Notice: Undefined offset: 3608 in /var/www/ on line 5961
General anaesthesia - Scholarpedia

General anaesthesia

From Scholarpedia
Axel Hutt et al. (2013), Scholarpedia, 8(8):30485. doi:10.4249/scholarpedia.30485 revision #135739 [link to/cite this article]
Jump to: navigation, search
Post-publication activity

Curator: Axel Hutt


There are many definitions of general anaesthesia , but at the most profound level it may be defined as a state in which the normal responses to the noxious stimulus of surgery have been blocked. These responses include: conscious awareness of the surgery, memory of surgery, autonomic activation, somatic movement, hyperalgesia, and immune depression (Sleigh, 2011).

Mostly general anaesthesia is mediated by drugs, but there are occasional reports of other less reliable options such as electro-anaesthesia (Lambooy, 1985) and hypnosis. Originally general anaesthesia was induced by a single drug, but in modern anaesthesia it is usual to give a combination of synergistic drugs so as to achieve the appropriate blend of general anaesthetic blockade, but with fewer side effects.


Origin of unconsciousness

The mechanisms by which general anaesthetics cause unconsciousness are only partially understood (Campagna, Miller et al. 2003). Coma is believed to be the result of either widespread inhibition of neocortical activity or reduction in information flow between different cortical regions (Alkire et al. 2008; Changeux 2012). This may occur from a direct drug effect on the neocortex, or secondary to a reduction in brain stem neuromodulator input into the cortex – which is required to keep the cortex in a depolarised active state. A number of neuromodulator substances and specific nuclei in the brain stem and hypothalamus have been identified as contributing to the maintenance of the wakeful state (Sanders et al. 2012). It is still unclear whether clinical anaesthesia occurs primarily as the result of direct cortical effects of the anaesthetic drugs or via indirect suppression via brain stem inactivation (Lu et al. 2008).

Classification of general anaesthetic drugs

The understanding of the action of general anaesthetic drugs is important to learn more about the origin of spectral changes in neural activity and more macroscopic phenomena such as the loss of consciousness. They can be classified into three groups according to their predominant molecular pharmacological effects (Grasshoff et al. 2006).

Intraveneous drugs

Intravenous gamma-amino-butyric-acid(GABA)ergic drugs (Garcia et al., 2010), such as etomidate, barbiturates, propofol, midazolam, cause the chloride channel of GABA-receptors to favour the open state, resulting in longer inhibitory postsynaptic currents at GABA synapses, and some hyperpolarisation in neuronal membranes associated with extra-synaptic GABA-chloride channels. In other words, they move the balance of excitation and inhibition in the cortex towards inhibition. These drugs are relatively potent agents with an effective concentrations of 5-50$\mu$M.

Volatile drugs

Examples for such drugs are isoflurane, halothane, desflurane and sevoflurane. They are not very potent agents (with an effective dose $>$300$\mu$M) and bind non-specifically to many proteins in the body. It is likely that their actions are also mediated largely by GABAergic and potassium channel opening effects (Franks and Lieb, 1994; Alkire et al., 2008). However they also have significant anticholinergic effects, N-methyl-D-Aspartate(NMDA) blockade, \(I_h\) current inhibition, gap junction closure, and presynaptic reduction of EPSP release.

Non-GABAergic drugs

This group of drugs includes typically blockers of NMDA, such as nitrous oxide, ketamine or cyclopropane. These drugs produce a qualitatively different type of anaesthesia characterised by a dissociated state, and an activated EEG that resembles the EEG seen in REM sleep. These drugs are believed to act via a reduction in excitatory synaptic transmission.

There are also a few miscellaneous compounds, in particular urethane. This drug is carcinogenic, and therefore not used clinically in human medicine. However it is well-established as drug in animal anaesthesia (Hara and Harris, 2002). It acts uniquely to reduce neuronal spike rate by opening a barium-sensitive potassium channel.

Experimental effects of anaesthetic drugs

EEG effects

Most GABAergic anaesthetic drugs cause EEG changes (San-juan et al., 2010) that are consistent with the thalamocortical hyperpolarising effects of these drugs; namely the loss of low amplitude-high frequency waves that characterize the wakeful state, and increasing predominance of activity in the alpha (8-16Hz) and delta (1-4Hz) wavebands (Bennett, Voss et al. 2009). Around the point of loss of consciousness the EEG shows an increase in amplitude in most frequency bands (the so-called “bi-phasic” effect (Kuizenga et al., 2001)). In higher concentrations these drugs cause a burst-suppression EEG pattern – which is (paradoxically) probably a state of increased cortical irritability (Kroeger and Amzica 2007) – and eventually for even higher concentrations a flat line EEG.

Non-EEG effects

Local Field Potentials

Apart from the effects on human EEG, intracranial recordings in animals have revealed how anesthetics affect neural populations more generally. The analysis of Local Field Potentials (LFP) in rat cortical areas and hippocampus under isoflurane anesthesia has discovered power reduction in the high-gamma frequency range (60Hz-120Hz) in all areas, whereas the power in the low-gamma frequency range (30Hz-60Hz) remained unchanged (Hudetz et al., 2011).

Cerebral blood flow

Moreover, the administration of the anesthetic propofol yields a global decrease of the brains' cerebral blood flow and regional flow decrements in the precuneus and the posterior cingulate (Alkire and Miller, 2005). These brain areas are implicated in the regulation of arousal and performance of conscious functions (Alkire et al., 2008).


Experimental studies also have revealed possible consequences on cognition. For instance isoflurane anesthesia in young rats and mice may lead to memory impairment and a reduced subsequent neurogenesis (Zhu et al., 2010). Even increased mortality after anesthesia has been reported (Monk et al., 2005).

During the transitions between consciousness and anaesthesia it is not uncommon for the patient to show abnormal movements or rigidity. There are a number of explanations for these signs of excitation - the most popular being preferential anaesthetic sensitivity of various inhibitory circuits.


Although anesthetic concentration is normally measured in the blood plasma, it takes about 2--4 min (Ludbrook et al., 1999) for the drug to diffuse across the blood-brain barrier to reach the presumed effect sites in the brain. To compensate for this delay, pharmacokinetics (PK) modelers introduce one or more hypothetical fluid volumes or compartments, linked by empirically-tuned rate constants, so that measured blood-plasma drug concentrations can be used to predict the effect-site concentrations in the brain (Roberts 2007). PK model tuning proceeds on the basis that if site-effect delay errors have been eliminated, then there should be no hysteresis separation between induction and emergence trajectories. In other words, the effect-site concentration at loss of consciousness (LOC) should match that at the moment of recovery of consciousness (ROC). The implicit assumption is that the recovery trajectory is a retracing of the induction trajectory. If in fact the entry and exit paths are distinct, as suggested by the phase transition model for cortical anesthesia (Steyn-Ross et al., 2004), then the PK approach of closing the hysteresis loop may represent an overcompensation. It is worth noting that PK corrections for anesthetic response have not been particularly successful (Coppens et al., 2010), and this may be due to patient-to-patient variability as well as an inherent neural inertia in brain response to anesthesia (Friedman et al., 2010).

Neural models

Few studies consider single neuron models to explain experimental data in the presence of anesthectics (Faulkner et al., 1999). Network models based on single neurons are successfull to describe experimental phenomena, such as paradoxical excitation (McCarthy et al., 2010), increases in the EEG \(\alpha\)-power (Ching et al., 2010), burst suppression (Ching et al., 2012) or cortical spindles (Destexhe et al., 1999). These biophysical models consider networks of up to several hundred neurons.

Rather than attempting to track the detailed biophysics of individual neurons, mean-field or continuum models of the cortex such as neural fields aim to describe the interactions of entire "populations" of neurons whose coordinated activity generates the electrical currents that are detectable with scalp electrodes (EEG/MEG).

Incorporating anesthetic action

At the molecular level, general anesthetic agents act directly on neurotransmitter-gated ion channels to suppress neural activity by either decreasing excitation, or increasing inhibition, or both. For instance, the GABAergic drug propofol has little effect on excitatory synapses but its primary effect is to increase negative charge transfer by prolonging the duration of the decay phase of the inhibitory postsynaptic time-dependent current (IPSC). Kitamura et al. (2002) found that peak-current amplitude was largely independent of propofol concentration in cortical neurons. By virtue of the lack of action on the excitatory receptors and its relatively clean response, i.e., lack of off-target effects, propofol is a popular choice for both modelers and anaestheologists.

For simplicity most modelers have followed Kitamura et al. (2002) and assumed a constant-height IPSC whose area scales linearly with propofol concentration. The IPSC has a pulsatile shape that is well approximated by a biexponential function with rise-rate $\beta$ and decay-rate \(\alpha\), i.e. \(\alpha \beta (\exp(\)-\(\alpha t)\)-\(\exp(\)-\(\beta t))/(\beta\)-\( \alpha)\), or by Rall's synaptic alpha-function \(\gamma^2 t \exp(\)-\(\gamma t)\) where \(\beta=\alpha=\gamma\), see Figure 1 1.
Figure 1: Anesthetic modulation of decay phase of (a) biexponential and (b) alpha-function postsynaptic response curves, normalized to unit height. Dimensionless parameter \(p\) represents anesthetic effect. Settings: (a) \((\alpha, \beta) = (50/p, \,\, 200)\)~s\(^{-1}\); (b) \(\gamma = \alpha = \beta = (120/p)\)~s\(^{-1}\) with \(p = 1, 2, 3\)

To introduce the effect of GABAergic anesthetics, we need to prolong the decay-phase of the response by a factor \(p\ge 1\) for which \(p-1\) is proportional to drug concentration. This can be done either by scaling the \(\gamma\) rate-constant of the second form: \(\gamma \to \gamma/p\) (Steyn-Ross et al., 1999), or by scaling the \(\alpha\) decay-rate in the biexponential form: \(\alpha \to \alpha/p\) (Bojak and Liley, 2005; Hutt and Longtin, 2010; Hindriks and van Putten, 2012). The latter approach has the advantage of allowing independent control of rise- and decay-rates. However, in all variants, the core concept --- that the area of the IPSC (and therefore the magnitude of the charge transferred in the synaptic cleft) should increase monotonically with anesthetic concentration---is retained.

Continuum models

The first continuum model for anesthesia (Steyn-Ross et al., 1999) includes the effect of GABAergic anesthetic as an alpha-function, cf. Figure 1 1(b). This model consists of two homogeneous populations of interacting excitatory and inhibitory neurons in which inhibitory synaptic coupling increases with anesthetic concentration. By locating numerically the equilibrium states of the model, it was established that there could be multiple cortical states for a given value of anesthetic concentration, leading to the possibility that the loss of consciousness (LOC) might correspond to a first-order switching transition from an activated high-firing state to a quiescent low-activity state of the cortex; see Figure 2 2 and 3(A). Moreover, because recovery of consciousness (ROC) occurs at a lower drug concentration than LOC, there is a predicted hysteresis separation between jump points (Steyn-Ross et al., 2004; Friedman et al., 2010).

Figure 2: Distribution of equilibrium firing rates \(Q\) (s\(^{-1}\) ) for a phase-transition model of anesthesia, plotted as a multivalued function of anesthetic inhibition $p$ versus subcortical excitation \(\Delta V^\text{rest}\) (in mV). Increasing anesthetic concentration moves the cortex along the red path on the high-firing upper branch (representing awake state), leading to a jump transition at the turning point onto the low-firing lower branch (representing the anesthetic state).

Subcortical stimulation of the cortical model (Steyn-Ross et al., 1999, 2004) with low-intensity white noise generated voltage fluctuations corresponding to explorations of the near-equilibrium state space. These fluctuations showed pronounced spectral slowing and increase in amplitude on close approach to the LOC and ROC jump points. Mathematically, the power surge in the solution of the model equations arises from the divergent critically-slowed fluctuations generated close to a saddle-node annihilation point (Steyn-Ross et al., 2006); see Figure 3 3. This provided the first putative explanation for the paradoxical boosts in EEG activity (the biphasic effect) that characterize most GABAergic anesthetic drugs during both induction and recovery (Kuizenga et al., 2001). However, this critical-slowing explanation for EEG spectral change is incomplete since the observed biphasic boosts during induction of anesthesia are frequency-specific, with the activity pattern following a systematic progression from high to low frequencies.

Figure 3: Simulation results for induction of anesthesia via slow increase in anesthetic effect \(p\). A. Cortical firing rate \(Q\) shows a downwards jump change at the turning point near \(p = 1.06\). B. Noise-induced fluctuations $\delta Q$ (black and gray traces) at two points on the cortical grid become significantly slower and larger on close approach to the saddle--node bifurcation at the turning point. C. Fluctuation power averaged across the cortical grid shows a pronounced surge as the point of induction (LOC: loss of consciousness) is approached.

Bojak and Liley (2005) investigated anesthesia under isoflurane and showed that, for suitable choices of physiological parameters, biphasic activity surges in EEG activity can be generated without requiring a phase transition between distinct neural states. They argue that the path from wakefulness to anesthesia is smoothly continuous, implying that the eventual recovery of consciousness, once the drug has been eliminated from the central nervous system, will proceed along a trajectory that is the reverse of that followed for induction. This suggests that, aside from pharmacokinetic delays, there should be no hysteresis effects, and the biphasic EEG peaks for entry and exit should occur at the same level of anesthetic concentration.

Molaee-Ardekani et al. (2007) advanced the earlier anesthesia models of Steyn-Ross et al. and Bojak and Liley by incorporating a slow ionic modulation of the firing rate function for the excitatory neural population. This model exhibits three distinct modes of activity: continuous firing in the up-state (awake); phasic firing (bursts followed by periods of silence) as the neurons cycle between high- and low-states (anesthetized); and continuous silence in the down-state (deep coma). Their numerical simulations show good qualitative agreement with EEG traces recorded from children undergoing desflurane anesthesia. To give an overview of the literature, Foster et al. (2008) provide a thorough review of mean-field cortical models.

Hutt and Longtin (2010) have developed a neural field model to investigate anesthetic-induced biphasic changes in EEG power. They derive analytical expressions for the EEG power spectra for both the case of a single solution (similar to the model of Bojak and Liley) and the multi-stable case (considered by Styen-Ross et al.). Their work demonstrates that biphasic power surges can occur in both cases. Thus their model can exhibit a first-order anesthetic phase transition similar to that of Steyn-Ross, and also a smoothly continuous transition like that of Bojak and Liley. This commonality is attributed to the fact that they share several major elements: (i) interactions between excitatory and inhibitory neural populations, (ii) a nonlinear (sigmoidal) mapping between soma voltage and firing rate and (iii) distinct response functions at excitatory and inhibitory synapses.

Up to these studies, population models for anesthesia have neglected contributions from subcortical structures such as the thalamus (Hudetz and Alkire (2011)). Hindriks and van Putten (2012) have considered a thalamocortical feedback in a mean-field model. The authors are able to demonstrate a boost in alpha and delta power during the early stages of anesthesia and explain it in terms of preferential propofol binding at cortical inhibitory neurons leading to disinhibition in the rest of the thalamocortical network. In this context, it is interesting to note that Hutt (2013) has shown that a purely cortical model assuming interactions of excitatory and inhibitory neurons can explain the power surge in either the \(\alpha\)-band or the \(\delta\)-band separately. This indicates that the two activity surges in the different bands may result from an interaction of two systems involving excitation and inhibition. The feedback loop between the thalamus and the cortex might represent such an interaction since the thalamus exhibits an excitation-inhibition circuit between relay and reticular structures.


Alkire, M. T. and Miller, J. (2005) General anesthesia and the neural correlates of consciousness. Prog. Brain Res. 150, 229-244.

Alkire, M. T., Hudetz, A. G., and Tononi G. (2008) Consciousness and anesthesia. Science 322 (5903):876-80.

Bojak, I., and Liley, D. T. J. (2005). Modeling the effects of anesthesia on the electroen- cephalogram. Phys. Rev. E 71: 041902.

Campagna, J. A., Miller, K. W., and Forman, S. A. (2003). Mechanisms of actions of inhaled anesthetics. N Engl J Med 348(21): 2110-24.

Changeux, J. P. (2012). Conscious processing: implications for general anesthesia. Curr Opin Anaesthesiol 25(4): 397-404.

Ching, S., Cimenser, A., Purdon, P.L., Brown, E.N., and Kopell, N.J. (2010). Thalamocortical model for a propofol-induced alpha rhythm associated with loss of consciousness. Proc. Natl. Acad. Sci. 107(52): 22665-70.

Ching, S., Purdon, P.L., Vijayan, S., Kopell, N., and Brown E.N. (2012). A neurophysiological-metabolic model for burst suppression. Proc Natl Acad Sci USA 109(8):3095-3100.

Coppens, M., Van Limmen, J.G.M., Schnider, T., Wyler, B., and Bonte, S. (2010). Study of the time course of the clinical effect of propofol compared with the time course of the predicted effect-site concentration: Performance of three pharmacokinetic-dynamic models. Br. J. Anaesth. 104(4):452-458.

Destexhe, A., Contreras, D., and Steriade M. (1999). Cortically-induced coherence of a thalamic-generated oscillation. Neuroscience 92: 427-443.

Franks, N. P., Lieb, W. R. (1994). Molecular and cellular mechanisms of general anaesthesia. Nature 367, 607–613.

Garcia, P. S., Kolesky, S. E., and Jenkins, A. (2010). General Anesthetic Actions on GABAA Receptors. Curr Neuropharmacol. 8(1): 2–9.

Grasshoff, C., Drexler B., Rudolph U. and Antkowiak B. (2006). Anaesthetic drugs: linking molecular actions to clinical effects. Curr Pharm Des 12(28): 3665-79.

Hara, K., and Harris, R. A. (2002). The Anesthetic Mechanism of Urethane: The Effects on Neurotransmitter-Gated Ion Channels. Anesthesia and Analgesia 94(2): 313-318.

Hindriks, R., and van Putten, M. J. A. M. (2012). Meanfield modeling of propofol-induced changes in spontaneous EEG rhythms. Neuroimage 60 (4): 2323–2334.

Hudetz A. and Alkire M. (2011), Thalamocortical system and anesthetic-induced unconsciousness, In: Neuroscientific Foundations of Anesthesiology, Mashour G.A. and Lydic R. (Eds.): 45-61.

Hudetz, A. G., Vizuete, J. A. and Pillay, S. (2011). Differential effects of isoflurane on high-frequency and low-frequency γ oscillations in the cerebral cortex and hippocampus in freely moving rats. Anesthesiology 114(3):588-95.

Hutt, A. and Longtin, A. (2010). Effects of the anesthetic agent propofol on neural populations. Cog. Neurodyn. 4 (1): 37–59.

Hutt, A. (2013) The anaesthetic propofol shifts the frequency of maximum spectral power in EEG during general anaesthesia: analytical insights from a linear model. Front. Comput. Neurosci. 7:2.

Faulkner, H. J., Traub, R.D., Whittington M.A. (1999) Anaesthetic/amnesic agents disrupt beta frequency oscillations associated with potentiation of excitatory synaptic potentials in the rat hippocampal slice. Br J Pharmacol. 128(8):1813-25.

Friedman, E. B., Sun, Y., Moore, J. T., Hung, H.-T., Meng, Q. C., Perera, P., Joiner, W. J., Thomas, S. A., Eckenhoff, R. G., Sehgal, A. and Kelz, M. B. (2010). A conserved behavioral state barrier impedes transitions between anesthetic-induced unconsciousness and wakefulness: evidence for neural inertia. PLoS One 5 (7): e11903.

Kitamura, A., Marszalec, W., Yeh, J. Z. and Narahashi, T. (2002). Effects of halothane and propofol on excitatory and inhibitory synaptic transmission in rat cortical neurons. J. Pharmacol. 304 (1): 162–171.

Kroeger, D., and Amzica, F. (2007). Hypersensitivity of the anesthesia-induced comatose brain. J Neurosci 27(39): 10597-607.

Kuizenga, K., Wierda, J. M. K. H., Kalkman, C. J., (2001). Biphasic EEG changes in relation to loss of consciousness during induction with thiopental, propofol, etomidate, midazolam or sevoflurane. Br. J. Anaesthesia 86, 354–360.

Lambooy, E. (1985). Electroanaesthesia or electroimmobilisation of calves, sheep and pigs by the Feenix Stockstill. Vet Q. 7(2):120-126.

Lu, J., Nelson L. E., Franks N., Maze M., Chamberlin N.L. and Saper C.B. (2008). Role of endogenous sleep-wake and analgesic systems in anesthesia. J Comp Neurol 508(4): 648-62.

Ludbrook, G. L., Upton, R. N., Grant, C. and Martinez, A. (1999). Prolonged dysequilibrium between blood and brain concentrations of propofol during infusions in sheep. Acta Anaesthesiol. Scand. 43(2): 206-211.

McCarthy, M.M., Brown, E.N. and Kopell, N.K. (2008). Potential network mechanisms mediating electroencephalographic beta rhythm changes during propofol-induced paradoxical excitation. J. Neurosci. 28(50): 13488-13504.

Molaee-Ardekani, B., Senhadji, L., Shamsollahi, M. B., Vosoughi-Vahdat, B., and Wodey, E. (2007). Brain activity modeling in general anesthesia: enhancing local mean-field models using a slow adaptive firing rate. Phys. Rev. E 76: 041911.

Monk, T. G., Saini, V., Weldon, B. C., and Sigl, J. C. (2005). Anesthetic management and one-year mortality after noncardiac surgery. Anesth. Analg 100:4–10.

Roberts, F. (2007). Pharmacokinetics and anaesthesia. Contin Educ Anaesth Crit Care Pain 7(1):25-29.

San-juan, D., Chiappa, K.H. and Cole A.J. (2010). Propofol and the electroencephalogram. Clin Neurophysiol. 121(7):998-1006.

Sanders, R. D., Tononi G., Laureys S. and Sleigh J.W. (2012). Unresponsiveness not equal unconsciousness. Anesthesiology 116(4): 946-59.

Sleigh, J. W. (2011). Depth of anesthesia: perhaps the patient isn't a submarine. Anesthesiology 115(6): 1149-50.

Steyn-Ross, M. L., Steyn-Ross, D. A., Sleigh, J. W. and Liley, D. T. J. (1999). Theoretical electroencephalogram stationary spectrum for a white-noise-driven cortex: Evidence for a general anesthetic-induced phase transition. Phys. Rev. E 60: 7299–7311.

Steyn-Ross, M. L., Steyn-Ross, D. A., and Sleigh, J. W., (2004). Modelling general anaesthesia as a first-order phase transition in the cortex. Prog Biophys Mol Biol 85 (2-3): 369–385.

Steyn-Ross, D. A., Steyn-Ross, M. L., Wilson, M. T. and Sleigh, J. W. (2006). White-noise susceptibility and critical slowing in neurons near spiking threshold. Phys. Rev. E 74: 051920.

Zhu, C., Gao, J., Karlsson, N., Li, Q., Zhang, Y., Huang, Z., Li, H., Kuhn, H. G., and Blomgren, K. (2010). Isoflurane anesthesia induced persistent, progressive memory impairment, caused a loss of neural stem cells, and reduced neurogenesis in young, but not adult, rodents. J. Cereb. Blood Flow Metab. 30(5):1017-30.

See also

Vegetative State

High-Conductance State


Thalamocortical oscillations

Up and down states

Recommended reading

Brown, E.N., Purdon, P.L., and Van Dort, C. (2011). General anesthesia and altered states of arousal: a systems neuroscience analysis. Annual Review of Neuroscience 34:601-628.

Foster, B. L., Bojak, I. and Liley, D. T. J. (2008). Population based models of cortical drug response: insights from anaesthesia. Cog. Neurodyn. 2 (4): 283–296.

Mashour G.A. and Lydic R. (2011). Neuroscientific Foundations of Anesthesiology. Oxford University Press, USA.

McCarthy, M. M., Ching, S., Whittington, M.A.,and Kopell, N. (2012). Dynamical changes in neurological disease and anesthesia. Current Opinion in Neurobiology 22(4):693-703.

Personal tools

Focal areas