Phase advance of the circadian sleep-wake cycle as an antidepressant

Circadian rhythms and psychiatry

There is currently a great deal of interest in whether or not circadian rhythms are capable of explaining certain psychiatric phenomena. This short review will briefly explain the concept of endogenous circadian rhythms and then go on to describe some of the attempts to relate this to psychiatry. It is not possible to give a complete discussion in the space available and interested readers are referred to two recent texts on circadian rhythms (Wever 1979, Minors and Waterhouse, 1981). The psychiatric aspects have been reviewed by Sampson and Jenner (1975).

Individual variations in the sleep of depression

Variations in the intensity or severity of affective disorders were evaluated relative to perturbations in nocturnal sleep physiology. Individual variations in polygraphic features of the sleep cycle based upon psychopathologic scale ratings were investigated in two constituencies (Ns = 6) for 8 hr during 1-3 consecutive nights. The constituencies consisted of twelve young adult (18-25 years) nonpsychotic unipolar depressed psychiatric patients with a primary affective illness and an age-matched normal healthy control group (N = 8). The severely versus mildly depressed patient subgroups scored significantly higher on the Hamilton, Beck and Zung psychopathologic rating scales, indicating a larger magnitude of depressive symptomatology. The average value for total time asleep was 6.1 hr in severely versus 7.8 hr among the mildly depressed patients and controls. EEG-sleep of the severely versus mildly depressed patients and controls contained significantly less stages 2 and 3. Although total time asleep was almost identical in the mildly depressed constituency compared with controls, patients accumulated significantly more of stages 2 and 3. Both patient subgroups exhibited a significantly shorter REM latency than controls. REM latency was reduced to a significantly lower level in the severely versus mildly depressed patients. A significant decrease of REM cycle duration occurred in the polygraphic sleep recordings of severely depressed patients compared with the age-matched controls. The shortened REM latencies indicate a disinhibition of neural processes that would normally delay appearance of the initial REM episodes during nocturnal sleep. The present study generally extends and confirms finding on nocturnal EEG-sleep disturbances in depression associated with the severity of affective illness, particularly the disrupted REM cycle and shorter REM latency.

Human brain contains vasopressin and vasoactive intestinal polypeptide neuronal subpopulations in the suprachiasmatic region

The suprachiasmatic nuclei (SCN) and retinohypothalamic tract ( RHT ) in the anterior hypothalamus have been postulated to play an important role in the timing of daily biological rhythms in mammals. Although physiological studies have described circadian rhythms in man, the presence of an RHT or SCN has not been conclusively demonstrated in the human brain. Immunocytochemical identification of distinct ventral vasoactive intestinal polypeptide (VIP) containing and dorsal vasopressin containing neuronal subpopulations in the human suprachiasmatic region provides correlative evidence of neuronal clusters which are homologous to discrete cell groups in the SCN of other mammalian species. Manipulation of the circadian system has been used to treat some affective illnesses and other physiological timing disorders. Characterization of the neural substrates underlying human circadian rhythms could be useful in the development of future treatment modalities and is essential for understanding normal human circadian organization.

Dichotic listening in the morning and evening

Normal subjects needed greater intensity to detect dichotic and monotic click stimuli in the evening compared to the morning. Also, the asymmetry for detecting dichotic clicks shifted from morning to evening. The morning pattern of left ear lead advantage was absent in the evening. In contrast, there was no morning versus evening change in either overall accuracy or right ear advantage for consonant-vowel discrimination.

Disorders of circadian body temperature rhythm in severely brain-damaged patients

Twenty-four hour patterns of body temperature (BT) were recorded during consecutive 3-10 day spans from 14 severely brain-damaged patients. Seven patients exhibited a normal circadian BT rhythm with an amplitude of more than 1 degree C and a normal phase position of the minimum BT being observed during the latter half of the nocturnal sleep. One patient with a dispersed type of sleep exhibited an extremely low amplitude of the BT rhythm with mean average 0.69 degrees C. In this patient, an 24-hr observation span was insufficient to detect the existence of a BT rhythm. Two patients manifested disturbance of period. As acrophase of the BT rhythm varied from day to day, the standard deviation (S.D.) of mean acrophase was extremely large. For these patients the light-dark cycle did not act as an entrainer because both had visual disturbance. A phase advance of the minimum BT was observed in four patients. The minimum BT appeared in the first half of nocturnal sleep. These three disturbances (amplitude, period and phase), were revealed only by longitudinal observation of the BT rhythm, indicating the importance of long-term observations over a sufficient period to make the nature of the rhythm disturbances clear. These disturbances were not related to that of the sleep-wake cycle, as two patients showed normal circadian BT rhythm in spite of their dispersed-type sleep.

Interindividual differences in a set of biological rhythms documented during the high arctic summer (79 degrees N) in three healthy subjects

The High Arctic summer with its permanent sunlight provides a situation in which one of the natural synchronizers, the light-dark alternation, is minimal. During the summers of 1981 and 1982 three healthy right-handed geographers who were performing field studies in Svalbard as part of their own research volunteered to document, 4-6 times per 24 hr for respectively 63, 141 and 147 days, a set of circadian rhythms: self-rated fatigue, oral temperature, grip strength of both hands, heart rate and times of awakening and retiring. Tests were performed before departure from France, in Svalbard (79 degrees N latitude) where their daily activities were often strenuous, and after returning to France. Time series were treated individually according to three methods: display of data as a function of time, cosinor analyses to quantify rhythm parameters, and spectral analyses to estimate component periods of rhythms. Circadian parameters such as period and acrophase of activity-rest, oral temperature and fatigue rhythms were not altered. On the other hand, the circadian rhythm in grip strength was altered: the period differed from 24 hr in one subject, while grip strength acrophase of the left, but not the right, hand of the other two subjects was phase shifted during the sojourn in Svalbard. A prominent circahemidian (about 12 hr) rhythm was observed in two subjects for their heart rate in Svalbard, while a prominent circadian rhythm (differing from exactly 24 hr) was observed in France associated with a small circahemidian component.

Circadian rhythm of plasma melatonin in endogenous depression

The circadian rhythm of plasma melatonin was investigated in normal men 18-30 years (N = 5), normal men 50-70 years (N = 5) and in six patients with endogenous depression. The environmental photoperiod was 11 hours. The subjects and patients were indoors with lights on from 07:00 until 23:00 hours. Blood samples were obtained every 4 hours over a 24 hour period, with additional sampling at 22:00 and 02:00 hours. Plasma melatonin was estimated by radioimmunoassay compared to both groups of controls. In the depressed patients, the levels of melatonin were low throughout the 24 hour period. The depressives had a delayed onset of the dark phase of the rhythm. The patients also showed peak melatonin levels occurring earlier than in the controls. Circadian rhythm of melatonin and therefore of its pacemaker may be altered in endogenous depression.

Critical interval hypotheses for depression

Two chronobiologic models are presented for the etiology of depression. The internal coincidence model suggests that a phase advance of the strong oscillator in reference to the weak oscillator causes depression. An external coincidence model suggests that depression is caused when the light/dark cycle or photoperiod provides too little illumination during a critical photosensitive interval, which might in turn occur early due to a phase advance. Thus, depression might be treated by drugs or other manipulations which delay the phase of internal circadian rhythms. Depression might also be treated with bright illumination during the critical photosensitive interval. Preliminary experiments suggest that bright artificial light does have antidepressant effects. The optimal times for light exposure and the most responsive patient groups have not yet been identified.

REM latency and depression: computer simulations based on the results of phase delay of sleep in normal subjects

A phase advance of the circadian rhythm of rapid eye movement (REM) sleep propensity relative to the sleep-wake cycle has been hypothesized to account for the abnormalities of REM sleep in depression. One implication of this hypothesis is that an acute phase delay of sleep in normal subjects should produce the same abnormalities of REM sleep. A further implication is that changes in REM sleep that occur in normal subjects with delay shifts of sleep of progressively greater magnitude describe those that occur during the course of a depressive episode. This hypothesis was tested by computer simulation using two equations which, given the data derived from normal subjects experiencing phase delays of sleep, generated the REM latencies expected during successive stages of a depressive episode. For severely depressed patients, the computer-generated data matched those found empirically. The known correlation between severity of depression and REM latency, and the REM latencies of recovered patients, are consistent with the hypothesis.

DST in depression is unaffected by altering the clock time of its administration

Circadian oscillators in major depressive illness may be phase advanced by several hours. We attempted to determine whether phase advance of the oscillator responsible for hypothalamic-pituitary-adrenal (HPA) function in depressives might influence the outcome of the overnight dexamethasone suppression test (DST). Six major depressives underwent DST with dexamethasone doses administered in a randomized fashion at 1900h and 2300h on separate evenings. Twenty-four hour cortisol secretory patterns basally and postdexamethasone were obtained for each subject. Postdexamethasone cortisol responses were similar for both the 1900h and 2300h dosage schedules in suppressors, nonsuppressors, and an early escape responder. We conclude that failure of the HPA axis to suppress normally with DST in major-depressive illness is a primary feature of neuroendocrine regulatory mechanisms rather than secondary to a posited phase advance of the related circadian oscillator.

The sleep therapies of depression

This paper reviews five different types of deliberate sleep-wake manipulations which are reported to have antidepressant effects: total sleep deprivation, partial sleep deprivation, a phase advance of the sleep periods, and REM deprivation. The effects of total sleep deprivation are best documented. Of 852 depressed patients studied, 493 or 57.9% improved following sleep deprivation. The REM deprivation procedure acts more slowly, but is of more lasting clinical value than the other forms. Partial sleep deprivation during the second half of the night may be as good as total sleep deprivation and better tolerated. The findings are reviewed in terms of psychological, neurophysiological, biochemical, and chronobiological perspectives.

Trends in research and treatment of affective disorders

Trends in research on affective disorders were reviewed. Studies which described the contribution of neurotransmitters and neuroendocrines to depression, unipolar and bipolar, and manic states were cited. Women more frequently suffer from affective disorders but women's response to different therapies sometimes is more favorable than men's. Age and married status influence sex differences in the incidence of affective disorders. Symptoms of pain and depression intertwine, and response to analgesia may emerge as an index of response to therapy for depression. Trends in pharmacotherapies for affective disorders were reviewed: lithium, tricyclic antidepressants, and MAOI drugs, as well as trends in the employment of ECT and sleep deprivation to relieve symptoms of affective disorders were reviewed.

Sleep pattern variations as a function of age in affective disorders

Polygraphic features of nocturnal sleep were evaluated in young adult psychiatric patients during acute unipolar depressive episodes. Averaged values and variability of polygraphic indexes were examined in 12 depressed patients under the age of 26 individually matched with a normal control group. Sleep was polygraphically recorded in the Laboratory for three consecutive nights from 12-8.00 a.m. Although average total time asleep was approximately equivalent (greater than 7.3 hr) between groups, depressives accumulated significantly: (i) less stage 4, (ii) more stage 1, (iii) vascillations among sleep stages, but (iv) most especially increased transitions into stage 1 and (v) intermittent wakefulness. The recorded sleep perturbations in young depressives were extremely variable across nights and among individuals. This was especially conspicuous across nights as reflected by significantly larger variability (SD) for: (i) transitions into stage 1, (ii) intermittent wakefulness and (iii) epsilon accumulations of stage 2. Variability (the SD) between individuals was also more substantial for: (i) total time asleep, (ii) stage 1, (iii) intermittent wakefulness, (iv) epsilon stage shifts and (v) intrusions into stage 1. The polygraphic recordings of young depressives contained anomalies reported for clinical pathologic states accompanied by physiological disregulation such as hypersomnia, narcolepsy and schizoaffective disorders. Polygraphic indexes reflecting the capacity (i) to remain asleep (means +/- SDs) and (ii) accumulate continuous sleep (SDs) indicated an imbalance of the 24-hr rest (sleep)--activity (waking) cycle was present in this constituency concomitant with affective distress. A comparison with selected cross-sectional polygraphic studies revealed that sleep cycle aberrations in young adult depressives were less intense than those which become exacerbated as a function of advanced age. By contrast to prepubertal children or postadolescent young adults who are depressed, elderly accumulate: (i) lower total sleep times, (ii) less proportions of stages 3-4 and (iii) remain awake longer. It is concluded that sleep-polygraphic anomalies in postadolescent depression are an attenuated form of the REM-NREM cycle perturabation endemic to affective disease occurring with advanced age or senescence.

Regulation of circadian rhythmicity

Daily rhythms in many behavioral, physiological, and biochemical functions are generated by endogenous oscillators that function as internal 24-hour clocks. Under natural conditions, these oscillators are synchronized to the daily environmental cycle of light and darkness. Recent advances in locating circadian pacemakers in the brain and in establishing model systems promise to shed light on the cellular and biochemical mechanisms involved in the generation and regulation of circadian rhythms.

Diurnal variation of urinary MHPG in unipolar and bipolar depressives

In eight bipolar depressives, 11 unipolar depressives, and 15 healthy controls urinary excretion of MHPG was measured at 3-h intervals over one 24-h period. Bipolars excreted smaller amounts of MHPG than unipolars and controls, especially at night. MHPG excretion was significantly dependent on time of day in the control group only. In the patients maximum excretion showed a tendency to occur earlier in the day than in controls. Minima were unaffected. There were indications that tricyclic antidepressants advance MHPG phases.

DSIP/DSIP-P and circadian motor activity of rats under continuous light

Daily intravenous injection of 30 nmol/kg DSIP (delta sleep-inducing peptide) in rats under constant illumination produced marked changes of their motor activity as compared to saline controls. Similar marked but distinctly different effects on the circadian pattern of locomotor behavior partially abolished by constant illumination were also obtained after repeated administration of 0.1 nmol/kg DSIP-P (the phosphorylated analogue of DSIP) which enhanced overall motor activity. In both instances the results additionally differed from those reported for a normal 12 hr light:dark cycle. The present results support the hypothesis that DSIP might primarily act by influencing circadian rhythmicity.

Clorgyline delays the phase-position of circadian neurotransmitter receptor rhythms

The number of alpha- and beta-adrenergic, muscarinic cholinergic, opiate, and benzodiazepine receptors in rat forebrain, and dopamine and benzodiazepine receptors in striatum, change throughout the day. The diurnal rhythms of these receptors were altered by treatment with the monoamine-oxidase inhibitor clorgyline: following treatment some or all rhythm characteristics of wave form, amplitude, 24-h mean, and phase, were affected. One common effect of treatment was a delay in phase-position of binding to alpha- and beta-adrenergic, opiate and benzodiazepine receptors. Additionally, the nocturnal elevation in pineal melatonin which normally returns to baseline at light onset, persisted 3 h into the light period after clorgyline administration. These biochemical observations extend behavioural findings that clorgyline can delay the phase-position of rodent nocturnal activity onset, and does so by slowing the central circadian pacemaker.

Psychiatric morbidity and time zone changes: a study of patients from Heathrow airport

In a two-year period, 186 patients were admitted from Heathrow Airport to the nearest psychiatric hospital. Affective illness was related to time zone change. Depression was diagnosed significantly more often on flights from east to west (P less than 0.012 east to west versus west to east; P less than 0.015 north to south combined with south to north versus east to west, Fisher's exact probability test, two tailed). Hypomania was inversely related to depression in an east to west comparison (P less than 0.025). No other associations with direction of travel were seen in other diagnoses. Ninety-three (50 per cent) were diagnosed as schizophrenic; 24 of these had been aimlessly wandering. Twenty patients had been admitted at least once before under similar circumstances. Schizophrenic patients from Heathrow constituted 20 per cent of the total number of schizophrenic patients admitted to the hospital during that period.

Clinical effects of sleep deprivation and clomipramine in endogenous depression

Three groups of ten hospitalized endogenously depressed patients were treated with clomipramine alone (CL), a combination of sleep deprivation and clomipramine (CL/SD), and a combination of sleep deprivation and placebo (PL/SD), respectively. The sleep deprivation (SD) regime consisted of two deprivations per week. Measurements by self, blind and nurse ratings were taken twice daily during the experimental period of 15 days. Self and blind ratings indicate that, on average, over the entire experimental period, CL/SD was the most effective treatment in alleviating depression. The results of the PL/SD treatment were not significantly different from those of CL. In contrast, the nurse ratings did not differentiate significantly between the CL/SD and CL groups. Hamilton depression ratings indicated significantly less depression after one week of treatment with CL/SD relative to the treatment with CL alone. After the second week, however, no such differential clinical gain was found for the CL/SD group. It is concluded that SD in combination with clomipramine as compared with CL alone induces a faster onset of improvement in depression.

Weather and infradian rhythms in self-reports of health, sleep and mood measures

Some individuals exhibit significant and sustained periodicities in their self-reports of physical well-being, mood, hours of sleep, anxiety and cognition. These infradian rhythms may be related to weather variables such as solar flux, barometric pressure, temperature and humidity. The time of year, or season, during which self-reporting is performed may predicate the infradian rhythms and their relationship to weather.

The dependence of onset and duration of sleep on th circadian rhythm of rectal temperature

The sleep-wake cycle and the circadian rhythm of rectal temperature were recorded in subjects who lived singly in an isolation unit. In 10 subjects, the freerunning rhythms remained internally synchronized, 10 other subjects showed internal desynchronization. Times of onset and end of bedrest ("sleep") were determined in each cycle and referred to the phase of the temperature rhythm. In the synchronized subjects, onset of sleep occurred, on the average, 1.34 h before the minimum of temperature, and end of sleep 6.94 h thereafter, with narrow distributions. The desynchronized subjects had a broad bimodal distribution of sleep onsets (peaks 6.3 and 1.3 h before the minimum); the duration of sleep varied between more than 15 h when sleep began about 10 h before the temperature minimum, and less than 4 h when sleep began several hours after the minimum. The dependence of sleep duration on body temperature is interpreted as a continuing action of the coupling forces between the two rhythms after mutual synchronization is lost.

REM sleep, naps, and depression

Continued interest in rapid eye movement (REM) sleep abnormalities in depression stimulated comparative studies on daytime naps versus nighttime sleep. In a group of 15 depressed patients, REM latencies in morning and afternoon naps were similar to the shortened REM onset at night. Although REM latency did not vary across the three times, the propensity for REM sleep appeared to be greater in the morning nap than in the afternoon nap and the early portion of nocturnal sleep. Finally, the data suggest that responders to tricyclic treatment tend to be poor sleepers during daytime naps.

Lithium promotes entrainment of rats to long circadian light-dark cycles

Manic-depressive patients may have an endogenous circadian oscillator (about 24-hour clock) which runs too fast or which seeks too advanced or early a phase. Lithium salts are a major treatment for manic-depressives and may work by slowing this overly fast clock. Previous experiments with blinded rats demonstrated that lithium could delay free-running circadian rhythms. In this experiment, male rats were exposed to 27- and 28-hour light-dark cycles which were too long to synchronize the rhythms of control animals. In these control animals, the interaction of a faster (about 24-hour) internal rhythm with a slower light-dark cycle produced a beating interaction. cyclic variations in activity were observed as a result. Measurement of wheel-running activity indicated that its circadian rhythm was significantly slowed in lithium-fed animals and became synchronized with the light-dark cycle. This illustrates and supports the hypothesis that an action of lithium may be to delay and resynchronize overly fast circadian rhythms.

Animal models and human depressive disorders

Clinical depressive disorders are complex in presentation, dissimilar in origins and course, and often pleomorphic in character. An adequate understanding of their origins, biological substrates, and amenability to established and novel forms of therapy demands biological and social interventions which cannot always readily or ethically be carried out in a clinical setting. One useful complementary approach to clinical research utilizes preclinical models for laboratory investigations in parallel. The present paper reviews current approaches to modelling depression using animals, with particular emphasis upon phylogenetic constraints, systematic validity and reliability, and nosological limitations. Preclinical models are useful and necessary adjuncts for adequately understanding depression in humans. However, their utility remains a direct function of a continuing dialogue between clinical and laboratory research, and demands scrupulous observation and methodological rigor on the part of both clinicians and experimental researchers.

Maintenance of therapeutic effects of total sleep deprivation by limitation of subsequent sleep. A pilot study

In a previous study it was shown that in endogenous depression repeated sleep deprivation (SD) during clomipramine treatment resulted in rapid improvement after each SD. However, except after the first night, relapses were observed after all sleep nights following SD (the "recovery nights"). The present pilot study had a therapeutic aim, namely to prevent these relapses. We were interested in whether limitation of sleep during the recovery nights might prevent these relapses and whether different amounts of sleep would have different effects on the course of mood after recovery sleep. Ten endogenously depressed patients were treated with clomipramine and with three SD's. Five patients slept for approximately 2 h, and the other five for about 5 h during the second and third recovery nights. On average, no relapse was shown by these patients and, in fact, there was some additional improvement after these nights, in contrast to the findings in the previous study, where patients were allowed unlimited sleep.

Specific norepinephrine and serotonin uptake inhibitors in man: a crossover study with pharmacokinetic, biochemical, neuroendocrine and behavioral parameters

Eight depressed patients with major affective illness were treated with both zimelidine, a selective serotonin-uptake inhibitor, and with desipramine, a selective norepinephrine uptake inhibitor, following a double-blind crossover design. At steady-state the active metabolite of zimelidine, norzimelidine, predominated in the CSF by a factor of 7 to 1 over parent drug. As predicted, even high concentrations of norzimelidine were not associated with decreased 3-methoxy-4-hydroxy-phenylglycol (MHPG) in the CSF. In the same individuals, desipramine concentrations were highly correlated with decreases of MHPG in the CSF. Despite specific effects on monoamine neurotransmitter systems which have been implicated in the control of neuroendocrine secretion, neither drug had consistent effects on plasma cortisol, luteinizing hormone, growth hormone or prolactin. Both drugs had a marked and unexpected common effect on the 24-hour rest-activity cycle. The excess activity during the normal rest period (0--700 hr.) which has been noted in severely depressed individuals was significantly reduced by both the serotonergic zimelidine and the noradrenergic desipramine. These findings suggest that effects on the rest-activity pattern may be a common pathway for antidepressant effect.

Nocturnal electrographic features of frequently changing-irregular sleep-wakefulness rhythms

Polygraphic characteristics of nocturnal sleep associated with frequently changing-irregular rest-activity schedules were investigated in healthy young adults. Two groups each of 12 male university students were classified according to a priori criteria as either: (a) controls who slept regularly for 7--8 hr at night or (b) whose retiring and arising times combined varied chronically +/- 1.5 hr. Sleep was recorded during three consecutive 8-hr nocturnal periods at fixed clock times. Polygraphic indices generally reflected greater discontinuity and fragmentation associated with the nocturnal sleep in the young adults whose 24-hr rest-activity cycle tended to be frequently changing-irregular. The significantly: (a) larger absolute quantities of (i) transitional stage 1 sleep, (ii) intermittent wakefulness and (b) increases stage shifts provided some indication that the intrasleep cycle becomes disturbed when rest-activity schedules follow no predictable pattern in the everyday environment. Despite, or because of, the enforced hour (11:30 p.m +/- 30 min) for retiring, it is possible that the capacity to fall asleep had become phase-delayed among subjects with irregular rest-activity schedules who experienced more initial wakefulness (on average) before sleep onset stage 1. Finally, the recorded sleep perturbations associated with frequently changing-irregular schedules were extremely variable across nights and among individuals. This was especially pronounced on a nightly basis as reflected by significantly larger variability (SDs): (a) in the latency to sleep onset, for (b) total time asleep, (c) intermittent wakefulness, and (d) the ultradian (90-min) REM cycle. Variability (the SD) between individuals was also more substantial for these same polygraphic measures at statistically significant levels.