Sleep, Anesthesia, and Consciousness

“And the LORD God caused a deep sleep to fall upon Adam, and he slept; and He took one of his ribs, and closed up the flesh in its place.”

Genesis (2:21)

Irrespective of whether the reader's beliefs are rooted in the Judeo-Christian tradition, the passage from Genesis suggests that the concept of surgical anesthesia as a deep sleep is ancient. Sleep is a ubiquitous metaphor in the modern practice of anesthesiology, where it is common to hear patients being told “You are going to sleep now” or “Time for a little nap” as anesthesia is being administered. Such phrases are reassuring, as sleep has positive connotations of restoration and is a state of reversible unconsciousness that every patient has experienced. The metaphor is no doubt rooted in the many phenotypic similarities of sleep and anesthesia.¹ Indeed, these two distinct states share a number of important traits such as hypnosis, amnesia, and immobility.²

In the mid-1990s, it was hypothesized that general anesthetics might preferentially act via neuronal mechanisms that have evolved to control sleep-wake states.³ Since that time, a number of sleep-wake nuclei in the brainstem and diencephalon have been demonstrated to be involved in anesthetic mechanisms.⁴ It has also become clear that there is a functional relationship between sleep and anesthesia. For example, sleep deprivation in animals accelerates the induction of general anesthesia and delays recovery,⁵ a phenomenon mediated in part by the adenosinergic system.⁶ In terms of sleep homeostasis, animals exposed to a prolonged infusion of propofol—a commonly used intravenous anesthetic—do not show signs of subsequent sleep rebound phenomena.⁷ Furthermore, propofol has been shown to satisfy the homeostatic need for sleep after deprivation as effectively as ad libitum sleep itself.⁸ But what is the link between the activity of anesthetic drugs such as propofol on sleep-wake nuclei and effects at the behavioral level? The study by Murphy et al.⁹ in the current issue of SLEEP may help lead us to an answer.

Using high-density electroencephalography, Murphy et al. have characterized and modeled slow waves during both propofol anesthesia and sleep.⁹ The use of 256 electroencephalographic channels empowered the investigators to achieve both the temporal resolution classically associated with neurophysiologic techniques as well as enough spatial resolution to identify the probable origin, propagation path, and involvement of propofol- and sleep-related slow waves. Although a prior study has compared spindle activity during propofol anesthesia and sleep,¹⁰ this is the first systematic assessment of slow waves in the two states. Numerous similarities were identified between anesthesia and sleep, including: slow wave activity in virtually all unconscious epochs, occurrence of the events in a large number of channels, consistency of frequency and distribution across subjects, and comparable scalp voltage topography. Most notable was the frequent origin of both propofol- and sleep-related slow waves in the insular and cingulate cortices. Slow waves in both states propagate posteriorly along the mesial highway of the anterior cingulate, cingulate, and posterior cingulate gyri. There were also notable differences in propofol slow wave generation and distribution, such as the occipital origin of some propofol-induced slow waves, less demarcated spatial distribution, and lack of the distinctive asymmetries found during sleep. Nonetheless, the authors present compelling evidence that the similarities argue for propofol anesthesia being a sleep-like state. It is of interest to note that a recent neuroimaging study of propofol-induced unconsciousness¹¹ led the authors of an accompanying editorial to the opposite conclusion.¹²

Despite the remarkable neurophysiologic homology of propofol anesthesia and sleep, an important question remains: do slow waves represent the business end of the functional relationship between sleep and anesthesia? Recent evidence suggests that while slow waves during anesthesia can, in fact, satisfy slow wave sleep homeostasis, they are not required. This is evidenced by the fact that the inhalational anesthetic desflurane titrated to either slow waves or isoelectricity satisfied slow wave homeostasis.¹³ Another recent study of isoflurane titrated to slow waves or burst suppression also concluded that anesthetic slow waves per se were not necessary for the satisfaction of sleep homeostasis.¹⁴ Furthermore, sevoflurane titrated to 50% burst suppression was recently shown to have a profound effect in attenuating non-rapid eye movement sleep rebound after total deprivation, with a concomitant slow wave homeostasis.¹⁵ Further investigation is required to determine whether slow waves are the functionally relevant component of general anesthesia that accounts for its effects on sleep homeostasis.

However, one of the most important findings of the study by Murphy and colleagues⁹ has significant implications not for sleep neurobiology but for the science of consciousness: gamma activity (25–40 Hz) during propofol anesthesia increased in the anterior and posterior cingulate cortices and its synchrony persisted. Although perhaps not universally true, it is arguable that a neural correlate of consciousness should be present during the waking state, absent or attenuated during anesthetic-induced unconsciousness, and present again after the recovery of consciousness. As such, the fact that gamma activity in the range of 25–40 Hz in these particular regions does not follow this pattern weakens the long-held hypothesis that “40 Hz” activity is a neural correlate of consciousness. The finding also presents a challenge to past studies of anesthetic mechanism related to the interruption of gamma activity.^16,17 However, it is important to note that the focus of research on anesthetic attenuation of gamma activity is shifting from 40 Hz to higher frequencies of gamma. The recent study by Breshears et al.¹⁸ used human electrocorticography to demonstrate a linear reduction of 75–105 Hz gamma power throughout induction of propofol anesthesia, with a return to baseline during recovery. Although the patients in that study had epilepsy, interpretations based on these direct recordings from the human cortex are supported by animal models showing that fast gamma activity is attenuated by anesthesia.¹⁹ The data on auditory evoked potentials alluded to by Murphy and colleagues⁹ will also be of interest given the role of the 40-Hz auditory steady state response,²⁰ as well as faster activity,²¹ in the assessment of anesthetic depth.

In conclusion, the study of Murphy et al.⁹ supports the hypothesis that general anesthesia is a sleep-like state, establishes a foundation for the further functional study of slow waves in the sleep-anesthesia connection, and directs us to a more refined perspective of the role of gamma activity in consciousness and anesthetic mechanisms.

DISCLOSURE STATEMENT

Dr. Mashour has indicated no financial conflicts of interest.