The frequency of alpha oscillations: Task-dependent modulation and its functional significance
Introduction
EEG alpha rhythm (8–12 Hz) is prominent over occipital-parietal cortices in humans during restful wakefulness. Like any oscillatory activity, alpha can be characterized by its amplitude (power) and frequency. Since its initial discovery by Berger (1929), extensive evidence has accumulated to suggest that alpha oscillations are involved in cognitive processing, including attention and working memory (Fink et al., 2005; Jensen et al., 2002; Jensen and Tesche, 2002; Klimesch et al., 1997). In particular, alpha power is considered a reliable indicator of cortical excitability (Jensen and Mazaheri, 2010; Lange et al., 2013) and can be modulated by higher-order brain areas to functionally inhibit or facilitate sensory information processing according to behavioral goals (Liu et al., 2014; Mulholland, 1968; Rajagovindan and Ding, 2011; Wang et al., 2016).
Alpha frequency, in contrast, has been viewed mainly as a trait variable. Individual differences in alpha frequency have been linked to differences in cognitive capabilities (Anokhin and Vogel, 1996; Grandy et al., 2013). An existing theory postulates that alpha frequency is a manifestation of an internal brain clock controlling the speed of information processing (Klimesch et al., 1996); the faster the internal clock (higher individual alpha frequency), the faster the information and cognitive processing (e.g., in memory retrieval and cognitive control). In addition, alpha frequency is known to change systematically over the lifespan (Bernhard and Skoglund, 1939), and is subject to the impact of neurological disorders (Bonanni et al., 2008). Perceptually, individual differences in alpha peak frequency has been shown to predict the temporal windows of the double-flash illusion, suggesting that the peak alpha frequency is the “fingerprint” that drives cross-modal impact on visual perception (Cecere et al., 2015). Similarly, the phasic differences of alpha rhythm has been shown to impact visual perception of rapid sequential stimuli (Minami and Amano, 2017; Valera et al., 1981).
To what extent alpha frequency can be modulated in a task-dependent manner? If such modulation occurs, what is the associated functional significance? These questions have yet to be fully addressed. Two earlier studies by Osaka and by Earle reported inconclusive findings. Osaka showed that alpha frequency increased with task difficulty (Osaka, 1984), but Earle reported that while in some tasks alpha frequency increased when task difficulty increased, in other tasks, the opposite was observed (Earle, 1988). Haegens et al., using an n-back working memory task, found that alpha frequency increased with task difficulty, i.e., higher working memory load was accompanied by higher alpha frequency (Haegens et al., 2014). A more recent study reported that alpha frequency changes according to the types of visual perception such as temporal stimulus integration and temporal stimulus segregation (Wutz et al., 2018). Despite these advances, many questions remain; in particular, the functional and behavioral relevance of alpha frequency modulation by task conditions remains to be further elucidated. Some of the tasks used in previous studies have temporally overlapping cognitive processes. For example, in the n-back task used by Haegens et al. (2014), cognitive processes such as sensory encoding, memory retention and memory retrieval are difficult to separate temporally, and each process may have a differential effect on alpha frequency (Jensen et al., 2002).
Alpha frequency modulation may also be considered in the context of how alpha interacts with other neural processes, such as theta oscillations (Sauseng et al., 2005; Scheeringa et al., 2009), beta oscillations (Yuan et al., 2010), and gamma oscillations (Voytek et al., 2010). In particular, in light of the proposal that gamma cycles are embedded within alpha (Osipova et al., 2008; Roux and Uhlhaas, 2014), alpha frequency modulation may become a means to flexibly increase and decrease the number of gamma cycles accommodated within a cycle of alpha to either facilitate or inhibit sensory processing.
In this study, we considered task-dependent modulation of alpha frequency, its functional significance, and the underlying neural correlates by conducting two experiments. In the first experiment, high-density EEG (128 channels) was recorded from 21 subjects performing a Sternberg working memory task. In this task, memory related cognitive processes such as encoding, retention and retrieval are well separated in time, making it well-suited for analyzing modulations of alpha frequency by distinct cognitive processes. Alpha frequency was estimated during different stages of the task and compared across memory loads. Functional significance was further assessed by correlating the modulations of alpha frequency with behavioral performance and other neural variables such as event-related brain responses. In the second experiment, we examined the neural sources underlying the modulation of scalp level alpha variables, namely, alpha power and alpha frequency. Simultaneous EEG-fMRI was recorded from a cohort of 59 subjects at rest. Utilizing the naturally occurring fluctuations of alpha frequency and alpha power, we examined the possible neural underpinnings of alpha frequency modulation and alpha power modulation by correlating them with the simultaneously recorded fMRI BOLD (Blood Oxygen Level Dependence) fluctuations from different regions of the visual cortex as well as the entire brain.
Section snippets
Procedure and paradigm
The experimental protocol was approved by the Institutional Review Board of the University of Florida (UF IRB). Twenty-one healthy individuals (age: 20 to 34; 3 women) with normal or corrected-to-normal vision gave written informed consent and participated in the study. On each trial of the working memory task, a digit set (cue set) of 1, 3 or 5 distinct numerical digits (0–9) was displayed for 1s (encoding). This was followed by a 3s period during which the subjects held the cue set in working
Behavioral results
Behavioral measures were analyzed using a mixed-effects linear model with performance metrics as the dependent variable, memory load (1, 3 and 5) as the independent variable, and subjects as the random effects. As shown in Fig. 1C and D, reaction time was significantly slower for higher memory load (p < 0.001), whereas accuracy was significantly lower for higher memory load (p < 0.001), in agreement with previous studies of the same paradigm (Jensen and Lisman, 1998; Sternberg, 1969).
Alpha power and frequency modulation by working memory load
Power
Main findings
We performed two experiments to examine the task-dependent modulation of alpha frequency and its functional significance as well as its neural substrate. In the first experiment, the analysis of EEG recorded from healthy subjects performing a Sternberg working memory task shows that: (1) alpha power decreased with memory load during encoding, increased with memory load during retention, and had no systematic relationship with memory load during retrieval, (2) alpha frequency decreased with
Disclosure of interest
The authors report no conflicts of interest.
Acknowledgment
This work was supported by US National Institutes of Health grant MH112206 and US National Science Foundation grant BCS1439188.
References (77)
- et al.
Electroencephalographic peak alpha frequency correlates of cognitive traits
Neurosci. Lett.
(2004) - et al.
EEG alpha rhythm frequency and intelligence in normal adults
Intelligence
(1996) - et al.
Individual differences in alpha frequency drive crossmodal illusory perception
Curr. Biol.
(2015) - et al.
EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis
J. Neurosci. Meth.
(2004) How and when the fMRI BOLD signal relates to underlying neural activity: the danger in dissociation
Brain Res. Rev.
(2010)- et al.
EEG alpha band dissociation with increasing task demands
Cognit. Brain Res.
(2005) - et al.
Thresholding of statistical maps in functional neuroimaging using the false discovery rate
Neuroimage
(2002) - et al.
Correlating the alpha rhythm to BOLD using simultaneous EEG/fMRI: inter-subject variability
Neuroimage
(2006) - et al.
Individual alpha peak frequency is related to latent factors of general cognitive abilities
Neuroimage
(2013) - et al.
Inter- and intra-individual variability in alpha peak frequency
Neuroimage
(2014)
Event-related desynchronization in the alpha band and the processing of semantic information
Cognit. Brain Res.
Quantitative EEG in normals and in patients with cerebral ischemia
Prog. Brain Res.
Where the BOLD signal goes when alpha EEG leaves
Neuroimage
Reference-free identification of components of checkerboard-evoked multichannel potential fields
Electroencephalogr. Clin. Neurophysiol.
Coupling between visual alpha oscillations and default mode activity
Neuroimage
Quantitative analysis of occipital EEG in different stages of Alzheimer's disease
Electroencephalogr. Clin. Neurophysiol.
Cholinergic inhibition in the cerebral cortex
Brain Res.
Working memory and neural oscillations: alpha–gamma versus theta–gamma codes for distinct WM information?
Trends Cognit. Sci.
The speed of alpha-band oscillations predicts the temporal resolution of visual perception
Curr. Biol.
Fronto-parietal EEG coherence in theta and upper alpha reflect central executive functions of working memory
Int. J. Psychophysiol.
Trial-by-trial coupling between EEG and BOLD identifies networks related to alpha and theta EEG power increases during working memory maintenance
Neuroimage
Slowing of EEG in Parkinson's disease
Electroencephalogr. Clin. Neurophysiol.
Sensitivity of human EEG alpha band desynchronization to different working memory components and increasing levels of memory load
Neurosci. Lett.
Perceptual framing and cortical alpha rhythm
Neuropsychologia
Estimating Granger causality after stimulus onset: a cautionary note
Neuroimage
Relation between P300 and event-related theta-band synchronization: a single-trial analysis
Clin. Neurophysiol.
Nonemotional features suppress early and enhance late emotional electrocortical responses to negative pictures
Biol. Psychol.
Negative covariation between task-related responses in alpha/beta-band activity and BOLD in human sensorimotor cortex: an EEG and fMRI study of motor imagery and movements
Neuroimage
Über das Elektrenkephalogramm des Menschen
Archiv für Psychiatrie und Nervenkrankheiten
On the alpha frequency of human brain potentials as a function of Age1
Skandinavisches Archiv Für Physiologie
EEG comparisons in early Alzheimer's disease, dementia with Lewy bodies and Parkinson's disease with dementia patients with a 2-year follow-up
Brain
Spontaneous alpha peak frequency predicts working memory performance across the age span
Int. J. Psychophysiol.
Cortical cell and neuron density estimates in one chimpanzee hemisphere
Proc. Natl. Acad. Sci. U.S.A.
Differential expression of muscarinic acetylcholine receptors across excitatory and inhibitory cells in visual cortical areas V1 and V2 of the macaque monkey
J. Comp. Neurol.
Task difficulty and EEG alpha asymmetry: an amplitude and frequency analysis
Neuropsychobiology
Cholinergic induction of oscillations in the hippocampal slice in the slow (0.5–2 Hz), theta (5–12 Hz), and gamma (35–70 Hz) bands
Hippocampus
Cholinergic inhibition of neocortical pyramidal neurons
J. Neurosci.
Motion opponency in visual cortex
J. Neurosci.
Cited by (50)
Individual differences in information processing during sleep and wake predict sleep-based memory consolidation of complex rules
2023, Neurobiology of Learning and MemoryThe neural mechanism of non-phase-locked EEG activity in task switching
2023, Neuroscience LettersCitation Excerpt :Similarly, Capizzi et al. (2020) found that the switch effect of alpha oscillations during the reactive control process (i.e., the alpha power in the switch trial was significantly reduced compared to the repeat trial) was also associated with memory retrieval [39]. Furthermore, alpha oscillations are also closely related to attention in task switching [47,48], such as switching attention between working memory representations [49], inhibiting potential distractors [47], and facilitating the maintenance of cortical excitability and processing of sensory information [50]. Some studies found that regardless of cue presentation [28,51,52] or target presentation [28], the alpha oscillation amplitude over the frontoparietal network was significantly reduced in switch trials compared to repeat trials.
No effects of prefrontal multichannel tACS at individual alpha frequency on phonological decisions
2022, Clinical NeurophysiologyAge- and movement-related modulation of cortical oscillations in a mouse model of presbycusis
2021, Hearing ResearchCitation Excerpt :Alpha power and peak alpha frequency decline with age in humans (Chiang et al., 2011; Harris and Dubno, 2017). Decline in alpha power and subsequent reduction in reliable phase estimation with age is predicted to affect stimulus detection thresholds and reaction times (Samuel et al., 2018; VanRullen et al., 2011), which can lead to difficulties with speech recognition in noise. The reduced alpha power may also mark elevated cortical load or arousal levels (Barry et al., 2007) and changes in cerebral perfusion (Cook et al., 1998) in the old mice (reviewed in Bazanova and Vernon, 2014).
EEG, MEG and neuromodulatory approaches to explore cognition: Current status and future directions
2021, Brain and CognitionCitation Excerpt :Suppression of alpha amplitude can, moreover, be observed upon sensory stimulation (Ikkai et al., 2016; Mo et al., 2013; Nishifuji et al., 2010; Pfurtscheller & Aranibar, 1977) as a function of processing load (Rajagovindan & Ding, 2011), and during movement tasks (Pfurtscheller & Da Silva, 1999). Spontaneous resting-state EEG fluctuations of alpha frequency are inversely associated with the visual BOLD activity, and higher alpha frequency preceding a cue/probe is associated with weaker cue/probe-evoked neural responses and longer RTs (Samuel et al., 2018). Altogether, these findings suggest that alpha-band activity may index cortical excitability, contributing to sensory-gating.