Drug effects on REM sleep include responses to medications that decrease and increase REM sleep and the withdrawal of REM-suppressing drugs

Rapid–eye-movement (REM) sleep is a state characterized by fast, low-voltage brain waves; involuntary muscle movements; rapid eye movements; irregular heart rate and respiration; vivid dreams; and a higher threshold of arousal. It lasts from 5 to 20 minutes, occurs at intervals of about 90 minutes, and occupies approximately 20% of adult sleep time overall. REM sleep is also known as active sleep, desynchronized sleep, fast-wave sleep, and paradoxical sleep.1 REM sleep is called paradoxical sleep because the brain is metabolically, physiologically, and psychologically active.1 During REM sleep, the brain neither depends on external stimuli for its increased activity nor expresses a motor output. This cyclic central activation during REM sleep is important to the brain’s metabolic processes.1 REM sleep is defined by electroencephalographic (EEG) activation, muscle atonia, and episodic bursts of rapid eye movements. The mental activity of REM sleep is associated with dreaming, based on the ability to recall vivid dreams reported after approximately 80% of arousals from REM.1 Inhibition of spinal motor neurons via brainstem mechanisms mediates suppression of postural motor tonus in REM sleep. A shorthand definition of REM sleep, therefore, is a highly activated brain in a paralyzed body.1

REM-sleep Induction
The brainstem is involved in REM induction. Brain activity is partly mediated through aminergic and cholinergic neuromodulatory subsystems of projecting neurons. The aminergic system is based in the brainstem nuclei, locus ceruleus (norepinephrine), and raphe nuclei (serotonin). Axons project to areas of the cortex and thalamus where they are hypothesized to control brain-activity states. Cholinergic nuclei located at the pons-midbrain junction contribute to brainstem control of the cortex and thalamus in the cholinergic system (acetylcholine). The waking state is dominated by the aminergic system of neuromodulation. This activity exhibits an inhibitory effect on the cholinergic system. REM sleep is marked by cholinergic activation and aminergic suppression. Decrease of aminergic firings frees the cholinergic system from its restraints and possibly triggers REM sleep. Consequently, there is a disinhibition of neurons in the brainstem, cortex, and limbic structures.1

EEG wave patterns originate in the pontine reticular formation (brainstem).1 Electrical bursts in cholinergic pons neurons propagate to the lateral geniculate nuclei and then to the occipital cortex and oculomotor nuclei. These initial waves are known as pontogeniculo-occipital (PGO) waves.2 Aminergic neurons modulate inhibition of the oculomotor nuclei and occipital cortex during wakefulness state. As aminergic activity ceases, activated PGO cholinergic neurons stimulate these areas. REM patterns directly reflect PGO cholinergic discharge patterns. PGO waves may be the internal source of visual stimuli present during REM dreaming. They are postulated to inform the visual system about eye movements to promote brain development and facilitate response to novel environmental stimuli.2

The anatomic substrates responsible for the generation of REM sleep are located within the pons.3 The tonic atonia that accompanies REM sleep is generated at the level of the pons, but involves a critical relay at the medulla.4 The prime candidates for the cholinergic cells that facilitate REM sleep are located in the dorsolateral tegmentum, the lateral dorsal tegmental (LDT) nuclei, and the pedunculopontine tegmental (PPT) group.1,5 These cells display increased firing rates before and during REM sleep, especially in association with PGO spikes. The LDT and PPT send projections to the thalamus that may inhibit the mechanisms responsible for sleep spindles and for transmitting d EEG waves to the medial pontine reticular formation (where application of cholinomimetic agents triggers REM sleep) and to the medulla (where they facilitate the atonia of REM sleep).1,5 The norepinephrine and serotonin responsive neurons in the locus ceruleus and dorsal raphe, respectively, lie close to the cholinergic neurons in LDT and PPT and apparently can inhibit them, either directly or indirectly, via g-aminobutyric acid interneurons.1,5 Cholinergic neurons exert a widespread and crucial role in the orchestration of REM sleep through their projections to the medial pontine reticular formation, medulla, and forebrain areas in the thalamus and basal forebrain.1,5

The reciprocal-interaction model of non-REM and REM sleep regulation suggests that the cycling, alternating pattern of non-REM and REM sleep is under the control of noradrenergic/serotonergic and cholinergic neuronal networks.6 Cholinomimetics like physostigmine and galanthamine provoked an earlier onset of REM sleep, whereas subchronic treatment with scopolamine, a cholinergic antagonist, only led to a heightening of REM density.6 Simultaneous administration of noradrenergic antagonists with a cholinergic agonist did not provoke a more pronounced REM-sleep advance.6 Comparative studies6,7 of the cholinergic agonist in patients with depression and schizophrenia revealed the most pronounced REM-sleep response in the depressed group. The REM-sleep response to cholinergic stimulation in depression did not, however, predict the treatment response to a different therapeutic strategy.6

f07a.jpg (29677 bytes)
Table 1. Interaction between serotonergic, noradrenergic, and acetylcholinergic neurons in the control of rapid–eye-movement sleep.

Memory and REM
Are memories processed or consolidated during REM sleep? There is a conflicting animal study8 of REM-sleep deprivation’s disruption of learning and memory. The three major classes of antidepressant drugs, monoamine oxidase inhibitors (MAOI), tricyclic antidepressants (TCA), and selective serotonin-reuptake inhibitors (SSRI), profoundly suppress REM sleep. The MAOIs virtually abolish REM sleep, and the TCAs and SSRIs have been shown to produce immediate reductions of 40% to 85% and sustained reductions of 30% to 50% in REM sleep.8 Despite marked suppression of REM sleep, these classes of antidepressants do not, on the whole, disrupt learning or memory.8 Recent functional imaging studies in humans have revealed patterns of brain activity in REM sleep that are consistent with dream processes, but not with memory consolidation.8 The authors proposed that the primary function of REM sleep is to provide periodic endogenous stimulation to the brain, which serves to maintain requisite levels of central nervous system activity throughout sleep.8 REM is the mechanism used by the brain to promote recovery from sleep. Cumulative evidence indicates that REM sleep serves no role in the processing or consolidation of memory.8

Polysomnographic recordings of depressed patients often reveal reduced slow-wave sleep, an early onset of the first episode of REM sleep, and increased phasic REM sleep.9 A deficit in serotonergic neurotransmission, a relative increase in pontine cholinergic activity, and, perhaps, an excess of nor-

adrenergic and corticotropin-releasing–hormone activity have been implicated in the pathogenesis of severe depression.9,10

The reduction in REM sleep produced by antidepressants11 may be an important part of their mechanism of action; however, the ability of new antidepressant compounds (such as nefazodone and moclobemide) to increase REM sleep throws doubt on this suggestion.11 The effects of antidepressants on slow-wave sleep are quite diverse; in general, TCAs increase slow-wave sleep, whereas SSRIs and MAOIs either reduce slow-wave sleep or produce no change.11 Sleep continuity is improved acutely following administration of antidepressants with sedating properties, such as certain TCAs, trazodone, and mianserin.11 Some nonsedating drugs (ritanserin and nefazodone) also improve sleep continuity, possibly through serotonin-receptor blockade.11

The primary changes seen in major depression in sleep architecture include shortened REM-sleep onset latency, increased REM density, reduced total sleep time, reduced sleep efficiency, increased awakenings, decreased slow-wave sleep, and a shift of slow-wave sleep from the first non-REM cycle to the second. Clinical response is better when an immediate, antidepressant-induced prolongation of REM-sleep latency, reduction in total REM-sleep time, and reduction in REM density are observed.12

Several types of pharmacological agents affect REM sleep through reciprocal inhibition and interaction. Some cholinomimetic agents and some aminergic antagonists augment REM sleep, whereas anticholinergic agents and aminergic agonists often attenuate or block it.1

Cholinergic agonists such as carbachol, bethanechol and neostigmine (a cholinesterase inhibitor) induce REM sleep.13 The administration of pharmacological agents antagonizing noradrenergic or serotonergic neurotransmission increases the occurrence of PGO waves, independently from REM sleep. Cholinomimetic administration increases the occurrence of PGO waves, as well as other components of REM sleep.2

Acetylcholine-receptor Subtypes
Scopolamine (a muscarinic antagonist) inhibits REM sleep in humans.14 REM sleep rebound has occurred when scopolamine was abruptly discontinued.

REM-sleep latency is short in depressed individuals. This may be related to upregulation of muscarinic neurotransmission or to diminished aminergic neurotransmission. Indeed, the cholinergic-aminergic imbalance hypothesis of depression suggests that depression results from an increased ratio of cholinergic to aminergic neurotransmission.10

Drugs interact with several types of cholinergic receptors that are generally classified as either nicotinic or muscarinic. The muscarinic receptors are subdivided into M1, M2, M3, M4, and M5 categories. M2-receptor agonists such as cisdioxolane and oxotremorine-M lead to rapid induction of REM sleep and an increased percentage of REM sleep, whereas M1 agonists have no effect on REM sleep.15 Hence, there is a cholinergic receptor specificity in the generation of REM sleep.

In human narcolepsy, REM-sleep latency is very short. An increased density of muscarinic receptors in the brain stem has been seen in animal studies16,17 of narcolepsy.

PGO spike generation is partially mediated by a nicotinic mechanism of cortical activation and measured as a desynchronized EEG, which represents an M1 component, and the muscle atonia seen during REM sleep is mediated by an M2 receptor.18

Noradrenergic Compounds
Many cells of the locus ceruleus cease firing during the transition from non-REM sleep to REM sleep (or during REM sleep). They regain their activity once REM sleep has terminated. The effects on norepinephrine are directly inhibitory. Drugs that decrease norepinephrine content or release also increase the amount of REM sleep, and compounds that increase norepinephrine activity decrease REM sleep.19

Amphetamine releases norepinephrine from nerve terminals and inhibits its reuptake. D-amphetamine decreases the percentage of REM sleep and the length of total sleep. Long-term treatment with amphetamines leads to a large REM-sleep rebound.20 During the first several weeks after withdrawal from cocaine or amphetamines there is hypersomnia, short REM-sleep latency, and increased amounts and percentage of REM sleep.21

Reserpine depletes tissue of norepinephrine, dopamine, and serotonin by preventing their entry into the vesicles whence they are released, thus increasing the amount of REM sleep and shortening its latency.22 Aminergic neurotransmitters inhibit REM sleep.

Norepinephrine systems are inhibitory to the generation of PGO spikes.23 Clonidine reduces PGO-spike activity.23 Norepinephrine-releasing cells in the locus ceruleus cease firing in association with PGO spikes.

Narcolepsy is characterized by excessive daytime sleepiness, cataplexy, sleep paralysis, and hypnogogic hallucinations (intrusion of REM sleep into the wakeful state). The norepinephrine and serotonergic systems appear to be involved because antidepressants, which potentiate both systems by blocking reuptake of amines, are effective in treating various symptoms of narcolepsy.24

Prazosin (an a1-adrenergic blocker) precipitates cataplexy, whereas clonidine, an a2-adrenergic agonist, decreases cataplectic attacks.23,24 Norepinephrine acts to suppress cataplexy.25

Adrenergic Receptor Subtypes
Adrenergic receptors are subdivided into a and b types. These receptors are further subclassified as a1, a2, b1, and b2 receptors.23 a1 Receptors are mainly found postsynaptically to the norepinephrine terminal, whereas a2 receptors are preferentially found in the terminal itself, and usually function to decrease the release of norepinephrine. Activation of a1 receptors should lead to positive norepinephrine effects, while activation of a2 receptors decreases the amount of norepinephrine released into the synaptic cleft and acts as an antagonist of norepinephrine actions. The a1 antagonist prazosin increases REM sleep.26 Intravenous administration of thymoxamine, an a1 receptor antagonist, increases REM sleep in humans.27 Activation of a1 receptors leads to increased wakefulness and a decrease in both non-REM and REM sleep. The a2 receptor agonist clonidine causes a modest decrease in REM sleep in humans.28 Yohimbine, an a2 antagonist, does decrease REM sleep in animal studies26,29 by increasing release of norepinephrine. b-Selective drugs such as propranolol have produced inconsistent (or absent) effects on REM sleep.26

Serotonergic Compounds
Serotonin has been proposed to be inhibitory to the state of REM sleep. Lysergic acid diethylamide (LSD) tends to increase REM sleep in humans,30 possibly because it shuts down serotonergic neurons, at least in the dorsal raphe of laboratory animals.31 Serotonin agonists such as fenfluramine, which releases serotonin, and fluoxetine, which is a potent reuptake blocker, decrease both total and REM sleep.32

Venlafaxine is an effective antidepressant. It has been approved for the treatment of generalized anxiety disorder.33 Venlafaxine was initially characterized as an inhibitor of both serotonin and norepinephrine uptake and was termed a dual-uptake inhibitor. Venlafaxine selectively inhibits serotonin uptake at low therapeutic doses and inhibits both serotonin and norepinephrine uptake at higher therapeutic doses.33

Neuropeptides and Other Possible Modulators
Vasoactive intestinal polypeptide increases REM sleep in animals, as does somatostatin.34 Reportedly, somatostatin overcomes the REM–sleep-suppressing effect of scopolamine.34

Drugs producing REM-sleep deprivation by arousal improve endogenous depression, according to Vogel et al.35 The arousal REM-sleep deprivation was large, persisted for weeks, and was followed by a REM rebound. The investigators found that all drugs producing arousal-type REM-sleep deprivation improved endogenous depression.35

Arousal-type REM sleep deprivation was produced by eight first-generation antidepressants (amitriptyline, clomipramine, clorgyline, desipramine, doxepin, imipramine, pargyline, and phenelzine); two first-generation antidepressants (iprindole and trimipramine) did not produce arousal-type REM sleep deprivation.35 Four second-generation antidepressants (amoxapine, butriptyline, viloxazine, and zimeldine) produced arousal-type REM-sleep deprivation and two (amineptine and trazodone) did not. Not all efficacious antidepressants produce arousal-type REM sleep deprivation.35 Addictive drugs (alcohol, amphetamines, barbiturates, other nonbenzodiazepine hypnotics, cocaine, and narcotics) did not produce arousal-type REM-sleep deprivation.35

Arousal-type REM-sleep deprivation is an antidepressant process through which drugs improve endogenous depression; however, arousal-type REM-sleep deprivation cannot be the only mechanism of action of all antidepressant drugs.35

Drug effects on REM sleep, in summary, include responses to drugs that decrease REM sleep (ethanol, TCAs, trazodone, SSRIs, MAOIs, lithium, amphetamines, methylphenidate, and clonidine); responses to drugs that increase REM sleep (nefazadone and reserpine); and responses to the withdrawal of REM-suppressing drugs.

Taj M. Jiva, MD, is clinical assistant professor of medicine, State University of New York at Buffalo, and a pulmonologist, intensivist, and sleep specialist at Buffalo Medical Group PC, NY.

1. Jones BE. Paradoxical sleep and its chemical/structural substrates in the brain. Neuroscience. 1991;40:637-656.
2. Callaway CW, Lydic R, Baghdoyan HA, Hobson JA. Pontogeniculooccipital waves: spontaneous visual system activity during rapid eye movement sleep. Cell Mol Neurobiol. 1987;7:105-149.
3. Jouvet M. The role of monoamines and acetylcholine containing neurons in the regulation of the sleep-waking cycle. Ergebnisse der Physiologie. 1972;14:166-308.
4. Magoun HW, Rhines R. An inhibitory mechanism in the bulbar reticular formation. J Neurophysiol. 1946;9:165-171.
5. Shiromani PJ, Armstrong DM, Gillin JC. Cholinergic neurons from the dorsalateral pons project to the medial pons: a WGA-HRP and choline acetyltransferase immunohistochemical study. Neurosci Lett. 1988;95:19-23.
6. Reimann D, Hohagen F, Fritsch-Montero R, et al. Cholinergic and noradrenergic neurotransmission: impact on REM sleep regulation in healthy subjects and depressed patients. Acta Psychiatrica Belgica. 1992;92:151-171.
7. Shiromani P, Gillin JC, Henriksen SJ. Acetylcholine and the regulation of REM sleep: basic mechanisms and clinical implications for affective illness and narcolepsy. Ann Rev Pharmacol Toxicol. 1987;27:137-156.
8. Vertes RP, Eastman KE. The case against memory consolidation in REM sleep. Behav Brain Sci. 2000;23:867-876.
9. Thase ME. Depression, sleep and antidepressants. J Clin Psychiatry. 1998;59:55-65.
10. Janowsky DS, El-Yousef MK, Davis JM. A cholinergic-adrenergic hypothesis of mania and depression. Lancet. 1972;II:632-635.
11. Sharpley AL, Cowen PJ. Effect of pharmacologic treatments on the sleep of depressed patients. Biol Psychiatry. 1995;37:85-98.
12. Gillin JC, Van Kammen DP, Graves J, Murphy D. Differential effects of D- and L- amphetamine on the sleep of depressed patients. Life Sci. 1975;17:1233-1240.
13. Hernandez-Peon R, Chavez-Ibarra G, Morgane PJ, et al. Limbic cholinergic pathways involved in sleep and emotional behavior. Exp Neurol. 1963;8:93-111.
14. Gillin JC, Sutton L, Ruiz C, et al. The effects of scopolamine on sleep and mood in depressed patients with a history of alcoholism and a normal comparison group. Biol Psychiatry. 1991;30:157-169.
15. Velazquez-Moctezuma J, Gillin JC, Shiromani PJ. Effects of specific M1, M2 muscarinic receptor agonists on REM sleep generation. Brain Res. 1989;503:128-131.
16. Kilduff TS, Bowersox SS, Kaitin KI, Baker TL, Ciaranello RD, Dement WC. Muscarinic cholinergic receptors and the canine model of narcolepsy. Sleep. 1986;9:102-106.
17. Faull KF, Zeller-DeAmicis LC, Radde L, et al. Biogenic amine concentrations in the brains of normal and narcoleptic canines: current status. Sleep. 1986;9:107-110.
18. Velazquez-Moctezuma J, Shalauta MD, Gillin JC, Shiromani PJ. Differential effects of cholinergic antagonists on REM sleep components. Psychopharmacol Bull. 1990;26:349-353.
19. Siegel JM, Rogawski MA. A function for REM sleep: regulation of noradrenergic receptor sensitivity. Brain Res Rev. 1988;13:213-233.
20. Rechtschaffen A, Maron L. The effect of amphetamine on the sleep cycle. Electroencephalogr Clin Neurophysiol. 1964;16:438-445.
21. Watson R, Hartmann E, Schildkraut JJ. Amphetamine withdrawal: affective state sleep patterns and MHPG excretion. Am J Psychiatry. 1972;129:39-45.
22. Coulter JD, Lester BK. Reserpine and sleep. Psychopharmacology. 1971;19:134-147.
23. Depoorte H. Adrenergic agonists and antagonists and sleep-wakefulness stages. In: Wauquier A, Gaillard JM, Monti JM, et al, eds. Sleep: Neurotransmitters and Neuromodulators. New York: Raven Press; 1985:79-92.
24. Mignot E, Guilleminault C, Bowersox S, Rappaport A, Dement WC. Role of central a-1 adrenoceptors in canine narcolepsy. J Clin Invest. 1988;82:885-894.
25. Aldrich MS, Rogers AE. Exacerbation of human cataplexy by prazosin. Sleep. 1989;12:254-256.
26. Hilakivi I, Leppavouri A. Effects of methoxamine, an a-1 adrenoceptor agonist, and prazosin, an a-1 antagonist, on the stages of the sleep-wake cycle in the rat. Acta Physiol Scand. 1984;120:363-372.
27. Oswald I, Adam K, Allen S, et al. a-Adrenergic blocker thymoxamine and mesoridazine both increase human REM sleep duration. Sleep Research. 1974;3:62.
28. Nicholson AN, Pascoe PA. Presynaptic a-2 adrenoceptor function and sleep in man: studies with clonidine and idazoxan. Neuropharmacology. 1991;30:367-372.
29. Hilakivi I. The role of b- and a-adrenoreceptors in the regulation of the stages of the sleep-waking cycle in the cat. Brain Res. 1983;227:109-118.
30. Muzio JN, Roffwarg HP, Kaufman E. Alterations in the nocturnal sleep cycle resulting from LSD. Electroencephalogr Clin Neurophysiol. 1966;21:313-324.
31. Haigler HJ, Aghajanian GK. Lysergic acid diethylamide and serotonin: a comparison of effects on serotonergic neurons and neurons receiving a serotonergic input. J Pharmacol Exp Ther. 1974;188:688-699.
32. Shiromani P, Gillin JC, Henriksen SJ. Acetylcholine and the regulation of REM sleep: basic mechanisms and clinical implications for affective illness and narcolepsy. Annu Rev Pharmacol Toxicol. 1987;27:137-156.
33. Roseboom PH, Kalin NH. Neuropharmacology of venlafaxine. Depress Anxiety. 2000;12:S20-S29.
34. Danguir J, De Saint-Hilaire-Kafi S. Scopolamine-induced suppression of paradoxical sleep is reversed by the somatostatin analogue SMS 201-995 in rats. Pharmacol Biochem Behav. 1988;30:295-297.
35. Vogel GW, Buffenstein A, Minter K, Hennessey A. Drug effects on REM sleep and on endogenous depression. Neurosci Biobehav Rev. 1990;14:49-63.