We are accustomed to think of sleep disturbances in mania and depression as symptoms; however, there is increasing evidence that sleep disturbances can be causes of these disorders, and that manipulation of sleep can be used as treatment. ? dramatic improvement in depression can be caused by total sleep deprivation, partial sleep deprivation in the second half of the night, and shifting the timing of sleep several hours earlier than usual. Sometimes, these sleep manipulations can also induce mania. Conversely, in patients who have improved after total sleep deprivation, sleep can eause a relapse into depression. These findings imply that changes in the timing and duration of sleep that occur during the natural course of affective illness1 may play a role in the pathogenesis of mania and depression. Furthermore, knowledge of the effects of sleep and sleep perturbations on affective state may be used by patients and their doctors to treat and to prevent affective episodes.
SLEEP DISRUPTION: TREATMENT FOR DEPRESSION
About 60% of patients with major depression experience a temporary remission if they are kept awake for one night.2,3 This effect of sleep deprivation is remarkable to observe. Patients who have been mired in deep depressions for months can remit completely within a few hours. Typically, when patients on our research ward are asked to deprive themselves of sleep, they protest, believing that lack of sleep is one of their most vexing problems and that sleep deprivation will only make them feel worse. Their negative expectations make their condition on the day following a night o\' sleep deprivation all the more striking. They are cheerful, talkative, and anxious to recount the experiences of the previous night. Often, they have spent the latter part of the night talking enthusiastically to the nursing staff, bustling about the ward doing chores, writing letters, and making plans for the future (Figure). It is strange for the patient to realize that during his depressive episode, when he went to sleep each night, he was but a few hours away from a complete remission, had he remained awake.
Sometimes, remissions induced by sleep deprivation are long lasting. Where this is the case, patients have sometimes chosen to use sleep deprivation as a self-treatment. More often, though, patients relapse after recovery sleep. For this reason, one night's total sleep deprivation has not proven to be useful in the treatment of most eases of depression. Modifications ?? the sleep deprivation procedure and the use of sleep deprivation in combination with drugs show promise as antidepressant treatments.
It is not necessary for a depressed patient to remain awake for the entire night to experience improvement after sleep deprivation. Pattuii sleep deprivation, permitting as much as four hours of sleep, can be as effective as total sleep deprivation.4 Also, the timing of sleeping and waking may be a critical factor. Partial sleep deprivation in the second half of the night is more effective than in the first half of the night.5 Furthermore, shifting the timing of sleep several hours earlier than usual (eg, 5 PM to 1 ?µ instead of 1 1 PM to 7 am) without sleep deprivation results in improvement in some patients.0 Thus, it appears that being awake in the latter half of the night (after I or 2 am), regardless of the amount of sleep obtained, is necessary for the antidepressant response. Unlike total sleep deprivation, partial sleep deprivation in the second half of the night or phase advance of the sleep period can be used for more than one night in succession. Therefore, it is feasible to use the procedures as ongoing treatments, and some patients have been able to maintain remissions for extended periods with repeated sleep manipulations.7
Wrist Motor Activity in Response to Sleep Deprivation
These approaches to treatment, which are still experimental, may prove useful in the management of depressed patients who do not wish to take, cannot tolerate, or do not respond to antidepressant medications, and who are motivated and interested in a drug- free approach. For example, a physician, with a hypomanic temperament and a family history of bipolar illness, became depressed at 40 years of age. After six months, he sought treatment, and after hearing about all the possible treatments, expressed interest in a nonpharmacologica! approach using sleep manipulations. After one night of partial sleep deprivation (sleep from 10 PM to 2 am), he experienced a complete remission of symptoms which he described as "miraculous." He continued to carry out the partial sleep deprivations for three or four nights a week, for several weeks, and maintained his improvement.
Another approach has been to use sleep manipulations in combination with pharmacological treatments of depression. Baxter et al have recently reported that lithium treatment, when initiated on the day after one night's total sleep deprivation, may help to sustain the antidepressant effects of the procedure after the patient returns to a normal sleep schedule.8 This could be thought of as an extension of lithium's capacity to prevent relapses when used prophylactically in remitted patients. Clomipramine, a tricyclic antidepressant, may have a similar effect when used in combination with sleep deprivation.9 Conversely, phase advance of the sleep period may potentiate antidepressant responses in patients who have not improved with medication treatments, according to Sack et al.10
SLEEP DISRUPTION: A CAUSE OF MANIA
There have been many reports that drugs and procedures used to treat depression can cause mania in bipolar patients." Sleep deprivation appears to be no exception in this regard (Figure). For example. Wehr et al deprived 12 depressed bipolar patients of sleep for one night and found that nine switched into mania or hypomania during the night or the following day.1 Three patients switched back into depression after recovery sleep, and six remained manic for days or weeks. Thus, induction of mania can be a side effect of sleep manipulations used to treat depression. When mania is mild, it may present no problem in clinical management, and may even be desirable, if it is not unduly prolonged. When mania is severe, it can be detrimental to the patient and may require aggressive treatment.
The capacity of sleep deprivation to induce mania in experimental conditions may help to explain how manic episodes begin during the natural course oí bipolar illness. A variety of psychological, environmental, pharmacological, behavioral, and medical factors have been reported to precipitate manic episodes.12 As discussed below, many of these factors disrupt sleep. Because sleep disruption can cause mania, it seems reasonable to propose that the sleep disruption caused by these factors actually mediates their mania-inducing effect.12 Some examples illustrate this hypothesis. There are many reports in the literature of bereavement manias, manias seemingly caused by the death of a loved one.12 Such manias seem paradoxical: the patient, instead of mourning, is elated and expansive. Such reactions have been interpreted in psychological terms as attempts to deny a painful loss. However, it is quite normal for acute grief to be accompanied by profound insomnia. In a bipolar patient, failure to sleep is very likely to induce mania. Thus, bereavement mania, according to this model, is not the product of a psychological defense mechanism, but is a predictable complication of insomnia, which in turn is a normal manifestation of grief.
Recently, there have been several reports that withdrawal of antidepressant medications can precipitate mania.12 On the face of it, this also seems paradoxical. However, insomnia is a common symptom of antidepressant withdrawal, and it might be responsible for the occurrence of mania in this situation.
Disruptions in routine due to role change, travel, various types of emergencies, etc, have been reported to induce mania, in some of these situations, such as overnight travel or sleeping in unfamiliar surroundings, sleep is disrupted, with predictable consequences for bipolar patients. For example, one evening a euthymic bipolar patient went to an emergency room because of a sore throat. She was kept waiting (and awake) until five o'clock in the morning. By then she had become so manic that involuntary hospitalization was necessary. The emergency room staff probably never realized that her mania may have been iatrogenic. One of our bipolar patients, while a student, stayed awake all night studying for a final exam and was hospitalized for acute mania shortly thereafter.
In the genesis of mania, psychological factors are sometimes confounded with sleep deprivation, as shown by the example of bereavement mania. A patient whose father had disappeared when she was a child learned of his whereabouts when she was a college student. He had moved to another city and remarried. She received word from his new wife that he was seriously ill and was invited to visit. When she arrived, she found that she liked her new-found stepmother and stayed awake most of the night talking to her about her father. The following day she became manic. One could formulate a psyehodynamic explanation for the occurrence of this mania, focusing on the reunion with the lost father, but sleep deprivation incidental to the patient's wish to talk to her stepmother may be a sufficient, if prosaic, explanation for the episode.
Mania and sleep loss may reinforce one another. Not only does sleep disruption sometimes cause mania, but mania also causes sleep disruption. This circular relationship could become self-reinforcing and escalate out of control, as appears to happen with mania in the clinical setting.
These observations about the impact of sleep and wakefulness on mania have important implications for the treatment and prevention of this condition. In a vulnerable patient, loss of sleep could serve as an early warning sign of possible impending mania. Mania could be treated with drugs and/or instructions that promote sleep. Mania might be prevented by educating patients and their families about the mania-inducing effects of sleep disruptions, and by intervening in situations that are likely to cause sleep deprivation or insomnia.
The mechanisms by which sleep perturbations improve depression and precipitate mania are unknown. The search for mechanisms is important for several reasons. First, the identification of mechanisms that mediate effects of sleep deprivation on mood is likely to increase our understanding of the pathogenesis of mania and depression. Second, knowledge of the biochemical substrates of such mechanisms could be used to develop new pharmacological treatments for these conditions. To the extent that such treatments could mimic or antagonize the clinical effects of sleep deprivation, they could be expected to act extremely rapidly incontrasi to existing drugs.
The search for mechanisms would be facilitated by a more precise definition of the conditions associated with sleep manipulations which are responsible for their antidepressant effects. As currently used, the procedures are like crude preparations of drugs which contain many extraneous elements. It may be that actual loss of sleep is not the active component of the sleep deprivation procedure. The procedure involves many changes in the patient's state and routine besides his level of consciousness. The patient is sitting or standing instead of lying, moving instead of resting, perhaps eating instead of fasting, engaging in social activities instead of withdrawing from them, and so on.
One consequence of the sleep deprivation procedure is that patients are exposed to light at night at a time when they would usually be sleeping in the dark. This is an important aspect of the procedure because light has recently been shown to have antidepressant effects in some depressed patients.13 Therefore, it must be determined whether exposure to light is a necessary condition, possibly even the effective principle, of sleep deprivation therapy. Our group conducted an experiment in which five patients were totally sleep deprived on two different occasions, once in very bright light and once in nearly total darkness.14 Patients experienced the same degree of improvement after both types of sleep deprivation, indicating that exposure to light at night is apparently not necessary for the antidepressant response to sleep deprivation. As discussed previously, there are some indications that the timing of sleep and wakefulness (ie, wakefulness in the second half of the night) is an important aspect of clinically effective sleep manipulation.
Another possibility is that the antidepressant effects of sleep deprivation depend on patients' being deprived of a particular stage of sleep. Vogel et al reported that selective REM sleep deprivation has antidepressant effects, while selective non-REM sleep deprivation does not.15 Differential effects on REM sleep might explain why partial sleep deprivation in the second half of the night is more effective than in the first half of the night. The amount of REM sleep that occurs at different times of the night is controlled by a process that exhibits a circadian rhythm.16 As a consequence of this rhythm the amount of REM sleep is highest late at night near dawn, and lowest early at night, near bedtime. Thus, sleep deprivation in the second half of the night deprives patients of REM sleep more than it does in the first half of the night.
Since sleep and wakefulness markedly alter secretory activity in several neuroendocrine systems, changes in these systems might mediate the effects of sleep deprivation on mood. Sleep deprivation stimulates Cortisol17 and thyrotropin (TSH)18 secretion and inhibits growth hormone19 and prolactin20 secretion. The stimulatory effects on hypothalamie-pituitary-thyroid (HPT) function are particularly interesting as possible mechanisms because HPT function has been reported to be deficient in some groups of patients21"23 and hormones of this axis may have antidepressant and mania-inducing effects.23
In attempting to discover the mechanism of action of sleep deprivation therapy, it would be helpful to know the normal biological function of sleep. Several different functions have been proposed. These include physiological functions, such as regeneration of tissues and chemical stores, and behavioral functions, such as avoidance of predators. Perhaps the most fruitful hypothesis is that sleep is an energy-conserving state like hibernation, estivation, and torpor.24 Many symptoms of mania and depression involve changes in physiology and behavior related to the acquisition, storage, and expenditure of energy. Besides sleep, these factors include appetite, eating behavior, weight, metabolism, activity and rest, exposure to and withdrawal from the environment, motivation and drive. If manicdepressive illness is a disturbance in the network of systems that regulates the energy balance of the organism, and if sleep has an energy-conserving function within this network, then sleep deprivation is a probe that touches the biological core of the illness.
Experimental investigations of the effects of sleep and wakefulness on manic-depressive illness show that patients' behavior and experience can regulate their clinical state and govern the course of their illness. These findings provide a model that may enable clinicians to understand, treat, and prevent the vicissitudes of their patients' illness, and they provide researchers with experimental paradigms (sleep deprivation and recovery sleep) that can be used to investigate the pathogenesis of affective states prospectively.
1. Wehr TA, Goodwin FK. Wirz-luslice A. et al: 48-hour sleep-wake cycles in manicdepressive illness. Arch Gen Psychiatry 1982; 59:559-565.
2. Pflug B. Tolle R: Disturbance of the 24hour rhythm in endogenous depression and the treatment ol endogenous depression by sleep deprivation. International Pharmacopsychiatry 1971; b: 187-196.
3. Gillin IC: The sleep therapies ol depression. Prog Neuropsychopharmacol liiol Psychiatry 1985: 7:351-364.
4. Schilgen B, Tölle R: Partial sleep deprivation as therapy for depression. Arch Gen Psychiatry 1980: 57:267-271.
5. Goetze U. Tölle R: Antidepressive Wirkung des partiellen Schlafcntzuges wahrend der 1. Hallte der Nacht. Psychiatr Clin (Hasel) 1981; 14:129-149.
6. Wehr TA, Wirz-|ustice A. Goodwin FK. el al: Phase advance of the circadian sleep-wake cycle as an antidepressant. Science 1 979 : 206: 710-715.
7. van Beinmcl AL. van den Hoofdakker RH: Maintenance of therapeutic effects of total sleep deprivation by limitation ot subsequent sleep. A pilot study. Acta I'sychiatr Scand 1981; 63:453-462.
8. Baxter LR: Can lithium carbonate prolong antidepressant effect oí sleep deprivation? Arch Gen Psychiatry 1985; 42:655.
9. Elsenga S. van den Hoofdakker RH: Clinical effects of sleep deprivation and clomipramine in endogenous depression. J Psychiatr Res 1985; 17:301-374.
10. Sack DA. Nurnburger J, Rosenthal NE. et al: The potentiation of antidepressant medications by phase-advance of lhe sleep-wake cycle. Am J Psychiatry 1985; 142:006-608.
11. Wehr TA. Goodwin FK: Can antidepressants precipitate mania and worsen the course of affective illness? Am J Psychiatry 1987; in press.
12. Wehr TA. Sack DA. Rosenthal NE: Sleep reduction as a linai common pathway in the genesis of mania. Am J Psychiatry 1987: 144:201-204.
13. Rosenthal NE, Sack DA. Carpenter CJ. et al: Antidepressant effects of light in seasonal affective disorder. Am J Psychiatry 1985: 142:163-170.
14. Wehr TA, Rosenthal NE, Sack DA. et al: Antidepressant effects of sleep deprivation in bright and dim light. Acta Psychiatr Scand 1985: 72:161-165.
15. Vogel GW. Vogel C. McAbee RS, et al: Improvement of depression by REM sleep deprivation. Arch Gen Psychiatry 1980: 57:247-253.
16. Weitzman ED, Nogeire C. Perlow M. et al: Effects of a prolonged 3-hour sleepwake cycle on sleep stage, plasma Cortisol, growth hormone and body temperature in man. J Clin Endocrinol Metab 1974; 38:1018-1070.
17. Weitzman ED, Czeisler CA. Zimmerman JC. et al: The sleep-wake pattern of Cortisol and growth hormone secretion during non-entrained (free-running) conditions in man. in Nijhoff M (ed): Circadian and Ultradian Variations of Pituitary Hormones in Man. Brussels. Elsevier. 1982.
18. Parker DC. Pekary AE. Hershman IM: Effect of normal and reversed sleep-wake cycles upon nyctohenteral rhythmicity of plasma thyrotropin: Evidence suggestive of an inhibitory influence in sleep. J Clin Endocrinol Metab 1976; 43:318-329.
19. Takahashi Y. Kipnis DM. Daughaday WH: Growth hormone secretion during sleep. J Clin Invest 1968; 47:2079-2090.
20. Parker DC. Rossman LG, Venderlan EF: Relation of sleep-entrained human prolactin release to REM-non-REM cycles. J Clin Endocrinol Metab 1974; 38:646-651.
21. Weeke A, Wecke J: Disturbed circadian variation of serum thyrotropin in patients with endogenous depression. Acta Psychiatr Stand 1978: 57:281-289.
22. Kjellman BF, Beckfriis J, Ljungren JG, et al: Twenty-four hour serum levels of TSH in affective disorders. Acta Psychiatr Stand 1984; 69:491-502.
23. Sack DA, James SP, Rosenthal NE, et al: Deficient nocturnal surge in TSH secretion during sleep and sleep deprivation in rapid cycling bipolar illness. Psychiatry Res 1987; in press.
24. Walker IM. Berger RJ: Sleep as an adaptation for energy conservation functionally related to hibernation and shallow torpor, in McConnell PS. Boer GJ, Romijn HJ. et al (eds): Progress in Brain Research. Amsterdam, Elsevier, vol 53, 1980.