Elsevier

Clinical Neurophysiology

Volume 116, Issue 12, December 2005, Pages 2719-2733
Clinical Neurophysiology

Invited review
Human EEG gamma oscillations in neuropsychiatric disorders

https://doi.org/10.1016/j.clinph.2005.07.007Get rights and content

Abstract

Due to their small amplitude, the importance of high-frequency EEG oscillations with respect to cognitive functions and disorders is often underestimated as compared to slower oscillations. This article reviews the literature on the alterations of gamma oscillations (about 30–80 Hz) during the course of neuropsychiatric disorders and relates them to a model for the functional role of these oscillations for memory matching. The synchronous firing of neurons in the gamma-band has been proposed to bind multiple features of an object, which are coded in a distributed manner in the brain, and is modulated by cognitive processes such as attention and memory. In certain neuropsychiatric disorders the gamma activity shows significant changes. In schizophrenic patients, negative symptoms correlate with a decrease of gamma responses, whereas a significant increase in gamma amplitudes is observed during positive symptoms such as hallucinations. A reduction is also observed in Alzheimer's Disease (AD), whereas an increase is found in epileptic patients, probably reflecting both cortical excitation and perceptual distortions such as déjà vu phenomena frequently observed in epilepsy. ADHD patients also exhibit increased gamma amplitudes. A hypothesis of a gamma axis of these disorders mainly based on the significance of gamma oscillations for memory matching is formulated.

Introduction

Since the discovery of the electroencephalogram (EEG) by Hans Berger, 1929, oscillatory patterns can be observed in the brain electrical activity (Berger, 1929). The most prominent oscillation in the spontaneous EEG exists in a frequency band of 8–12 Hz, which was considered by Berger as the basic rhythm and was named α-rhythm. Alpha oscillations appear with well noticeable amplitudes between 10 and 50 μV and have multiple cognitive correlates (Basar et al., 1997). The chronologically next identified frequency range between 12 and 30 Hz was named by Berger consequently with the Greek letter β. Faster oscillations in the human EEG between 30 and 80 Hz could later be identified and were named as gamma activity (Chatrian et al., 1960). The slower waves below the alpha range were first named as δ and afterwards were divided into the delta (0–4 Hz) and theta (4–8 Hz) ranges. Since the amplitudes of the EEG oscillations decrease with increasing frequencies (Fig. 1), higher frequency bands such as the omega range (80–120 Hz) were identified later. Today it is known that oscillations with frequencies up to 600 Hz exist in the EEG (Curio et al., 1994).

The functional significance of brain activity in the alpha, theta and delta frequency bands and event-related oscillations within this frequency range were discovered relatively early, because these slow waves can easily be observed in the EEG (Basar, 1980, Basar et al., 1997, Demiralp and Basar, 1992, Steriade et al., 1990). Oscillations in the beta band and in higher frequency bands could, however, be revealed adequately only with the use of special amplifiers and analysis techniques due to their small amplitudes.

An important property of brain oscillations in the brain is that they can show phase relations either in terms of a phase synchronisation among various oscillators or a phase-coupling to a transient event. The discussion of these properties is important for understanding possible functional meanings of oscillatory activity in the brain (Varela et al., 2001). Within this framework, three types of phase synchrony can be described for brain electrical signals: inter-neuronal, inter-electrode and inter-trial.

The inter-neuronal phase synchronisation represents a local synchronisation among neurons in a relatively small area of some millimeters, such that their membrane potentials oscillate in phase and/or they fire synchronously with each other. This type of synchrony can be investigated by measuring either the membrane potentials of a number of single cells simultaneously or by measuring multiunit activity and local field potentials using electrodes in close proximity. On the other hand, this type of synchrony is very basic for the generation of the EEG, because only synchronous activity of a large set of neurons can reach the skull and therefore, each EEG electrode records the spatial sum of such synchronous activity of a large number of oscillating neurons or neural circuits (Basar, 1980, Steriade et al., 1990). Therefore, EEG oscillations result from brain activity with high inter-neuronal synchrony. However, either more neurons with the same degree of inter-neuronal synchrony or the same number of neurons with a higher degree of inter-neuronal synchrony may lead to increased EEG amplitudes.

On the other hand, phase synchrony may also exist on a larger scale, and the electrical signals from distant electrodes may contain coherent oscillations with 0 or constant phase shift (Varela et al., 2001). Such synchronous activity among distant parts of the brain yields high values (close to 1) in the coherence function. This type of synchrony is important in the detection of anatomically distant, however, functionally closely related brain structures that are interacting either through simultaneous communication with the same sub-cortical structure (0 phase shift) or among each other through long cortico-cortical connections (constant phase shift). It is called inter-electrode synchrony and measured via coherence functions.

The third type of phase synchronisation, inter-trial synchrony, is especially important in the analysis of event-related oscillations, which can only be analyzed through statistical approaches on the data when the same event is repeated a number of times (trials). In this case, two different types of oscillatory responses can be obtained (Basar, 1980, Herrmann et al., 2005). The so-called ‘evoked’ oscillations are phase-coupled to the triggering event, hence start within a preferred range of phase angles after the external event. This type of responses result in clear oscillations in the average of the trials, even if no amplitude enhancement is present in single trials. A second type of event-related oscillations that are termed ‘induced’ responses show an amplitude increase in single trials, however, do not occur with a constant phase lag after the triggering event. Therefore, they are not observable in the averaged response, in case many trials have been averaged. However, they can be quantified by transforming the single trials into the frequency domain first, and then averaging the transformed data after removing the phase information. Due to their different temporal dynamics, both types of event-related oscillations may correspond to different functions, although their generation might depend on the same or similar circuits.

In recent years, a special interest for oscillations in the gamma frequency range has emerged in the neurosciences (Basar-Eroglu et al., 1996). The interest for these oscillations depends on the fact that gamma activity is closely correlated with cognitive functions (Engel et al., 2001).

Various gamma phenomena can be grouped into the following categories (Galambos, 1992): Spontaneous gamma oscillations, that contribute a fraction of the total EEG/MEG power at any given moment have been explained by thalamocortical resonant synaptic interactions (Llinas and Ribary, 1992) and have been assumed to reflect the consciousness level. On the other hand, a number of studies based on multiunit activity or field potential measurements in sensory cortices of various species showed induced gamma oscillations that follow a sensory stimulus, but are not phase-locked to the stimulus (Engel et al., 1992), i.e. do not show any inter-trial phase synchronisation. Induced gamma activity has been interpreted to reflect feature-binding processes and to generate a neural representation of the stimulus by integrating the different features encoded in different neuronal maps. However, not only the features that are represented by synchronously firing neurons are integrated to a coherent object, but also object representations and related motor activity are integrated by the gamma oscillations (Roelfsema et al., 1997). Evoked gamma responses that are most consistently recorded from human scalp are occurring in an earlier time window than induced gamma oscillations and are phase-locked to the stimulus, i.e. show inter-trial phase synchrony. It has also been demonstrated that simple sensory stimuli evoke such gamma responses in the cortex and subcortical structures of the animal brain (Basar, 1980, Demiralp et al., 1996). Steady-state gamma oscillations are the driven electrical responses, of the brain obtained by the application of repetitive stimuli like clicks, tone pips or an amplitude modulated tone which reach their maximum at repetition rates around 40 Hz (Galambos et al., 1981). Later, Pantev et al. (1991) used magnetoencephalographic measurements to reveal that the generators of the steady-state gamma response (SSR) and the transient gamma band response overlap and concluded that the SSR to auditory stimuli is essentially the overlapping sum of the successive transient gamma band responses.

Spontaneous gamma activity is recorded without any task for the subjects/patients. In order to record steady-state responses, a repetitive stimulation (e.g. visual or auditory) needs to be supplied, which repeats at the frequency of interest, i.e. 40 Hz. Evoked and induced gamma oscillations occur as transient event-related oscillations (EROs) in experimental paradigms exploring cognition, which at the same time evoke event-related potentials (ERPs, cf. Fig. 2).

While occurring within different contexts and with different temporal dynamics and relations to sensory and cognitive events, there is evidence that spontaneous, steady-state, evoked and induced gamma oscillations might be generated by the same neural circuits (Basar, 1980, Herrmann, 2001, Pantev et al., 1991). Basar (1980) used intracranial measurements of field potentials to demonstrate that brain structures respond to transient stimuli with enhancement of oscillations within those frequency ranges, which are present in the power spectrum of the spontaneous electrical activity of that brain structure. This phenomenon, which the author calls response susceptibility, shows a close relationship between the generators of spontaneous and evoked or induced oscillations. Later, Herrmann (2001) demonstrated that resonance phenomena existed in visual steady-state responses at the same frequencies where evoked oscillations usually occur. Additionally, the sources of auditory steady-state responses were located in primary auditory cortex (Gutschalk et al., 1999, Herdman et al., 2002), which is also true for the sources of event-related gamma activity revealed by intracranial recordings in monkeys (Brosch et al., 2002) and humans (Crone et al., 2001). Along the same lines, visual steady-state responses were located in human visual cortex (Hillyard et al., 1997, Müller et al., 1997), where also event-related gamma activity was found in intracranial recordings in monkeys (Fries et al., 2001, Rols et al., 2001) as well as in humans (Tallon-Baudry et al., 2005). Furthermore, the cognitive correlates of evoked and induced gamma responses generally overlap (Debener et al., 2003, Engel et al., 2001, Fries et al., 2001, Herrmann et al., 2004, Herrmann et al., 2004, Tiitinen et al., 1993, Yordanova et al., 1997, Yordanova et al., 1997). Therefore, it seems plausible to assume that similar generation mechanisms are responsible for both types of gamma activity. On this basis, we will not separate the different measurement modalities and review the different types of gamma oscillations within a general frame.

Section snippets

Gamma activity reflects memory matching

A number of studies have shown that gamma oscillations are modulated by a variety of cognitive processes such as attention, object recognition, and working memory (Debener et al., 2003, Fries et al., 2001, Herrmann and Mecklinger, 2001, Herrmann et al., 2004, Herrmann et al., 2004, Tiitinen et al., 1993, Yordanova et al., 1997, Yordanova et al., 1997). Therefore, gamma activity is assumed to reflect an integration mechanisms of the brain (Herrmann et al., 2004, Herrmann et al., 2004).

Neurons in

Changes in gamma activity in neuropsychiatric disorders

As these high frequency oscillations with small amplitudes can be better registered with modern technology, they were intensively investigated only recently. It turned out, that there are significant interindividual deviations in gamma activity that correlate with cognitive parameters (Strüber et al., 2000). Additionally, it was demonstrated that an excess or deficiency of this ‘cognitive‘ activity is characteristic for certain pathological conditions. Both phenomena were to be expected, if

Gamma oscillations in epilepsy

Epilepsy is a functional disorder of the brain due to excessive neuronal discharges. In ICD-10 (International Classification of Diseases, 10th Revision, Geneva: World Health Organization, 1992) this disorder is classified under the disorders of the nervous system (G40). Epilepsy is often accompanied by psychiatric symptoms (Bruton et al., 1994, Mace, 1993, Mendez et al., 1993). Epilepsy can be differentiated from other seizure-like disorders through the use of the EEG. During a seizure the EEG

Gamma oscillations in ADHD

ICD-10 lists ADHD as a hyperkinetic disorder (F90) and it is characterized by usually starting within the first 5 years of life. Patients often exhibit a deficiency in persevering for occupations that require a cognitive effort. Additionally, the patients have a tendency to change from one activity to another without finishing any of them. DSM-IV (Diagnostic and Statistical Manual of Mental Disorders, Washington: American Psychiatric Association, 1994) divides ADHD into predominantly

Gamma oscillations in Schizophrenia

Schizophrenia (DSM-IV: 295.x, ICD-10: F20.x) is generally characterized by a disturbance of perception (illusions and hallucinations), disorganized thoughts, and ‘blunted’ or ‘flat’ affects (McCarley et al., 1993). One of the central deficits in this disease is considered to be a lack of the integration of sensory input with the information stored in memory (Gray, 1998). This integration disorder gets especially evident, when the patients suffer from hallucinations, during which the lack of a

Gamma oscillations in Alzheimer's disease

Alzheimer's disease (AD) is the most frequent type of primary degenerative dementias affecting 5–10% of the population above the age of 65 (ICD-10: F.00). Amnesia is the earliest, most frequent and severe symptom in AD. Clinically, the disease is characterized by progressive amnesia, which is followed by a slow decline in all cognitive domains resulting in global dementia (DSM-IV, 294.1x). AD is caused by the degeneration of nerve-cells in the brain and their replacement by elements known as

The hypothesis of a gamma-axis of neuropsychiatric disorders

While the data reported above reflects the state of the art without any integrative interpretation, we will now suggest a hypothesis that integrates these findings in a speculative but functionally meaningful manner. For the understanding of neuropsychiatric disorders, it is important to have a model of the correctly working cognitive functions and their changes during pathology (Gordon, 2001).

The above-mentioned results demonstrate vividly that the significance of gamma activity in

Concluding remarks

As a conclusion, it should be shortly discussed here what type of hypotheses the proposed gamma-axis of psychiatric disorders generates, how they could be verified, and what type of analyses could be carried out for their verification. On the one hand, a physiological e.g. neuronal model could explain the relationships between different disorders, which are already observed such as the increased risk for epileptic seizures in ADHD patients. On the other hand, we can line up further hypotheses.

Acknowledgements

Tamer Demiralp's stay in Magdeburg (Germany) during the preparation of this manuscript was supported by the Alexander von Humboldt Foundation. We want to thank Stefanie Junge and Daniel Lenz for preparing some of the figures.

References (166)

  • G. Buzsaki

    Theta oscillations in the hippocampus

    Neuron

    (2002)
  • G.E. Chatrian et al.

    Depth electrographic study of a fast rhythm evoked from the human calcarine region by steady illumination

    Electroencephalogr Clin Neurophysiol

    (1960)
  • B.A. Clementz et al.

    Aberrant brain dynamics in schizophrenia: delayed buildup and prolonged decay of the visual steady-state response

    Brain Res Cogn Brain Res

    (2004)
  • N.E. Crone et al.

    Induced electrocorticographic gamma activity during auditory perception

    Clin Neurophysiol

    (2001)
  • G. Curio et al.

    Localization of evoked neuromagnetic 600 Hz activity in the cerebral somatosensory system

    Electroencephalogr Clin Neurophysiol

    (1994)
  • T. Demiralp et al.

    Theta rhythmicities following expected visual and auditory targets

    Int J Psychophysiol

    (1992)
  • A.K. Engel et al.

    Temporal coding in the visual cortex: new vistas on integration in the nervous system

    Trends Neurosci

    (1992)
  • J. Gallinat et al.

    Reduced oscillatory gamma-band responses in unmedicated schizophrenic patients indicate impaired frontal network processing

    Clin Neurophysiol

    (2004)
  • E. Gordon

    Integrative psychophysiology

    Int J Psychophysiol

    (2001)
  • M.F. Green et al.

    Visual masking as a probe for abnormal gamma range activity in schizophrenia

    Biol Psychiatry

    (2003)
  • T. Gruber et al.

    Selective visual-spatial attention alters induced gamma band responses in the human EEG

    Clin Neurophysiol

    (1999)
  • A. Gutschalk et al.

    Deconvolution of 40 Hz stead-state fields reveals twooverlapping source activities of the human auditory cortex

    Clin Neurophysiol

    (1999)
  • A.R. Haig et al.

    Gamma activity in schizophrenia: evidence of impaired network binding?

    Clin Neurophysiol

    (2000)
  • C.S. Herrmann et al.

    Cognitive functions of gamma activity: memory match and utilization

    Trends Cogn Sci

    (2004)
  • L.E. Hong et al.

    Evoked gamma band synchronization and the liability for schizophrenia

    Schizophr Res

    (2004)
  • Y. Izaki et al.

    Evoked prefrontal gamma oscillation by hippocampal train stimulation in anesthetized rats

    Neurosci Lett

    (2003)
  • J. Kaiser et al.

    Dynamics of gamma-band activity in human magnetoencephalogram during auditory pattern working memory

    Neuroimage

    (2003)
  • S. Karakas et al.

    Early gamma response in human neuroelectric activity is correlated with neuropsychological test scores

    Neurosci Lett

    (2003)
  • A. Keil et al.

    Effects of emotional arousal in the cerebral hemispheres: a study of oscillatory brain activity and event-related potentials

    Clin Neurophysiol

    (2001)
  • J. Kissler et al.

    MEG gamma band activity in schizophrenia patients and healthy subjects in a mental arithmetic task and at rest

    Clin Neurophysiol

    (2000)
  • W. Klimesch

    EEG alpha and theta oscillations reflect cognitive and memory performance: a review and analysis

    Brain Res Brain Res Rev

    (1999)
  • W. Klimesch et al.

    Theta oscillations and the ERP old/new effect: independent phenomena?

    Clin Neurophysiol

    (2000)
  • W. Klimesch et al.

    Episodic retrieval is reflected by a process specific increase in human electroencephalographic theta activity

    Neurosci Lett

    (2001)
  • T. Koenig et al.

    Decreased EEG synchronization in Alzheimer's disease and mild cognitive impairment

    Neurobiol Aging

    (2005)
  • G.P. Krishnan et al.

    Steady state visual evoked potential abnormalities in schizophrenia

    Clin Neurophysiol

    (2005)
  • N. Laporte et al.

    Cognitive epilepsy: ADHD related to focal EEG discharges

    Pediatr Neurol

    (2002)
  • K.H. Lee et al.

    Synchronous gamma activity: a review and contribution to an integrative neuroscience model of schizophrenia

    Brain Res Brain Res Rev

    (2003)
  • S.C. Li et al.

    Aging cognition: from neuromodulation to representation

    Trends Cogn Sci

    (2001)
  • I. Lussier et al.

    Memory and attention deficits in drug naive patients with schizophrenia

    Schizophr Res

    (2001)
  • J. Ma et al.

    Relation between hippocampal gamma waves and behavioral disturbances induced by phencyclidine and methamphetamine

    Behav Brain Res

    (2000)
  • J. Ma et al.

    Metabotropic glutamate receptors in the hippocampus and nucleus accumbens are involved in generating seizure-induced hippocampal gamma waves and behavioral hyperactivity

    Behav Brain Res

    (2002)
  • C. Madler et al.

    Sensory information processing during general anaesthesia: effect of isoflurane on auditory evoked neuronal oscillations

    Br J Anaesth

    (1991)
  • B.K. Madras et al.

    The dopamine transporter: relevance to attention deficit hyperactivity disorder (ADHD)

    Behav Brain Res

    (2002)
  • C. Babiloni et al.

    Abnormal fronto-parietal coupling of brain rhythms in mild Alzheimer's disease: a multicentric EEG study

    Eur J Neurosci

    (2004)
  • E. Basar

    EEG-brain dynamics

    (1980)
  • H. Berger

    Über das Elektrenkephalogramm des Menschen

    Archiv für Psychiatrie und Nervenkrankheiten

    (1929)
  • N.N. Boutros et al.

    Replication and extension of P50 findings in schizophrenia

    Clin Electroencephalogr

    (1991)
  • M. Brosch et al.

    Stimulus-related gamma-oscillations in primate auditory cortex

    J Neurophys

    (2002)
  • A.S. Brown

    A review of the deja vu experience

    Psychol Bull

    (2003)
  • C.J. Bruton et al.

    Epilepsy, psychosis, and schizophrenia: clinical and neuropathologic correlations

    Neurology

    (1994)
  • Cited by (460)

    View all citing articles on Scopus
    View full text