Elsevier

NeuroImage

Volume 51, Issue 1, 15 May 2010, Pages 365-372
NeuroImage

Association of individual resting state EEG alpha frequency and cerebral blood flow

https://doi.org/10.1016/j.neuroimage.2010.02.024Get rights and content

Abstract

Cognitive task performance differs considerably between individuals. Besides cognitive capacities, attention might be a source of such differences. The individual's EEG alpha frequency (IAF) is a putative marker of the subject's state of arousal and attention, and was found to be associated with task performance and cognitive capacities. However, little is known about the metabolic substrate (i.e. the network) underlying IAF. Here we aimed to identify this network. Correlation of IAF with regional Cerebral Blood Flow (rCBF) in fifteen young healthy subjects revealed a network of brain areas that are associated with the modulation of attention and preparedness for external input, which are relevant for task execution. We hypothesize that subjects with higher IAF have pre-activated task-relevant networks and thus are both more efficient in the task-execution, and show a reduced fMRI-BOLD response to the stimulus, not because the absolute amount of activation is smaller, but because the additional activation by processing of external input is limited due to the higher baseline.

Introduction

There are considerable and relevant differences in perceptual and cognitive individual capacities that vary to some degree with the subject's state, but that also appear to represent to a large degree stable individual trait markers. One of the most interesting neurobiological correlates of individual cognitive resources is the individual EEG alpha frequency (IAF). The alpha rhythm lies in the range of 8 to 13 Hz but its peak frequency varies from person to person (Klimesch, 1999, Posthuma et al., 2001, van Beijsterveldt and Boomsma, 1994). The IAF has been shown to exhibit specific personal characteristics with high intra-individual stability (Binnie et al., 2003, Fernandez et al., 1993), which is supported by twin studies reporting a high genetic component of the IAF (Smit et al., 2006). Specifically, the IAF has been shown to be associated with various cognitive features such as task performance, (working) memory capacity (Klimesch, 1997, Klimesch et al., 1997, Lebedev, 1994, Richard Clark et al., 2004), reaction time (Klimesch et al., 1996, Surwillo, 1963) and speed of information processing (Klimesch et al., 1996) while its role as a marker for overall intelligence is still debated (Anokhin and Vogel, 1996, Doppelmayr et al., 2002, Jausovec and Jausovec, 2000, Posthuma et al., 2001). Furthermore, individual variations in IAF could also be related to attentional demands and/or arousal as well as cognitive preparedness (Angelakis et al., 2004, Klimesch et al., 1993).

Changes in the EEG frequency are also related to changes in the state of vigilance, respectively, attentional level of the person. Besides the pronounced frequency-band changes that occur when switching from a relaxed resting state (alpha band) to a state of mental effort (beta band: 13–40 Hz/and gamma band: above 40 Hz) or during the transition from the awake to sleep stages (delta and theta bands: below 8 Hz), already slight changes within the alpha frequency band can indicate a shift in vigilance or attention (Kubicki et al., 1979).

Although the IAF is an interesting biomarker of trait-like individual cognitive resources, there is still little knowledge about its neurobiological substrate. To investigate individual differences in state or trait markers such as the IAF, a method that provides absolute, quantitative and reproducible values of brain metabolism is necessary. Whilst the fMRI BOLD signal represents only relative changes in the blood's oxygenation level without absolute information about Cerebral Blood Flow (CBF), Cerebral Blood Volume (CBV) and energy metabolism (Brown et al., 2007, Buxton, 2005, Buxton et al., 1998Buxton et al., 2004, Davis et al., 1998, Leontiev and Buxton, 2007, Leontiev et al., 2007, Obata et al., 2004)., Positron Emission Tomography (PET) and Arterial Spin Labelling (ASL) meet those requirements (reproducibility: PET: Bartlett et al., 1988; ASL: Jahng et al., 2005). The fMRI BOLD nevertheless is suited to investigate intra-individual temporal fluctuations in EEG oscillations (de Munck et al., 2009, Esposito et al., 2009, Goldman et al., 2002, Jann et al., 2009). Furthermore, interpretations of fMRI BOLD experiments are almost always based on the relative differences in metabolic energy consumption associated to neuronal activity between two task conditions, while neglecting the neuronal activity and the related high energy consumption during the baseline state. However, it has been reported that those incremental energies found between task and baseline are small (for cognitive tasks less than 1%) compared to the larger total baseline energies (Hoge and Pike, 2001, Hyder et al., 2006, Kida and Hyder, 2006, Ogawa et al., 1993). The high baseline energy metabolism maintains the subject in a state that enables him/her to respond to external stimuli as demonstrated by anesthesia studies (Alkire, 2008, Katoh et al., 2000). Thus, understanding the high energies in the baseline seems to be prerequisite to understand and interpret the small energy increments evoked by a task. Hence, in order to investigate the baseline state and its inter-individual differences, fixed state-variables (putatively the IAF) and absolute measures of brain metabolism (i.e. CBF as surrogate for it) are mandatory (Leontiev and Buxton, 2007, Liau and Liu, 2009, Liau et al., 2008).

Up to date, there are several studies that investigated the relationship between EEG alpha rhythm variations and baseline regional CBF (rCBF) differences. However, most of these studies focused on the magnitude of alpha oscillations within the whole frequency band (Danos et al., 2001, Dierks et al., 2000, Larson et al., 1998, Oakes et al., 2004, Sadato et al., 1998). Only a few studies considered variations in the alpha peak frequency. Early studies revealed a relationship between mean global CBF or metabolism and overall EEG mean frequency (Ingvar et al., 1979, Kuschinsky, 1993). More recently, Alper et al. (2006) performed a regionally more detailed study. They reported on topographically distinct correlation patterns between EEG frequencies and five different, pre-selected brain compartments. The mean CBF signals for the thalamus, the dorsolateral prefrontal cortex, the medial frontal cortex and the left respectively right temporal lobe were calculated. These mean CBF values then were correlated to several EEG parameters (among them the mean frequency within the alpha band) at each electrode site. For all five compartments except the thalamus most electrodes showed a positive correlation between the mean alpha frequency and the mean CBF, although with different topographic distributions. Yet, the authors commented on the findings for the magnitude but not the mean frequency. Another study related to IAF and CBF was conducted by Koch et al. (2008). They investigated the predictive value of the IAF on the neuronal and vascular response during visual stimulation. Based on their findings they argue that IAF differences might partly explain the recently reported inter-individual variability in combined EEG-fMRI studies (Goncalves et al., 2006).

Nevertheless, to our knowledge, there is no report on whole-brain, voxelwise correlations of rCBF with the IAF. Hence, the structures of the networks that are associated with the IAF are still largely unknown. The aim of the current study was to identify those networks. This required on one side a method to reliably estimate absolute individual rCBF in a voxel-wise and quantitative way, and on the other side, (in order to keep the complexity of the results low) a physiologically sensitive, but simple, global index of the IAF. The former was provided by ASL as discussed above, for the latter we chose to base our analysis on the position of the global center of gravity (CoG) in frequency. The usage of a descriptor that takes into account the global distribution of the alpha frequency was initially proposed by Klimesch et al. (1996). He argues that most likely the IAF represents the synchronized state of a mixture of different alpha rhythms with distinct spatial distributions, rather than IAF reflecting just one single process. Alternatively, the IAF peak might be a feature that marks the overlap between the oscillatory processes in the upper and the lower alpha bands (Klimesch et al., 1996). While the lower alpha band is associated with attentional processes the upper band is thought to be related to memory processes (Klimesch, 1997). However, how many and what rhythms exactly contribute to the IAF is not yet unravelled. In either case, methods taking into accounts the global features of the alpha frequency were suggested to be more suitable to deal with a population of different alpha frequencies (Klimesch, 1996). Moreover, since the method is inferring the measure from the dynamics of extended EEG scalp fields, it can theoretically be related to extended intra-cerebral neural networks.

In summary, the IAF is a potential marker for the subject's cognitive capacity and/or present state and represents inter-subject variability. However, little is known about related baseline metabolic differences. A state of higher vigilance represents a condition of elevated alertness. In this condition a subject would rapidly shift its attention to a new external stimulus. Such a shift of attention could be most efficiently performed when the capacity to react (i.e. metabolic reactivity) would be at maximum level. Under this assumption, we would hypothesize to find a positive correlation between IAF and baseline CBF. Moreover, we would expect to observe correlations in areas usually involved in task execution, i.e., areas involved in the modulation of attention, preparedness to external stimuli (Angelakis et al., 2004) or working memory processes.

Section snippets

Subjects

Seventeen healthy volunteers were recruited from the University of Bern. Two subjects had to be excluded, one due to a benign cyst and the other because it did not exhibit a clear alpha rhythm. The remaining fifteen subjects (8 female/7 male; age ± SD 26.07 ± 2.55 years) were included in the study. All of them gave their written informed consent. All subjects refrained from caffeine, alcohol and nicotine for at least 6 h before the experiment. Any neurophysiological or psychiatric disorder or

EEG alpha frequency

The values of the IAF represented by APF and CoG were highly correlated (r = 0.85, p < 0.001) and shared more than 72% of the variance (Fig. 1). Nevertheless, the APF is limited to the frequency bins that are defined by the duration and the sampling rate of the epoch for which the FFT is calculated. The CoG is not limited to such bins and therefore can provide a continuous quantifier. Therefore, we used only the CoG for the correlation analysis with the ASL data.

Correlation analysis between subjects' mean rCBF and CoG

The correlation analysis between the

Discussion

Measures such as resting CBF and the individual alpha frequency (IAF) represent basic features of the individuals ‘baseline’ state. Both features are currently subject of interest since they have been demonstrated to influence respectively bias task performance. However, only little is known whether there is a direct association between a subject's IAF and variability in regional CBF. Therefore, we targeted the present study on inter-individual variability in regional CBF and its relationship

Conclusions

In conclusion, we observed inter-individual regional differences of baseline CBF associated with IAF. Both measures are features of the individual's baseline state and are argued in the literature to influence task performance. Our findings represent novel detailed information about inter-individual differences in regional baseline perfusion and delineate a network that has previously been related to monitoring of the environment and modulation of attention and arousal and to define the

Acknowledgments

We would like to thank Mara Kottlow, Karin Zwygart and Verena Beutler for their assistance in data acquisition. We furthermore acknowledge Jiri Wackermann's help in methodological issues that came up during the review process and the reviewers for their helpful comments. This study was partly financed by the Swiss National Science Foundation grant 320000-108321/1.

References (90)

  • EspositoF. et al.

    Distributed analysis of simultaneous EEG-fMRI time-series: modeling and interpretation issues

    Magn. Reson. Imaging

    (2009)
  • EspositoF. et al.

    Independent component model of the default-mode brain function: Assessing the impact of active thinking

    Brain Res. Bull.

    (2006)
  • FoxM.D. et al.

    Intrinsic fluctuations within cortical systems account for intertrial variability in human behavior

    Neuron

    (2007)
  • GoncalvesS.I. et al.

    Correlating the alpha rhythm to BOLD using simultaneous EEG/fMRI: inter-subject variability

    Neuroimage

    (2006)
  • HaierR.J. et al.

    Regional glucose metabolic changes after learning a complex visuospatial/motor task: a positron emission tomographic study

    Brain Res.

    (1992)
  • HamzeiF. et al.

    The influence of extra-and intracranial artery disease on the BOLD signal in FMRI

    Neuroimage

    (2003)
  • HogeR.D. et al.

    Oxidative metabolism and the detection of neuronal activation via imaging

    J. Chem. Neuroanat.

    (2001)
  • JannK. et al.

    BOLD correlates of EEG alpha phase-locking and the fMRI default mode network

    Neuroimage

    (2009)
  • KlimeschW.

    Memory processes, brain oscillations and EEG synchronization

    Int. J. Psychophysiol.

    (1996)
  • KlimeschW.

    EEG-alpha rhythms and memory processes

    Int. J. Psychophysiol.

    (1997)
  • KlimeschW.

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

    Brain Res. Brain Res. Rev.

    (1999)
  • KlimeschW. et al.

    Brain oscillations and human memory: EEG correlates in the upper alpha and theta band

    Neurosci. Lett.

    (1997)
  • KlimeschW. et al.

    Upper alpha ERD and absolute power: their meaning for memory performance

    Prog. Brain Res.

    (2006)
  • KochS.P. et al.

    Individual alpha-frequency correlates with amplitude of visual evoked potential and hemodynamic response

    Neuroimage

    (2008)
  • LaskarisN.A. et al.

    Single-trial variability in early visual neuromagnetic responses: an explorative study based on the regional activation contributing to the N70 m peak

    Neuroimage

    (2003)
  • LeontievO. et al.

    Reproducibility of BOLD, perfusion, and CMRO2 measurements with calibrated-BOLD fMRI

    Neuroimage

    (2007)
  • LeontievO. et al.

    CBF/CMRO2 coupling measured with calibrated BOLD fMRI: sources of bias

    Neuroimage

    (2007)
  • LiauJ. et al.

    Inter-subject variability in hypercapnic normalization of the BOLD fMRI response

    Neuroimage

    (2009)
  • LiauJ. et al.

    Caffeine reduces the activation extent and contrast-to-noise ratio of the functional cerebral blood flow response but not the BOLD response

    Neuroimage

    (2008)
  • MantiniD. et al.

    Large-scale brain networks account for sustained and transient activity during target detection

    Neuroimage

    (2009)
  • NeubauerA.C. et al.

    Intelligence and neural efficiency: further evidence of the influence of task content and sex on the brain–IQ relationship

    Brain Res. Cogn. Brain Res.

    (2005)
  • ObataT. et al.

    Discrepancies between BOLD and flow dynamics in primary and supplementary motor areas: application of the balloon model to the interpretation of BOLD transients

    Neuroimage

    (2004)
  • OgawaS. et al.

    Functional brain mapping by blood oxygenation level-dependent contrast magnetic-resonance-imaging—a comparison of signal characteristics with a biophysical model

    Biophys. J.

    (1993)
  • RestomK. et al.

    Cerebral blood flow and BOLD responses to a memory encoding task: a comparison between healthy young and elderly adults

    Neuroimage

    (2007)
  • Richard ClarkC. et al.

    Spontaneous alpha peak frequency predicts working memory performance across the age span

    Int. J. Psychophysiol.

    (2004)
  • SmitC.M. et al.

    Genetic variation of individual alpha frequency (IAF) and alpha power in a large adolescent twin sample

    Int. J. Psychophysiol.

    (2006)
  • SurwilloW.W.

    The relation of simple response time to brain-wave frequency and the effects of age

    Electroencephalogr. Clin. Neurophysiol.

    (1963)
  • VanniniP. et al.

    Task demand modulations of visuospatial processing measured with functional magnetic resonance imaging

    Neuroimage

    (2004)
  • VanniniP. et al.

    Failure to modulate neural response to increased task demand in mild Alzheimer's disease: fMRI study of visuospatial processing

    Neurobiol. Dis.

    (2008)
  • WackermannJ.

    Towards a quantitative characterisation of functional states of the brain: from the non-linear methodology to the global linear description

    Int. J. Psychophysiol.

    (1999)
  • WangJ. et al.

    To smooth or not to smooth? ROC analysis of perfusion fMRI data

    Magn. Reson. Imaging

    (2005)
  • AlkireM.T.

    Probing the mind: anesthesia and neuroimaging

    Clin. Pharmacol. Ther.

    (2008)
  • AngelakisE. et al.

    EEG neurofeedback: a brief overview and an example of peak alpha frequency training for cognitive enhancement in the elderly

    Clin. Neuropsychol.

    (2007)
  • BartlettE.J. et al.

    Reproducibility of cerebral glucose metabolic measurements in resting human subjects

    J. Cereb. Blood Flow Metab.

    (1988)
  • BazanovaO.M. et al.

    The use of individual EEG peculiarities for increase of neurofeedback efficiency

    Zh. Nevrol. Psikhiatr. Im. S. S. Korsakova

    (2006)
  • Cited by (95)

    View all citing articles on Scopus
    View full text