The main focus of our present work is the phenomenon of slow waves, considered as a quintessence of sleep. Slow waves are also known as delta waves according to the classical EEG terminology, originating from William Gray Walter (1910–1977) in the 1930s (Walter 1936). He has performed their topographic mapping and recognized the link between brain lesions and local delta activity (meaning disease, degeneration, and death). The strong link between slow waves and sleep was first reported, also in the 1930s, by Alfred Lee Loomis (1935). He managed to make all-night EEG records of human sleep in his laboratory. This was an unbelievable technical achievement at those times as even Harvard researchers were unable to perform such records. Loomis characterized the EEG stages of sleep: the stage characterized by slow waves was called random, as the waves were considered as such in his view. He also described the increase in the amplitude of delta waves during the course of wake-sleep transition until reaching deep sleep, and most importantly he also recognized the reactivity of the slow waves and made a correct description on K-complexes evoked by knocking on the sleepers’ door (K-knock).

Thus, slow waves were considered mainly as signs of low arousal level (Loomis et al. 1935) or brain lesion (Walter 1936). This idea was present in the work of Giuseppe Moruzzi (1910–1986) and Horace Winchell Magoun (1907–1991) also who described the ascending reticular activating system in the brainstem and characterized it as follows: “The evidence given above points to the presence in the brain stem of a system of ascending reticular relays whose direct stimulation activates or desynchronizes the EEG, replacing high-voltage slow waves with low-voltage fast activity. This effect is exerted generally upon the cortex and is mediated, in part, at least, by the diffuse thalamic projection system” (Moruzzi and Magoun 1949). High-voltage slow waves were already considered as signs of sleep or low arousal in these works. This kind of “negative” view on slow waves (i.e., being signs of lack of something) persisted and was even more emphasized during and after the discovery of REM sleep in humans (Aserinsky and Kleitman 1953) or paradoxical sleep in cats (Jouvet et al. 1959). It was the appearance of fast waves instead of slow ones during sleep that seemed interesting in those times.

However, scientific interest has changed during the next few years. The search for the molecular sleep factor (S-factor) by the Pappenheimer group stimulated the search for some reliable physiological measure of sleep deprivation-induced sleep rebound. It was recognized that the best measure is sleep EEG slow wave activity. They wanted to introduce into these studies the time spent in those sleep phases characterized by EEG slow waves as well as the increase in their amplitude, instead of focusing on low-voltage fast activity (Pappenheimer et al. 1975). Similar approaches characterized the research on humoral sleep regulation (Schoenenberger et al. 1977).

Although the homeostatic regulation of slow wave sleep was already known in those studies, it was Wilse B. Webb who explicitly stated that the amount of stage 4 sleep was related to the amount of pre-sleep wakefulness (Webb and Agnew 1971). Hints on the potential recuperative role of slow wave sleep together with its precise homeostatic regulation changed the view of sleep researchers and have led to the generally accepted theories: slow waves are not merely passive expressions of reduced cortical activity due to brain lesions or low arousal; rather, they represent active self-organizational forms of neuronal activity, serving recuperative and off-line information processing functions.

We are witnessing that the hitherto very poorly understood sleep EEG graphoelements partially described earlier come step-by-step to life and fill up by meaning. They behave as an organic system with biological role, like the pieces of wood transformed to Pinocchio, a living child with full conscience in the hands of master Geppetto, in the famous tail.

We learned from the Steriade group that different rhythms (spindles, delta, and <1-Hz slow waves on one hand and fast frequencies on the other) prevail in coalescence during slow wave sleep.

Within the cortical <1-Hz slow activity during the so-called up states (surface-positive half wave) even ripple-like (50–200 Hz) fast activity and during the “down states” (surface-negative half wave), an interruption of synaptic and neural activity alternates (Fig. 7.2). “The entire cortical network can swing rhythmically between extremely different microstates, ranging from wakefulness-like network activation to functional disconnection in the space of a few milliseconds” (Massimini et al. 2003). The “black-hole”-like down state may provide a rest phase and restoration for the most vulnerable and recuperation demanding frontal cognitive activity.

In the work of Csercsa et al. (2010) intracortical laminar local field potential gradient, multiple-unit and single-unit activities were recorded during slow wave sleep, related to simultaneous electrocorticography, and were analyzed by current source density and spectral methods. We found that slow wave activity in humans reflects a rhythmic oscillation between widespread cortical activation and silence. Cortical activation was demonstrated as increased wideband (0.3–200 Hz) spectral power including virtually all bands of cortical oscillations, increased multiple- and single-unit activities, as well as powerful inward transmembrane currents, mainly localized to the supragranular layers. Neuronal firing in the “up state” was sparse, and the average discharge rate of single cells was less than expected from animal studies. Synchronization patterns of action potentials across all cortical layers suggested that any layer could initiate firing at the onset of up states.

These findings provide strong direct experimental evidence that slow wave activity in humans is characterized by hyperpolarizing currents associated with suppressed cell firing, alternating with high levels of oscillatory synaptic/transmembrane activity associated with increased cell firing. Our results emphasize the major involvement of supragranular layers in the genesis of slow wave activity.

The recurrent cortical disfacilitation during down states together with high-level cortical activity during up states has a serious consequence for sensory transmission. The main one is that the thalamocortical sensory processing during deep slow wave sleep is not steady (Massimini et al. 2003). This could be shown by experiments where the averaging of sensory evoked potentials was grouped according to the up and down states of <1-Hz slow wave oscillation in deep sleep of animals. By this method, the amplitudes of cortical somatosensory evoked responses were found to be different: high, although not reaching wakefulness-like levels, during the up states (confirming the possibility of detecting meaningful events even in deep sleep) and low during the disfacilitation of the down state. This lack of stationarity may impair sensory integration during deep slow wave sleep.

The activation of certain brain structures in slow wave sleep has been recorded by high-tech neuroimaging methods recently.

The advent of functional neuroimaging allowed the registration of regional cerebral blood flow (SPECT) and metabolic (PET scanning) and later neuronal activity BOLD (fMRI) changes also for sleep studies.

PET studies using/18F/-fluorodeoxyglucose as a tracer showed that global cerebral glucose metabolism in humans is lowest in stages 2–3 in NREM sleep, in contrast to REM sleep in which it is similar to the waking state (Maquet et al. 1990). Later, the resolution of PET methods substantially improved both in time and in space, and a more reliable determination of the regional differences during sleep has become possible (Braun et al. 1997; Hofle et al. 1997; Nofzinger et al. 1997; Maquet and Phillips 1998; Maquet 2000). A significant decrease (greater than in the rest of the brain) in rCBF has been shown during NREM sleep in the brainstem (dorsal pons and mesencephalon), thalamus, basal ganglia, basal forebrain, and hypothalamus and in the orbitofrontal and anterior cingulate cortex, as well as the precuneus.

The fMRI studies offer sophisticated knowledge about BOLD activation patterns related to sleep (Czisch et al. 2002, 2004, 2009; Kaufmann et al. 2006; Schabus et al. 2007; Dang-Vu et al. 2008). Several brain regions’ activities were found to be reduced during each NREM sleep stage. Most of these regions are located in the frontal lobes, but some regions of the thalamus, limbic system (anterior and posterior cingulum), as well as wider areas of the temporal, parietal, and occipital lobes, and the midbrain, hypothalamus, and mamillary bodies have also been shown to have decreased activity (Kaufmann et al. 2006). Certain patterns of activation sequences were described along the process of falling asleep, like loss of alertness during expectation of environmental stimuli as well as during deepening sleep. The decrease of insular activity gained significance related to interoceptive awareness. Special attention was paid to the crucial hypothalamic region showing globally decreased activity reflecting to the decreased firing of wake-promoting neurons. It is interesting that “some pronounced synchronous BOLD activity changes” were observed in the hypothalamus in NREM that correlated with a simultaneous activity in wider cortical areas resembling to the activity in the pathways of the ascending arousal system during NREM sleep. Kaufmann et al. (2006) emphasize the role of a hypothalamic-driven network regulation both in the initiation and maintenance of sleep. They observed a decreased activity of the hypothalamic regions throughout NREM sleep stages, reflecting the reduced activity of wake-promoting neurons. Due to limited spatial resolution, differentiating hypothalamic subregions was not possible. However, “some pronounced synchronous BOLD activity changes that correlated with the hypothalamus occurred: limbic structures, regions of the frontal and parietal cortex, the basal forebrain and brainstem showed a similar time course of activation as hypothalamic regions did.” Kaufman et al. interpreted this pattern of activation as resembling to the pathway of the ascending arousal system which sends projections from the brainstem and posterior hypothalamus throughout the forebrain. These findings are congruent with the role of arousal-driven phasic activity, as described in this book. Phasic increases were observed in regional brain activity in relation to discrete events as slow waves or delta waves, compared to baseline sleep activity (Dang-Vu et al. 2008). Delta waves (>1 Hz) were associated with frontal responses. Significant increases in activity were associated with the <1-Hz waves in several cortical areas, including the inferior frontal, medial prefrontal, and precuneus and posterior cingulate areas. The <1-Hz slow waves are associated with significant activity in the parahippocampal gyrus, cerebellum, and brainstem, whereas delta waves are related to frontal regions compared to baseline activity. No decrease of activity (negative BOLD signal) was observed. A partial overlap was seen between the default network of wakefulness and responses associated with slow wave oscillation. They unexpectedly found significant responses associated with slow oscillations in the midbrain and pontine tegmentum, especially in regions of cholinergic and aminergic nuclei. This particular finding again points to traces of arousal activity during NREM sleep, the importance of which was highlighted in Chap. 6

Using high-density EEG and source modeling methods, sleep slow waves occurring spontaneously and evoked by transcranial magnetic stimulation were studied (Murphy et al. 2009). It was found that some areas of the cortex are differentially involved in slow waves and that sleep slow waves can be locally regulated. Slow waves as a group were associated with large currents in the medial, middle, and inferior frontal gyri, the anterior cingular, the precuneus, and posterior cingular regions. This study again emphasized that areas showing maximal involvement in slow waves also show considerable overlap with the default network which is paradoxically implicated in monitoring the external environment and can be altered by sleep deprivation.

Further understanding of the role of slow wave oscillation and particularly of the up and down states in plastic changes during sleep would be crucial; however, spatial resolution of fMRI does not allow us to reveal the time and space structure of cortical up and down states during NREM sleep changes.

We have shown that K-complexes express a double nature (Janus-faced): they show responsivity to sensory (mainly acoustic) stimuli and behave as slow waves following the homeostatic decay increasing after sleep deprivation and being related to the deepness of the actual sleep cycle. Furthermore, the studies of Amzica and Steriade (1998) and later Cash et al. (2009) proved that the main surface-negative large N550 component of K-complexes is characterized by an important decrease of unit discharges and synaptic activity in the frontal cortex similar to the “down state” of the frontal slow oscillation during NREM sleep (Fig. 3.4).

According to the fMRI studies of Riedner et al. (2011), even the spontaneous K-complexes wear, at their initial segment, the traces of multisensory processing. Therefore, we may consider K-complexes as the most characteristic prototypes of the “reactive slow wave activity.” Although K-complexes can be well obtained by the averaging also in deep slow wave sleep (Fig. 3.3), they are more frequent in stage 2.

K-complexes are essential constituents of CAP sequences, mainly in the A₁ phase events. At the same time A₁ events occur much more frequently during the descending slopes of the first cycles, but both spontaneous and elicited K-complexes have higher frequency during the “A” slope. This difference in distribution is even more pronounced if we consider the behavior of evoked K-complexes under ­continuous random sensory stimulation (Halász 1982). While less reactive slow activity can be elicited in the form of CAP A₁ phases during the “A” slopes of the first cycles, the responsivity in the form of singular K-complexes is more preserved compared to the “D” slopes.

Thus, K-complex occurs more frequently during stage 2 sleep and during “A” slopes of cycles and seems to be more linked to the arousal compared to other sleep slow waves. If we involve into the group of reactive sleep graphoelements the classical vertex sharp transients, a continuum of these graphoelements can be outlined more clearly. In superficial sleep (stage 1–2), sensory (mainly acoustic) stimuli evoke the vertex sharp wave, which can be considered as an exaggerated form of acoustic evoked potential, without any important slow wave constituent. Later, in stage 2, the stimulation evokes K-complexes instead of vertex waves, with recognizable remnants of the acoustic evoked potential, but the essential part is the slow wave complex (N350-N550-P900, see Fig. 3.3). Beginning from stage 2, in deeper NREM sleep, the evoked response becomes more complex in the form of CAP A₁, which incorporates K-complexes, slow waves, and spindles, while the original evoked potential compartment is hardly visible. The response to sensory stimulation shifts from a multimodal evoked potential-type response to a nonspecific bilateral, frontally dominant slow wave response.

The recent functional neuroimaging results seem to confirm this continuum. Stern et al. (2011), investigating vertex sharp wave generation by fMRI, have shown that BOLD positive activations of regions principally encompass the primary sensorimotor cortical regions for vision, hearing, and touch. Earlier in this book, we detailed the findings of Riedner et al. (2011) who had highlighted the initial evoked potential compartments before the slow wave components of K-complexes. The fMRI counterparts of sleep slow waves were discussed in detail in Sect. 6.2.

The special situation of K-complex in sensory information workup and in fulfilling trophotropic and plastic functions of NREM sleep is nicely supported by the findings of Jahnke et al. (2012). In their fMRI study, they found that K-complex is associated with positive BOLD signal changes in subcortical (brainstem and thalamus), cerebellar, sensory (auditory, visual, and sensorimotor) motor, and midline (anterior and midcingular gyri, precuneus) areas and regions which form part of the so-called default mode network (DMN) (prefrontal cortex, inferior parietal lobule, precuneus). Negative BOLD signal changes were observed in bilateral insular cortices. The primary auditory cortex was the first cortical region to be influenced during K-complex.

These findings confirm “the central role of the thalamus in the mediation of cortically generated K-complex leading to sleep spindles and slow waves” and also that “K-complex permits a low level information processing in one hand and a sleep protective counteraction on the other.” The DMN-type activation is interpreted in line with the “sentinel hypothesis,” describing DMN activity “in the context of a broad low level focus of attention when one – like a sentinel – monitoring the external world for unexpected events.” This version of the Hamletian question (“to wake or not to wake”) is very near to our interpretation (Halász 2005).

Summing up, within the continuum of reactive sleep graphoelements, K-complex seems to be situated between the evoked potential type and reactive slow wave type responses (Table 7.2). The functional significance of this Janus-faced character is that during K-complex, the brain decides “not to wake up” and compensates the disturbing effect of the incoming stimulus by producing slow wave activity.