What Are Gamma Waves (And How Are They Relevant to Clinical Psychiatry)?

Abstract

This review raises the importance of understanding the functions of gamma waves in the practice of clinical psychiatry. Measured by electroencephalograph (EEG), gamma waves are those electrical waves between 30 Hz and 200 Hz, and are relevant to psychiatry because of their ability to influence other brain electrical activity, and thence have associations with various mental disorders. To inform readers, gamma waves are defined and described, and some examples of their association with psychiatric disorders are briefly recounted. Four ways in which gamma waves might modulate brain activity and influence behaviour, potentially leading to a comprehensive model of gamma and its association with mental illness, are discussed. Inconsistencies in definition, measurement, and findings from gamma wave research are also described, and suggestions made for overcoming these limitations in future research.

1. Psychiatry and the Brain

At the beginning of the current century, several major position papers urged that psychiatry should consider itself as based in neuroscience [1,2,3,4]. There has also been some comment that social aspects of human interaction also need to contribute to that perspective [5,6], so the emerging focus of clinical psychiatry has been upon understanding the way the social environment influences the brain within a biopsychosocial model of behaviour [7]. Central to this model is understanding how the brain works and how it influences the observable behaviour that is the target of most clinical assessment regimes [8,9]. As such, neuroimaging techniques have been increasingly used to detect normal and abnormal brain structure and functions that are associated with those observable symptoms of psychiatric disorders [10,11]. Some of those techniques (such as Magnetic Resonance Imaging: MRI) focus upon brain structure [12,13], others use the electrical activity of the brain as measured via electroencephalography (EEG) to attempt to identify the neural correlates of psychiatric disorders [14,15]. These are often accompanied by advanced computing resources that extend understanding of brain activity from surface electrodes to deeper brain structures [16,17]. In these models, waves of brain electrical activity can help to identify the kinds of cognitive activities that may be occurring within the brain, and how these relate to observable behaviour/symptomatology.

For example, scalp-measured ‘alpha’ wave activity (8–13 Hz) is commonly but not exclusively associated with relaxed, eyes-closed mental states [18]. Faster (13–30 Hz) activity (referred to as ‘beta’ waves) is often associated with active mental processing, mental concentration and/or problem-solving [19], although these associations depend on task demands, region, and individual factors. The relative power (i.e., the amplitude or strength of brain wave activity within these bands) of these two wave forms has been used to detect different cognitive states that may be associated with various disorders [14,20,21]. One relatively understudied but potentially influential wave form is described by the term ‘gamma’ waves, which are classified as occurring between 30 Hz and 200 Hz [22], although that range has sometimes been divided into ‘classic’ gamma (30–80 Hz) and ‘high’ gamma (80–200 Hz) [23]. Fitzgerald and Watson [22] argued that 30 Hz to 200 Hz gamma waves held major potential for the identification and understanding of depression [22]. Other psychiatric disorders that have been similarly linked with gamma include schizophrenia [24,25], anxiety, and autism [26], as well as a range of neurological disorders [27,28,29]. The potential for gamma to contribute to our understanding of psychiatric disorders lies with its (i) characteristics, (ii) functions, and (iii) and associations with relevant cognitive behaviour. However, despite the potential value of these aspects, gamma wave activity is relatively unfamiliar to medical, psychiatric, psychological, and mental health specialists [22], arguing for a clarification of gamma, and its relevance to these fields. Firstly, it is necessary to understand what gamma is, where it is initiated, how it is related to mental illness, and how research might be conducted into gamma’s role in clinical psychiatry. These issues are the foci of this review. As a caveat, this review is not intended to be exhaustive in the same way as a meta-analysis or a systematic review, nor to delve into the challenges of gamma wave research or the overall limitations of reliable EEG-based biomarkers. Instead, the intention here is to bring gamma to the attention of the clinician, explain its origins, and how it may be related to psychiatric illness, but not to include such details as specific experimental setups, task designs, or EEG analysis pipelines.

2. What Are Gamma Waves?

Gamma waves, named because of their wave-like variability in amplitude of the power of the electrical signal emitted by neurons between 30 Hz and 200 Hz, were first measured in humans by Berger in 1924, using the EEG [30]. Gamma waves can be observed across multiple brain regions [31] and usually occur in sudden bursts which may appear to be random and short-lived [32]. They may be initiated by clusters of neurons sometimes referred to as circuit generators [33], based on the interaction between excitatory and inhibitory neurons as a result of ionic variations within synaptic zones [32] (see Figure 1). As well as possibly emerging from long-range (inter-areal) synchronization of neural networks, gamma waves may be indicative of the coordinated neural activity across brain regions that enables system-based communication [34]. As described by Jia and Kohn [23] “a prominent gamma rhythm provides a signature of engaged networks” (p. 1), although gamma at a specific site on the scalp may be a response to a particular stimulus or nearby neural spiking [35]. For example, when measured from the higher cortex, gamma power is directly associated with cognitive activities that entail working memory [36] and learning [37].

The finding that gamma waves can be correlated with local and distant neural centres may reflect systematic firing of adjacent and distal neural circuits [38,39]. These possibly serve as an indicator of various cognitive processes and synaptic signaling [40], enhanced intercellular and system-wide communication [41], coordinated neural spike firing, entrainment, and inhibitory/excitatory cycling, sensory inputs parsing [26], and attentional focus [42]. However, gamma waves differ from other frequency-band waves in that they are not associated with specific cognitive states and regions in the way (for example) alpha waves are often linked with relaxation, and beta waves may often occur during concentrated thinking. By contrast, gamma waves have been associated with a range of different cognitive states including vegetative states [43], heightened arousal or cognitive processing [26,44], and neural synchrony [45] (i.e., the correlation of neural activity in different brain regions), suggesting that they have a complex and multi-faceted influence on neural functioning.

This variability in the associations between gamma waves and cognitive state implies that gamma per se does not perform specific tasks but rather may facilitate those tasks as they are performed by separate sets of neurons. That is, as explained by Nyhus and Curren some time ago [46], connections between distinct brain regions enable the formation of “complex and dynamic brain networks” (p. 1023). Local populations of neurons process information at a basic content level, but it is neural assemblies that include multiple sets of these local neural populations that enable more complex cognitive tasks to be achieved. These large neural assemblies may be initiated by slowly changing connections based upon neuroanatomy, but also by transient connections that are facilitated by phase synchronization of neural waves across particular neural assemblies in different brain regions [47]. These waves represent opportunities for the connections between neural assemblies so that, for example, when assemblies A and B each oscillate to a firing state, they are more likely to connect, producing a more complex set of neural assemblies, and supporting more complex cognitive activity. If assembly A reaches the firing state while assembly B does not, then connectivity is unlikely [48,49].

The connections between neural assemblies may occur at a local level, where the close connections are occasioned during faster, higher frequency brain wave activity (e.g., gamma), or at more distant levels, when the connections are supported by slower frequency brain waves (e.g., theta) [50]. There may also be situations where the slower wave activity provides a basis or framework for faster wave activity that leads to connectivity and transmission of signals between assemblies [51,52]. Despite the distance involved, it is the timing of individual assemblies’ waves that is the key to the likelihood of connectivity. Concurrent with this model of synchrony of neural activity enabling connectivity between neural assemblies, it has been established for some time that increased gamma power is a key influence upon that synchronization process [53,54,55].

3. Gamma’s Influence on Mental Health and Illness

Investigation of this complex and multi-faceted model of neural functioning has produced data regarding the positive correlation between gamma and mental health. For example, regular gamma waves may be indicative of particular states of psychological health via their association with neuromodulatory functions across brain networks [33,40,44,56]. Emerging preclinical and early human evidence suggests that stimulation of activity in the range of gamma frequencies may have a therapeutic effect on several brain disorders [31,57]. These direct associations between gamma wave activity and cognitive function are exemplified by Gray et al.’s [58] finding of a correlation between gamma activity of about 40 Hz and the binding of the features of a visual stimulus in the cat visual cortex, leading to the positing of the ‘binding-by-synchrony’ hypothesis [59]. That is, synchronized gamma waves contribute to coordination of different neural assemblies’ responses to particular aspects of a visual stimulus, enabling the organism to develop a ‘picture’ of the entire object rather than just a part of it [59]. This finding was confirmed in more detail by Han et al. [60], who noted that different aspects of the information from the visual stimulus were carried by different gamma wave sub-bands, referred to by Han et al. as low gamma (25 Hz to 40 Hz), medium gamma (40 Hz to 65 Hz), and high gamma (65 Hz to 85 Hz). Han et al.’s finding that these three sub-bands of gamma were generated from different neural circuits further emphasizes the complex nature of the gamma wave and its influence on healthy neural processing. Uhlhaas et al. [61] described gamma as low (30 Hz to 60 Hz) and high (60 Hz to 200 Hz) on the basis of specific associations with different brain functions and cognitive activities, but it is reasonable to also enquire about the possibility of further breaking down gamma into (for example) 20-Hz ranges (i.e., 30 Hz to 50 Hz, etc.), and investigating these from different brain sites as a means to ‘map’ gamma according to frequency and association with neural and behavioural processes. This inconsistency in the literature about defining gamma wave sub-frequencies is a major contributor to the lack of clear overall research findings (see Section 4: The trouble with gamma, etc.).

As expected from the direct association between gamma waves and neurocognitive function described above, significant direct correlations have been observed between ‘irregular’ gamma wave signals and disorders of cognition and/or sensory processing [29,31,43]. Some ‘irregular’ aspects of the gamma signal that contribute to this association are deficits in the amplitude and synchrony of gamma, perhaps initiated by alterations in the GABAergic neurotransmitter system [61]. Uhlhaas and Singer [62] have described this process as being responsible for the cognitive disruption witnessed in schizophrenia. There is a body of evidence that the lack of synchrony in firing between relevant (i.e., contributing some aspects of the complex neural function underlying a specific cognitive activity) assemblies of neurons disturbs the necessary level of connectivity between those neural assemblies of neurons responsible for complex neurocognitive behaviour [29,33,44], resulting in the kind of disordered thinking that is a characteristic of schizophrenia [9].

This review provides only a brief overview of the association between gamma and psychiatric disorders, but there have been several reviews of that association [see, for example 64 for a review of 31 studies of the association between gamma and depression] [63]. Other reviews have focused on the roles of gamma in other psychiatric disorders [28,64], providing a cogent argument for the relevance of gamma to an understanding (and perhaps treatment) of those disorders. However, all of those reviews share common concerns regarding inconsistencies in definitions of gamma, methods of measurement, and data analysis that can lead to inconsistent results. These are key issues that require addressing before this field can proceed to advantage and must be understood when interpreting and evaluating research on gamma and psychiatric disorders.

4. The Trouble with Gamma: Inconsistencies in Definition, Measurement, and Findings

As mentioned in many reviews, some studies of the link between gamma waves and neural function have been inconclusive. Directly correlating gamma wave variables with cognitive processes in humans is a complex task, with disagreement concerning how gamma waves relate to behaviour [38,42]. Since first being described as high frequency signals potentially emerging from central cortical regions along the midline [65], debate has ensued around the relevance and characterization of gamma waves. For example, it has been argued that, owing to their ubiquity across brain regions, gamma waves may reflect little more than electrochemical epiphenomena [66], i.e., as a consequence of electrochemical interactions which emerge through cortical excitation without specific cognitive function [38,43].

In an effort to explore gamma at more than a unitary frequency range (i.e., between 30 Hz and 200 Hz), it may be that subdividing gamma into narrower frequency bands might provide stronger evidence for the association between some sub-bands of gamma and specific cognitive tasks. Some evidence supporting that suggestion has been reported in regard to perceptual function. For example, Han et al. [67] noted that at least two distinct gamma peaks occurring at different sub-bands (30 Hz to 40 Hz, and 50 Hz to 70 Hz) were associated with visual processing and were likely to relate to different components of visual information. This conclusion was supported by Guan et al. [29], who found that three sub-frequencies of gamma (low: 25 Hz to 40 Hz; medium: 40 Hz to 65 Hz; and high: 65 Hz to 90+ Hz) were attributable to the primary visual cortex where they likely reflected distinct aspects of visual information processing. Supporting the ‘sub-frequency’ argument, a recent review noted that, although some gamma-wave characteristics have been described and linked to higher-order cognitive processing, differences in terms of how gamma frequencies are defined has led to inconsistent results [42], producing disagreement regarding which cognitive processes (if any) are associated with gamma [38]. Furthermore, confounding of results due to data misinterpretation, imprecise signal processing, or the lack of a priori criteria defining the exact frequency bands under investigation have generated additional controversies that further cloud understanding of gamma [33,42].

Perhaps reflecting the overall lack of agreement regarding research protocols for gamma and mental illness, there has also been some controversy regarding the recommended break-down of gamma into sub-band frequencies (mentioned above) when investigating relevant cognitive variables [29,68]. In fact, some researchers have advocated disregarding sub-frequencies of gamma and instead accepting that the complexity and apparent randomness of gamma as a correlate of complex cognitive behaviour is simply a critical component of gamma waves per se e.g., [69]. Disagreements about gamma band definition [42], signal processing [33], and the ways in which event-related, evoked (i.e., changes in gamma power in response to a specific stimulus), induced (i.e., changes in gamma power that are not in response to a specific stimulus) and spontaneous gamma wave activity are measured [42,43] have also been suggested as responsible for some of the inconsistent research findings reported [42]. It is therefore important to acknowledge at an overview level what is known about gamma and psychiatric disorders.

5. What Do We Do Know About Gamma?

Despite the disagreements and controversies regarding research methodology issues mentioned above, significant advances in computer-aided imaging have enabled some valuable findings to be established. For example, exact low-resolution brain electromagnetic tomography (eLORETA) [17], and brain-machine interfaces [33] have allowed for the linking of perceptual [70], cognitive [43], and behavioral [26,44] traits with gamma-band activity. eLORETA provides relatively coarse localization data, which might be interpreted as network-level inferences rather than precise foci. Improvements in terms of instrumentation [42], signal processing [33], optogenetic tracing [38,70], disease modelling [43,71], pharmacological intervention [22], source tracing [26,72], and machine-based learning [67] have aided discovery of gamma’s wider properties. Multimodal technologies such as Gamma Entrainment Using Sensory Stimuli [73] have further revealed the complex psychobiological dynamics involved [44]. Further refinement of methodology is needed for this field to produce the kind of definitive outcomes necessary for clinical practice, but research on one major aspect of gamma’s influence on the brain appears to be moving towards that outcome—that of gamma as an agent for neuromodulation (i.e., the alternation of neural activity) (see Figure 2 for some hypothesized pathways that are targets for future testing and verification).

5.1. Gamma Waves as an Agent for Neuromodulation

Computational modelling and high-resolution neuroimaging techniques have led to the discovery that gamma waves may be associated with neuromodulation within the central nervous system [33,40]. This is a crucial insight, enabling researchers to better understand how gamma waves could reflect the neural substrates that drive homeostasis, with implications for cognitive behaviour and mental health. Although gamma waves occur under a wide variety of conditions [38], they appear to change during situations of intense and demanding cognitive arousal [26], multisensory processing [57], social interaction [74,75], hippocampal and sensory circuit function, memory-making [69], social affiliation [45,76], and perceptual processes [44], plus olfactory [77], visual [43,67,70], and auditory discrimination ([67]. All of these conditions share the characteristic of requiring a degree of modulation of neural processing.

Although gamma waves have been inconsistently associated with specific psychiatric disorders, the major phenomena they may indicate include neural synchrony [39], coherence [41], and modulation [40]. When viewed as indicators of a complex, network-wide framework, gamma waves appear to serve as a key window into the types of neural activities that underly cognition, perception and movement [69]. These, in turn, are reflective of the circuit functions from which perception, subjective discrimination, emotional regulation and attentional focus emerge [42,68,78]. In this way, gamma waves indicate neural processing, intercellular signaling [39,40], and active modulation of behavioural traits [44]. These processes may include communication across brain regions [39], cross-frequency coupling [69], and neurotransmitter expression [22,40,43], which can drive the interplay of neural and neurochemical catalysts within an overall brain/cognitive system [22,44,79,80]. Under these models, gamma waves appear to be indicative of an integrative [33,43] and neuromodulatory function [44] that has implications for mental health. This role is discussed below in four ways in which gamma may be active in neuromodulation.

5.1.1. Modulation Through Synchrony and Coherence

Synchrony refers to the coordinated timing of brainwave activity between different brain regions, considered to be an indicator of functional connectivity and, potentially, network-wide integration [39,81,82]. Evidence supports the hypothesis that gamma waves reflect the synchronous peak firing of local neurons, allowing for the seamless neural processing that is necessary for neural plasticity and adaptation to environmental demands [32,39,40,41,42]. As building blocks of a universal neural code [43], gamma waves provide insight into local and network-wide functions which ‘bind’ neural populations in ways that augment cognitive processing [41]. Unlike other waves, the significance of gamma waves may lie less with detectable changes within specific frequency band characteristics and more in their capacity to suggest how synchronous brain activity is modulated at both local and global levels [33,69].

Coherence is a measure of synchronization [41], recording the consistency of phasic cycles between signals and across neural populations. Coherence reveals how effectively brain electrical activity is coordinated, driving summation within neural clusters and efficient transmission of relevant signals across synaptic spaces, which may explain how attentional focus and intracellular communication unite for instantaneous and accurate cognition [41] while integrating simultaneously occurring, multisensory stimuli [43,72]. High coherence is an indicator of tight coordination between diverse brain regions, allowing neural inputs to summate more effectively, which in turn leads to better aligned action potentials [39]. Coherence makes communication selective, effective and precise, and may be linked with gamma [41,44].

5.1.2. Modulation Through Cross-Frequency Coupling

Gamma waves are also known to be representative of cross-frequency coupling, allowing for the definition of relationships between oscillatory features and particular frequency bands of the brain’s electrical activity [83,84]. Elements such as phase-phase, phase-power, and phase-amplitude coupling show how waves at different frequencies interact with each other, and are thought to serve as an indicator of cognitive processes including attention, plasticity, socialization and learning [32,70,85,86]. For example, gamma-theta wave coupling is indicative of hippocampal-cortical network processes that support information-processing, learning, memory and the regulation of emotion [87,88]. In addition, gamma-beta wave coupling has been directly linked to social affiliation and caregiving [89]. Gamma frequency coupling may be ubiquitously present across brain regions, possibly indicating episodic and semantic memory, attention, emotion, dreaming, and imagination [90], all of which are correlates of good mental health.

5.1.3. Neurochemical Modulation

Gamma waves may also impact neurotransmitter expression [39,40], as indicated by pharmacological studies that relate gamma activity to neurochemical expression [22,40]. For example, gamma activity may play a causal role in the therapeutic uptake of ketamine [91], as well as help to identify which depressed patients might benefit most from serotonergic versus noradrenergic antidepressants [92]. Recent studies appear to suggest a degree of bi-directionality in the relationship between gamma and psychiatric disorders, whereby neurotransmitters and gamma-wave activity (a) regulate neurophysiological function via cross-frequency coupling [40], and (b) coordinate network connectivity via electrochemical means [93]. For example, Weiss et al. (2023) [40] found that the noradrenergic, cholinergic, and dopaminergic neuromodulatory systems appear to respond to coupled theta-gamma waves as part of the circuitry that underlies working memory and cognitive function. Watanabe et al. (2024) [93] argued that these same systems alter wave activity when viewed across temporal scales with moment-by-moment changes in monoamine levels reflected in concurrent alteration in neural wave activity.

Importantly, gamma-band activity has been linked with social affiliation and oxytocin (OT) expression [45,80,94,95,96]. Pratt et al. [89] reported that OT expression changes both mother–child synchrony and gamma power in the temporal/insular regions, and Feldman [45] also reported increased OT expression and gamma wave power in bio-behavioural synchrony studies. OT expression is also known to be directly involved in multiple behaviours linked to mental health, such as the formation of social memories [94], tendencies toward prosocial behaviours [97], and signaling within social neurocircuits [98]. Together, elements of the oxytocinergic system may also be key mechanisms that drive gamma wave activity to protect or stabilize psychological integrity [96].

5.1.4. Mechanical Modulation

Finally, recent evidence has emerged supporting the notion that the stochastic, low power and localized characteristics of gamma waves, together with their network-wide synchronizing function, may reveal a unique neuromodulatory mechanism with broad, physiological significance [33,44]. For example, endogenous gamma waves have been shown to be associated with regulation of blood and glymphatic flow and microglial balance, and may indicate states of metabolic regulation [99] and cerebral hemodynamics [44]. All of these are supported by the discovery that the breakdown of gamma rhythmicity contributes to the kind of neural degeneration and circuit dysfunction that is associated with brain disorders [29,31,44].

6. Future Research Directions

Research to date linking gamma waves to the kind of cognitive processes that are relevant to clinical psychiatry diagnostic and treatment models has largely focused on spectral analyses, yielding conflicting results. For example, increases in gamma power have been observed in heightened cognitive processing [43], social bonding [89], and increased anxiety [26], as well as during advanced states of relaxation [100]. As noted by Fitzgerald and Watson [22], mixed results have been reported for the association between depression and the strength and direction of gamma wave activity over different brain regions, gamma frequencies measured, and when participants were undergoing different tasks and had their depression measured by different self-report scales. These findings, while appearing to represent contradictions, are more likely the outcome of a lack of agreement or specificity regarding methodological issues and need more systematic and detailed investigation.

As a further example of the need for researcher agreement regarding methodological details, although variations in spectral power have been attributed to wave changes, these changes may not be wave-like in origin [33]. Each gamma wave is a compound, field-driven feature built from rhythmic patterns of neuronal spiking and synaptic excitation/inhibition, characterized by qualities of low amplitude, synchrony, spontaneity and stochastic expression [38]. Gamma waves may arise from endogenous sources as well as exogenous periodic stimuli capable of triggering rhythmicity via synchronistic, frequency-generating mechanisms [27,101]. As such, they generate complex signals with both periodic and aperiodic features, so that commonly used data-processing tools such as Fourier transforms and power spectral densities may obscure some of the relevant gamma characteristics [42]. That is, changes in spectral power could appear to reflect behavioural processes but may in fact be an expression of coherence [72], neurochemical interaction [40], a behavioural proxy [33], or even signal-processing anomalies [42] rather than a cognitive event. This unresolved definition requires greater attention to specific signal extraction methods such as gamma sub-band delineation and comparison of power spectral densities with single-trace recordings. Further refinement of these measurement strategies may also flow on to more precise targetting of treatments to increase gamma activity in specific brain regions via (for example) transcranial stimulation procedures [102]. Finally, there is a growing research literature describing the rôle of gamma in such areas as brain–computer interfaces, robotic control, sensory-motor coordination, and cognitive modelling which may be seen to describe the general principles of gamma activity [103] and which may repay attention for the clinican who wishes to understand some of the mechanisms which underly gamma.

7. Conclusions

Some initial foundations exist for our understanding of gamma and its role in psychiatry. For example, power-based investigations still offer important insights. Changes in gamma band power have been able to provide a proxy of behaviour, and many potential diagnostic applications may be based on that information. However, to realize its potential as a major diagnostic and therapeutic tool in clinical psychiatry, further research is required into the relationship between gamma waves and mental health, and multimodal analyses combining spectral characteristics with coherence, frequency-coupling, topographical and neurochemical variables. When grounded in the associations of gamma with neuromodulation described above, and working within some wider inter-study agreement regarding definition of gamma in terms of different frequency ranges, results of those investigations have the potential to clarify the function and role of gamma as a promising biomarker family for network-level dysfunction and treatment monitoring.