Inducible and neuronal nitric oxide synthases (NOS) have complementary roles in recovery sleep induction

Sleep homeostasis is the process by which recovery sleep is generated by prolonged wakefulness. The molecular mechanisms underlying this important phenomenon are poorly understood. We have previously shown that nitric oxide (NO) generation increases in the basal forebrain (BF) during sleep deprivation (SD). Moreover, both NO synthase (NOS) inhibition and a NO scavenger prevented recovery sleep induction, while administration of a NO donor during the spontaneous sleep-wake cycle increased sleep, indicating that NO is necessary and sufficient for the induction of recovery sleep. Next we wanted to know which NOS isoform is involved in the production of recovery sleep. Using in vivo microdialysis we infused specific inhibitors of NOS into the BF of rats during SD, and found that an inhibitor of inducible NOS (iNOS), 1400W, prevented non-rapid eye movement (NREM) recovery, while an inhibitor of neuronal NOS (nNOS), L-N-propyl-arginine, decreased REM recovery but did not affect NREM recovery. Using immunoblot analysis we found that iNOS was not expressed during the spontaneous sleep-wake cycle, but was induced by prolonged wakefulness (increased by 278%). A known iNOS inducer, lipopolysaccharide, evoked an increase in sleep that closely resembled recovery sleep, and its effects were abolished by 1400W. These results suggest that the elevation of NO produced by induction of iNOS in the BF during prolonged wakefulness is a specific mechanism for producing NREM recovery sleep and that the two NOS isoforms have a complementary role in NREM and REM recovery induction.     

Sleep-related neuronal discharge in the basal forebrain of cats

Although evidence suggests that the basal forebrain contains a hypnogenic mechanism, putative sleep-promoting neural elements within this area have not been identified. We examined basal forebrain neuronal activity during waking, non-rapid-eye-movement (NREM) sleep, REM sleep and various transition states. Based on state-related discharge rates. 3 cell types were defined. Thirty-nine of 83 cells were classified as waking-active, i.e. waking discharge rates were greater than 2 times NREM sleep rates. Twenty-three of 82 cells were classified as state-indifferent (waking and NREM rates differed by a factor of less than 2). NREM sleep discharge rates of the remaining 20 cells were greater than 2 times waking rates. These were labeled sleep-active cells. Discharge rates of these cells during epochs of alert waking were low, averaging less than 1 spike/s. Maximal discharge rates occurred during NREM sleep, averaging 9.44 spikes/s. Increased discharge of sleep-active cells anticipated sleep onset; cells had an average discharge rate of 6.60 spikes/s during transitions between waking and NREM sleep. Sleep-active cells were confined to the ventral basal forebrain, in the horizontal limb of the diagonal bands of Broca, substantia innominata, entopeduncular nucleus and ventral globus pallidus. These areas overlap, in part, with those where chemical, thermal and electrical stimulations evoke sleep, and where lesions suppress sleep. Based on location and discharge pattern we consider sleep-active cells candidates for mediating some of the sleep-promoting functions of the basal forebrain.     

Cognitive flexibility across the sleep-wake cycle: REM-sleep enhancement of anagram problem solving

Flexible or 'fluid' cognitive processes are regarded as fundamental to problem solving and creative ability, requiring a specific neurophysiological milieu. REM-sleep dreaming is associated with creative processes and abstract reasoning with increased strength of weak associations in cognitive networks. REM sleep is also mediated by a distinctive neurophysiological profile, different to that of wake and NREM sleep. This study compared the performance of 16 subjects on a test of cognitive flexibility using anagram word puzzles following REM and NREM awakenings across the night, and waking performances during the day. REM awakenings provided a significant 32% advantage in the number of anagrams solved compared with NREM awakenings and was equal to that of wake time trials. Correlations of individual performance profiles suggest that REM sleep may offer a different mode of problem solving compared with wake and NREM. When early and late REM and NREM awakening data were separated, a dissociation was evident, with NREM task performance becoming more REM-like later in the night, while REM performance remained constant. These data suggest that the neurophysiology of REM sleep represents a brain state more amenable to flexible cognitive processing than NREM and different from that in wake, and may offer insights into the neurocognitive properties of REM-sleep dreaming.     

Local energy depletion in the basal forebrain increases sleep

Sleep saves energy, but can brain energy depletion induce sleep? We used 2,4‐dinitrophenol (DNP), a molecule which prevents the synthesis of ATP, to induce local energy depletion in the basal forebrain of rats. Three‐hour DNP infusions induced elevations in extracellular concentrations of lactate, pyruvate and adenosine, as well as increases in non‐REM sleep during the following night. Sleep was not affected when DNP was administered to adjacent brain areas, although the metabolic changes were similar. The amount and the timing of the increase in non‐REM sleep, as well as in the concentrations of lactate, pyruvate and adenosine with 0.5–1.0 mm DNP infusion, were comparable to those induced by 3 h of sleep deprivation. Here we show that energy depletion in localized brain areas can generate sleep. The energy depletion model of sleep induction could be applied to in vitro research into the cellular mechanisms of prolonged wakefulness.     

Forerunners of REM sleep

The development of sleep research can be divided into two main periods. The first one was initiated in 1863 by the first systematic measurement of the depth of sleep, the second in 1953 by the discovery of recurrent episodes of rapid eye movements in sleep. The main methodological procedure in the first of these two periods was the measurement of a single physiological variable, while beginning with long-term measurements of the electroencephalogram (EEG) in sleep, multi-channel, polygraphic recording became the method of choice for sleep studies. Although rhythmic changes in the ultradian frequency range of one to 2?h were observed early in many variables during sleep (movements, autonomic functions, penile erections), the recognition of the existence of two different states of sleep (rapid eye movement (REM) and non-rapid eye movement (NREM sleep)) was contingent upon a 'synthetic' view, which focus on the coalescence of multiple variables. The dual concept of sleep organization evolved stepwise in parallel to the rapid growth of neurophysiological knowledge and techniques in the first half of the 20th century, culminating in the discovery of REM sleep.     

Hypothalamic control of sleep

A sleep-promoting function for the rostral hypothalamus was initially inferred from the presence of chronic insomnia following damage to this brain region. Subsequently, it was determined that a unique feature of the preoptic hypothalamus and adjacent basal forebrain is the presence of neurons that are activated during sleep compared to waking. Preoptic area "sleep-active" neurons have been identified by single and multiple-unit recordings and by the presence of the protein product of the c-Fos gene in the neurons of sleeping animals. Sleep-active neurons are located in several subregions of the preoptic area, occurring with high density in the ventrolateral preoptic area (vlPOA) and the median preoptic nucleus (MnPN). Neurons in the vlPOA contain the inhibitory neuromodulator, galanin, and the inhibitory neurotransmitter, GABA. A majority of MnPN neurons activated during sleep contain GABA. Anatomical tracer studies reveal projections from the vlPOA and MnPN to multiple arousal-regulatory systems in the posterior and lateral hypothalamus and the rostral brainstem. Cumulative evidence indicates that preoptic area neurons function to promote sleep onset and sleep maintenance by inhibitory modulation of multiple arousal systems. Recent studies suggest a role for preoptic area neurons in the homeostatic aspects of the regulation of both rapid eye movement (REM) and non-REM (NREM) sleep and as a potential target for endogenous somnongens, such as cytokines and adenosine.     

Changes in the thermal characteristics of hypothalamic neurons during sleep and wakefulness

The characteristics of the mammalian thermoregulatory system are dependent upon arousal state. During NREM sleep thermoregulatory mechanisms are intact but body temperature is regulated at a lower level than during wakefulness. In REM sleep thermoregulatory effector mechanisms are inhibited and thermal homeostasis is severely disrupted. Thermosensitivity of neurons in the preoptic/anterior hypothalamus (POAH) was determined for behaving kangaroo rats (Dipodomys deserti) during electrophysiologically defined wakefulness, NREM sleep and REM sleep to elucidate possible neural mechanisms for previous findings of state-dependent changes in thermoregulation. Thirty cells were tested during at least two arousal states. During wakefulness, 70% of the recorded cells were sensitive to changes in local temperature, with the number of warm-sensitive (W) cells outnumbering cold-sensitive (C) cells by 1.6:1. In NREM sleep, 43% of the cells were thermally sensitive, with the ratio of W:C remaining the same as in wakefulness. In REM sleep only two cells were thermosensitive (both W). The decrease in neuronal thermosensitivity of POAH cells during REM sleep parallels findings of inhibition of thermoregulatory effector responses during REM, although further work is necessary to determine the source and nature of the inhibition.     

What in sleep is for memory

Since the seminal research by Jenkins and Dallenbach in the 1920s, it has been well proven that sleep has a major effect on the memory of pre-sleep material. However, there is still sparse knowledge about exactly which features of sleep have the most impact. Studies which examined separately the role of non-rapid eye movement (NREM) sleep and rapid eye movement (REM) sleep provided largely controversial results and aroused harsh scientific debate, and the investigation of the link of specific sleep patterns to different memory systems (e.g. declarative vs. procedural) did not fully reconcile these inconsistencies. New research perspectives have been proposed in recent years to overcome the limits of the previous 'single state' approach. Psychological, neurophysiological and neuroanatomical data have recently suggested that NREM and REM sleep both play a part in memory consolidation. We here present the hypothesis that NREM and REM are complementary for memory processes during sleep, thanks to their close interaction within the NREM-REM cycle, and discuss experimental data which prove the critical role of the sleep cycle for the morning recall of verbal material.     

State transitions between wake and sleep, and within the ultradian cycle, with focus on the link to neuronal activity

The structure of sleep across the night as expressed by the hypnogram, is characterised by repeated transitions between the different states of vigilance: wake, light and deep non-rapid eye movement (NREM) sleep, and rapid eye movement (REM) sleep. This review is concerned with current knowledge on these state transitions, focusing primarily on those findings that allow the integration of data at cellular level with spectral time-course data at the encephalographic (EEG) level. At the cellular level it has been proposed that, under the influence of circadian and homeostatic factors, transitions between wake and sleep may be determined by mutually inhibitory interaction between sleep-active neurons in the hypothalamic preoptic area and wake-active neurons in multiple arousal centres. These two fundamentally different behavioural states are separated by the sleep onset and the sleep inertia periods each characterised by gradual changes in which neither true wake nor true sleep patterns are present. The results of sequential spectral analysis of EEG data on moves towards and away from deep sleep are related to findings at the cellular level on the generating mechanisms giving rise to the various NREM oscillatory modes under the neuromodulatory control of brainstem-thalamic activating systems. And there is substantial evidence at cellular level that transition to and from REM sleep is governed by the reciprocal interaction between cholinergic REM-on neurons and aminergic REM-off neurons located in the brainstem. Similarity between the time-course of the REM-on neuronal activity and that of EEG power in the high beta range (approximately 18-30 Hz) allows a tentative parallelism to be drawn between the two. This review emphasises the importance of the thalamically projecting brainstem activating systems in the orchestration of the transitions that give rise to state progression across the sleep-wake cycle.     

Prostaglandin D2 and sleep/wake regulation

Prostaglandin (PG) D₂ is the most potent endogenous sleep-promoting substance. PGD₂ is produced by lipocalin-type PGD synthase localized in the leptomeninges, choroid plexus, and oligodendrocytes in the brain, and is secreted into the cerebrospinal fluid as a sleep hormone. PGD₂ stimulates DP₁ receptors localized in the leptomeninges under the basal forebrain and the hypothalamus. As a consequence, adenosine is released as a paracrine sleep-promoting molecule to activate adenosine A_2A receptor-expressing sleep-promoting neurons and to inhibit adenosine A₁ receptor-possessing arousal neurons. PGD₂ activates a center of non-rapid eye movement (NREM) sleep regulation in the ventrolateral preoptic area, probably mediated by adenosine signaling, which activation inhibits the histaminergic arousal center in the tuberomammillary nucleus via descending GABAergic and galaninergic projections. The administration of a lipocalin-type PGD synthase inhibitor (SeCl₄), DP₁ antagonist (ONO-4127Na) or adenosine A_2A receptor antagonist (caffeine) suppresses both NREM and rapid eye movement (REM) sleep, indicating that the PGD₂-adenosine system is crucial for the maintenance of physiological sleep.     

Effects of hypocretin (orexin) neuronal loss on sleep and extracellular adenosine levels in the rat basal forebrain

Neurons containing the neuropeptide hypocretin (HCRT, orexin) are localized only in the lateral hypothalamus, from where they innervate multiple regions implicated in arousal, including the basal forebrain. HCRT activation of downstream arousal neurons is likely to stimulate release of endogenous factors. One such factor is adenosine, which in the basal forebrain increases in level with wakefulness and decreases with sleep, and is hypothesized to regulate the waxing and waning of sleep drive. Does loss of HCRT neurons affect adenosine levels in the basal forebrain? Is the increased sleep that accompanies HCRT loss a consequence of higher adenosine levels in the basal forebrain? In the present study, we investigated these questions by lesioning the HCRT neurons with HCRT‐2–saporin (HCRT‐2–SAP) and measuring sleep and extracellular levels of adenosine in the basal forebrain. In separate groups of rats, the neurotoxin HCRT‐2–SAP or saline was administered locally to the lateral hypothalamus, and 80 days later adenosine and sleep were assessed. Rats given the neurotoxin had a 94% loss of HCRT neurons. These rats woke less at night, and had more rapid eye movement sleep, which is consistent with HCRT hypofunction. These rats also had more sleep after brief periods of sleep deprivation. However, in the lesioned rats, adenosine levels did not increase with 6 h of sleep deprivation, whereas an increase in adenosine levels occurred in rats without lesion of the HCRT neurons. These findings indicate that adenosine levels do not increase with wakefulness in rats with a HCRT lesion, and that the increased sleep in these rats occurs independently of adenosine levels in the basal forebrain.     

Baseline and post-deprivation recovery sleep in SCN-lesioned rats

In humans, advancing age alters sleep patterns, reducing high voltage NREM sleep, sleep bout length, and delta power during NREM sleep. Although the mechanism by which these alterations occur is unknown, age-related changes in normal circadian processes may play a role. Increased age produces histological and functional changes in the suprachiasmatic nucleus (SCN), and alters the amplitude and phase of circadian rhythms. To examine the relationship between SCN function and age-related changes in sleep, we produced radiofrequency (RF) lesions of the SCN in rats of different ages and examined sleep behavior before and after sleep deprivation. Three-, 12- and 18-month-old rats received RF or sham lesions of the SCN. After verifying loss of circadian rhythm, 24-h EEG/EMG/temperature recordings were made in dim light before and after 24 h of sleep deprivation using the disk-over-water method. Age-related changes in NREM sleep, sleep bout length, and delta EEG power persisted despite SCN lesions. SCN lesions in all age groups increased baseline NREM sleep by 4% and NREM delta power by 15%, and decreased REM sleep by 10%. Although SCN lesions initially produced more REM and NREM sleep during recovery, 24-h values did not differ. Deteriorating SCN function is unlikely to cause the characteristic changes in sleep that occur with age. Our data also imply that an intact SCN slightly inhibits NREM sleep in the rat. Changes in NREM sleep and delta EEG power during recovery in lesioned rats suggest that the SCN may influence homeostatic regulation.     

No phylogeny without ontogeny — a comparative and developmental search for the sources of sleep-like neural and behavioral rhythms

A comprehensive review is presented of reported aspects and putative mechanisms of sleep-like motility rhythms throughout the animal kingdom.It is proposed that ’rapid eye movement(REM)sleep’ be regarded as a special case of a distinct but much broader category of behavior,’rapid body movement(RBM)sleep’,defined by intrinsically-generated and apparently non-purposive movements.Such a classification completes a 2×2 matrix defined by the axes sleep versus waking and active versus quiet.Although ’paradoxical’ arousal of forebrain electrical activity is restricted to warm-blooded vertebrates,we urge that juvenile or even infantile stages of development be investigated in cold-blooded animals,in view of the many reports of REM-like spontaneous motility(RBMs)in a wide range of species during sleep.The neurophysiological bases for motorically active sleep at the brainstem level and for slow-wave sleep in the forebrain appear to be remarkably similar,and to be subserved in both cases by a primitive diffuse mode of neuronal organization.Thus,the spontaneous synchronous burst discharges which are characteristics of the sleeping brain can be readily simulated even by highly unstructured neural network models.Neuromotor discharges during active sleep appear to reflect a hierarchy of simple relaxation oscillation mechanisms,spanning a wide range of spike-dependent relaxation times,whereas the periodic alternation of active and quiet sleep states more likely results from the entrainment of intrinsic cellular rhythms and/or from activity-dependent homeostatic changes in network excitability.     

Dream experience during REM and NREM sleep of patients with complex partial seizures

This study examined the effectiveness of the cognitive processes underlying dreaming in patients with complex partial seizures (CPS), by assessing the frequency of recall and the structural organization of dreams reported after awakenings provoked alternately during REM and stage 2 NREM sleep on 12 cognitively unimpaired CPS-patients (six with epileptic focus in the right hemisphere and six in the left one). Each patient was recorded for three consecutive nights, respectively, for adaptation to the sleep laboratory context, for polysomnography and for dream collection. The frequency of dream recall was lower after stage 2 NREM sleep than REM sleep, regardless of the side of epileptic focus, while the length and structural organization of dreams did not significantly differ in REM and NREM sleep. However, the length of story-like dreams was influenced by global cognitive functioning during REM sleep. These findings indicate that in CPSs-patients the elaboration of dream experience is maintained in both REM and NREM sleep, while the access to information for conversion into dream contents and the consolidation of dream contents is much less effective during NREM rather than during REM sleep. Further studies may distinguish between these two possibilities and enlighten us as to whether the impaired memory functioning during NREM sleep is a side effect of anticonvulsant treatment.     

Short-term effects of fluoxetine and trifluoromethylphenylpiperazine on electroencephalographic sleep in the rat

Fluoxetine and trifluoromethylphenylpiperazine (TFMPP) were studied for their short-term effects on electroencephalographic sleep in male rats. Following single injection, each drug produced a sizeable, dose-related suppression of rapid-eye-movement (REM) sleep that persisted for 4-5 h (fluoxetine, 0.625-5 mg/kg; TFMPP, 0.10-1.25 mg/kg). TFMPP also consistently increased non-REM (NREM) sleep during the second hour after drug injection, though this effect was not dose-related (it was seen at all doses tested). Fluoxetine produced small effects on NREM sleep that varied non-systematically with dose and time after drug injection. TFMPP, but not fluoxetine, also increased at all doses the number of delta waves per minute of NREM sleep in the second hour. A structural analog of TFMPP that is inactive at serotonin (5-HT) receptors [4-(m-trifluoromethylphenyl)piperadine; LY97117] was also tested, and found to be devoid of effects on NREM and REM sleep. Both fluoxetine (a 5-HT reuptake blocker) and TFMPP (a 5-HT agonist) enhance transmission across 5-HT synapses, though by different mechanisms. Because they have the common effect of suppressing REM sleep, and in a dose-related manner, the data support the notion that 5-HT neurons in the brain, when active, can suppress REM sleep.     

Parasomnias: An Updated Review

Parasomnias are abnormal behaviors emanating from or associated with sleep. Sleepwalking and related disorders result from an incomplete dissociation of wakefulness from nonrapid eye movement (NREM) sleep. Conditions that provoke repeated cortical arousals, or promote sleep inertia lead to NREM parasomnias by impairing normal arousal mechanisms. Changes in the cyclic alternating pattern, a biomarker of arousal instability in NREM sleep, are noted in sleepwalking disorders. Sleep-related eating disorder (SRED) is characterized by a disruption of the nocturnal fast with episodes of feeding after an arousal from sleep. SRED is often associated with the use of sedative-hypnotic medications; in particular, the widely prescribed benzodiazepine receptor agonists. Recently, compelling evidence suggests that nocturnal eating may in some cases be a nonmotor manifestation of Restless Legs Syndrome (RLS). rapid eye movement (REM) Sleep Behavior Disorder (RBD) is characterized by a loss of REM paralysis leading to potentially injurious dream enactment. The loss of atonia in RBD often predates the development of Parkinson??s disease and other disorders of synuclein pathology. Parasomnia behaviors are related to an activation (in NREM parasomnias) or a disinhibition (in RBD) of central pattern generators (CPGs). Initial management should focus on decreasing the potential for sleep-related injury followed by treating comorbid sleep disorders. Clonazepam and melatonin appear to be effective therapies in RBD, whereas paroxetine has been reported effective in some cases of sleep terrors. At this point, pharmacotherapy for other parasomnias is less certain, and further investigations are necessary.     

REM and NREM sleep as natural accompaniments of the evolution of warm-bloodedness

Divergence of primitive sleep into REM and NREM states is thought to have occurred in the nocturnal Triassic ancestors of mammals as a natural accompaniment of the evolution of warm-bloodedness. As ambient temperatures during twilight portions of primitive sleep traversed these evolving ancestors' core temperature, mechanisms of thermoregulatory control that employ muscle contractions became superfluous. The resulting loss of need for such contractions during twilight sleep led to muscle atonia. With muscle tone absent, selection favored the persistence of the fast waves of nocturnal activity during twilight sleep. Stimulations by these waves reinforce motor circuits at the increasing temperatures of evolving warm-bloodedness without leading to sleep-disturbing muscle contractions. By these and related interlinked adaptations, twilight sleep evolved into REM sleep. The daytime period of sleep became NREM sleep. The evolution of NREM and REM sleep following this scenario has implications for sleep's maintenance processes for long-term memories. During NREM sleep, there is an unsynchronized, uncoordinated stimulation and reinforcement of individual distributed component circuits of consolidated memories by slow wave potentials, a process termed ‘uncoordinated reinforcement’. The corresponding process during REM sleep is the coordinated stimulation and reinforcement of these circuits by fast wave potentials. This action temporally binds the individual component circuit outputs into fully formed memories, a process termed ‘coordinated reinforcement’. Sequential uncoordinated and coordinated reinforcement, that is, NREM followed by REM sleep, emerges as the most effective mechanism of long-term memory maintenance in vertebrates. With the evolution of this two-stage mechanism of long-term memory maintenance, it became adaptive to partition sleep into several NREM-REM cycles, thereby achieving a more lengthy application of the cooperative sequential actions.     

New neurons in the adult brain: the role of sleep and consequences of sleep loss

Research over the last few decades has firmly established that new neurons are generated in selected areas of the adult mammalian brain, particularly the dentate gyrus of the hippocampal formation and the subventricular zone of the lateral ventricles. The function of adult-born neurons is still a matter of debate. In the case of the hippocampus, integration of new cells in to the existing neuronal circuitry may be involved in memory processes and the regulation of emotionality. In recent years, various studies have examined how the production of new cells and their development into neurons is affected by sleep and sleep loss. While disruption of sleep for a period shorter than one day appears to have little effect on the basal rate of cell proliferation, prolonged restriction or disruption of sleep may have cumulative effects leading to a major decrease in hippocampal cell proliferation, cell survival and neurogenesis. Importantly, while short sleep deprivation may not affect the basal rate of cell proliferation, one study in rats shows that even mild sleep restriction may interfere with the increase in neurogenesis that normally occurs with hippocampus-dependent learning. Since sleep deprivation also disturbs memory formation, these data suggest that promoting survival, maturation and integration of new cells may be an unexplored mechanism by which sleep supports learning and memory processes. Most methods of sleep deprivation that have been employed affect both non-rapid eye movement (NREM) and rapid eye movement (REM) sleep. Available data favor the hypothesis that decreases in cell proliferation are related to a reduction in REM sleep, whereas decreases in the number of cells that subsequently develop into adult neurons may be related to reductions in both NREM and REM sleep. The mechanisms by which sleep loss affects different aspects of adult neurogenesis are unknown. It has been proposed that adverse effects of sleep disruption may be mediated by stress and glucocorticoids. However, a number of studies clearly show that prolonged sleep loss can inhibit hippocampal neurogenesis independent of adrenal stress hormones. In conclusion, while modest sleep restriction may interfere with the enhancement of neurogenesis associated with learning processes, prolonged sleep disruption may even affect the basal rates of cell proliferation and neurogenesis. These effects of sleep loss may endanger hippocampal integrity, thereby leading to cognitive dysfunction and contributing to the development of mood disorders.     

Band-specific electroencephalogram and brain cooling abnormalities during NREM sleep in patients with winter depression.

BACKGROUND: We previously reported that delta wave activity and facial skin temperatures, an index of brain cooling activity, were both abnormal during sleep in patients with winter depression (SAD). Because other electroencephalographic (EEG) frequencies may also convey relevant thermal, homeostatic, and circadian information, we sought to spectrally analyze delta, theta, alpha, and sigma frequencies during sleep from 23 patients with SAD and 23 healthy control subjects. METHODS: We computed means for delta, theta, alpha, and sigma power during both NREM and REM sleep. We also generated 22 cross-correlation functions for each group by crossing facial and rectal temperature with each other, as well as with delta, theta, alpha, and sigma frequencies. RESULTS: We found that delta, theta, and alpha frequency activities were all increased during NREM, but not REM sleep, in patients with SAD. In addition, there were significant and abnormal cross-correlations between facial temperatures and delta and theta frequencies during NREM sleep in patients with SAD. CONCLUSIONS: Patients with winter depression exhibit correlated abnormalities of sleep homeostasis and brain cooling during NREM sleep. Their EEG profiles during NREM sleep resemble the EEG profiles of subjects who have been sleep deprived. Further studies of NREM sleep homeostasis in patients with SAD seem warranted.