The Deployment of Intersensory Selective Attention: A High-density Electrical Mapping Study of the Effects of Theanine

Abstract

: Ingestion of the nonproteinic amino acid theanine (5-N-ethylglutamine) has been shown to increase oscillatory brain activity in the so-called alpha band (8-14 Hz) during resting electroencephalographic recordings in humans. Independently, alpha band activity has been shown to be a key component in selective attentional processes. Here, we set out to assess whether theanine would cause modulation of anticipatory alpha activity during selective attentional deployments to stimuli in different sensory modalities, a paradigm in which robust alpha attention effects have previously been established.
: Electrophysiological data from 168 scalp electrode channels were recorded while participants performed a standard intersensory attentional cuing task.
: As in previous studies, significantly greater alpha band activity was measured over parieto-occipital scalp for attentional deployments to the auditory modality than to the visual modality. Theanine ingestion resulted in a substantial overall decrease in background alpha levels relative to placebo while subjects were actively performing this demanding attention task. Despite this decrease in background alpha activity, attention-related alpha effects were significantly greater for the theanine condition.
: This increase of attention-related anticipatory alpha over the right parieto-occipital scalp suggests that theanine may have a specific effect on the brain's attention circuitry. We conclude that theanine has clear psychoactive properties, and that it represents a potentially interesting, naturally occurring compound for further study, as it relates to the brain's attentional system.

Full text

The Deployment of Intersensory Selective
Attention: A High-density Electrical
Mapping Study of the Effects of Theanine
Manuel Gomez-Ramirez, BA,*†Beth A. Higgins, BS,†
Jane A. Rycroft, PhD,‡Gail N. Owen, PhD,‡Jeannette Mahoney, BA,†x
Marina Shpaner, BS,*†and John J. Foxe, PhD*†
Abstract
Objective:
Ingestion of the nonproteinic amino acid theanine
(5-N-ethylglutamine) has been shown to increase
oscillatory brain activity in the so-called alpha
band (8Y14 Hz) during resting electroencephalo-
graphic recordings in humans. Independently,
alpha band activity has been shown to be a key
component in selective attentional processes.
Here, we set out to assess whether theanine would
cause modulation of anticipatory alpha activity
during selective attentional deployments to stim-
uli in different sensory modalities, a paradigm in
which robust alpha attention effects have previ-
ously been established.
Methods:
Electrophysiological data from 168 scalp electrode
channels were recorded while participants per-
formed a standard intersensory attentional cuing
task.
Results:
As in previous studies, significantly greater alpha
band activity was measured over parieto-occipital
scalp for attentional deployments to the auditory
modality than to the visual modality. Theanine
ingestion resulted in a substantial overall decrease
in background alpha levels relative to placebo
while subjects were actively performing this
demanding attention task. Despite this decrease
in background alpha activity, attention-related
alpha effects were significantly greater for the
theanine condition.
Conclusion:
This increase of attention-related anticipatory
alpha over the right parieto-occipital scalp sug-
gests that theanine may have a specific effect on
the brain’s attention circuitry. We conclude that
theanine has clear psychoactive properties, and
that it represents a potentially interesting, natu-
rally occurring compound for further study, as it
relates to the brain’s attentional system.
Key Words: theanine, attention, attentional
deployment, alpha oscillations, multisensory, EEG
(Clin Neuropharmacol 2007;30:25Y38)
Theanine (5-N-ethylglutamine) is a natu-
rally occurring nonproteinic amino acid
that, along with caffeine and cathecins, is 1
of 3 major constituents of tea (Camellia
sinensis). Found only in tea and one rare
mushroom variety (Xerocomus badius), it is
relatively easy to extract from tea, and it is
not uncommon for it to be consumed in its
purified form as a dietary supplement, with
many of the commercial concerns proffering
claims regarding beneficial psychological
and physiological effects. Regrettably, the
vast majority of these claims remain to be
properly substantiated in clinical trials. One
claim that seems to have some reasonable
measure of physiological support, however,
relates to potential effects on the resting
brain state. That is, ingestion of theanine
has been shown to increase oscillatory
brain activity in the so-called alpha band
(8Y14 Hz) when subjects are in a passive
resting state.
1
Indeed, this brain rhythm has
traditionally been associated with a relaxed
state.
2
This finding of increased alpha activ-
ity because of theanine ingestion was
recently corroborated by Nobre et al at the
University of Oxford (personal communica-
tion May, 2003). An increase in alpha activity
implies that theanine or one of its metabo-
lites is likely psychoactive and warrants
further investigation.
Activity in the alpha band has classi-
cally been associated with the general state
DOI: 10.1097/01.WNF.0000240940.13876.17
25
Original Article CLINICAL
NEUROPHARMACOLOGY
Volume 30, Number 1
January /February 2007
*Program in Cognitive
Neuroscience, Department of
Psychology, The City College of
the City University of New York,
New York, NY; †The Cognitive
Neurophysiology Laboratory,
Nathan S. Kline Institute for
Psychiatric Research, Program in
Cognitive Neuroscience and
Schizophrenia, Orangeburg, NY;
‡Unilever Beverages Global
Technology Centre, Colworth
House, Unilever R&D Colworth,
Sharnbrook, Bedfordshire, UK;
and xFerkauf Graduate School of
Psychology, Albert Einstein
College of Medicine, Bronx, NY.
Address correspondence and
reprint requests to: John J. Foxe,
PhD, The Cognitive
Neurophysiology Laboratory,
Nathan S. Kline Institute for
Psychiatric Research, Program in
Cognitive Neuroscience and
Schizophrenia, 140 Old
Orangeburg Road, Orangeburg,
NY 10962; E-mail: foxe@nki.
rfmh.org
Supported by a grant from the
Unilever Beverages Global
Technology Centre.
Copyright Ó2007 by Lippincott
Williams & Wilkins
Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

of mental alertness or arousal,
3
implying in
turn that theanine may exert effects on the
level of alertness or attentiveness of humans.
Indeed, anecdotal reports from tea drinkers
often emphasize changes in general arousal
and relaxation state. An open question,
however, is whether the effects of theanine
on oscillatory brain mechanisms are rela-
tively nonspecific, limited to more general
arousal states during periods of rest or
passivity, or whether this compound also
operates in a more specific manner on
attentional control systems of the brain
when subjects are engaged in more demand-
ing and mentally taxing tasks.
Previous studies have shown that oscil-
lations in the alpha band are not simply
associated with brain arousal states, but that
these oscillatory potentials are also key
components in selective attentional pro-
cesses.
4Y15
That is, directed deployment of
alpha oscillatory activity seems to underlie
the suppression of distracting information in
visual space during highly demanding atten-
tion tasks. For example, it has been shown
that there is a highly specific pattern of alpha
oscillatory activity when humans deploy
their attention to stimulation in one sensory
modality (eg, audition) while attempting to
exclude or suppress inputs from an interfer-
ing sensory modality (eg, vision).
5,7,8
In one such study, Foxe et al
5
used
symbolic cue stimuli to instruct subjects to
attend to a given sensory modality (either
audition or vision). After a 1-second delay, a
compound audiovisual multisensory stimulus
appeared, and subjects performed a very
demanding task solely within the cued
modality while attempting to suppress
inputs from the other distracting modality.
It was shown that deploying attention to
impending stimuli in the auditory modality
resulted in a robust increase in alpha power
during the late phase of the cue-target
interval (CTI), that is, the attentional deploy-
ment phase, over parieto-occipital scalp
regions. It was proposed that this oscillatory
alpha enhancement reflected anticipatory
gating or suppression of visual processing
by parieto-occipital structures known to be
involved in attentional switching and disen-
gagement within the visual modality.
5,6,16,17
A subsequent study showed that similar
alpha processes were deployed to locations
in space where distracters might appear
during a demanding visuospatial selective
attention task,
6
reinforcing the role of alpha
processes in attentional deployment and
distracter suppression.
10,12
The present study set out to assess
whether theanine would cause modulation
of anticipatory alpha activity during selective,
attentional deployments to stimuli in differ-
ent sensory modalities. Here, subjects per-
formed a task similar to that of Foxe et al
18,19
in which they were instructed to deploy
their attention to a given modality (auditory
or visual) on a trial-by-trial basis. By inspect-
ing the alpha band activity during the CTI,
the effects of theanine on intersensory atten-
tional deployments were assessed. A simple
hypothesis was tested. We reasoned that
because theanine is now known to affect
overall alpha band activity during the resting
state, it might also enhance (or indeed
suppress) alpha-based attentional mecha-
nisms. If the latter is the case, this would
provide direct evidence that theanine does
not simply have a general effect on arousal
but can affect more specific attentional brain
states. Furthermore, if theanine is found to
enhance attentional processes, this com-
pound might well represent a potentially
attractive, naturally occurring, intervention
for attentional disorders.
METHODS
Participants
Fifteen (eight women) neurologically
healthy, paid volunteers (mean age, 27.8
years, SD T6.5 years, range, 18Y36 years)
participated. All participants provided writ-
ten informed consent, and the institutional
review board of the Nathan Kline Research
Institute approved the procedures. All par-
ticipants reported normal or corrected-
to-normal vision, and all were right-hand
Gomez-Ramirez et al
Ó2007 Lippincott Williams & Wilkins
CLINICAL
NEUROPHARMACOLOGY
Volume 30, Number 1
January /February 2007
26
Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

dominant as assessed by the Edinburgh Hand-
edness Inventory.
20
Subjects were required
to refrain from drinking any caffeine-based
products (such as soft drinks, soda, coffee,
or tea) for at least 24 hours before the day
of testing. Subjects’ neurological status
was determined by conducting a shortened
version of the Structured Clinical Interview
for Diagnostic and Statistical Manual of
Mental Disorders, Fourth EditionYText
Revision.
Experimental Paradigm
The sequence of events in a typical
trial is illustrated in Figure 1. A trial com-
menced with the onset of the cue stimulus
(S1) instructing participants as to which
modality they should attend. Stimulus onset
asynchrony between S1 and S2 (ie, the cue
and the subsequent potential target) was
1200 milliseconds. The intertrial interval
(ITI), that is, the period between each pair
of cue-target stimuli, randomly varied
between 1400 and 2400 milliseconds. A
central fixation cross (black and 1-degree
angle) remained on the screen throughout
the experiment, and participants were
instructed to maintain fixation at all times.
All visual stimuli were displayed on a gray
background, whereas all auditory stimuli
were presented through headphones. Partic-
ipants completed a minimum of 15 blocks of
trials, on each of the 2 days of testing. Each
block contained a total of 100 S1 to S2 pairs,
giving an average block run time of 5
FIGURE 1. A, Schematic illustration of the paradigm. Each trial commenced with the presentation of
an auditory symbolic cue (S1) whereby the Aud (for auditory) or Vis (for visual) were presented
across headphones to the subjects. A delay period of 800 milliseconds followed the cue, after which
1 of 4 types of S2 stimuli would appear: auditory stimuli alone, visual stimuli alone, catch trials
(where no physical stimulus was presented), or compound audiovisual multisensory stimuli. Subjects
were required to respond with a push button to targets within the cued modality. B, An example of
a visual target trial. A pair of Gabor patches was presented bilaterally with the orientation of one
slightly different to the other. Target trials occurred on only 20% of visual S2s. On the other 80% of
trials, the orientation of both patches was identical. The orientation difference was psychophysically
calibrated on an individual basis such that subjects could only identify 80% of targets correctly.
Theanine and Intersensory Attention CLINICAL
NEUROPHARMACOLOGY
Volume 30, Number 1
January /February 2007
27
Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

minutes. The specific details of the stimuli
themselves follow.
S1 (Cues)
The cue stimulus (S1) was presented
auditarily and was either the word ‘‘AUD’’ or
‘‘VIS,’’ spoken by a male speaker and pre-
sented bilaterally via Sennheiser HD600
headphones ([Sennheiser Headphones, Old
Lyme, Conn] duration, 400 milliseconds; 10-
millisecond rise/fall, 80 dB sound pressure
level). These word cues indicated the sen-
sory modality (auditory or visual) to be
attended when the second stimulus (S2)
arrived. Specifically, the pseudoword
‘‘AUD’’ instructed participants to deploy
their attention to the auditory modality,
whereas the pseudoword ‘‘VIS’’ instructed
theparticipantstoattendtothevisual
modality. These cue-stimuli appeared in
random order throughout the experiment.
S2 (Targets)
The S2 consisted of an auditory-alone
(pair of tones, 20%), visual-alone (pair of Gabor
patches, 20%), audio-visual compound (pair of
tones and Gabor patches, 40%), or a null
(catch) stimulus (no S2 stimulus, 20%). The
auditory S2 stimulus consisted of a pair of
binaural tones (eg, 2000 or 2100 Hz, 80 dB
sound pressure level, 100-millisecond duration
each, 5-millisecond rise/fall, 5-millisecond
interval between tones). That is, the pair of
tones was presented in a rapid sequence with
auditory stimulation lasting for a total of 205
milliseconds. The visual stimulus consisted of
a pair of Gabor patches (4.5 degrees in
diameter, centered 2.5 degrees left and right
of fixation, 100-millisecond duration). The
audiovisual compound stimuli were a combi-
nation of the previously described auditory
and visual stimuli. The null stimulus consisted
of a period of fixation only. All visual stimuli
were presented on an Iiyama VisionMaster
Pro502 21-inch computer monitor on a gray
background.
On 85% of visual S2 trials, the pair of
Gabor patches was identical, and no overt
response was required (ie, these were non-
target standards). On 85% of auditory S2
trials, the pair of tones was also identical,
and no response was required. On the other
15% of visual S2 trials, the orientations of the
left and right Gabor patches were slightly
different (see later), and subject responded
with a push button (ie, these were target
stimuli) when they had been cued to the
visual modality. On 15% of auditory trials,
the 2 tones were of slightly different pitch
andalsorequiredaresponsewhenthe
auditory modality had been cued. The reader
should note that the occurrence of a target
in each modality was independent of the
other modality, such that the probability of
occurrence of a double target (ie, a bisensory
target) was approximately 1%.
This scenario led to 4 possible target
types: (1) unisensory-auditory (ie, when the
cue instructed the subject to attend to the
auditory modality, and only an auditory S2
stimulus occurred); (2) unisensory-visual, (3)
multisensory-auditory (ie, when the cue
instructed the subject to attend to the
auditory modality, and an audiovisual com-
pound multisensory S2 stimulus occurred),
and (4) multisensory-visual.
Procedure
At the beginning of each experimental
day, all participants were given either a mixed
solution of the theanine substance or a
placebo drink. The mixed drink solution
consisted of a mixture of 250 mg of powdered
clear theanine with 200 mL of room temper-
ature water. The placebo drink consisted only
of the 200 mL of water (ie, approximately a
cup of water). The day of drinking the
theanine solution was counterbalanced
across participants. Note that theanine is
colorless and tasteless in a water solution
and subjects were at chance in guessing
whether they were taking the active com-
pound or simply water.
On each experimental day, before the
start of the testing phase, all participants
performed a psychophysical test that equa-
ted the participant’s performance to an 80%
level on both the auditory and visual tasks to
Gomez-Ramirez et al
Ó2007 Lippincott Williams & Wilkins
CLINICAL
NEUROPHARMACOLOGY
Volume 30, Number 1
January /February 2007
28
Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

be used herein. This psychophysical test,
known as the Up-Down Transformed Rule,
is a rapid method that assures the experi-
menters that the level of task difficulty for
each task will be equivalent across all
participants.
21
Measurements
Continuous EEG data, digitized at 512
Hz, were acquired through the ActiveTwo
Biosemi electrode system (Active Two Bio-
semi System: Amsterdam, Netherlands) from
168 scalp electrodes. BioSemi replaces the
‘‘ground’’ electrodes used in conventional
EEG systems with 2 separate electrodes; the
so-called Common Mode Sense active elec-
trode and the Driven Right Leg passive
electrode. These 2 electrodes form a feedback
loop, which drives the average potential of the
subject (the Common Mode voltage) as close
as possible to the analog<to<digital converter
(ADC) reference voltage in the AD box (the
ADC reference can be considered as the
amplifier ‘‘zero’’). With the Biosemi system,
every electrode or combination of electrodes
can be assigned as the ‘‘reference,’’ and this is
done purely in software after acquisition. A
detailed description of the referencing con-
ventions used by this active electrode system
can be found at the following Web site: http://
www.biosemi.com/faq/cms&drl.htm.
All data were re-referenced to a midline
prefrontal scalp site after acquisition. After
each recording session, before the electrode
cap was removed from the subject’s head,
the 3-dimensional coordinates of the electro-
des with reference to anatomical landmarks
on the head (nasion, preauricular notches)
were digitized using a Polhemus magnetic
3-dimensional digitizer (Polhemus: La Jolla,
Calif). This allowed the investigators to
ensure that electrode locations were consis-
tent between test days for all subjects. The
EEG was recorded continuously and epoched
and averaged off-line. Trials with blinks
and large eye movements were rejected off-
line on the basis of horizontal and vertical
electrooculogram recordings. An artifact
rejection criterion of T100 2V was used at
all other electrode sites to exclude periods
of high EMG and other noise transients.
Data Analysis Strategy
Accepted trials were epoched for the
period around the onset of the S1 cues
(j300 milliseconds prestimulus to 1200
milliseconds poststimulus). Note that only
the S1 stimuli are analyzed for the present
study. The baseline was defined as the mean
voltage from 300 to 0 millisecond before the
onset of S1. Separate averages were made
for the 2 possible variants of the S1 stimuli
(cue-AUD and cue-VIS). We inspected oscil-
latory activity in the alpha band (8Y14 Hz
1
)
during the CTI. Alpha band activity was
characterized in this period by the temporal
spectral evolution (TSE) technique, which
provides an index of induced alpha activity
as a function of time (see Fu et al
7
for a full
description of the method). All statistical
analyses were performed on these induced
alpha oscillations. The TSE waveforms are
derived by the following method:
&Individual (single trial) stimulus-locked
epochs are band-pass filtered after artifact
rejection (Butterworth zero-phase, 8Y14
Hz, 48 dB/octave).
&Filtered epochs are then full-wave rectified
(ie, all negative data points are made
positive).
&Rectified waveforms are then averaged.
Repeated-measures analysis of variance
(ANOVA) was used to statistically test for
effects over a region of interest (ROI)
defined by 4 electrode sites over the right
parieto-occipital scalp (the ROI chosen here
was based on previous work from this
laboratory, where maximal alpha-dependent
attention effects have been defined
5
). Factors
were attention condition (auditory vs visual)
and treatment (theanine vs placebo). The
1
The exact band-pass that constitutes the alpha band is not
consistent across the literature and could be considered
somewhat arbitrary. In fact, the centre frequency of
alpha is quite variable across individuals and for most it
tends to be in the 10Y12 Hz range.
22
As such, the band-
pass chosen here of 8Y14 Hz nicely spans this range.
Theanine and Intersensory Attention CLINICAL
NEUROPHARMACOLOGY
Volume 30, Number 1
January /February 2007
29
Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

epoch chosen for testing here (700Y1100
milliseconds) was also based on our previous
work,
5Y7
where it was shown that the alpha
effect was maximal in the latter part of the
CTI in the period preceding the onset of the
S2 (ie, for approximately 400 milliseconds).
The dependent measure was an integrated
area measure across this 400-millisecond
epochVthat is, how far was the TSE wave-
form displaced from zero.
A secondary analysis of the relative
attentional differences between treatment
conditions in induced oscillations was also
performed. Specifically, we derived an atten-
tional modulation index. This technique is a
useful metric to normalize the data set with
regard to background alpha levels. Modula-
tion indices on the induced alpha oscillations
were derived by the following method:
&Step 1: An average background alpha
activity value was computed for each treat-
ment condition across the 1200-millisecond
CTI.
&Step 2: After computing this average, the
difference in alpha amplitude between the
attend-auditory and attend-visual condition
during the 400-millisecond period from
700 to 1100 milliseconds was divided by
the appropriate mean value computed in
step 1.
RESULTS
Performance Data
Percent correct-hit responses (accu-
racy) and reaction time (RT) mean values are
shown in Figure 2. Two 2 4 repeated-
measures ANOVAs with factors of treatment
(placebo vs theanine) and S2 target type
FIGURE 2. Behavioral data. Hit rates are plotted for each treatment in the upper graph. Mean
reaction times for correct responses are plotted in the lower graph. Asterisks illustrate statistically
significant effects between conditions.
Gomez-Ramirez et al
Ó2007 Lippincott Williams & Wilkins
CLINICAL
NEUROPHARMACOLOGY
Volume 30, Number 1
January /February 2007
30
Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

(unisensory-auditory, multisensory-auditory,
unisensory-visual, multisensory-visual) were
computed for both the accuracy and RT data.
For accuracy, a main effect of stimulus
type was found (F
3,45
= 12.364, PG0.01),
driven by a considerable performance decre-
ment that was seen for the multisensory-
auditory stimulus type during both treat-
ment conditions. The reader will recall that
the auditory task is identical between the
unisensory-auditory and multisensory-auditory
conditions, so the presence of distracting
visual information in the latter case had a
large impact on performance accuracy. No
other effects reached significanceVthat is,
accuracy was not found to be different in
any condition between treatments (ie, pla-
cebo vs theanine).
For the RT data, a main effect of stimulus
type was found (F
1,14
= 12.23, PG0.01)
caused mainly by faster RTs to the unisensory-
visual stimuli, again observed during both
treatment types. An interaction of treatment
stimulus type (F
3,45
= 5.96, PG0.05) was
also seen. This was because RTs were consid-
erably slower during the theanine treatment
than placebo for both the unisensory-auditory
and multisensory-auditory stimulus types. No
other significant effects were found.
Induced Alpha Band Activity
Alpha band activity during the atten-
tional deployment phase is illustrated in
Figure 3. Depicted are average 8- to 14-Hz
TSE waveforms across the four electrode
sites chosen for testing over the right
parieto-occipital scalp region. It can be seen
that the divergence in the alpha band activity
starts to occur at approximately 400 milli-
seconds and is sustained until the onset of
the S2 stimulus at 1200 milliseconds, where
the largest difference is observed.
A repeated-measures ANOVA per-
formed on these TSE waveforms revealed a
FIGURE 3. Alpha band oscillatory activity is selectively modulated by the deployment of anticipatory
attention to different sensory modalities. The TSE waveforms, averaged from the 4 scalp sites
illustrated over the right parieto-occipital scalp, are plotted for placebo (upper panel) and theanine
conditions (lower panel). A sustained divergence in TSE amplitude is seen starting at approximately
400 milliseconds after cue. In both treatment conditions, the alpha band activity is significantly
greater when subjects have been cued to attend selectively to impending auditory stimulation (blue
trace) compared with visual stimulation (red trace). However, this sustained divergence is significantly
greater during the theanine condition compared with the placebo condition.
Theanine and Intersensory Attention CLINICAL
NEUROPHARMACOLOGY
Volume 30, Number 1
January /February 2007
31
Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

main effect of treatment (F
1,14
= 9.112, PG
0.01), indicating that the overall alpha activ-
ity in the late sustained period is significantly
reduced by the theanine solution. Figure 3
shows that this drop in the alpha power is
present across the whole CTI, not just in the
late sustained period.
The ANOVA also revealed a main effect
of attention (F
1,14
= 9.776, PG0.01). A trend
toward an interaction of treatment atten-
tion was also observed (F
1,14
= 3.056, P=
0.1). No other effects reached significance.
In the second phase of our analysis, we
derived an attention modulation index. After
the computation of these indices, a paired
ttest between the treatment conditions
(collapsed across attention conditions) was
computed on the same ROI mentioned
previously. The ttest revealed a significant
difference (t
14
=j2.465, P= 0.027), indicat-
ing that the overall amplitude of the alpha
attention effect as a function of background
alpha was significantly greater for the the-
anine treatment day than the placebo day.
To investigate the topographic distribu-
tion of the induced alpha effect, the TSE
response of the attend-visual condition was
subtracted from the TSE response to the
attend-auditory condition. This subtraction
results in a derived response representing only
the differential activity between attention
conditions. After this subtraction, the resulting
waveform was modeled using the minimum
norm solution technique of the BESA 5.01
software (Brain Electric Source Analysis, Gra¨-
felfing, Germany. Available at: www.besa.de).
The topographic analysis of the
induced alpha band oscillatory effect reveals
a distribution over bilateral parieto-occipital
regions, with a predominance of activity
over the right hemisphere (Fig. 4). The
theanine treatment condition elicits a greater
differential and sustained alpha effect over
the period.
DISCUSSION
The present study showed that thea-
nine consumption caused substantial effects
on the brain’s generation of the alpha
oscillatory rhythm (8Y14 Hz) while subjects
performed a demanding intersensory selec-
tive attention task. First, theanine consump-
tion resulted in a substantial decrease in
overall background alpha power, regardless
of which sensory modality was being
attended. Second, theanine caused a relative
amplitude increase of a previously defined
alpha attention effect associated with the
deployment of intersensory selective atten-
tion.
5,7
These earlier studies showed that
cuing attention to the auditory features of an
imminent, compound audiovisual stimulus
resulted in significantly higher alpha ampli-
tude in the period preceding the onset of
this stimulus than when attention was cued
to the visual features. Here, we found that
this difference in anticipatory alpha activity
was relatively greater when subjects had
ingested theanine, suggesting that theanine
may have enhanced this well-characterized
attention effect. This anticipatory alpha
activity is generated within the right parieto-
occipital cortex, a critical node of the brain’s
attention circuitry, further suggesting that
theanine, rather than having a more general
effect on arousal levels, may be having more
specific effects on the brain’s attention
circuitry. These and other implications of
the present data are discussed in what
follows.
A Resting Versus Active
Alpha Paradox
The motivation for the present study
was that theanine has been shown to
increase ongoing tonic alpha power during
the resting or passive state. That is, the 2
previous studies of the neurophysiology of
theanine in humans have simply recorded
ongoing EEG while subjects sat passively and
were not performing any cognitive task.
1
That theanine might enhance alpha activity
is of real interest, as increased alpha ampli-
tude may have implications for general
cognitive performance abilities. For exam-
ple, it has been shown that children with
Gomez-Ramirez et al
Ó2007 Lippincott Williams & Wilkins
CLINICAL
NEUROPHARMACOLOGY
Volume 30, Number 1
January /February 2007
32
Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

an impoverished educational background
and low reading and writing scores show
generally reduced resting alpha power.
23
The same appears to be true of children
with attention deficit/hyperactive disorder
who show generally reduced alpha power,
24
and evidence also suggests that this reduc-
tion can be normalized through medica-
tion.
25,26
A relationship between increasing
alpha across practice sessions and improved
performance on a verbal working memory
task has also been shown.
27
Most recently, it
has been shown that performance on a
visuospatial attention task was predicated
on the amplitude of preceding anticipatory
alpha.
13
In a design closely modeled on the
original one used by Worden et al,
6
subjects
were cued by a central arrow to attend to
either the left or right hemifields for an
impending visual target. Reaction times to
detected targets were faster on the attended
side, and for both left and right targets, RTs
varied as a function of the amplitude of
anticipatory alpha, with faster RTs to right
targets when alpha was greater over the
right hemisphere (lower on the left) and
faster RTs to left targets when alpha was
greater on the left (lower on the right).
Overall, there seems to be growing evidence
that increased tonic alpha power (ie, during
the resting state) may be related to better
cognitive performance abilities, and that the
strength of alpha processes during more
active tasks may be a predictor of perform-
ance on demanding attentional tasks.
Although the present study was
designed to specifically assess alpha atten-
tion effects, 1 surprising and apparently
paradoxical outcome was that overall back-
ground power was significantly lower in the
theanine treatment condition compared with
placebo. It is important to point out that this
reduction in alpha power is only seen while
subjects are engaged in a very demanding
cognitive task. In this sense, our study is
entirely different from the previous EEG
studies. Unfortunately, in the present study,
a measure of tonic alpha was not taken
during a resting condition as in the 2
previous studies, which is why a direct com-
parison between these results and previous
work is not possible. It should also be noted
FIGURE 4. Topographic analysis using the minimum norm solution reveals a clear predominance
of alpha oscillatory activity over the right parieto-occipital region. The theanine treatment condi-
tion (lower field) elicits a greater differential and sustained alpha effect compared with placebo
(upper field).
Theanine and Intersensory Attention CLINICAL
NEUROPHARMACOLOGY
Volume 30, Number 1
January /February 2007
33
Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

that despite the overall drop in background
alpha observed here, this had no effect on
the subjects’ ability to effectively deploy
alpha attention mechanisms because the
differential alpha effect was greater in the
theanine condition. That is, although back-
ground alpha dropped in amplitude, concur-
rent attention-related alpha processes were
enhanced. As such, these results may suggest
that overall background alpha and attention-
related alpha processes represent distinct
cognitive functions, a possibility meriting
further investigation.
Another consideration here is the
makeup of the subject group in our study
relative to the previous studies, which were
carried out in Japan
1
and England (Nobre
Laboratory). The young American adults
who served in our study were not regular
tea drinkers, and as such, had little to no
previous exposure to theanine. This is
unlikelytohavebeenthecaseinthe
previous studies as tea drinking is consider-
ably more prevalent in both countries than it
is in the United States. Levels of previous and
ongoing exposure to theanine are factors
that will have to be accounted for in future
studies of this compound.
Visual Dominance During
Multisensory Target Trials
Onedramaticeffectseenherein
behavioral performance (for both the pla-
cebo and theanine conditions) was a sub-
stantial reduction in performance accuracy
for the auditory task when visual stimuli
were concurrently presentedVthat is, dur-
ing the audiovisual multisensory target con-
dition. The reader will recall that both the
auditory and visual tasks were psychophysi-
cally equated before any recordings were
made, with performance calibrated such that
all subjects detected exactly 80% of targets in
both sensory modalities. However, this pre-
test psychophysical calibration was only
conducted for the unisensory stimulus con-
ditions. Behavioral results show that the
manipulation was fully effective as perfor-
mance remained at or very close to 80%
for both unisensory stimulus conditions
across both treatment days. The same result
extends to the multisensory-visual targets,
where subjects also achieved about 80%
accuracy, indicating that the presence of
simultaneous distracting auditory inputs had
no effect on performance of the visual task.
The reverse was not the case. Subjects’
performance dropped substantially for the
auditory task when concurrent distracting
visual inputs were present (from 80% to
approximately 50%).
This finding has implications for 1
prominent view in attentional theory. A
central tenet of modern theory is that
selective attention is necessary because of
the limited informational processing capacity
of the brain.
28
That is, selective attention
functions to filter the overwhelming quantity
of information constantly impinging on our
senses, allowing for preferential processing
of a subset of these inputs. Implicit in this
construct is the notion that stimuli that are
not in the attentional focus receive reduced
processing. The existence of a limited
capacity is seen in divided attention tasks,
where subjects monitor 2 streams simulta-
neously, and clear deficits in performance
within a given stream are seen because of
interference from the second stream.
29,30
However, some have claimed that this capacity
limitation only occurs when the 2 streams of
information are within a single sensory modal-
ity, and that no such capacity limit is present
when the streams of information occur within
separate sensory systems. For example, in a
seminal article, Duncan et al
31
used the so-
called attentional blink (AB) paradigm to
show that there was no interference
between modalities. In the AB paradigm,
subjects are typically required to respond to
2 target stimuli that occur in relatively close
temporal proximity within a stream of dis-
tracters. Identification of the first target in
the sequence produces a sustained reduction
in the ability to identify the second target
stimulus when the 2 targets are presented
within a single modalityVthat is, when all
Gomez-Ramirez et al
Ó2007 Lippincott Williams & Wilkins
CLINICAL
NEUROPHARMACOLOGY
Volume 30, Number 1
January /February 2007
34
Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

stimuli are only visual (or only auditory). In
contrast, Duncan et al
31
showed that when a
pair of targets was presented in 2 different
modalities, one auditory and the other visual
(or vice-versa), no such interference effects
were found. Similar results for audiovisual
pairings have been found
2
.
32,33
This finding
led Duncan et al
31
to propose that ‘‘visual
attention to one simple target does not
restrict concurrent auditory attention to
another,’’ and to conclude that there was
no restricted attentional capacity between
sensory modalities.
The present results argue otherwise.
Concurrent visual stimulation, despite being
completely irrelevant to the task at hand,
caused a substantial decrement in subjects’
abilities to perform the auditory discrimina-
tion task, suggesting that supramodal (supra-
sensory) attentional systems do in fact have a
limited capacity. In this respect, our results
are reminiscent of early work by Colavita,
35
Colavita and Weisberg,
36
and Sinnett et al.
37
The question becomes why such a
strong interference effect is seen in these
data but not in the AB paradigms of Duncan
et al.
31
One possible explanation is that a
degree of automatic multisensory integration
of the simultaneous auditory and visual
constituents of our multisensory targets
occurred. Indeed, Duncan et al
31
explicitly
pointed out that their auditory and visual
stimulation streams were clearly separable,
and they were careful to specify that other
circumstances might apply for the case
where the auditory and visual events ema-
nated from the same object (ie, when multi-
sensory integration might occur). Our
auditory and visual stimuli did occur in
temporal register, although it should be
emphasized that they were presented in
different spatial locations (headphones vs
computer screen), and that they had no
natural relationship with each other. Fur-
thermore, subjects were required to explic-
itly treat them as separate stimulus streams
to be attended differentially, and this separa-
bility would have been reinforced by the fact
that only 40% of the S2 stimuli were actually
bisensory. Nonetheless, some degree of
automatic multisensory integration of the
auditory and visual components of S2 may
well have occurred, and there is precedence
in the literature for integration of even such
basic stimuli.
38Y41
Another obvious question that arises is
why the interference effect is found to be
unidirectional, only affecting performance
on the auditory task when targets were
bisensory, with visual performance remain-
ing unaffected. A striking phenomenon in
the study of perceptual models of multi-
sensory stimuli is the tendency of visual
information to dominate other competing
sensory stimuli. Jordan
42
and Posner et al
43
have argued that visual dominance phenom-
ena arise from brain mechanisms designed to
compensate for the ‘‘low alerting capability
of visual signals.’’ The premise is that
because of this intrinsically lower state of
alertness in the visual system, the attentional
system is biased toward vision when there is
a probability that reliable information will be
provided in this modality. As a consequence,
this visual biasing would result in the with-
drawal of processing resources from other
senses. It should be pointed out though that
Posner et al
43
proposed that this general
attentional biasing would only occur when
the subjects are likely to receive reliable
inputs from the visual modality. To test this,
Klein (unpublished data, 1974) presented
subjects with random trials of visual, soma-
tosensory, or multisensory somatovisual
stimuli. He found that subjects responded
fastest to the multisensory stimuli and slow-
est to the visual-alone stimuli. These results led
Posner et al
43
to propose that in a conflict
situation (ie, when vision, proprioception,
and/or audition provide discrepant informa-
tion), visual inputs will tend to dominate the
brain’s attentional system, unless the subject
is aware that responses based on visual
2
Note that an AB was in fact found by Soto-Faraco et al
34
between the visual and somatosensory modalities,
suggesting that capacity limitations may vary across
sensory combinations.
Theanine and Intersensory Attention CLINICAL
NEUROPHARMACOLOGY
Volume 30, Number 1
January /February 2007
35
Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

information will be disadvantageous.
43
Both
Jordan
42
and Posner et al
43
predicted that
visual dominance would no longer prevail
under circumstances where vision does not
provide adequate information.
The present results simply do not align
with this view. A strong visual dominance
effect persists, even when the visual stimuli
are known by subjects to be immaterial to
the task at hand, suggesting that endogenous
attentional processes cannot fully supersede
exogenous attentional capture by visual
inputs. Our data suggest that there is indeed
an attentional capacity limit for streams of
information occurring across 2 sensory
modalities. The contention that there is no
attentional capacity limit between sensory
modalities is also not supported by data from
patients with spatial neglect syndrome. For
example, cross-sensory extinction between
vision and touch in neglect patient has been
shown,
44
again suggesting a central supra-
modal attentional bottleneck.
45,46
Similarly,
extinction between audition and touch has
been described.
47
Theanine Slows RTs to
Auditory Stimuli
Our results also showed a marked
slowing of RTs for all auditory conditions
when subjects had ingested theanine (by as
much as 60 milliseconds), although there
was no change in performance accuracy.
This fairly dramatic slowing of responses was
unique to the auditory modality, as RTs to all
visual stimuli remained unaffected by thea-
nine. One possibility is that theanine may
cause subjects to be more deliberate when
confronted with such stimuli, although why
this would only occur for the auditory
stimuli is not clear, and if subjects have
become more deliberate, they certainly have
not benefited in terms of accuracy from such
a change. Whatever the case, and this needs
more study, theanine appears to differen-
tially affect the auditory and visual sensory
modalities, causing a measurable slowing of
performance in the auditory domain. One
possibility is that theanine may increase
susceptibility to visual dominance as de-
scribed previously.
CONCLUSIONS
High-density electrical mapping showed
that theanine ingestion had robust effects on
alpha oscillatory activity in the cortex of
healthy adult subjects while they performed
a highly demanding intersensory attention
task. We found a robust decrease in back-
groundalphaactivitybutaconcomitant
increase in attention-related alpha effects;
that is, during the late sustained period of
an attentional deployment phase, while sub-
jects prepared to attend to either the visual or
auditory modality for an impending multi-
sensory target, differential alpha deployments
were observed with greater amplitude than
was seen during a placebo control condi-
tion. Topographic mapping confirmed that
these alpha processes were generated in the
posterior parietal cortices, especially over
the right hemisphere. Thus, theanine, as well
as affecting more general brain rhythms
associated with cortical arousal, was found
to have more specific effects on known
nodes of the visual attention circuit. We
conclude that theanine has clear psycho-
active properties, and that it represents a
potentially interesting, naturally occurring
compound for further study as it relates to
the brain’s attentional system.
ACKNOWLEDGMENT
The authors thank Dr. Simon Kelly for
his detailed reading of an earlier text and
many useful editorial comments.
REFERENCES
1. Juneja LR, Chu DC, Okubi T, et al. L-theanineVa
unique amino acid of green tea and its relaxation
effect in humans. Trends Food Sci Technol 1999;10:
199Y204.
2. Pfurtscheller G. Event-related synchronization
(ERS): an electrophysiological correlate of cortical
areas at rest. Electroencephalogr Clin Neurophysiol
1992;83:62Y69.
3. Klimesch W, Doppelmayr M, Russegger H, et al.
Induced alpha band power changes in the human
EEG and attention. Neurosci Lett 1998;244:73Y76.
Gomez-Ramirez et al
Ó2007 Lippincott Williams & Wilkins
CLINICAL
NEUROPHARMACOLOGY
Volume 30, Number 1
January /February 2007
36
Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

4. Vanni S, Revonsuo A, Hari R. Modulation of the
parieto-occipital alpha rhythm during object
detection. J Neurosci 1997;17:7141Y7147.
5. Foxe JJ, Simpson GV, Ahlfors SP. Parieto-occipital
~10 Hz activity reflects anticipatory state of visual
attention mechanisms. Neuroreport 1998;9:
3929Y3933.
6. Worden MS, Foxe JJ, Wang N, et al. Anticipatory
biasing of visuospatial attention indexed by
retinotopically specific alpha-band EEG increases
over occipital cortex. J Neurosci 2000;20:1Y6.
[RC63].
7. Fu KG, Foxe JJ, Murray MM, et al.
Attention-dependent suppression of distracter
visual input can be cross-modally cued as indexed
by anticipatory parieto-occipital alpha-band
oscillations. Cogn Brain Res 2001;12:145Y152.
8. Bastiaansen MC, Brunia CH. Anticipatory attention:
an event-related desynchronization approach. Int J
Psychophysiol 2001;43:91Y107.
9. Bastiaansen MC, Bocker KB, Brunia CH, et al.
Event-related desynchronization during anticipatory
attention for an upcoming stimulus: a comparative
EEG/MEG study. Clin Neurophysiol 2001;112:
393Y403.
10. Kelly SP, Lalor E, Reilly RB, et al. Visual spatial
attention tracking using high-density SSVEP data
for independent brain-computer communication.
IEEE Trans Neural Syst Rehabil Eng 2005;13:
172Y178.
11. Kelly SP, Lalor EC, Reilly RB, et al. Alpha-based
suppression mechanisms facilitate sustained
visuo-spatial attention during continuous bilateral
stimulation. J Neurophysiol 2006;95:3844Y3851.
12. Sauseng P, Klimesch W, Stadler W, et al. A shift of
visual spatial attention is selectively associated with
human EEG alpha activity. Eur J Neurosci 2005;22:
2917Y2926.
13. Thut G, Nietzel A, Brandt SA, et al. Alpha-band
electroencephalographic activity over occipital
cortex indexes visuospatial attention bias and
predicts visual target detection. Journal of Neurosci
2006;26:9494Y9502.
14. Yamagishi N, Goda N, Callan DE, et al. Attentional
shifts towards an expected visual target alter the
level of alpha-band oscillatory activity in the
human calcarine cortex. Cogn Brain Res 2005;25:
799Y809.
15. Rihs T, Michel C, Thut G. Spatial distribution of
alpha-band oscillatory EEG activity during covert
visuo-spatial orienting is retinotopically specific for
8 visual field positions. Brain Topogr 2005;18:24.
16. Posner MI, Walker JA, Friedrich FJ, et al. Effects of
parietal injury on covert orienting of attention.
J Neurosci 1984;4:1863Y1874.
17. Posner MI, Petersen SE. The attention system of the
human brain. Ann Rev Neurosci 1990;13:25Y42.
18. Foxe JJ, Simpson GV. Biasing the brain’s attentional
set: II. Effects of selective intersensory attentional
deployments on subsequent sensory processing.
Exp Brain Res 2005;166:393Y401.
19. Foxe JJ, Simpson GV, Ahlfors SP, et al. Biasing the
brain’s attentional set: I. Cue driven deployments of
intersensory selective attention. Exp Brain Res
2005;166:370Y392.
20. Oldfield RC. The assessment and analysis of
handedness: the Edinburgh inventory.
Neuropsychologia 1971;9:97Y113.
21. Wetherill GB, Levitt H. Sequential estimation of
points on a psychometric function. Br J Math Stat
Psychol 1965;18:1Y10.
22. Doppelmayr M, Klimesch W, Pachinger T, et al.
Individual differences in brain dynamics: important
implications for the calculation of event-related
band power. Biol Cybern 1998;79:49Y57.
23. Schmid RG, Tirsch WS, Scherb H. Correlation
between spectral EEG parameters and intelligence
test variables in school-age children. Clin
Neurophysiol 2002;113:1647Y1656.
24. Clarke AR, Barry RJ, McCarthy R, et al. Effects of
methylphenidate on EEG coherence in
attention-deficit/hyperactivity disorder. Int J
Psychophysiol 2005;58:4Y11.
25. Clarke AR, Barry RJ, McCarthy R, et al. Effects of
stimulant medications on the EEG of children with
attention deficit/hyperactivity disorder
predominantly inattentive type. Int J Psychophysiol
2003;47:129Y137.
26. Song DH, Shin DW, Jon DI, et al. Effects of
methylphenidate on quantitative EEG of boys with
attention-deficit hyperactivity disorder in
continuous performance test. Yonsei Med J 2005;
46:34Y41.
27. Smith ME, McEvoy LK, Gevins A. Neurophysiological
indices of strategy development and skill
acquisition. Cogn Brain Res 1999;7:389Y404.
28. Broadbent DE. Perception and Communication.
New York, NY: Oxford University Press, 1958.
29. Posner MI, Snyder CR, Davidson BJ. Attention and
the detection of signals. J Exp Psychol 1980;109:
160Y174.
30. Duncan J. Similarity between concurrent visual
discriminations: dimensions and objects. Percept
Psychophys 1993;54:425Y430.
31. Duncan J, Martens S, Ward R. Restricted attentional
capacity within but not between sensory modalities.
Nature 1997;19:808Y810.
32. Potter MC, Chun MM, Banks BS, et al. Two
attentional deficits in serial target search: the visual
attentional blink and an amodal task-switch deficit.
J Exp Psychol Learn Mem Cogn 1998;24:979Y992.
33. Soto-Faraco S, Spence C. Modality-specific auditory
and visual temporal processing deficits. Q J Exp
Psychol 2002;55:23Y40.
34. Soto-Faraco S, Spence C, Fairbank K, et al. A
crossmodal attentional blink between vision and
touch. Psychon Bull Rev 2002;9:731Y738.
35. Colavita FB. Human sensory dominance. Percept
Psychophys 1974;16:409Y412.
36. Colavita FB, Weisberg D. A further investigation of
visual dominance. Percept Psychophys 1979;16:
409Y412.
37. Sinnett S, Soto-Faraco S, and Spence C. The Colavita
effect revisited: sensory dominance with complex
stimuli. Proceedings of the 6th Annual Meeting of
the International Multisensory Research Forum;
2005; Rovereto, Italy; pp. 51.
38. Molholm S, Ritter W, Murray MM, et al. Multisensory
auditory-visual interactions during early sensory
Theanine and Intersensory Attention CLINICAL
NEUROPHARMACOLOGY
Volume 30, Number 1
January /February 2007
37
Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

processing in humans: a high-density electrical
mapping study. Cogn Brain Res 2002;14:121Y134.
39. Molholm S, Ritter W, Javitt DC, et al. Multisensory
visual-auditory object recognition in humans: a
high-density electrical mapping study. Cereb Cortex
2004;14:452Y465.
40. Foxe JJ, Schroeder CE. The case for feedforward
multisensory convergence during early cortical
processing. Neuroreport 2005;16:419Y423.
41. Schroeder CE, Foxe JJ. Multisensory contributions
to low-level, ‘‘Unisensory’’ processing. Curr Opin
Neurobiol 2005;15:454Y458.
42. Jordan TC. Characteristics of visual and
proprioceptive response time in the learning of a
motor skill. Q J Exp Psychol 1972;24:536Y543.
43. Posner MI, Nissen MJ, Klein RM. Visual dominance:
an information-processing account of its origins and
significance. Psychol Rev 1976;83:157Y171.
44. Mattingley JB, Driver J, Beschin N, et al. Attentional
competition between modalities: extinction
between touch and vision after right hemisphere
damage. Neuropsychologia 1997;356:867Y880.
45. Rapp B, Hendel SK. Principles of cross-modal
competition: evidence from deficits of attention.
Psychon Bull Rev 2003;10:210Y219.
46. Bueti D, Costantini M, Forster B, et al. Uni- and
cross-modal temporal modulation of tactile
extinction in right brain damaged patients.
Neuropsychologia 2004;42:1689Y1696.
47. Ladavas E, Pavani F, Farne A. Auditory peripersonal
space in humans: a case of auditory-tactile
extinction. Neurocase 2001;72:97Y103.
Gomez-Ramirez et al
Ó2007 Lippincott Williams & Wilkins
CLINICAL
NEUROPHARMACOLOGY
Volume 30, Number 1
January /February 2007
38
Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.