
CHAPTER 5
INDUCING CIRCADIAN DISTURBANCES IN PARTIAL
ENVIRONMENTAL ISOLATION: A RAPID 9h ADVANCE
PHASE-SH HFT AFTER A GRADUAL 9h DELAY SHIN
Inducing Circadian Rhythm Disturbances in Partial Environmental Isolation: A Rapid 9h Advance Phase-shift After a Gradual 9h Delay Shift.
5.1 INTRODUCTION
Field studies on phase-shifting and other strategies to alleviate the problems
associated with jet lag and shift work are expensive and difficult to control.
Laboratory simulations provide a more controlled environment with which to study
these circadian rhythm disorders in detail. However, one disadvantage of laboratory
simulations is that the subjects are completely isolated from a normal environment,
whereas field studies allow the subject to be exposed to a whole range of time cues
which may be important in the normal adaptation process in everyday life. In order
to simulate conditions as true to life as possible, an experimental model was
developed without complete environmental isolation allowing subjects to adapt to
rapid shifts in local time in their normal environment.
This experiment was a crossover study which investigated the effects of bright
light and melatonin in facilitating the readaptation of the circadian system and
alleviating the associated behavioural problems after a rapid 9h advance shift in local
time. Comparison of these treatment strategies will be discussed in Chapter 6. It is
the purpose of this chapter to describe in detail an experimental model which
gradually forces a 9h delay phase-shift in the circadian system under controlled
environmental conditions and allows a rapid 9h advance shift of external time cues
to be induced when the subjects resume their baseline sleep/wake schedule. The
readaptation of the subjects to their normal environment will be described (under
placebo treatment). A combination of exposure to bright light and darkness was
160
timed to gradually delay circadian rhythms over a number of days in a similar way
to that described in Chapter 4.
5.2 SUBJECTS and METHODS
5.2.1 Subjects
8 healthy volunteers (4M: 4F), aged 22-26, under no medication and all non-
smokers were recruited from research students of the University of Surrey who
followed similar sleep/wake schedules and lifestyles. Using the Horne and Östherg
questionnaire (1976) to determine morning or evening type individuals, no definite
morning or evening types were detected (scores > 70 or < 31). However, two
subjects were more of the morning type (scores of 63 and 64) and three were rather
of the evening type (scores of 36,41 and 41)
The subjects refrained from heavy exercise and restricted their alcohol
consumption (maximum 2 units per day) throughout the experiment.
5.2.2 Study Design
This experiment was a double-blind, placebo-controlled crossover study with
four different treatments administered after the rapid 9h advance phase-shift. The 4-
stage study was performed from November 1992 and April 1993. At least two weeks
separated the start of each stage. Each stage began on the same day of the week and
spanned the first two weeks of the month.
161
Each stage comprised the following:
DAYS (D)
-2 -1 0123456789 10 11
(Changing ; Stable Baseline Slow, forced
Regular sleep/wake 9h delay cycle and normal phase-shift environment T
Readaptation in normal environment with baseline sleep/wake cycle.
Rapid 9h advance phase-shift
Baseline: For 3 days prior to the phase-shift (days (D) -2 to 0), the subjects were
required to maintain a regular sleep/wake schedule, retiring to bed at approximately
2330h and waking at 0800h, remaining in the dark (< 1 lux) between these times.
Gradual 9h delay phase-shift: Immediately following the baseline days, the
subjects were exposed to 9h pulses of moderately bright (average 1200 lux at eye
level), full spectrum light at progressively 3h later times over the first three days (D1-
3) and then at the same times on D4-5 as for D3. The periods of light exposure were
as follows:
D 1: 1800-0300h, D2: 2100-0600h, D3,4,5: 2400-0900h.
Each period of light treatment was followed by 8 hours of imposed darkness
(< 1 lux) when they slept. During the remaining hours of each day subjects were
in dim light (< 300 lux, indoors) or moderately bright light (< 1000 lux, outdoors
162
with sunglasses) until dusk (1630/1700h). From dusk until the light treatment period,
subjects remained in domestic intensity light (50-300 lux at eye level). Throughout
the phase-shift (D1-5) meal-times were shifted progressively by 3 hours on days 1-3
and then maintained at the same times on D4-5 as for D3, in line with the imposed
light/dark and sleep/wake cycles. Subjects were allowed to choose their activity (e. g.
reading, writing, watching videos) as long as they glanced regularly (at least every
5 minutes) at the light source. The subjects were continuously supervised throughout
these periods. The patterns of exposure to bright light, natural ambient light and
darkness are summarized in Figure 5.1.
9h rapid advance phase-shift: Immediately following the last light/dark
treatment on day 6, subjects were required to assume the baseline sleep/wake
schedule in their normal environment, inducing a rapid 9h advance phase-shift in
external time cues, i. e. retiring to bed at 2330h and waking at 0800h for 6 days (D6-
11). The subsequent readaptation of the subjects in their normal environment was
then observed. Subjects took a placebo pill (lactose-gelatin capsule) just before the
desired bedtime at 2300h on D6-8. In addition, from 0800-1200h on D7 and D8
subjects were exposed to controlled lighting of between 50-300 lux. Subjects were
asked to avoid bright sunlight and wore sunglasses when outdoors (filtering out
>90% of the light). At dusk, they were required to remain in ordinary domestic
lighting (50-300 lux) until bedtime (and darkness). Each subjects ambient lighting at
work and home was monitored using a digital light meter (see section 2.4).
Subjects were only allowed to sleep during the periods of darkness throughout
this study. Naps were forbidden.
163
Figure 5.1 Diagrammatic design of the study
-2 -1
0 II 2 3 4 5 6 7
8
9
10
11
I I I
I
24: 00 04: 00 08: 00 12: 00 16: 00 20: 00 24: 00 Time of day
Diagrammatic design of the study showing patterns of exposure to bright light (O
,- 1200 lux), darkness/sleep (m) and natural ambient light (0). After a3 day baseline (D-2 to DO), a gradual 9h delay phase-shift was induced
over 3 days (D 1-3) and then the 9h shift was maintained for the following 2 days (D4-5). On day 6, subjects resumed their baseline sleep/wake schedule in a normal environment, thus experiencing a rapid 9h advance phase-shift of local time cues.
164
5.2.3 Sampling
Urine: Sequential 4-hourly urine samples (8-hourly when asleep) were collected
throughout the study. All urine was collected throughout each four hour period and
the bladder completely emptied at the end of that collection period. See section 2.2
for sample collection and storage.
Temperature: Rectal temperature was recorded continuously every 6 minutes
throughout the trial (see section 2.5).
5.2.4 Behavioural Measurements
Every 2h throughout the whole trial during waking hours, mood parameters
(alertness, cheerfulness and calmness) were rated on 10cm visual analogue scales
(VAS: Drowsy *-º Alert, Depressed Happy and Tense*--, - Calm, respectively)
and performance tasks (low- and high-load search and memory tests: SAM-1 and
SAM-5, respectively) were carried out every. In addition, daily sleep logs and
activity diaries were kept. For 3-4 days prior to the trial subjects practised the
performance tests regularly. Behavioural measurements were recorded using Psion
organisers. Sections 2.7 describes these behavioural measurements in detail.
5.2.5 aMT6s assays
For methods of analysis of urinary aMT6s, see section 2.3.4. The interassay
coefficients of variation were 7,4 and 3.3% at 2.9,22.3 and 42.7 ng/ml (n=37).
165
The minimum level of detection was 0.4 ± 0.22 ng/ml (n=37).
5.3 STATISTICAL ANALYSIS
The 4 legs of the trial involved 4 different treatments administered after the
rapid 9h advance phase-shift. During the baseline and gradual 9h delay phase-shift,
these legs will be referred to as the pre-treatment groups.
5.3.1 Temperature and aMT6s rhythms
Repeated measures ANOVA (factors: pre-treatment group, day of study and
time of day) followed by post-hoc analysis (Duncan's Multiple Range Test) were
applied to raw aMT6s and temperature data, and the cosinor-derived parameters of
acrophase and amplitude for 24h temperature and aMT6s profiles over the baseline,
gradual phase-delay shift and readaptation periods (under placebo treatment). For
analysis by ANOVA, the temperature data were averaged over each hour. Degrees
of shift were determined by comparison with the individual's mean baseline.
5.3.2 Mood and performance rhythms
24h profiles for each mood and performance parameter were analysed by
cosinor analysis over 3-day windows for more reliable estimates of acrophase and
amplitude (baseline: D-2 to DO, changing phase-shift: D1-3, stable phase-shift: D3-5,
readaptation period: D7-9 and D9-11) followed by ANOVA (for repeated measures).
166
An estimate of entrainment to the 24h day (and establishment of cosinor rhythmicity
with a period of 24h) is provided by (1) the percentage number of subjects with
significant 24h cosine fits and (2) the percentage variance accounted for by a cosine
curve, see section 2.8. Where the mood and performance parameters exhibited
significant cosinor rhythmicity over the baseline period, all comparisons are made
with respect to individuals baseline values. Repeated measures ANOVA was also
used to establish a time of day effect over the baseline period.
5.3.3 Sleep data and daily means for mood and performance
All sleep, mood and performance data were calculated as a percentage of or
difference from individuals own baseline. Daily means for every mood and
performance parameter were then determined. The performance parameters accuracy,
response time and efficiency were defined as the percentage of correct responses, time
taken to complete the task and accuracy/response time, respectively. Response time
data were normalised by square root determination.
5.4 RESULTS
One subject did not conform to the protocol and was therefore excluded from
all analysis.
Cosinor analysis gave most significant fits for 24h temperature data (100%
at p<0.005) followed by urinary aMT6s (84% at p<0.05). However, the percentage
variance accounted for by a cosine curve was > 80% in the 24h aMT6s profiles.
167
5.4.1 Ambient light monitoring
At home in the evening, light intensity did not exceed 180 lux (average 110
lux). Maximum natural bright light exposure (up to 10 000 lux) commonly occurred
around 0800-0900h, 1200-1400h and 0500-0700h corresponding to travel to work,
lunch and travel home, respectively. However, the light intensity always remained
less than < 1000 lux with sunglasses when outside and more commonly <500 lux.
In the work environment, the general light intensity was approximately 490 lux (range
55-1000 lux). However, whilst working at the laboratory bench or at the computers,
the average light intensity was reduced to 170 lux (range 55-500 lux).
5.4.2 Baseline
No baseline differences for any sleep, mood, aMT6s or temperature
parameters were evident. Of the performance variables, a small but significant fall
in SAM-1 accuracy was found (p<0.05, as 1% fall, D-2 to DO). A significant time
of day effect was established by ANOVA (repeated measures) for aMT6s,
temperature, alertness (Figures 5.2,5.3 and 5.4, respectively) and performance
response time (p<0.0001). One subject did not exhibit any significant cosine rhythm
or time of day effect as determined by cosinor analysis and repeated measures
ANOVA. This subject was discarded from the analysis of SAM-5 performance
variables.
168
I
iIIIIiiii 000000000 00000000 00 10 d' NO 00 Np rN P" P-1 r-+
(`U"u) Sum
ýi O
`..
0 0 N ý
O O N ri
ý, ý, o .r öý A,
ýýä r. 0)
.0 qj 0ýi
ýF ed NNý
0. iý ^ Sro 4- .a rI nj P" -d "'' "O - 'C t2.. +'
O rF1, ýb-
.,, vu ý"yý u .ý
K, ca - ;, =
to 1-4 Ö ý'° ybý'
iý
0 0 N ý
N
OQ O
ö GÖ ýQ ýý -I- 1-11
.äQ
ööv .ýö N
ocäc,
ýý aý a ý oml i-. a,
öQ N ý
O ÖQ
N ý
oý oQ N
ý
N Oý OQ N ý
= J--
169
Figure 5.3 Temperature profiles before and during a gradual 9h delay phase-shift
37.60-1
37.40-
, -, 37.20-
37.00-
36.80- E H 36.60-
36.40
36.20 -
36.00 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 D-2 D-1 DO D1 D2 D3 D4 D5 D6 D7
Clock time/ Day of study
Circadian profiles for core body temperature over a3 day baseline (D-2 to DO) and a gradual 9h delay phase-shift (D 1-5) for one of the four trial legs. There were no differences between
the four legs. A significant phase-delay was indicated by ANOVA for repeated measures (time of day by day of study interaction, D5 versus baseline, p<0.001). A decline in
amplitude was observed over the phase-shift with return to baseline values on D5. Bars
show patterns of exposure to bright light (1200 lux, =), darkness (< 1 lux, ý) and natural ambient light (®). Temperature was recorded continously every 6 minutes then averaged and plotted every hour as the mean of 7 subjects (+ SEM).
170
Figure 5.4 Alertness profiles before and during a gradual 9h delay phase-shift
6 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00
D-2 D-1 DO D1 D2 D3 D4 D5 D6
Clock time/ Day of study
15-
14-
13-I
12-
11-
10
9
8
T
i
t
ý
T,
.T
7
Daily profiles for self-rated alertness over a3 day baseline (D-2 to DO) and a gradual 9h delay phase-shift (D 1-5) for one of the four trial legs. There were no differences between the four legs. Using cosinor acrophase estimates, over 3-day windows, there was a significant phase-delay of 10.62 +/- 0.96h (mean +/- SD, D3-5 compared to D-2 to DO, p<0.0001, ANOVA). =Bright light exposure (1200 lux),
m darkness (< 1 lux), ® normal ambient light. Values represent mean (+ SEM, n=7) alertness rated every 2h during wakefulness.
171
5.4.3 Gradual, forced 9h delay phase-shift
Using moderately bright light combined with darkness (sleep) it was possible
to force a progressively larger delay phase-shift in the temperature and aMT6s
rhythms (max. degree of shift, D5 compared to mean baseline: 9.13 ± 0.83h and
9.09 ± 1.06h, respectively), Figure 5.5. A comparable shift in the alertness and
performance response time rhythms was also evident, Figures 5.6-5.8 (degree of
shift, D3-5 compared to baseline: 10.62 ± 0.96h, 10.51 ± 1.33h [SAM-1] and 9.24
± 1.88h [SAM-5], respectively). A delay phase-shift was also indicated with
ANOVA on the raw aMT6s and temperature data, Figures 5.2 and 5.3 (time of day
by day of study interaction, p<0.0001). Significant suppression of the temperature
and aMT6s rhythm amplitudes was observed over the forced phase-shift (ANOVA,
p<0.0001) with return to baseline values on D4-5 for aMT6s and D5 for
temperature, Figures 5.2 and 5.3.
No important detrimental effects on mood or sleep were evident, Figures 5.4,
5.9 and 5.10. However, a small but significant increase in sleep latency (p=0.025)
and corresponding fall in sleep duration (p=0.0017) was observed during the phase-
shift (mean ± SD: 8.1 ± 13.1 minutes, increase in sleep latency). This was mainly
attributable to the first two nights of the phase-shift. Daily mean alertness and
calmness significantly increased by 3.0 ± 1.0% and 7.4 ± 2.1%, respectively,
during the shift (D1-5) with the greatest improvement associated with the latter end
of the shift.
172
Figure 5.5 Acrophases of aMT6s and temperature rhythms during a gradual 9h delay phase-shift
G) rA cd
. J. - C. O ýO
s. ý,
"ý aý +r 0. ý C) H
14: 00-
12: 00-
10: 00-
08: 00-
06: 00-
04: 00-
02: 00-
02: 00-
24: 00-
22: 00-
20: 00-
18: 00-
16: 00-
14: 00-
-2 -1 012345
T----I--- T
-2 -1 0 : ý..:. ý
T
1 I
T
2 T
3 T
4
Baseline Delay phase-shift
Day of study
i
5 I
Acrophase estimates for (a) aMT6s and (b) core body temperature ('masked') over a three day baseline (D-2 to DO) and during a slow 9h delay phase-shift (D1-5) for the 4
pre-treatment legs of the trial (Bright light, , Placebo, o, Melatonin + Bright light, "
and Melatonin, £ ). A highly significant phase-delay was indicated by ANOVA (for
repeated measures, p<0.0001). Degree of shift (D5 compared to mean baseline) for
aMT6s was 9.13 +/- 0.83h, and for temperature was 9.09 +/- 1.06h (mean +/- SD). Values represent mean +/- SEM, n=7.
173
Figure 5.6 Adaptation of the alertness rhythm during a gradual delay phase-shift
02: 00- 24: 00- 22: 00- 20: 00- 18: 00- 16: 00- 14: 00-
:ýý 100ý
.uöO Gý ++ 0
Ö ý"r "ý
0 "N
ý
ý
\ 0
80
60 40
20
0 70 60 50 40 30 20 ]0 0
a)
C)
1
I
A ol
II
.1
I I
D-2 to 0 DI-3 D3-5 Day of study (3-day window)
a) Cosinor-derived alertness acrophase estimates using 3-day windows during baseline (D-2 to DO) and a gradual 9h delay phase-shift (D 1-5) for the 4 pre- treatment legs of the trial (Bright light, , Placebo, 0, Melatonin + Bright fight, " and Melatonin, A ). An estimate of entrainment (to a 24h cosine rhythm) is provided by (b) % number of subjects with significant cosine fits (set at p<0.05) and (c) % variance accounted for by a cosine curve. A significant phase-delay of 10.62 +/- 0.96h (mean +/- SD) was obtained during the forced delay shift (p<0.0001, ANOVA, D3-5 compared to baseline). Values for Figs. (a) and (c) represent mean +/- SEM (n=7).
174
Figure 5.7 Adaptation of the SAM-1 response time performance rhythm during a gradual delay phase-shift
Cl) r. + ti: i
04: 00- 02: 00- 24: 00- 22: 00- 20: 00- 18: 00- 16: 00- 14: 00-
aý ý "W
Oý Uý
V
'5 I
ý
fý4 \° 0
a)
T T
D-2 to 0 DI-3 D3-5 Day of study (3-day window)
i
1
a) Cosinor-derived SAM-1 response time acrophase estimates using 3-day windows during baseline (D-2 to DO) and a gradual 9h delay phase-shift (D1-5) for the 4 pre- treatment legs of the trial (Bright light, U, Placebo, 13, Melatonin + Bright light, "
and Melatonin, A ). An estimate of entrainment (to a 24h cosine rhythm) is provided by (b) % number of subjects with significant cosine fits (set at p<0.05) and (c) %
variance accounted for by a cosine curve. A significant phase-delay of 10.51 +/- 1.33h (mean +/- SD) was obtained during the forced delay shift (p<0.0001, ANOVA, D3-5 compared to baseline). Values for Figs. (a) and (c) represent mean +/- SEM (n=7).
175
Figure 5.8 Adaptation of the SAM-5 response time performance rhythm during a gradual delay phase-shift
04: 00- 02: 00- 24: 00- 22: 00- 20: 00- 18: 00- 16: 00- 14: 00-
a)
T i
D-2 to 0 DI-3 D3-5
Day of study (3-day window)
i
i
a) Cosinor-derived SAM-5 response time acrophase estimates using 3-day windows during baseline (D-2 to DO) and a gradual 9h delay phase-shift (D1-5) for the 4 pre- treatment legs of the trial (Bright light, , Placebo, Q, Melatonin + Bright light, " and Melatonin, A ). An estimate of entrainment (to a 24h cosine rhythm) is provided by (b) % number of subjects with significant cosine fits (set at p<0.05) and (c) % variance accounted for by a cosine curve. A significant phase-delay of 9.24 +/- 1.88h (mean +/- SD) was obtained during the forced delay shift (p<0.0001,
ANOVA, D3-5 compared to baseline). Values for Figs. (a) and (c) represent mean +/- SEM (n=7).
176
Figure 5.9 Daily means for mood before and during a gradual 9h delay phase-shift
H y N :+^
aý C ý "ý
ýy RS
ND
Eý .: r Ri
A
V1 y N
N ...
Vy
0) N ý
`ý
ot.
90 ý1
115-
110-
105-
100-
95
T I T
-2 -1 012345 115-
110-
1051
100
9s1
H y ý
ýý
ej - V) ýH
ýý .ý A
90
-2 -1 012345
115-
110-
105-
100-
95
90 II I I -2 -1 0
Baseline
I I
I
I
2
i
i
3
I
T
4
Delay phase-shift Day of study
T
5 I
Daily means for (a) alertness, (b) cheerfulness and (c) calmness over a three day baseline (D-2 to DO) and during a gradual 9h delay phase-shift (D1-5) for the 4
pre-treatment legs of the trial (Bright light, , Placebo, [] , Melatonin + Bright
light, " and Melatonin, " ). No deleterious effect of the gradual 9h delay shift on any mood parameter. Alertness and calmness significantly increased during the phase- shift (p<0.0001 and p<0.05, respectively, ANOVA). Values represent mean +/- SEM, n=7.
ý
177
Figure 5.10 Subjective sleep measures during a gradual forced 9h delay phase-shift
y
3° ce * ? DO C7
öý C -v o ... 1ý ý. ý A
-3 -2 -1 012345
C)
-3 -2 -1 012345 Baseline Delay phase-shift
Night of study
Subjective measures of (a) sleep quality, (b) duration of night awakenings and (c)
sleep latency over a4 night baseline (nights -3 to 0) and during a gradual 9h delay
phase-shift (nights 1-5) for the 4 pre-treatment legs of the trial (Bright light, , Placebo, [3, Melatonin + Bright light, " and melatonin, " ). Small but significant increase in sleep latency was observed during the phase-shift (p<0.05, mainly attributable to night 1).
70
-3 -2 -1 012345
140- 130- 120- 110- 100- 90- 80-
30
20
10
0 -10
-20
178
Figure 5.11 Daily means for SAM-1 performance measures before and during a gradual forced 9h delay phase-shift
V
E
oa yN
ii. ý.
y CG
IZ
O
C `/
ý ý .ý A
110-,
105 t
100-ý
95 1 I i I I i T I
-2 -1 012345
104 ý b)
102 ý
looý 98-1
1
-2 -1 012345 94
96
105
100
95
90
Baseline Delay phase-shift
Day of study
Faster
response times
I,
Daily means for SAM-1 (a) efficiency (b) accuracy and (c) response time over a three day baseline (D-2 to DO) and during a gradual 9h delay phase-shift (D1-5) for the 4 pre-treatment groups (Bright light,
, Placebo, [], Melatonin + Bright light, " and Melatonin, f ). Significant fall in accuracy, and increase in response time and efficiency during the phase-shift (p<0.001, ANOVA). Values
represent mean +/- SEM, n=7.
179
Figure 5.12 Daily means for SAM-5 performance measures before and during a gradual forced 9h delay phase-shift
110-7
105 -
100 -1
95
90 J
110-7 105-
100- (L)
ýý 95- 90-
85-
80
i I I I I T I T
-2 -1 012345
b)
iT
-2 -1 012345
Faster
response times
-2 -1 012345 Baseline Delay phase-shift
-1 I
Day of study
Daily means for SAM-5 (a) efficiency (b) accuracy and (c) response time over a three day baseline (D-2 to DO) and during a gradual 9h delay phase-shift (D1-5) for the 4 pre-treatment groups (Bright light, , Placebo, 0, Melatonin + Bright light, " and Melatonin, " ). Significant fall in accuracy, and increase in response time during the phase-shift (p<0.001, ANOVA). Values represent mean +/- SEM, n=6.
180
SAM-1 and SAM-5 response time quickened significantly (3.7 ± 1.2% and
2.8 ± 1.5%, p<0.0001) although at the expense of accuracy which fell (SAM-1:
2.35 ± 0.38%, p=0.028; SAM-5: 4±1.7%, p=0.0002). Overall performance
efficiency improved for SAM-1 (2.9 ± 0.3%, p=0.0007) and remained unchanged
for SAM-5, Figures 5.11 and 5.12, respectively. The fall in performance accuracy
improved towards the latter end of the shift.
5.4.4 Adaptation to a rapid 9h advance phase-shift (placebo treatment)
After the rapid 9h advance shift of external time cues on day 6a large and
highly significant (p<0.0001) reduction in sleep duration (4 1.6 ± 0.36h), quality
(1 24 ±5%, Figure 5.13) and increase in the duration of night awakenings (t 52.6
± 14.7 minutes, Figure 5.13) was apparent, nights 6-11. Sleep latency and the
number of night awakenings remained virtually unchanged, nights 6-11.
The daily means for cheerfulness and calmness deteriorated (7.5 ± 3% and
6.1 ± 3%, respectively, D7-11) but the major deleterious effects were evident with
alertness (p<0.001,11.5 ± 2.7%), Figure 5.14.
Significant decrements in daily mean performance accuracy (SAM-1: 4.0 f
0.58%, SAM-5: 11.12 ± 1.55%) and efficiency (SAM-1: 4.2 ± 1.17%, SAM-5:
11.12 ± 1.2%) were evident (p5 0.0003, D7-11), Figure 5.15. Slower reaction
times (SAM-1: t 0.45 ± 0.97 minutes, SAM-5: t 0.76 ± 1.07 minutes, D7-11)
were also apparent. The performance response time and alertness rhythms did not
reestablish a normal baseline pattern by day 11 (see Figures 6.8,6.11 and 6.12).
181
Figure 5.13 Adaptation of sleep to a rapid 9h advance phase-shift
130 -1 a) i inI ý 11V
rr 0 90 i --L
li T
U5 ä 70- '10
50-
30 6789 10 11 Mean
Baseline
y
0ý
WN
N)
k2 iz "O 9 lýI u
E
171 ý
0 ý
240- 200- 160- 120- 80- 40- 0-
35,
25 H
W 15 ýy
ý .05 E -5
-15-'
6789 10 11 Mean Baseline
C)
EI01® ý ý
Mean 6789 10 11 Baseline
Night of study
Subjective measures of (a) sleep quality, (b) duration of night awakenings and (c)
sleep latency showing adaptation to a rapid 9h advance phase-shift (on day 6) in a normal environment. Significant deleterious effects on sleep quality and duration of night awakenings were evident (p<0.0001, ANOVA for repeated measures, nights 6- 11 compared to baseline) with no significant effect on sleep latency. Values represent mean +/- SEM, n=7.
182
Figure 5.14 Adaptation of mood to a rapid 9h advance phase-shift 110
(A ää
_N y
Cd cd
4. ) CCG
"ý ý
110-
100-
90 i -I-
80 6789 10 11 Mean
Baseline
T
T 89 10 11
Day of study Mean
Baseline
Subjective measures of (a) alertness, (b) calmness and (c) cheerfulness during adaptation to a rapid 9h advance phase-shift (on day 6) in a normal environment. Significant deleterious effects on alertness were seen (p<0.001, ANOVA for repeated measures, D7-11 compared to baseline). Values represent mean +/- SEM, n=7.
9 10 11 Mean Baseline
183
Figure 5.15 Adaptation to a rapid 9h advance phase-shift : daily mean performance
105 -,
4
I III 89 10 11
- 7-Mean
Baseline
1
ý 105 ..,
103 a aý
101
99
97 .,.., Q 95
- 7-Mean
Baseline
IIIIII 6789 10 11 C)
4 T
6789 10 11 Mean ý--l
Day of study Baseline
ý
Slower
response times
Daily means for performance measures of (a) efficiency, (b) accuracy and (c)
response time for SAM-1 ( ) and SAM-5 (0) tasks showing adaptation to a rapid 9h advance phase-shift (on day 6) in a normal environment. Significant decrements in efficiency and accuracy were evident (p<0.001, ANOVA for repeated measures, D7-11 compared to baseline). Values
represent mean +/- SEM (SAM-1: n=7; SAM-5: n=6).
184
Figure 5.16 Adaptation of aMT6s and temperature rhythms to a rapid 9h advance phase-shift
0
.9 ý Q) ý cd
.ý a 0- C. ) Co
cl
ý
18: 00 a)
16: 0
6789 10 11 Mean Baseline
14: 00-
12: 00-
10: 00-
08: 00-
06: 00-
04: 00-
02: 00-
i Lri-----4-1
13: 00 6789 10 11
17: 00-
ýs: oo 15: 00-
0
Mean Baseline Day of study
Acrophase estimates for (a) aMT6s and (b) core body temperature ('masked') showing adaptation to a rapid 9h advance phase-shift (on day 6) in a normal environment (placebo treatment stage). aMT6s acrophase remained significantly delayed compared to baseline (p<O. 01, D7-11,
repeated measures ANOVA) whilst the 'masked' temperature rhythm apparently reentrained by D7. Values represent mean +/- SEM, n=7.
185
Figure 5.17 Adaptation of the temperature rhythm to a rapid 9h advance phase-shift
37.60-
37.40-
37.20-
37.00-
36.80-
36.60-
36.40-
36.20-
36.00 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00
D5 D6 D7 D8 D9 D10 D11
Clock time / Day of study
12: 00 Baseline
(D-1)
Adaptation of the circadian rhythm for core body temperature after a rapid 9h advance phase-shift (on day [D] 6) in a normal environment (placebo stage). Although the calculated peak of the rhythm indicates reentrainment by D7 (see Figure 5.16), the rhythm amplitude remained significantly reduced (p<0.0001, D7- 11 compared to baseline, ANOVA). Normal baseline patterns are not established until at least D 10. Bars show patterns of exposure to darkness (< 1 lux, M), natural ambient light (= ) and bright light (1200 lux, o ). Temperature was recorded continously every 6 minutes then averaged and plotted every hour as the mean of 7 subjects (+ SEM).
186
Figure 5.18 Direction of reentrainment of individual aMT6s rhythms after a rapid 9h advance phase-shift
24: 00
22: 00
20: 00
18: 00
16: 00
14: 00
12: 00
10: 00
08: 00
06: 00
04: 00
02: 00 56789
Day of study
-T------r 10 11
i
Mean Baseline
(+/- SEM, n=7)
Acrophase estimates for individual circadian aMT6s profiles showing the direction of reentrainment after a rapid 9h advance phase-shift (on day 6) in
a normal environment (placebo treatment). At least two subjects show an antidromic effect where they reentrain to a 9h advance shift by delay.
187
Temperature acrophase was delayed significantly from baseline on D6 only
(p<0.01, Figure 5.16) although the rhythm amplitude remained significantly lower
on D6-10 (p < 0.05), Figure 5.17. aMT6s acrophase remained significantly delayed
from baseline (p<0.01) on days 6-11. This was also indicated with ANOVA on the
raw aMT6s data where a time of day by day of study interaction was still significant
on D II compared to DO (p<0.0001), Figure 5.16.
The direction of the subjects natural entrainment of the aMT6s rhythm varied.
At least 2 subjects reentrained by phase-delay and at least 2 subjects phase-advanced,
Figure 5.18.
5.5 DISCUSSION
5.5.1 Gradual 9h delay phase-shift
Using a controlled environment, an approximate 9h delay phase-shift was
induced in the core body temperature and urinary aMT6s rhythms, two commonly
used physiological rhythm markers of the biological clock. A similar degree of shift
was also identified with the behavioural rhythms of self-rated alertness and
performance response time. The contemporary shift in the sleep/wake cycle clearly
influenced the calculated estimates of acrophase for mood and performance rhythms
since this data was only collected during wakefulness. In addition, exogenous
influences may have modified cosinor estimates of the temperature rhythm.
However, comparable shifts were seen with the aMT6s rhythm which is minimally
influenced by exogenous factors apart from the acute suppressant effects of bright
188
light, see section 1.3.
The maximum degree of shift induced (approx. 9h) was assessed on the last
day of the gradual delay phase-shift under bright light conditions (D5). However, the
aMT6s rhythm had reestablished the normal baseline amplitude and shape on D4-5
and visual inspection of the aMT6s raw data confirms the magnitude of the maximum
shift. A mean delay shift of 7.5h (± 1.5h, mean ± SD) was obtained on day 6,
when subjects returned to the normal environment in conditions of darkness and dim
light with no masking influences. This provides a reliable estimate of the shift
induced allowing for the natural adaptation of the circadian system after one day.
The gradual delay shift in environmental and social (e. g. meals) time cues,
together with the suitably timed bright light/darkness were effective in promoting the
adaptation of sleep, mood and performance efficiency to the 9h delay shift. No
important detrimental effects on sleep and mood were apparent. Small (2-4%) but
significant deleterious effects on performance accuracy were observed, mainly over
the changing shift period (D1-3), which stabilised when the 9h delay shift was
maintained for another two days. The accelerated response time seen over this period
was unlikely to be a practice effect since tests were performed regularly before each
stage of the trial and no observable trend was seen over the baseline. Overall,
performance efficiency remained either unchanged or slightly improved. In addition
to the effects of bright light on the circadian system, bright light exposure may have
directly enhanced mood and performance (Campbell and Dawson, 1990; French et
al, 1990; Myers and Badia, 1993).
189
5.5.2 Rapid 9h advance phase-shift
The major and significant deleterious effects on behaviour occurred after
subjects resumed their baseline sleep pattern in their normal environment. This rapid
9h advance shift of external time cues can be related to a 9h eastward transmeridian
flight. Using this direct analogy, a flight at 2300h from Los Angeles to Paris after
a 9h flight crossing 9 time zones (forward in time) would arrive in Paris at 1700h
local time or 0800h Los Angeles time. Having adopted local time, travellers would
retire to bed at approximately 2300-2400h (1700-1800h Los Angeles time) to try and
get sufficient sleep before rising the following morning for work at 0900h. This
schedule was simulated in the present experiment. Trying to sleep at inappropriate
circadian phases inevitably resulted in some sleep disruption or loss because the body
was not prepared to sleep until 0500-0600h local time, corresponding to 2300-2400h
Los Angeles time. Travel fatigue and sleep loss whilst flying may have compound
these problems in real life.
Sleep patterns, mood and performance variables and the endogenous aMT6s
rhythm took at least 5 days to reestablish normal baseline patterns in agreement with
field based studies (Klein and Wegmann, 1980; Fbvre-Montange et al, 1981,
Wegmann et al, 1983; Arendt et al, 1987) and laboratory simulations (Samel et al,
1991). The temperature rhythm apparently took two days reentrain. However,
environmental and behavioural masking were likely to be responsible for this effect.
Since the natural reentrainment of circadian rhythms after rapid and large phase-shifts
can take many days or weeks (Aschoff et al, 1975; Klein and Wegmann, 1980;
Wegmann et al, 1983), and since the zeitgeber strength of natural light exposure was
190
controlled, it was unlikely that in every individual the rhythm shifted by 9h in one
day. The significant reduction in rhythm amplitude suggests that the rhythmic
functions of core body temperature had not reentrained properly even after 5 days.
One potential way of eliminating the masking effects on the endogenous circadian
component of the temperature rhythm is a mathematical procedure (Minors and
Waterhouse, 1992). This method has been used to "purify" the temperature profiles
and will be discussed in Chapter 7.
Usually, the circadian system reentrains in the direction of the induced shift
and environmental zeitgebers (Klein and Wegmann, 1980; Wegmann et al, 1983).
However, the direction of shift in the urinary aMT6s rhythm to the rapid 9h advance
change of external time cues varied in different individuals even with similar
environmental zeitgebers. Unfortunately, it was not possible to assess properly the
direction of entrainment for some subjects and the time taken for complete
reentrainment to occur due to the limited number of days studied. However, some
individuals showed a definite tendency to reentrain to the rapid 9h advance shift by
delay, a so-called antidromic effect (Klein and Wegman, 1980) where individuals
delay the circadian system by 15h to complete the 24h cycle of entrainment. This
effect has been observed in field studies over 8-10 time zones flying east (advancing
time) (Mills et al, 1978; Colquhoun, 1979; Klein and Wegmann, 1980; Gander et al,
1989; Spencer et al, 1994) and in laboratory based studies after phase-advances of 6h
(Wever, 1980; Moline et al, 1992).
The natural tendency for circadian rhythms to delay may explain the fact that
some individuals apparently reentrained by phase-delay after a rapid and large phase-
advance. Reasons for the antidromic effect remain unknown. Individual differences
191
in urinary aMT6s acrophase or "morning/evening" type personality scores did not
correlate with the direction of natural reentrainment. However, the small sample size
precludes the determination of any reliable relationship.
The advantage of field studies on travellers and shift workers is the exposure
to a whole range of external time cues and psychological influences. However, there
are important disadvantages. Susceptibility to jet lag is highly individual therefore
crossover experiments afford the best design. However, field studies are expensive
and difficult to control since standardising social and environmental stimuli over each
stage is virtually impossible. There are also difficulties in the collection of
biochemical data. Laboratory studies provide a means of measuring a whole range
of physiological and behavioural variables in detail and in a controlled environment.
But the laboratory environment may not give a true reflection of the normal
readaptation process in real life. For example, Wegmann et al (1986) described how
sleep quality improved in the quieter isolated environment of the laboratory in
comparison to real life. Allowing subjects to readapt in a normal environment
overcomes this problem. This experimental model therefore acts as a compromise
between field studies and studies in complete environmental isolation.
Using a controlled environment it was possible to force a progressively larger
delay phase-shift in circadian rhythms with no significant detrimental effect on sleep
and mood and only minimal effects on performance. There were no differences
between the four pre-treatment stages suggesting that this design is effective for
crossover experiments. The subsequent and rapid 9h advance shift of external time
cues induced major and significant problems related to behaviour and physiology.
This model provides an appropriate method to study adaptation strategies after rapid
192
shifts in local time. The following Chapter describes the effects of melatonin and
bright light on adaptation to a rapid 9h advance phase-shift in the second half of this
experiment.
193
CHAPTER 6
INFLUENCE OF MELATONIN ON ADAPTATION TO A
LARGE RAPID ADVANCE PHASE-SHIFT IN THE
PRESENCE OF CONFLICTING BRIGHT LIGHT
Influence of Melatonin on Adaptation to a large Rapid Advance Phase-shift in the Presence of Conflicting Bright Light
6.1 INTRODUCTION
When exogenous melatonin is administered to humans, at appropriate times
and in entrained conditions, it will phase-shift its own endogenous circadian rhythm
(Arendt et at, 1985a; Lewy et at, 1992; Zaidan et at, 1994) and the alertness/fatigue
rhythm (Arendt et at, 1984). These properties together with its somnolent and sleep
inducing capabilities (Cramer et at, 1974; Vollrath et at, 1981; Waldhauser et at,
1990; Dollins et at, 1993,1994) suggest that melatonin may provide a means of
expediting the adaptation of circadian rhythms following time-zone transitions or
shiftwork similar to the ability of bright light (Czeisler et at, 1990; Eastman, 1991)
and help improve sleep, performance and mood at the appropriate times.
In both field studies and a laboratory simulation melatonin has been shown to
relieve subjective feelings of jet lag and facilitate the reentrainment of endogenous
circadian rhythms (Arendt et at, 1987; Arendt and Aldhous, 1988; Claustrat et at,
1992; Nickelsen et at, 1991b; Petrie et at, 1989,1993 and Samel et at, 1991).
Additionally, during rotating shift schedules melatonin administration favours the
reentrainment of the body temperature rhythm (Armstrong et at, 1986) and improves
sleep and alertness (Folkard et at, 1993). However, in previous studies not all
subjects responded beneficially to melatonin treatment (Arendt et at, 1985a; Mallo et
at, 1988; Skene et at, 1988). This may be due to the incorrect timing of melatonin
administration or perhaps inappropriate exposure to natural ambient light.
195
Light exposure to treat jet lag can be timed to induce phase-advances or phase-
delays depending on its position with regard to the temperature minimum (Czeisler
et at, 1989; Minors et at, 1991, van Cauter et at, 1994, section 1.4.3). However,
in real life it is inconvenient and difficult to control natural light exposure. This may
occur at inappropriate times for the most expeditious reentrainment of circadian
rhythms and may counteract any beneficial influence of melatonin.
In field studies, it is difficult to standardise conditions in a crossover design
experiment. However, in a controlled environment it is possible to force a 9-10 hour
delay phase-shift and then release the subjects back into their normal environment,
simulating a rapid 9-10 hour eastward time-zone shift (see chapter 5).
This present study was designed primarily to study the effects of morning
bright light exposure potentially encountered by time-zone travellers or shift workers
and the consequence it may have on melatonin treatment administered near bedtime,
after a rapid 9h advance shift of external time cues, i. e. after a simulated eastward
transmeridian flight over 9 time-zones. Bright light exposure was timed to phase-
delay the biological clock and the melatonin treatment to facilitate the adaptation by
phase-advancing.
A second aim was to identify whether the beneficial influence of melatonin
on behaviour corresponds to the phase-shifting or resynchronising effects on the
endogenous "clock" in a placebo-controlled randomised crossover trial in the same
environment.
The effects of melatonin or placebo on adaptation to a rapid 9h advance phase-
shift were investigated, in the presence and absence of bright light exposure, in a4
leg, randomised, placebo-controlled, double blind crossover study.
196
6.2 SUBJECTS and METHODS
6.2.1 Subjects
See section 5.2.1
6.2.2 Study design
Figure 6.1 shows a diagrammatic representation of the study. Section 5.2.2 describes
the initial part of the study in detail. In summary, volunteers were initially subjected
to a gradual, forced synchronised phase-delay of 9h using a combination of bright
light and imposed darkness (and sleep) in a controlled environment (D1-D5). The
adaptation to a subsequent rapid 9h advance shift in the subjects normal environment
was then studied under 4 different treatments.
Treatments:
Immediately following the maximum phase-delay (9h) induced by last
light/dark treatment, subjects were released back into their normal environment and
required to assume the baseline sleep/wake schedule, i. e. retiring to bed at 2330h and
waking at 0800h for 6 days (D6-1 1). The subsequent readaptation of the subjects in
their normal environment was then observed under 4 different treatments consisting
of 1) melatonin + dim light (MT), 2) placebo + dim light (P), 3) melatonin + bright
light (MT + BL) and 4) bright light + placebo (BL). Melatonin (5mg in gelatin-
lactose capsules) or placebo (vehicle) was administered just before the desired bedtime
at 2300h on D6-8. Bright light treatment occurred between 0800-1200h on D7-8 (dim
197
Figure 6.1 Diagrammatic design of the study
I
-1 I
0 1
2 3 4
5
6
7 8
9
10
11
I
+1 +1
I I
I
---F- ýIIIýIII., I 24: 00 04: 00 08: 00 12: 00 16: 00 20: 00 24: 00
Time of day
After a baseline period (days [D] -2 to 0), this study design was used to induce an initial, gradual 9h delay phase-shift (D 1-5) in volunteers using a combination of exposure to bright light (- 1200 lux, o) and darkness (ý). Subjects then resumed their normal baseline sleep/wake cycle in a normal environment on day 6 inducing a rapid 9h advance phase-shift of local time cues. In this 4 leg crossover study the readaptation was studied under four different treatments : melatonin, placebo, melatonin+bright light and bright light. ® Time of bright light treatment (2000 lux). + Time of melatonin (5mg) or placebo administration. O Natural ambient light exposure.
198
light, < 150 lux; bright light, 2000 lux). Subjects avoided bright sunlight by wearing
sunglasses at all times when outdoors (transmitting < 10% of the light). At dusk,
they were required to remain in ordinary domestic lighting (50-300 lux) until bedtime
(and darkness). Each subjects' ambient lighting at work and home was monitored,
see section 5.4.1.
Subjects were only allowed to sleep during the periods of darkness throughout
this study. Naps were forbidden.
6.2.3 Sampling
See section 5.2.3.
6.2.4 Behavioural Measurements
See section 5.2.4
6.2.5 Hormone assays
See section 5.2.5
199
6.3 STATISTICAL ANALYSIS
6.3.1 aMT6s and core body temperature profiles
Comparisons between treatments for the aMT6s rhythm were made on the
post-treatment days 10 and 11 where exogenous melatonin (aMT6s) had cleared from
the urine. On these days cosinor analysis gave 100 % fits at p<0.05. Temperature
profiles gave 100% fits over the treatment and post-treatment periods. Repeated
measures ANOVA followed by Duncan's post-hoc test were used to determine
differences between treatments for the raw aMT6s profiles on D 10-11 and cosinor-
derived rhythm parameters of aMT6s (D10-11) and temperature (D7-11).
6.3.2 Mood and performance daily means and sleep data
Treatment effects on the daily means for mood and performance parameters
and sleep variables were identified using repeated measures ANOVA (factors: day of
study and treatment), D7-11 (or nights 6-11 for sleep) followed by Duncan's multiple
range test for comparison between treatments.
6.3.3 Mood and performance rhythms.
For comparison between treatments, 24h profiles for each mood and
performance parameter were analysed by cosinor analysis over 3-day windows for
more reliable estimates of acrophase (treatment: D7-D9_ and post-treatment: D9-D11)
200
followed by ANOVA (for repeated measures, factors: day of study and treatment,
D7-D11) and Duncan's multiple range test. An estimate of reentrainment to the 24h
day (and reestablishment of baseline cosinor rhythmicity with a period of 24h) is
provided by (1) percentage number of subjects with significant 24h cosine fits and (2)
percentage variance accounted for by a cosine curve, see section 2.8. All
comparisons are made with respect to individuals' baseline values.
6.4 RESULTS
One subject did not conform to the protocol and was therefore excluded from
the analysis. One subject did not exhibit a significant cosine rhythm (from cosinor
analysis) or time of day effect (repeated measures ANOVA) for SAM-5 performance
response time and was excluded from the analysis of this parameter.
The phase position of the light and melatonin treatments with respect to the
temperature and melatonin rhythms was determined on D6 during the placebo stage.
Measurement of aMT6s and temperature on this day occurred in conditions of dim
light (< 200 lux) or darkness with the subjects resting or asleep the entire 24h
period. The midpoint of light treatment occurred 2.94 ± 1.21h (mean ± SD) before
the temperature minimum. Melatonin administration occurred 12.4 ± 1.21h (mean
± SD) before the temperature minimum and 7.4 ± 1.21h (mean ± SD) before the
calculated melatonin onset in plasma.
6.4.1 Ambient light monitoring
See section 5.4.1
201
6.4.2 aMT6s
Repeated measures ANOVA on raw aMT6s levels on D10-11 identified a
significant time of day by treatment interaction (p=0.0018, Figure 6.2) where MT
and MT + BL treatments favoured faster reentrainment of the aMT6s rhythm over
placebo and bright light which remained delayed. No effect of treatment on aMT6s
acrophase was evident (D10-11, p>0.1), Figure 6.3. Two subjects (AT5 and BT5)
delayed with MT and MT + BL treatment whereas all others advanced, Figure 6.3.
This differing effect of melatonin on certain subjects did not correlate with the
direction of individuals natural entrainment (cf Figure 5.18) or baseline acrophase.
BL delayed the aMT6s acrophase in 6 out of 7 subjects.
6.4.3 Temperature
No differences in temperature acrophase were seen between treatments
(Figures 6.4 and 6.5), although a significant time by treatment interaction for rhythm
amplitude (p=0.018) was evident where melatonin treatment favoured a faster return
to baseline amplitude over placebo and bright light.
6.4.4 Mood
There was a trend for both melatonin groups to improve daily mean alertness
and cheerfulness over placebo (p=0.06), Figure 6.6. BL improved alertness and
cheerfulness transiently over placebo, during the treatment period. There were no
significant effects on calmness. The alertness rhythm reestablished a normal baseline
cosine rhythm under the melatonin treatment groups with complete reentrainment,
Figures 6.7 and 6.8.
202
Figure 6.2 Effect of melatonin and bright light on aMT6s profiles during adaptation to a rapid 9h advance phase-shift
r-. ý
ý C
ý ý
ýO F-+ ý
22 20 18 16 14 12 10 8 6 4 2 0
1800 1600 1400 1200 1000 800 600 400 200
0
1800 1600 1400 1200 1000 800 600 400 200
0 12: 00 12: 00
D5 D6 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00
D7 D8 D9 D 10 D II
Clock time / Day of study
Adaptation of the circadian profiles of urinary aMT6s after a rapid 9h advance phase-shift under 4 different treatments: 1) melatonin (MT, A ), 2) placebo (P, 0 ), 3) melatonin + bright light (MT+BL, ") and 4) bright light (BL, ). Melatonin treatment (? ) was timed to phase-advance (2300h on D6,7 and 8)
. Bright light exposure (Q ) was timed to phase-
delay (0800-1200h on D7 and D8). A significant time of day by treatment interaction (p<O. 002) on D 10-11 (exogenous melatonin days excluded) where both melatonin treatment groups facilitated faster reentrainment of the aMT6s rhythm over BL and P. M Darkness from 2330h to 0800h (< 1 lux), D6-11 .m Natural ambient light. Values represent mean +/- SEM (n=7) and are plotted with reference to the midpoint of the urine collection period.
203
Figure 6.3 Effect of melatonin and bright light on adaptation of aMT6s acrophase to a rapid 9h advance phase-shift
20: 00-
18: 00-
16: 00-
14: 00-
12: 00-
10: 00-
08: 00-
06: 00
04: 00
n=7
56789 10 11 Mean baseline
22: 00 b)
18: 00 .ý" ý ý
1 AAA a-_----"--- 1fF: UV I
2
56789 10 11 Mean baseline
02: 00 1ýT
., t -I n=2
-:: -.
-* -- -4:: z I 06: 00 -1 \I n=5
ý r_
10: 001
Day of study
Acrophase estimates for urinary aMT6s during adaptation to a rapid 9h advance phase-shift (on day 6) showing (a) effect of 4 different treatments on group mean (n=7): melatonin (MT, A), placebo ( ), melatonin + bright light (MT+BL, ")
and bright light (13 ), no significant effect of treatment, repeated measures ANOVA), and (b) two subjects who delayed with MT (A) and MT+BL (0) treatments compared to all others who advanced with MT (A) and MT+BL ("). Exogenous melatonin days excluded from the analysis (- --). Values
represent mean +/- SEM. 4 Bright light exposure. Melatonin administration.
204
ý
d L
E ae L ý
a E aý ý aý s ý w O fl 0
*rw M 920 CC Jz
V (A
s fl ao`ý
"ý o,
"v a fl
m-o O6
sý O
,T IIIII eý
"II: NÖ
oý0 d i. t- t- t- t- IsO ýMMMMM
... W
rl
ýyv
ý ý: ... ýv Gq ýH NýCw 146 "--~ ý~ý 4) -` -
o--4 oQ N . -ý
,-1
ý^k" o
ýO&d) ...
-d ý
ID
.iý 0sm . °Q
O ^' a' + .+ oA ri ýý. aý
O\ A
> .+w ýýý
0 0 CA ""
4--ý
4--ý
T'ý-'ý'T'ý CD ooO ýO ýNO
\O MMMM
(D. ) amjuiadwaZ
0 0 --4
0 0 N ý
0 0
`ý M
: -. 00 +ý 0 Ii
ý N Ao "ý
a. ýxý Öy
CL. Öý
A- H
Q ed Z N O. O
ý GQ
.. r V1
WO
ýo ca ýo w. ý ý,
ý .ý 00 ýl V AýýýS
n A
ýD Q
A 0 0 N . --ý
GM >, 10 - (-l -= MNC i'
i aý ä, E0 ýýýýn ý 5. Vu
G E
- -- a, = ; Fl :. = P. Iv . r: --t. e
3 ý I- at U %- o a.
y ýý ý P. cd wý
^ I- aýC
ý+ .+0"' «+ Ci ýy
.C0 '% E Q :.: 4)
-- :x_w C 'a 00 150
ýOvý-ý
U
U 4 '4 U «+
NH ö
,1 .0v Q...., a : 6,
205
Figure 6.5 Effect of melatonin and bright light on adaptation of the temperature cosinor rhythm to a rapid 9h advance phase-shift
02: 00 -1
24: 00
22: 00-
20: 00 -ý
5 14: 00 -1-T
16: 00-
ý 18: 00
0.1-
o-
-0.1-
-0.2-
-0.3
-0.4
b)
5
9 10 11 III
1
+
- r--Mean
baseline
f
789 10 11 Mean Day of study baseline
(a) acrophase and (b) amplitude estimates for the core body temperature rhythm during adaptation to a rapid 9h advance phase-shift (on day 6) under 4 different treatments: melatonin (MT, A ), placebo (O), melatonin + bright light (MT+BL, ") and bright light ( ). The rhythm apparently reentrained for all treatment groups by D7. A time by treatment interaction (repeated measures ANOVA, p<0.02) for rhythm amplitude indicated faster adaptation with MT treatment. Values represent mean +/- SEM, n=7.
4 Bright light exposure. M Melatonin administration.
206
Under bright light and placebo treatments, the normal baseline rhythm remained
completely disrupted. The low % rhythm and low number of subjects with significant
cosine fits with cosinor analysis prevented accurate estimates of acrophase to be
determined.
6.4.5 Performance
An effect of treatment proved significant for daily mean performance
efficiency only (SAM-1, p=0.022 and SAM-5, p=0.0044, D7-11, Figures 6.9 and
6.10) where both MT and MT + BL significantly (p < 0.01) countered the decrements
in daily means under P treatment. BL alone improved these ratings over P (p <0.01)
but only transiently.
No consistent and reliable cosine rhythm could be fitted to accuracy and
efficiency data (see section 5.4.2). Both MT and the MT+BL treatment groups
facilitated the reentrainment of the response time rhythms for both SAM-1 and SAM-
5, Figures 6.11 and 6.12, respectively. Lack of fit and low % rhythm precluded
estimates of reentrainment for bright light and placebo groups.
6.4.6 Sleep
MT and MT+BL improved sleep duration (p < 0.05), quality (p < 0.01, Figure
6.13) and duration of night awakenings (reduced, p <0.05, Figure 6.13) over P and
BL, N6-11. No differences were found between MT and MT+BL and between P
and BL for any sleep parameter.
207
Figure 6.6 Influence of melatonin and bright light on adaptation of mood to a rapid 9h advance phase-shift
ý -T--l
120-]
110-
100-
90-
80--
)
mm4mý
8
Y
9 10 11
T Ir 10 11 6789
Day of study
Daily means for (a) alertness and (b) cheerfulness during adaptation to a rapid advance phase-shift (on day 6) under 4 different treatments: melatonin (" ), placebo (a), melatonin + bright light (") and bright light (0). There was a trend for both melatonin groups to improve alertness and cheerfulness over placebo (repeated measures ANOVA, p=0.06). ý Melatonin administration days. 4 Bright light exposure days
208
ý .. +
rA V2 0 ý
d ý. ý aý ý ý. + 0 0 ý
fl., = lea "0 a aa v, d dý sm ý" ea
äo ci
c> . IM Cý
ýý C =""
r ý ý s, ý bý
(j"ý
ý O ý
(OZ- I `ajEas a0oput, junstn) 2llTILi ssaü). IajE antIaafqns
209
aý .ý ý aý c`ý Cý
, --r OA O
0 o '-' oA N
0 N
0
N
O' A
00 Q
14t ^ +r b. 'ý1' 'Ö
,ý .b. -' ä
ý
10 a rý bE
Ov
ý+ y ý+
;' "++ G, 2 ý O ý "ý d °" ý w$ E
.2 .ýö 23 -0 ä
rn E co ý, ý,
'b M
ce ýe 4. A
ä . -ý f QI "A 4) .c -v
Nüo ö ^ý ý
"° c^ö
äýýý
r,
Figure 6.8 Influence of melatonin and bright light on adaptation of the alertness cosinor rhythm to a rapid 9h advance phase-shift
loo- 80- 60-
1c0ý 40
20-
22: 00-
20: 00-
18: 00-
16: 00-
14: 00
a)
k 6
ý
II
0-i 70- 60- 50- 40- 30- 20- 10-
0
b) ýS
r-- --
ý: A
'
n^
I
C)
m
i
I
D7-9 D9-11 Baseline
Day of study
a) Cosinor-derived self-rated alertness acrophase estimates using 3-day windows during adaptation to a rapid 9h advance phase-shift (on day [D ] 6) under 4 different treatments: melatonin (MT, AL ), Placebo (P, 0 ), melatonin + bright light (MT+BL, ") and bright light (BL, ). An estimate of reentrainment (to a 24h cosine rhythm) is
provided by (b) % number of subjects with significant cosine fits (set at p<0.05) and (c) % variance accounted for by a cosine curve. Complete reentrainment was evident with MT and MT+BL treatments. Lack of fit and low % rhythm precluded reliable estimates of reentrainment for BL and P treatments, see Figs. (b) and (c). Values for Figs. (a) and (c) represent mean +/- SEM (n=7). D7-9: treatment period, D9-1 1: post- treatment.
210
Figure 6.9 Influence of melatonin and bright light on the adaptation of SAM-1 performance to a rapid 9h advance phase-shift
E
aý H G". 0
aý
102-
98-
>, ý 94- ý ý A
90-
C)
III 6789 10 11
Day of study
Slower response
times
Daily mean performance measures for SAM-1 (a) efficiency, (b) accuracy and (c)
response time during adaptation to a rapid advance phase-shift (on day 6) under 4 different treatments: melatonin (A ), placebo (Q ), melatonin + bright light (9) and bright light ( ). Significant treatment effect for efficiency only (p<0.05, repeated measures ANOVA, D7-11) where placebo treatment was significantly worsethan MT (p<0.01) and MT+BL and BL treatments (p<0.05), Duncan's multiple range post-hoc.
ý Melatonin administration days. 4 Bright light exposure days. Values represent mean +/- SEM, n=7.
211
Figure 6.10 Influence of melatonin and bright light on the adaptation of SAM-5 performance to a rapid 9h advance phase-shift
ý
110-
105-
100-
95-7
90
85
80=
80
109
--ý 105- 101-
0 ýo 97
93 89- 85-
Iý4 8
ý-ý -¬
T-T I
9 10 11
11 6789 10 11
Day of study
Slower response
times
Daily mean performance measures for SAM-5 (a) efficiency, (b) accuracy and (c)
response time during adaptation to a rapid advance phase-shift (on day 6) under 4 different treatments: melatonin (& ), placebo (o ), melatonin + bright light ( ") and bright light ( ). Significant treatment effect for efficiency (p<0.005, repeated measures ANOVA, D7-11) where placebo treatment was significantly worse than MT and MT+BL (p<0.01) and BL treatments (p<0.05), Duncan's multiple range post-hoc. ý Melatonin administration days. Q Bright light exposure days. Values represent mean +/- SEM, n=6.
212
Figure 6.11 Influence of melatonin and bright light on the adaptation of SAM-1 performance cosinor rhythm to a rapid 9h advance phase-shift
02: 00
22: 00
is: ooý
1a: oo- lnn-ý
j> --
b) f
'i I-- T
JD' 16.. b CD
Vý
O "y
ý
ý
\° 0
avv ýýº_
80- 60-
f---" 40- 20-
o-t- 60 C) 50 40 30 20 10
0
I I
D7-9 D9-11 Day of study
i m
i
Baseline
a) Cosinor-derived SAM-1 response time acrophase estimates using 3-day windows during adaptation to a rapid 9h advance phase-shift (on day [D ] 6) under 4 different treatments: melatonin (MT, A ), Placebo (P, Q ), melatonin + bright light (MT+BL, ") and bright light (BL, ). An estimate of reentrainment (to a 24h cosine
rhythm) is provided by (b) % number of subjects with significant cosine fits (set at p<0.05) and (c) % variance accounted for by a cosine curve. Significant reentrainment was evident with MT and MT+BL treatments. Lack of fit and low % rhythm precluded reliable estimates of reentrainment for BL and P treatments, see Figs. (b) and (c). Values for Figs. (a) and (c) represent mean +/- SEM (n=7). D7-9: treatment period, D9-11: post-treatment.
213
Figure 6.12 Influence of melatonin and bright light on the adaptation of SAM-5 performance cosinor rhythm to a rapid 9h advance phase-shift
24: 00-1
22: 00
zo: oo1 18: 00 16: 00 1 14: 00
H 100 ý2 80 H .ý t± tA
J a
$ö 60
,u". ý 40
0 20 0
0- 60- 50- 40-
20- 10-
0
b) A 0-A
C)
I
. oý
I
D7-9 D9-11
Day of study
i S m
i
Baseline
a) Cosinor-derived SAM-5 response time acrophase estimates using 3-day windows during adaptation to a rapid 9h advance phase-shift (on day [D ] 6) under 4 different treatments: melatonin (MT, A ), Placebo (P, o ), melatonin + bright light (MT+BL, ") and bright light (BL, ). An estimate of reentrainment (to a 24h cosine rhythm) is
provided by (b) % number of subjects with significant cosine fits (set at p<0.05) and (c) % variance accounted for by a cosine curve. Significant reentrainment was evident with MT and MT+BL treatments. Lack of fit and low % rhythm precluded reliable estimates of reentrainment for BL and P treatments, see Figs. (b) and (c). Values for Figs. (a) and (c) represent mean +/- SEM (n=6). D7-9: treatment period, D9-11: post- treatment.
214
Figure 6.13 Influence of melatonin and bright light on the adaptation of sleep to a rapid 9h advance phase-shift
Better Quality
Longer Duration
Longer Latency
789 Night of study
10 11
Subjective sleep measures of (a) sleep quality (SQ), (b) duration of night awakenings (DNA) and (c) sleep latency during adaptation to a rapid advance phase- shift (on day 6) under 4 different treatments: melatonin (A), placebo (O), melatonin + bright light (" ) and bright light ( ). Significant treatment effect for SQ and DNA (p<0.01, repeated measures ANOVA, nights 6-11) where both melatonin treatment groups improved ratings significantly over placebo and bright light (p<0.05, Duncan's post-hoc). ý Melatonin administration days. Q Bright light
exposure days. Values represent mean +/- SEM, n=7.
215
6.5 DISCUSSION
Time-zone travellers and shift workers are often exposed to natural bright light
at times which may oppose the quickest adaptation of the circadian system. For
example, after flying east to a destination 9 time-zones away, the quickest adaptation
is generally to phase-advance the circadian system. However, working or relaxing
in bright light the following morning may delay the biological clock according to
phase-response curve theory (Czeisler et al, 1989; Minors et al, 1991, van Cauter et
al, 1994, section 1.4.3, Figure 1.5), preventing the quickest adaptation, and
prolonging the effects of "jet lag" in some individuals whose natural direction of
entrainment is to advance. Eastman and Miescke (1990) described how evening light
exposure of about approximately 2000 lux facilitated the entrainment of the
temperature rhythm in 74% of the subjects to a 26h sleep/wake schedule but only
25% of the subjects adapted with morning light exposure of the same intensity.
Therefore, natural bright light exposure may oppose the phase-shifting effects of
melatonin and negate any beneficial influence of melatonin on the circadian clock and
behaviour.
One of the aims of this study was to investigate the effects of melatonin timed
to facilitate the reentrainment of circadian rhythms by phase-advancing and exposure
to sufficiently intense morning bright light suitably timed to phase-delay. Of the four
treatments, morning bright light alone phase-delayed the aMT6s rhythm relative to
placebo in 6 out of 7 subjects. No such effect could be shown with the temperature
rhythm, presumably due to masking. Interestingly, bright light treatment did not
improve or worsen sleep and even improved daily mean performance scores and
216
mood ratings compared to placebo. However, this effect proved transient, in most
cases ameliorating mood and performance only over the treatment period. In
addition, the alertness and performance response time rhythms did not reestablish
normal baseline patterns throughout the treatment and post-treatment periods. This
suggests that the improvements seen with bright light were most likely due to direct
psychostimulating effects (Campbell and Dawson, 1990; French et al, 1990; Myers
and Badia, 1993). The combination of melatonin and bright light treatment was as
effective as melatonin treatment alone at improving behaviour. In summary, bright
light, which phase-delayed the aMT6s rhythm, did not cause any great detrimental
effects on performance, sleep or mood relative to placebo and even had transient
beneficial effects on mood and performance.
The responsiveness of subjects to the phase-shifting effects of melatonin was
variable. The aMT6s rhythm delayed in 2 subjects with melatonin treatment whereas
the others advanced irrespective of the individuals natural direction of reentrainment.
Other studies have also reported inconsistent effects of melatonin on the biological
clock. Arendt et al (1985a) studied the effects of 2mg melatonin given at 1700h daily
for one month in spring or 3 weeks in autumn, without environmental control. The
endogenous melatonin rhythm (on the last day of the dose) was found to advance in
only 5 out of 11 subjects in Autumn. However, the endogenous and exogenous
components could not be easily distinguishable. In spring, the melatonin rhythm (on
the day after the last dose) advanced in only 2 out of 12 subjects. Mallo et al (1988)
found no phase-shift in the endogenous melatonin rhythm during administration of
8mg at 2200h for four days, but found an advance 3 days after the final melatonin
dose. Such inconsistencies may be due to the exposure of natural bright light
217
opposing the phase-shifting effect of melatonin in the latter studies or individual
differences in circadian clock time. In the present study, there was no indication of
significant differences in individual circadian clock time as shown by similar
temperature minimums and aMT6s acrophases during the baseline and at the end of
the gradual delay phase-shift.
Using the melatonin phase-response curve (PRC) of Lewy et al (1992) as a
guide (see Figure 1.5), melatonin was first given approximately 7.5 hours prior to the
expected melatonin onset time in plasma which should have fallen within the phase-
advance portion of the PRC for every individual. A recent report by Zaidan et al
(1994) also predicts an advance phase-shift with melatonin administration at the time
used in the present study.
Previous studies have shown improvements in sleep patterns and alertness with
either no effect on performance (Arendt et al, 1987; Samel et al, 1991) or small
deleterious effects (Folkard et al, 1993) with melatonin treatment. However, baseline
differences and learning effects in these experiments prevented studying performance
variables effectively. In the present study, important improvements in performance
were observed with melatonin treatment after a large phase-shift in external time
cues.
Melatonin, timed to theoretically phase-advance its own endogenous rhythm
was consistently effective at improving sleep, mood and performance parameters even
in the presence of bright light exposure timed to counteract the melatonin treatment.
Using cosinor analysis, alertness and response time rhythms completely reestablished
normal baseline patterns within 1-2 days (during treatment) with both melatonin and
melatonin + bright light treatments unlike bright light alone and placebo treatments
218
which took at least 5 days.
In contrast, there was no significant adaptation of the aMT6s rhythm
immediately after treatment withdrawal, although a significant time by treatment
interaction (ANOVA) in the raw aMT6s data indicated that both melatonin groups
favoured faster reentrainment over placebo and bright light treatments. Thus it seems
likely that the beneficial effects of melatonin on sleep and performance may not be
mediated through the entrainment of the endogenous "clock" and all circadian
rhythms. This is also suggested in a study where melatonin stabilised sleep patterns
and improved mood in a blind man without entrainment of the temperature or cortisol
rhythms (Folkard et al, 1990).
Moreover, the improvements in behaviour observed in this study were
obtained without a pre-shift (i. e. `pre-flight') treatment paradigm as used in other
studies (Arendt et al, 1986; Arendt and Aldhous, 1988; Arendt et al, 1987; Petrie et
al, 1989; Samel et al, 1991) and only 3 days post-shift (i. e. `post-flight') treatment.
In field studies where melatonin treatment was administered only after the flight, self-
rated jet lag still improved (Claustrat et al, 1992; Petrie et al, 1993). This treatment
schedule becomes of great practical benefit since daytime treatment prior to travel
may give rise to problems of increased drowsiness in some individuals (Lino et al,
1993, see Chapters 9-10 and Table 2.1).
In summary, all subjects responded similarly with immediate improvements
in behaviour with melatonin alone and when used in combination with bright light,
despite 2 subjects who phase-delayed and 5 who phase-advanced. Disturbances in
sleep which are the most common complaint of shift workers on rotation (Kogi, 1985)
may be the cause of the general feeling of discomfort, tiredness and poor performance
219
common to jet lag and shift work more than disturbances in circadian rhythms. The
beneficial effect of melatonin on mood and performance coincides with improvements
in sleep and may be directly related to sleep quality. However, an effect on the
timing mechanisms of sleep, performance and mood cannot be excluded since the
improvements were not transient.
In conclusion, suitably timed melatonin treatment significantly alleviates
behavioural disturbances associated with large shifts of external time cues in the
presence or absence of bright light timed to counteract the phase-shifting properties
of melatonin. This effect of melatonin is not wholly mediated through an effect on
the biological clock or the reentrainment of all circadian rhythms.
220
CHAPTER 7
PURIFICATION OF THE CORE BODY TEMPERATURE
RHYTHNI AFTER SIMULATED TIME ZONE TRAVEL
Purification of the core body temperature rhythm after simulated time-zone
travel
7.1 INTRODUCTION
The circadian rhythms of melatonin and its urinary metabolite aMT6s, together
with core body temperature are commonly used phase markers of the endogenous
body clock. However, these rhythms are susceptible to powerful masking (disturbing)
influences which can distort the rhythm and result in the unreliable estimation of
circadian time. Although the melatonin/aMT6s rhythms appear to be minimally
affected by exogenous influences apart from light and posture (Lewy et al, 1980;
Deacon et al, 1994), core body temperature is susceptible to many exogenous factors.
For example, sleep causes body temperature to fall (Mills et al, 1978b; Wever,
1985b; Minors and Waterhouse, 1987) and activity, upright posture and bright light
increases temperature (Moog and Hildebrandt, 1989; Dawson and Campbell, 1991;
Minors and Waterhouse, 1989; Myers and Badia, 1993).
Measured rhythms are therefore the product of two components. One is
governed by the endogenous body clock, the other is dependent upon exogenous
influences, often described as "masking" effects. The size of the masking effect will
depend upon the relative strengths of the endogenous and exogenous components
acting on the circadian rhythm. A rhythm which is highly susceptible to exogenous
factors (such as the body temperature rhythm) will apparently adapt faster to changes
in local time cues than rhythms which are more strongly influenced by the slow
reentrainment of the internal body clock (such as the melatonin rhythm).
222
Attempts to separate these components may therefore provide more reliable
information about the behaviour of the endogenous body clock. There are two main
techniques for separating exogenous influences from the endogenous component. The
first involves a "constant routine" procedure which eliminates or at least minimises
exogenous masking effects. Subjects are kept awake for at least 24h in an
environment of constant lighting, ambient temperature and humidity, and are given
identical and small snacks at frequent and regular intervals in either a recumbent
(Czeisler et al, 1989) or sedentary position with minimal movement (Mills et al,
1978a).
The second procedure involves changing the daylength in isolated
environments free from everyday external time cues. There is a limited daylength to
which the endogenous body clock can normally entrain to (23-27h) beyond which it
will free-run with its natural endogenous m 25h period (Wever, 1983). By extending
or shortening the day beyond this range of natural entrainment, it will effectively
separate the two components. For example, requiring subjects to live on a 21h day,
will allow the endogenous rhythm to free-run with a periodicity of approximately 25h.
By observing the exogenous component (with a periodicity of 21h) over all phases of
the endogenous cycling rhythm, it is possible to determine the degree of masking due
to activity, sleep, posture and meals, etc (Wever, 1979; Minors and Waterhouse,
1982).
The constant routine procedures are impractical and laborious to study
circadian rhythms on a day to day basis when investigating adaptation to time-zone
shift or shift work. One of the major problems is depriving the subjects of sleep.
However, an estimate of the masking effect can be determined by comparing the
223
circadian rhythm in a normal environment with that obtained under constant routine
conditions. The difference between the two rhythms can then be used as a measure
of the masking effect (Wever, 1985b). Then by simply subtracting the exogenous
increment, a more reliable estimate of the endogenous behaviour of the body clock
can be determined in a normal environment. This does assume that the endogenous
and exogenous components are additive and independent of each other, however.
One mathematical procedure which employs this idea has been developed by Minors
and Waterhouse (1992,1993).
It is the purpose of this chapter to describe the results obtained from using
such a mathematical model to uncover the endogenous behaviour of the body clock
after simulated time-zone travel, using the core body temperature rhythm as a
circadian phase marker. Using comparisons with the raw temperature data and
urinary aMT6s, the usefulness of this model will be assessed.
Dr J. M. Waterhouse and Dr D. S. Minors kindly performed the "demasking"
procedure on the raw temperature data. I then performed the subsequent analysis.
7.2 METHODS
The experiment is described in detail in Chapters 5 and 6. In summary, after
a 3-day baseline (days [D] 1-3), 7 healthy volunteers were initially subjected to a
gradual 9h delay phase-shift in partial isolation using a combination of exposure to
bright light and darkness/sleep (over 5 days, Dl-5). Subjects then resumed their
baseline sleep/wake cycle in a normal environment inducing a rapid 9h advance
224
phase-shift in local time cues and simulating a 9h eastward time-zone shift (on D6).
In this 4-leg crossover study the readaptation was studied under four different
treatments, 1) melatonin, 2) placebo, 3) bright light and 4) melatonin and bright light.
Core body temperature was recorded every 6 minutes throughout the whole
study and averaged into 30 minute bins for analysis.
7.2.1 Purification procedure
This "demasking" procedure relied on the individuals activity diaries, in which
they recorded the type of activity during half hour segments throughout each day (see
section 2.5). In addition, sleep logs were used to determine periods of recumbency
(lying) and sleep. The main categories used for "purification" were `sleeping',
`lying', `relaxing' (e. g watching TV, reading), ̀ sitting alert' (e. g. sitting working at
a computer or writing at desk) and ̀ light activity'( e. g. walking). Subjects refrained
from vigorous activity (e. g. playing squash). `Lying' was taken as the first 30
minutes in bed, and other times indicated on their logs as in bed but not asleep.
The principle of the purification procedure was as follows. Depending on the
type of activity that occurred during the previous half hour, temperature data were
either increased or decreased by 0.1°C increments or remained unchanged. For every
adjustment made in the temperature data, the purified 24h data set was compared with
the standard endogenous 24h profile obtained under constant routine conditions. This
standard profile was compiled by pooling data from approximately 80 subjects and
was taken to represent the normal population as a whole (Minors and Waterhouse,
1992). By superimposing the two profiles and by altering the size of the masking
225
effects by 0.1°C increments, the best-fitting profile which produced the minimum
summed square deviations between the purified data set and the standard profile was
determined.
All temperature data recorded when the subjects were lying were unchanged.
"Lying" was taken to be the standard postural position/activity where no masking
effects were present since the standard endogenous profile was obtained when the
subjects were inactive and recumbent in constant routine conditions. Temperature
data recorded when subjects were sleeping were increased and data recorded when
subjects were relaxing (but sitting/erect), sitting alert or walking were reduced.
This procedure was performed a number of times to cater for all the
combinations of possible fits. For example, under the `sleep' category, every time
the subject recorded being asleep (which depresses body temperature), the
temperature data were increased in 0.1°C increments from 0 to 1.0°C. This means
that the purification procedure needed to be performed 10 times for this parameter.
For temperature data recorded whilst walking (light activity category, corresponding
to an elevation in temperature), values were reduced from 0 to -1. IOC. This means
that the purification procedure needed to be performed 11 times for this category.
Therefore, just using these two categories the purification procedure was performed
10 x 11 times to determine the best-fit. This was repeated for every subject for every
day of the trial. All five categories were used.
226
7.3 STATISTICAL ANALYSIS
Cosinor analysis was used to determine the rhythm parameters of acrophase
and amplitude (section 2.8) for both the raw "masked" and purified temperature
profiles.
In order to obtain more reliable estimates, for each individual the cosinor-
derived parameters were grouped on identical days during baseline and gradual delay
phase-shift over the 4 legs of the study for body temperature under masked conditions
and after purification, and for aMT6s for comparison.
7.3.1 Baseline (D-2 to DO)
Differences in purified temperature acrophase and amplitude estimates over
the baseline were assessed by one way ANOVA. Comparisons between temperature
parameters before and after purification were analysed by Student's paired t-test to
determine the effect of purification on acrophase and amplitude values obtained under
normal entrained conditions.
7.3.2 Gradual 9h delay phase-shift (D1-5)
Acrophase and amplitude values for masked and purified temperature rhythms
(and aMT6s acrophases) were calculated as a difference from individuals' mean
baseline. Subjects were then grouped to compare (a) degree of shift between purified
temperature, raw temperature and aMT6s rhythms and (b) temperature amplitude
227
values before and after purification, using repeated measures ANOVA (factors: day
of study and type of rhythm). Regression correlation was used to determine any
relationship in the rate of adjustment between the rhythms.
7.3.3 Rapid 9h advance phase-shift (D6-11)
Comparisons of the normal rate of adaptation between aMT6s, purified
temperature and masked temperature were assessed under placebo treatment.
Due to the uneven number of subjects with significant cosine fits (and with
variable estimates of acrophase, e. g. see Table 7.1) repeated measures ANOVA was
not appropriate. One way ANOVAs were used to determine differences in the degree
of adaptation between the aMT6s rhythm, and the temperature rhythm before and
after purification across each day. The effect of treatment on adaptation of the
purified temperature rhythm was determined in a similar way.
228
7.4 RESULTS
Only the 24h data sets with significant cosine fits (p<0.05) were used. 86%
of the purified 24h data sets were significant at p<0.05 compared to 99.5% of the
masked temperature data.
7.4.1 Normal baseline conditions
No significant differences in the purified or raw temperature estimates of
acrophase and amplitude were apparent over the baseline. The purified estimates of
acrophase were significantly phase-delayed by 1.15 ± 0.13h relative to the raw
estimates (p<0.02, masked rhythm = 16.16 ± 0.06h, purified rhythm = 17.31 ±
0.15h, mean ± SEM, n =7). The amplitude of the rhythm was significantly reduced
by approximately 50% after purification (p<0.0001, masked = 0.52 ± 0.02°C,
purified = 0.26 ± 0.02°C, mean ± SEM), Figure 7.1.
7.4.2 Adaptation to a gradual 9h delay phase-shift
The purified temperature rhythm was significantly slower to adjust to the
gradual 9h delay phase-shift than the masked temperature rhythm by an average of
0.7 ± 0.4h per day and the aMT6s rhythm by 1.05 ± 0.31h per day (D1-5, mean
± SEM, treatment and time by treatment interaction effect, p<0.05), Figures 7.2
and 7.3. The maximum degree of shift in the purified temperature rhythm (compared
229
Figure 7.1 Purification of the core body temperature rhythm
11-. U 37.8
37.4
37
36.6
36.2
35.8
-0.54
Light activity - Sitting alert
Relaxing
Lying
Sleeping
c)
24: 00 08: 00 16: 00 24: 00
Time of day
r-77-'r'" 08: 00
Purification of the core body temperature rhythm for one subject during a baseline day in a normal environment showing (a) the temperature rhythm under masked conditions () and after purification (-), (b) the degree of masking effect from
sleep and activity and (c) subjects activity diary. Arrow indicates calculated peak time of the rhythm.
230
Figure 7.2 Comparison of the rate of adjustment between the aMT6s and temperature (raw and purified) rhythms to a gradual 9h
delay shift
10-
81 ý
ý ý6 ý aý b ý.. ý - 0
-
I
2
0 IIIIII
1 234 Day of study
5 6
Degree of delay phase-shift (from acrophase estimates) in the aMT6s rhythm (f ), and the core body temperature rhythm under masked conditions (0 )
and after purification (U) during adaptation to a gradual 9h delay phase- shift (days 1 to 5). The purified temperature rhythm was significantly slower to adjust than the masked temperature rhythm by a mean (+/- SEM) 0.7 +/- 0.4h per day and the aMT6s rhythm by 1.05 +/- 0.31h per day (time by treatment interaction, p<0.05). Values represent mean +/- SEM, n=7 subjects.
231
Figure 7.3 Relationship between the rate of adaptation of purified and raw temperature rhythms and aMT6s rhythm to a gradual 9h delay
phase-shift
a) r-0.887, p<0.0001 y=1.03 (+/- 0.09) + 0.56 (0.53)
b) r-0.765, p<0.0001 y--0.77 (+/- 0.11) + 0.37 (0.74)
IIIII
02468 10 24
Degree of shift in the purified temperature rhythm (h)
ýý vý ý
ýý4ýý - 47
6
Lt1m/" r-0.921, p<0.0001
8 10
21 1 jýb y=1.07 (+/- 0.08) -0.77 (0.52)
1
0 2468 10
Degree of shift in the aMT6s rhythm (h)
Relationship between the degree of shift during the gradual 9h delay phase-shift of (a)
purified and raw temperature rhythms, (b) purified temperature and aMT6s rhythms and (c) raw temperature and aMT6s rhythms. Regression values inside brackets are +/- SEM.
232
to mean baseline) on D5 was 8.14 ± 0.38h and D6 was 8.98 ± 1.33h (mean ±
SEM). There were no significant differences in the degree of shift obtained between
aMT6s, purified temperature and raw temperature rhythms on D6 with only a
tendency for the purified temperature rhythm to be slightly delayed compared to the
raw temperature and aMT6s rhythms on D5 (p=0.06).
The temperature rhythm amplitude under masked conditions was significantly
reduced by 0.18 ± 0.02°C (D1-5, p <0.005) relative to the purified estimates, Figure
7.4. After purification the rhythm amplitude showed no significant deviation from
baseline.
7.4.3 Adaptation to rapid 9h advance shift
The masked temperature rhythm apparently adjusted to a rapid 9h advance
phase-shift in one day whereas even after 5 days (D11) the purified temperature and
aMT6s rhythms were not fully adjusted (p<0.05, significantly different from
individuals baseline, n =7), Figure 7.5, Table 7.1.
233
Figure 7.4 Effect of purification on the temperature rhythm amplitude during a gradual 9h delay phase-shift
I I 012345
Day of study
-0.24
Amplitude of the core body temperature rhythm under masked conditions (0)
and after purification ( ) during adaptation to a gradual delay 9h phase-shift (days 1 to 5). The temperature rhythm amplitude under masked conditions was significantly reduced compared to the rhythm after demasking (ANOVA, p<0.005). Values represent mean +/- SEM, n=7.
234
Figure 7.5 Effect of purification on the rate of adjustment of the temperature rhythm to a rapid 9h advance phase-shift
i
0
2-
4-
6
8-
loý
456789 10 11 Day of study
I IIIII
Degree of advance phase-shift (from acrophase estimates) in the aMT6s rhythm (f ), and the core body temperature rhythm under masked conditions ( 0) and after purification ( ) during adaptation to a rapid 9h advance phase-shift (on day 6). 5 days after the shift (day 11) the purified temperature and aMT6s rhythms were not fully adjusted (p<0.05) unlike the masked temperature rhythm which was apparently adjusted by day 7/8. Values represent mean +/- SEM (only data from subjects with significant (p<0.05) cosine fits used, n=7 for masked temperature and aMT6s, for purified temperature mean n=5.4 +/- 1.7 (+/- SD)).
235
Table 7.1. Individual difference in purified temperature acrophase estimates (on
hours) from mean baseline (D-2 to DO) during adaptation to a rapid 9h advance
phase-shift (on day 6)
Subject
Day of study ABCDEFG
6 9.97 9.4 6.55 9.5 7.9 10.4 9.13
7 11.77 ns ns ns 2.8 ns 0.73
8 5.37 ns 8.35 3.7 2.8 ns ns
9 2.57 10.4 0.35 7.9 6.2 3.3 0.63
10 ns ns ns 8.1 8.6 10.1 0.83
11 4.57 6.8 2.65 7.9 3.8 1 1.13
ns: non-significant fits
The temperature rhythm amplitude under masked conditions apparently did not
adjust properly until the fourth and fifth days after the advance shift (D10 and D11),
Figure 7.6. After purification amplitude estimates were not significantly affected.
Of the four treatments, melatonin and melatonin + bright light treatments
facilitated the reentrainment of the purified temperature rhythm with complete
reentrainment by D II. Bright light and placebo treatments remained delayed.
However, no significant treatment effect could be determined. No treatment
differences in rhythm amplitude were apparent.
236
Figure 7.6 Effect of purification on the rate of adjustment of the temperature rhythm amplitude to a rapid 9h advance phase-shift
o. i ý Ö 0-
, ý. -0.2- 0
-0.37 ä
-0.4 I I IIIII 456789 10 11
Day of study
The amplitude of the core body temperature rhythm under masked conditions (0) and after purification ( ) during adaptation to a rapid 9h advance phase- shift (on day 6). Values represent mean +/- SEM (only data from subjects with significant (p<0.05) cosine fits used, n=7 for masked temperature and aMT6s, for purified temperature mean n=5.4 +/- 1.7 (+/- SD)).
237
Figure 7.7 Effect of melatonin and bright light on adaptation of the purified body temperature rhythm
04
02ý
24-
22
20-
18-
16-
14 I -FE II ---F
56789 10 11
Day of study
4
T
mean baseline
Acrophase estimates for the purified core body temperature rhythm during
adaptation to a rapid 9h advance phase-shift (on day 6) under 4 different treatments: melatonin (A), placebo (O), melatonin + bright light (" )
and bright light (m). Values represent mean +/- SEM (only subjects data
with significant (p<0.05) cosine fits used, mean n= 5.7 +/-1 (+/- SD)). No
sý ficant treatment effect (ANOVA). Bright light exposure. = Melatonin administration.
238
7.5 DISCUSSION
There are some assumptions of this purification procedure which are described
in detail by Minors and Waterhouse (1992) and are outlined here. The endogenous
component of the rhythm is considered to reflect the internal pacemaker whereas the
exogenous component is assumed to be largely a result of the subjects' lifestyle. Both
are independent and the exogenous component amplifies the endogenous rhythm. It
is assumed that every individuals endogenous circadian rhythm reflects the standard
endogenous rhythm derived from constant routine experiments in period and shape,
i. e. one which closely represents a cosine curve. Furthermore, masking effects are
assumed not to be dependent on circadian phase. However, there is some evidence
to suggest that the masking effect of sleep on temperature is greater on the descending
phase of the endogenous temperature rhythm than at other times (Wever, 1985b;
Mills et al, 1978b). The effects of posture, light activity and exercise at different
circadian phases are considered to be small or insignificant (Moog and Hildebrant,
1986; Minors and Waterhouse, 1989).
7.5.1 Normal baseline conditions
In entrained (baseline) conditions, the exogenous component accounted for
approximately 50% of the temperature rhythm amplitude, suggesting that the ratio
between the endogenous and exogenous components is approximately 1: 1 in normal
entrained conditions. This ratio agrees with that determined by Folkard (1989). The
phase of the rhythm was also delayed after purification. The endogenous temperature
239
rhythm peaked around 1700h (in the presence of little or no masking effects).
However, elevated temperature values due to activity during the day especially in the
morning caused the cosine curve to fit slightly differently such that the calculated
peak of the rhythm was estimated to be earlier in the day, for example see Figure
7.1.
7.5.2 Adaptation to a gradual 9h delay phase-shift
During the slow 9h delay phase-shift, designed so that behavioural and
physiological rhythms shifted in synchrony, the adjustment of the purified temperature
rhythm was slower compared to the masked rhythm. The imposed delay shift in the
sleep/wake cycle delayed the exogenous masking effects such that activity occurred
during times when body temperature was naturally low (elevating the values) and
sleep occurred during times when body temperature was naturally high (lowering the
values). This effectively shifted the acrophase estimates in the direction of the
imposed phase-shift. The degree of masking due to sleep and activity could be
clearly observed in the amplitudes of the temperature rhythm before and after
purification. After purification no significant reduction in the amplitude of the
temperature rhythm was observed compared to the raw temperature profiles.
The maximum degree of shift induced was assessed on the last day of the
gradual delay phase-shift. However, comparison with the shift in the aMT6s rhythm
may be more reliable on day 6 since the masking effects of bright light on melatonin
concentrations (e. g. Lewy et al, 1980) and body temperature (French et al, 1990;
Myers and Badia, 1993) may have influenced acrophase estimates during the shift.
240
On this day (D6) subjects were exposed to conditions of dim light (<200 lux) or
darkness. It is assumed that the magnitude of the actual induced delay shift in the
aMT6s rhythm was greater than or equal to the measured shift on this day. The
degree of shift was comparable to that obtained for both the purified and raw
temperature rhythms, indicating that 9h shifts can be induced in major endogenous
physiological circadian rhythms without disruption using this experimental model.
7.5.3 Adaptation to a rapid 9h advance phase-shift
After the rapid 9h advance phase-shift (simulated eastward time-zone shift) and
observing the readaptation without treatment, masking caused the raw temperature
rhythm to reentrain within one to two days. However, it was clear that the rhythm
amplitude had not established its normal baseline pattern even after 5 days. This
could be explained by the inappropriate phase relationship between the endogenous
and exogenous components of the temperature rhythm as explained above. The
temperature rhythm after purification showed slow apparent readaptation similar to
the aMT6s rhythm. Of the known masking influences on the aMT6s rhythm, the
masking effects of ambient bright light were controlled and considered minimal
during readaptation (intensity of light <500 lux). The rate of adjustment of the
aMT6s rhythm was therefore considered to be a reliable indication of circadian time
and an appropriate reference rhythm marker for the temperature purification method.
The aMT6s rhythm adjusted at a steady and slow rate. By the fifth day the
rhythm had only readapted by approximately 2.5h (with great individual variations,
241
see Figure 5.18). In comparison, after purification the temperature rhythm apparently
"phase jumped" by about 4h immediately after the rapid advance shift followed by
a more gradual adaptation. After five days (D 11), the purified temperature rhythm
had adjusted by approximately 5h. When this temperature purification procedure was
used to determine the adaptation of nurses to night shift work, a similar "phase jump"
was observed on the first switchover day, after which a slow and gradual adaptation
to the new schedule was observed (Minors and Waterhouse, 1993). This "phase-
jump" may suggest that some masking effects were still present in these studies. In
another simulated time-zone shift, where the exogenous component was taken into
account using a similar mathematical model, no immediate "phase jump" was evident
(Folkard et al, 1991). Instead, a gradual adaptation, calculated to be approximately
30 minutes per day (2.5h after five days) was observed. By determining the
endogenous component of the temperature rhythm in this way the rate of adaptation
of the internal clock was found to be smaller compared to the rate of about lh per
day determined by other studies under normal conditions (Klein and Wegmann, 1980;
Wegmann et al, 1983), using a version of the constant routine procedure which
allows sleep (Moog and Hildebrandt, 1989) and using a mathematical simulation (Tsai
et al, 1988).
Unfortunately, in the present study it was not possible to determine accurately
the rate of adjustment to the rapid 9h advance phase-shift due to the lack of subjects
with significant cosine fits and highly variable acrophase estimates. A number of
reasons may account for this. Firstly, not all masking effects may have been
removed. The categories of masking used concentrated on activity/sleep and posture.
Other masking factors may be important. For example, the masking effects of hot
242
showering or bathing, stress and heavy meals on body temperature has not been fully
investigated. These masking effects may well be attenuated in normal life compared
to the effects of activity and sleep. However, the purification method reduces the
amplitude of the rhythm and hence weak masking factors under normal conditions
may have more pronounced effects on cosinor-derived rhythm parameters after the
stronger masking factors are removed. The purification process is therefore
influenced by the number of categories of masking used. As Minors and Waterhouse
(1992) suggest, by increasing the number of categories used overestimation of the
degree of shift is less likely to occur.
However, even if all sources of variation due to masking were removed, the
chance of error from other sources such as temperature measurement and inaccurate
activity diaries still exist and would result in more erratic acrophase estimates. There
were indeed cases where the activity diaries did not coincide with the raw data.
Computerised activity monitors may provide a more reliable reference source for
future experiments. A more reliable estimate of circadian time was achieved by
averaging individuals acrophase and amplitude values on the same days during
baseline and the gradual delay shift over the 4-leg crossover study.
It was difficult to assess correctly the cosinor-derived parameters of the
purified temperature rhythm during the treatment period due to the temperature-
elevating effects of bright light (French et al, 1990; Myers and Badia, 1993) and
temperature-lowering effects of melatonin (Chapter 9 and 10). Treatment effects on
the aMT6s rhythm were compared after exogenous melatonin had cleared, on days
10-11. The purified temperature rhythm exhibited a similar acrophase profile to that
observed with the aMT6s rhythm for each treatment, Figure 6.2. Under bright light
243
and placebo treatments the endogenous temperature rhythm remained phase-delayed,
whereas both melatonin treatment groups facilitated the adaptation of the temperature
rhythm with complete adjustment on D11. These results, however, do remain
tentative for the reasons outlined above. For a detailed account of the effects of
treatment on adaptation to a rapid 9h advance shift, see Chapter 6.
There is little doubt that masking causes apparent overestimates of the degree
of shift leading to inaccurate estimates of circadian time. An example of the
problems encountered in interpreting masked data comes from fractional
desynchronisation studies where the daylength was progressively shortened or
lengthened until the endogenous rhythms free-ran (Wever, 1983; Folkard et al, 1984;
Minors et al, 1988). Such protocols allowed deductions as to whether a particular
endogenous rhythm was more likely to be governed by the "strong" oscillator with
a limited range of entrainment or by a weaker pacemaker which has a much greater
range of entrainment. Temperature apparently adjusted more easily to longer or
shorter days than the urinary melatonin rhythm (Folkard et al, 1984; Minors et al,
1988). However, taking into account the exogenous component, the temperature
rhythm followed a similar path to the aMT6s rhythm (Minors and Waterhouse, 1992).
Although constant routine protocols are effective in minimising masking
influences, they are inappropriate for studying the detailed adjustment of circadian
rhythms in every day life in the presence of normal zeitgebers. A major problem of
constant routine protocols is sleep loss. It is not known whether this effect or indeed
the procedure itself imposes a masking effect. It is quite possible that fatigue
associated with sleep loss may affect subjective ratings of sleep and mood and
244
performance measurements. Moog and Hildebrandt (1986) have, however, adapted
the constant routine protocol to allow sleep when desired.
In conclusion, masking can lead to overestimations of the degree of adaptation
of the body clock using the core body temperature rhythm as a circadian phase
marker. This method of temperature purification should prove useful by allowing a
more accurate estimation of circadian time of the endogenous body clock in normal
everyday life.
245
CHAPTER 8
INDUCING CIRCADIAN RHYTHM DISTURBANCES IN
PARTIAL ENVIRONMENTAL ISOLATION: ADAPTATION
TO A 9h DELAY PHASE-SHIFT AFTER A GRADUAL 9h
ADVANCE SHIFT
Inducing Circadian Rhythm Disturbances in Partial Environmental Isolation: Adaptation to a Rapid 9h Delay Phase-Shift After a Gradual 9h advance Shift.
8.1 INTRODUCTION
In field studies, the length of time required for readjustment tends to be faster
after westbound transmeridian travel, which requires the circadian system to delay,
than eastbound travel requiring a phase-advance. An average reentrainment speed of
lh per day when travelling east and 1.5h per day when travelling west over 5-11
time-zones has been estimated in field conditions (Aschoff et al, 1975; Klein and
Wegmann, 1980). These are estimates encompassing different circadian rhythms,
conditions, subjects and the whole readaptation period. However, this direction
difference seems to be independent of whether the flight was outward bound or
homeward bound or whether the flight occurred at night or during the day (Klein,
1970; Wegmann et al, 1970; Klein et al, 1972a, 1972b).
The present study was designed to simulate a rapid 9h westward bound
transmeridian flight using a similar experimental model to that described in Chapter
5 but in the opposite direction. The readaptation of various circadian rhythms and
behavioural functions were studied in detail in a normal environment and compared
to the simulated 9h eastward transmeridian flight described before.
This trial was a collaborative project in association with final year project
students L. Bhattacharjee, A. Curnow, N. Lawrence, Z. Wright, A. Tasoula. It
included a separate study investigating the response of various gut hormones to a set
meal before and after a 9h advance phase-shift. Sampling details from the latter study
247
will be provided since frequent and heavy blood collections were required and were
important considerations when interpreting behavioural measurements.
I will report here on the analyses I have performed.
8.2 SUBJECTS and METHODS
8.2.1 Subjects
6 healthy volunteers (3M: 3F), aged 20-24 years (mean ± SD, 20.7 ±2
years), under no medication and all non-smokers, followed similar sleep/wake
schedules and lifestyles prior to the trials. They refrained from heavy exercise and
restricted their alcohol consumption (maximum 2 units per day) throughout the
experiment.
8.2.2 Study Design
-3
The design was as follows:
-2 -1 0
Baseline
Regular sleep/wake cycle and normal environment
DAYS (D)
23456 7
III : hanging 1 Stable
Slow, forced 9h advance phase-shift
8 9 10
I Readaptation in normal environment with baseline sleep/wake cycle.
Rapid 9h delay phase-shift
248
Baseline: For 4 days prior to the phase-shift (days (D) -3 to 0), the subjects were
required to maintain a regular sleep/wake schedule, retiring to bed at approximately
2330h and waking at 0730h, remaining in the dark (< 1 lux) between these times.
Gradual 9h advance phase-shit: Immediately following the baseline days, the
subjects were exposed to 10h pulses of moderately bright (average 1200 lux at eye
level), full spectrum light at progressively 3h earlier times over the first three days
(D1-3) and then at the same times on D 4-5 as for D3. The periods of light exposure
were as follows:
Dl: 0430-1430h, D2: 0130-1130h, D3,4,5: 2230-0830h.
Except for D1, each period of light treatment was preceded by 8 hours of
imposed darkness (< 1 lux) when they slept. On DI, light exposure was preceded
by 5h of darkness/sleep. During the remaining hours of each day, subjects were in
dim light (< 300 lux, indoors) or moderately bright light (< 1000 lux, outdoors with
sunglasses) until dusk (1630/1700h). Throughout the phase-shift (D1-5) meal-times
were shifted progressively earlier by 3 hours on days 1-3 and then maintained at the
same times on D4-5 as for D3, in line with the imposed light/dark and sleep/wake
cycles. Subjects were allowed to choose their activity (e. g. reading, writing,
watching videos) as long as they glanced regularly (at least every 5 minutes) at the
light source. The patterns of exposure to bright light, natural ambient light and
darkness are summarized in Figure 8.1.
249
Figure 8.1 Diagrammatic design of the study
I
I
I
I
I
I
I
I
END II I ITITII TI III IT'71Tr,
24: 00 04: 00 08: 00 12: 00 16: 00 20: 00 24: 00
Time of day
Diagrammatic design of the study showing patterns of exposure to bright light (0
,- 1200 lux), darkness/sleep (=) and natural ambient light (r m). After a4 day baseline (D-3 to DO), a gradual 9h advance phase-shift was induced over 3 days (D1-3) and then the 9h shift was maintained for the following 2 days (D4-5). On day 6, subjects resumed their baseline
sleep/wake schedule in a normal environment, thus experiencing a rapid 9h delay phase-shift of local time cues.
250
9h rapid delay phase-shift: Immediately following the last light/dark treatment on
day 6, subjects were required to assume the baseline sleep/wake schedule in their
normal environment, inducing a rapid 9h delay phase-shift in external time cues, i. e.
retiring to bed at 2330h and rising at 0730h for 5 days (D6-10). The subsequent
readaptation of the subjects in their normal environment was then observed. Subjects
were asked to avoid bright sunlight and wore sunglasses when outdoors (filtering out
>90% of the light). At dusk, they were required to remain in ordinary domestic
lighting (50-300 lux) until bedtime (and darkness).
Subjects were only allowed to sleep during the periods of darkness throughout
this study. Naps were forbidden.
8.2.3 Sampling
Urine: Sequential 4-hourly urine samples (8-hourly when asleep) were collected
throughout each stage. All urine was collected throughout each four hour period and
the bladder completely emptied at the end of that collection period. See section 2.2.
Blood: Blood samples (16mis) were collected frequently (6 x 15mins. followed by
4X 30 mins. followed by 3x 60 mins. ) from 1315h until 1945h on days 0 and 6.
Temperature: Rectal temperature was recorded continuously every 6 minutes
throughout the trial as described in section 2.5.
251
8.2.4 Behavioural Measurements
These were measured as described in section 5.2.4.
8.2.5 aMT6s assay
Urinary aMT6s was measured by L. Withey (University of Surrey) as
described in section 2.3.4. The interassay coefficients of variation were 13.2,13.5
and 12.7% at 3.3,23.7 and 57 ng/ml. The minimum level of detection was 0.21 f
0.13 ng/ml (mean ± SD).
8.3 STATISTICAL ANALYSIS
8.3.1 Temperature and aMT6s rhythms
Urine collections were not complete due to ill health of 2 of the investigators.
58% of the 24h collections were complete. D-2, D4 and D9 were used for the
analysis of urinary aMT6s since 5 subjects had complete 24h urine collections on
these days.
Repeated measures ANOVA (factors: day of study and time of day) were
applied to raw temperature (averaged every hour) and urinary aMT6s data. Cosinor-
derived parameters of acrophase and amplitude for both rhythms were subjected to
ANOVA (factor: day of study), over the baseline, gradual delay phase-shift and
readaptation periods. Degrees of shift were determined by comparison with
252
individuals mean baseline.
8.3.2 Mood and performance rhythms
These were analysed as described in section 5.3.24h profiles for each mood
and performance parameter were subjected to cosinor analysis over 1-day windows
and then 3-day windows for more reliable estimates of acrophase and amplitude
(baseline: D-3 to D-1, changing phase shift: D1-3, stable phase shift: D3-5,
readaptation period: D6-8 and D8-10).
8.3.3 Sleep data and daily means for mood and performance
Daily means for every mood and performance parameter were determined.
All sleep, mood and performance data were calculated as a percentage of, or
difference, from individuals own mean baseline (D-3 to D-1). Both DO and the
following night were omitted for this calculation since heavy blood sampling and a
shortened night may have compromised behavioural parameters. All data were then
analysed by repeated measures ANOVA (factor: day of study) over baseline (D-3 to
D-1), gradual advance phase-shift (D-3 to D5) and the rapid delay phase-shift (D6-
10), compared to baseline (D-3 to D-1).
253
8.4 RESULTS
Cosinor analysis did not consistently and reliably fit 24h profiles of mood and
performance data even during the baseline. No reliable inferences could therefore be
made regarding the cosinor-derived estimates of phase and amplitude of mood and
performance rhythms. One subject did not carry out a sufficient number of mood
ratings and performance tests to determine accurately daily means and was therefore
excluded from all mood and performance analysis.
Cosinor analysis gave 100% (at p<0.005) significant fits for 24h temperature
data. The percentage variance accounted for by a cosine curve for all aMT6s profiles
on D-2, D4 and D9 were > 80%.
8.4.1 Baseline
No significant baseline differences for any sleep, mood, aMT6s or temperature
parameters were evident (p > 0.1). A significant time of day effect (p < 0.001) was
established by ANOVA for aMT6s, body temperature and alertness (except for D-3),
Figures 8.2,8.3 and 8.4, respectively.
254
Figure 8.2 Temperature profiles before and during a gradual 9h advance phase-shift
37.60
37.40-ý
37.201 0 ý 37.00
Cd 36.80
ý H 36.60-
36.40 -
36.20 -
I
36.00 IIII
12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 D-3 D-2 D-1 DO Dl D2 D3 D4 D5 D6
Clock time/ Day of study
Circadian profiles for core body temperature over a4 day baseline (D-3 to DO) and a gradual 9h advance phase-shift (D 1-6). A significant phase-advance was indicated by ANOVA for repeated measures (time of day by day of study interaction, p<0.001, D-1
versus D5). A decline in amplitude was observed over the phase-shift (p<0.05). Bars show
patterns of exposure to bright light (1200 lux, o), darkness (< 1 lux, M) and natural ambient light (®). Temperature was recorded continously every 6 minutes then averaged and plotted every hour as the mean of 6 subjects (+ SEM).
255
Figure 8.3 Adaptation of the urinary aMT6s rhythm to a gradual 9h advance phase-shift
0730- 1130- 1530- 1930- 2230- 0330- 1130 1530 1930 2230 0330 0730
Time (hour of day)
Mean circadian profiles of urinary aMT6s for 5 subjects (+ SEM) showing phase-advance on the fourth day (D4, ) of the gradual 9h advance phase-shift compared to baseline (D-2, [3). Phase-advance indicated by repeated measures ANOVA (time of day by day of study interaction,
p<0.02). Darkness from 2330 to 0730h on D-2 and from 1430 to 2230h on D4.
256
Figure 8.4 Alertness profiles before and during a gradual 9h advance phase-shift
ý
13 ---,
12H ýo
i
c3- 11 ý UU ýý
aý
> ct Üý
cqs ý
ý
10
9H
8H
7
.T
IT
N
r I 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00
D-3 D-2 D-1 DO D1 D2 D3 D4 D5
4 Clock time/ Day of study
Daily profiles for self-rated alertness over a4 day baseline (D-3 to DO) and a gradual 9h advance phase-shift (D 1-5). = Bright light exposure (1200 lux), - darkness (< I lux), ® normal ambient light. Values represent mean (+ SEM), n=5 every 2h during wakefulness. I Blood sampling day. A phase-advance was indicated by
repeated measures ANOVA (time of day by day of study interaction, baseline versus D5, p<0.0001).
257
8.4.2 Adaptation to a gradual 9h advance phase-shift
Using individuals as their own controls the degree of advance shift in the
aMT6s rhythm was 9.16 ± 1.22h (mean ± SEM, n =5, D4 compared to D-2),
Figure 8.5. The masked temperature rhythm shifted by a similar degree, 8.63 ±
0.57h (mean ± SEM, n=6, D5 compared to mean baseline), Figure 8.5. An
advance phase-shift was also indicated with ANOVA on the raw aMT6s, temperature
and alertness data, Figures 8.2,8.3 and 8.4 (time of day by day of study interaction:
alertness: p<0.001 (n=5, D5 versus mean baseline), temperature: p<0.001 (n=6,
D5 versus mean baseline) and aMT6s: p<0.02 (n=5, D4 versus D-2). Significant
suppression of the temperature rhythm amplitude was observed over the forced phase-
shift (ANOVA, p<0.05) with clear disruption of the 24h profile on D1-3, Figure
8.2. No significant suppression of the aMT6s rhythm was evident on D4 compared
to baseline.
No significant deleterious effects on daily mean values were evident for any
mood or performance variables, Figures 8.6 and 8.7. However, the daily alertness
profiles were clearly disrupted on D0-3, Figure 8.4. More striking behavioural
disturbances were evident with subjective sleep measurements, over the initial few
days of the phase-shift, Figure 8.8. Significantly longer duration of night awakenings
and poorer sleep quality were evident (p < 0.01, nights 1-5 compared to baseline).
258
Figure 8.5 Acrophases of aMT6s and temperature rhythms during a gradual 9h advance phase-shift
a) 06: 00
04: 00-
02: 00-
24: 00-
22: 00
20: 00-
18: 00- -2
18: 00
16: 00 -I
14: 00
12: 00
10: 00 -ý
-3 -2 -1 06: 00
08: 00 -
Baseline
4
0 I
V-- 12345 Day of study
Phase-shift I
Acrophase estimates for (a) aMT6s and (b) core body temperature ('masked') before
and during a gradual 9h advance phase-shift (DO-5). A highly significant phase- advance was indicated by ANOVA (p<0.0001) for both temperature and aMT6s rhythms. Mean (+/-SEM) degree of shift: aMT6s = 9.16 +/- 1.22h (D4 from D-2),
temperature = 8.63 +/- 0.57h (D5 from mean baseline). Values represent mean +/- SEM (n=6 for temperature, n=5 for aMT6s).
259
Figure 8.6 Adaptation of mood to a gradual 9h advance phase- shift
Cl) W G)
aý 03
Aý ý . ,.., ý Q
115-
Ilp 110 0 rA, -, 105
aU rA Q ý 100- >1 Ä(Uý 95
90 85 80
aý aý
co
1 1). 0
>, r r. ed Q
-3 -2 -1 0 b)
I
+
3 2
r%
45
III-IIII
-3 -2 -1 012345
-2 -1 baseline T
34 II
phase-shift
Day of study
Daily means for (a) alertness, (b) cheerfulness and (c) calmness over a three day baseline (D-3 to DO) and during a gradual 9h advance phase-shift (D 1-5). No
significant deleterious effects due to the phase-shift or blood sampling were evident on any mood parameter (ANOVA, p>0.1). Values represent mean +/- SEM, n=5. T Blood sampling day.
115-
110-
105-
100-1
95-
90-
260
Figure 8.7 Adaptation to a gradual 9h advance phase-shift : daily mean performance
120 -ý
110-
100- Eý
90- ... A °'
80 II T I I I I I i
-3 -2 -1 012345
E
Cd A
IIIIIIIII
-3 -2 -1 012345
Slower
response times
Baseline
1
i
234
Phase-shift
Day of study
i
Daily means for performance measures of (a) efficiency, (b) accuracy and (c) response time for SAM-1 ( ) and SAM-5 (0) tasks showing baseline (day, D-3 to DO)
and adaptation to a gradual 9h advance phase-shift (D 1-5). No significamt deleterious effects of the gradual phase-shift on any performance parameter were evident (repeated measures ANOVA). Values represent mean +/- SEM (n=5).
T Blood sampling day.
ýý ý
>N
261
Figure 8.8 Adaptation of sleep to a gradual 9h advance phase-shift
40
30
20
10
0 -10
-20
a) Sleep latency
I
I
fflugý
-3 -2 -1 012345 -3 -2 -1 012345
c) Number of night awakenings 1.5
1-
0.5-
0 ýID i ý,
-0.5ý
-1 -'
d) Sleep quality
-3 -2 -1 012345
TT -3 -2 -1 012345
Night of study Night of study
Subjective measures of sleep showing baseline values (night-3 to -1) and adaptation to a gradual 9h advance phase-shift (night 0 to night 5). 4 Blood sampling day. Significantly increased duration of night awakenings and poorer quality sleep were evident during the phase-shift (p<O. 01, repeated measures ANOVA).
b) Duration of night awakenings 150 130
110 90
70 50
30 10
-10
130-
120-
u 110 .0 100 -
90
80J
70
262
8.4.3 Adaptation to a rapid 9h delay phase-shift
After the subsequent rapid 9h delay shift of external time cues on day 6 no
significant disturbances in any mood or performance parameters were apparent,
Figures 8.9-8.12. Sleep latency was significantly reduced (p <0.005, nights 6-10)
compared to baseline. The alertness rhythm apparently reentrained by D7 (ANOVA:
no significant time of day by day of study interaction, compared to baseline, p>0.1),
Figure 8.9. The temperature rhythm reestablished a normal baseline pattern on D10
(5 days after the shift), Figure 8.13. Reentrainment was confirmed by ANOVA on
cosinor-derived acrophase estimates, Figure 8.14 (p>0.1, D 10 compared to
baseline). aMT6s acrophase estimates on the fifth day (D10) were significantly
earlier by 2.28 ± 0.49h (mean ± SEM) than baseline values. However, ANOVA
performed on the raw aMT6s data could not confirm a significant phase-advance on
D 10 compared to D-2, Figure 8.15.
The direction of the subjects natural entrainment of the body temperature
rhythm was to delay in at least 4 subjects, Figure 8.16.
263
Figure 8.9 Alertness profiles during adaptation to a rapid 9h delay phase-shift
ý Cq O ýN
6I
ý vý «S UU
U Uý
ý
> cd Ü
`v cqs
ýý = ^1 rn
13 -,
12-
11-
10-
9-
8-
7-
vý 5 6-
5-
12: 00 12: 00 12: 00 12: 00 12: 00
D6 D7 D8 D9 D10
Clock time/ day of study
baseline (D-2)
Daily profiles for self-rated alertness showing adaptation to a rapid 9h delay phase- shift (on day 6) in a normal environment. ý darkness (< 1 lux), ® normal ambient light exposure. The alertness rhythm apparently readapted by D7 (ANOVA: no significant time of day by day of study interaction, p>0.1). Values
represent mean (+ SEM, n=5) alertness rated every 2h during wakefulness.
264
Figure 8.10 Adaptation of mood to a rapid 9h delay phase-shift
6
(A H aý c ý
a) U
as a) P
105-I
95-
85 0
y y ef
130
6789 10 Mean
7
)
10 Mean Baseline
I
75 -L--l
120 -
110 0
100- .ýa ýý
. -' A ,. 90- i1I11
6789 10 + Day of study
Baseline
I Mean
Baseline
Subjective measures of (a) alertness, (b) cheerfulness and (c) calmness showing adaptation to a rapid 9h delay phase-shift (on day 6) in a normal environment. No
significant deleterious effects on mood were evident (ANOVA). Values represent mean +/- SEM, n=5. T Blood sampling day.
265
Figure 8.11 Adaptation to a rapid 9h delay phase-shift : daily mean performance
ý U C G) Ü
ý
Uý
ý .c E6 ý \° A `.
U cd 4..
U U
ýU
JD . --i "ý ý
A `/
115-
110-
105-
100-
95-
90-
85--l
115-
110- 105-
100- 95-
90- 85- 80-
a)
6789 10
I I T
10
- -r-Mean
Baseline
+
- r--Mean
Baseline
4 Slower
response times
T 67
I i
89
8 T Day of Study
9 10 Mean Baseline
Daily means for performance measures of (a) efficiency, (b) accuracy and (c)
response time for SAM-1 ( ) and SAM-5 (0) tasks showing adaptation to a rapid 9h delay phase-shift (on day 6) in a normal environment. No significant deleterious effects on any performance parameter were evident (ANOVA for repeated measures, D6-10 compared to baseline). Values represent mean +/- SEM (n=5). t Blood sampling day.
266
Figure 8.12 Adaptation of sleep to a rapid 9h delay phase-shift
a) Sleep latency b) Duration of night awakenings 30
-30 J -10 6789 10
c) Number of night awakenings 3-
2.5- 2-
1.5-
0.5- T10
-0.5 -1 1
6789 10 T
Night of study
Mean baseline
a) a ..,,
H
ý O
i- Mean
baseline
6789 10
d) Sleep quality
T
789 10
Night of study
Mean baseline
Mean baseline
Subjective measures of sleep showing adaptation to a rapid 9h delay phase-shift (on day 6) in a normal environment. No significant deleterious effects on any sleep parameter were evident. Sleep latency was significantly reduced (p<0.0005, ANOVA, nights 6-11 compared to baseline). 4 Blood sampling day
267
Figure 8.13 Adaptation of the temperature rhythm to a rapid 9h delay phase-shift
37.60-
37.40-
37.20-
37.00-
36.80-
36.60-
36.40-
36.20
Y
u 36.00 ý
12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 12: 00 D6 D7 D8 D9 D 10 Baseline
(D-1) Clock time / Day of study
Adaptation of the circadian rhythm for core body temperature after a rapid 9h delay
phase shift (on day [D] 6) in a normal environment. Amplitude was significantly reduced on D6 and D7 compared to baseline (p<0.05). Bars show patterns of exposure to darkness (< 1 lux, ý ), natural ambient light (® ) and bright light (1200lux, =). Temperature was recorded continously every 6 minutes then averaged and plotted every hour as the mean of 6 subjects (+ SEM).
268
Figure 8.14 Adaptation of aMT6s and temperature rhythms to a rapid delay phase-shift
a) 06: 00-
04: 00
02: 00
24: 00
22: 00
20: 00
18: 00 456789 10
4
T
Baseline
4 N
. sý a o. -. ýö
u
4 V----T- T 56789 10 Baseline
Day of study
Acrophase estimates for (a) aMT6s and (b) core body temperature ('masked') showing adaptation to a rapid 9h delay phase-shift (on day 6) in a normal environment. Values
represent mean +/- SEM (n=6 for temperature, n=5 for aMT6s). Masked temperature rhythm apparently readapted by 10 (ANOVA, p>0.1). The aMT6s rhythm on D10 was 2.28 +/- 0.49h (mean +/- SEM) earlier compared to baseline (p<0.05, ANOVA).
269
Figure 8.15 Adaptation of the urinary aMT6s rhythm five days after a rapid 9h delay phase-shift
t C ý rA
ýo ý
ý
.ý ý
0730- 1130
1130- 1530- 1930- 1530 1930 2330
Time (hour of day)
2330- 0730
Circadian profiles for urinary aMT6s showing readaptation by the fifth day (D 10, ®)
after a rapid 9h delay phase-shift, compared to baseline (D-2, Q ). Reentrainment of the aMT6s rhythm was indicated by repeated measures ANOVA (no significant time of day by day of study interaction, p>0.1). However, acrophase estimates showed that the rhythm was still approx. 2h phase-advanced on D10 compared to baseline (p<0.02, ANOVA, see Figure 8.14). Darkness from 2330-0730h (ý ). Values represent mean + SEM, n=5.
270
Figure 5.16 Direction of reentrainment of individual masked temperature rhythms after a rapid 9h delay phase-shift
20: 001
1s: oo
16: 00
14: 00
12: 00
10: 00
08: 00
06: 00 Mean
Baseline
I I T I T i
56789 10
Day of study
Individual acrophase estimates for the masked temperature rhythm showing reentrainment by delay in at least 4 subjects after a rapid 9h delay phase-shift (on day 6) in a normal environment.
271
8.5 DISCUSSION
8.5.1 Adaptation to a gradual 9h advance phase-shift
The light/dark treatment was chosen to advance the circadian system gradually
without causing desynchrony between circadian rhythms and the external time cues.
In the present study, using a 21h sleep/wake schedule over 3 days, with bedtime and
waketime 3h earlier each day, the morning light treatment was aimed to hit the phase-
advance portion of the phase-response curve each day (Czeisler et al, 1989; Minors
et al, 1991; van Cauter et al, 1994). The following 2 days (D4 and 5) restored a
normal 24h sleep/wake schedule with further light treatment to reinforce the 9h
advance shift. Normally a 21h day is beyond the range of natural entrainment (23-
27h) (Wever, 1979; Minors and Waterhouse, 1982). However, bright light (3000
lux) has been found to increase the range of entrainment (19-31h) of circadian
rhythms (Wever et al, 1983; Wever, 1989).
In the present study, the bright light and darkness/sleep schedule was effective
at inducing and maintaining a 9h advance phase-shift in the aMT6s and masked
temperature rhythms by the end of the shift. However, significantly increased sleep
disturbances and poorer sleep quality were evident over the first few days of the
gradual phase-shift. Daily profiles of subjective alertness and body temperature were
also adversely affected initially.
Sleep loss on the night of DO may have have been an important determinant
of behavioural disturbances on D1 (Carskadon and Dement, 1979,1981), however,
subjective sleep, mood and performance measures had apparently recovered and
272
adapted by the end of the gradual advance phase-shift.
An advance phase-shift of 0.6-2.2h in the melatonin rhythm has been
demonstrated after exposure to a single pulse of bright light (3000 lux) between 0300-
0900h (Buresovä et al, 1991) and it is possible that the light treatment in this present
study was not of sufficient intensity or duration to promote the immediate adaptation
to the 21 h advancing sleep/wake schedule each day over 3 days. Gallo and Eastman
(1993) showed poor adaptation to a 22h `daylength' for 5 days during a gradual 10h
advance phase-shift, using bright light (2000-4000 lux) for 2h after waking. Indeed
much previous work underlines the difficulty in forcing the circadian system to phase-
advance. However, the present study shows that a three cycle advance of bright light
and darkness/sleep together with `reinforcement' days (D4-5) can induce large phase-
shifts in the circadian system with minimal disturbances in subjective sleep, mood and
performance by the end of the treatment. The relative effectiveness of this design
over that of Gallo and Eastman may be due to the greater duration of light exposure
and additional days to reinforce the shift. In addition, the gradual advance shift was
controlled using partial environmental isolation whereas Gallo and Eastman carried
out the study in field conditions in the presence of competing everyday 24h
zeitgebers.
8.5.2 Adaptation to a rapid 9h delay phase-shift
On day 6 the subjects returned to the normal environment and resumed their
baseline sleep/wake cycle. This rapid 9h delay phase-shift in local time cues can be
related to a westward bound transmeridian flight. Westward bound flights are usually
273
day flights. A flight at 1130h from Paris to Los Angeles after a 12h flight over 9
time zones, would arrive in Los Angeles at 1430h local time or 2330h Paris time (LA
time 9h delayed with respect to Paris time). This schedule was simulated in the
present experiment where subjects ̀boarded the flight' at 0230h clock time on day 6
(or 1130h body time). `On arrival', having adopted the new local time schedules in
their normal environment, sleep was delayed by 9h. Thus sleep eventually occurred
at an abnormal circadian phase after an unusually long time awake. Yet subjective
sleep measures were not compromised. Sleep latency was significantly reduced,
probably as a result of prior sleep deprivation.
No significant deleterious effects on mood or performance measures were
evident. This is in marked contrast to the 9h rapid advance phase shift described in
Chapter 5, where behavioural patterns took at least 5 days to establish normal
baseline patterns.
The masked body temperature rhythm completely adjusted after 5 days (on
D 10) when the aMT6s rhythm was approximately 2h earlier than baseline values.
This is in striking contrast to the rapid 9h advance phase-shift (Chapter 5) where
urinary aMT6s and core body temperature (purified) rhythms were still M 6.5h and
4h phase-advanced by the 5th day, respectively. Reentrainment of individuals
temperature rhythm followed a phase-delay path in at least 4 subjects. Masking due
to the imposed shift in the sleep/wake/activity cycle clearly affected the temperature
rhythm in the remaining 2 subjects since a phase-shift of 10h or more, in either an
advance or delay direction, after one day was unlikely.
This present study underlines the preferred direction of reentrainment of the
circadian system and confirms previous findings from field studies that greater
274
disruption of physiology and behaviour occurs after eastward (phase-advance shift)
rather than after westward flights (phase-delay shifts) (Aschoff et al, 1975; Klein and
Wegmann, 1980; Graeber et al, 1986; Gander et al, 1989). This experimental model
also provides an appropriate and inexpensive model to study adaptation mechanisms
of various biological parameters to rapid and large delay shifts in local time cues in
a normal environment.
275
CHAPTER 9 ACUTE EFFECTS OF A SINGLE MELATONIN TREATMENT
ON CIRCADIAN PHASE, BODY TEMPERATURE AND
BEHAVIOUR
Acute effects of a single melatonin treatment on circadian phase, body
temperature and behaviour
9.1 INTRODUCTION
Human circadian rhythms can become dissociated from one another and from
the environmental zeitgebers, for example after rapid time-zone travel and during
night shift work and this can lead to detrimental disturbances in human physiology
and behaviour (see Chapters 5 and 8). Alleviation of these problems is thought to
accompany faster adaptation of circadian rhythms and since several reports show that
the pineal hormone melatonin has both phase-advancing and phase-delaying properties
(Arendt et al, 1985a; Zaidan et al, 1994; Lewy et al, 1992; McArthur et al, 1991;
Sack et al, 1987), melatonin has been proposed as a treatment for circadian rhythm
disorders (Arendt et al, 1986; Lewy et al, 1992; Arendt, 1993,1994).
The mechanism through which melatonin may achieve this effect is unknown
but it may act directly on the biological clock to regulate the timing of the circadian
system. Putative melatonin receptors have been located in the human SCN (Reppert
et al, 1988), which is the major rhythm generating system of the hypothalamus
(Rusak and Zucker, 1979; Ralph et al, 1990), and in vitro melatonin, together with
its specific agonist S20098, can directly shift the rhythm of electrical activity in the
suprachiasmatic nucleus (McArthur et al, 1991, Mason et al, 1993). However, as
described in chapter 7,5mg melatonin significantly facilitated the adaptation of mood,
performance and sleep to a rapid 9h advance phase-shift within one day before
277
significant reentrainment of the biological clock occurred (using the urinary aMT6s
rhythm as a phase marker). Therefore other mechanisms of melatonin action cannot
be ruled out.
Exogenous melatonin has acute effects on body temperature (Carman et al,
1976; Cagnacci et al, 1992; Dollins et al, 1993,1994) and sleepiness (Cramer et al,
1974; Vollrath et al, 1981; Arendt et al, 1984; Tzischinsky et al, 1992a; Dollins et
al, 1993,1994), which may prove to be important in the spectrum of therapeutic
effects.
In humans, melatonin has been used in repeated daily doses to achieve a
phase-shift (Arendt et al, 1985a; Lewy et al, 1992) and to facilitate readaptation after
time-zone transitions (Chapter 7). The acute phase-shifting effects of melatonin have
received little attention until recently when intravenous administration of melatonin
at physiological concentrations was found to have time-dependent phase-shifting
effects on the endogenous melatonin rhythm (Zaidan et al, 1994).
The present study investigates the relationship between the acute phase-shifting
effects of an oral 5mg dose of melatonin -a dose commonly used in the treatment of
circadian rhythm disorders - and its effects on core body temperature and behaviour
in a normal environment.
278
9.2 SUBJECTS AND METHODS
9.2.1 Subjects
8 healthy males, aged 23-28 years (mean ± SD: 25.6 ± 1.8 years) continued
their normal daily activities during the study, except where specified, and refrained
from heavy exercise and alcohol consumption.
Using the Home and Österberg questionnaire to determine morning and
evening type individuals, 2 definite evening types (scores > 70) and 2 moderate
morning types were detected.
9.2.2 Study design
Melatonin (5mg) or placebo was administered to all subjects in a randomised,
double-blind, placebo-controlled, cross-over study. Both stages of the study began
on the same day of the week and were performed over 5 days (D1-5) with at least one
week separating the start of each stage. The volunteers maintained a regular
sleep/wake cycle, (bedtime at 2400h, awaken at 0800h, darkness between these times,
<1 lux), and natural environment (Winter, 52°N) throughout the experiment. From
dusk (approximately 1600h) until 2400h throughout the experiment, subjects remained
in dim artificial lighting (< 100 lux) and refrained from exposure to bright sunlight
at all other times, wearing sunglasses when outside (transmitting < 10 % of the light
intensity).
D1-2 served to establish a common sleep/wake cycle and accustom subjects
279
to rectal thermometers. At 1700h on D3, subjects received either melatonin (5mg in
lactose-gelatin capsules) or placebo (vehicle), administered with water.
All subjects consumed the same standard meals at set times on D3-4.
9.2.3 Sampling and measurements
Rectal temperature was recorded continuously every 6 minutes from 0800h on
D2 to 0800h on D5. Saliva samples were collected at 30 minute intervals from 1600h
until 2330h on D3 and D4, for the determination of melatonin. See section 2.2 for
details of collection and storage of samples. Subjects remained seated throughout the
saliva sampling period, except for brief periods (< 5 minutes) if urination/defecation
was necessary.
Subjective alertness was rated on a visual analogue scale (VAS: " Very drowsy
"--º Very alert") every 30 minutes from 1600h bedtime on D3 and every hour from
0800h -º bedtime on D4. Daily sleep logs were kept detailing quality (VAS),
duration, latency and number and duration of night awakenings. Section 2.7
describes these subjective assessments in detail.
9.2.4 Hormone assay
Salivary melatonin was measured by Mrs Judie English (University of Surrey)
as described in section 2.3.3. The interassay coefficients of variation were 6.7%,
8.7%, 4.9% and 3.4% at 10.4,19.6,35.5 and 56.8 pg/ml. The minimum detection
limit of the assay was 0.95 ± 0.43 pg/ml (n =10, mean ± SD).
280
9.3 STATISTICAL ANALYSIS
Differences in body temperature and alertness between melatonin and placebo
after 1700h on D3 were assessed by repeated measures ANOVA and area under the
curve (AUC) analysis followed by Student's paired t-test. The phase of the core body
temperature rhythm from 0800h on D4 to 0800h on D5 was determined by cosinor
analysis (section 2.8). The onset of melatonin secretion was determined one day after
treatment on D4 as the time after which two consecutive values exceeded 4 pg/ml.
Changes in melatonin were determined by repeated measures ANOVA. Any
relationship between the temperature suppression, exogenous salivary melatonin levels
and degree of phase-shift in melatonin onset time was assessed by Pearson's
correlation. Sleep quality was determined as a percentage of a baseline night (night
before treatment) and differences from placebo assessed by paired t-test.
9.4 RESULTS
The melatonin was unmeasurable in one subject due to food contamination.
This subject was therefore discarded from the analysis of salivary melatonin.
9.4.1 Acute effects of melatonin treatment (D3)
Salivary melatonin levels: Mean salivary melatonin concentrations are illustrated
in Figure 9.1. Salivary melatonin levels reached a maximum at 1730h in all subjects
(4291.71 ± 2823pg/ml, mean ± SD).
281
Figure 9.1 Saliva concentrations of melatonin after ingestion of 5mg of melatonin taken orally in gelatin/lactose capsules
1 RE Ir OF 1P OF 95 10
16: 00 18: 00 20: 00
Time (hours)
I11 1-7
22: 00 24: 00
Mean (+/- SEM) saliva melatonin profiles for 7 subjects after administration of 5mg melatonin (0) or placebo (0) at 1700h.
282
Figure 9.2 Acute effects of 5mg melatonin on body temperature and alertness
20 17: 00 19: 00 21: 00 23: 00 01: 00
tMeal
Time (hour)
Immediate and significant suppression of a) core body temperature (p=0.02) and b) subjective alertness (p=0.009) after administration of melatonin () at 1700h compared to placebo () (area under the curve analysis from 1700-2400h, followed by paired t-test). All values are mean +/- SEM, n=8.
283
Temperature and alertness: Melatonin clearly induced a transient but significant
suppression of body temperature between 1700-2400h (p=0.02) with a concomitant
effect on subjective alertness (p=0.009), Figure 9.2. ANOVA identified a significant
treatment effect between 1812h and 2030h (p=0.046) with a mean (± SD) fall in
temperature of 0.26 ± 0.19°C. The minimum temperature achieved following
melatonin administration was observed at 1934h ± 49 minutes (mean ± SD).
9.4.2 Delayed effects of melatonin (one day after treatment, D4)
Melatonin onset: A significant time by treatment interaction was observed
(p<0.01), whereby melatonin advanced the melatonin onset time in saliva by 1.14
± 0.19h (p < 0.001), Figure 9.3.
Temperature: The observed minimum (nadir) of the body temperature rhythm
advanced with melatonin treatment compared to placebo, Figure 9.4. This was also
indicated using cosinor-derived acrophases over the period from 0800h on D4 to
0800h on D5, where the rhythm tended to advance by 0.83 ± 0.35h (p=0.053;
acrophase after placebo: 17.01 ± 1.25h, after melatonin: 16.18 ± 1.32h, mean t
SD).
Alertness: There were no significant differences in the phase of the rhythm or
daily mean values.
284
Figure 9.3 Acute phase-shifting effects of 5mg melatonin treatment on the endogenous melatonin rhythm one day
after treatment
0-ý- 17: 00
T ---F- 19: 00
I-
Time (hour)
TMeal
Phase-advance in the salivary melatonin onset time after treatment at 1700h the previous evening with 5mg melatonin (p) or placebo ( ). A significant phase-advance was indicated by ANOVA for repeated measures (time by treatment interaction, p<0.008). Melatonin treatment advanced the onset time by 1.14 +/- 0.19h (mean +/- SEM), p<0.001, paired t-test. Values
represent mean +/- SEM, n=7.
I II --I 21: 00.23: 00
285
Figure 9.4 Influence of melatonin on the body temperature nadir during the second night after late afternoon treatment
24: 00 02: 00 04: 00 06: 00 08: 00
Time (hours)
Mean (+/- SEM) core body temperature profiles for 8 subjects during the night following D4 after treatment with either 5mg melatonin (- ) or placebo (-) at 1700h on D3. A phase-advance and elevation of the temperature nadir was observed with melatonin treatment. Temperature was recorded continuously and plotted every 6 minutes.
9
36.90 Al
36.80 -I 1 36.70=
36.60-
36.50-
36.40-
: 3'
3 T
7
ý
.,. J- . ý 3630
36.20
ý- -ýr! "ý ...
I I
E! #'
z. ý 36.10
ý
.ý
�I
36.00
; fr
1 ýWýW
286
Figure 9.5 Effect of late afternoon melatonin treatment on sleep the following nights
180
160 ý-.
140
120 ý 100 z .ý 80
60
cz 40
20
0 First Second
Night following treatment
Subjective sleep quality on the first (N3) and second (N4) nights following 5mg melatonin ( ) or placebo treatment ( 0) treatment on D3 at 1700h. Melatonin
significantly improves subjective sleep quality over placebo on the first night after treatment (p=0.006) with a similar tendency on the second night (p=0.053) - paired t-test. Values are mean +/- SEM, n=8.
287
Sleep: Sleep quality on the first night after melatonin treatment (N3) significantly
improved by 70 ± 18.4% (p<0.05) and also showed a tendency to improve (by 50
± 21.6%, p =0.053) on the night following D4, Figure 9.5. No significant effects
on other sleep parameters were evident.
9.4.3 Relationship between temperature suppression, exogenous melatonin
levels and degree of shift.
The degree of individual melatonin-induced temperature suppression (using
AUC analysis from 1700-2400h) was positively correlated to the degree of shift in
melatonin onset time in all but one subject (r=0.853, p=0.031), Figure 9.6. This
subject slept, periodically, for 3 hours after receiving melatonin treatment and showed
an exceptional fall in temperature (0.65°C) which was more than 2SD from the mean.
This was presumably due to masking by sleep. There was no significant correlation
between the individual maximum of exogenous melatonin concentrations and the
degree of temperature suppression.
288
Figure 9.6 Relationship between individual degree of phase- shift and temperature suppression
2-
1.8-
1.6-
1.4-
1.2-
1
0.8-
0.6-
0.4-
0.2-
0 -0.5
r-0.853 p=0.031
0 0.5 1 1.5 2 2.5 3 3.5 4
Temperature suppression (° Cx minutes, AUC from 1700-2400h)
Relationship between the individual degree of phase shift in the salivary melatonin (MT) onset time (on D4) and the magnitude of temperature suppression (from AUC analysis from 1700-2400h on D3)
289
9.5 DISCUSSION
In a normal environment a single 5mg melatonin treatment phase-advanced its
own endogenous rhythm 24h after treatment, with a similar effect in core body
temperature. Repeated doses of oral melatonin are therefore not necessary to see an
effect on the human biological clock. Recently, the acute phase-shifting effects of
intravenous administration of physiological levels of melatonin have been described
(Zaidan et al, 1994). The observation that a phase-shift in the endogenous melatonin
rhythm was still apparent 24h after treatment and after exogenous melatonin had
cleared, indicates that a single pulse of melatonin has a prolonged effect on the phase-
adjustment of the clock. This resetting of the biological clock is comparable to the
work of McArthur et al (1991) who demonstrated, in vitro, that a shift in the rhythm
of electrical activity of the SCN could still be seen for at least 24h after melatonin
treatment.
In the present study, the induced phase-shift in the salivary melatonin onset
was preceded by the immediate suppression of core-body temperature and subjective
alertness. The timing of the fall in temperature and alertness correspond closely to
the reported increae in sleep propensity after 5mg melatonin given in the late
afternoon (Tzischinsksy et al, 1992a).
Immediate increases in drowsiness and deleterious effects on performance have
recently been observed after oral ingestion of low doses of melatonin in the morning
(0.1-10mg at 1145h, Dollins et al, 1994). In the present study, transient sleepiness,
and in one case periodic sleep, during the late afternoon and early evening did not
compromise sleep during the subsequent night. In addition, no deleterious effects on
290
alertness were observed the following day. Considering the short half-life of
melatonin (Aldhous et al, 1985) and the transient nature of its acute drowsiness and
temperature-lowering effects, these sedative-like qualities could not account for the
delayed effects on sleep quality during the subsequent nights. An effect on the timing
of sleep onset/offset was unlikely to be seen since these times were fixed. However,
an effect of melatonin on the timing mechanisms of sleep cannot be ruled out, in
addition to any transient sedative ability. Administration of 5mg melatonin at midday
can delay the so-called nocturnal "sleep gate" where the sleep propensity normally
rises steeply around 2300h (Tzischinsky et al, 1992a). This delayed effect is in
addition to an increase in sleep propensity shortly after administration. In this present
study, melatonin may have indirectly affected sleep quality by synchronising the
endogenous sleep/wake rhythm to the imposed sleep/wake schedule and environmental
zeitgebers, i. e. strengthening the phase relationship between them.
The individual magnitude of melatonin-induced temperature suppression
correlated highly with the degree of phase-shift in the melatonin onset time in saliva,
such that the greater the suppression the larger the induced phase-shift. Cagnacci et
al (1993) and Strassman et al (1991) have provided evidence that melatonin
counteracts the stimulatory effect of bright light on body temperature at night. In
view of the present results, it is possible that the phase-shifting effects of melatonin
and bright light may be related to their ability to acutely change core body
temperature in humans. The next chapter follows on from this preliminary study to
investigate in detail the dose-relationship of melatonin between the acute phase-
shifting properties and its immediate effect on temperature and behaviour.
291
CHAPTER 10
DOSE-RESPONSE EFFECTS OF MELATONIN ON
CIRCADIAN PHASE, BEHAVIOUR AND TEMPERATURE
SUPPRESSION
Dose-response effects of melatonin on circadian phase, behaviour and temperature suppression
10.1 INTRODUCTION
In humans, melatonin has phase-shifting effects on the internal biological clock
(Arendt et at, 1985a; Lewy et at, 1992; Zaidan et at, 1994) which are preceded by
acute effects on temperature and behaviour (see Chapter 9; Deacon et at, 1994).
Melatonin has acute temperature lowering properties (Carman et at, 1976; Cagnacci
et at, 1992; Dollins et at, 1993,1994, Deacon et at, 1994; Chapter 9) and the
stimulatory effects of bright light on temperature are mediated through the
suppression of melatonin (Cagnacci et at, 1993; Strassman et at, 1991). Melatonin
also has acute transient sleep-inducing effects (Vollrath et at, 1981; Arendt et at,
1984; Dollins et at, 1993,1994) and time-dependent effects on the sleep propensity
function (Tzischinsky et at, 1992a).
It is possible that there is a direct relationship between the acute effects of
melatonin and its ability to induce phase-shifts in the biological clock.
In the previous chapter, the degree of individual phase-shift in the endogenous
melatonin rhythm after 5mg of oral melatonin at 1700h was shown to relate to the
magnitude of temperature suppression. The present study was designed to investigate
the dose-response relationship between acute effects of oral melatonin on temperature,
sleep, mood and performance and the subsequent phase-shift in core body temperature
and melatonin rhythms. It will also help determine the optimal doses of melatonin
required to treat circadian rhythm disorders.
293
10.2 SUBJECTS AND METHODS
10.2.1 Subjects
6 healthy volunteers (3M: 3F), aged 23-34 years (27.2 ± 3.7 years, mean ±
SD), under no medication and all non-smokers were recruited from research students
and staff of the University of Surrey. They followed similar sleep/wake schedules
and lifestyles. No definite morning or evening types were detected (scores > 70 or
< 31) using the Home and Östberg questionnaire (1976). However, one subject
tended to be a morning type (score of 63) and two were rather of the evening type
(scores of 41 and 42).
Except where specified, subjects continued their normal daily activities during
the experimental period but refrained from heavy exercise and alcohol consumption.
Female subjects were screened to check the stage and regularity of their menstrual
cycles.
10.2.2 Preparation of melatonin/placebo
300mg melatonin was dissolved in 30 ml absolute ethanol to give 10mg/ml.
This solution was then diluted as follows:
1) 5mls diluted to 50m1 with corn oil to give lmg/ml
2) 5mls of 0. ing/ml diluted to 50m1 with corn oil plus 10% absolute ethanol to give
0.1mg/ml
294
3) 5mls of 0.1 mg/ml diluted to 50m1 with corn oil plus 10 % absolute ethanol to give
0.01mg/ml.
Each dose was 5mls of the above to give 5mg, 0.5mg and 0.05mg melatonin.
Placebo consisted of the equivalent amount of corn oil with 10% absolute ethanol.
The melatonin or placebo dose in corn oil was suspended by thorough mixing in 50m1
semi-skimmed milk before administration.
10.2.3 Study design
Melatonin (0.05mg, 0.5mg, 5mg) or placebo was administered to all subjects
in a randomised, double-blind cross-over study. The 4 stage study was divided into
two periods and took place during the first 11 days of February and March 1994 at
52° N. The study was designed such that the female subjects began each period of
the study 1 to 2 days into the follicular phase of their menstrual cycle, to minimise
any possible influence of the menstrual cycle on circadian temperature and melatonin
rhythm parameters (Nakayama et al, 1992; Wetterberg et al, 1976). Each stage
began on the same day of the week and was performed over 5 days (D1-5) with 1
week separating the start of each stage in each period.
Subjects were required to maintain a regular sleep/wake cycle, retiring to bed
between 2330-2400h and rising between 0730-0800h. They remained in darkness
(< 1 lux) between 2400-0800h. From dusk (approximately 1800h) until 2400h they
remained in dim artificial lighting (<50 lux) throughout the experiment. Subjects
were asked to avoid exposure to bright sunlight throughout the study and wore
sunglasses when outside.
295
Days (D)1-2 served to establish a common sleep/wake (and light/dark) cycle,
accustom the subjects to rectal thermometers and practice performance tasks every 2h.
On D3, after fasting for 4h, subjects received either a single dose of melatonin
(0.05mg, 0.5mg or 5mg) or placebo at 1700h and remained seated throughout the
evening.
On D4 subjects remained sedentary, by watching TV or reading/writing to
minimise the influence of activity on temperature. The volunteers received the same
standard breakfast, lunch and evening meals at set times on D3 (0900h, 1300h and
2130h) and D4 (0900h, 1300h and 1900h). The meal was later on D3 to allow 6.5h
for the acute effects of melatonin on body physiology and behaviour to be measured
"unmasked" in the absence of meals.
10.2.4 Sampling
Rectal temperature was recorded every 6 minutes from 0800h on D2 until
0800h on D5.
An indwelling cannula was inserted into the forearm of each subject, under
local anaesthetic, between 1600-1645h on D3 and 1700-1745h on D4. Blood samples
(5m1) were obtained every 30 minutes from 1700-2300h on D3 and from 1800-2330h
on D4. A total of 510 ml of blood was removed from each person over the whole
experiment.
Subjects remained seated for at least 30 minutes before blood sampling
commenced on D3 and D4 and remained seated throughout the entire sampling
period, except for brief periods (<5 minutes) for urination/defecation if necessary,
296
immediately after blood sampling. This was necessary to control for postural effects
on melatonin concentrations (see Chapter 3). Sampling times differed on D3 and D4
to minimise blood sampling but to provide a sufficient number of sample points to 1)
obtain an adequate profile of exogenous melatonin in plasma on D3 and 2) determine
the melatonin onset time on D4.
10.2.5 Behavioural measurements
Daily sleep logs were recorded on waking in order to estimate sleep onset,
offset, latency and the number and duration of night awakenings. Sleep quality was
rated on a visual analogue scale (VAS; Best ever -a--º Worst ever). Subjective
alertness was also rated on a VAS (Drowsy *--> Alert) every 30 minutes from 1700-
2400h on D3, every hour from 0800-2400h on D4 and at 0800h on D5.
Performance tests in the form of 1- and 5-target search and memory tasks
(SAM-1 and SAM-5, respectively) were carried out every hour from 1700-2400h on
D3, every 2 hours from 0800-2400h on D4 and on waking at 0800h on D5. On D1-
2, subjects practised the tests regularly, approximately every 2/3h.
It was necessary to measure mood and performance as frequently as possible
after melatonin/placebo administration on D3 to assess acute transient effects of
treatment, but without overwhelming the subjects and compromising subsequent
behavioural measurements. To enable subjects to continue as normal a routine as
possible on D4 and to prevent performance `boredom', the interval between
performance tests and alertness ratings was doubled.
Section 2.7 describes these behavioural measurements in detail.
297
10.2.6 Melatonin assay
Plasma melatonin was measured as described in section 2.3.2. The interassay
coefficients of variation were 8.5,7 and 6.2% at 33.3,64 and 228 pg/ml (n=10).
The minimum level of detection (defined as the concentration at 2SD from the mean
at maximum binding) was 4.8 ± 0.43 pg/ml (mean ± SEM).
10.3 STATISTICAL ANALYSIS
The performance parameters accuracy, response time and efficiency were
defined as the percentage of correct responses, time taken to complete the task and
accuracy/response time, respectively. Response time data were normalised by square
root determination.
Comparisons were made between treatments for acute effects on D3 and for
delayed effects on D4.
Acute effects of melatonin on D3
Core body temperature, plasma melatonin, alertness and performance were
expressed as area under the curve (AUC) from 1700-2400h (timed to encompass the
whole acute effect of treatment). Alertness and performance data were calculated as
a percentage difference from the value at 1700h for each individual. Grouped
differences were then assessed by one-way ANOVA and Duncan's multiple range
298
post-hoc test.
Delayed effects of melatonin (one day after treatment, D4-5)
Plasma melatonin was analysed by two-factor ANOVA for repeated measures
(factors: time and treatment) followed by Duncan's multiple range test to identify
specific differences between treatments. The onset of plasma melatonin secretion in
plasma was defined as the time when plasma concentrations exceeded twice the
minimum detection limit of the assay, when two consecutive values exceeded this
amount.
Alertness and performance data were calculated for each individual at each
time point as a percentage difference from the daily mean (D4). Daily means of
percentage differences for each individual were then determined. All sleep data were
calculated as a difference from the baseline night (N2, i. e. the night of D2). Group
differences for all sleep parameters and performance and alertness daily means were
analysed by one-way ANOVA followed by Duncan's multiple post-hoc test.
The phases of the core body temperature, alertness and performance rhythms
were determined by cosinor analysis (0800h on D4 to 0800h on D5, section 2.8) and
assessed for treatment differences by one-way ANOVA and Duncan's multiple range
test.
The shift in the temperature nadir during the night following D4 (0200-0800h)
was assessed by repeated measures ANOVA (factors: time and treatment).
Pearson's correlation was used to determine any relationship between
temperature suppression, exogenous plasma melatonin levels and the degree of phase-
299
shift in the temperature and melatonin rhythms.
10.4 RESULTS
10.4.1 Acute effects of melatonin treatment (D3)
Plasma melatonin levels: Mean plasma melatonin concentrations are illustrated in
Figure 10.1. Plasma melatonin levels significantly increased with increasing doses
of melatonin treatment (p<0.001) and in a linear fashion (r=0.72, p<0.0001). A
maximum (mean ± SEM), which was significantly different from 1700h for all
doses, (p<0.02) was reached at 1730h for 0.05 mg (118 ± 37 pg/ml) and 5 mg
melatonin (18495 ± 3326 pg/ml) and at 1800h for 0.5mg melatonin (1327 ± 491
pg/ml). Normal nocturnal levels of melatonin (< 200pg/ml) were achieved 3.16 ±
0.30h (mean ± SEM) after 0.5 mg melatonin. Only two subjects showed
physiological concentrations before the last blood sample at 2300h after 5mg
melatonin.
Temperature: Melatonin clearly induced an acute, transient, but significant
suppression of core body temperature which was dependent upon the dose given
(p <0.001), Figure 10.2.0.5mg and 5mg melatonin were significantly more effective
than 0.05mg melatonin and placebo (p<0.01). Mean ± SEM fall in temperature
(from 1700-2400h) was: 0.02 ± 0.02°C, 0.1 ± 0.01°C and 0.14 ± 0.02°C for
0.05mg, 0.5mg and 5mg melatonin, respectively, compared to placebo. The times
of the temperature minima were not significantly different from each other (p > 0.1):
300
0.05mg = 19: 16 ± 12 minutes, 0.5mg = 19: 14 ±4 minutes and 5mg = 19: 30 ±
10 minutes. The maximum suppression compared to placebo was: 0.05mg = 0.18
± 0.03°C, 0.5mg = 0.3 ± 0.03°C and 5mg = 0.35 ± 0.02°C, (mean ± SEM).
Alertness: Subjective alertness decreased with increasing doses of melatonin
(p<0.001), Figure 10.3.0.5mg and 5mg melatonin treatments were significantly
more effective at reducing alertness than 0.05mg melatonin and placebo treatments
(p < 0.05), Table 10.1.
Performance: SAM-5 accuracy and efficiency (Figures 10.4 and 10.5) decreased with
increasing doses of melatonin (p<0.01 and p<0.001, respectively), with a trend
towards slower response times (Table 10.1). 5mg melatonin treatment was
significantly more deleterious than all other treatments (p<0.05), Table 10.1. No
significant treatment effects were observed on SAM-1 performance parameters
(Figures 10.4 and 10.5 and Table 10.1).
301
Figure 10.1 Plasma concentrations of melatonin after ingestion of low doses of melatonin taken orally in corn oil.
100000
10000 1
1000 1
100,
101
1 I I I IIII 17100 18 1990 20 2200 22 223D0
Time (hours)
Mean (+/- SEM) plasma melatonin profiles for 6 subjects after administration of 5mg (0), 0.5mg (f), 0.05mg (p) of melatonin or placebo ( ) at 1700h on D3.
302
Figure 10.2 Acute dose-dependent suppression of body temperature by melatonin
37.4
37.3
37.2
37.1
37
36.9
36.8
16: 00 t 18: 00 36.7 -1--T---1
20: 00 2: 00 24: 00
Meal Melatonin administration
Time (hours)
Mean core body temperature for 6 subjects after an oral dose of placebo ( ), 0.05mg (0), 0.5mg (f), 5mg (Q) melatonin at 1700h (recorded every 6 minutes, plotted every 12 minutes for clarity). The higher the dose the greater the suppression (ANOVA: p<0.001, using the difference from placebo after AUC analysis from 1700-2400h).
Average SEM = +/- 0.090C.
303
Figure 10.3 Dose-dependent decrease in subjective alertness immediately after melatonin treatment
105-
ý .c O ° 95 . -ý
85-
7 75-
65 ý
55 IIIIIII
16: 00 * 18: 00 20: 00 22: 00 24: 00 1ºdel.,
Ilý.. w...
ý
"ri,,, e ýfn,
»rný Melatonin Meal I uiiv, `11vU1 CO)
administration
Subjective alertness ratings after an oral dose of placebo ( ), 0.05mg (0 ), 0.5mg (A ), 5mg (0) melatonin at 1700h. Alertness decreased with increasing doses of melatonin (ANOVA: p<0.001, after AUC analysis from 1700-2400h). 0.5mg and 5mg were significantly more deleterious than 0.05mg and placebo (p<0.05, Duncan's multiple range test). Values represent mean +/- SEM (n=6).
304
Table 10.1 Acute effects of melatonin treatment (D3)
Effects of low oral doses on alertness and performance, mean (SEM), % difference from
placebo with AUC analysis (1700-2400h).
Performance
Alertness
0.05mg
-7.0 (4.0)a, b
SAM-5 Accuracy -4.7 (9.6)al
SAM-5 Response Time* +0.9 (4.0)al
SAM-5 Efficiency -3.1 (8.6)
SAM-1 Accuracy -1.5 (8.0)
SAM-1 Response Time* -2.2 (2.9)
SAM-1 Efficiency +0.5 (7.4)
0.5mg
-18.0 (. 03)a
-13.4(6.3)b
+ 2.7 (2.8)bl
-15.9 (6.1)
-5.7 (8.5)
+0.7(2.3)
-6.1 (9)
5mg
-21.0 (3.0)b
-31.6 (6.7)ß1, b
+ 2.1(301, b1
-34.2 (6.9)
-9.5 (3.3)
+3.4(2.4)
-13.5 (2.1)
* "+" indicates slower response time, "-" faster response time
Melatonin doses with the same letters represent significant differences between
treatments (Duncan's multiple range test): a and b: p<0.05, al and bl: p<0.01,
Melatonin
305
Figure 10.4 Acute dose-dependent effects of melatonin on performance accuracy
130-1
120 -a
110-
100ý
90 -a
80-I
70
160-1
140 -
120-
100-I
80 -
60-1
40
17: 00 19: 00 21: 00 23: 00
b)
17: 00 19: 00 21: 00 23: 00
Time (hour) tMeal
Accuracy scores for a) low- (SAM-1) and b) high- (SAM-5) memory capacity tests, after oral administration of either placebo ( ), 0.05mg melatonin (MT) (0), 0.5mg MT(A) or 5mg MT (O)at 1700h. Significant treatment effect for SAM-5 only (ANOVA: p<0.01 after AUC analysis 1700-2400h): accuracy scores after 5mg melatonin were significantly worse than all other treatments. Values
represent mean +/- SEM, n=6.
306
Figure 10.5 Acute dose-dependent effects of melatonin on performance efficiency
120
110-
100-
90-
80 -ý
70 f- 17: 00 19: 00
17: 00 19: 00 21: 00 23: 00
Time (hour) tMeal
I 21: 00
I -Tý 23: 00
I
Efficiency scores for a) low- (SAM-1) and b) high- (SAM-5) memory capacity tests, after oral administration of either placebo ( ), 0.05mg melatonin (MT) (0 ), 0.5mg MT (f) or 5mg MT (Q) at 1700h. Significant treatment effect for SAM-5 only (ANOVA: p<0.001 after AUC analysis 1700-2400h): efficiency scores after 5mg melatonin were significantly worse than all other treatments. Values represent mean +/- SEM, n=6.
307
10.4.2 Delayed effects of melatonin treatment (one day after
treatment, D4)
Melatonin onset: A significant time by treatment effect (p < 0.0001) was seen, where
increasing doses of melatonin advanced the plasma melatonin onset time relative to
placebo by 0.36 ± 0.13h, 0.69 ± 0.15h and 1.43 ± 0.16h (mean ± SEM) for
0.05mg, 0.5mg and 5mg melatonin, respectively, Figure 10.6. The 5mg melatonin
treatment significantly (p<0.05) advanced the onset time more than all other
treatments.
Temperature: There was a clear and dose-dependent phase-advance in the nadir of the
temperature rhythm during the night following D4 (ANOVA for repeated measures,
p<0.02, Figure 10.7). This was also evident by cosinor analysis over the period on
D4 to D5 (0800h on D4 to 0800h on D5). A significantly greater advance phase-shift
was seen with increasing doses of melatonin using differences in acrophase from
placebo (p<0.0001). The rhythm advanced further (p<0.01) with 0.5mg and 5mg
melatonin treatments compared to 0.05mg, and 0.5mg treatment compared to 5mg
melatonin (p<0.05), Table 10.2.
Alertness: There was no significant treatment effect on the phase of the alertness
rhythm or on daily mean values one day after treatment. However, there was a trend
for melatonin to advance the rhythm with increasing doses compared to placebo
(p=0.067), Figure 10.8, Table 10.2.
308
Performance: No significant performance rhythm could be established (cosinor
analysis) and no treatment effect on daily mean values was seen.
Sleep: Figure 10.9 shows the effect of low doses of oral melatonin on various sleep
parameters on the night following treatment (D3). On the first night only, a
significant treatment effect was seen with an increase in sleep quality
(p < 0.05), a decrease in the duration of night awakenings (p < 0.05) and an earlier
sleep onset time (p <0.01), Table 10.2. The 5mg treatment proved more effective
than all other treatments (p<0.05). The quality of sleep improved by = 30%, 43%
and 61 % for 0.05mg, 0.5mg and 5mg melatonin treatment respectively, compared to
placebo. On the second night (night following D4), there were no significant
differences between groups in any sleep parameter.
309
Figure 10.6 Dose-response relationship between oral melatonin and degree of induced phase-shift in melatonin rhythm one day after
treatment
01- 1800
IIII 1 ý-1 1900 2000 2100 2200 2300 2400
Time (hour)
Phase-advance in plasma melatonin onset time after treatment at 1700h the previous evening with 5mg (0 ), 0.5mg (f), 0.05mg (0) melatonin or placebo ( ). A significant dose-response phase-shifting effect identified by ANOVA (for repeated measures, time by treatment interaction: p<0.0001). Melatonin treatment advanced the onset time by 0.36 +/- 0.13h, 0.69 +/- 0.15h and 1.43 +/- 0.16h (mean +/- SEM) for 0.05mg, 0.5mg and 5mg melatonin, respectively. Values represent mean +/- SEM, n=6.
310
Figure 10.7 Dose-dependent phase-advance in the temperature nadir during the second night after melatonin treatment
36.37 02: 00 03: 00 04: 00 05: 00 06: 00 07: 00 08: 00
Time (hours)
Mean core body temperature profiles for 6 subjects during the night following D4 (from 0200h to 0800h) - one day after treatment with either placebo ( ), 0.05mg
melatonin (MT) (0), 0.5mg MT (f) or 5mg MT (Q) treatment. Rectal temperature was recorded every 6 minutes, averaged and plotted every 30 minutes for clarity. There was a dose-dependent phase-advance in the nadir of the rhythm (ANOVA for repeated measures: time by treatment interaction, p<0.02). This was confirmed by cosinor analysis (from 0800h on D4 to 0800h on D5, ANOVA: p<0.0001). Mean (+/- SEM) difference in phase-shift from placebo: 0.05mg melatonin (MT) _ -0.27 +/- 0.16h, 0.5mg MT = 0.48 +/- 0.12h and 5mg MT = 1.0 +/- 0.09h.
311
Figure 10.8 Delayed effects of low oral doses of melatonin on the alertness rhythm one day after treatment
130 -1
120-
ý-.
ý V E ý
ý . r, ý 'd
110
100 1 '011 ý
N N
aý
90 -
80
704
60 08: 00 12: 00 16: 00 20: 00 24: 00 Time (hours)
Mean subjective alertness ratings (calculated as a% of individuals daily mean) for 6 subjects on D4, one day after treatment (at 1700h, on D3) with either placebo (-- ), 0.05mg melatonin (MT) (.... ), 0.5mg MT () or 5mg MT (_ ). * p<0.05: significant treatment effect seen at 1000h (ANOVA),
where placebo had significantly (p<0.05) lower ratings than 0.5mg and 5mg melatonin treatment. Using cosinor-derived acrophases, there was a trend for the rhythm to phase-advance (ANOVA: p=0.07). Mean (+/- SEM) difference in phase from placebo was an advance shift of 0.37 +/- 0.31h with 0.05mg MT, 0.97 +/- 0.35h with 0.5mg MT and 1.6 +/- 0.3h with 5mg MT.
312
Table 10.2 Delayed effects of melatonin treatment (D4, one day after treatment)
Effect of low oral doses of melatonin on the magnitude of temperature suppression
(%) and degree of advance phase-shift (hours) in the rhythms of body temperature,
melatonin, alertness and sleep. Mean (SEM), difference from placebo.
Parameter Melatonin
Temperature
suppression (%)
0.05mg 0.5mg 5mg
1.64 (1.23)a, b 7.32 (0.96)a 10.2 (1.59)b
Shift (h) in:
Body Temperature -0.27(0.16)a, bl 0.48 (0.12)a, c 1.0 (0.09)b1, c
Plasma melatonin
onset time 0.36(0.13)a1 0.69 (0.15)b1 1.43(0.16)11, b1
Alertness 0.37 (0.31) 1.07 (0.43) 1.6 (0.3)
Sleep onset 0.28 (0.08)a. b 0.35 (0.1)a 0.59 (0.14)b
Sleep offset 0.32 (0.2) 0.39 (0.16) 0.49 (0.18)
Melatonin doses with the same letters represent significant differences between treatments
(Duncan's multiple range test): a, b and c: p<0.05, al and bl: p<0.0 1,
313
Figure 10.9 Dose-dependent effects on sleep on the first night following melatonin treatment
24.8
24.6
24.4
24.2
24
23.8
23.6
a) Sleep Onset
c) Duration of Night Awakenings
25 VJ
.ý
ý u
. ý--ý
cd ýc ý O
15
-15
t) C
.ý ý aý ý cd
-0 0
b) Sleep Offset
d) Sleep Quality
150 r 130
110
90
70
50
Subjective sleep measurements for the first night following late afternoon treatment with placebo (0), 0.05mg ( ), 0.5mg (Q), 5mg (M ) melatonin. There were significant treatment effects for advancing sleep onset (p<0.01), improving sleep quality (p<0.05) and reducing the duration of night awakenings (p<0.05), where the 5mg treatment was significantly more effective than all other treatments, ANOVA. Values represent mean +/- SEM, n=6.
314
10.4.3 Relationship between individual degree of phase-shift and
magnitude of temperature suppression
Table 10.2 shows the mean magnitude of temperature suppression and the degree
of shift in the temperature, melatonin, alertness and sleep rhythms.
A highly significant correlation existed between the individual magnitude of
temperature suppression and both the degree of phase-shift in the plasma melatonin onset
time (r=0.77, p<0.0005) and the temperature rhythm (r=0.549, p<0.05), Figure
10.10, but not the alertness rhythm. The magnitude of the phase-shift of the temperature
and melatonin rhythms also correlated closely (r=0.699, p<0.001).
10.4.4 Relationship between individual exogenous plasma melatonin
levels, temperature suppression and degree of phase-shift
Individual exogenous melatonin concentrations in plasma on D3, as assessed by
AUC, correlated with the magnitude of temperature suppression (r=0.468, p <0.05) and
the degree of phase-shift in the melatonin rhythm (r=0.782, p<0.0002), but not the
degree of phase-shift in the body temperature rhythm (r=0.462, p=0.053).
315
Figure 10.10 Relationship between individual degree of phase- shift and temperature suppression
a) i.
r-0.77 p<0.0002
-1-I iIIii
-4 048 12 16
a) ii.
r-0.55 p<0.02
-0.5 1 is
-4 04 8 12 16
Temperature suppression ( 0C
x minutes, AUC from 1700-2400h, difference from placebo)
-0.5
I i ti I
0 0.5 1 1.5 Shift in temperature rhythm
(hours, difference from placebo)
a) Relationship between the individual magnitude of melatonin-induced temperature suppression and the shift in i) plasma melatonin onset time, ii) body temperature rhythm. b) Relationship between the individual degree of melatonin-induced phase- shift in the temperature rhythm and plasma melatonin onset time.
316
10.5 DISCUSSION
The previous chapter describes how the individual degree of phase-shift in the
endogenous melatonin rhythm relates to the magnitude of temperature suppression. The
present study confirms this finding, where greater temperature suppression was associated
with a larger phase-shift. In addition, a dose-response relationship between the dose of
oral melatonin, temperature suppression and phase-shift in the endogenous melatonin
rhythm has been established. With increasing doses of melatonin a greater magnitude of
temperature suppression and degree of phase-shift in the endogenous melatonin rhythm
was achieved. The rhythm of core body temperature shifted in a similar fashion and
correlated with the shift in the endogenous melatonin rhythm, but the relationship was
less marked. Body temperature is more susceptible to masking by exogenous influences
such as meals, activity and sleep (Wever, 1985b; Minors and Waterhouse, 1989) than the
melatonin rhythm (Lewy, 1983). The design of the study controlled for the two major
masking influences on melatonin, light and posture (Lewy et at, 1980; Deacon et at,
1994). Although the subjects ambient lighting and activity were strictly controlled during
the evening of D3 and throughout D4, some masking of the temperature rhythm was still
possible.
There is considerable support for a thermoregulatory role for melatonin
(Strassman et at, 1991; Cagnacci et at, 1992,1993; Badia et at, 1993). The peak of
melatonin secretion at night corresponds closely to the nadir in body temperature and
Cagnacci et at (1992) have suggested that about 40% of the diurnal variation in core body
temperature may be the direct consequence of the acute effects of endogenous melatonin.
Whether or not these phenomena are representative of physiological mechanisms is a
317
question of interest. In the present study and in previous work (Chapter 9, Deacon et al,
1994), the time of maximum suppression of body temperature occurred approximately
2h after the observed peak in (physiological and pharmacological) levels of exogenous
melatonin. This 2h lag is consistent with the relationship between the endogenous
melatonin and body temperature rhythms before and after a forced phase-shift with bright
light (Shanahan and Czeisler, 1991). Both melatonin and body temperature shifted in a
similar direction and magnitude with the melatonin fitted maximum preceding the
temperature minimum by nearly 2h. These changes were observed in constant routine
conditions, i. e. in the absence of external time cues with either constant, or no disturbing
(masking) influences. Furthermore, a recent study has demonstrated that the core body
temperature rhythm is exactly inverse to the urinary aMT6s rhythm under similar
constant routine conditions (Kräuchi and Wirz-Justice, 1994). Considering the time taken
for hepatic metabolism and renal excretion, the urinary aMT6s rhythm will be delayed
by m 1-1.5h compared to the melatonin rhythm in plasma (Matthews et al, 1991; Deacon
and Arendt, 1994). However, this time lag between the melatonin rise and temperature
fall is not apparent in some situations. Wehr et al (1993) demonstrated, under constant
routine conditions, a phase dissociation between the melatonin and temperature rhythms
with the body temperature rhythm declining before the rise of the melatonin rhythm when
subjects were transferred to a winter from a summer photoperiod. Taken together, these
results suggest that the primary effect of endogenous melatonin may not be direct
suppression of the nighttime fall in body temperature but is more probably an indirect
role in regulating the timing and amplitude of body temperature.
The exact mechanism and site of action by which melatonin influences body
temperature is unknown. It is possible that melatonin acts centrally in the preoptic area
318
and anterior hypothalamus - the main thermoregulatory centres (Hellon, 1972; Hensel,
1973; Satinoff, 1978) - and/or peripherally to adjust heat loss. In rats, a major site of
heat loss occurs in the brain arteries of the circle of Willis and tail arteries where high
affinity melatonin receptors have been located (Viswanathan et al, 1990; Seltzer et al,
1992). A low level of specific melatonin binding has been reported in the preoptic area
of the hypothalamus in rodents (Williams et al, 1989). In addition, high affinity
melatonin binding sites have been found in other areas of the mammalian hypothalamus
including the SCN and paraventricular nucleus (PVN) (review: Stankov et al, 1993; see
section 1.2.11). Direct and indirect neural and anatomical associations between these
sites and the preoptic area have been described (review: Watts et al, 1991; see section
1.2.11), and suggest possible pathways through which melatonin (and the SCN) may
modulate the rhythmicity of body temperature.
The plasma melatonin concentrations achieved with different doses varied from
normal nocturnal night time levels with 0.05mg to 2 orders of magnitude higher than
physiological concentrations with 5mg. There was a log-linear relationship between
plasma melatonin levels and the degree of phase-shift in the endogenous melatonin
rhythm. In addition, the magnitude of temperature suppression related to plasma
melatonin levels, although the results were less striking. There was also a trend towards
a relationship between the degree of phase-shift in the core body temperature rhythm and
plasma melatonin levels. The range and direction of shift observed in plasma melatonin
onset was comparable to that obtained by Zaidan et al (1994) with a single physiological
melatonin infusion from 16: 00-19: 00h (shift in onset time = 21 ± 50 minutes, mean f
SD) and Lewy et al (1992) after 4 days with 0.5mg/day (shift in onset time " 1h).
The acute sedative-like effects of melatonin have been known for many years
319
(Cramer et al, 1974; Vollrath et al, 1981). Recently, Dollins et at (1994) have reported
an increase in sleepiness and deleterious effects on performance after administration of
0. lmg and 0.3mg melatonin at 1145h when volunteers were tested in a controlled
environment from 0930-1730h and suggested that endogenous melatonin may be a direct
factor in sleep-inducing mechanisms. The present study, also in a controlled
environment, showed a dose-dependent significant decrease in alertness and SAM-5
efficiency only with 0.5mg and 5mg melatonin but not with 0.05mg. The latter was the
only dose to achieve normal physiological concentrations and argues against a
physiological sleep-inducing role for melatonin. Rather, as discussed previously (Arendt
et al, 1984; Tzischinsky et al, 1992a), the results could be interpreted on the basis that
melatonin influences the timing mechanisms of sleep and wakefulness. The fact that the
night's sleep immediately following melatonin treatment was markedly enhanced, in a
dose-dependent manner compared to placebo, with an earlier sleep onset and a trend
towards earlier sleep offset also supports this contention. There was also a trend for the
self-rated alertness rhythm to shift to an earlier time the following day. Even without
taking into account the sleep ̀inertia' which influences alertness during the first hour or
two after waking (Folkard and Akerstedt, 1989; Akerstedt and Folkard, 1990; Dijk et al,
1994), there was a clear elevation in alertness in the morning which was dose-related.
Arendt et al (1984) have reported on the delayed effects of 2mg melatonin given
at 1700h on self-rated `fatigue' in a normal environment. The response was not
immediate in all subjects and achieved group significance after 4 days and suggests an
effect on the timing mechanisms of fatigue. Tzischinsky et al (1992a) have demonstrated
a time-dependent delayed effect of melatonin on the nocturnal sleep gate. These studies
lend further support for an effect of melatonin on the timing mechanisms of sleep and
320
wakefulness.
In consideration of the present results and of the other studies described above,
acute changes in core body temperature induced by melatonin may be a primary event
in phase-shifting mechanisms.
321
CHAPTER 11
GENERAL DISCUSSION
General Discussion
11.1 DETERMINING THE BEHAVIOUR OF THE ENDOGENOUS BODY
CLOCK
When developing models to study circadian rhythm disturbances in a normal
environment, it is important to use a circadian rhythm marker that gives a reliable
indication of the behaviour of the endogenous body clock.
11.1.1 Masking
Most measured circadian rhythms are influenced by external factors which
disturb or mask the endogenous component of the rhythm and this can lead to
inaccurate estimates of the phase and amplitude of the internal biological clock. It
is therefore important to consider any factors that may influence the endogenous
rhythmicity of measured of circadian rhythms.
Melatonin and core body temperature are extensively used as circadian phase
markers. Of the exogenous factors known to alter melatonin levels in humans, light
is well recognised (Lewy et al, 1980). The influence of gravity on melatonin
concentrations has now been demonstrated in this thesis.
Melatonin in plasma and saliva is concentrated by moving from a supine to a
standing position, and diluted after an opposite change in posture. Like other plasma
constituents, the effect can be explained through the influence of gravity which causes
plasma volume to fall on standing and increase on lying down (e. g. Tan et al, 1973).
323
Potential masking effects on the melatonin rhythm due to light and posture can
therefore be avoided by controlling the exposure to light and ensuring subjects
maintain the same posture for at least 20 minutes prior to sampling. Controlling
potential masking influences on body temperature is more difficult. Sleep, activity,
posture and bright light are all known to influence core body temperature (Mills et
al, 1978b; Wever, 1985b; Moog and Hildebrant, 1986; Minors and Waterhouse,
1987,1989; Dawson and Campbell, 1991; Myers and Badia, 1993). Masking
influences on the core body temperature rhythm can be eliminated or minimised by
the use of constant routine protocols (Mills et al, 1978a; Czeisler et al, 1989).
However, this procedure is inappropriate for the day to day study of circadian
rhythms in a normal environment.
An alternative method for separating the two components was developed by
Minors and Waterhouse (1992,1993). This mathematical procedure was employed
to purify the core body temperature rhythm after simulated time-zone travel (in
collaboration with Drs Waterhouse and Minors). Masking appeared to lead to
overestimates in the rate of adaptation of the endogenous circadian pacemaker with
the purified data showing slower adaptation. Such methods may provide a means of
obtaining more reliable estimates of the behaviour of the endogenous body clock and
allow data to be collected under normal conditions.
11.1.2 Different markers of melatonin secretion
Melatonin is considered to be a reliable marker of the endogenous body clock
and a number of methods are now used to determine the circadian parameters of
324
melatonin secretion. There are obvious difficulties in the frequent collection of blood
in order to obtain the complete circadian profile of melatonin. Measuring only the
onset of its secretion in blood as a phase marker reduces blood loss and does not
compromise sleep in most situations (Lewy et al, 1989). However, this approach
gives no information on amplitude or offset of melatonin secretion. Non-invasive
techniques have also been employed to measure the melatonin rhythm in saliva and
its metabolite, aMT6s, in urine and have been variously compared under normal
conditions (section 4.5). The usefulness of these methods in determining circadian
parameters of melatonin secretion had not previously been evaluated after a phase-
shift. This thesis describes the close relationship between plasma melatonin onset and
the calculated peak times of melatonin in plasma and saliva, and aMT6s in urine and
suggests that all indices give reliable information about the behaviour of the
endogenous melatonin circadian rhythm. If saliva is not sampled during the sleep
phase cosinor estimates of the melatonin rhythm are inadequate.
When the melatonin rhythm adapts to an advance shift, onset times in plasma
and saliva will usually occur during the sleep phase and thus cannot be measured.
Offset times may be appropriate in these situations. However, there is some evidence
for different mechanisms affecting the onset and offset of melatonin secretion
(Illnerovä and Vanecek, 1982; Illnerovä et al, 1989; Skene et al, 1994). The
melatonin offset as a phase marker therefore needs further evaluation. The most
appropriate index of melatonin secretion clearly depends upon the experimental
design.
325
11.2 AN EXPERIMENTAL MODEL TO STUDY ADAPTATION
STRATEGIES
On arrival at their destination, time-zone travellers encounter a pattern of
sunrise and sunset, activity and social schedules shifted in time. The endogenous
circadian timing system is slow to adapt to the new time cues, and until the correct
phase relationship between biological rhythms and external zeitgebers is established,
a host of physiological and behavioural problems are manifest. A similar problem
is encountered with shift workers operating to new work schedules out of phase with
the normal light/dark cycles, other competing zeitgebers and their endogenous
biological clock.
11.2.1 Adaptation to gradual phase-shifts
The circadian system usually takes 0.5-1.5h/day to adjust to new local time
cues (Aschoff, 1975; Klein and Wegmann, 1980; Folkard et al, 1991). However, by
increasing the strength of the zeitgebers at suitable times it is possible speed up the
rate of adjustment.
In a partially controlled environment, suitably timed bright light (" 1200 lux)
and darkness combined with a shift in the sleep/wake cycle successfully promoted the
adaptation of various physiological and behavioural functions to a gradual 9h delay
and advance phase-shift within 3-5 days. No major detrimental effects on body
physiology and behaviour were evident throughout the whole adjustment to such a
delay phase-shift. This method may therefore provide a means of facilitating
326
adaptation to new time-zones or shift work. For example, by using this paradigm for
night shift workers starting a new shift from midnight to 0900h, individuals could be
shifted or "nudged" progressively later by 3h a day for 3 days (perhaps over the
weekend) before the first night shift, with further light treatment on the night shift to
reinforce the 9h phase-shift.
This "nudging" technique has also been employed by Eastman (1987,1991)
using a 26h light/dark schedule (i. e. daylength progressively lengthened by 2h/day)
with bright light to phase-shift circadian rhythms to the desired time. It is important
to note that in the studies conducted by Eastman and colleagues, strict control of
natural ambient light after sleep was necessary to achieve this effect.
Both these techniques may be impractical for individuals not wishing to
undergo such an elaborate sleep/wake and light/dark schedule prior to the new shift
or new time-zone. An alternative method involves the use of much brighter intensity
light (=10 000 lux) centred directly over the temperature minimum; a time at which
the circadian system is most sensitive to large phase-shifts (Czeisler et al, 1989).
This technique has been employed in a shift work study by Czeisler et al (1990) using
bright light during the night shift together with darkness during the day and with no
treatment administered prior to the new shift. Complete adaptation of circadian
variables and improvements in alertness and performance were evident within 6 days.
However, in a study comparing the bright light "nudge" technique with the rapid shift
method, Eastman (1987) found that although the latter technique induced a faster
phase-shift, sleep and alertness were compromised. Clearly, further research is
needed to evaluate the most effective and practical method of promoting adaptation
to new time schedules.
327
11.2.2 Adaptation to rapid shifts - simulation of jet lag and
maladaptation to shift work
Although laboratory simulations can provide detailed information about the
adaptation process of the circadian system to changes in local time cues, there is still
a considerable difference to the real life situation. The critical feature of
environmental isolation studies is the elimination of every day zeitgebers. Treatment
strategies to facilitate adaptation to shift work or new time-zones must be effective
in the presence of competing zeitgebers which are present in normal every day life.
After successful adaptation to a gradual 9h delay or advance phase-shifts, the
subjects in these studies returned to their normal environment and baseline sleep/wake
cycle, thus experiencing a rapid 9h delay or advance phase-shift in local time cues.
The subsequent disturbances in behavioural and physiological functions was
comparable to that found in field conditions after transmeridian flight. For shift
workers, this is equivalent to a fully-adapted night shift worker returning to a day
shift. This experimental model therefore provides an effective and inexpensive
method of simulating aspects of jet lag and shift work so that the effectiveness of
various treatment strategies can be assessed in real life.
11.2.3 Preferred direction of adaptation
The faster rate of adaptation observed with westward (delayed time)
transmeridian flights compared to travelling east (advanced time) in field studies (e. g.
Klein and Wegmann, 1980) has been confirmed by our studies comparing the rate of
328
adjustment to both rapid and gradual 9h advance and delay phase-shifts. The most
striking observation was that the direction of the phase-shift was more important than
the speed of the shift.
Whereas bright light was effective in adjusting the circadian system to a
gradual 9h delay, it was less successful in the adaptation to a gradual 9h advance
phase-shift; sleep, mood and performance were initially compromised in the latter
case. In addition, physiological and behavioural variables took far longer to adapt to
a rapid 9h advance phase-shift compared to a rapid 9h delay shift. This underlines
the preferred direction of reentrainment of the biological clock and confirms that the
direction of the rotating shift schedule is important to consider when designing new
work timetables. Studies directly comparing the direction of shift rotation have
shown that delaying the shift schedule (i. e. mornings - afternoons -* nights -º
mornings) reduces fatigue and improves sleep compared to advancing them (i. e.
nights -º afternoons -º mornings -º nights) (Landen et al 1981; Epstein et al, 1991;
Barton and Folkard, 1993).
11.3 EFFECTS OF BRIGHT LIGHT AND MELATONIN DURING
ADAPTATION TO RAPID PHASE-SHIFTS.
Can bright light counteract the beneficial effects of melatonin ?
Light and melatonin, suitably applied, can reset the timing of the endogenous
biological clock, but their actions are opposite in phase. Time-zone travellers and
shift workers are often exposed to bright light at times which oppose rapid adaptation.
It is conceivable that such exposure may compromise the beneficial effects of
329
melatonin treatment. This possibility was therefore investigated after a rapid 9h
advance phase-shift.
Melatonin (5mg), timed to phase-advance, clearly improved subjective sleep,
alertness and performance and facilitated the adaptation of the aMT6s rhythm, after
a rapid 9h advance shift of external time cues both in the presence or absence of
bright light (2000 lux) timed to counteract the phase-shifting properties of melatonin.
This observation has practical benefits for time-zone travellers and shift workers since
they may carry out their normal activities in natural ambient light (at least up to 2000
lux) during melatonin treatment. Although melatonin proved to be the dominant
`zeitgeber' in this study, stronger intensities of light may prove more effective in
opposing the effects of melatonin treatment.
11.4 REPEATED MELATONIN ADMINISTRATION: quickest adaptation by
changing times of administration?
Repeated doses of melatonin were not necessary to observe phase-shifting
effects on the biological clock. A single melatonin treatment (0.05mg to 5mg) was
sufficient to induce phase-shifts in the endogenous circadian rhythms of core body
temperature and melatonin secretion. The use of repeated melatonin administration
at the same times each day for circadian rhythm disorders may be questionable.
Lewy et al (1992) have described a phase-response curve for melatonin after repeated
administration of melatonin for 4 days. However, the current results suggest that the
phase-response curve would shift after each melatonin treatment. Therefore, repeated
treatment at the same times over many days or weeks could conceivably be ineffective
330
or counterproductive depending on the part of the PRC on which treatment falls. If
the phase-response curve shifted with each treatment such that the following treatment
fell within the zones of little or no response, or across the inflexion point where
phase-delays are induced instead of phase-advances and vice versa, melatonin may
prove to have little beneficial effect.
Clearly repeated administration of melatonin at the same clock times is
effective when administered just before the desired bedtime and may provide a
constant entraining effect (e. g. Arendt et al, 1987; Folkard et al, 1990; Jan et al,
1994). However, changing the time of administration each day so that the treatment
falls within the zone of maximum phase-shift may speed up the resynchronization
process for jet travellers and shift workers.
11.5 MECHANISM OF ACTION OF MELATONIN
11.5.1 Phase-shifting effect on the biological clock
The most commonly reported effects of melatonin treatment in circadian
rhythm disturbances have been improvements in mood and sleep (e. g. Arendt et al,
1987; Folkard et al, 1990; Jan et al, 1994). Redman et al (1983) suggested that
melatonin's effect on rest (sleep)/activity may occur via a phase-resetting action on
the biological clock. Certainly, as has been demonstrated in this thesis and by Zaidan
et al (1994) single melatonin treatments can phase-shift the endogenous clock, but by
less than 1.5h in entrained conditions. However, after a rapid 9h advance phase-shift,
subjective sleep improved significantly on the first night following administration (one
331
hour before bedtime) with no significant adaptation of the urinary aMT6s or purified
temperature rhythms. It seems unlikely that this immediate effect on sleep was
attributable to an effect on the endogenous clock and suggests that other mechanisms
are involved.
11.5.2 `Hypnotic' effects
It has been suggested that the beneficial effects of melatonin on sleep are due
to its `hypnotic' properties similar to that seen with benzodiazepines, and that
endogenous melatonin is involved in normal physiological sleep inducing mechanisms
(Waldhauser et al, 1990; Dollins et al, 1993,1994). However, studies investigating
alterations in sleep architecture have been conflicting and any changes seen were
minor (see section 1.2.10). At a dose commonly used to treat circadian rhythm
disorders, 5mg melatonin has been found to slightly increase REM latency (James et
al, 1987) in normal individuals and reduce the latency from bedtime to stage 2 sleep
in patients suffering from delayed sleep phase syndrome, suggesting a phase-resetting
and initiation mechanism (Dahlitz et al, 1991).
Benzodiazepines have proved beneficial in facilitating sleep adaptation after
transmeridian flight (Nicholson et al, 1985; Donaldson and Kennaway, 1991).
Melatonin has been shown to interact with the GABAergic system (Rudeen et al,
1980; Rosenstein and Cardinali, 1986; Coloma and Niles, 1988) through which the
benzodiazepines are likely to exert their sedative and anxiolytic properties (Enna and
Mähler, 1987). Although melatonin may enhance these properties of
benzodiazepines, it does not itself behave like a benzodiazepine (Guardiola-Lemaltre
332
et at, 1992). In rodents sedative and anxiolytic activity occurs at much higher doses
than the levels needed to entrain circadian rhythms (Tobler et al, 1994). Sedative
effects of melatonin were only significant at pharmacological doses in the studies
described in this thesis.
It must also be noted that the acute drowsiness observed in the experiments
described in this thesis and of Dollins et al (1993,1994) using low oral doses of
melatonin may be related to the experimental conditions whereby subjects seated in
a warm room in dim light and were asked to perceive sleepiness more clearly than
when active. Other studies in a normal environment have shown that administration
of 2mg melatonin at 1700h induced clear drowsiness only after 4 days (Arendt et al,
1984,1985a), which suggests an effect on the timing mechanisms of sleep rather than
a sedative effect. In addition, in previous experiments using 5mg melatonin to
facilitate adaptation to a new time-zone, subjects could not usually distinguish
melatonin from placebo (Arendt et at, 1987; Samel et al, 1991). Clearly the activity
of the subject may easily override any melatonin-induced sleepiness.
Melatonin-induced drowsiness may simply be a result of temperature lowering
effects (Dawson and Encel, 1993). The sedative-like effects of melatonin on mood
and performance followed the same time course as the suppression of body
temperature. Certainly there is strong evidence for a thermoregulatory role for
melatonin (Strassman et al, 1991; Cagnacci et al, 1992,1993; Badia et al, 1993) and
a strong relationship between the free-running rhythms of sleep propensity, body
temperature and melatonin (Nakagawa et al, 1992).
Although, it seems unlikely that endogenous melatonin induces physiological
sleep, melatonin may prepare the body for sleep in some way. With pharmacological
333
doses, melatonin may exert its beneficial effects through a combination of sedation,
temperature regulation and phase-shifting mechanisms.
11.6 MASKING OR CUE FOR ENTRAINMENT ?
Although external masking by the subject's behaviour and environment is
considered undesirable by those interested in the pattern of the internally generated
rhythm, some types of masking may well be important in normal physiology as a part
the adaptive homeostatic mechanisms of the body, enabling the body to respond to
environmental changes.
Aschoff (1987) suggests that external masking from zeitgebers (e. g. from light
and sleep) augments endogenous circadian rhythms and strengthens the temporal
relationship with other circadian rhythms and the environment. In addition to this
direct effect on the endogenous clock, Minors and Waterhouse (1989) underline the
importance of "internal masking". In this case, physiological or biochemical changes
occurring within the body as result of external factors could affect the circadian
timing system. For example, the acute changes in body temperature in response to
melatonin or bright light may provide additional input distinct from direct
entraining/phase-shifting effects. When external time cues such as the light/dark
cycle and sleep/wake schedule are in phase with the internal body clock, they may
strengthen the circadian timing system. When out of phase, acute masking effects of
the external zeitgebers such as the light/dark and sleep/wake cycles are in phase with
the new shift in local time, a time to which the body needs to adapt to. Clearly
masking from exogenous zeitgebers may provide an appropriate time cue. Suitably
334
timed exogenous melatonin may have similar effects.
11.7 RELATIONSHIP BETWEEN THE ACUTE EFFECTS AND PHASE-
SHIFTING PROPERTIES OF MELATONIN AND BRIGHT LIGHT
The intensity of light required to shift the timing of circadian rhythms is
enough to cause the suppression of melatonin secretion (Arendt, 1988; Lewy et al,
1980; McIntyre et al, 1989). The phase-shifting ability of bright light is associated
with a short term fall in the body temperature rhythm amplitude (Czeisler et al, 1989;
Jewett et al, 1991). Cagnacci and co-workers (1992) have suggested that this
reduction in amplitude is similar to that obtained with atenolol suppression of
melatonin and speculate that the suppression of melatonin may be involved in the
phase-shifting mechanisms of bright light.
The phase-response curves for light (using core body temperature and
melatonin as circadian phase markers, Czeisler et al, 1989; Minors et al, 1991; van
Cauter et al, 1994) and melatonin (using its own endogenous rhythm as a marker,
Lewy et al, 1992; Zaidan et al, 1994) are nearly opposite in phase. Furthermore,
through indirect and direct evidence, the acute effects of bright light and melatonin
on body temperature have been shown to be inversely related (Strassman et al, 1991;
Badia et al, 1993; Campbell and Dawson, 1990).
This thesis describes the dose-dependent suppression of core body temperature
with a concomitant fall in alertness following melatonin treatment during the day.
Conversely, Myers and Badia (1993) describe the dose-dependent stimulation of body
temperature and alertness with bright light. In the latter study, light-induced
335
stimulation of temperature and alertness occurred only at light intensities found to
suppress melatonin secretion and after the melatonin onset time. Furthermore, bright
light exposure during the day does not appear to elevate body temperature (Murphy
et al, 1991; French et al, 1990).
These acute effects of bright light on body temperature are only clear during
the late evening, nighttime hours and early morning (i. e. during periods of darkness),
and at times when the circadian system is susceptible to the phase-shifting effects of
bright light. Unfortunately, the acute effects of exogenous melatonin on temperature
have not been studied during the nighttime hours in the presence of endogenous
melatonin. However, administration during the day causes suppression of body
temperature (Cagnacci et al, 1992; Dollins et al, 1993,1994; Deacon et al, 1994).
It would be interesting to establish any time-dependency of the acute effects of
melatonin on body temperature. The suggestion is that melatonin would only be
effective at times when phase-shifts can be induced. This thesis has clearly shown
that prior to its phase-shifting action on the endogenous body clock, melatonin has
acute dose-related effects on body temperature, and that these acute effects are highly
correlated to the degree of shift in the endogenous biological clock. A recent study
indicates that slow changes in ambient temperature at night can phase-shift the
temperature rhythm (Dewasmes et al, 1994), strengthening the suggestion that
thermoregulatory responses may be involved in phase-shifting mechanisms.
In consideration of the results shown in this thesis and of the other studies
described, acute changes in body temperature by bright light and melatonin may be
a primary event in phase-shifting mechanisms.
336
11.8 FUTURE WORK
Given that bright light can induce large phase-shifts in the endogenous clock
which are nearly opposite in phase to those of melatonin, a combination of
treatments, when suitably timed, may reinforce each other and provide the most
effective phase-shifting strategy. Future experiments should consider the use bright
light (of varying intensities) in combination with melatonin to facilitate the gradual
and rapid adaptation to advance phase-shifts. This direction of shift clearly induces
the greatest behavioural problems and circadian rhythm disturbances.
A complete intensity-response curve showing the degree of shift with varying
intensities and duration of light should be carried out to determine the optimum
conditions for efficient phase-shifting. Relating the degree of shift to the magnitude
of melatonin suppression, stimulation of temperature and alertness, will strengthen the
suggestion that light-induced effects on physiology and behaviour are mediated
through changes in melatonin and body temperature.
It is still not known whether the duration of exogenous melatonin treatment
is important. Clearly oral doses of fast-release melatonin do not reflect the nocturnal
profile of melatonin secretion. Intravenous infusions (Strassman et al, 1987; Zaidan
et al, 1994) and sustained release preparations (Beni s et al, 1993) which mimic
nocturnal melatonin secretion should provide a useful tool to study the role of
endogenous melatonin in thermoregulation and phase-shifting mechanisms. The
ability of a single melatonin treatment to phase-shift the endogenous body clock
indicates the importance of establishing a PRC based on single oral treatments similar
to that described by Zaidan et al (1994) using intravenous infusions. For clinical
337
purposes, a PRC based on single oral melatonin treatment would be of great practical
benefit. The most effective dose of melatonin for phase-shifting the endogenous body
clock was 5mg. This particular oral dose has been used extensively to treat circadian
rhythm disorders (Arendt et al, 1987; Petrie et al, 1989; Folkard et al, 1990; Dahlitz
et al, 1991; Tzischinsky et al, 1993). It will be of interest to study higher doses in
future experiments.
As single melatonin treatments can shift the endogenous body clock and
therefore the PRC within one day, future experiments should compare the phase-
shifting effects of repeated daily melatonin administration at the same times with a
schedule which alters the time of melatonin treatment daily to take account of the
altered PRC each day. Changing the time of administration should provide the
optimum phase-shifting efficiency. However, such a treatment paradigm may be
impractical when the desired administration time falls within the sleep phase, although
delayed release preparations may prove beneficial in these situations. Furthermore,
melatonin administration during the day may induce undesirable drowsiness.
Melatonin treatment strategies should therefore be evaluated with these considerations
in mind.
Melatonin can induce temperature suppression and drowsiness and poor
performance. Drugs which inhibit the secretion of melatonin may therefore reverse
this effect. There is indirect evidence for this. B-antagonists, which suppress
melatonin synthesis, are associated with elevated body temperature at night (Cagnacci
et al, 1992) and disturbed sleep together with reduced alertness during the day
(Dimenas et al, 1990, McAinsh and Cruickshank, 1990).
In contrast, it would be interesting to see whether drugs or situations which
338
alter body temperature affect melatonin production. There is limited evidence to
suggest that non-steroidal anti-inflammatory drugs reduce melatonin production in
humans (Bird et al, 1986; Murphy et al, 1993) and elevate body temperature at night
(Badia et al, 1993; Murphy et al, 1993). It would be interesting to see if preventing
melatonin-induced temperature suppression hinders the phase-shifting properties of
melatonin.
It is not known whether melatonin acts peripherally to adjust body temperature
or centrally within the thermoregulatory centres of the hypothalamus. It is possible
that the acute effects of melatonin on temperature provide information to the SCN
indirectly via thermoregulatory responses. This in turn may modulate the timing and
amplitude of the temperature rhythm as part of the "adaptive homeostatic"
mechanisms of the body. The close anatomical and neural association between the
thermoregulatory centres in the hypothalamus and the SCN, the presence of putative
melatonin receptors within the hypothalamus, together with the close association
between melatonin-induced temperature suppression and its phase-shifting effects,
suggest the involvement of hypothalamic control of body temperature in phase-shifting
responses and provides a new direction for future research into circadian phase-
shifting mechanisms.
339
PUBLICATIONS AND PRESENTATIONS
Publications and Presentations
Deacon S. J. and J. Arendt, 1994. Phase-shifts in melatonin, 6-sulphatoxymelatonin
and alertness rhythms after treatment with moderately bright light at night. Clinical Endocrinology, 40 (3), 413-420. Part of this work was presented at
the European Society for Pineal Research, Copenhagen, Denmark, July, 1993.
Deacon S. J. and J. Arendt, 1994. Posture influences melatonin concentrations in
plasma and saliva in humans. Neuroscience letters, 167,191-194. This work
was presented at the 184th meeting for the Society for Endocrinology jointly
with the Endocrine Section of the Royal Society of Medicine, London,
November, 1993.
Deacon S. J., J. English and J. Arendt, 1994. Acute phase-shifting effects of
melatonin associated with temperature suppression in humans. Neuroscience
Letters, 178,32-34. This work was presented at the 4th meeting for the
Society for Research on Biological Rhythms, Florida, USA, May, 1994.
Deacon S. J. and J. Arendt, 1994. Forced phase-shifts in partial environmental isolation. In: Work Hours, Sleepiness and Accidents, Stress Research
Reports, No. 248, T. Akerstedt and G. Kecklund (eds. ), Karolinksa Institute,
Stockholm, Sweden. This work was presented at the conference on "Work
Hours, Sleepiness and Accidents", Stockholm, Sweden, September, 1994.
Deacon S. J., J. English and J. Arendt, 1994. Does the phase-shifting ability of
melatonin relate to core body temperature suppression ? In: Work Hours,
Sleepiness and Accidents, Stress Research Reports, No. 248, T. Akerstedt and G. Kecklund (eds. ), Karolinksa Institute, Stockholm, Sweden. This work was
presented at the conference on "Work Hours, Sleepiness and Accidents",
Stockholm, Sweden, September, 1994.
341
Deacon S. J. and J. Arendt, 1994. Use of melatonin to adapt to phase-shifts, I.
Melatonin counters sleep problems after a large advance shift in external time
cues in spite of ambient light conditions, abstract. Fourth meeting for the
Society for Research on Biological Rhythms, Florida, USA, May, 1994.
Deacon S. J. and J. Arendt, 1994. Use of melatonin to adapt to phase-shifts, II.
Mood and performance after a large advance shift in external time cues,
abstract. Fourth meeting for the Society for Research on Biological Rhythms,
Florida, USA, May, 1994.
Deacon S. J. and J. Arendt, 1993. Decrements in performance and sleep in response
to a phase-shifting light stimulus (1200 lux) at night. J. Interdisciplinary
Cycle Research, 24,264-266. This work was presented at the European
Society for Chronobiology, Bath, September, 1993.
Deacon S. J., J. Arendt and J. English, 1993. Posture: a possible masking factor of
the melatonin circadian rhythm - Comparison of plasma and salivary melatonin levels in high frequency sampling. Proceedings of the International
Symposium on the Pineal Gland, Paris, September 1992. In: Melatonin and
the Pineal Gland - From Basic Science to Clinical Application, Y.
Touitou, J. Arendt & P. PBvet (eds. ), pp. 387-390, Elsevier, New York.
Skene D. J., S. Deacon and J. Arendt, 1994 (in press). Light and melatonin
responses in humans. Proceedings of the Fourth International Scientific
conference on "Work with Display Units" Italy, October 1994.
Submitted (1994)
Deacon S. J. and J. Arendt. Melatonin-induced temperature suppression and its phase-
shifting effects correlate in a dose-dependent manner in humans. Brain
Research.
342
REFERENCES
Abalan F., Hui Bon Hoa A., Guillonneau D., Ruedas E., Calache M., Ellison W., Rigal F., Perey F., Sousselier M. and Bourgeois M. Effect of posture on total cortisol plasma concentration. Horm. Metab. Res. 24: 595-596,1992.
Abo T., Kawate T., Itoh K., and Kumagai K. Studies on bioperiodicity of the immune response, I: Circadian rhythms of human T, B, and K cell traffic in the peripheral blood. J. Immun. 126: 1360-1363,1981.
Akerstedt T. Sleepiness as a consequence of shift work. Sleep 11: 17-34,1988.
Akerstedt T. and Folkard S. A model of human sleepiness. In: Sleep'90, J. Home (ed. ), Pontenagel Press, Bochum, pp 210-313,1990.
Akerstedt T., Froberg J. E., Friberg Y. and Wetterberg L. Melatonin excretion, body temperature and subjective alertness during 64h of sleep deprivation. Psychoneuroendocrinology 4: 219-225,1979.
Akerstedt T. and Gillberg M. Sleep disturbances and shift work. In: Night and Shift Work: Biological and Social Aspects, A. Reinberg, N. Vieux and P. Andlauer (eds. ), Oxford, Pergamon Press, pp 127-137,1981.
Akerstedt T., Knutsson A., Alfredsson L. and Theorell T. Shiftwork and cardiovascular disease. Scand. J. Work Environ. Health 10: 409-414,1984.
Akerstedt T., Pätkai P. and Dalgren K. Field studies of shiftwork II. Temporal patterns in psychological activation in workers alternating between night and day work. Ergonomics 20: 621-631,1977.
Akerstedt T. and Torsvall L. Shiftwork: shift-dependent well-being and individual differences. Ergonomics 24: 265-273,1981.
Aldhous M. E. and Arendt J. Radioimmunoassay for 6-sulphatoxymelatonin in urine using an iodinated tracer. Ann. Clin. Biochem. 25: 298-303,1988.
Aldhous M., Franey C., Wright J. and Arendt J. Plasma concentrations of melatonin in man following oral adsorption of different preparations. Br. J. Clin. Pharm. 19: 517-521,1985.
Alvarez B., Dahlitz M. J., Vignau J. and Parkes J. D. The delayed sleep phase syndrome: clinical and investigative findings in 14 subjects. J. Neurol. Neurosurg. Psychiatry 55: 665-670,1992.
Anton-Tay F. Melatonin: effects on brain function. In: Serotonin: Biochemistry and Behavioural and Clinical Studies. Advances in Biochemical Psychopharmacology, E. Costa, G. L. Cessa and M. Sandler (eds. ), Raven Press, New York, vol. 11, pp 315- 324,1974.
344
Arendt J. Role of the pineal gland and melatonin in seasonal reproductive function in mammals. Oxford Rev. Reproductive Biology 8: 266-320,1986.
Arendt J. Melatonin. Clin. Endocrinol. 29: 205-229,1988.
Arendt J. Use of melatonin in circadian rhythm disturbances. Biogenic Amines 9: 469-471,1993.
Arendt J. Clinical perspectives for melatonin and its agonists. Biol. Psychiatry 35: 1- 2,1994.
Arendt J. and Aldhous M. Further evaluation of the treatment of jet lag by melatonin: a double blind crossover study. Ann. Rev. Chronopharmacol. 5: 53-55, 1988.
Arendt J., Aldhous M. and Marks V. Alleviation of jet-lag by melatonin: preliminary results of a controlled double blind trial. Br. Med. J. 292: 1170,1986.
Arendt J., Aldhous M. and Wright J. Synchronization of a disturbed cycle in a blind man by melatonin treatment. Lancet 1: 772-773,1988.
Arendt J., Aldhous M., English J., Marks V. Marks M. and Folkard S. Some effects of jet lag and their alleviation by melatonin. Ergonomics 30: 1379-1393,1987.
Arendt J., Bojkowski C., Folkard S., Franey C., Marks V., Minors D., Waterhouse J., Wever R. A., Wildgruber C. and Wright J. Some effects of melatonin and the control of its secretion in humans. In: Photoperiodism, Melatonin and the Pineal Gland, D. C. Evered and Clark S. (eds. ), Ciba Foundation Symposium 117, Pitman, London, pp 266-283,1985a.
Arendt J., Bojkowski C., Franey C., Wright J. and Marks V. Immunoassay of 6- hydroxymelatonin sulphate in human plasma and urine: abolition of the urinary 24- hour rhythm with atenolol. J. Clin. Endocrinol. Metab. 60: 1166-1173,1985b.
Arendt J., Borb6ly A. A., Franey C. and Wright J. The effects of chronic, small doses of melatonin given in the late afternoon on fatigue in man: a preliminary study. Neurosci. Lett. 45: 317-321,1984.
Arendt J. and Broadway J. Phase response of human melatonin rhythms to bright light in Antarctica. J. Physiol. 377: 68P, 1986.
Arendt J., Brown W. B., Forbes J. M. and Marsden A. Effect of pinealectomy on immunoassayable melatonin in sheep. J. Endocrinol. 85: 1-2P, 1980.
Arendt J., Wirz-Justice A. and Bradtke J. Annual rhythm of serum melatonin in man. Neurosci. Leu. 7: 327-330,1977.
345
Arendt J., Wirz-Justice A., Bradtke J. and Kornemark M. Long-term studies on immunoreactive human melatonin. Ann. Clin. Biochem. 16: 307-312,1979.
Arendt J., Hampton S., English J., Kwasowski P. and Marks V. 24-hour profiles of melatonin, cortisol, insulin, C-peptide and GIP following a meal and subsequent fasting. Clin. Endocrinol. 16: 89-95,1982.
Argoblast B., Lubanovic W., Halberg F., Cornelissen G. and Bingham C. Chronobiologic serial sections of several orders. Chronobiologia 10: 59-68,1983.
Armstrong S. M., Cassone V. M., Chesworth M. J., Redman J. R. and Short R. V. Synchronization of mammalian circadian rhythms by melatonin. J. Neural Trans. (Suppl. ) 21: 375-394,1986.
Armstrong S. M., McNulty O. M., Guardiola-Lemaitre B. and Redman J. R. Successful use of S20098 and melatonin in an animal model of delayed sleep-phase syndrome (DSPS). Pharmacol. Biochem. Behav. 46: 45-49,1993.
Armstrong S. M. and Redman J. Melatonin administration: effects on rodent circadian rhythms. In: Photoperiodism, Melatonin and the Pineal, D. C. Evered and S. Clark (eds. ). Ciba Foundation Symposium 117, Pitman, England, pp 188-202, 1985.
Aschoff J. Circadian rhythms in man: a self-sustained oscillator with an inherent frequency underlies human 24-hour periodicity. Science 148: 1427-1432,1965.
Aschoff J., Hoffman K., Pohl H. and Wever R. Reentrainment of circadian rhythms after phase-shifts of the zeitgeber. Chronobiologia 2: 23-78,1975.
Aschoff J. Masking of circadian rhythms by zeitgebers as opposed to entrainment. In: Proceedings of the 18th International Conference on Chronobiology, ISC Leiden, Holland, July 1987.
Aschoff J. and Wever R. The circadian system of man. In: Handbook of Behavioural Neurobiology, Biological Rhythms, J. Aschoff (ed. ), New York, Plenum, Vol. 4,1981.
Attanasio A., Borrelli P. and Gupta D. Circadian rhythms in serum melatonin from infancy to adolescence. J. Clin. Endocrinol. Metab. 61: 388-390,1985.
Axelrod J. and Reisine T. D. Stress hormones: their interaction and regulation. Science 224: 452-459,1984.
Axelrod J. and Zatz M. The ß-adrenergic receptor and the regulation of circadian rhythms in the pineal gland. In: Biochemical Actions of Hormones, G. Litwack (Ed. ) Vol. 4. Academic Press, New York, pp 249-268,1977.
346
Axelrod J. and Weissbach H. Enzymatic O-methylation of N-acetylserotonin to melatonin. Science 131: 1312,1960.
Axelrod J. The pineal gland: A neurochemical transducer. Science 184: 1341-1348, 1974.
Badia P., Culpepper J., Myers B., Boecker M. and Harsh J. The immediate effect of bright and dim light on body temperature. Sleep Res. 19: 386,1990a.
Badia P., Myers B., Boecker M. and Murphy P. Effects of bright and dim light on body temperature, brain activity and behaviour. Psychophysiology 27 (Suppl. 4A): S 15,1990b.
Badia P., Myers B., Boecker M. and Harsh J. Bright light effects on body temperature, alertness, EEG and behaviour. Physiol. Behav. 50: 583-588,1991.
Badia P., Myers B. and Murphy P. J. Melatonin and thermoregulation. In: Melatonin: Biosynthesis, Physiological Effects, and Clinical Applications, H. -S. Yu and R. J. Reiter (Eds. ), CRC Press, Florida, pp 349-364,1993.
Bartness T. J., Powers J. B., Hastings M. H., Bittman E. L. and Goldman B. D. The timed infusion paradigm for melatonin delivery: what has it taught us about the melatonin signal, its reception, and the photoperiodic control of seasonal responses? J. Pineal Res. 15: 161-190,1993.
Barton J. and Folkard S. Advancing versus delaying shift systems. Ergonomics 36: 59-64,1993.
Bartsch H. and Bartsch C. Effect of melatonin on experimental tumors under different photoperiods and times of administration. J. Neural. Trans. 52: 269-279, 1981.
Bassi C. J. and Powers M. K. Daily fluctuations in the detectability of dim light by humans. Physiol. Behav. 38: 871-877,1986.
Beck-Friis J., Von Rosen D., Kjellman B. F., Ljunggren J. G. and Wetterberg L. Melatonin in relation to body measures, sex, age, season and the use of drugs in patients with affective disorders and healthy subjects. Psychoneuroendocrinology 9: 261-277,1984.
Beck-Friis J., Borg G. and Wetterberg. Rebound increase of nocturnal serum melatonin levels following evening suppression by bright light exposure in healthy men: Relation of cortisol levels and morning exposure. Ann. N. Y. Acad. Sci. 453: 371-375,1985.
347
B6n6s L., Brun J., Claustrat B., Degrande G., Ducloux N., Geoffriau M., Horribre F., Karsenty H. and Lagain D. Plasma melatonin and 6-sulphatoxymelatonin kinetics after transmucosal administartion to humans. In: Melatonin and the Pineal Gland - From Basic Science to Clinical Application, Y. Touitou, J. Arendt and P. Pdvet (eds. ), Elsevier Science Publishers, Amsterdam, pp 347-350,1993.
Berk M. L. and Finklestein J. A. An autoradiographic determination of the afferent projections of the suprachiasmatic nucleus of the hypothalamus. Brain Res. 226: 1-13, 1981.
Bird H., Othman H., Smith J. A. and Suddall. Effect of ibruprofen on human serum melatonin. J. Neural Trans. 21: 485,1986.
Bittman E. L., Dempsey J. and Karsch F. J. Pineal melatonin secretion drives the reproductive response to daylength in the ewe. Endocrinology 113: 2276-2283,1983a.
Bittman E. L., Karsch F. J. and Hopkins J. W. Role of the pineal gland in ovine photoperiodism: regulation of seasonal breeding and negative feedback effects of estradiol upon luteinizing hormone secretion. Endocrinology 113: 329-336,1983b.
Blask D. E., Lemus-Wilson A. M., Wilson S. Y. and Cos S. Neurohormonal modulation of cancer growth by pineal melatonin. In: Melatonin and the Pineal Gland - From Basic Science to Clinical Application, Y. Touitou, J. Arendt, P. Privet (eds. ), Excerta Medica, Amsterdam, pp303-310,1993.
Blask D. E. The pineal: an oncostatic gland? In: The Pineal Gland, Reiter R. J. (ed. ). Raven Press, New York, pp 253-284,1984.
Blood M. L., Sack R. L., Sexton G. and Lewy A. J. A comparison of the melatonin onset and acrophase as markers for circadian phase. Abstract 143. Fourth Meeting, Society for Research on Biological Rhythms, Florida, USA, May 1994.
Bojkowski C. J. 6-sulphatoxymelatonin as an index of pineal function in human physiology. PhD thesis, University of Surrey, Guildford, Surrey, UK, 1988.
Bojkowski C. J., Aldhous M. E., English J., Franey C., Poulton A. L., Skene D. J. and Arendt J. Suppression of nocturnal plasma melatonin and 6-sulphatoxymelatonin by bright and dim light in man. Horm. Metab. Res. 19: 437-440,1987a.
Bojkowski C. J. and Arendt J. Annual changes in 6-sulphatoxymelatonin excretion in man. Acta Endocrinologica 117: 470-476,1988
Bojkowski C. J. and Arendt J. Factors influencing urinary 6-sulphatoxymelatonin, a major melatonin metabolite, in normal human subjects. Clin. Endocrinol. 33: 435- 444,1990.
348
Bojkowski C. J., Arendt J., Shih M. C. and Markey S. P. Melatonin secretion in humans assessed by measuring its metabolite, 6-sulphatoxymelatonin. Clin. Chem. 33: 1343-1348,1987b.
Bonnefond C., Martinet L., Lesieur D., Adam G. and Guardiola-Lemaltre B. Characterisation of S-20098, a new melatonin analog. In: Melatonin and the Pineal Gland - From Basic Science to Clinical Application, Y. Touitou, J. Arendt and P. Pev6t (eds. ), Excerta Medica, Amsterdam, pp 127-130,1993.
Bonnet M. H., Dexter J. R., Gillin J. C., James S. P., Kripke D., Mendelson W. and Mitler M. The use of triazolam in phase-advanced sleep. Neuropsychopharmacology 1: 225-234,1988.
Bougrine S., Cabon P., Ignazzi G., Geoffriau M., Brun J. and Claustrat B. Phase delay of rhythm of 6-sulphatoxy-melatonin-excretion by bright light improves sleep and performance. In: Melatonin and the Pineal Gland - From Basic Science to Clinical Application, Y. Touitou, J. Arendt and P. Pevet (eds. ), Excerta Medica, Amsterdam, pp 219-223,1993.
Boulos Z., Rosenwasser A. M. and Terman M. Feeding schedules and the circadian organization of behaviour in rats. Behav. Brain Res. 1: 39-65,1980.
Brainard G. C., Richardson B. A., King T. S. and Reiter R. J. The influence of different light spectra on the suppression of pineal melatonin in the Syrian hamster. Brain Res. 294: 333-339,1984.
Brainard G. C., Lewy A. J., Menaker M., Fredrickson R. H., Miller L. S., Weleber R. G., Cassone V. and Hudson D. Effect of light wavelength on the suppression of nocturnal plasma melatonin in normal volunteers. Ann. N. Y. Acad. Sci. 453: 376-378, 1985.
Brainard G. C., Lewy A. J., Menaker M., Fredrickson R. H., Miller L. S., Weleber R. G., Cassone V. and Hudson D. Dose-response relationship between light irradiance and the suppression of plasma melatonin in human volunteers. Brain Res. 454: 212-218,1988.
Broadway J., Arendt J. and Folkard S. Bright light phase shifts the human melatonin rhythm during the Antarctic winter. Neurosci. Lett. 79: 185-189,1987.
Brownstein M. and Axelrod J. Pineal gland 24 hour rhythm in norepinephrine turnover. Science 184: 163-165,1974.
Brun J., Claustrat B., Harthe C. and David L. Night/day variations and daily excretion of melatonin and 6-sulphatoxymelatonin from infancy to adulthood. &p. Clin. Endocrinol. 11: 215-219,1992.
349
Brzezinski A., Lynch H. J., Seibel M. M., Deng M. H., Nader T. M. and Wurtman R. J. The circadian rhythm of plasma melatonin during the normal menstrual cycle and in amenorrheic women. J. On. Endocrinol. Metab. 66: 891-895,1988.
Brzezinski A., Seibel M. M., Lynch H. J., Deng M. -H. and Wurtman R. J. Melatonin in human preovulatory follicular fluid. J. Clin. Endocrinol. Metab. 64: 865-867, 1987.
Bunnell D. E., Triber S. P., Phillips N. H. and Berger R. J. Effects of evening bright light exposure on melatonin, body temperature and sleep. J. Sleep Res. 1: 17-23, 1992.
Buresovä M., Dvoräkovä M., Zvolsky P. and Illnerovä H. Early morning bright light phase advances the human circadian pacemaker within one day. Neurosci. Lett. 121: 47-50,1991.
Buresovä M., Dvoräkovä M., Zvolsky P. and Illnerovä H. Human circadian rhythm in serum melatonin in short winter days and in simulated artificial long days. Neurosci. Lett. 136: 173-176,1992.
Byerely W. F., Risch S. C., Gillin J. C., Janowsky D. S., Parker D., Rossman L. G. and Kripke D. F. Biological effect of light. Prog. Neuro. Psychopharmacol. Biol. Psychiat. 13: 683-686,1989.
Cagnacci A., Elliot J. A. and Yen S. C. C. Amplification of pulsatile LH secretion by exogenous melatonin in women. J. On. Endocrinol. Metab. 73: 210-212,1991.
Cagnacci A., Elliot J. A. and Yen S. C. C. Melatonin: a mjor regulator of the circadian rhythm of core temperature in humans. J. Clin. Endocrinol. Metab. 75: 447-452,1992.
Cagnacci A., Soldani, R. and Yen S. C. C. The effect of light on core body temperature is mediated by melatonin in women. J. Clin. Endocrinol. Metab. 76: 1036-1038,1993.
Campbell S. S., Stroud A. and Lebowitz S. Effects of bright light exposure on shift work adaptation in middle-aged subjects. Abstract 21. Fourth Meeting, Societyfor Research on Biological Rhythms, Florida, USA, May 1994.
Campbell S. S. and Dawson D. Enhancement of nighttime alertness and performance with bright ambient light. Physiol. Behav. 48: 317-320,1990.
Cardinali D. P., Lynch H. J. and Wurtman R. J. Binding of melatonin to human and rat plasma proteins. Endocrinology 91: 1213-1218,1972.
Cardinali D. P. and Ritta M. N. Role of prostaglandins in neuroendocrine junctions: Studies in the pineal gland and the hypothalamus. Neuroendocrinology 36: 152-160,
350
1983.
Cardinali D. P. and Vacas M. I. Feedback control of pineal function by reproductive hormones -a neuroendocrine paradigm. J. Neural Trans. 13 (Suppl. ): 175-201, 1978.
Carlson L. L., Weaver D. R. and Reppert S. M. Melatonin signal transduction in hamster brain: inhibition of adenyl cyclase through a pertussis toxin-sensitive G protein. Endocrinology 125: 2670-2676,1989.
Carman J. S., Post R. M., Buswell R. and Goodwin F. K. Negative effects of melatonin on depression. Am. J. Psychiatry 133: 1181-1186,1976.
Carr D. B., Reppert S. M., Bullen B., Skrinar G., Beitins I., Arnold M., Rosenblatt M., Martin J. B. and McArthur J. W. Plasma melatonin increases during exercise in women. J. Endocrinol. Metab. 53: 224-225,1981.
Carskadon M. A. and Dement W. C. Effects of total sleep loss on sleep tendency. Percept. Motor Skills 48: 495-506,1979.
Carskadon M. A. and Dement W. C. Cumulative effects of sleep restriction on daytime sleepiness. Psychophysiology 18: 107-113,1981.
Cassone V. M. Melatonin and the suprachiasmatic nucleus. In: Suprachiasmatic Nucleus: The Mind's Clock, D. C. Klein, R. Y. Moore and S. M. Reppert (eds. ). Oxford Press, New York, pp 309-323,1991.
Cassone V. M., Chesworth M. J. and Armstrong S. M. Dose-dependent entrainment of rat circadian rhythms by daily injection of melatonin. J. Biol. Rhythms 1: 219-229, 1986.
Cavallo A. Plasma melatonin rhythm in normal puberty: interactions with age and pubertal stages. Neuroendocrinology 55: 372-379,1992.
Cavallo A. Melatonin and human puberty: current perspectives. J. Pineal Res. 15: 115-121,1993.
Champney T. H., Craft C. M., Webb S. M. and Reiter R. J. Hormonal modulation of pineal melatonin synthesis in rats and Syrian hamsters: effects of adrenalectomy and corticosteroid implants. J. Neural Trans. 64: 67-79,1985a.
Champney T. H., Holtorf A. P., Craft C. M. and Reiter R. J. Hormonal modulation of pineal melatonin synthesis in rats and Syrian hamsters: effects of streptozocin- induced diabetes and insulin injections. Comp. Biochem. Physiol. 83: 391-395,1985b.
351
Checkley S. A., Corn T. H., Glass I. B., Thompson C. Franey C. and Arendt J. Neuroendocrine and other studies of the mechanism of anti-depressant action of desipramine. Ciba Foundation Symposium 123, Pitman, London, pp 126-147,1986.
Checkley S. A., Murphy D. G., Abbas M., Marks M., Winton F., Palazidou E., Murphy D. M., Franey C. and Arendt J. Melatonin rhythms in seasonal affective disorder. Br. J. Psychiatry 163: 332-337,1993.
Checkley S. A., Winton F., Corn T., Franey C. and Arendt J. Effects of a phosphodiesterase inhibitor (rolipram) on the urinary excretion of 6- sulphatoxymelatonin in man. J. Psychopharmacol. 1: 20-22,1987.
Chik C. L., Ho A. K. and Klein D. C. Ethanol inhibits dual receptor stimulation of pineal cAMP and cGMP by vasoactive intestinal peptide and phenylephrine. Biochem. Biophys. Res. Commun. 147: 145-151,1987.
Chik C. L., Liu Q. -Y., Girard M., Karpinski E. and Ho A. K. Inhibitory action of ethanol on L-type Cat+-dependent guanosine 3', 5'-monophosphate accumulation in rat pinealocytes. Endocrinology 131: 1895-1902,1992.
Clarke J. D. and Coleman G. J. Persistent meal-associated rhythms in SCN-lesioned rats. Physiol. Behav. 36: 105-113,1986.
Claustrat B., Brun J., David M., Sassolas G. and Chazot G. Melatonin and jet lag: confirmatory result using a simplified protocol. Biol. Psychiatry 32: 705-711,1992.
Claustrat B., Chazot G., Brun J., Jordan D. and Sassolas G. A chronobiological study of melatonin and cortisol secretion in depressed subjects: plasma melatonin, a biochemical marker in major depression. Biol. Psychiatry 19: 1215-1228,1984.
Clodore M., Foret J., Benoit 0., Touitou Y., Aguirre A., Bouard G. and Touitou C. Psychophysiological effects of early morning bright light exposure in young adults. Psychoneuroendocrinology 15: 193-205,1990.
Cole R. J. and Kripke D. F. Amelioration of jet lag by bright light treatment: effects on sleep consolidation. Sleep Res. 18: 411,1989.
Coloma F. M. and Niles L. P. Melatonin enhancement of PH]-GABA and [3H]- muscimol binding in rat brain. Biochem. Phamacol. 37: 1271-1274,1988.
Colquhoun W. P. Phase shift in temperature rhythm after transmeridian flight, as related to pre-flight phase angle. Int. Arch. Occup. Environ. Health 42: 149-157, 1979.
Colquhoun W. P., Blake M. J. F. and Edwards R. S. Experimental studies of shift- work, III. Stabilized 12-hour shift systems. Ergonomics 12: 865-882,1969.
352
Colquhoun W. P. and Folkard S. Personality differences in body temperature rhythm, and their relation to its adjustment to night work. Ergonomics 21: 811-817,1978.
Commentz J. C., Fischer P., Stegner H., Winkler P., Helmke K. and Willig R. P. Pineal calcification does not affect melatonin production. J. Neural Trans. (Suppl. 21): 481,1986.
Comperatore C. A. and Krueger G. P. Circadian rhythm desynchronosis, jet lag, shift lag and coping strategies. Occup. Med. 5: 323-341,1990.
Costa G., Lievore F., Casaletti G., Gaffuir E. and Folkard S. Circadian characteristics influencing interindividual differences in tolerance and adjustment to shift work. Ergonomics 32: 373-385,1989.
Cowen P. J., Bevan J. S., Gosden B. and Elliot S. A. Treatment with ß-adrenoceptor blockers reduces plasma melatonin concentration. Br. J. Clin. Pharmacol. 19: 258- 260,1985.
Cramer H., Rudolph J. and Consbruch U. On the effects of melatonin on sleep and behaviour in man. In: Serotonin - New Vistas: Biochemistry and Behavioural and Clinical Studies. Advances in Biochemical Psychopharmacology, E. Costa, G. L. Gessa, M. Sandler (eds. ), Vol. 11. Raven Press, New York, pp 187-191,1974.
Czeisler C. A., Moore-Ede M. C., Regestein Q. R., Kisch E. S., Fang V. S. and Ehrlich E. N. Episodic 24-hour cortisol secretory patterns in patients awaiting elective cardiac surgery. J. Clin. Endocrinol. Metab. 42: 273-283,1976.
Czeisler C. A., Allan J. S., Strogatz S. H., Ronda J. M., Sanchez R., Rfos C. D., Freitag W. O., Richardson G. S. and Kronauer R. E. Bright light resets the human circadian pacemaker independent of the timing of the sleep/wake cycle. Science 233: 667-671,1986.
Czeisler C. A., Dumont M., Duffy J. F., Steinberg J. D., Richardson G. S., Brown E. N., Sanchez R., Rios C. D. and Ronda J. M. Association of sleep-wake habits in older people with changes in output of circadian pacemaker. Lancer. 340: 933-936, 1992.
Czeisler C. A., Johnson M. P., Duffy J. F., Brown E. N., Ronda J. M. and Kronauer R. E. Exposure to bright light and darkness to treat physiologic maladaptation to night work. New Engl. J. Med. 322: 1253-1259,1990.
Czeisler C. A., Kronauer R. E., Strogatz S. H., Allan J. S., Duffy J. F., Jewett M. E., Brown E. N. and Ronda J. M. Bright light induction of strong (type 0) resetting of the human circadian pacemaker. Science 244: 1328-1333,1989.
353
Czeisler C. A., Moore-Ede M. C. and Coleman R. M. Resetting circadian clocks: applications to sleep disorders medicine and occupational health. In: Sleep/Wake Disorders: Natural History, Epidemiology, and Long-term Evolution, C. Guilleminault and E. Lugaresi (eds. ), Raven Press, New York, pp 243-260,1983.
Czeisler C. A., Moore-Ede M. C. and Coleman R. M. Rotating shift work schedules that disrupt sleep are improved by applying circadian principles. Science 217: 460- 463,1982.
Czeisler C. A., Richardson G. S., Zimmerman J. C., Moore-Ede M. C. and Weitzman E. D. Entrainment of human circadian rhythms by light-dark cycles: a reassessment. Photochem. Photobiol. 34: 239-247,1981.
Daan S. and Lewy A. J. Scheduled exposure to daylight: a potential strategy to reduce "jet lag" following transmeridian flight. Psychopharmacol. Bull. 20: 566-568, 1984.
Dahlitz M., Alvarez B., Vignau J., English J., Arendt J. and Parkes J. D. Delayed sleep phase syndrome response to melatonin. Lancet 337: 1121-1124,1991.
Dawes C. Circadian rhythms in human salivary flow rate and composition. J. Physiol. 220: 529-545,1972.
Dawson D. and Campbell S. S. Timed exposure to bright light improves sleep and alertness during simulated night shifts. Sleep 14: 511-516,1991.
Dawson D. and Encel N. Melatonin and sleep in humans. J. Pineal Res. 15: 1-12, 1993.
Deacon S. and Arendt J. Posture influences concentrations in plasma and saliva in humans. Neurosci. Lett. 167: 191-194,1994.
Deacon S. and Arendt J. Phase-shifts in melatonin 6-sulphatoxymelatonin and alertness rhythms after treatment with moderately bright light at night. Clin. Endocrinol. 40: 413-420,1994.
Deacon S. English J. and Arendt J. Acute phase-shifting effects of melatonin associated with suppression of core body temperature in humans. Neurosci. Lett. 178: 32-34,1994.
Demisch K., Demisch L., Bochnik H. J., Nickelsen T., Althoff P. H., Schoffling K. and Rieth R. Melatonin and cortisol increase after fluvoxamine. Br. J. On. Pharmacol. 22: 620-622,1986.
354
D6sir D., Van Cauter E., Fang V. S., Matrino E., Jadot C., Spire J. C., Noel P., Refetoff S., Copinschi G. and Golstein J. Effects of "jet lag" on hormonal patterns, 1. Procedures, variations in total plasma proteins, and disruption of adrenocorticotropin-cortisol periodicity. J. Clin. Endocrinol. Metab. 52: 628-641, 1981.
Dewasmus G., Nicolas A., Rodriguez D., Salame P., Eschenlauer R., Joly D. and Muzet A. Human core temperature minimum can be modified by ambient thermal transients. Neurosci. Lett. 173: 151-154,1994.
Dijk D. -J., Beersma D. G. M., Daan S. and Lewy A. J. Bright morning light advances the human circadian system without affecting NREM sleep homeostasis. Am. J. Physiol. 256: R106-R111,1989.
Dijk D. -J., Cajochen C. and Borb6ly A. A. Effect of a single 3-hour exposure to bright light on core body temperature and sleep in humans. Neurosci. Lett. 121: 59- 62,1991.
Dijk D. -J. and Czeisler C. A. Paradoxical timing of the circadian rhythm of sleep propensity serves to consolidate sleep and wakefulness in humans. Neurosci. Lett. 166: 63-68,1994.
Dijk D. -J., Duffy J. F. and Czeisler C. A. Circadian and sleep/wake dependent aspects of subjective alertness and cognitive performance. J. Sleep Res. 1: 112-117, 1992.
Dijk D. -J., Visscher C. A., Bloem G. M., Beersma D. G. M. and Daan S. Reduction of human sleep duration after bright light exposure in the morning. Neurosci. Lett. 73: 181-186,1987.
Dimenas E., Kerr D. and Macdonald I. Beta-adrenoceptor blockade and CNS-related symptoms: a randomized, double-blind, placebo-controlled comparison of metaprolol CR/ZOK, atenolol and propanolol LA in healthy subjects. J. Clin. Pharmacol. 30 (Suppl. 2): S82-90,1990.
Dollins A. B., Lynch H. J., Wurtman R. J., Deng M. H., Kischka K. U., Gleason R. E. and Harris H. R. Effect of pharmacological daytime doses of melatonin on human mood and performance. Psychopharmacology 112: 490-496,1993.
Dollins A. B., Zhdanova IN., Wurtman R. J., Lynch H. J. and Deng M. H. Effect of inducing nocturnal serum melatonin concentrations in daytime on sleep, mood, body temperature, and performance. Proc. Natl. Acad. Sci. USA. 91: 1824-1828,1994.
Donaldson E. and Kennaway D. J. Effects of temazepam on sleep, performance, and rhythmic 6-sulphatoxymelatonin and cortisol excretion after transmeridan travel. Aviat. Space Environ. Med. 61: 654-660,1991.
355
Drennan M., Kripke D. F. and Gillin J. C. Bright light can delay human temperature rhythm independent of sleep. Am. J. Physiol. 1: 2,1989.
Drucker-Colin R. Fetal suprachiasmatic nucleus transplants: diurnal rhythm recovery of lesioned rats. Brain Res. 311: 353,1984.
Dubocovich M. L. Luzindole (N-0774): a novel melatonin receptor antagonist. J. Pharm. Exper. Ther. 246: 902-910,1988.
Duncan M. J., Takahashi J. S. and Dubocovich M. L. Characteristics and autoradiographic localization of 2-['u]iodomelatonin binding sites in Djungarian hamster brain. Endocrinology 125: 1011-1018,1989.
Earnest D. J. and Turek F. W. Neurochemical basis for the photic control of circadian rhythms and seasonal reproductive cycles: role for acetylcholine. Proc. Natl. Acad. Sci. USA, 82: 4277-4281,1985.
Eastman C. I. Bright light in work-sleep schedules for shift workers: application of circadian rhythm principles. In: Temporal Disorders in Human Oscillatory Systems, L. Rensing, U. an der Heiden and M. C. Mackey (eds. ), Springer-Verlag, New York, pp 176-185,1987.
Eastman C. I. Squashing versus nudging circadian rhythms with artificial bright light: solutions for shift work? Perspect. Biol. Med. 34: 181-195,1991.
Eastman C. I. High-intensity light for circadian adaptation to a 12-h shift of the sleep schedule. Am. J. Physiol. 263 (Reg. Intgr. Comp. Physiol. 32): R428-R436,1992.
Eastman C. I., Gallo L. C., Lahmeyer H. W. and Fogg L. F. The circadian rhythm of temperature during light treatment for winter depression. Biol. Psychiatry 34: 210- 220,1993.
Eastman C. I. and Miescke K. -J. Entrainment of circadian rhythms with 26-h bright light and sleep/wake schedules. Am. J. Physiol. 259 (Reg. Integr. Comp. Physiol. 28): R1 189-R 1197,1990.
Ebadi M. and Govitrapong P. Orphan transmitters and their receptor sites in the pineal gland. J. Neural Trans. 13 (Suppl. ): 97-114,1976.
Ebisawa T., Karne S., Lerner M. R. and Reppert S. M. Expression of a high-affinity melatonin receptor from Xenopus dermal melanophores. Proc. Natl. Acad. Sci. USA 91: 6133-6137,1994.
Ekman A. -C., Leppäluoto J., Huttunen P., Aranko K. and Vakkuri 0. Ethanol inhibits melatonin secretion in healthy volunteers in a dose-dependent randomized double blind cross-over study. J. Clin. Endocrinol. Metab. 77: 780-783,1993.
356
Elliot A. L., Mills J. N., Minors D. S. and Waterhouse J. M. The effect of real and simulated time-zone shifts upon the circadian rhythms of body temperature, plasma 11-hydroxycorticosteroids, and renal excretion in human subjects. J. Physiol. 221: 227-257,1972.
English J., Hampton S., Dunne M. and Marks V. The effect of timing of a standard meal on plasma cortisol levels in man. Horn. Metab. Res. 14: 106,1982.
English J., Middleton B. A., Arendt J. and Wirz-Justice A. Rapid, direct measurement of melatonin in saliva using an iodinated tracer and solid phase second antibody. Ann. Clin. Biochem. 30: 415-416,1993.
Enna S. J. and Möhler H. Gamma-aminobutyric acid (GABA) receptors and their association with benzodiazepine recognition sites. In: Psychopharmacology: The Third Generation of Progress, H. Y. Meltzer (ed. ). Raven Press, New York, pp 265- 272,1987.
Epstein R., Tzischinsky P. and Lavie P. Sleep-wake cycle in rotating shift workers: effects of changing from phase advance to phase delay rotation. 20th International conference on chronobiology, Tel Aviv, Israel, 1991.
Evans J. I. The effect on sleep of travel across time. Clin. Trials J. 7: 64-75,1970.
Evans J. I., Christie G. A., Lewis S. A., Daly J. and Moore-Robinson M. Sleep and time zone changes. Arch. Neurol. 26: 36-48,1972.
Feinberg I. Changes in sleep cycle patterns with age. J. Psychiat. Res. 10: 283-306, 1974.
Fellenberg A. J., Phillipou G. and Seamark R. F. Urinary 6-sulphatoxymelatonin excretion during the human menstrual cycle. Clin. Endocrinol. 17: 71-75,1982.
Fernandes G., Carandente F., Halberg E., Halberg F. and Good R. A. Circadian rhythm in activity of lympholytic natural killer cells from spleens of Fischer rats. J. Immun. 123: 622-625,1979.
Fernandes G., Halberg H., Yunis E. J. and Good R. A. Circadian rhythmic plaque- forming cell response of spleens from mice immunized with SRBC. J. Immun. 117: 962-966,1976.
Ferrier I. N., Arendt J., Johnstone E. C. and Crow T. J. Reduced nocturnal melatonin secretion in chronic schizophrenia: relationship to body weight. On. Endocr nol. 17: 181-187,1982.
357
Fevre-Montange M., Van Cauter E., Refetoff S., D6sir D., Tourniaire J. and Copinschi G. Effects of "jet lag" on hormonal patterns, II. Adaptation of melatonin circadian periodicity. J. Clin. Endocrinol. Metab. 52: 642-649,1981.
Folkard S. The pragmatic approach to masking. Chronobiol. Int. 6: 55-64,1989.
Folkard S. and Akerstedt T. Towards the prediction of alertness on abnormal sleep/wake schedules. In: Vigilance and Performance in Automated Systems, A. Coblentz (ed. ), Kluwer Academic Publishers, Dordrecht, pp 287-296,1989.
Folkard S., Arendt J., Aldhous M. and Kennett H. Melatonin stabilises sleep onset time in a blind man without entrainment of cortisol or temperature rhythms. Neurosci. Lett. 113: 193-198,1990.
Folkard S., Arendt J. and Clark M. Can melatonin improve shift workers tolerance of the night shift? Some preliminary findings. Chronobiol. Inter. 10: 315-320,1993.
Folkard S., Minors D. S. and Waterhouse J. M. Is there more than one circadian clock in humans? Evidence from fractional desynchronization studies. J. Physiol. (London) 257: 341-356,1984.
Folkard S., Minors D. S. and Waterhouse J. M. "Demasking" the temperature rhythm after simulated time zone transitions. J. Biol. Rhythms 6: 81-91,1991.
Folkard S. and Monk T. H. Individual differences in the circadian response to a weekly rotating shift system. In: Night and Shift Work: Biological and Social Aspects, A. Reinberg, N. Vieux and P. Andlauer (eds. ), Pergamon Press, Oxford, pp 367-374,1981.
Folkard S., Monk T. H. and Lobban M. C. Short and long-term adjustment of circadian rhythms in `permanent' night nurses. Ergonomics 21: 785-799,1978.
Folkard S., Wever R. A. and Wildgruber C. M. Multi-oscillator control of circadian rhythms in human performance. Nature. 305: 223-225,1983.
Franey C., Aldhous M., Burton S., Checkley S. A. and Arendt J. Acute desipramine stimulates melatonin and 6-sulphatoxymelatonin production in man. Br. J. Clin. Pharmacol. 22: 73-79,1986.
Fraser S., Cowen P., Franklin U., Franey C. and Arendt J. A direct radioimmunassay for melatonin. Clin. Chem. 29: 396-399,1983.
French J., Hannon P. and Brainard G. C. Effects of bright illuminance on body temperature and human performance. Ann. Rev. Chronopharm. 7: 37-40,1990.
358
Gaddy J. R., Rollag M. D. and Brainard G. C. Pupil size regulation of threshold of light-induced melatonin suppression. J. Clin. Endocrinol. Metab. 77: 1398-1401, 1993.
Gallo L. C. and Eastman C. I. Circadian rhythms during gradually delaying and advancing sleep and light schedules. Physiol. Behav. 53: 119-126,1993.
Gander P. H., Myhre G., Graeber C., Anderson H. T. and Lauber J. K. Adjustment of sleep and the circadian temperature rhythm after flights across nine time zones. Aviat. Space Environ. Med. 60: 733-743,1989.
Goodwin F. K. National Institute of Health Advisory Report on Light Therapy Devices. Soc. Light Treatment Biol. Rhythms. 5: 4-7,1992.
Graeber R. C. Alterations in performance following rapid transmeridian flight. In: Rhythmic Aspects of Behaviour, F. M. Brown and R. C. Graeber (eds. ), Hillsdale, NJ: Lawrence Erlbaum Assoc. pp 173-212,1982.
Graeber R. C., Dement W. C., Nicholson A. N., Sasaki M. and Wegmann H. -M. International cooperative study of air crew layover sleep operational summary. Aviat. Space Environ. Med. 57 (Suppl. 12): B10-13,1986.
Green D. J. and Gillette R. Circadian rhythm of firing rate recorded from single cells in the rat suprachiasmatic nucleus slice. Brain Res. 245: 198-200,1982.
Greenberg L. H. and Weiss B. ß-adrenergic receptors in aged rat brain: Reduced number and capacity of the pineal gland to develop supersensitivity. Science 201: 61- 63,1978.
Griffiths P. A., Folkard S., Bojkowski C., English J. and Arendt J. Persistent 24-h variations of urinary 6-hydroxymelatonin sulphate and cortisol in Antarctica. Experientia. 15: 430-432,1986.
Groos G. and Hendriks J. Circadian rhythms in electrical discharge of rat suprachiasmatic neurons recorded in vitro. Neurosci. Lett. 34: 283-288,1982.
Guardiola-Lemaitre B., Lenbgre A. and Porsolt R. D. Combined effects of diazepam and melatonin in two tests for anxiolytic activity in the mouse. Pharmacol. Biochem. Behav. 41: 405-408,1992.
Hagan R. D., Diaz F. J. and Horvath S. M. Plasma volume changes with movement to supine and standing positions. J. Appl. Physiol. 45: 414-418,1978.
Haimov I., Laudon M., Zisapel N., Souroujon M., Nof D., Shlitner A., Herer P., Tzischinsky 0. and Lavie P. Sleep disorders and melatonin rhythms in elderly people. Br. Med. J. 309: 167,1994.
359
Hannon P., Brainard G., Gibson W., French J., Arnall D., Brugh L., Crank L., Fleming S., Hanifin J. and Rollag M. Bright light suppresses melatonin and improves cognitive performance during nighttime hours in humans. Sleep Res. 21: 377,1992.
Hariharasubramanian N. and Nair N. P. V. Circadian rhythm of plasma melatonin and cortisol during the menstrual cycle. In: The Pineal Gland: Endocrine aspects, Advances in Biosciences, G. M. Brown and S. D. Wainright (eds. ), Pergamon Press, Oxford, 53, pp 31-35,1985.
Hastings M. H., Mead S. M., Vindlacheruvu R. R., Ebling F. J. P., Maywood E. S. and Groose J. Non-photic phase shifting of the circadian activity rhythm of Syrian hamsters: relative potency of arousal and melatonin. Brain Res. 591: 20-26,1992.
Hauty G. T. and Adams T. Phase shifting of the human circadian system. In: Circadian Clocks, Aschoff J. (ed. ), North Holland, Amsterdam, pp 413-425,1965.
Hellon R. F. Central thermoreceptors and thermoregulation. In: Handbook of Sensory Physiology, E. Neil (ed. ). Springer, Heidelberg, Vol. 3, part 1,1972.
Henden T., Stokkan K. A., Reiter R. J., Nonaka K. O., Lerchl A. and Jones D. J. The age-associated reduction in pineal ß-adrenergic receptor density is prevented by life- long food restriction in rats. Biol. Signals. 1: 70-74,1991.
Hensel H. Neural processes in thermoregulation. Physiol. Rev. 53: 948-1017,1973.
Hildebrandt G. and Stratmann I. Circadian system response to night work in relation to the individual circadian phase position. Arch. Occup. Environ. Health 43: 73-83, 1979.
Hirata F., Hayaishi 0., Tokuyama T. and Senoh S. In vitro and in vivo formation of two new metabolites of melatonin. J. Biol. Chem. 249: 1311-1313,1974.
Honma K., Honma S. and Wada T. Entrainment of human circadian rhythms by artificial bright light cycles. Experientia 43: 572-574,1987.
Honma K. and Honma S. A human phase response curve for bright light pulses. Jap. J. Psychiatry Neurol. 42: 167-168,1988.
Honma S., Kanematsu N., Katsuno Y. and Honma K. Light suppression of nocturnal pineal and plasma melatonin in rats depends on wavelength and time of day. Neurosci. Lett. 147: 201-204,1992.
Horne J. A. and Östberg 0. A self-assessment questionnaire to determine morningness-eveningness in human circadian rhythms. Inter. J. Chronobiol. 4: 97- 110,1976.
360
Howell H. E., Yous S., Lesieur D., Guardiola B. and Morgan P. High affinity melatonin agonists based on napthalene nucleus. J. Endocrinol. (Suppl. ) 131: 38, 1991.
Huether G. B., Poeggeler B., Reimer A. and George A. Effect of tryptophan administration on circulating melatonin levels in chicks and rats: evidence for stimulation of melatonin synthesis and release in the gastrointestinal tract. Life Sci. 51: 945-953,1992.
Husdan H., Rapoport A. and Locke S. Influence of posture on the serum concentration of calcium. Metabolism 22: 787-797,1973.
Iguchi, H., Kato, K-I. and Ibayashi H. Age-dependent reduction in serum melatonin concentrations in healthy human subjects. J. Clin. Endocrinol. Metab. 55: 27-29, 1982.
Illernovä H. Entrainment of mammalian circadian rhythms in melatonin production by light. Pineal Res. Rev. 6: 173-217,1988.
Illernovä H., Backstrom M., Saapf. J., Wetterberg L. and Vancbo B. Melatonin in rat pineal gland and serum, rapid parallel decline after light exposure at night. Neurosci. Lett. 9: 189-193,1978.
Illnerovä H., Buresovä M., Nadvfdkovä J., Dvoräkovä M. and Zvolsky P. Maintenance of circadian phase adjustment of the human melatonin rhythm following artificial long days. Brain Res. 626: 322-326,1993.
Illnerovä H., Trentini G. P. and Maslova L. Melatonin accelerates re-entrainment of the circadian rhythm of its own production after an 8-h advance of the light-dark cycle. J. Comp. Physiol. 166: 97-102,1989.
Illernovä H. and Vanacek J. Pineal rhythm in N-acetyltransferase activity in rats under different artificial photoperiods and in natural daylight in the course of a year. Neuroendocrinology 31: 321-326,1980.
Illernovä H. and Vanacek J. Two-oscillator structure of the pacemaker controlling the circadian rhythm of N-acetyltransferase in the rat pineal gland. J. Comp. Physiol. 145: 539-548,1982.
lllernovä H. and Vanacek J. Circadian rhythm in inducibility of rat pineal N- acetyltransferase after light pulses at night. Control by a morning oscillator. J. Comp. Physiol. 154: 739-744,1984.
Illernovä H., Zvolsky P. and Vanacek J. The circadian rhythm in plasma melatonin concentration of the urbanised man: The effect of summer and winter time. Brain Res. 328: 186-189,1985
361
Inouye S. T. and Kawamura H. Persistence of circadian rhythmicity in mammalian hypothalamic ̀island' containing the suprachiasmatic nucleus. Proc. Natl. Acad. Sci. (USA) 76: 5962-5966,1979.
James S. P., Mendelson W. B., Sack D. A., Rosenthal N. E. and Wehr T. A. The effect of melatonin on normal sleep. Neuropsychopharmacology 1: 41-44,1987.
James S. P., Sack D. A., Rosenthal N. E. and Mendelson W. B. Melatonin administration in insomnia. Neuropsychopharmacology 3: 19-23,1990.
James S. P., Wehr T. A., Sack D. A., Parry B. L. and Rosenthal N. E. Treatment of seasonal affective disorder with light in the evening. Br. J. Psychiatry 147: 424-428, 1985.
Jan J. E., Espezel H. and Appleton R. E. The treatment of sleep disorders with melatonin. Develop. Med. Child Neurol. 36: 97-107,1994.
Jacobsen F. M., Wehr T. A., Skwerer R. A., Sack D. A. and Rosenthal N. E. Morning versus midday phototherapy of seasonal affective disorder. Am. J. Psychiatry 144: 1301-1305,1987.
Jewett M. E., Kronauer R. E. and Czeisler C. A. Light-induced suppression of endogenous circadian amplitude in humans. Nature 350: 59-62,1991.
Johnson R. F., Smale L., Moore R. Y. and Morin L. P. Lateral geniculate lesions block circadian phase-shift responses to a benzodiazepine. Proc. Natl. Acad. Sci. 85: 5301-5304,1988.
Jones R. L., McGeer P. L. and Greiner A. C. Metabolism of exogenous melatonin in schizophrenic and non-schizophrenic volunteers. Clin. Chim. Acta. 26: 281-285, 1969.
Joss J. M. P. Autoradiographic localisation of sites of uptake of 1H]-melatonin in the brains of the lizard and the lamprey. In: Pineal Function, C. D. Matthews and R. F. Seamark (eds. ), Elsevier/North Holland Press, Amsterdam, pp. 223-234,1981.
Kato M., Honma K., Shigemitsu T. and Shiga Y. Circularly polarized 50-Hz magnetic field exposure reduces pineal gland and blood melatonin concentrations of Long-Evans rats. Neurosci. Lett. 166: 59-62,1994.
Kawate T., Hinuma S. and Kumagi K. Studies on the bioperiodicity of the immune response: II. Co-variations of murine T and B cells and the role of corticosteroid. J. Immun. 126: 1364-1367,1981.
Keefe D. L., Earnest D. J., Nelson D., Takahashi J. S. and Turek F. W. A cholinergic antagonist mecamylamine blocks the phase-shifting effects of light on the circadian rhythm of locomotor activity in the golden hamster. Brain Res. 403: 308-312,1987.
362
Kennaway D. J., Blake P. and Webb H. A. A melatonin agonist and N-acetyl-N2- formyl-5-methoxykynurenamine accelerates the reentrainment of the melatonin rhythm following a phase advance of the light-dark cycle. Brain Res. 495: 349-354,1989.
Kennaway D. J. and Royles P. Circadian rhythms of 6-sulphatoxymelatonin, cortisol and electrolyte excretion at the summer and winter solstices in normal men and women. Acta Endocrinol. 113: 450-456,1986.
Kennaway D. J. and Van Dorp C. F. Free-running rhythms of melatonin, cortisol, electrolytes, and sleep in humans in Antarctica. Am. J. Physiol. 260 (Reg. Integr. Comp. Physiol. 29) :R 1137-1144,1991.
Kitay J. I. Pineal lesions and precocious puberty: a review. J. Clin. Endocrinol. Metab. 14: 622-625,1954.
Klein D. C. Circadian rhythms in the pineal gland. In: Endocrine Rhythms, Kreiger D. T. (ed. ). Raven Press, New York, pp 203-223,1979.
Klein D. C., Berg G. R. and Weller J. L. Melatonin synthesis: Adenosine 3', 5'-
monophosphate and norepinephrine stimulate N-acetyltransferase. Science 168: 979- 980,1970.
Klein D. C., Smoot R., Weller J. L., Higa S., Markey S. P., Creed G. J. and Jacobowitz D. M. Lesions of the paraventricular nucleus area of the hypothalamus disrupt the suprachiasmatic spinal cord circuit in the melatonin rhythm generating system. Brain Res. Bulletin. 10: 647-652,1983.
Klein K. E., Bruner H., Gunther E., Jovy D., Mertens J., Rimpler A. and Wegmann H. -M. Psychological and physiological changes caused by desynchronization following transzonal air travel. In: Aspects of Human Efficiency: Diurnal Rhythm and Loss of Sleep, W. P. Colquhoun (ed. ), English Universities Press, London, pp 295- 305,1972a.
Klein K. E., Bruner H., Holtmann H., Rehme H., Stolze J., Steinhoff W. D. and Wegmann H. -M. Circadian rhythm of pilots' efficiency and effects of multiple time zone travel. Aerospace Med. 41: 125-132,1970.
Klein K. E. and Wegmann H. -M. Significance of circadian rhythms in aerospace operations. AGARDograph No. 247, Advisory Group for Aerospace Research and Development, NATO, pp 1-60, Technical Editing and Reproduction Ltd., London, 1980.
Klein K. E., Wegmann H. -M. and Hunt B. I. Desynchronization of body temperature and performance circadian rhythm as a result of outgoing and homegoing transmeridian flights. Aerospace Med. 43: 119-132,1972b.
363
Klein D. C. and Weller J. L. Indole metabolism in the pineal gland: A circadian rhythm in N-acetyltransferase. Science 169: 1093-1095,1970.
Klein T., Martens H., Dihk D. -J., Kronauer R. E., Seely E. W. and Czeisler C. A. Circadian sleep regulation in the absence of light perception: chronic non-24-hour circadian rhythm sleep disorder in a blind man with a regular 24-hour sleep-wake schedule. Sleep 16: 333-343,1993.
Klemfuss H. and Clopton P. L. Seeking tau: a comparison of six methods. J. Interdiscipl. Cycle Res. 24: 1-16,1993.
Kogi K. Introduction to the problem of shift work. In: Hours of Work -Temporal Factors in Work Scheduling, S. Folkard and T. H. Monk (eds. ), John Wiley and Sons, New York, pp 165-184,1985.
Kopin I. J., Pare C. M. B., Axelrod J. and Weissbach H. The fate of melatonin in animals. J. Biol. Chem. 236: 3072-3075,1961.
Koster-van Hoffen G. C., Mirmiran M., Bos N. P. A., Witting W., Delagrange P. and Guardiola-Lemaltre B. Effects of a novel melatonin analogue on circadian rhythms of body temperature and activity in young, middle-aged, and old rats. Neurobiol. Aging 14: 565-569,1993.
Kräuchi K. and Wirz-Justice A. Circadian rhythm of heat production, heart rate, and skin and core body temperature under unmasking conditions in men. Am. J. Physiol. 267 (Reg. Integr. Comp. Physiol. 36) R819-R829,1994.
Krierger D. T., Hauser H. and Krey L. C. Suprachiasmatic nuclear lesions do not abolish food-shifted circadian adrenal and temperature rhythmicity. Science 197: 398-399,1977.
Kripke D. F. Phase-advance theories for affective illnesses. In: Circadian Rhythms in Psychiatry, T. A. Wehr and F. K. Goodwin (eds. ), Boxwood Press, California, pp 41-69,1983.
Laitinen J. T., Castren E., Vakkuri 0. and Saavedra J. M. Diurnal rhythm of melatonin binding in the rat suprachiasmatic nucleus. Endocrinology 124: 1585-1587, 1989.
Laitinen J. T. and Saavedra J. M. Characterization of melatonin receptors in the rat area postrema: modulation of affinity with cations and guanine nucleotides. Neuroendocrinol. 126: 2110-2115,1990.
Landen R. O., Vikström A. O. and Öberg B. Social and psychological effects related to the order of shifts. In: Stress Research Reports, no. 126, Karolinska Institute, Stockholm, Sweden, pp 30-32,1981.
364
Lehman M. N., Silver R., Gladstone W. R., Kahn R. M., Gibson M. and Bittman E. L. Circadian rhythmicity restored by neural transplant. Immunocytochemical characterization of the graft and its integration with the host. J. Neurosci. 7: 1626- 1638,1987.
Leiberman H. R., Waldhauser F., Garfield G., Lynch H. J. and Wurtman R. J. Effects of melatonin on human mood and performance. Brain Res. 323: 201-207,1984.
Lerchl A., Nonaka K. O. and Reiter R. J. Pineal gland "magnetosensitivity" to static magnetic fields is a consequence of induced electric currents (eddy currents). J. Pineal Res. 10: 109-116,1991.
Lewy A. J. Biochemistry and regulation of mammalian production. In: The Pineal Gland, R. M. Relkin (ed. ). Elsevier/North Holland, New York, pp 77-128,1983.
Lewy A. J., Ahmed S., Latham Jackson J. M. and Sack R. L. Melatonin shifts human circadian rhythms according to a phase-response curve. Chronobiol. Int. 9: 380-392, 1992.
Lewy A. J. and Newsome D. A. Different types of melatonin circadian secretory rhythms in some blind subjects. J. Clin. Endocrinol. Metab. 56: 1103-1107,1983.
Lewy A. J., Nurnberger J. I., Wehr T. A., Pack D., Becker L. E., Powell R. L. and Newsome D. A. Supersensitivity to light: possible trait marker for manic depressive illness. Am. J. Psychiat. 142: 725-727,1985.
Lewy A. J. & Sack R. L. The dim light melatonin onset as a marker for circadian phase position. Chronobiol. Int. 6: 93-102,1989.
Lewy A. J., Sack R. L., Muller S and Hoban T. M. Antidepressant and circadian phase-shifting effects of light. Science 235: 352-354,1987.
Lewy A. J., Sack R. L. and Singer C. M. Treating phase-typed chronobiologic sleep and mood disorders using appropriately timed bright artificial light. Psychopharmacol. Bull. 21: 368-437,1985.
Lewy A. J., Wehr T. A., Goodwin F. K., Newsome D. A. and Markey S. P. Light suppresses melatonin secretion in humans. Science 210: 1267-1269,1980.
Lieberman H. R., Waldhauser F., Garfiled G., Lynch H. J. and Wurtman R. J. Effects of melatonin on human mood and performance. Brain Res. 323: 201-207, 1984.
Lincoln G. A., Almeida O. F. X. and Arendt J. Role of melatonin and circadian rhythms in seasonal reproduction in rams. J. Reprod. Fertil. 30(Suppl. ): 23-31,1981.
365
Lino A., Silvy S., Condorelli L. and Rusconi A. C. Melatonin and jet lag: Treatment schedule. Biol. Psychiatry 34: 587-588,1993.
Lissoni P., Resentini M., Mauri R., De Medici C., Morabito F., Esposti D., Di Bella L., Rossi D. and Parravicini L. Effect of acute injection of melatonin on the basal secretion of hypophyseal hormones in prepubertal and pubertal healthy subjects. Acta. Endocrinol. 111: 305-311,1986.
Lynch H. J., Rivest R. M., Ronsheim P. M. and Wurtman R. J. Light intensity and the control of melatonin secretion in rats. Neuroendocrinology 33: 181-185,1981.
MacFarlane J. G., Cleghorn J. M., Brown G. M. and Strainer D. L. The effects of exogenous melatonin on the total sleep time and daytime alertness of chronic insomniacs: a preliminary study. Biol. Psychiatry 30: 371-376,1991.
Maestroni G. J. M. The immunoneuroendocrine role of melatonin. J. Pineal Res. 14: 1-10,1993.
Mallo C., Zaidan R., Faure A., Brun J., Chazot G. and Claustrat B. Effects of a four-day nocturnal melatonin treatment on the 24h plasma melatonin, cortisol and prolactin profiles in humans. Acta Endocrinol. 119: 474-480,1988.
Markey S. P. and Buell P. E. Pinealectomy abolishes 6-hydroxymelatonin excretion by male rats. Endocrinology 111: 425-426,1982.
Markey S. P., Higa S., Shih M., Danforth D. N. and Tamarkin L. The correlation between human plasma melatonin levels and urinary 6-hydroxymelatonin excretion. Clin. Chim. Acta 150: 221-225,1985.
Mason R., Delagrange P. and Guardiola B. Melatonin and S-20098 mediated resetting of rat suprachiasmatic circadian clock neurones in vitro. Br. J. Pharmacol. 108 (Suppl. ): 32P, 1993.
Matthews C. D., Guerin M. V. and Wang X. Human plasma melatonin and urinary 6-sulphatoxymelatonin: studies in natural annual photoperiod and extended darkness. Clin. Endocrinol. 35: 21-27,1991.
McAinsh J. and Cruickshank J. M. Beta-blockers and central nervous system side effects. Pharmacol. Ther. 46: 163-197,1990.
McArthur A. J., Gillette M. U. and Prosser R. A. Melatonin directly resets the rat suprachiasmatic circadian clock in vitro. Brain Res. 565: 158-161,1991.
McIntyre I. M., Norman T. R., Burrows G. D. and Armstrong S. M. Melatonin rhythm in human plasma and saliva. J. Pineal Res. 4: 177-183,1987.
366
McIntyre I. M., Norman T. R., Burrows G. D. and Armstrong S. M. Human melatonin suppression by light is intensity dependent. J. Pineal Res. 6: 149-156,1989a.
McIntyre I. M., Norman T. R., Burrows G. D. and Armstrong S. M. Human response to light at different times of the night. Psychoneuroendocrinology 14: 187-193, 1989b.
Medanic M. and Gillette M. U. NPY phase-shifts the circadian rhythm of neuronal activity in the rat SCN. Proc. Soc. Res. Biol. Rhythms, P104,1992.
Meijer J. H., van der Zee E. A. and Dietz M. Glutamate phase-shifts circadian activity rhythms in hamsters. Neurosci. Lett. 86: 177-183,1988.
Menendez-Pelaez A., Howes K. A., Gonzalez-Brito A. and Reiter R. J. N- acetyltransferase activity, hydroxyindole-O-methyltransferase activity, and melatonin levels in the Harderian glands of the female Syrian hamster. Changes during light/dark cycle and the effects of 6-chlorophenylalanine administration. Biochem. Biophys. Res. Commun. 145: 1231-1238,1987.
Menendez-Pelaez A. and Reiter R. J. Distribution of melatonin in mammalian tissues: The relative importance of nuclear versus cytoslic localization. J. Pineal Res. 15: 59- 69,1993.
Miles A. Melatonin: perspectives in the life sciences. Life Sciences, 44: 375-385, 1989.
Miles A., Philbrick D. R. S., Shaw D. M., Tidmarsh S. F. and Pugh A. J. Salivary melatonin estimation in clinical research. Clin. Chem. 31: 2041-2042,1985.
Miles A., Philbrick D. R. S., Thomas D. R. and Grey J. E. Diagnostic and clinical implications of plasma and salivary melatonin assay. On. Chem. 33: 1295-1297, 1987.
Miles L. E. M., Raynal D. M. and Wilson M. A. Blind man living in normal society has circadian rhythms of 24.9 hours. Science 198: 421-423,1977.
Mills J. N., Minors D. S. and Waterhouse J. M. The physiological rhythms of subjects living on a day of abnormal length. J. Physiol. (London) 268: 303-326,1977.
Mills J. N., Minors D. S. and Waterhouse J. M. Adaptation to abrupt time shifts of the oscillator(s) controlling human circadian rhythms. J. Physiol. (London) 285: 455- 470,1978a.
Mills J. N., Minors D. S. and Waterhouse J. M. The effects of sleep upon human circadian rhythms. Chronobiologia 5: 4-27,1978b.
367
Minors D. S. and Waterhouse J. M. Endogenous and exogenous components of circadian rhythms when living on a 21h day. Int. J. Chronobiol. 8: 31-48,1982.
Minors D. S. and Waterhouse J. M. The problem of masking and some ways to deal with it. In: Chronobiology and Chronomedicine, G. Hildebrandt, R. Moog and F. Raschke (eds. ), Peter Lang, Frankfurt, pp 119-135,1987.
Minors D. S. and Waterhouse J. M. Mathematical and statistical analysis of circadian rhythms. Psychoneuroendocrinology 13: 443-464,1988.
Minors D. S. and Waterhouse J. M. Masking in humans and some attempts to solve it. Chronobiol. Int. 6: 29-53,1989.
Minors D. S. and Waterhouse J. M. Investigating the endogenous component of human circadian rhythms: a review of some simple alternatives to constant routines. Chronobiol. Int. 9: 55-78,1992.
Minors D. S. and Waterhouse J. M. Separating the endogenous and exogenous components of the circadian rhythm of body temperature during night work using some "purification" models. Ergonomics 36: 497-507,1993.
Minors D. S., Waterhouse J. M., Hume K., Marks M., Arendt J., Folkard S. and Akerstedt T. Sleep and circadian rhythms of temperature and urinary excretion on a 22.8hr "day". Chronobiol. Int. 5: 65-80,1988.
Minors D. S., Waterhouse J. M. and Wirz-Justice A. A human phase-response curve to light. Neurosci. Lett. 133: 36-40,1991.
Mitler M. M., Carskadon M. A., Czeisler C. A., Dement W. C., Dinges D. F. and Graeber R. C. Catastrophes, sleep, and public policy: consensus report. Sleep 11: 100-109,1988.
Moldofsky H. S., Musisi S. and Phillipson E. A. Treatment of a case of advanced sleep phase syndrome by phase delay chronotherapy. Sleep 9: 61-65,1986.
Moline M. L., Pollak C. P.. and Hirsch E. Effects of bright light on sleep following an acute phase advance. Sleep Res. 18: 432,1989.
Moline M. L., Pollak C. P., Monk T. H., Lester L. S., Wagner D. R., Zendell S. M., Graeber C., Salter C. A. and Hirsch E. Age-related differences in recovery from simulated jet lag. Sleep 15: 28-40,1992.
Monk T. H. Advantages and disadvantages of rapidly rotating shift schedules -a circadian viewpoint. Hum. Factors 28: 553-557,1986.
368
Monteleone P., Maj M., Fusco M., Orazzo C. and Kemali D. Physical exercise at night blunts the nocturnal increase of plasma melatonin levels in healthy humans. Life Sci. 47: 1989-1995,1990.
Monteleone P., Maj M., Fuschino A. and Kemali D. Physical stress in the middle of the dark phase does not affect light-depressed plasma melatonin levels in humans. Neuroendocrinology. 55: 367-371,1992.
Moog R. and Hildebrandt G. Comparisons of different causes of masking effects. In: Night and shiftwork: Long term effects and their prevention, M. Haider, M. Koller and R. Cervinka (eds. ), Lang, Frankfurt, pp 131-140,1986.
Moog R. and Hildebrandt G. Adaptation to shift work - experimental approaches with reduced masking effects. Chronobiol. Int. 6: 65-75,1989.
Moore R. Y. Retinohypothalamic projections in mammals: A comparative study. Brain Res. 49: 403-409,1973.
Moore R. Y. and Klein D. C. Visual pathways and the central neural control of a circadian rhythm in pineal serotonin N-acetyltransferase activity. Brain Res. 71: 17- 34,1974.
Moore R. Y. and Speh J. C. GABA is the principal neurotransmitter of the circadian system. Neurosci. Lett. 150: 12-116,1993.
Moore-Ede M. C., Czeisler C. A. and Richardson G. S. Circadian timekeeping in health and disease. Part I. Basic properties of circadian pacemakers. New Engl. J. Med. 308: 469-476,1983.
Morgan P. J., Lawson W., Davidson G. and Howell H. E. Guanine nucleotides regulate the affinity of melatonin receptor on ovine pars tuberalis. Neuroendocrinology 50: 359-362,1989a.
Morgan P. J., Lawson W., Davidson G. and Howell H. E. Melatonin inhibits cyclic AMP production in cultured ovine pars tuberalis cells. J. Mol. Endocrinol. 3: R5-8, 1989b.
Morgan P. J. and Williams L. M. Central melatonin receptors: implications for a mode of action. Experentia. 45: 955-965,1989.
Morris M., Lack L. and Barrett J. The effect of sleep/wake state on nocturnal melatonin excretion. J. Pineal Res. 9: 133-138,1990.
Mrosovsky N. and Salmon P. A. A behavioural method for accelerating re- entrainment of rhythms to new light-dark cycles. Nature 330: 372-373,1987.
369
Murphy P., Myers B., Badia P. and Harsh J. The effects of bright light on daytime sleep latencies. Sleep Res. 20: 465,1991.
Murphy P., Myers B., Wright K., Boecker M. and Badia P. Nonsteroidal anti- inflammatory drugs and melatonin levels in humans. Sleep Res. 22: 372,1993.
Myers B. L. and Badia P. Immediate effect of different light intensities on body temperature and alertness. Physiol. Behav. 54: 199-202,1993.
Nair N. P. V., Hariharasubramanian N., Pilapil C., Isaac I. and Thavundayil J. X. Plasma melatonin - an index of brain aging in humans? Biol. Psychiatry 21: 141-150, 1986.
Nakagawa H., Sack R. L. and Lewy A. J. Sleep propensity free-runs with the temperature, melatonin and cortisol rhythms in a totally blind person. Sleep 15: 330- 336,1992.
Nakayama K., Yoshimuta N., Sasaki Y., Kadokura M., Hiyama T., Takeda A., Sasaki M. and Ushijima S. Diurnal rhythm in body temperature in different phases of the menstrual cycle. Jap. J. Psychiat. Neurol. 46: 235-237,1992.
Namboodiri M. A., Sugden D., Klein D. C. and Mefford I. N. 5-hydroxytryptophan elevates serum melatonin. Science 221: 659-661,1983.
Nelson W., Tong Y. L., Lee, J. -K. and Halberg F. Methods for cosinor- rhythmometry. Chronobiologia 6: 305-323,1979.
Neuwelt E. A. and Lewy A. J. Disappearance of plasma melatonin after removal of a neoplastic pineal gland. New Engl. J. Med. 19: 1132-1135,1983.
Nicholson A. N., Roth T. and Stone B. M. Hypnotics and air crew. Aviat. Space Environ. Med., 56: 299,1985.
Nickelsen T., Demisch L., Demisch K., Radermacher B. and Schoffling K. Influence of subchronic intake of melatonin at various times of the day on fatigue and hormonal levels: a placebo-controlled, double-blind trial. J. Pineal Res. 6: 325-334, 1989.
Nickelson T., Samel A., Maaß H., Vejvoda M., Wegmann H. and Schöffling K. Circadian patterns of salivary melatonin and urinary 6-OH-melatoninsulphate before and after a9 hour time-shift. In: Advances in Experimental Medicine and Biology: Kyrnurenine and serotonin pathways, R. Schwarz, S. N. Young and R. R. Brown (eds. ), Plenum Press, New York, pp 493-496,1991a.
370
Nickelson T., Lang A. and Bergau L. The effect of 6-, 9- and 11-hour time shifts on circadian rhythms: adaptation of sleep parameters and hormonal patterns following the intake of melatonin or placebo. In: Advances in Pineal Research, Arendt J., P6vet P. (eds. ), vol. 5, John Libbey, London, pp 303-306,1991b.
Nowak R., McMillen I. C., Redman J. and Short R. V. The correlation between serum and salivary melatonin concentrations and urinary 6-hydroxymelatonin sulphate excretion rates: two non-invasive techniques for monitoring circadian rhythmicity. Clin. Endocrinol. 27: 445-452,1987.
Olcese J. M. The neurobiology of magnetic field detection in rodents. Pro. Neurobiol. 35: 325-330,1990.
Olcese J. M., Reuss S. and Vollrath L. Evidence for the involvement of the visual system in mediating magnetic field effects on the pineal function in the rat. Brain Res. 333: 382-384,1985.
Olcese J. M., Reuss S. and Vollrath L. Geomagnetic field detection in rodents. Life Sci. 42: 605-613,1988.
Orth D. N. and Island D. P. Light synchronization of the circadian secretory rhythm in plasma cortisol (17-OHCS) concentration in man. J. Clin. Endocrinol. 29: 479- 486,1969.
Owen J. and Arendt J. Melatonin suppression in human subjects by bright light and dim light in Antarctica: time and season-dependent effects. Neurosci. Lett. 137: 181- 184,1992.
Palm L., Blennow G. and Wetterberg L. Correction on non-24-hours sleep/wake cycle by melatonin in a blind retarded boy. Ann. Neurol. 29: 336: 339,1991.
Pang S. F. Melatonin concentrations in blood and pineal gland. Pineal Res. Rev. 3: 115-159,1985.
Pang S. F., Brown G. M., Grota L. J., Chambers J. W. and Rodman R. L. Determination of N-acetylserotonin and melatonin activities in the pineal gland, retina, Harderian gland, brain and serum of rats and chickens. Neuroendocrinology 23: 1-13, 1977.
Parfitt A. G. and Klein D. C. Sympathetic nerve endings in the pineal gland protect against stress-induced increase in N-acetyltransferase activity. Endocrinology 99: 840- 851,1976.
Parfitt A. G., Weller J. L. and Klein D. C. Beta adrenergic blockers decrease adrenergically stimulated N-acetyltransferase activity in pineal glands in organ culture. Neuropharmacology 15: 353-358,1976.
371
Petrie K., Conaglen J. V., Thompson L. and Chamberlain K. Effect of melatonin on jet lag after long haul flights. Br. Med. J. 298: 705-707,1989.
Petrie K., Dawson A. G., Thompson L. and Brook R. A double-blind trial of melatonin as a treatment for jet lag in international aircrew. Biol. Psychiatry 33: 526- 530,1993.
Petterborg L. J., Kjellman B. F., Thalen B. E. and Wetterberg L. Effect of a 15 minute light pulse on nocturnal serum melatonin levels in human volunteers. J. Pineal Res. 10: 9-13,1991a.
Petterborg L. J., Thaldn B. E., Kjellman B. F. and Wetterberg L. Effect of melatonin replacement on serum hormone rhythms in a patient lacking endogenous melatonin. Brain Res. Bull. 27: 181-185,1991b.
Pevet P. The different classes of proteic and peptidic substances present in the pineal gland. In: The Pineal Gland and its Endocrine Role, Axelrod J., Fraschini F. and Velo F. P. (eds. ), Plenum Press, New York, pp 113-149,1983a.
Pevet P. The 5-methoxyindoles different from melatonin: their effects on the sexual axis. In: The Pineal Gland and its Endocrine Role, Axelrod J., Fraschini F. and Velo F. P. (eds. ), Plenum Press, New York, pp 331-348,1983b.
Pierpaoli W. and Regelson W. Pineal control of aging: Effect of melatonin and pineal grafting on aging mice. Proc. Natl. Acad. Sci. 91: 787-791,1994.
Pickard G. E., Ralph M. R. and Menaker M. T. The intergeniculate leaflet partially mediates effects of light on circadian rhythms. J. Biol. Rhythms. 2: 35-56,1987.
Pittendrigh C. S. and Daan S. A functional analysis of circadian pacemakers in nocturnal rodents. V. Pacemaker structure: a clock for all seasons. J. Comp. Physiol. 106: 333-355,1976.
Poeggeler B., Reiter R. J., Tan D. -X., Chen L. -D. and Manchester L. C. Melatonin, hydroxyl radical mediated oxidative damage and aging: a hypothesis. J. Pineal Res. 14: 151-169,1993.
Prosser R. A., Miller J. D. and Heller C. H. A serotonin agonist phase-shifts the circadian clock in the suprachiasmatic nuclei in vitro. Brain Res. 534: 336-339,1990.
Riad-Fahmy D., Read G. F., Walker R. F. and Griffiths K. Steroids in saliva for assessing endocrine function. Endocr. Rev. 3: 367-395,1982.
Raikhlin N. T., Kvetnoy I. M. and Tolkachev V. N. Melatonin may be synthesized in enterochromaffin cells. Nature 255: 344,1975.
372
Ralph M. R., Foster R. G., Davis F. C. and Menaker M. Transplanted suprachiasmatic nucleus determines circadian period. Science 247: 975-978,1990.
Redman J. R. and Armstrong S. M. Reentrainment of rat circadian activity rhythms: effects of melatonin. J. Pineal Res. 5: 203-215,1988.
Redman J. R., Armstrong S. M. and Ng K. T. Free-running activity rhythms in the rat: entrainment by melatonin. Science 219: 1089-1091,1983.
Redman J. R. and Guardiola-Lemaitre B. The melatonin agonist, S 20098: effects on the rat circadian system. In: Melatonin and the Pineal Gland - From Basic Science to Clinical Application, Y. Touitou, J. Arendt and P. Pdvet (eds. ), Excerta Medica, Amsterdam, pp 127-130,1993.
Reebs S. G. and Mrosovsky N. Effects of induced wheel running on the circadian activity rhythms of Syrian hamsters: entrainment and phase response curves. J. Biol. Rhythms 4: 39-48,1989.
Refinetti R. Laboratory instrumentation and computing: comparison of six methods for the determination of the period of circadian rhythms. Physiol. Behav. 54: 869- 875,1993.
Reghunandanan V., Reghunandanan R. and Singh P. I. Neurotransmitters of the suprachiasmatic nucleus: role in the regulation of circadian rhythms. Prog. Neurobiol. 41: 647-655,1993.
Reinberg A., Vieux N., Ghata J., Chaument A. -J. and Laporte A. Is the rhythm amplitude related to the ability to phase shift circadian rhythms of shift workers? J. Physiol. (Paris) 74: 405-409,1978.
Reinberg A., Andlauer P., De Prins J., Malbecq W., Vieux N. and Bourdeleau P. Desynchronization of the oral temperature circadian rhythm and intolerance to shift work. Nature 308: 272-274,1984.
Reiter R. J. The melatonin message: Duration versus coincidence hypotheses. Life Sci. 46: 2119-2131,1987.
Reiter R. J. The pineal and its hormones in the control of reproduction in mammals. Endocrinol. Rev. 1: 109-131,1980.
Reiter R. J. Some perturbations that disturb the circadian melatonin rhythm. Chronobiol. Int. 9: 314-321,1992a.
Reiter R. J. The ageing pineal gland and its physiological consequences. BioEssays, 14: 169-175,1992b.
373
Reiter R. J., Poeggeler B., Tan D. -X., Chen L. -D., Manchester L. C. and Guerrero J. M. Antioxidant capacity: a novel action not requiring a receptor. Neuroendocrinol. Lett. 15: 103-116,1993.
Reiter R. J., Steinlechner B., Richardson B. A. and King T. S. Differential response to pineal melatonin levels to light at night in laboratory-raised and wild-captured 13- lined ground squirrels (Spermophilus tridecemlineatus). Life Sci. 32: 2625-2629, 1983.
Reppert S. M., Weaver D. R., Rivkees S. A. and Stopa E. G. Putative melatonin receptors in a human biological clock. Science 242: 78-81,1988.
Riad-Fahmy D., Read G. F., Walker R. F. and Griffiths K. Steroids in saliva for assessing endocrine function. Endocr. Rev. 3: 367-395,1982.
Riggs L. A. Light as a stimulus for vision. In: Vision and Visual Perception, C. H. Graham, N. R. Bartlett, J. L. Brown., Y. Hsia, C. G. Mueller and Riggs L. A. (eds. ), Wiley and Sons, New York, pp 1-38,1965.
Rivkees S. A., Carlson L. L. and Reppert S. M. Guanine nucleotide-binding protein regulation of melatonin receptors in lizard brain. Proc. Natl. Acad. Scl. USA. 86: 3882-3886,1989.
Röjdmark S., Wilkner J., Adner N., Andersson D. E. H. and Wetterberg L. Inhibition of melatonin secretion by ethanol in man. Metabolism 42: 1047-1051, 1993.
Rollag M. D., Morgan R. J. and Niswender G. D. Route of melatonin secretion in
sheep. Endocrinology. 102: 1-8,1978.
Ronkainen H., Vakkuri O. and Kauppila A. Effects of physical exercise on the serum concentration of melatonin in female runners. Acta Obstet. Gynecol. Scand. 65: 827-829,1986.
Rosenstein R. E. and Cardinali D. P. Melatonin increases in vivo GABA accumulation in rat hypothalamus, cerebellum, cerebral cortex and pineal gland. Brain Res. 398: 403-406,1986.
Rosenthal N. E., Sack D. A., Gillin J. C., Lewy A. J., Goodwin F. K., Davenport Y., Mueller P. S., Newsome D. A. and Wehr T. A. Seasonal affective disorder. A discription of the syndrome and preliminary findings with light therapy. Arch. Gen. Psychiatry 41: 72-79,1984.
Rosenthal N. E., Jacobsen F. M., Sack D. A., Arendt J., James S. P., Parry B. L. and Wehr T. A. Atenolol in seasonal affective disorder: a test of the melatonin hypothesis. Am. J. Psychiatry 145: 52-56,1988.
374
Rosenthal N. E., Levendosky A. A., Skwerer R. G., Vanderpool J. R., Kelly K. A., Hardin T., Kasper S., DellaBella P. and Wehr T. A. Effects of light treatment on core body temperature in seasonal affective disorder. Biol. Psychiatry 27: 39-50, 1990.
Rudeen P. K., Philo R. C. and Symmes S. K. Antiepileptic effects of melatonin in the pineal-ectomized mongolian gerbil. Epilepsia 21: 149-154,1980.
Rusak B. The role of the suprachiasmatic nuclei in the generation of circadian rhythms in the golden hamster, Mesocricetus auratus. J. Comp. Physiol. Psychol. 118: 145-164,1977.
Rusak B. and Zucker I. Neural regulation of circadian rhythms. Physiol. Rev. 59: 449-526,1979.
Rutenfranz J., Knauth P. and Angersbach D. Sleep disturbances and shift work. In: The Twenty-Four Hour Workday: Proceedings of a Symposium on Variations in Work- Sleep Schedules, L. C. Johnson, D. I. Tepas, Colquhoun W. P. and M. J. Colligan (eds. ), Cincinnati, pp 221-259,1981.
Sack R. L., Blood M. L. and Lewy A. J. Melatonin rhythms in night shift workers. Sleep 15: 434-441,1992.
Sack R. L., Lewy A. J., Blood M. L., Stevenson J. and Keith D. Melatonin administration to blind people: phase advances and entrainment. J. Biol. Rhythms 6: 249-261,1991.
Sack R. L. Lewy A. J. and Hoban T. M. Free-running melatonin rhythms in blind people: phase shifts with melatonin and triazolam administration. In: Temporal Disorder in Human Oscillatory Systems, L. Rensing, U. an den Heiden and Mackey M. C. (eds. ), Springer, Heidelberg, 1987.
Sack R. L., Lewy A. J., Erb D. L., Vollmer W. M. and Singer C. M. Human melatonin production decreases with age. J. Pineal Res. 3: 379-388,1986.
Samel A., Wegmann H. -M. and Vejvoda M. Resynchronization of circadian rhythms after different types of 9-h advance shifts. Abstract 18. Fourth Meeting, Soc. Res. Biol. Rhythms, Florida, USA, May 1994.
Samel A., Gundel A. and Wegmann H. -M. Bright light exposure in the morning and in the afternoon after 6h advance shift. Third Meeting, Abstract 3, Soc. Res. Biol. Rhythms, Florida, 1992.
Samel A., Wegmann H. -M. and Vejvoda M. Response of the circadian system to 6° head-down tilt bed rest. Aviat. Space Environ. Med. 64: 50-54,1993.
375
Samel A., Wegmann H. -M., Vejvoda M., Maaß H., Gundel A. and Schütz M. Influence of melatonin treatment on human circadian rhythmicity before and after a simulated 9-hr time shift. J. Biol. Rhythms. 6: 235-248,1991.
Sasaki M., Kurosaki Y., Onda M., Yamaguchi 0., Nishimura H., Kashimura K. and Graeber R. C. Effects of bright light on circadian rhythmicity and sleep after transmeridian flight. Sleep Res. 18: 442,1989.
Satinoff E. Neural organization and evolution of thermal regulation in mammals. Science 201: 16-22,1978.
Satinoff E., Li H., Tcheng T. K., Liu C., McArthur A. J., Medanic M. and Gillette M. U. Do the suprachiasmatic nuclei oscillate in old rats as they do in young ones? Am. J. Physiol 265 (Reg. Integr. Comp. Physiol. 34): R1216-R1222,1993.
Sawaki Y., Nihonmatsu I. and Kawamura H. Transplantation of the neonatal suprachiasmatic nuclei into rats with complete bilateral suprachiasmatic lesions. Neurosci. Res. 1: 67-72,1984.
Schaad N. C. and Klein D. C. Characterization of r2-adrenergic receptors of rat pinealocytes. Endocrinology 130: 2804-2810,1992.
Seigel P. V., Gerathewohl S. J. and Mohler S. R. Time zone effects. Science 164: 1249-1255,1969.
Seltzer A., Viswanathan M. and Saavedra J. M. Melatonin-binding sites in brain and caudal arteries of the female rat during the estrous cycle and after estrogen administration. Endocrinology 130: 1896-1902,1992.
Shanahan T. L. and Czeisler C. A. Light exposure induces equivalent phase shifts of the endogenous circadian rhythms of circulating plasma melatonin and core body temperature in men. J. Clin. Endocrinol. Metab. 73: 227-235,1991.
Shanahan T. L., Neri D. and Czeisler C. A. The melatonin rhythm observed during three months of rotating shift work. In: Work Hours, Sleep and Sleepiness: Stress Research Reports, T. Akerstedt and G. Kecklund (eds. ), p119, Sept. 1994.
Simonneaux V., Ebadi M. and Bylund D. B. Identification and characterization of 012D-adrenergic receptors in bovine pineal gland. Mol. Pharmacol. 40: 235-241,1991.
Skene D. J., Aldhous M. E. and Arendt J. Melatonin, jet-lag and the shift-wake cycle. In: Sleep: Transient Insomnia, Shift-Work and Jet Lag, J. Home (ed. ), Gustav Fischer Verlag, Stuttgart, New York, pp 39-41,1988.
Skene D. J., Bojkowski C. J. and Arendt J. Comparison of the effects of acute fluvoxamine and desipramine administration on melatonin and cortisol production in humans. Br. J. Clin. Pharmac. 37: 181-186,1994.
376
Smale L., Michels K. M., Moore R. Y. and Morin L. P. Destruction of the hamster serotonergic system by 5,7-DHT: effects on circadian rhythm phase, entrainment and response to triazolam. Brian Res. 515: 9-19,1990.
Spencer M. B., Nicholson A. N., Pascoe P. A. and Rogers A. S. Adaptation of the temperature rhythm to a 10 hour eastward time zone change. Abstract 17, Fourth Meeting, Soc. Res. Biol. Rhythms, Florida, May 1994.
Stanberry L. S. Das Gupta T. K. and Beattie C. W. Photoperiodic control of melanoma growth in hamsters: influence of pinealectomy and melatonin. Endocrinology 113: 469-475,1983.
Stankov B., Biella G., Panara C., Capsoni S., Lucini V., Fautek J., Cozzi B. and Fraschini F. Melatonin signal transduction and mechanism of action in the central nervous system: using the rabbit cortex as a model. Endocrinology 130: 2152-2159, 1992.
Stankov B. and Fraschini F. High affinity melatonin binding sites in the vertebrate brain. Neuroendocrinol. Lett. 15: 149-164,1993.
Stehle J. H., Foulkes N. S., Molina C. A., Simonneaux V., P6vet P. and Sassone-Corsi P. Adrenergic signals direct rhythmic expression of transcriptional repressor CREM in the pineal gland. Nature 365: 314-320,1993.
Stephan F. K., Berkley K. J. and Moss R. L. Efferent projections of the rat suprachiasmatic nucleus. J. Neurosci. 6: 2625-2641,1981.
Stephan F. K., Swann J. M. and Sisk C. L. Anticipation of 24-hr feeding schedules in rats with lesions of the suprachiasmatic nucleus. Behav. Biol. 25: 346-363,1979.
Stoker D. J. and Wynn V. Effect of posture on the plasma cholesterol levels. Br. Med. J. 1: 336-338,1966.
Stokkan K. A. Reiter R. J., Nonaka K. O., Lerchl A., Yu B. P. and Vaughan M. K. Food restriction retards aging of the pineal gland. Brain Res. 545: 66-72,1991.
Strassman R. J., Peake G. T., Qualls C. R. and Lisansky E. J. Model for the study of the acute effects of melatonin in man. J. Clin. Endocrinol. Metab. 65: 847-852, 1987.
Strassman R. J., Qualls C. R., Lisansky E. J. and Peake G. T. Elevated rectal temperature produced by all-night bright light is reversed by melatonin infusion in man. J. App!. Physiol. 76: 2178-2182,1991.
Sugden D. Pharmacological effects of melatonin in mouse and rat. J. Pharmacol. Exp. 7her. 227: 587-591,1983.
377
Sugden D., Namboodiri M. A. A., Klein D. C., Pierce J. E., Grady R. K. Jr. and Mefford I. Ovine pineal «, -adrenoreceptors: Characterisation and evidence for a functional role in the regulation of serum melatonin. Endocrinology 116: 1960-1967, 1985a.
Sugden D., Vanacek J., Klein D. C., Thomas T. P. and Anderson W. Activation of protein kinase C potentiates isoprenaline-induced cyclic AMP accumulation in rat pinealocytes. Nature 314: 359-361,1985b.
Sugden L., Sugden D. and Klein D. C. Essential role of calcium influx in the adrenergic regulation of cAMP and cGMP in rat pinealocytes. J. Biol. Chem. 261: 11608-11612,1987.
Swanson L. W. In: Handbook of Chemical Neuroanatomy, A. Björklund, T. Hökfelt and L. W. Swanson (eds. ), Elsevier, Amsterdam, pp 1-124,1987.
Takahashi J. S. Circadian clocks 4 la CREM. Nature 365: 299-230,1993.
Tamarkin L., Baird C. J. and Almeida O. F. X. Melatonin: a coordinating signal for mammalian reproduction? Science 227: 714-720,1985.
Tamarkin L., Danforth D., Licher A., De Moss E., Cohen M., Chabner B. and Lippman M. Decreased nocturnal plasma melatonin peak in patients with estrogen receptor positive breast cancer. Science 216: 1003-1005,1982.
Tan D. -X., Chen L. -D., Poeggeler B., Manchester L. C. and Reiter R. J. Melatonin: a potent endogenous hydroxyl radical scavenger. J. Endocrinol. 1: 57-60,1993.
Tan M. H., Wilmshurst E. G., Gleason R. E. and Soeldner J. S. Effect of posture on serum lipids. New Engl. J. Med. 289: 416-419,1973.
Terman M., Reine C. E., Rafferty B., Gallin P. F. and Terman J. S. Bright light therapy for winter depression: potential ocular effects and theoretical implications. Photochem. Photobiol. 51: 781-792,1990.
Terman M. and Terman J. S. A circadian pacemaker for visual sensitivity? Ann. NY. Acad. Sc!. 453: 147-161,1985.
Terman M., Terman J. S., Quitkin F. M., Cooper T. B., Lo E. S., Gorman J. M. Stewart J. W. and Mcgrath P. J. Response of the melatonin cycle to phototherapy for seasonal affective disorder. J. Neural Trans. 72: 147-165,1988.
Terman M. and Terman J. S. Light therapy for winter depression: Report to the Depression Guidelines Panel, P. H. S. Agency for Health Care Policy and Research. New York, NY, New York State Psychiatric Institute, 1990.
378
Terzolo M., Revelli A., Guidetti D., Piovesan A., Cassoni P., Paccotti P., Angeli A. and Massobrio M. Evening administration of melatonin enhances the pulsatile secretion of prolactin but not of LH and TSH in normally cycling women. Clin. Endocrinol. 39: 185-191,1993.
Tetsuo M., Poth M. and Markey S. P. Melatonin metabolite excretion during childhood and puberty. J. Clin. Endocrinol. Metab. 55: 311-313,1982.
Thompson C., Franey C., Arendt J. and Checkley S. A. A comparison of melatonin secretion in depressed patients and normal subjects. Br. J. Psychiatry 152: 260-265, 1988.
Thorington L. Spectral, irradiance, and temporal aspects of natural and artificial light. Ann. MY Acad. Sci. 453: 28-54,1985.
Tobler I., Jaggi K. and Borbely A. A. Effects of melatonin and the melatonin receptor agonist S-20098 on the vigilance states, EEG spectra, and cortical temperature in the rat. J. Pineal Res. 16: 26-32,1994.
Tominaga K., Shibata S., Ueki S. and Watanabe S. Effects of a 5-HTIA receptor agonists on the circadian rhythm of wheel running activity in hamsters. Eur. J. Pha»nacol. 214: 79-84,1992.
Totterdell P. and Folkard S. In situ repeated measures of affect and performance facilitated by use of a handheld computer. Behav. Res. Instr. Comp. 24: 545-553, 1992.
Touitou Y., Fevre M., Bogdan A., Reinberg A., De Prins J., Beck H. and Touitou C. Patterns of plasma melatonin with aging and mental condition: stability of nycthemeral rhythms and differences in seasonal variations. Acta Endocrinol. 106: 145-151,1984.
Touitou Y., Fevre M., Lagoguey M., Carayan A, Bogdan A., Reinberg A., Beck H., Cesselin F. and Touitou C. Age and mental health related circadian rhythms of plasma levels of melatonin, prolactin, luteinizing hormone and follicle stimulating hormone. J. Endocrinol. 91: 467-175,1981.
Touitou Y., Motohashi Y, Reinberg A., Touitou C., Bourdeleau P., Bogdan A. and Auzkby A. Effect of shiftwork on the night-time secretory patterns of melatonin, prolactin, cortisol and testosterone. Eur. J. Appl. Physiol. 60: 288-292,1990.
Troiani M. E., Reiter R. J., Vaughan M. K., Oaknin S. and Vaughan G. M. Swimming depresses nighttime melatonin content without changing N-acetyltransferase activity in the rat pineal gland. Neuroendocrinology 47: 55-60,1987.
379
Troiani M. E., Reiter R. J., Tannenbaum M. G., Puig-Domingo M., Guerrero J. M. and Menendez-Pelaez A. Neither the pituitary gland nor the sympathetic nervous system is responsible for eliciting the large drop in elevated rat pineal melatonin levels due to swimming. J. Neural Trans. 74: 149-160,1988.
Tsai T. -H., Okumura M., Yamasaki M. and Sasaki T. Simulation of jet lag following a trip with stopovers by intermittent schedule shifts. J. Interdiscipl. Cycle Res. 19: 89-96,1988.
Turek F. W. and Losee-Olson S. A benzodiazepine used in the treatment of insomnia phase-shifts the mammalian circadian clock. Nature 321: 167-168,1986.
Tzischinsky 0., Dagan Y and Lavie P. The effects of melatonin on the timing of sleep in patients with delayed sleep phase syndrome. In: Melatonin and the Pineal Gland - From Basic Science to Clinical Application, Y. Touitou, J. Arendt and P. Pevet (eds. ), Excerta Medica, Amsterdam, pp 351-354,1993.
Tzischinsky 0., Lavie P. and Pal I. Time-dependent effects of 5mg melatonin on the sleep propensity function. J. Sleep Res. I: P8,1992a.
Tzischinsky 0., Pal I., Epstein R., Dagan Y. and Lavie P. The importance of timing in melatonin administration in a blind man. J. Pineal Res. 12: 105-108,1992b.
Vakkuri 0. Diurnal rhythm of melatonin in human saliva. Acta Physiol. Scand. 124: 409-412,1985.
Vakkuri 0., Leppaluoto J. and Kauppila A. Oral administration and distribution of melatonin in human serum, saliva and urine. Life Sci. 37: 489-495,1985.
Vakkuri 0., Tervo J., Luttinen R., Ruotsalainen H., Rahkamara E. and Leppaluoto J. A cycle isomer of 2-hydroxymelatonin: a novel metabolite of melatonin. Endocrinology 120: 2453-2459,1987.
Van Cauter E., Sturis J., Byrne M. M., Blackman J. D., Leproult R., Ofek G., L'Hermite-Baleriaux M., Refetoff S., Turek F. W. and Van Reeth O. Demonstration of rapid light-induced advances and delays of the human circadian clock using hormonal phase markers. Am. J. Physiol. 266: E953-963,1994.
Vanecek J., Sugden D., Weller J. L. and Klein D. C. Atypical synergistic a, - and 8- adrenergic regulation of 3', 5'-monophosphate and guanosine 3', 5'-monophosphate in rat pinealocytes. Endocrinology 116: 2167-2173,1985.
Vanecek J. and Illnerovä H. Night pineal N-acetyltransferase activity in rats exposed to white or red light pulses of various intensity and duration. Ezperientia 38: 1318- 1320,1982.
380
Van Reeth 0., Sawyers D., Leproult R., Olivares E. and Lapeyre G. S-20098, a melatonin agonist, has hypothermic effects after single morning oral administration in humans. Abstract 146, Fourth Meeting, Soc. Res. Biol. Rhythms, Florida, May 1994.
Van Reeth 0., Zhang Y., Reddy A., Zee P. and Turek F. W. Aging alters the entraining effects of an activity-inducing stimulus on the circadian clock. Brain Res. 607: 286-292,1993.
Van Reeth 0., Sturis J., Byrne M. M., Blackman J. D., L'Hermite-Baleriaux M., Leproult R., Oliner C., Refetoff S., Turek F. W. and Van Cauter E. Nocturnal exercise phase delays circadian rhythms of melatonin and thyrotropin secretion in normal men. Am. J. Physiol. 266: E964-974,1994.
Van Reeth 0. and Turek F. W. Stimulated activity mediates phase shifts in the hamster circadian clock induced by dark pulses or benzodiazepines. Nature 339: 49- 51,1989.
Vaughan G. M., Macdonald S. D., Jordan R. M., Allen J. P., Bell R. and Stevens E. A. Melatonin, pituitary function and stress in humans. Psychoneuroendocrinology 4: 351- 362,1979
Vining R. F. and McGinley R. A. Flux of steroids between blood and saliva. In: Frontiers of Oral Physiology, Y. Kawamura and D. B. Ferguson (eds. ), Karger, Basel, pp2l-32,1984.
Viswanathan M. J., Laitinen J. and Saavedra J. Expression of melatonin receptors in arteries involved in thermoregulation. Proc. Natl. Acad. Sci. 87: 6200-6203,1990.
Vollrath L., Semm P. and Gammel G. Sleep induction by intranasal application of melatonin. Biosciences 29: 327-329,1981.
Voordouw B. C. G., Euser R., Verdonk R. E. R., Alberda B. T., De Jong F. H., Drogendjik A. C., Fauser B. C. J. M. and Cohen M. Melatonin and melatonin combinations alter pituitary-ovarian function in women and can inhibit ovulation. J. Clin. Endocrinol. Metab. 74: 108-117,1992.
Waldhauser F., Saletu B. and Trinchard-Lugan I. Sleep laboratory investigations on hypnotic properties of melatonin. Psychopharmacology 100: 222-226,1990.
Waldhauser F., Steger H. and Vorkapic P. Melatonin secretion in man and the influence of exogenous melatonin on some physiological and behavioural variables. Adv. Pineal Res. 2: 207-223,1987.
Waldhauser F., Vierhapper H. and Pirich K. Abnormal circadian melatonin secretion in night-shift workers. New Engl. J. Med. Dec 18: 1614,1986
381
Waldhauser F., Weiszenbacher G., Tatzer E., Gisinger B., Waldhauser M., Schemper M. and Frisch H. Alterations in nocturnal serum melatonin levels in humans with growth and aging. J. Clin. Endocrinol. Metab. 66: 648-652,1988.
Walsh J. K., Schweitzer P. K., Anch A. M., Muehlbach M. J., Jenkins N. A. and Dickins Q. S. Sleepiness/alertness on a simulated night shift following sleep at home with triazolam. Sleep 14: 140-146,1991.
Wass J. A. L., Jones A. E., Rees L. H. and Besser G. M. hCGß producing pineal choriocarcinoma. Clin. Endocrinol. 17: 423-431,1982.
Watts A. G. The efferent projections of the suprachiasmatic nucleus: anatomical insights into the control of circadian rhythms. In: Suprachiasmatic Nucleus: The Mind's Clock, D. C. Klein, Moore R. Y. and Reppert S. M. (eds. ), Oxford, New York, pp 77-106,1991.
Watts A. G., Swanson L. W. and Sanchez-Watts G. Efferent projections of the suprachiasmatic nuclues I: studies using retrograde transport of Phaseolus vulgaris leucoagglutinatinin in the rat. J. Comp. Neurol. 258: 204-229,1987.
Weaver D. R., Rivkees S. A. and Reppert S. M. Localization and characterization of melatonin receptors in rodent brain by in vitro autoradiography. J. Neurosci. 9: 2581- 2590,1989.
Weaver D. R., Stehle J. H., Stopa E. G. and Reppert S. M. Melatonin receptors in human hypothalamus and pituitary: implications for circadian and reproductive responses to melatonin. J. Clin. Endocrinol. Metab. 76: 295-301,1993.
Webley G. E. and Leidenberger F. The circadian pattern of melatonin and its positive relationship to progesterone in women. J. Clin. Endocrinol Metab. 63: 323-328, 1986.
Wegmann H. -M., Bruner H., Jovy D., Klein K. E., Marbargar J. P. and Rimpler A. Effects of transmeridian flights on the diurnal excretion patterns of 17- hydroxycorticosteroids. Aerospace Med. 41: 1003-1005,1970.
Wegmann H. -M., Klein K. E., Conrad B. and Esser P. A model for prediction of resynchronization after time-zone flights. Aviat. Space Environ. Med. 54: 524-527, 1983.
Wegmann H. -M., Gundel A., Naumann M., Samel A., Schwartz E. and Vejvoda M. Sleep, sleepiness, and circadian rhythmicity in aircrews operating on transatlantic routes. Aviat. Space Environ. Med. 57(12, suppl. ): B53-64,1986.
Wehr T. The durations of human melatonin secretion and sleep respond to changes in daylength (photoperiod). J. Clin. Endocrinol Metab. 73: 1276-1280,1991)
382
Wehr T. A., Jacobsen F. M., Sack D. A., Arendt J., Tamarkin L. and Rosenthal N. E. Phototherapy of seasonal affective disorder: time of day and suppression of melatonin are not critical for its antidepressant effects. Arch. Gen. Psychiatry 43: 870-875, 1986.
Wehr T. A., Moul D. E., Barbato G., Giesen H. A., Seidel J. A., Barker C. and Bender C. Conservation of photoperiod-responsive mechanisms in humans. Am. J. Physiol. 265 (Reg. Integr. Comp. Physiol. 34): R846-857,1993.
Weiland N. G. and Wise P. M. Aging progressively decreases the densities and alters the diurnal rhythms of alpha-1 adrenergic receptors in selected hypothalamic regions. J. Endocrinol. 126: 2392-2397,1990.
Weissbach H., Redfield B. G. and Axelrod J. The enzymatic acetylation of serotonin and other naturally occurring amines. Biochim. Psych. Acta. 54: 190-192,1961.
Weitzman E. D., Czeisler C. A., Coleman R. M., Spielman A. J., Zimmerman J. C. and Dement W. Delayed sleep phase syndrome: a chronobiological disorder with sleep- onset insomnia. Arch. Gen. Psychiatry 38: 737-746,1981.
Welsh M. G. Pineal calcification: structural and functional aspects. Pineal Res. Rev. 3: 41-68,1985.
Wetterberg 1., Aperia B., Beck-Friis J., Kjellman B. F., Ljunggren J. -G., Nilsonne A., Petterson U., Tham A. and Umden F. Melatonin and cortisol levels in psychiatric illness. Lancet ii, 100,1982.
Wetterberg 1., Arendt J., Pannier 1., Sizonenko P. C., van Donselaar W. and Heyden T. Human serum melatonin changes during the menstrual cycle. J. On. Endocrinol Metab. 42: 185-188,1976.
Wetterberg L. Melatonin in humans: physiological and clinical studies. J. Neural Trans. (suppl. ) 13: 389-310,1978.
Wever R. A. The circadian system of man. Results of experiments under temporal isolation. New York: Springer-Verlag, 1979.
Wever R. A. Phase shifts of human circadian rhythms due to shifts of artificial zeitgebers. Chronobiologia 7: 303-327,1980.
Wever R. A. Fractional desynchronization of human circadian rhythms. A method of evaluating entrainment limits and functional independencies. Pflagers Arch. 395: 128-137,1983.
Wever R. A. Use of light to treat jet lag: differential effects of normal and bright artificial light on human circadian rhythms. Ann. N. Y. Acad. Sci. 453: 282-304, 1985a.
383
Wever R. A. Internal interactions within the human circadian system: the masking effect. Experientia 41: 332-342,1985b.
Wever R. A. Characteristics of circadian rhythms in human functions. J. Neural Trans. (Suppl. ) 21: 323-373,1986.
Wever R. A. Light effects on human circadian rhythms: a review of recent Andechs experiments. J. Biol. Rhythms 4: 161-185,1989.
Wever R. A., Polasek J. and Wildgruber C. M. Bright light effects human circadian rhythms. Pflügers Arch. 396: 85-87,1983.
Whillock M., Clark I. E., McKinlay A. F., Todd C. D. and Mundy S. J. Ultraviolet levels associated with the use of fluorescent general lighting UV-B and UV-B lamps in the workplace and home. National Radiation Protection Board Technical Report 221, H. M. S. O., London, 1988.
Wickland C. R. and Turek F. W. Phase-shifting effects of acute increases in activity on circadian locomotor rhythms in hamsters. Am. J. Physiol. 261: R1109-R1117, 1991.
Wilkinson M., Arendt J., Bradtke J. and de Ziegler D. Determination of a dark- induced increase of pineal N-acetyltransferase activity and simultaneous radioimmunoassay of melatonin in pineal, serum and pituitary tissue of the male rat. J. Endocrinol. 72: 243-244,1977.
Williams L. M., Morgan P. J., Hastings H., Lawson W., Davidson G. and Howell H. E. Melatonin receptor sites in the Syrian hamster brain and pituitary. Localization and characterization using ['1I] iodomelatonin. J. Neuroendocrinol. 1: 315-320,1989.
Winget C. M., DeRoshia C. W., Markley C. L. and Holley D. C. A review of human physiological and performance changes associated with desynchronosis of biological rhythms. Aviat. Space Environ. Med. 55: 1085-1096,1984.
Wirz-Justice A. and Anderson J. L. Morning light exposure for the treatment of winter depression: the one true light therapy? Psychopharmacol. Bull. 26: 511-520, 1990.
Wirz-Justice A., Graw P., Kräuchi K., Gisin B., Arendt J., Aldhous M. and Pöldinger W. Morning or night-time melatonin is ineffective in Seasonal Affective Disorder. J. Psychiat. Res. 24: 129-137,1990.
Wirz-Justice A., Graw P., Kräuchi K., Gisin B., Jochum A., Arendt J., Fisch H. -U. Buddeberg C. and Pöldinger W. Light therapy in seasonal affective disorder is independent of the time of day or circadian phase. Arch. Gen. Psychiatry 50: 929- 937,1993.
384
Wirz-Justice A. and Richter R. Seasonality in biochemical determinations: A source of variance and a clue to the temporal incidence of affective illness. Psychiatr Res. 1: 53-60,1979.
Wolfe D., Potter L., Richardson K. and Axelrod J. Localizing tritiated norepinephrine in sympathetic axons by electron microscopic autoradiography. Science 138: 440-442,1962.
Wright J., Aldhous M., Franey C., English J. and Arendt J. The effects of exogenous melatonin on endocrine function in man. Clin. Endocrinol. 24: 375-382, 1986.
Wu W., Reiter R. J., Troiani M. E. and Vaughan G. M. Elevated daytime rat pineal and serum melatonin levels induced by isoprotenol are depressed by swimming. Life Sci. 41: 1473-1479,1987
Wurtman R. J., Axelrod J. and Barchas J. D. Age and enzyme activity in the human pineal. J. Clin. Endocrinol. Metab. 24: 299-301,1964.
Wurtman R. J., Axelrod J. and Phillips L. S. Melatonin synthesis in the pineal gland: control by light. Science 142: 1071-1073,1963.
Yaga K., Reiter R. J. and Richardson B. A. Tryptophan loading increases daytime serum melatonin levels in intact and pinealectomized rats. Life Sci. 52: 1231-1238, 1993.
Yanovski J., Witcher J., Adler N., Markey S. and Klein D. C. Stimulation of the paraventricular nucleus of the hypothalamus elevates urinary 6-hydroxymelatonin during daytime. Brain Res. Bull. 19: 129-133,1987.
Young I. M., Francis P. L., Leone A. M. Stovell P. and Silman R. E. Constant pineal output and increasing body mass account for declining melatonin levels during human growth and sexual maturation. J. Pineal Res. 5: 71-86,1988.
Young I. M., Leone R. M., Francis P., Stovell P. and Silman R. E. Melatonin is metabolised to N-acetylserotonin and 6-hydroxymelatonin in man. J. Clin. Endocrinol Metab. 60: 114-119,1985.
Yous S., Andrieux J., Howell H. E., Morgan P. J., Renard P., Pfeiffer B., Lesieur D. and Guardiola-Lemaitre B. Novel napthalenic ligands with high affinity for the melatonin receptor. J. Med. Chem. 35: 1484-1486,1992.
Zaidan R., Geoffriau M., Brun J., Taillard J., Bureau C., Chazot G. and Claustrat B. Melatonin is able to influence its secretion in humans: description of a phase- response curve. Neuroendocrinology 60: 105-112,1994.
385
Zee P., Rosenberg R. S. and Turek F. W. Effects of aging on entrainment and rate of resynchronization of circadian locomotor activity. Am. J. Physiol. 263 (Reg. Integr. Comp. Physiol. 32): R1099-R1 103,1992.
386
Clinical Endocrinology (1994) 40,413-420
Phase-shifts in melatonin, 6-sulphatoxymelatonin and alertness rhythms after treatment with moderately bright light at night
Stephen J. Deacon and Josephine Arendt School of Biological Sciences, University of Surrey, Guildford, Surrey, UK
(Received 7 April 1993; returned for revision 23 June 1993; finally revised 23 July 1993; accepted 17 August 1993)
Summary
OBJECTIVES Shift work and rapid travel across several time zones leads to desynchronization of Internal circa- dian rhythms from the external environment and from each other with consequent problems of behaviour, phy- siology and performance. Field studies of travellers and shift workers are expensive and difficult to control. This investigation concerns the simulation of such rhythm disturbance in a laboratory environment. The main objec- tives are to assess the ability of controlled exposure to moderately bright light and darkness/sleep to delay circadian rhythms in volunteers without environmental isolation and, secondly, to evaluate the use of different Indices of melatonin (MT) secretion together with self- rated alertness as marker rhythms. PATIENTS Six normal volunteers aged 22-26 years (mean ± SD 24.3±1.4). DESIGN Subjects were exposed to the following periods of moderately bright light (1200 lux) on three consecutive days in early December 1991: Day (D)1: 2000-0200 h, 02: 2200-0400 h and D3: 2400-0600 h. Each period was followed by 8 hours of darkness (< I lux). Hourly blood, sequential 4-hourly urine (8-hourly when asleep) and hourly saliva (except when asleep) samples were taken throughout a 24-hour period on DO (baseline), D4 (1 day post-light treatment) and D7 (4 days post-light treatment). During waking hours, subjective alertness was rated every 2 hours on a visual analogue scale. MEASUREMENTS MT was measured in plasma and saliva, and Its metabolite, 6-sulphatoxymelatonin (aMT6s), was measured in urine. MT, aMTBs and alertness scores were analysed by ANOVA and a cosinor analysis program. RESULTS A delay shift was present in the aMT6s, plasma
Correspondence: Professor Josephine Arendt, School of Biological Sciences, University of Surrey, Guildford GU2 5XH, Surrey, UK. Fax: 0483 576978.
MT and salivary MT rhythms (degree of shift: 2.67±0.3 h (P<0.001, n=5); 2.35±0.29 h (P<0.001, n=6); and 1.97±0.32 h (P<0.01, n=6), mean ±SEM, respectively) 1 day post-light treatment compared to baseline. Adaptation to the Initial phase position was apparent by the 4th post- treatment day. Significant correlations were obtained between plasma MT onset (degree of shift: 3.12±0.74 h (P < 0.001, n=6, mean ± SEM)) and the acrophases (calcu- lated peak times) of plasma MT (P<0.001), salivary MT (P < 0.05) and urinary aMT6s (P < 0.01). A significant phase delay In the alertness rhythm was also evident 1 day post- treatment (3.08 ± 0.67 h (P < 0.01, n=6, mean ± SEM)) with adaptation by the 2nd post-treatment day. CONCLUSIONS This study suggests that these methods of determining MT secretion are comparable and give reli- able assessments of the MT circadian phase position even after a phase-shift. Significant phase-shifts of similar magnitude can be induced In both MT and alertness rhythms using moderate intensity bright light at night.
Rapid time zone change and shift work can lead to conflict between the new external time cues (zeitgebers) and the internal biological clock with consequent physiological disturbances and deleterious effects on, for example, sleep and performance (Seigel et al., 1969; Fevre-Montange et al., 1981; Winget ei al., 1984; Touitou et al., 1990). Pharmaco- logical and non-pharmacological interventions to promote adjustment to rapid change of external time cues have been proposed (Lewy et al., 1985; Arendt et al., 1986; Czeisler et al., 1990; Redfern, 1992).
Suitably timed bright light of sufficient intensity and duration will induce phase-shifts in physiological variables such as core body temperature, cortisol and melatonin secretion (Czeisler et al., 1986; Broadway et al., 1987; Lewy et al., 1987; Minors et al., 1991; Shanahan & Czeisler, 1991). In constant routine ('unmasked') conditions, the magnitude and direction of shift of melatonin and core body tempera- ture are similar (Shanahan & Czeisler, 1991) suggesting a common control mechanism. Under normal environmental conditions, however, MT is less influenced by external factors (other than light, Lewy et a!. (1980)), compared to other circadian rhythms such as core body temperature and cortisol secretion. One primary aim of this study was to devise a simple, rapid and inexpensive method to study the
413
414 S. J. Deacon and J. Arendt
adaptation of circadian (about 24-hour) rhythms during and after a phase-shift without complete environmental isola-
tion. In this initial study a combination of moderately bright light and subsequent imposed darkness was used to induce a phase-shift and MT was employed as a circadian rhythm marker together with self-rated alertness.
Melatonin has been variously measured and compared in plasma, saliva, and its metabolite aMT6s in urine (Markey et al., 1985; Vakkuri, 1985; Bojkowski et a!., 1987; McIntyre et al., 1987; Nowak et a!., 1987) under normal environmental conditions. The onset of its secretion in dim light has been
extensively used by Lewy and co-workers (e. g. Lewy & Sack, 1989) as a circadian phase marker. A second aim of this study was to assess the relative validity of these various approaches in determining the circadian phase position under normal and phase-shifted conditions.
Subjects and methods
Subjects
Six healthy volunteers (3M: 3F), aged 22-26, under no medication (except oral contraceptives) and all non- smokers, were recruited from research students and staff of the University of Surrey. All subjects gave informed consent and the study protocol was approved by the Ethical Committee of the University of Surrey. Prior to the study all volunteers underwent standard biochemical and haemato- logical tests.
They refrained from heavy exercise and caffeine-contain- ing drinks for 12 hours before sampling of body fluids and restricted their alcohol consumption (maximum 2 units per day) throughout the experiment.
Study design
The study was performed in early December 1991. Consis- tent with other related studies (e. g. Czeisler et al., 1986), individual responses were evaluated in comparison to their own baseline values.
Baseline. For one week prior to light treatment (days (D) -7 to 0), the subjects were required to maintain a regular sleep/ wake schedule, retiring to bed at 2330 h and waking at 0800 h, remaining in the dark (< I lux) between these times.
Light treatment. Immediately following the baseline week, the subjects were exposed to the following periods of moderately bright, full spectrum light (average 1200 lux at eye level, irradiance 492 pW/cm2; True-Lite, Duro-Test Corporation, Fairfield, NJ, USA, obtained through Full Spectrum Lighting, High Wycombe, Buckinghamshire, UK)
Clinical Endocrinology (1994) 40
on 3 consecutive days: D1: 2000-0200 h, D2: 2200-0400 h, D3: 2400-0600 h. Each period of light treatment was followed by 8 hours of imposed darkness (< I lux). During the remaining hours of each day strict light control was not imposed until dusk (1630 h), but the subjects were asked to refrain from periods of exposure to bright sunlight. From dusk until the light treatment period, subjects remained in domestic intensity light (cool, white fluorescent, 150-300 lux
at eye level, irradiance 45-90 pW/cm2). Throughout the light treatment (Dl-3) meal-times were shifted progressively by 2 hours each day in line with the imposed light/dark cycle. Subjects were allowed to choose their activity (e. g. reading, writing, watching videos) as long as they glanced periodically (at least every 5 minutes) at the light source. They were continuously supervised throughout these periods.
Post-light treatment. Immediately following the last light treatment, subjects were required to assume the baseline
sleep/wake schedule, i. e. retiring to bed at 2330 h and waking at 0800 h for 4 days (D4-8). The subsequent readaptation of the subjects in their normal environment was then observed. Subjects were allowed to sleep only during the periods of darkness throughout this study. Naps were not allowed.
Sampling
During body fluid sampling and light treatment the subjects remained in the Clinical Investigation Unit at the University
of Surrey. Blood, urine and saliva samples were taken on the last baseline day (DO), the first post-light treatment day (D4) and the fourth post-treatment day (D7). All blood and saliva samples obtained throughout the dark-phase were taken in the recumbent position and daytime samples after sitting for at least 10 minutes.
Blood. An indwelling cannula was inserted into the forearm of each subject, under local anaesthetic, between 1630 and 1700 h. Blood samples (5 ml) were obtained hourly from 1700 h for 24 hours on each sampling day and collected into lithium heparin tubes. The plasma was separated within 20 minutes by centrifugation (2500 g for 10 minutes) and stored at - 20° C until analysis. Sampling during darkness was performed in dim red light (< I lux). A total of 396 ml of blood was removed from each person.
Saliva. 2-3 ml were obtained hourly from 1700 h for 24 hours, except during sleep, on the sampling days (following 2
minutes of chewing on Parafilm), collected into labelled
plastic vials and frozgn at -20°C until analysis. Meals and drinks were taken just after saliva sampling and a thorough
mouthwash taken afterwards to reduce any possible interfer-
ence with salivary MT estimation.
Clinical Endocrinology (1994) 40
Urine. Sequential 4-hourly urine samples (8-hourly when asleep) were collected from 1600 h throughout 24 hours on DO, 4 and 7. Each subject voided at 1600 h on each sampling day. All urine was then collected throughout each 4-hour
period and the bladder completely emptied at the end of that collection period. The volume of each sample was noted and a 5-ml aliquot from each sample was collected into a labelled
vial and frozen at -20°C until analysis.
Behavioural measurements
Throughout the entire experimental period, alertness was rated every 2 hours on a 10-cm visual analogue scale (VAS, `Drowsy. -. Alert'), except when asleep. In addition, daily
sleep logs were kept and performance tasks were carried out. The effects of an induced phase-shift on various sleep and performance parameters will be reported elsewhere.
Hormone assays
Plasma MT levels were determined by direct RIA using the method described by Fraser et al. (1983), with antiserum (no. G/S/704-8403) obtained from Stockgrand Ltd (University of Surrey, Guildford, UK). The interassay coefficients of variation were 10.4,15.2 and 9% at 97,142 and 905 pmol/l (n=10). The minimum detection limit (defined as the concentration at 2 SD from the mean at maximum bound counts) was 26 pmol/l.
Salivary MT was determined using a GCMS-validated modification (English et a!., 1993) of the direct plasma MT RIA described by Fraser et al. (1983) using an iodinated tracer, with antiserum no. 19540/16876 from Stockgrand Ltd. The interassay coefficients of variation from routine quality control samples were 15.3,11.7 and 5.5% at 105,169 and 288 pmol/I (n = 6). The minimum level of detection was 13.8 pmol/l.
Urinary aMT6s was determined using the method de- scribed by Aldous and Arendt (1988) with antiserum (no. HP/S/1118-23884) from Stockgrand Ltd. The interassay coefficients of variation were 16.8,11.6 and 5.8% at 7.2,63.3 and 132 nmol/l (n = 6). The minimum level of detection was 1.2 nmol/l.
Statistical analysis
The raw hormonal data were subjected to two-way ANOVA
with replication. Subsequently, a cosinor analysis program (written by Dr D. S. Minors, University of Manchester, UK) was used to derive the rhythm characteristics of both the alertness and hormonal data. ANOVA followed by Dun-
can's Multiple Range Test, or Student's t-test where appro-
Phase-shifts in melatonin and alertness rhythms 415
priate, were applied to the cosinor-derived parameters of acrophase (calculated peak time), amplitude and daily mean (for alertness). All data below the limit of detection were excluded from the analysis, except for the purposes of the cosinor program where values lower than this limit were assigned a value of half the minimum detection limit of the assay.
The onset and offset of plasma MT secretion were defined as twice the minimum detection limit of the assay when two consecutive values either exceeded or were below this amount. The relations between concentrations of MT in
saliva and plasma, and aMT6s in urine, and between their respective acrophases and plasma MT onset time, were determined by linear regression analysis and Pearson's
correlation. Mean daily alertness scores were also determined for each
individual and calculated as a percentage of their own baseline. All values quoted are mean±SEM unless stated otherwise.
Results
No subjects reported any oral lesion or discomfort during the
study and no saliva sample appeared to be contaminated with blood. On four of the baseline days (D -6 to - 3) one subject suffered from flu-like symptoms which subsequently resulted in reduced alertness and disrupted his sleep pattern. The alertness data obtained on these days were excluded from the analysis.
The profiles of urinary aMT6s, plasma MT and salivary MT are shown in Fig. 1.
Inspection of the aMT6s data indicated a marked phase delay on the first post-light treatment day (D4) in five out of six subjects. One subject showed a threefold increase in the amplitude of aMT6s on D4 with no phase change and, in contrast to all the other subjects, a poor correlation with plasma and salivary MT. This subject was consequently excluded from the analysis of aMT6s.
A significant time-dependent effect of treatment (ANOVA-phase delay) was shown by two measures of MT secretion (aMT6s, P<0.01; plasma MT, P<0.05) but not with salivary MT, on the first post-light treatment day (D4), which was no longer present on the fourth post-light treatment day (D7) (P<0-l).
Cosinor analysis gave most significant fits of plasma MT (17 out of 18 profiles at P<0.001), followed by salivary MT (14 out of 18 at P<0.05) and urinary aMT6s (10 out of 15 at P<0.1) reflecting the missing data during sleep for both salivary MT and aMT6s. However, the percentage of the variance accounted for by a cosine curve was greater than 81 % in the aMT6s profiles.
416 S. J. Deacon and J. Arendt
` D
3
2
I
1600- 2000- 2330- 0800- 1200- 2000 2330 0800 1200 1600
Clock time
Clinical Endocrinology (1994) 40
1800 2200 0200 0600 1000 1400
Clock time
Fly. 1 Circadian profiles for a, urinary aMT6s; b, plasma MT; and c, salivary MT showing (i) phase delay on , the first post-light treatment day (D4) compared to Q, baseline (DO) and (ii) readaptation by , &, the fourth post-light treatment day (D7), i. e. no significant difference from baseline (DO). Darkness from 2330 to 0800 h. *P<0.05, **P<0.01 significant time-dependent treatment effect: phase delay (two-way ANOVA with replication). Values represent mean±SEM (n=6; urinary aMT6s: n=5).
Using the acrophase data derived from the cosinor program, a significant phase delay (ANOVA) was present in
all measures of MT secretion on the first post-light treatment day. The mean peak time was shifted by 2.67±0.3 h (n=5) for aMT6s (P<0. OO1), 2.35±0.29 h (n=6) for plasma MT (P<0.001) and 1.97±0.32 h (n=6) for salivary MT (P<0.01), l day post-light treatment (Fig. 2). Plasma MT
onset also showed a significant delay on D4 (3.12±0.74 h,
n=6, P<0.001). These shifts did not significantly differ from
each other (P> 0.1). The acrophase estimates and MT onset time for the fourth post-light treatment day (D7) were not significantly different from baseline (P>0.1, ANOVA), but
were significantly different from D4 (P<0.05, ANOVA), indicating re-entrainment by D7. The plasma MT offset time
shifted to a lesser degree (1.33 +0-35 h, n= 6) compared to its onset time.
Table I shows the correlation of acrophases between plasma and salivary MT and urinary aMT6s, together with plasma MT onset times. The urinary aMT6s acrophase occurred on average 1.45 and 2.09 hours after the plasma and salivary MT acrophases, respectively, and 5.6 hours after the plasma MT onset.
There was a direct correlation between plasma and salivary MT in samples taken simultaneously (r=0.758, P<0-001) before the light treatment, immediately after the phase-shift (r=0.792, P<0.001) and on D7 (r=0.576, P<0.01). The difference between plasma and salivary MT levels remained constant throughout the study. Salivary MT
Clinical Endocrinology (1994) 40
80
60
c(i)
1
0 1800 2200 0200 0600 1000 1400
Clock time
Fiq. 1c
levels were estimated to represent a mean ± SD of 40 ± 6.9%
of the corresponding plasma levels throughout the baseline day (DO), 441 ±4.9% over the first post-light treatment day
and 37.4±4.6% on the fourth post-light treatment day, to
give an overall mean±SD value of 40.4±5.3%. The degree of shift for aMT6s on D4 was significantly
correlated (P>0.05) to the pretreatment amplitude of the aMT6s rhythm, such that the greater the amplitude the smaller the phase-shift. No such correlations were found to be statistically reliable with plasma or salivary MT rhythms. No significant change in the rhythm amplitude was apparent after the phase-shift with any of the measures of MT secretion.
Alertness
No significant difference in the timing of the alertness acrophase was found throughout the baseline days (mean
Phase-shifts in melatonin and alertness rhythms 417
0200
[Soo
1600
ý Y U 0 U
1400
4,
0800
c 0
0600
0 E a 0 n
0400 m O jr n O V
4 0200
2400
2200
04
* **
TI1I
7
Days Fig. 2 Acrophase estimates for ", urinary aMT6s; Q, plasma MT; f, salivary MT; , plasma MT onset; and 0, alertness on the baseline day (DO), I day post-light treatment (D4) and 4 days post-light treatment (D7). Darkness from 2330 to 0800 h. Acrophase estimates are represented as mean ± SEM (n=6, n=5 for aMT6s). **P<0. O1, ***P<0.001: significant difference from baseline (ANOVA). D7 was significantly different from D4 (P<0.05) but not from DO for all variables. This suggests a significant phase-shift on the Ist post-light treatment day and re- entrainment by the 4th post-light treatment day (2nd-post treatment day for the alertness rhythm (Fig. 3)).
acrophase 1484±0.28 h). There was a reliable condition effect (P < 0-0 1, ANOVA), with a clear phase delay in the
rhythm (shift: 3-08+0-67h, P<0.01, n=6, ANOVA; 1 day
post-light treatment compared to DO). This effect is shown in
Fig. 2. The alertness rhythm re-entrained by the second post-
418 S. J. Deacon and J. Arendt
Table 1 Correlation and linear regression analysis of acrophase estimates between plasma MT, salivary MT and urinary aMT6s, together with the time of plasma MT onset. y=mean (±SEM)x +mean (±SEM)
Plasma MT/urinary aMT6s y =0"82 (±0-18)x + 1-45 (±0"93) r=0"754, P<0"001, n= 15
Plasma MT/salivary MT y=0-60 (±0.2)x+2"35 (±0"93)
r=0"6, P<0.01, n= 18
Urinary aMT6s/salivary MT y=0"78 (±0.19)x+2"09 (±0-78) r=0.718, P<0"001, n= 18
Plasma MT onset/plasma MT
y=0.62 (f0"I)x+5"06 (±0"23)
r=0"849, P<0"001, n= 18
Plasma MT onset/urinary aMT6s y=0.7 (±0"14)x+5"6 (±0"23)
r=0.69, P<0"01, n= 15
Plasma MT onset/salivary MT
y=039 (t0"15)x+4"55 (±0-24)
r=0S3, P<0"05, n=18
2400
2200
2000
1800
1600
1400
12001 1, III. IIIIIiIII
-7 -5 -3 -I I357
Day #f
Fig. 3 Alertness acrophase estimates (mean ± SEM, n=6; D-6 to D-3, it = 5) throughout an 8-day baseline (D- 7 to DO), a 3-day light treatment period (D1-3) and a 4-day post-light treatment (D4-7). 'P<0.05, "'P<0.01: significant difference from baseline (Duncan's multiple range test). A significant phase-shift can be
observed on D3 and D4 with re-entrainment by D5. f Bright light
administration.
Clinical Endocrinology (1994) 40
light treatment day (Fig. 3). No reliable difference in
amplitude was apparent although the daily mean alertness fell significantly during the first day of the light treatment (P<0.01, Duncan's multiple range). The degree of shift in
the alertness rhythm on D4 (compared to DO) was not significantly different from the shift in urinary aMT6s, plasma MT, salivary MT and the plasma MT onset time (P>0.1).
Discussion
One object of this study was to assess the usefulness of non- invasive methods of determining MT secretion in the form of salivary MT and urinary aMT6s before and after a phase- shift. The close relationship between plasma MT onset as a phase marker, and the calculated peak times of MT in
plasma and saliva, and aMT6s in urine, suggests that all these approaches give valid information.
The light treatment schedule was chosen to enhance the
natural tendency of circadian rhythms in the majority of people to delay. During the light treatment it was not possible to assess correctly the phase position of circadian
rhythms due to the direct, acute effects (`masking') of light. For example, with sufficiently bright light MT is suppressed
and core body temperature is increased (Lewy et al., 1980; Badia et al., 1990). For this reason, the immediate post-light treatment period provides the nearest estimate to the actual
phase-shift induced. During this period (D4), during the baseline (DO), and after readaptation (D7) the MT onset and mean peak values occurred in conditions of dim light or darkness. In such conditions, it is assumed that the measured
phase position of MT represented the actual phase position
on those days, and that the magnitude of the actual induced
shift was greater than or equal to the measured shift on D4.
In order to minimize sleep disturbance which may have
compromised behavioural measurements, saliva and urine
sampling were avoided during the sleep-phase. There was thus some loss of definition in circadian profiles, particularly for salivary MT. In spite of this problem a comparable phase-shift in the salivary MT rhythm was obtained. Clearly, however, the saliva data would have been greatly enhanced by night-time sampling. The salivary MT onset time was not a useful index of the rhythm since a high proportion (67%) of onset times occurred during the sleep-phase after the phase- shift.
In this present study the ratio of salivary MT levels to
plasma MT levels remained constant and the correlation remained highly significant even immediately after the
phase-shift. These results agree well with the other reports (Vakkuri, 1985; McIntyre et al., 1987; Nowak et al., 1987)
Clinical Endocrinology (1994) 40
where samples were obtained throughout 24 hours in a normal environment. McIntyre and colleagues (1987) also showed a good correlation (r = 0.92, P<0.005) when plasma MT and salivary MT levels were suppressed with one hour of bright light between 2400 and 0100 h. Nickelson et al. (1991)
suggested that the relationship between salivary MT and urinary aMT6s may change following a 9-hour forced phase- shift. There is no evidence for this phenomenon in this present study. However, the induced shift was much smaller and more frequent saliva collections were made, at least during waking hours.
The practical advantage of urinary aMT6s is that it
provides an integrated measure for MT production over the night-time period. However, due to limited frequency of sampling it is not able to measure subtle changes in the rhythms. The delayed aMT6s acrophase compared to that of plasma MT is presumably due to both hepatic metabolism and renal excretion mechanisms.
A clear phase delay in the alertness rhythm was evident, comparable to that of MT. This may imply a causal relationship as suitably timed MT administration can modify alertness (Arendt et al., 1987). However, in this study it is more likely that the treatment induced a shift in central rhythm generating mechanisms common to both variables. Whilst the contemporary shift in the timing of meals (social zeitgebers) and the sleep/wake cycle may have had some effect, previous work indicates (for review see Eastman, 1990) that bright light has the more powerful rapid phase shifting ability. A significant decline in daily mean alertness was apparent only on the day of the first light treatment. It is likely that this was due to blood sampling the previous night and throughout the day. On all other days bright light treatment, blood sampling and the effects of the phase-shift did not affect the amplitude or daily mean alertness.
Although field studies have the advantage of being able to expose the individual to the whole range of environmental and social time cues, and associated psychological factors, it is expensive and difficult to standardize these influences in a number of subjects, with the collection of biochemical data
an obvious problem. The use of a three-cycle stimulus of
moderate intensity bright light (1200 lux), suitably timed and
coupled with periods of imposed darkness (<I lux) and
sleep, clearly induces significant phase-shifts in a major
endogenous hormonal circadian rhythm (MT) and a major behavioural rhythm (self-rated alertness). It should therefore be possible to simulate circadian rhythm disorders and
related problems such as jetlag and shiftwork, similar to those seen in field studies, in an inexpensive and controlled
environment which is convenient for biochemical sampling
and where the subjects are left in their normal environment for most of the time.
Phase-shifts in melatonin and alertness rhythms 419
Acknowledgements
S. J. Deacon was supported by a studentship from the MRC
with further support from Stockgrand Ltd (University of Surrey, Guildford, Surrey, UK). We would like to thank Dr J. Wright for medical assistance, J. English and B. Middleton for support and help and Dr S. Folkard and P. Totterdell for invaluable analytical advice.
References
Aldous, M. E. & Arendt, J. (1988) Radioimmunoassay for 6-
sulphatoxymelatonin in urine using an iodinated tracer. Annals of Clinical Biochemistry, 25,298-303.
Arendt, J., Aldous, M., English, J., Marks, V., Arendt, J. H., Marks, M. & Folkard, S. (1987) Some effects of jet lag and their alleviation by melatonin. Ergonomics, 30,1379-1393.
Arendt, J., Aldous, M. & Marks, V. (1986) Alleviation of jet lag by melatonin: preliminary results of controlled double-blind trial. British Medical Journal, 292,1170.
Badia, P., Culpepper, J., Myers, B., Boecker, M. & Harsh, J. (1990) The immediate effect of bright and dim light on body temperature. Sleep Research, 19,386.
Bojkowski, C. J., Arendt, J., Shih, M. C. & Markey, S. P. (1987) Melatonin secretion in humans assessed by measuring its metabo- lite, 6-sulphatoxymelatonin. Clinical Chemistry, 33,1343-1348.
Broadway, J., Arendt, J. & Folkard, S. (1987) Bright light phase shifts the human melatonin rhythm during the Antarctic winter. Neuroscience Letters, 79,185-189.
Cziesler, C. A., Allan, J. S., Strogatz, J. S., Ronda, J. M., Sandrez, R., Rios, C. D., Freitag, W. O., Richardson, G. S. & Kronauer, R. E. (1986) Bright light resets the human circadian pacemaker inde- pendent of the timing of the sleep-wake cycle. Science, 233, 667-671.
Czeisler, C. A., Johnson, M. P., Duffy, J. F., Brown, E. M., Ronda, M. S. & Kronauer, R. E. (1990) Exposure to bright light and darkness to treat physiologic maladaptation to night work. New England Journal of Medicine, 322,1253-1259.
Eastman, C. A. (1990) Circadian rhythms and bright light: recom- mendations for shift work. Work Stress, 4,245-260.
English, J., Middleton, B. A., Arendt, J. & Wirz-Justice, A. (1993) Rapid, direct measurement of melatonin in saliva using an iodinated tracer and solid phase second antibody. Annals of Clinical Biochemistry 30,415-416.
Fevre-Montange, M., Van Cauter, E., Refetoff, S., Desir, J., Tourniaire, J. & Copinschi, G. (1981) Effects of 'jetlag' on hormonal patterns. II Adaptation of melatonin circadian periodi- city. Journal of Clinical Endocrinology and Metabolism, 52(4), 642-649.
Fraser, S., Cowen, P., Franklin, U., Franey, C. & Arendt, J. (1983) A direct radioimmunoassay for melatonin. Clinical Chemistry, 29, 396-399.
Lewy, A. J. & Sack, R. L. (1989) The dim light melatonin onset as a marker for circadian phase position. Chronobiology International, 6(1), 93-102.
Lewy, A. J., Sack, R. C., Muller, S. & Hoban, T. M. (1987) Antide- pressant and circadian phase-shifting effects of light. Science, 235, 352-354.
Lewy, A. J., Sack, R. L. & Singer, C. M. (1985) Treating phase-typed
420 S. J. Deacon and J. Arendt
chronobiologic sleep and mood disorders using appropriately timed bright artificial light. Psychopharmacology Bulletin, 21, 368-437.
Lewy, A. J., Wehr, T. A., Goodwin, F. K., Newsome, D. A. & Markey, S. P. (1980) Light suppresses melatonin secretion in humans. Science, 210,1267-1269.
Markey, S. P., Higa, S., Shih, M., Danforth, D. N. & Tamarkin, L. (1985) The correlation between human plasma melatonin levels and urinary 6-hydroxymelatonin excretion. Clinica Chimica Acta, 150,221-225.
McIntyre, I. M., Norman, T. R., Burrows, G. D. & Armstrong, S. M. (1987) Melatonin rhythm in human plasma and saliva. Journal of Pineal Research, 4,177-183.
Minors, D. S., Waterhouse, J. M. & Wirtz-Justice, A. (1991) A human phase-response curve to light. Neuroscience Letters, 133, 36-40.
Nickelson, T., Samel, A., Maaß, H., Vejvoda, M., Wegmann, H. & Schöfliing, K. (1991) Circadian patterns of salivary melatonin and urinary 6-OH-melatoninsulphate before and after a9 hour time- shift. In Advances in experimental medicine and biology: kynure- nine and serotonin pathways (eds R. Schwarz, S. N. Young & R. R. Brown), pp. 493-496. Plenum Press, New York.
Nowak, R., McMillen, I. C., Redman, J. & Short, R. V. (1987) The
Clinical Endocrinology (1994) 40
correlation between serum and salivary melatonin concentrations and urinary 6-hydroxymelatonin sulphate excretion rates: two non-invasive techniques for monitoring circadian rhythmicity. Clinical Endocrinology, 27,445-452.
Redfern, P. H. (1992) Can pharmacological agents be used effectively in the alleviation of jetlag? Drugs, 42,146-153.
Seigel, P. V., Gerathewohl, S. J. & Mohler, S. R. (1969) Time-zone effects. Science, 164,1249-1255.
Shanahan, T. L. & Czeisler, C. A. (1991) Light exposure induces equivalent phase shifts of the endogenous circadian rhythms of circulating plasma melatonin and core body temperature in men. Journal of Clinical Endocrinology and Metabolism, 73(2), 227-235.
Touitou, Y., Motohashi, Y., Reinberg, A., Touitou, C., Bourdeleau, P., Bogdan, A. & Auzeby, A. (1990) Effect of shiftwork on the night-time secretory patterns of melatonin, prolactin, cortisol and testosterone. European Journal of Applied Physiology, 60,288- 292.
Vakkuri, 0. (1985) Diurnal rhythm of melatonin in human saliva. Acta Physiologica Scandinavia, 124,409-412.
Winget, C. M., DeRoshia, C. W., Markley, C. L. & Holley, D. C. (1984) A review of human physiological and performance changes associated with desynchronosis of biological rhythms. Aviational Space and Environmental Medicine, 55,1085-1096.
ELSEVIER Neuroscience Letters 178 (1994) 32-34 NEUROSCIENCE LETTERS
Acute phase-shifting effects of melatonin associated with suppression of core body temperature in humans
S. Deacon, J. English, J. Arendt* Endocrinology and Metabolism Group, School of Biological Sciences, University of Surrey, Guildford, Surrey GU2 5XH, UK
Received 25 May 1994; Revised version received 29 June 1994; Accepted 6 July 1994
Abstract Appropriately administered melatonin is able to phase shift circadian rhythms, to induce transient sleepiness and to suppress core
body temperature. The relationships between these phenomena have not been fully explored. In a double-blind, placebo-controlled crossover study, 8 healthy males maintained a regular sleep-wake cycle in a natural environment throughout. From dusk until 2400 h on days (D) 1-4 subjects were in dim artificial lighting (< 100 lux) with darkness (< 1 lux) from 2400-0800 h. Sunglasses were worn during the day when outside. At 1700h on D3 either melatonin (5 mg) or placebo was administered. Saliva samples were collected at 30 min intervals, 1600-2400 h on D3 and D4, and subjective alertness rated at 30 min intervals from 1600-2400 h on D3 and hourly from 0800-2400 h D4. Sleep quality was rated on nights 2,3 and 4 and core temperature was recorded throughout. Melatonin induced a significant suppression of temperature and alertness peaking 2.5 h after the dose, together with improved sleep quality on the night of D3 and a phase advance of the endogenous melatonin rhythm on D4. The degree of phase shift was related to the amount of temperature suppression in 6 of 7 subjects with detectable melatonin, suggesting that temperature suppression is an integral part of the phase-shifting mechanism.
Key words: Melatonin; Temperature; Circadian rhythm
The pineal hormone melatonin has been proposed as a treatment for circadian rhythm disorders such as jet lag and maladaptation to shift work [1]. Several reports have indicated that melatonin has both phase-advancing and phase-delaying effects on the internal biological clock [3,6,11,13,14] and that the successful alleviation of dis- turbed sleep and mood associated with a rapid shift of local time cues is accompanied by a quicker adaptation of various circadian rhythms [2,15]. Melatonin also has acute transient effects on body temperature [4,7,9], sub- jective sleepiness [7,9] and the sleep propensity function [16]. The mechanism of action is unknown, but the rhyth- mic electrical activity of the suprachiasmatic nucleus (the, major central rhythm generating system of the hypothal- amus) can be shifted in vitro by melatonin [13]. Human studies have generally used daily timed administration of melatonin to induce a cumulative phase shift [3,11 ]. Only one report to date has considered the acute phase shifting
*Corresponding author. Fax: (44) 483-576 978.
effects of melatonin using its own endogenous rhythm as a phase marker [6]. We report here that even in a normal environment a single oral melatonin treatment can phase-advance its own endogenous rhythm (arguably the best marker of the circadian clock) within 24 h, with a similar tendency in core body temperature, and that this is preceded by immediate suppression of core body tem- perature and subjective alertness. These effects were ac- companied by an improvement in subjective sleep quality.
Eight healthy men aged 23-28 years, taking no medi- cation, were recruited from members of staff and post- graduates of the University of Surrey. All volunteers gave informed consent and the study was approved by the University of Surrey Advisory Committee on Ethics. Except where specified, subjects continued their normal daily activities during the experimental period but re- frained from heavy exercise and alcohol consumption. Melatonin (5 mg) or placebo was administered to all subjects in a randomised, double-blind cross-over study. Each leg of the study began on the same day of the week and was performed over 5 days (D1-D5) with at least a
0304-3940/94/$7.00 © 1994 Elsevier Science Ireland Ltd. All rights reserved SSDI 0304-3940(94)00530-3
S. Deacon et al. lNenroscience Leiters 178 (1994) 32-34
week between treatments, during winter at 52°N. Sub- jects were required to keep a regular sleep--wake cycle, sleeping and/or in darkness (< I lux) from 2400 0800 h,
and from dusk (approximately 1630h) until 2400 h they remained in dim artificial lighting (< 100 lux) throughout the experiment. Sunglasses were worn during the day
when outside. Rectal temperature was recorded continu- ously every 6 min using Squirrel Minilogger devices (Grant Instruments, Cambridge, UK) from 0800 h on D2 to 0800 h on D5.
Days I and 2 served to establish a common sleep wake cycle and to accustom subjects to the rectal ther- mometers. On D3 subjects received either a single dose
of oral melatonin (5 mg in lactose-gelatine capsules) or %placebo (vehicle) at 1700 h. From 1600 h on D3 and D4 they remained seated, except for brief (less than 5 min) periods if urination/defecation was necessary. They re- ceived the same standard meal at 2130 h on D3 and D4. At 30 min intervals from 1600 h until 2400 h on D3 and D4 saliva samples were collected after rinsing the mouth with tap water and subjective alertness scores (visual analogue scale-drowsy-alert') were rated. Alertness was also rated hourly from 0800-2400 h on D4. Sleep quality was rated on a visual analogue scale (best ever-worst ever) on awakening for the nights of D2, D3 and D4.
Melatonin was measured in saliva by GCMS-vali- dated radioimmunoassay (limit of detection, 0.95 ± 0.43
pg/ml, mean ± S. D., n= 10) [10]. Differences in core body temperature, alertness and melatonin were assessed by two-way analysis of variance (ANOVA) for repeated measures (factors: time, treatment) and area under the
curve (AUC) analysis followed by Student's paired t-test. The phase of the core body temperature rhythm on D4- 5
was determined by cosinor analysis (program written by Dr. D. S. Minors, University of Manchester, UK) fol-
lowed by paired Student's t-test. The onset of melatonin secretion was determined on D4 and was defined as the time when concentrations exceeded 4 pg/ml, when 2 con- secutive values exceeded this amount. Any relationship between the magnitude of temperature suppression (AUC) and degree of phase shift of melatonin onset time was determined by regression correlation. Differences in
sleep quality between placebo and melatonin, deter-
mined as a percentage of a baseline night (D2) were assessed by paired Student's t-test.
Melatonin clearly induced an acute, transient but sig- jr nificant suppression of body temperature between 1700
and 2400 h and alertness between 1700 and 2100 h (AUC
analysis, paired t-test, P=0.02) (Fig. 1). ANOVA identi- fied a significant treatment effect on temperature be-
tween 1812 and 2030 h (F = 5.02, df = 1,242, P<0.05)
with a mean (±S. D. ) fall in temperature of 0.26 ± 0.19°C, together with a significant fall in alertness by 31.1 ± 1.1% (mean ± S. D. ) between 1700-2130 h (F= 6.67, df = 1,12, P<0.05). Salivary melatonin was undetectable in one subject, however in 7 subjects a sig-
37 .2
320
36 8
36.6
a)
36.4
T b) 100
80
60
40
Placebo Melatonin
;z
217: 00 19: 00 21: 00 23: 00 01: 00
t Melatonin
--6 Mclatonin
-0- Placebo
Time (hour)
Fig. I. Time course of the suppression of core body temperature, meas-
ured every 6 min (a) and subjective alertness measured every 30 min (b) following a single oral administration of melatonin (5 mg) at 1700 h. Temperature, 1700-2400 h, alertness 1700- 2100 h: AUC analysis, paired I-test, P=0.02. Temperature: repeated measures ANOVA be-
tween 1812 and 2030 h, F= 5.02, dl'= 1.242, P<0.05. The following day, onset of mclatonin secretion was advanced after receiving mela- tonin compared to placebo (c), ANOVA, time by treatment, P=0.0078, shift 1.14 ± 11.49 h (mean ± S. D. ). All values are
mean ± S. E. M., n=8 (n =7 for salivary mclatonin).
nilicant time by treatment erect (ANOVA, F=2.73, df= 10,60, P<0.01) was seen, where melatonin ad- vanced the melatonin onset time relative to placebo by
1.14 ± 0.49 h (mean ± S. D. ) on D4, the day following
administration. Melatonin concentrations in saliva reached a maximum at 1730 h in all subjects (4363 ± 2621 pg/m1, mean ± S. D. ), whereas the mini- mum temperature following melatonin administration was found at 1934 h± 49 min (mean ± S. D. ). There was no significant correlation between the individual maxi- mum salivary melatonin concentration and the degree of temperature suppression. A recent report [9] suggests that a dose response relationship exists between the
exogenous dose of melatonin and the amount of temper-
ature suppression. In the present study the limited frequency of saliva sampling precludes accurate correla- tions.
T Time (hour) Meal
34 S. Deacon et aL /Neuroscience Letters 178 (1994) 32-34
The distortion of the temperature rhythm by mela- tonin during the evening and night of day 3 invalidates any phase estimates. However the calculated peak time (acrophase) of the temperature rhythm from 0800 h on D4 to 0800 h on D5 also tended to phase advance (acro- phase after placebo, 17.01 ± 1.25 h, after melatonin, 16.18 ± 1.32 h, mean ± S. D., P=0.053). One subject slept, intermittently, for 3h after receiving melatonin and showed an exceptionally large drop in temperature (0.65°C), presumably due to masking. Excluding this subject there was a significant correlation between the magnitude of melatonin-induced temperature suppres- sion from 1812-2030 h and the degree of phase shift in melatonin onset time the following day in 6 subjects with detectable melatonin (r = 0.843, P=0.035). Expressed as a percentage of the baseline night, sleep quality on the night of D3 was improved by 70 ± 52% (mean ± S. D., P<0.05, n= 8) and also showed a tendency to improve (by 50 ± 61%, P=0.053, n= 8) on the night of D4. There were no significant changes in alertness on D4.
The timing of the fall in temperature and alertness correspond closely to the reported increase in sleep pro- pensity after 5 mg melatonin given in the late afternoon [16]. This acute `masking' is followed by a significant advance in the timing of the melatonin onset time in saliva the following evening which may be related to the degree of temperature suppression. The temperature rhythm itself is more susceptible than melatonin to ex- ogenous influences [17]. Thus, whilst it is possible to avoid masking of the melatonin rhythm by control of bright light exposure and posture [8,12], phase shifts in the temperature rhythm may have been obscured in a normal environment. It is clear that acute induction of early evening drowsiness by melatonin does not compro- mise and indeed appears to enhance sleep quality during the subsequent night.
Whilst the maximum of the endogenous melatonin rhythm corresponds closely to the nadir in core body temperature (e. g. [4]), the maximum suppression of tem- perature obtained using a pharmacological dose of mela- tonin in this study occurred approximately 2h after treatment. The observed effect may not therefore be rep- resentative of endogenous mechanisms.
In previous experiments using 5 mg melatonin to has- ten adaptation to phase shift, subjects could not nor- mally distinguish treatment from placebo [2,15]. The acute drowsiness, and in one case intermittent sleep re- corded here, and in recent experiments using even lower doses of melatonin [9], may be related to the experimen- tal conditions whereby subjects seated in a warm room in dim light, or indeed requested to try to sleep, perceive transient sleepiness more clearly than when active.
Cagnacci et al. [4] reported acute suppressant effects of melatonin when given in divided doses in the morning. They and others have provided evidence [5] that mela- tonin counteracts the stimulatory effect of bright light on
body temperature at night. In view of the present results it would not be unreasonable to suggest that the primary event in the induction of a phase shift by bright light or by melatonin is an acute change in body temperature. If this proves to be the case it provides substantial scope for investigation of phase-shifting mechanisms via the cen- tral hypothalamic control of body temperature.
[1] Arendt, J., Clinical perspectives for melatonin and its agonists, Biol. Psychiatry, 35 (1994) 1.
[2] Arendt, J., Aldhous, M., English, J., Marks, V., Arendt, J. H., Marks, M. and Folkard, S., Some effects of jet lag and their alleviation by melatonin, Ergonomics, 30 (1987) 1379-1393.
[3] Arendt, J., Bojkowski, C., Folkard, S., Franey, C., Minors, D., Waterhouse, J., Wever, R. A., Wildgruber, C., Wright, J. and Marks, V., Some effects of melatonin and the control of its secre ll
tion in man. In D. Evered and S. Clark (Eds. ), Ciba Foundation Symposium 117, Photoperiodism, Melatonin and the Pineal, Pitman, London, 1985, pp. 266-283.
[4] Cagnacci, A., Elliott, J. A. and Yen, S. S. C., Melatonin: A major
[51
regulator of the circadian rhythm of core body temperature in humans, J. Clin. Endocrinol. Metab., 75 (1992) 447-452. Cagnacci, A., Soldani, R. and Yen, S. S. C., The effect of light on core body temperature is mediated by melatonin in women, J. Clin. Endocrinol. Metab., 76 (1993) 1036-38.
[6] Claustrat, B., Presented at the Philipe Laudat Conference on Neureobiology of Circadian and Seasonal Rhythms: Animal and Clinical Studies, Strasbourg/Bischenberg, France, 1991.
[7] Dawson, D. and Encel, N., Melatonin and sleep in humans, J. Pineal Res., 15 (1993) 1-12.
[8] Deacon, S. and Arendt, J., Posture influences melatonin concen- trations in plasma and saliva in humans, Neurosci. Lett., 167 (1994) 191-194.
[9] Dollins, A. B., Zhdanova, I. V., Wurtman, R. J., Lynch, H. J. and Deng, M. H., Effect of inducing nocturnal serum melatonin con- centrations in daytime on sleep, mood, body temperature, and performance, Proc. Natl. Acad. Sci. USA, 91 (1994) 1824-1828.
[10] English, J., Middleton, B., Arendt, J. and Wirz-Justice, A., Rapid, direct measurement of melatonin in saliva using an iodinated tracer and solid phase second antibody, Ann. Clin. Biochem., 30 (1993) 415-416.
[11] Lewy, A. J., Ahmed, S., Latham Jackson, J. M. and Sack, R. L., Melatonin shifts human circadian rhythms according to a phase- response curve, Chronobiol. Int., 9 (1992) 380-392.
[12] Lewy, A. J., Wehr, T. A., Goodwin, F. K., Newsome, D. A. and Markey, S. P., Light suppresses melatonin secretion in humans, Science, 210 (1980) 1267-1269.
[13] McCarther, A. J., Gillette, M. U. and Prosser, R. A., Melatonin directly resets the rat suprachiasmatic clock in vitro, Brain Res., 565 (1991) 158-161.
[14] Sack, R. L., Lewy, A. J. and Hoban, T. M., Free-running melatonin rhythms in blind people: phase shifts with melatonin and Tria- zolam administration. In D. Evered and S. Clark (Eds. ), Temporal Disorder in Human Oscillatory Systems, Springer, Heidelberg, ' 1987, pp. 219-224.
[15] Samel, A., Wegman, H. M., Vejvoda, M. and Maas, H., Influence
of melatonin treatment on human circadian rhythmicity before and after a simulated 9 hour time shift, J. Biol. Rhythms, 6 (1991) 235-248.
[16] Tzischinsky, 0., Lavie, P. and Pal, I., Time-dependent effects of 5 mg melatonin on the sleep propensity function, J. Sleep Res., 1 Suppl. 1 (1992) 234.
[17] Wever, R. A., Characteristics of circadian rhythms in human func- tions. In R. J. Wurtman and R. J. Waldhauser (Eds. ), Melatonin in Humans, J. Neural Transm., Suppl. 21 (1986) 323-374.
Neuroscience Letters, 167 (1994) 191-194 191 © 1994 Elsevier Science Ireland Ltd. All rights reserved 0304-3940/94/$ 07.00
NSL 10234
Posture influences melatonin concentrations in plasma and saliva in humans
Stephen Deacon, Josephine Arendt* Endocrinology and Metabolism Group, School of Biological Sciences, University of Surrey, Guildford, Surrey GU2 SXH, UK
(Received 21 october 1993; Revised version received 8 December 1993; Accepted 10 December 1993)
Key words: Posture; Masking; Melatonin; Circadian rhythm; Saliva
Melatonin is extensively used as a circadian marker rhythm and thus any factors influencing its concentrations other than endogenous rhythmicity must be assessed. We report here the effects of posture on melatonin concentrations in plasma and saliva. The study was performed during the rising phase of the melatonin rhythm between 21: 00 and 01: 30 It. From 23: 00 h, 7 healthy subjects remained sitting until 24: 00 h and in dim light (< 10 lux) until 01: 30 h. From 24: 00 h, they assumed a different postural position in one of the following stages: Stage 1 a: 24: 00-00: 30 It - standing, 00: 30-01: 00 It - supine, 01: 00-01: 30 It - standing; Stage lb - reverse of Stage la, Stage 2a: 24: 00-01: 30 h- supine; and Stage 2b: 24: 00-01: 30 h- standing. Blood and saliva samples were obtained every hour from 21: 00-24: 00 h and then every 10 min from 24: 00-01: 30 It. Plasma and salivary melatonin concentrations, measured by radioimmunoassay, increased when moving from a supine to a standing position and decreased when these positions were reversed. These changes, as seen with other blood components, can be explained through the influence of gravity which causes a decrease in plasma volume on standing and an increase in plasma volume on lying down.
The force of gravity causes significant changes in the human circulatory system and with the continual chang- ing of posture in everyday life the influence of gravity on various blood parameters becomes apparent. In healthy subjects, when an upright position is adopted after lying recumbent (supine), the plasma volume is reduced and the concentration of various plasma components, espe- cially that of the proteins and of the constituents bound to them, increases. This rise in plasma constituent con- centration is in the order of 10-20% (e. g. ref. 7) and the effect is reversible on reversing the postural position. Such studies suggest that other plasma constituents such as hormones and metabolites might show similar changes in plasma levels.
The pineal hormone melatonin is secreted into the blood in a highly stable and reproducible pattern within individuals [2]. The rhythm is endogenous, and governed by an internal circadian pacemaker: the suprachiasmatic nucleus of the hypothalamus. An important feature of melatonin secretion is that, with the exception of light [8], it is less susceptible to behavioural and environmental masking (disturbing) influences, than other overt (meas- ured) circadian rhythms such as core body temperature and cortisol secretion. Masking influences include sleep,
'Corresponding author. Fax: (44) 483-509712.
waking, meals and stress and can result in the unreliable estimation of the circadian time of the internal biological clock. For this reason the melatonin circadian rhythm is thought to provide an appropriate and reliable marker of the endogenous biological horologue under normal con- ditions, as well as in constant routine and in other iso- lated environments.
Although melatonin, a lipophilic molecule, easily fil- ters through capillary membranes, up to 70% of mela- tonin may be weakly bound to albumin [3] which is not readily diffusible, and albumin-bound hormones do not pass into the surrounding tissues [10]. It is therefore pos- sible that a shift in posture may cause a change in plasma melatonin levels. Salivary melatonin is thought to closely reflect the changes in plasma melatonin levels [9] and since saliva is becoming a routine method of determining melatonin secretion it is important to consider the effects of posture on salivary melatonin levels as well. It is the purpose of this investigation to study the possible mask- ing influence of posture on melatonin levels.
Seven healthy volunteers (3M: 4F), aged 24-35 years (26.3 ±4 years, mean ± S. D. ), under no medication (ex- cept oral contraceptives), one of whom was a smoker, were recruited from members of staff and research post- graduates of the University of Surrey. All volunteers gave informed consent and the study was approved by
SSD/ 0304-3940(93)E0858-S
192
the University of Surrey Advisory Committee on Ethics. They refrained from heavy exercise, drinking alcoholic beverages or those containing caffeine, for 12 h prior to, and during each stage of the study. On each day of the study, the subjects moderated their fluid intake to no more than 1.5L daily.
The 4-day study was divided into two periods (stages I and 2) of 2 days, I week apart, and was performed in October during the rising phase of the melatonin rhythm, between 21: 00 h and 01: 30 h. On all 4 days subjects re- mained in a sitting position under domestic lighting (150-300 lux) from 21: 00 to 23: 00 h. From 23: 00 to 01: 30 h, subjects remained sitting (until 24: 00 h) and awake under dim light (< 10 lux). After this baseline pe- riod, the treatments were applied within each stage as a randomised crossover design. From 24: 00 h, the subjects assumed a different postural position as follows:
Stage la: 24: 00 h- standing, 00: 30 h- supine, 01: 00 h
- standing; Stage 1 b: 24: 00 h- supine, 00; 30 h- standing, 01: 00 h- supine; Stage 2a: 2400-01: 30 h- supine; Stage 2b: 2400-01: 30 h- standing. Brief and slow movement was allowed throughout each stage.
Blood samples (5 ml, collected by dim red torch light
after 23: 00 h; < 10 lux) were taken hourly via an indwell- ing forearm cannula, together with hourly saliva samples (after chewing on parafilm), from 21: 00 to 24: 00 h and then every 10 min from 24: 00 h until 01: 30 h. Only water was allowed just after saliva sampling, to reduce any pos- sible interference with salivary melatonin estimation. Blood samples were collected into lithium heparin tubes
and centrifuged (2,500 xg for 15 min), the plasma sepa- rated within 25 min of collection and stored at -20°C until analysed for melatonin using the method of Fraser
et al. [6]. The interassay coefficients of variation were 12.4%, 9% and 7.3% at 19.4,36.8 and 191 pg/ml (n = 10). The minimum detection limit was 3.8 ± 0.24 pg/ml (mean ± S. E. M. ). Saliva samples were collected into
minivials and stored at -20°C until analysed for mela- tonin using the method of English et al. [5]. The interas-
say coefficients of variation were 16%, 10%, 9% and 6%
at 8.7,20,36 and 58 pg/ml (n = 8). The minimum level of detection was 1.9 ± 0.15 pg/ml (mean ± S. E. M. ).
All plasma and salivary melatonin values were calcu- lated as a percentage change in melatonin levels from the initial point of assuming a different postural position using the subjects as their own controls. Since each stage was of a random crossover design, any differences be- tween standing and supine positions were compared be- tween stages la and lb and between stages 2a and 2b. Area under the curve (AUC) analysis followed by Stu- dent's paired t-tests (two-tailed) were used to identify sig- nificant differences between postures. Paired compari- sons were also made at each 10-min sampling point
120
100
80
! ýj 60 C G)
NC ^O C 40
O 20 ^ ý0
-20
A.
2400/0 10 20 30/0 10 20 30/0 10 20 30 Time (h/mins)
so
60
40 y ?ý 20 ýý ýc ý'ý
Ü ie ý6 -40
B.
2400/0 0 20 30/0 10 20 30/0 10 20 30 Time (h/mins)
Fig. 1. Percentage change in A. plasma melatonin and B. salivary mela- tonin levels within 30 min in alternate postural positions from 24: 00 h. Mean ± S. D., n=7. *P < 0.05, **P < 0.01, ***P < 0.001: significant differences between standing and supine positions (stage la (": versus stage lb (Q), paired t-test, two-tailed). Mean difference in melatonin levels (± S. D. ) between standing and supine: 24: 00-00: 30 h (plasma: P<0.05,35.2 ± 21.3%: saliva: P<0.01,42 ± 27%), 00: 30-01: 00 h (plasma: P<0.0 1,33.5 ± 16.2%; saliva: 30.9 ± 34.4%) and 01: 00-01: 30 h (plasma: P<0.01; 36.7 ± 17.5%; saliva: 31.9 ± 44.8%), significance
levels determined from AUC analysis followed by paired t-test.
between standing and supine using Student's paired t- tests (two-tailed). Correlations between plasma and sali- vary melatonin levels were carried out using Pearson's correlation. All values lower than the detection limit of the assay were excluded from the calculations. This cor- responded to 5% of the plasma melatonin data and 6% of the salivary melatonin data. All values are quoted as mean ± S. D. unless stated otherwise.
No subject reported any oral lesion and no blood con- tamination was observed in the saliva samples. All sub- jects completed stage 1. Four subjects completed stage 2 and one further subject completed stage 2a and 60 min of stage 2b.
Fig. 1 shows the percentage change in melatonin levels within 30 min periods in alternate postural positions from 24: 00 h. Changing from standing to supine and back to standing (stage 1 a), and likewise when the posi- tions were reversed (stage 1 b), caused marked differences in percentage change of both plasma and salivary mela- tonin levels. AUC analysis identified significant differ- ences between standing and supine positions with greater
0 ;G2 -20
193
- Q
.-
ý
I 1ý(1
100 50
0 Eý -50
ý-I oo ý 'A 100
ý -- 50
<D 0
clu -50
-100 200
150
100
50 0
-50
y
t I1-11"ight St: utdinb/Supinc
%P
2100 2300 10 20 30 40 50 60 70 80 90 22002400/0
Upright Stand ing/Supinc
250
150
50
-100 0
300
200
100
0
_inn 2100 23°° 10 20 30 40 50 60 70 80 90 22002400/0
Time (h/mins)
Time (h/mins) Fig. 2. Individual profiles showing percentage change in plasma melatonin levels over a 90 min period from 24: 00 h. Significant difference (P < 0.01) over 90 min between supine (A, stage 2a ) and standing (0, stage 2b ) positions for plasma melatonin only (n = 4; AUC analysis, paired t-test). Mean difference in melatonin levels ± S. D. -- plasma: 18.6 ± 14.8%, saliva: 25.7 ± 46.3%. The changes in plasma melatonin levels reached significance
(P < 0.05. paired t-test, n= 4) at 20,40,50,60 and 70 min between stages 2a and b.
increases associated with the standing position (Fig. 1). Significant changes between postures were observed within 10 min in all three time periods for plasma mela- tonin, Fig. lA. Differences in salivary melatonin levels between standing and supine positions were less signifi- cant, Fig. 1 B.
Fig. 2 shows the individual profiles for the percentage change in plasma melatonin levels over a 90 min period between stages 2a and 2b. Differences between standing and supine positions were significant for plasma mela- tonin only (P < 0.01) with AUC analysis. The changes in salivary melatonin levels were more variable and proved not to be significant.
Good correlation between plasma and salivary mela- tonin levels was obtained during hourly sampling (r = 0.83, P<0.0001, y=0.59 [± 0.01] x-2 [±1.2]) from 21: 00 to 24: 00 h. Although still highly significant, the correlation between the two melatonin mediums was not as good during short term sampling (r = 0.555, P<0.0001, y=0.23 [± 0.02] x+5.7 [± 0.9]). Salivary
II. I..
melatonin levels were estimated to represent 46.2 ± 18.1% and 40.4 ± 18.7% during hourly and 10 min sampling, respectively, of the corresponding plasma levels.
When moving from a supine position to a standing position, blood volume becomes redistributed from the central pool within the chest, to the lower limbs where small molecules pass through the capillary membranes into the interstitial space by diffusion. The plasma vol- ume is reduced, thus increasing the concentration of non- filtrable blood constituents such as red cells, and proteins together with their bound components [7], including al- bumin-bound hormones [10].
The results from this study indicate that changing from the standing to the supine position caused a marked fall in plasma and salivary melatonin levels which was reversed on returning to the standing position once again. When these positions were reversed over the same period of time an opposite effect of similar magnitude was seen. These differences were significant for plasma
194
melatonin within 10 min as calculated between stages 1a
and 2b and within 20 min comparing stages 2a and 2b. These changes showed no reliable (P > 0.1) progressive change with time indicating that the changes in plasma melatonin levels associated with the effects of posture were rapid and remained consistent after approximately 20 min. This agrees well with other posture-related stud- ies affecting various blood constituents where little change occurred after 20 min (e. g. ref. 7).
The effects on salivary melatonin levels were more var- iable. This may be associated with frequent and forced sampling (every 10 min) which probably changes sali- vary composition and secretory mechanisms of the gland [4], modifying the true melatonin concentration in saliva. Although the overall proportion of melatonin levels in saliva to plasma, throughout hourly and 10 min sam- pling, remained comparable, the correlation fell consid- erably throughout 10 min sampling.
Abalan et al. [1] recently noted a difference of 32.6% ± 34.3 (mean ± S. D., range: -20% to 94%) in total plasma cortisol - another circadian phase marker - on standing for 40 min compared to remaining supine, during the rising phase of the cortisol rhythm. A similar difference in plasma and salivary melatonin levels was observed in this study even though sampling was more frequent and over different periods of time.
Although considerable interindividual variation was evident, these results indicate that posture has a masking effect on plasma and salivary melatonin levels. Changes due to posture may prove to be important when inter- preting data from subjects in different postural positions, for example, comparing bedridden patients and healthy controls, and may have implications in situations where the body is exposed to other environments where body fluid shifts are also experienced, such as spaceflights, water immersion and exercise. Changes in melatonin and cortisol levels due to posture may alter circadian parame- ters such as daily means and amplitudes and influence- acrophase values. In fact Samel et al. [11] studied the effects on the circadian system of 7 days of bed rest with a constant sleep-wake cycle and a regular photoperiod. Although neither melatonin or cortisol were measured, small changes in rhythm acrophases were evident for noradrenalin, adrenalin, potassium, heart rate and urine volume and significant changes in rhythm amplitudes
were observed. These changes may be attributable to in- activity and posture.
This study indicates the necessity for standardising the timing of blood sampling and indeed saliva sampling with prior knowledge of the postural position and the time in that position. Care should also be taken in the interpretation of results from high frequency saliva sam- pling.
S. Deacon was supported by a studentship from the MRC with further support from Stockgrand Ltd. (Uni- versity of Surrey, Guildford, UK). We would like to thank Dr J. Wright and especially Dr. V. Kulkani for medical assistance, J. English and B. Middleton for help and support.
I Abalan, F., Hui Bon Hoa, A., Guillonneau, D., Ruedas, E., Ca- lache, M., Ellison, W., Rigal, F., Perey, F., Sousselier, M. and Bour- geois, M., Effect of posture on total cortisol plasma concentration, Horm. Metab. Res., 24 (1992) 595-596.
2 Arendt, J., Melatonin (review), Clin. Enäocrinol., 29 (1988) 205- 229.
3 Cardinali, D. P., Lynch, H. J. and Wurtman, R. J., Binding of mela- tonin to human and rat plasma proteins, Endocrinology, 91 (1972) 1213-1218.
4 Dawes, C., Circadian rhythms in human salivary flow rate and composition, J. Physiol., 220 (1972) 529-545.
5 English, J., Middleton, B. A., Arendt, J. and Wirz-Justice, A., Rapid direct measurement of melatonin in saliva using an iodinated tracer and solid phase second antibody, Ann. Clin. Biochem., 30 (1993) 415-416.
6 Fraser, S., Cowen, P., Franklin, U., Franey, C. and Arendt, J., A direct radioimmunoassay for melatonin, Clin. Chem., 29 (1983) 396-399.
7 Hagan, R. D., Diaz, F. J. and Horvath, S. M., Plasma volume changes with movement from supine to standing positions, J. Appl. Physiol., 45(3) (1978) 414-418.
8 Lewy, A. J., Wehr, T. A., Goodwin, F. K., Newsome, D. A. and Markey, S. P., Light suppresses melatonin secretion in humans, Sci- ence, 210 (1980) 1267-1269.
9 Nowak, R., McMillen, I. C., Redman, J. and Short, R. V., The corre- lation between serum and salivary melatonin concentrations and urinary 6-hydroxymelatonin secretion rates: two noninvasive tech- niques for monitoring human circadian rhythmicity, Clin. Endocri- nol., 27 (1987) 445-452.
10 Riad Fahmy, D., Read, G. F., Walker, R. F. and Griffiths, K., Sali- vary progesterone for assessing ovarian function, J. Clin. Chem. Biochem., 19 (1981) 812.
11 Samel, A., Wegmann, H. -M. and Vejvoda, M., Response of the circadian system to 6° head-down tilt bed rest, Aviat. Space. Envi- ron. Med., 64 (1993) 50-54.