Missoner, Et Al 2006

7/28/2019 Missoner, Et Al 2006

1/10

DOI 10.1007/s00702-005-0443-9

J Neural Transm (2006) 113: 14771486

Frontal theta event-related synchronization: comparison

of directed attention and working memory load effects

P. Missonnier1;2, M.-P. Deiber2, G. Gold4, P. Millet1, M. Gex-Fabry Pun3,L. Fazio-Costa4, P. Giannakopoulos2;5, and V. Ibanez1

1 Neuroimaging Unit,2 Service of Geriatric Psychiatry, and

3 Unit of Clinical Research, Department of Psychiatry,University Hospitals of Geneva, 1225 Geneva, and

4 Department of Geriatrics, University Hospitals of Geneva, 1226 Geneva, and5 Service of Old Age Psychiatry, University Hospitals of Lausanne, 1008 Prilly, Switzerland

Received July 11, 2005; accepted December 8, 2005Published online April 11, 2006; # Springer-Verlag 2006

Summary. Early studies showed that long-term encoding and retrieval of new informa-tion is associated with modulation of thetheta rhythm. More recently, changes in theta

power amplitude over frontal electrode siteswere reported during working memory, yettheir relative significance in regard to at-tentional and memory processes remains un-clear. Event-related synchronisation responsesin the 47.5 Hz theta EEG frequency bandwas studied in 12 normal subjects performingfour different tasks: two working memorytasks in which load varied from one (1-backtask) to two (2-back task) items, an oddballdetection (attention) task and a passive fixa-

tion task. A phasic theta increase was ob-served following stimulus apparition on allelectrode sites within each task, with longerculmination peak and maximal amplitudeover frontal electrodes. Frontal theta event-related synchronization (ERS) was of higheramplitude in the 1-back, 2-back and detec-tion tasks as compared to the passive fixationtask. Additionally, the detection task elicited

a larger frontal and central theta ERS than the2-back task. By analyzing theta ERS charac-teristics in various experimental conditions,the present study reveals that early phasic

theta response over frontal regions primarilyreflects the activation of neural networks in-volved in allocation of attention related totarget stimuli rather than working memoryprocesses.

Keywords: Working memory, EEG, event-related synchronisation, theta oscillations,attention.

Introduction

Brain oscillations reflect the synchronizationof neuronal cell assemblies that participatein local or long-distance network systems.Alteration of oscillatory brain temporal dy-namics can occur in different EEG frequencybands as a result of functional modulationduring perceptive or cognitive tasks. Amongthese task-related oscillations, the humantheta activity has been implicated in memory


2/10

performance (Doppelmayr et al., 2000;Klimesch, 1999), after hippocampal thetarhythm was originally described in memoryprocessing in the rodent (OKeefe, 1993;

Winson, 1978). However, task-related humantheta activity is widely recorded over thescalp and cannot be univocally attributed tothe hippocampal region (for a review seeKahana et al., 2001).

Klimesch and colleagues reported anincrease in theta power during encoding andretrieval of new information but not duringsemantic memory performance (Klimesch,1999; Klimesch et al., 1997), suggesting thattheta activity may not be involved in all typesof long-term memory processes. Recently,the influence of short-term memory, mainlyworking memory, on scalp theta activityhas received increasing interest. Using task-related power or time frequency analysis,several authors have reported an increase offrontal theta amplitude as a function of work-ing memory load (Gevins et al., 1997; Jensenand Tesche, 2002; McEvoy et al., 2001).Krause et al. (2000) specifically examinedthe temporal dynamics of theta activity dur-ing classical working memory n-back tasks

by applying the event-related synchronisation(ERS) and desynchronisation (ERD) analysisprocedure (Pfurtscheller, 1992, 2001). Theyobserved that theta ERS amplitude was larg-er over frontal electrodes sites during thememory condition with the highest memorycomponent (2-back task) and suggested thatfrontal theta synchronization could be anindex of working memory load.

Working memory is by definition closelyrelated to attentional processes. Indeed, fo-cused attention to incoming stimulus ismandatory to ensure correct encoding, a fun-damental step for further successful workingmemory processes. Thus, the influence ofattention on theta modulation in workingmemory cannot be ignored. There is evidencefor strong evoked theta activity during neu-ropsychological paradigms involving focusedattention (Basar-Eroglu et al., 1992). Further-

more, results from animal studies suggestthat theta oscillations recorded in motor andprefrontal cortex could participate to the at-tention system (Demiralp et al., 1994).

The aim of this study was to determinethe respective contribution of working mem-ory and attentional processes to theta ac-tivity. We used the ERD=ERS procedure(Pfurtscheller and Aranibar, 1977) to ana-lyze the EEG power in the theta frequencyband during two visual n-back workingmemory tasks and an oddball detection task.In addition, a passive fixation task was usedto determine whether theta activity couldbe elicited in absence of any specific taskdemand.

Subjects and methods

Participants

Twelve healthy, cognitively intact right-handedvolunteers (aged 2231, 26.75 2.86; 5 women,7 men) participated in this study. All participantshad normal or corrected-to-normal visual acuity,and none reported a history of head injury, neurolog-ical or psychiatric disorders. All participants weremedication free and none exhibited alcohol or drugabuse. Informed consent was obtained from all sub-

jects. The study was approved by the Ethical Com-mittee of the University Hospitals of Geneva, andwas in line with the Helsinki Declaration.

Experimental design

The subjects, comfortably seated, watched a computer-controlled display screen at a distance of 57 cm. Theyviewed pseudo-random sequences of consonant andvowels common to the French alphabet, and presseda computer-controlled button with their right index fin-ger as soon as a target appeared (response trials). Fornon-target stimuli, no motor response was required (no-

response trials). In relation to the task, targets weredefined either according to the oddball (rare event) orto the n-back design (Fig. 1).

Stimuli consisted of white letters, Arial font(2 2.5 visual angle), with 10% grey noise, embed-ded in a 50% random noise grey rectangular back-ground patch (6 6.7 visual angle). They werepresented in the centre of the screen for 0.5 s, separatedby 5 s intervals (onset to onset) during which a dothelped subjects maintain fixation.

1478 P. Missonnier et al.


3/10

Four tasks were used. In the detection task, sequen-tial letters (target) or background patches without let-ters (non-target) were presented. In the 1-back task, thetarget was any letter identical to the one immediatelypreceding it. In the 2-back task, the target was any letterthat was identical to the one presented two trials back.Thus, working memory load increased from detection(memory-free) to 1-back (moderately demanding) and2-back (highly demanding). In the passive fixation task,letter series identical to the 2-back task were presented,but subjects were unaware of the nature of the task and

watched the series passively.Each task was tested in three stimulus sequences,

composed of 30 images each (7 targets), adding up to90 trials per task (21 targets). Before each sequence,subjects were informed about the nature of the forth-coming task. The protocol started with three sequencesof the passive fixation task. A first detection tasksequence was followed by a sequence of the 1-backtask, three sequences of the 2-back task, two sequencesof the 1-back task, and two sequences of the detection

task. Reaction time and performance were systemati-cally recorded, but no feedback on performance wasprovided.

Electrophysiological recordings

Continuous EEG (Micromed, Brain Quick system98, Treviso, Italy) was recorded using twenty surfaceelectrodes placed over the scalp according to the1020 international system, with linked earlobes asreference. Skin impedance was kept below 5 k.Electrophysiological signals were sampled at 1024 Hz,with a lower cut-off of 0.3 Hz and a upper cut-off of100 Hz (DC amplifiers Micromed). The electrooculo-gram (EOG) was recorded using two pairs of bipolarelectrodes in both vertical and horizontal directions.Single pulses (TTL) synchronized with stimulus onsetwere recorded and used off-line to segment the con-tinuous EEG data into epochs time-locked to stimulus

onset.

Data processing

EEG data were analyzed using NeuroScan software(NeuroScan Inc., Herndon, VA, USA). EEG signalswere corrected for ocular artifacts using a thresholdreduction algorithm (NeuroScan Inc.). The signal wasdigitally filtered in the 47.5 Hz theta frequency band,using narrow band pass filter (48dB=octave). TheEEG data were automatically scanned for contamina-tion by muscular or electrode artifacts, and visuallyinspected to control for other artifacts. EEG data were

split into epochs of 4500 ms, starting 2500 ms beforestimulus onset. The filtered epochs were then squaredto obtain absolute spectral power magnitude, expressedin mV2. Theta power data were analyzed from trialswith correct responses only. Due to the reduced numberof response trials and to increase the signal-to-noiseratio, response and no-response trials were averaged,after controlling that there was no significant differencein theta power between them. Taking into account itswidespread scalp distribution, theta power was initiallymeasured for each subject in the following regions:frontal (F4, Fz, F3), central (C4, Cz, C3), parietal(P4, Pz, P3), and occipital regions (O2, Oz, O1). Then,data from the three electrodes of each region were

averaged. Within each task, the culmination peak ofthe theta event-related synchronization (ERS) was mea-sured in the 0500 ms time window after stimulus onsetand the ERS magnitude was normalized to account forinter-electrode and inter-subject variations in absolutevalues. Normalized ERS magnitude was computed asthe percentage of the difference between the meanpower during a chosen reference period (Baseline;1100ms to 100 ms relative to stimulus onset) andthe ERS maximal value in the 0 to 500 ms post stimulus

Fig. 1. Schematic representation of the four tasks. Inthe passive fixation task, letter series identical to the2-back working memory task are presented, but thesubject receives no instruction about the nature ofthe task. In the detection task, the subject must pressa pushbutton as soon as a letter is presented. In the1-back and 2-back tasks, the subject must decidewhether the letter is identical to the one presentedone trial back (1-back task) or two trials back (2-backtask). Stimulus duration 500 ms, interstimulus inter-

val (ISI) 5 s

Frontal theta event-related synchronization 1479


4/10

interval (Maximum), according to the followingformula:

%ERS Maximum Baseline=Baseline 100

Statistical analysis

Differences in behavioral performance were analyzedwith the Wilcoxon non-parametric test. Reaction time(RT) differences were assessed using a one-way re-peated measures ANOVA, with task (passive, detection,1-back, 2-back) as within-subject factor. A logarithmic(log) transformation was performed on theta %ERS inorder to obtain a gaussian distribution of the values.Averaged culmination peak and log %ERS data overfrontal, central, parietal and occipital electrode loca-tions were used for statistics. Statistical analysis wasperformed on ERS culmination peak and log %ERSvalues separately using a two-way repeated measures

ANOVA, with task (passive, detection, 1-back, 2-back)and electrode site (Frontal, Central, Parietal, Occipital)as within-subject factors. Post-hoc analysis was per-formed using Fishers Least Square Difference (LSD)test (Milliken and Johnson, 1984). Statistical thresholdwas set at p


5/10

Fig. 2. Time course of mean event-related synchronisation (ERS) for the 47.5 Hz EEG frequency band as afunction of task and electrode site (n 12). The y-axis depicts average theta power in mV2



6/10

active task (Fisher LSD, detection vs. pas-sive fixation, p


7/10

storage and recall (Klimesch et al., 1994).However, working memory tasks also in-volve a major attentional component, and itremains unclear to which extent theta activity

reflects an attentional framework for memoryintegration, rather than pure working mem-ory processes. We attempted to resolve thisissue by designing appropriate experimentalconditions to dissociate the contribution offocused attention and working memory loadto the genesis of theta activity. In addition,we introduced a fixation task involving mini-mal focused attention and cognitive demand,in order to evaluate the incidence of passivevisual stimulation on theta activity.

Performance was very close across tasks,and thus a similar number of correct re-sponses was analyzed for each task. Further-more, no significant difference in theta ERSwas observed between response and no-response trials, confirming the absence oftheta reactivity to cerebral motor processesrelated to memory matching. Consequently,we opted to pool both types of trials for anal-ysis in order to increase our signal to noiseratio. Theta power was of similar low levelfor all tasks in the reference period preceding

stimulus presentation, suggesting that al-though tasks varied in attention and memorycontent, they involved a comparable amountof neuronal assemblies oscillating at thetafrequency before visual stimulation. Phasictheta ERS started briskly with stimulus onsetover all electrode sites, reaching a maximumat frontal and central electrodes within500 ms after stimulus. This observation isconcordant with previous observation report-ing stimulus phase-locked theta activity turn-ing on during cognitive task and off betweentrials, and culminating early after stimuluspresentation (Caplan et al., 2001; Kahanaet al., 2001; Raghavachari et al., 2001).

The comparison of theta ERS betweentasks led to three main findings. First, frontalphasic theta ERS was found of smalleramplitude in the passive fixation task as com-pared to the two memory tasks and the de-

tection task. This observation supports thehypothesis of a link between task demandand amplitude of frontal theta power. Phasictheta synchronisation emerging in relation to

task demand supports the role of theta activ-ity in the integration of the incoming visualinformation for further cognitive process(Bastiaansen and Hagoort, 2003). Second,the amplitude of theta ERS was larger atfrontal and central electrodes during the de-tection task as compared to the 2-back task,suggesting that attentional rather than mem-ory processes contribute to enhance thetaactivity. Third, we did not observe any sig-nificant difference in theta power between2-back and 1-back tasks, with theta ERStending to be of smaller amplitude in the2-back task. This result contrasts with thestudy of Krause and colleagues who reporteda greater theta ERS at anterior electrode sitesin a visual 2-back letter task compared to a1-back task (Krause et al., 2000). An impor-tant methodological difference relative to theanalysis frequency range might explain suchdiscrepancy. While we chose to analyze thewhole theta 47.5 Hz band, their results wereobtained in the upper 68 Hz theta band, and

observation of their figure reveals strongsimilarities between upper theta and loweralpha power time course. Lower alpha bandERD is generally admitted to reflect generaltask demand, and showed a stronger reactiv-ity to the 2-back task in Krauses study.

Our results suggesting that theta activityprimarily reflects attentional processes, it islegitimate to wonder which neuroanatomicalsubstrate could subtend this widespread thetaoscillatory activity. Pioneer studies on cere-bral attention networks in non-human pri-mates may bring us some clues (Baleydierand Mauguiere, 1985; Mesulam, 1981). Theparietal gyrus, prefrontal cortex, and corticallimbic structures (Mesulam, 1981) weredemonstrated to participate to a distributednetwork of directed attention, together withthe medial pulvinar nucleus (Baleydier andMauguiere, 1985), which would play a key



8/10

role in interconnecting the cerebral structures(Baleydier and Mauguiere, 1985). Indeed,this nucleus projects to and receives inputsfrom the cortical structures of the network.

Most importantly, it receives visual inputsfrom occipital cortex and superior colliculus,and connects to the parahippocampal struc-tures (Baleydier and Mauguiere, 1985, 1987),forming an essential link with hippocampusfor further memory processing (Van Hoesenand Pandya, 1975; Van Hoesen et al., 1979).The early rise and widespread distribution oftheta activity following visual stimulationwould be consistent with the activation ofthe directed attention network via extrastriatepathways, reaching the pulvinar through thesuperior culliculus and being subsequentlydistributed to all components of the network.While it cannot be ruled out that the networkfor directed attention might be activatedafter visual information has reached the stri-ate cortex, the fast increase of theta powershortly after visual stimulus better fits withthe former hypothesis. However, the highpass filters used for theta frequency bandanalysis may produce a shift of theta activityonset, and it is thus difficult to definitely con-

clude on the anatomical pathways involved.Our data support the hypothesis that pha-

sic theta activity may reflect the activation ofthe distributed network for directed attentionwithin the extrapersonal space. Visual stimuliwould activate the network as a functionof task characteristics. Low level attentiontasks, such as passive fixation task, wouldrequire fewer network resources and resultin weaker theta response. Tasks demandingmore attention would recruit enlarged oscil-lating neuronal populations, resulting in en-hanced theta response. The finding of smallertheta ERS amplitude during the 2-backworking memory task as compared to thedetection task could be interpreted as a real-location of resources for memory function-ing, resulting in a net reduction of thetaactivity. Indeed, cognitive theories of work-ing memory underline the interdependence of

attention and working memory processes(Baddeley, 1981, 1992), which have beenshown to share in part common cerebral net-works (LaBar et al., 1999).

We employed the conventional event-related band power measure, which doesnot differentiate between evoked and inducedcomponents. The evoked component corre-sponds to the EEG power phase-locked tothe stimulus, while the induced componentrefers to oscillations modulated by the stim-ulus but not responding in a phase-lockedmanner (Klimesch, 1999; Tallon-Baudryand Bertrand, 1999). The characteristics oftheta ERS recorded in the present paradigmsuggest the contribution of both evoked andinduced components. Theta ERS recordedover the parieto-occipital regions culminatedearlier than fronto-central theta ERS, anddid not reveal any task-related effect. Thisposterior theta activity is likely to haveclose correspondence with evoked visualresponses, as recently demonstrated byKlimesch and colleagues (Klimesch et al.,2004). In contrast, the task dependency ofthe fronto-central theta ERS argues for thedifferent nature of this activity. This hypoth-

esis is supported by previous EEG studiesindicating an independence of frontal andparietal cortical brain regions to the genesisof the theta ERS response during cognitivetasks (Bastiaansen et al., 2002; Jensen andTesche, 2002; Raghavachari et al., 2001;Sederberg et al., 2003). However furthermethodological investigation is necessary todetermine whether frontal theta ERS is in-deed dominated by induced oscillations.

In conclusion, our results suggest that theearly theta power increase during workingmemory tasks is not mainly related to long-term encoding and=or retrieval, but ratherrepresents an EEG correlate of focused atten-tion to target stimuli.

Acknowledgements

This project was funded by the Swiss National Founda-tion for Scientific Research, grant 3100-59110.99 (PG),

1484 P. Missonnier et al.


9/10

the Research and Development HUG grant 03-I-15(VI) and the Jeeroome Tissieeres Foundation, Stiftung fuurKlinische Neuropsychiatrische Forschung and Ernstand Lucie Schmidheing Foundation (PG).

ReferencesBaddeley A (1981) The concept of working memory: a

view of its current state and probable future devel-opment. Cognition 10: 1723

Baddeley A (1992) Working memory. Science 255:556559

Baleydier C, Mauguiere F (1985) Anatomical evidencefor medial pulvinar connections with the posteriorcingulate cortex, the retrosplenial area, and theposterior parahippocampal gyrus in monkeys.J Comp Neurol 232: 219228

Baleydier C, Mauguiere F (1987) Network organiza-tion of the connectivity between parietal area 7,

posterior cingulate cortex and medial pulvinarnucleus: a double fluorescent tracer study inmonkey. Exp Brain Res 66: 385393

Basar-Eroglu C, Basar E, Demiralp T, Schurmann M(1992) P300-response: possible psychophysiologi-cal correlates in delta and theta frequency channels.A review. Int J Psychophysiol 13: 161179

Bastiaansen M, Hagoort P (2003) Event-induced thetaresponses as a window on the dynamics of memory.Cortex 39: 967992

Bastiaansen MC, van Berkum JJ, Hagoort P (2002)Event-related theta power increases in the humanEEG during online sentence processing. Neurosci

Lett 323: 1316Caplan JB, Madsen JR, Raghavachari S, Kahana MJ(2001) Distinct patterns of brain oscillations under-lie two basic parameters of human maze learning.J Neurophysiol 86: 368380

Demiralp T, Basar-Eroglu C, Rahn E, Basar E (1994)Event-related theta rhythms in cat hippocampusand prefrontal cortex during an omitted stimulusparadigm. Int J Psychophysiol 18: 3548

Doppelmayr M, Klimesch W, Schwaiger J, Stadler W,Rohm D (2000) The time locked theta re-sponse reflects interindividual differences in hu-man memory performance. Neurosci Lett 278:141144

Gevins A, Smith ME, McEvoy L, Yu D (1997) High-resolution EEG mapping of cortical activationrelated to working memory: effects of task diffi-culty, type of processing, and practice. CerebCortex 7: 374385

Jensen O, Tesche CD (2002) Frontal theta activity inhumans increases with memory load in a workingmemory task. Eur J Neurosci 15: 13951399

Kahana MJ, Seelig D, Madsen JR (2001) Theta returns.Curr Opin Neurobiol 11: 739744

Klimesch W (1999) EEG alpha and theta oscillationsreflect cognitive and memory performance: areview and analysis. Brain Res Brain Res Rev29: 169195

Klimesch W, Doppelmayr M, Pachinger T, Ripper B

(1997) Brain oscillations and human memory: EEGcorrelates in the upper alpha and theta band. Neu-rosci Lett 238: 912

Klimesch W, Schack B, Schabus M, Doppelmayr M,Gruber W, Sauseng P (2004) Phase-locked alphaand theta oscillations generate the P1-N1 complexand are related to memory performance. Brain ResCogn Brain Res 19: 302316

Klimesch W, Schimke H, Schwaiger J (1994) Episodicand semantic memory: an analysis in the EEG thetaand alpha band. Electroencephalogr Clin Neuro-physiol 91: 428441

Krause CM, Sillanmaki L, Koivisto M, Saarela C,Haggqvist A, Laine M, Hamalainen H (2000)The effects of memory load on event-relatedEEG desynchronization and synchronization. ClinNeurophysiol 111: 20712078

LaBar KS,Gitelman DR, ParrishTB, Mesulam M (1999)Neuroanatomic overlap of working memory andspatial attention networks: a functional MRI com-parison within subjects. Neuroimage 10: 695704

McEvoy LK, Pellouchoud E, Smith ME, Gevins A(2001) Neurophysiological signals of workingmemory in normal aging. Brain Res Cogn BrainRes 11: 363376

Mesulam MM (1981) A cortical network for directedattention and unilateral neglect. Ann Neurol 10:

309325Milliken GA, Johnson DE (1984) Designed experi-

ments, vol 1. Van Nostrand Reinhold, New YorkOKeefe J (1993) Hippocampus, theta, and spatial

memory. Curr Opin Neurobiol 3: 917924Pfurtscheller G (1992) Event-related synchronization

(ERS): an electrophysiological correlate of corticalareas at rest. Electroencephalogr Clin Neurophy-siol 83: 6269

Pfurtscheller G (2001) Functional brain imaging basedon ERD=ERS. Vision Res 41: 12571260

Pfurtscheller G, Aranibar A (1977) Event-relatedcortical desynchronization detected by power mea-surements of scalp EEG. Electroencephalogr Clin

Neurophysiol 42: 817826Raghavachari S, Kahana MJ, Rizzuto DS, Caplan JB,

Kirschen MP, Bourgeois B, Madsen JR, Lisman JE(2001) Gating of human theta oscillations by aworking memory task. J Neurosci 21: 31753183

Sederberg PB, Kahana MJ, Howard MW, Donner EJ,Madsen JR (2003) Theta and gamma oscillationsduring encoding predict subsequent recall. JNeurosci 23: 1080910814



10/10

Tallon-Baudry C, Bertrand O (1999) Oscillatorygamma activity in humans and its role in objectrepresentation. Trends Cogn Sci 3: 151162

Van Hoesen G, Pandya DN (1975) Some connectionsof the entorhinal (area 28) and perirhinal (area 35)

cortices of the rhesus monkey. I. Temporal lobeafferents. Brain Res 95: 124

Van Hoesen GW, Rosene DL, Mesulam MM (1979)Subicular input from temporal cortex in the rhesusmonkey. Science 205: 608610

Wetzel W, Ott TA, Matthies H (1977) Post-traininghippocampal rhythmic slow activity (theta) eli-

cited by septal stimulation improves memory con-solidation in rats. Behav Biol 21: 3240

Winson J (1978) Loss of hippocampal theta rhythmresults in spatial memory deficit in the rat. Science201: 160163

Authors address: P. Missonnier, Department ofPsychiatry, Neuroimaging Unit, University of GenevaHospitals, Ch. du Petit-Bel-Air 2, 1225 Geneva,Switzerland, e-mail: [email protected]

1486 P. Missonnier et al.: Frontal theta event-related synchronization

Documents

Missoner, Et Al 2006