Content uploaded by Efthymios Angelakis
Author content
All content in this area was uploaded by Efthymios Angelakis on Apr 01, 2014
Content may be subject to copyright.
College of Arts and Sciences
Drexel E-Repository and Archive (iDEA)
http://idea.library.drexel.edu/
Drexel University Libraries
www.library.drexel.edu
The following item is made available as a courtesy to scholars by the author(s) and Drexel University Library and may
contain materials and content, including computer code and tags, artwork, text, graphics, images, and illustrations
(Material) which may be protected by copyright law. Unless otherwise noted, the Material is made available for non
profit and educational purposes, such as research, teaching and private study. For these limited purposes, you may
reproduce (print, download or make copies) the Material without prior permission. All copies must include any
copyright notice originally included with the Material. You must seek permission from the authors or copyright
owners for all uses that are not allowed by fair use and other provisions of the U.S. Copyright Law. The
responsibility for making an independent legal assessment and securing any necessary permission rests with persons
desiring to reproduce or use the Material.
Please direct questions to archives@drexel.edu
This article was downloaded by:[Kounios, John]
On: 23 January 2007
Access Details: [subscription number 769609799]
Publisher: Psychology Press
Informa Ltd Registered in England and Wales Registered Number: 1072954
Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
The Clinical Neuropsychologist
Publication details, including instructions for authors and subscription information:
http://www.informaworld.com/smpp/title~content=t713721659
EEG Neurofeedback: A Brief Overview and an Example
of Peak Alpha Frequency Training for Cognitive
Enhancement in the Elderly
First Published on: 25 October 2006
To link to this article: DOI: 10.1080/13854040600744839
URL: http://dx.doi.org/10.1080/13854040600744839
Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf
This article maybe used for research, teaching and private study purposes. Any substantial or systematic reproduction,
re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly
forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents will be
complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be
independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings,
demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or
arising out of the use of this material.
© Taylor and Francis 2007
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
EEG NEUROFEEDBACK: A BRIEF OVERVIEW AND AN
EXAMPLE OF PEAK ALPHA FREQUENCY TRAINING FOR
COGNITIVE ENHANCEMENT IN THE ELDERLY
Efthymios Angelakis
1
, Stamatina Stathopoulou
1
,
Jennifer L. Frymiare
2
, Deborah L. Green
1
, Joel F. Lubar
3
,
and John Kounios
1
1
Department of Psychology, Drexel University, Philadelphia, PA,
2
Department
of Psychology, University of Wisconsin-Madison, Madison, WI, and
3
Department of Psychology, University of Tennessee, Knoxville, TN, USA
Neurofeedback (NF) is an electroencephalographic (EEG) biofeedback technique for
training individuals to alter their brain activity via operant conditioning. Research has
shown that NF helps reduce symptoms of several neurological and psychiatric disorders,
with ongoing research currently investigating applications to other disorders and to the
enhancement of non-disordered cognition. The present article briefly reviews the fundamen-
tals and current status of NF therapy and research and illustrates the basic approach with
an interim report on a pilot study aimed at developing a new NF protocol for improving cog-
nitive function in the elderly. EEG peak alpha frequency (PAF) has been shown to correlate
positively with cognitive performance and to correlate negatively with age after childhood.
The present pilot study used a double-blind controlled design to investigate whether training
older individuals to increase PAF would result in improved cognitive performance. The
results suggested that PAF NF improved cognitive processing speed and executive function,
but that it had no clear effect on memory. In sum, the results suggest that the PAF NF
protocol is a promising technique for improving selected cognitive functions.
Keywords: Biofeedback; Cognitive aging; Cognitive enhancement; EEG; Neurofeedback; Neuroplasticity;
Peak alpha frequency
INTRODUCTION
Neurofeedback (NF, also called EEG biofeedback) is an electroencephalo-
graphic (EEG) operant-conditioning training technique that helps individuals learn
to control or change their brain activity. It is used to treat a variety of neurologi cal
and psychological conditions and to increase cognitive perfor mance in nonclinical
individuals. Routine conditions treated with NF include Attention Deficit Hyperac-
tivity Disorder (ADHD), anxiety, epilepsy, and addictive disorders. Traumatic brain
Address correspondence to: John Kounios, Department of Psychology, Drexel University, 245 N.
15th Street, MS 626, Philadelphia, PA 19102-1192, USA. E-mail: john.kounios@gmail.com
Accepted for publication: April 6, 2006. First published online October 25, 2006.
# 2006 Taylor & Francis Group, LLC, an informa business
The Clinical Neuropsychologist, 21: 110–129, 2007
http://www.psypress.com/tcn
ISSN: 1385-4046 print=1744-4144 online
DOI: 10.1080/13854040600744839
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
injury (TBI), learning disabilities, depression, and schizophrenia are current ly be ing
investigated as potential candidates (Monastra, 2003).
The present article consists of two parts. The first part briefly reviews the his-
tory and current state of NF in the treatment of psychological or neurological dis-
orders and discusses its future promise. This review is a selective overview of NF
research; for detailed or specialized reviews, see Lubar, 1991, 1997, 2003; Monastra,
2003; Moore, 2000; Nash, 2000; Sterman, 1996, 2000; Thatcher, 2000; Trudeau,
2000. The second part of this article presents an example in the form of a preliminary
report on work toward a new NF protocol intended to enhance cognitive function,
particularly in elderly participants.
HISTORICAL OVERVIEW
The electroencephalogram is produced by synchronous postsynaptic potentials
from thousands to millions of neurons, and is usually recorded at the scalp, althoug h
intracranial EEG is sometimes recorded. When amplified, digitized, and plotted, the
raw EEG signal appears as a complex oscillatory pattern. This complex signal can be
filtered to isolate narrow frequency bands (defined in Hz) that reflect specif ic brain
sources and functions (Duffy, Iyer, & Surwillo, 1989).
NF was discovered and developed concurrently by independent researchers for
the treatment of different pathological conditions. In the late 1960s, Sterman and his
colleagues accidentally discovered that cats trained to produce 12–15 Hz activity
over their rolandic cortex were more resistant to substance-induced epileptic seizures
than non- trained cats, and in the early 1970s these researchers demonstrated this
phenomenon in humans as well (for a review, see Sterman, 2000). This training
has been termed SMR NF, for sensorimotor rhythm, since the 12–15 Hz EEG activit y
trained is recorded over, and is typical of, primary sensorimotor cortex. In the 1970s,
Twemlow and Bowen (1976) first reported on the potential of NF to increase alpha
(8–13 Hz) and theta (4–7 Hz) magnitude as a treatment for alcoholism, though the
first controlled study to support these findings was published 13 years later (Peniston
& Kulkosky, 1989). Posterior a lpha activity is associated with relaxed consciousness;
alpha ‘‘blocking’’ (i.e., reduction) is associated with alertness and active processing
(Berger, 1929; Penfiel d & Jasper, 1954; Pilgreen, 1995). Central-posterior theta is
associated with drowsiness (Duffy et al., 1989, p. 111). In the late 1 970s, Hardt
and Kamiya (1978) reported on the efficacy of alpha neurofeedback to reduce or
increase anxiety, depending on the direction of the training (i.e., alpha magnitude
inversely proportional to anxiety). Around the same time, similar work by Garrett
and Silver (1976) found alpha enhancement to reduce test anxiety. Both alpha and
alpha=theta NF is administered at the vertex of the head (electrode CZ according
to the International 10= 20 System; Homan, 1998). Although these anxiolytic effects
were found in one study to be related to percei ved success at the training but inde-
pendent of direction of training (Plotkin & Rice, 1981), later research supported the
directional effects of alpha training in generalized anxiety, showing decreased versus
increased heart-rate reactivity to stress for alpha enhancers and supressors respect-
ively (Rice, Blanchard, & Purcell, 1993).
In the mid 1970s, based on the observation that 12–14 Hz (SMR) enhancement
with 4–7 Hz inhibition (often seizures are 4–7 Hz) NF over rolandic cortex reduced
EEG NEUROFEEDBACK 111
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
convulsions and related motor activity in epileptics, (Lubar & Shouse, 1976; Shouse
& Lubar, 1979) successfully applied this treatment to reduce hyperact ivity in a
hyperkinetic child, which showed a reversal effect when SMR was suppress ed in
an ABA crossover design. Since then, subsequent studies have shown the utility
and success of theta=SMR NF for the treatment of ADHD, and have added beta
(16–20 Hz) enhancement for improvement of attention (Linden, Habib, & Radijevic,
1996; Lubar, 1991; Lubar & Lubar, 1984; Lubar, Swartwood, Swartwood, &
O’Donnell, 1995). The rationale for this treatment is based on studies that showed
abnormally lower beta and greater theta activity in children with ADHD as
compared to controls (Bresnahan, Anderson, & Barry, 1999; Chabot, di Michele,
Prichep, & John, 2001; Chabot & Serfontein, 1996; Mann, Lubar, Zimmerman,
Miller, & Muenchen, 1992; Monastra, Lubar, & Linden, 2001).
A number of studies have shown the potential of NF to improve cognitive per-
formance in healthy individuals as well. Beatty, Greenberg, Deibler, and O’Hanlon
(1974) trained healthy young adults to either augment or suppress occipital theta
(3–7 Hz) activity based on findings that occipital theta amplitude is negatively corre-
lated with vigilance. After only two 1-hour NF sessions, they found improved and
worsened performance, respectively, in a monitoring task. Rasey, Lubar, McIntyre,
Zoffuto, and Abbott (1996) trained four college students to increase beta (16–22 Hz)
while suppressing theta-alpha (6–10 Hz) amplitude at the vertex (CZ). After 20 ses-
sions, two of the participants learned to control their EEG according to the require-
ments, whereas the other two failed. Learners, but not non-learners, showed
improvement in attention as measured by a continuous performance Go=No-Go test.
Recently, several studies have focused on the specificity of the EEG frequency
to be trained and its effects on cognition. Vernon and colleagues (2003) trained 32
medical students to either augment SMR (12–15 Hz) activity or augment theta
(4–7 Hz) activity, while suppressing neighboring frequencies. Only the SMR group
managed to change their EEG and improve semantic working memory and focused
attention, whereas the theta group neither changed their EEG nor improved cogni-
tive performance. In a similar recent study, researchers from the same group (Egner
& Gruzelier, 2004) found frequency-specific cognitive effects of NF. They showed
that training college students to increase SMR amplitude improved their perceptual
sensitivity (d
0
) and reduced omission errors and reaction time variability, whereas
training others to increase beta (15–18 Hz) amplitude increased their reaction time
speed and the amplitude of the P300 event-related potential (ERP). Both of the latter
studies included control waiting-list groups that showed no effe cts. Finally, slow cor-
tical potential (SCP) (<1 Hz) NF over the left hemisphere has been shown to speed
up or slow down lexical decisions when individuals are trained to produce negative
and positive SCP shifts, respectively (Pulvermueller, Mohr, Schleicert, & Veit, 2000).
As in Rasey et al. (1996) and Vernon et al. (2003), individuals who failed to change
their SCP showed no cognitive changes.
NF sessions typically last for 1 hour or less, including participant preparation
time plus 20–40 minutes of NF, and are usually administered twice per week. The
number of sessions needed for treatment varies substantially from individual to indi-
vidual, depending, among other things, on the condition being treated, the indivi-
dual’s learning success, and the severity of the condition. Sterman (2000) reports
using 25 sessions to treat epileptic seizures. Lubar (1991) suggests that 40–80 sessions
112 EFTHYMIOS ANGELAKIS ET AL.
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
are needed to treat ADHD. In contrast, improvement in patients with anxiety
disorders has been reported with only eight NF sessions (see Moore, 2000, for a
review). However, in our experience with epilepsy and ADHD, some clients may
need many more sessions, for at least a year or more, to show learning and signifi-
cant improvement. Moreover, since NF is an operant conditioning technique, it may
well follow the laws of schedules of reinforcement in its patterns of acquisition and
extinction. Therefore, occasional post-treatment ‘‘boosting’’ sessions are recom-
mended (Lubar, 2003).
The rationale for the development of NF protocols has been based on EEG
and neuroimaging research on correlates of brain pathology (e.g., ADHD,
depression, TBI), accidental discovery (e.g., epilepsy); or neurophysiological corre-
lates of cognitive states (e.g., anxiety, substance abuse). Sometimes more than one
NF protocol is found to be effective for the same syndrome. For example, in
addition to SMR training, SCP-NF is found to reduce seizure activity in epileptic
patients (Kotchoubey, Strehl, Holzapfel, Schneider, Blankenhorn, & Birbaumer,
1999). Some propose the comparison of clients’ EEG to normative EEG databases
in order to individualize NF according to each client’s abnormalities (Thatcher,
1999). This is partly based on the rationale that identical symptoms may be due to
different underlying pathologies. A good example is that attentional problems
may be due to a variety of conditions, including ADHD and depression. Even single
diagnoses such as ADHD have been shown to have subtypes of quite distinct EEG
patterns (Chabot et al., 2001). However, EEG abnormality does not necessarily sig-
nify pathology, just as normal EEG does not guarantee healthy brain function.
Therefore, caution and experience must be applied in such decisions, just as with
any other medical or psychological treatment. A combination of standardized pro-
tocol, EEG normative database comparison, experience, and e xpert consultation will
maximize the probability of treatment success.
Clinical efficacy of NF varies across studies. Treatment protocols for epilepsy
and ADHD seem to have the strongest empirical support, followed by those
developed to treat anxiety and substance abuse. Treatments for depression, schizo-
phrenia, traumatic brain injury, learning disabilities, Tourette’s and chronic fatigue
syndromes, and autism are all under investigation (Jarusiewicz, 2002; Monastra,
2003). Long-term effects of NF have been reported after 6 months and up to 10 years
post-treatment (Lubar, 2003; Monastra, 2003; Trudeau, 2000). The existing research
therefore suggests that NF is a promising technique for treating a number of disor-
ders. And though most of the important studies establishing the effectiveness of NF
for treatment have not combined it with other forms of therapy, NF should not be
considered a panacea or sole treatment for all sympt oms of the conditions treated.
Other concurrent forms of care are recommended, including psychotherapy, family
therapy, group support, and medication, whenever needed (Lubar, 2003). Further-
more, there is a need for more research to investigate the specificity of NF versus
placebo, as well as the specificity of EEG frequency and scalp location for obtaining
and maximizing therapeutic results.
Success of NF can be assessed with various objective and subjective measures,
including standar dized tests and inventories, self-reports, and reports from family,
educators, or employers. Post-NF EEG ‘‘normalization’’ is another measure to
assess learning success and validate NF as a specific treatment. However, although
EEG NEUROFEEDBACK 113
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
post-NF EEG changes often correspond to the training protocol, some times they do
not, even if clinical improvement is achieved (Sterman, 2000). It is important to
remember that the adjustment of brain function is not always easy or simple with
NF, just as it is not always easy or simple with medication. For example, in an
NF experiment involving ch ildren with learning disabilities, Fernandez and collea-
gues (2003) found EEG changes not only in the frequencies trained (4–13 Hz) but
also in slower and faster frequencies (ranging from 3–19 Hz) and at electrode sites
other than the ones trained, suggesting a large-scale reorganization of EEG activity.
Lubar (2003) reports more than 1200 organizations worldwide (including
private clinics and research centers=laboratories) utilizing NF for treating ADHD
alone, based on information from NF equipment manufacturers. There are several
professional organizations that promote education and research on NF, includi ng
the International Society for Neuronal Regulati on and its chapters worldwide (www.
isnr.org), the Association for Applied Psychophysiology and Biofeedback (www.
aapb.org), the Electroencephalography and Clinical Neuroscience Society
(www.ecns.com), and the Biofeedback Society of California (www.biofeedbackcali-
fornia.org) based in the United States. In addition, the Biofeedback Certificat ion
Institute of America (www.bcia.org) tests and certifies clinicians who provide NF
services to the public.
EEG Neurofeedback certification from the Biofeedback Certification Institute
of America (BCIA) is considered the standard in the field by the Association for
Applied Psychophysiology and Biofeedback. Certification is available to profes-
sionals that meet the degree and licensing requirements in BCIA approved health
care fields. Additional requirements include a 36-hour didactic EEG biofeedback
education, covering areas from neuroanatomy to treatment planning and pro-
fessional conduct, 25 contact hours with a BCIA-approved mentor to review 10 per-
sonal and 100 patient sessions, and a 3-hour written examination. Recertification is
necessary at 4-year intervals to ensure that providers continue to uphold BCIA ethi-
cal practices and monitor new developments in the field.
PAF NF: A PROPOSED PROTOCOL FOR COGNITIVE ENHANCEMENT
One of the most prominent EEG phenomena is the alpha rhythm, an oscil-
lation in the range of 8–13 Hz with an average peak of 10–11 Hz in healthy adults.
Peak alpha frequency (PAF) corresponds to the discrete frequency with the highest
magnitude within the alpha range, and is known to be slower in children and the eld-
erly, although it also varies across individuals (Klimesch, 1997; Posthuma, Neale,
Boomsma, & de Geus, 2001). While the amplitude of alpha rhythm oscillations is
associated with relaxation, anxiety, and internal focus (discussed above), the fre-
quency of this oscillation has been positively correlated with mental performance
at all ages, both in healthy individuals and in individuals with neurological con-
ditions (for reviews, see Angelakis, Lubar, Stathopoulou, & Kounios, 2004b;
Klimesch, 1997). Among other correlates, PAF is invers ely correlated with age after
the age of 20 (Kopruner, Pfurtscheller, & Auer, 1984), and is lower in individuals
with Alzheimer’s disease compared to healthy matched controls (Klimesch, 1997).
Although previous studies have shown the effectiveness of neurofeedback in
the treatment of various neurological and psychiatric disorders such as epilepsy
114 EFTHYMIOS ANGELAKIS ET AL.
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
(Sterman, 2000) and ADHD (L ubar & Lubar, 1999), by altering EEG amplitude in
the alpha, beta, or theta frequencies, to our knowledge no published study to date
has investiga ted the effects of training individuals to change the peak frequency of
their alpha rhythm. The present pilot study investigated whether training elderly
individuals to restore (i.e., increase) their PAF to the levels of younger people would
result in improved cognitive performance. As a control for this experimental treat -
ment, two other neurofeedback conditions were employed, one to train an increase
in alpha amplitude, and one pseudo-neurofeedback placebo condition in which the
participant was told she was having real neurofeedback but was actually viewing
playback of another participant’s session. It was expected that, out of the three treat-
ments, PAF NF would result in the largest improvements in cognitive performance,
whereas alpha amplitude NF and pseudo-NF would result in either no improvement
or small improvement s in cognitive performance due to placebo effects. In total, six
healthy senior citizens have participated to date. Therefore, the present study serves
as a pilot for further research, and preliminary results were interpreted in term s of
common patterns within subgroups.
METHOD
Participants
Four healthy elderly women and two healthy elderly men were recruited from the
greater Philadelphia area. Participants were pseudo-randomly assigned to the two
groups, one experimental (EF1, EM, EF2) and one control (CF1, CM, CF2), matched
for sex and education (E: experimental, C: control, F: female, C: male; e.g., EF1 ¼ first
experimental female). The age of the participants ranged from 70 to 78 years (EF1: 74,
EM: 73, EF2: 75, CF1: 70, CM: 78, CF2: 74). Participants were paid a base wage of
$10=hour, plus an extra $2.5=hour for each training session in which they showed
improvement. In addition, participants received a bonus of $2.5 for each hour of
participation upon completion of all training sessions (40 maximum) and final assess-
ment. Informed written consent was obtained from all participants. The study was
approved by the Internal Review Board of the University of Pennsylvania.
Equipment and Instrumentation
For pre- and post-training EEG measurement, a 128-channel electrode cap
designed according to the extended International 10–20 System of electrode place-
ment was used in conjunction with MANSCAN RECORDER, an amplifier and
recording software system from SAM Technology (http:==www.manscaneeg.
com=DataAcquisition
page.htm). EEG was recorded at 250 samples per second
(frequency bandpass: 0.05–100 Hz), referenced to digitally linked mastoids.
Neurofeedback administration requires a similar amplifier and recording soft-
ware system to amplify and digitize the signal from the electrodes and to specify the
protocol of neurofeedback training, respectively. Protocols can be designed utilizing
the various training screens and scripts, which provide visual feedback to parti-
cipants and designate the training parameters. Although neurofeedback is typically
administered at one electrode site, many biofee dback hardware=software packages
provide the versatility to monitor psychophysiological activity with as many as 40
EEG NEUROFEEDBACK 115
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
electrodes in several modalities. In this study, neurofeedback was administered at
parietal-occipital midline electrode POz with a BioGraph=ProComp þ biofee dback
system (Thought Technology Ltd.).
Assessment Instruments and Tasks
Neuropsychological and personality assessment. A short questionnaire
was used to assess neurological and psychological state and current self-rating of
cognitive function. Participants were administered the State-Trait Personality Inven-
tory (STPI) (Spielberger et al., 1979), and a number of standard psychometric tests,
including the Digit Span (Wechsler, 1995), the Word List Memory Task (Welsh,
Butters, Hughes, Mohs, & Heyman, 1991), the Stroop test (Str oop, 1935), the
Passage Comprehension test (Woodcock & Johnson, 1977), Raven’s Standard Prog-
ressive Matrices (Raven, 1960), and the Logical Memory, Faces, Verbal Paired
Associates, Family Pictures, and Visual Reproduction subtests of the Wechsler
Memory Scale-III (Wechsler, 1997).
Cognitive tasks during pre- and post-training EEG. Two tasks were
employed during pre- and post-NF EEG assessment, the ‘‘n-back’’ task and a
‘‘Go=No-Go’’ oddball task. The n-back task had three separate blocks of 5 minutes
each and 1-minute practice before each block. Single letters were presented on a
computer monitor for 250 milliseconds at a rate of 2.5 seconds and were preceded
by a warning tone by 500 ms. During the first block (n 1), participants were asked
to respond (by pressing a mouse button with their right index finger) whenever the
current letter matched the one before it. During the second block (n 2), they were
asked to respond when the current letter matched the one two positions before it; the
third block (n 3) required matching with the letter three positions before. Match-
ing targets were pseudo-randomly present ed 14% of the time.
The Go=No-Go oddball task had three separate identical blocks of 5 minutes
each and 1 minute of practice before the first block. It presented three different tones
through loudspeakers at a random rate of 1.0–1.3 seconds. A high-pitched tone (target)
was presented 20% of the time, a lower-pitched (standard) tone was presented 65% of
the time, and a complex, novel sound was presented 15% of the time. Participants were
required to respond with a mouse button-press only to the high-pitched target tones.
Procedure
Neuropsychological and personality assessment. EEG, personality, and
cognitive assessment were performed twice: before and after the completion of NF
training. Personality and cognitive evaluation was admini stered by two independent
psychometricians, one for all of the pre-training and the other for all of the post-
training evaluations. Both administrators were blind to the participants’ train ing
condition. After a brief orientation, participants were administered the short
questionnaire, the STAI, and the psychometric tests. Personality and cognitive
assessment lasted for approximately 1.5 hours and was split into two sessions on
different days before NF, whereas it was administered in one session with an inter-
mediate break after NF.
116 EFTHYMIOS ANGELAKIS ET AL.
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
EEG acquisition. Participants were fitted with the electrode cap and impe-
dances at all channels were reduced to below 20,000 Ohms. EEG assessment involved
six recordings in the following order: a 3-minute eyes-closed resting baseline (ECB)
block, a 3-minute eyes-open rest ing baseline (EOB) block; three 5-minute n-back
task blocks (1back, 2back, 3back); three 5-minute Go=No-Go oddball-task blocks
(OB1, OB2, OB3); a 3-minute eyes-open post-task rest (PTRO) block; and a 3-
minute eyes-closed post-task rest (PTRC) block.
Neurofeedback training. NF training was administered once or twice per
week for a total of 31–36 sessions. Among the experimental participants, EF1 pa rti-
cipated in 35 sessions, EM in 31, and EF2 in 36. Among the controls, CF1 partici-
pated in 35 sessions, CM in 36, and CF2 in 32. Experimental participants were
trained to increase their PAF, whereas two controls were trained to increase their
alpha amplitude. In both cases (frequency and amplitude measures), alpha was
defined as 8–13 Hz. The third control (CF2) was shown playback of one of the other
experimental participant’s (EF1) sessions. None of the participants knew which con-
dition they were in, nor did they know there were experimental and control con-
ditions. They were all told that different training protocols were tested to see
which one, if any, might provide cognitive improvement. After training, all parti-
cipants reported the belief that they had be en in an experimental group, except for
the pseudoneurofeedback participant CF2, who reported finding the feedback con-
fusing and ineffective.
Sessions lasted approximately 1 hour and included 10 minutes of preparation,
2 minutes of EOB EEG recording, 8 minutes of NF, a 3-minute break watching a
documentary video (or reading after the tenth session), 8 minutes of NF, a 3-minute
break as before, an d 8 final minutes of NF. A reference electrode was attached to,
and the training electrode was attached to a posterior midline site (POz). The NF
system provided auditory and visual feedback for increasing the EEG training mea-
sure (PAF or alpha amplitude) while kee ping the other (alpha amplitude or PAF)
from rising above baseline levels. Participants were instructed to try to note and uti-
lize the particular strategy that yielded the desired effect, and were verbally encour-
aged when they performed well. All participants received approximately equal
amounts of encouragement by the experimenter. At the end of each session, parti-
cipants were asked to describe their subjective experience.
Data Analysis
EEG data were analyzed by a ‘‘non-blind’’ experimenter. Fast Fourier Trans-
form (FFT) was computed using a bandpass pre-filter of 5–15 Hz. Eye -blink artifacts
were eliminated using an adaptive filter constructed with EMSE 5.0 (www.source-
signal.com) separately for each participant, and additional artifacts were excised
by visual inspection. For resting conditions, all 128 channels were analyzed, but
for task conditions only limited midline channels were analyze d due to excessive arti-
facts at lateral and frontal sites. At the time this article was prepared, only the EOB,
PTRO, and 1-back conditions had been analyzed. All results discussed are descrip-
tions of the data pattern and are not statistically derived conclusions, due to the
small number of participants.
EEG NEUROFEEDBACK 117
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
RESULTS
NF Performance
All three experimental participant s learned to increase their PAF, as shown by
the trend lines (smoothed with a 6-point running-mean filter) in Figure 1, and both
alpha controls learned to increase their alpha amplitude, as shown in Figure 2. Both
figures show normalized percent-time above training threshold.
Self-Reports
During experimental training (PAF NF), two participants (EF1, EM) showed
higher PAF when they concentrated, relaxed, ‘‘thought everything is going well in life,’’
had pleasant thoughts, thought of food, traveling, funny stories, or ‘‘crazy things their
children had done,’’ said the alphabet backwards to themselves, or thought of foods
that start with the letter Z. A third experimental participant (EF2) increased both alpha
amplitude and PAF when she thought of people who cared for her, had pleasant
thoughts, emptied her mind, relaxed, sang simple songs to herself, imagined calling
friends to help her, or thought about ‘‘an appointment today.’’ These experimental
participants had lower PAF when they thought of problems, were frustrated, had
not participated for some time, or were drowsy.
During control trai ning (alpha amplitude), participants (CF1, CM) showed
more alpha amplitude when they reported that they were relaxed or sleepy, thought
Figure 1 Changes in peak alpha frequency across sessions for three PAF-NF participants. Normalized
trendlines (smoothed with a 6-point running mean) show changes in estimated PAF (computed using per-
cent time above training threshold).
118 EFTHYMIOS ANGELAKIS ET AL.
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
of what had to be done for the week, Christmas presents, cooking, upcoming vaca-
tions, actors, singers, jazz music, or their adolescence, concentrated on the ‘‘smiley
face’’ feedback screen, thought of nothing particular, or had a blank mind. These
participants showed less alpha amplitude when they reported that they had worries,
problems, or unpleasant memories, thought about war or injustice, thought about
his wife at the hospital, tried hard, focused on something, let their mind wander,
or had not participated recently. Finally, the pseudo-NF control participant (CF2)
reported not being able to understand what thoughts yield positive feedback on
the screen.
All PAF participants rated themselves as thinking faster, averaging a 1-point
(19%) improvement (EF1: 1.0; EM: 1.5; EF2: 0.5) after NF training. Two control
participants rated themselves as thinking slightly faster (CF1: 0.5; CF2: 0.5), and
one as thinking more slowly (CM: 1.5), averaging a 0.16-point slowing (8%).
Moreover, no PAF participant reported a change in concentration level, whereas
two controls rated themselves lower in concentration; CF1 by 1.5 points (27%),
and CF2 by 1.0 point (14%).
STPI
After training, two PAF and one active control participants scored higher in
Trait Depression. EF1 scored higher by 7 points (70%), EM by 2 points (11%),
and CF1 by 9 points (90%). Moreover, two controls scored lower in Trait Curiosity
Figure 2 Changes in alpha magnitude across sessions for two alpha-magnitude NF participants. Normal-
ized trendlines (smoothed with a 6-point running mean) showing changes in estimated alpha magnitude
(computed using percent time above training threshold).
EEG NEUROFEEDBACK 119
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
(CF1 6, 21%; CF2 4, 12.5%), and the pseudo-NF control (KCF2) scored 5
points higher in Trait Anger (28%).
Memory
Verbal memory tests with consistent pre=post training patterns within groups
showed no change for experimental participants, and mixed results for controls. In
the word memory list, CF1 scored higher by 2 points (29%), CM scored higher by 4
points (80 % ), and CF2 scored higher by 2 points (25%), whereas in the recall stories
task (first part), CF1 scored lower by 8 points (75%), CM scored lower by 1 point
(8%), and CF2 scored lower by 4 points (31%).
Visual memory tests showed worsening for the experimental participants, an d
improvement for the active controls after training. Specifically, in the Visual Repro-
duction task, CF1 scored higher by 4 points (9.3%), and CM scored higher by 11
points (12.7%), whereas EF1 scored lower by 18 points (20%), EM scored lower
by 24 points (36%), and EF2 scored lower by 15 points (18%).
Stroop
After training, two out of the three PAF participants increased their response
speed and accuracy, whereas the two active controls decreased their speed. Of the
PAF participants, EF1 scored 17 seconds faster (12%) with 7 more hits (7 %), and
EM scored 25 seconds faster (16%) with 5 more hits (7%). From the active controls,
CF1 scored 12 seconds slower (8%), and CM 12 seconds slower (11%), whereas the
pseudo-NF control (CF2) did not change her speed. However, EF2 (a PAF partici-
pant) did not fit this pattern. In contrast to the other two PAF participants, she
decreased her speed by 49 seconds (43%), and her hits by 6 (8%).
N-back Task
There was no overall improvement in the total number of correct responses or
false alarms for either group. However, during the third, and most demanding block
(3-back), all PAF participants showed decreased performance (17% average), and
all controls showed increased performance (þ16.6% average). Across blocks, while
PAF participants increased their average reaction time (RT) by 14%, all PAF part-
icipants improved their RT stability (coefficient of variation: 53%). In contrast,
controls increased their average RT by 4% and demonstrated worsened RT stability
(coefficient of variation: þ18%).
Go/NoGo Task
After NF training, all PAF participants and one active control decreased their
RT, whereas the two other controls increased it. On average, PAF participants
decreased their RT by 6%, whereas controls (on average) kept it unchanged. How-
ever, PAF participants increased their errors by 186% whereas controls increased
them by only 50 %.
120 EFTHYMIOS ANGELAKIS ET AL.
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
Table 1 summarize s some general trends from the results on cognitive perform-
ance. In sum, after NF training, experimental (PAF) participants reported thinking
faster, whereas control s reported slight worsening in concentration. Psychometric
testing comparing pre- to post-NF performance suggested that PAF training wor-
sened performance in visual reproduction, improved speed and accuracy in the
Stroop test (only two of the three participants), worsened performance in the 3-back
task, improved reaction-time consistency in the n-back task, and increased errors in
the oddball task more than for the control participants. Control participants were
found to improve in word recall, worsen in story recall, improve in visual repro-
duction, become slower in the Stroop test, score better in the 3-back task, lose
reaction-time consistency in the n-back task, and increase their oddball-task errors,
although much less than the experimental participants.
EEG Results
Overall, across electrodes, two experimental and two control participants
increased their alpha magnitude after the completion of NF training, whereas one
control participant (CF1) reduced her alpha magnitude and one experimental par-
ticipant (EF1) did not change her alpha magnitude after NF training.
Figure 3 reveals an interesting pattern of PAF results. For both the EOB rest-
ing EEG (top row) and for EEG during the 1-back task (bottom row), PAF training
for the exp erimental group was associated with an increase in PAF at frontal sites,
but not at posterior sites, with little or no observable PAF changes for the control
group. An increase in PAF was not unexpected for the experimental group. How-
ever, it was not expected to occur at frontal sites rather than the posterior site used
for PAF training. These results suggest that the increase in PAF resulting from trai n-
ing was not occurring in posterior brain areas but was instead occurring in frontal
Table 1 Average change per group (experimental vs. control) in different cog-
nitive domains after neurofeedback training
Experimental Control
Speed of processing
Self-reported speed of thinking þ –
Speed in Stroop þ –
Reaction time in Go=No-Go þ
Reaction time in n-back –
Executive function
Self-reported concentration –
Performance in Go=No-Go – –
Performance in Stroop þ
Reaction time stability in n-back þ –
Memory
Performance in 3-back – þ
Visual Memory – þ
Memory for words þ
Story recall –
‘‘ þ ’’: improvement, ‘‘–’’: worsening.
EEG NEUROFEEDBACK 121
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
areas. The lone posterior electrode site used during training was apparently detecting
(via electrical volume conduction) an increase in frontal PAF. This suggests that pos-
terior PAF is not as malleable as frontal PAF.
DISCUSSION
Effectiveness of PAF NF
The present pilot study examined the effects of a new NF protocol on the cog-
nitive performance of healthy elderly participants. In contrast to existing NF proto-
cols that involve training ch anges in EEG amplitude, this new protocol focused on
changing peak frequency. To control for non-specific effects, two alternative con-
ditions were employed, one active NF and one pseudo-NF placebo.
Both experimental and active control participants showed progressive learning
in their NF protocols, as shown in Figures 1 and 2. Even though the tw o training
protocols targeted the same frequency band, the two protocols contrasted each
other, as each protocol inhibited reward for increases in the dimension of the other.
Specifically, the PAF NF protocol withheld reward if alpha magnitude increased,
and the alpha magnitude NF protocol withheld reward if PAF increased.
Figure 3 Peak alpha frequency (PAF) for the experimental and the control group, before and after neu-
rofeedback (NF) training, for frontal (Fz and FCz) and posterior (Pz and POz) midline sites. Top left:
eyes-open baseline, experimental group. Top right: eyes-open baseline, control group. Bottom left: 1-back
task, experimental group. Bottom right: 1-back task, control group.
122 EFTHYMIOS ANGELAKIS ET AL.
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
Interestingly, preliminary results suggest that the cognitive effects of the two
protocols also contrasted. The PAF protocol suggested general improvement in
speed of processing and executive function, and the alpha magnitude pro tocol
showed improvement in memory, whereas each protocol showed a general decrease
in the cognitive dimensions for which the other protocol showed improvement (see
Table 1). Moreover, the pseudo-NF protocol showed either similar cognitive changes
to those of the active control protocol (i.e., alpha magnitude), or no changes.
The results described above suggest not only that PAF NF is feasible, but also
that it has specific effects on cognitive performance. The latter point is supported by
both the contrasting effects of the active control NF protocol and the fewer effects of
the pseudo-NF condition. In other words, PAF NF showed apparent cognitive
improvement in areas that alpha amplitude NF did not, and vice versa, whereas both
of these NF protocols showed great er cognitive effects than did the pseudo-NF pla-
cebo condition. Although previous studies have repeatedly shown changes specific to
the employed protocol (e.g., Hardt & Kamiya, 1978; Lubar, 1991; Sterman, 2000),
the present study discriminated between frequency-specific EEG effects on cognition
by controlling for all possible non-specific factors of NF training, such as training
environment and procedure, electrode position, frequency band, verbal coaching,
number of sessions, and individual differences. To some extent, the present study
also controlled for experi menter bias, since cognitive performance before and after
NF training was assessed by ‘‘blind’’ experimenters who were not aware of each par-
ticipant’s NF condition. In a recent study, Egner and Gruzelier (2004) have similarly
showed the specificity of EEG NF frequency on behavioral changes by showing dif-
ferent cognitive effects for individuals trained to increase the amplitude of neighbor-
ing frequencies (12–15 Hz versus 15–18 Hz).
EEG Changes and their Relationship to Behavioral Changes
Overall, EEG changes after training did not fit all protocol-specific expecta-
tions. First, we were surprised to find post-NF alpha magnitude increases
independent of NF co ndition. Specifically, two experimental and two control parti-
cipants showed increased alpha magnitude during the post-training recording. The
likely explanation for this is that participants were anxious during the initial EEG
recording due to the unfamiliarity of the procedure, whereas they likely felt more
relaxed during the post-training recording. While participants were shown a short
video explaining the EEG procedure, future studies should consider including a more
elaborative pre-training session to familiarize participants with the EEG and NF
procedures and to minimize their anxiety.
Second, even though all training was applied posteriorly (electrode site POz),
the experimental group showed post-training PAF increases frontally (but not pos-
teriorly), whereas the c ontrol group showed little or no PAF changes at any electrode
site. Specifically, post-NF PAF topography shifted more anteriorly for the experi-
mental group only (see Figure 4). In previous studies, we (Angelakis et al., 2004b)
have shown that frontal PAF topography correlates with cognitive preparedness,
as well as strong correlations between PAF and response control (Angelakis, Lubar,
& Stathopoulou, 2004a). Furthermore, the apparent increase in frontal PAF is con-
sistent with improvement in measures of executive function. In sum, these results
EEG NEUROFEEDBACK 123
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
suggest that posterior PAF may not be as trainable as frontal PAF. Future studies
can address this possibility by using multiple electrode sites during training. In this
way, the training protocol can be modified to encourage posterior PAF increa ses
while discouraging frontal PAF increases. In addition, future studies can directly
encourage frontal (but not posterior) PAF increases.
The specific effects of PAF NF on processing speed support the idea that PAF
reflects speed of information access (Klimesch, 1997). However, this increased speed
did not appear to improve memory functions as would be predicted by the model sug-
gested by Klimesch (1997), so this idea must be approached with caution until other
studies measure the specific effects of PAF changes on different types of cognitive
speed. In contrast, the present study suggested that memory improvement occurred
in the alpha magnitude control group that was specifically trained to not increase
Figure 4 Topographic maps of resting EOB peak alpha frequency (PAF) before and after neurofeedback
(NF) training. Top panel: experimental (PAF) participants; bottom panel: control (alpha magnitude) parti-
cipants. Each map is a representation of a view of the top of the head, with the nose at the top of the map, and
the left and right ears on the left and right sides of the map, respectively. Orange and white areas show the
scalp regions exhibiting the greatest alpha EEG amplitude for that condition. Each map is individually scaled
to the maximum and minimum values in that map in order to highlight the topographic distribution of PAF.
See online article for colour version of this Figure.
124 EFTHYMIOS ANGELAKIS ET AL.
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
PAF. This resul t came as a surprise, since we expected alpha magnitude NF to show
lesser (if an y) cognitive improvement compared to PAF NF. Some studies (e.g.,
Garrett & Silver, 1976) have reported decreases in test anxiety after alpha amplitude
NF, which may have selectively affected memory functions in the present study.
As further research increases our understanding of subtle differences between
the relationships between variant EEG frequencies and mental functions, we may
be better able to address questions such as how theta=SMR NF and PAF NF might
overlap. The present study extended previous findings (Angelakis et al., 2004a) that
PAF is positively related to response control, by suggesting decreased reaction time
variability after PAF training. However, similar effects have been reported in train-
ing younger individuals to reduce their theta=SMR ratio (Egner & Gruzelier, 2004).
Moreover, response control is one of the symptoms of ADHD that respond to
theta=SMR training (Lubar & Lubar, 1984). Since the theta (4–8 Hz) and SMR
(12–15 Hz) frequency bands are immediately adjacent to the alpha band (8–13 Hz),
it is worth considering whether there is some overlap between these two protocols,
given that pushing PAF toward a higher frequency may actually increase upper
alpha and decrease lower alpha magnitude (Klimesch, 1997).
Previous research has shown that frequency-specifi c NF trai ning can be less
than specific in its effects on the underlying EEG (Fernandez et al., 2003). This is
not surprising, given that EEG frequencies are produced by different populations
of neurons within the same system, and relevant parts of this system may well inter-
act with and reorganize other parts, just as neurotransmitter-specific drugs may
affect levels and functions of other neurotransmitters, such as selective serotonin
reuptake inhibitors (SSRIs) that affect regula tion of norepinephine as well (Sulser,
1989). Future studies may address these questions by directly comparing related
NF protocols according to their effects on specific mental functions, and by examin-
ing potential overlap between such protocols. A logical goal, then, would be to com-
pare the efficacy of these protocols in terms of the number of sessi ons required to
produce a desired effect, as well as in terms of ancillary effects of each protocol.
Limitations of the Study
The present study has severa l limitations. First, due to the small sample size,
these findings can serve only as pilot data suggesting further research. Second, the
individuals who participated in the study were highly motivated and high-functioning
seniors. This has both benefits and drawbacks: Less motivated and lower-functioning
individuals may learn more slowly, but may have more room for improvement.
Finally, even though the exploratory nature of this study did not allow for coaching
specific to each participant (since conditions had to be identical for both groups), the
present results suggest that some form of PAF training of healthy elderly individuals
has promise. Furthermore, even with the limited sample size involved, the present
data support the idea that EEG biofeedback is protocol-specific, resulting in quite dif-
ferent effects depending on the frequency and modality of training.
CONCLUSIONS
In general, NF has been shown to be a promising technique for a variety of
psychological and neurological disorders, including epilepsy, anxiety, ADHD, and
EEG NEUROFEEDBACK 125
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
possibly TBI, depression, learning disabilities, and autistic disorde rs. Moreover, NF
has been shown to help healthy individuals improve their cognitive performance.
According to the results reported here, PAF NF in healthy elderly individuals
may improve cognitive processing speed and executive function, but have little or no
effect on memory. In contrast, alpha amplitude NF may improve verbal, visual, and
working memory, but worsen speed of processing and executive function.
The present study illustrated the process of developing a new NF protocol
from theoretical conceptualization to protocol design, to EEG and behavioral
results, and showed that NF is an EEG frequency-specific technique that can be con-
trolled for non-specific placebo effects. As our understanding of the relationship
between specific components of the EEG and mental function grows, new NF pro-
tocols may attempt to treat mental conditions not addressed previously. This will be
the product of combining clinical observation, new knowledge of the EEG, and
experimentation on NF.
ACKNOWLEDGMENTS
The authors would like to thank Benjamin Alterman for helping with neuro-
feedback protocol development, Chris Ochner for administering and scoring psycho-
metric tests, and Allen Osman for contributing lab resources and helpful insights.
This study was supported by grant DC-04818 to John Kounios from the National
Institute of Deafness and Other Communication Disorders.
REFERENCES
Angelakis, E., Lubar, J. F., & Stathopoulou, S. (2004a). Electroencephalographic peak alpha
frequency correlates of cognitive traits. Neuroscience Letters, 371, 60–63.
Angelakis, E., Lubar, J. F., Stathopoulou, S., & Kounios, J. (2004b). Peak alpha frequency:
An electroencephalographic measure of cognitive preparedness. Clinical Neurophysiology,
15, 887–897.
Beatty, J., Greenberg, A., Deibler, W. P., & O’Hanlon, J. F. (1974). Operant control of occipi-
tal theta rhythm affects performance in a radar monitoring task. Science , 183, 871–873.
Berger, H. (1929). Uber das Elektrenkephalogramm des Menschen. Archiv fu
¨
r Psychiatrie, 87,
527.
Bresnahan, S. M., Anderson, J. W., & Barry, R. J. (1999). Age-related changes in quantitative
EEG in attention-deficit=hyperactivity disorder. Biological Psychiatry, 46, 1690–1697.
Chabot, R. J., di Michele, F., Prichep, L., & John, E. R. (2001). The clinical role of computer-
ized EEG in the evaluation and treatment of learning and attention disorders in children
and adolescents. Journal of Neuropsychiatry and Clinical Neuroscience, 13, 1–16.
Chabot, R. J., & Serfontein, G. (1996). Quantitative electroencephalographic profiles of chil-
dren with attention deficit disorder. Biological Psychiatry, 40, 951–963.
Duffy, F. H., Iyer, V. G., & Surwillo, W. W. (1989). Clinical electroencephalography and topo-
graphic brain mapping: Technology and practice. New York: Springer-Verlag.
Egner, T., & Gruzelier, J. (2004). EEG Biofeedback of low beta band components: Frequency-
specific effects on variables of attention and event-related brain potentials. Clinical Neu-
rophysiology, 115, 131–139.
Fernandez, T., Harmony, T., diaz-Comas, L., Herrera, W., Aboytes, G., Fernandez-Bouzas, A.
et al. (2003). Neurofeedback effect on EEG current sources. Poster presented at the annual
meeting of the Cognitive Neuroscience Society, New York City, NY.
126 EFTHYMIOS ANGELAKIS ET AL.
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
Garrett, B. L. & Silver, M. P. (1976). The use of EMG and alpha biofeedback to relieve test
anxiety in college students. In I. Wickramasekera (Ed.), Biofeedback, behavior therapy,
and hypnosis. Chicago, IL: Nelson-Hall.
Hardt, J. V., & Kamiya, J. (1978). Anxiety change through electroencephalographic alpha
feedback seen only in high anxiety subjects. Science, 201, 79–81.
Homan, R. W. (1998). The 10–20 electrode system and cerebral location. American Journal of
EEG Technology, 28, 269–279.
Jarusiewicz, B. (2002). Efficacy of neurofeedback for children in the autistic spectrum: A pilot
study. Journal of Neurotherapy, 6, 39–50.
Klimesch, W. (1997). EEG-alpha rhythms and memory processes. International Journal of
Psychophysiology, 26, 319–340.
Kopruner, V., Pfurtscheller, G., & Auer, L. M. (1984). Quantitative EEG in normals and in
patients with cerebral ischemia. In G. Pfurtscheller, E. J. Jonkman, & F. Lopes da Silva
(Eds.), Brain ischemia: Quantitative EEG and imaging techniques (pp. 29–35). Amsterdam:
Elsevier.
Kotchoubey, B., Strehl, U., Holzapfel, S., Schneider, D., Blankenhorn, V., & Birbaumer, N.
(1999). Control of cortical excitability in epilepsy. In H. Stefan, F. Andermann, P. Chauvel,
& S. Shorvon (Eds.), Advances in neurology (Vol. 8). Philadelphia, PA: Lippincott Williams
& Wilkins.
Linden, M., Habib, T., & Radijevic, V. (1996). A controlled study of the effects of EEG bio-
feedback on cognition and behavior of children with attention deficit disorder and learn-
ing disabilities. Biofeedback and Self-Regulation, 21, 35–49.
Lubar, J. F. (1991). Discourse on the development of EEG diagnostics and biofeedback
for Attention-Deficit=Hyperactivity Disorders. Biofeedback and Self-Regulation, 16,
201–225.
Lubar, J. F. (1997). Neocortical dynamics: Implications for understanding the role of neuro-
feedback and related techniques for the enhancement of attention. Applied Psychophysiol-
ogy and Biofeedback, 22, 111–126.
Lubar, J. F. (2003). Neurofeedback for the management of attention-deficit disorders. In
M. S. Schwartz & F. Andrasik (Eds.), Biofeedback: A practitioner’s guide (pp. 409–437).
New York: The Guilford Press.
Lubar, J. F., & Lubar, J. O. (1984). Electroencephalographic biofeedback of SMR and beta
for treatment of attention deficit disorders in a clinical setting. Biofeedback and Self-
Regulation, 9, 1–23.
Lubar, J. F., & Lubar, J. O. (1999). Neurofeedback assessments and treatment for
attention deficit=hyperactivity disorders. In J. R. Evans & A. Abarbanel (Eds.), Introduc-
tion to quantitative EEG and neurofeedback (pp. 103–143). San Diego, CA: Academic
Press.
Lubar, J. F., & Shouse, M. N. (1976). EEG and behavioral changes in a hyperkinetic child
concurrent with training of the sensorimotor rhythm (SMR). Applied Psychophysiology
and Biofeedback, 1, 293–306.
Lubar, J. F., Swartwood, M. O., Swartwood, J. N., & O’Donnell, P. H. (1995). Evaluation of
the effectiveness of EEG neurofeedback training for ADHD in a clinical setting as
measured by changes in T.O.V.A. scores, behavioral ratings, and WISC-R performance.
Biofeedback and Self-Regulation, 20, 83–99.
Mann, C. A., Lubar, J. F., Zimmerman, A. W., Miller, C. A., & Muenchen, R. A. (1992).
Quantitative analysis of EEG in boys with attention-deficit-hyperactivity disorder: Con-
trolled study with clinical implications. Pediatric Neurology, 8, 30–36.
Monastra, V. J. (2003). Clinical applications of electroencephalographic biofeedback. In M. S.
Schwartz & F. Andrasik (Eds.), Biofeedback: A practitioner’s guide (pp. 438–463). New
York: The Guilford Press.
EEG NEUROFEEDBACK 127
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
Monastra, V. J., Lubar, J. F., & Linden, M. (2001). The development of a quantitative elec-
troencephalographic scanning process for attention deficit-hyperactivity disorder:
Reliability and validity studies. Neuropsychology, 15, 1336–1344.
Moore, N. C. (2000). A review of EEG biofeedback treatment for anxiety disorders. Clinical
Electroencephalography, 31, 1–6.
Nash, J. K. (2000). Treatment of attention deficit hyperactivity disorder with neurotherapy.
Clinical Electroencephalography, 31, 30–37.
Penfield, W., & Jasper, H. (1954). Epilepsy and the functional anatomy of the human brain.
Boston, MA: Little Brown.
Peniston, E. G., & Kulkosku, P. J. (1989). Alpha-theta brainwave training and beta endorphin
levels in alcoholics. Alcoholism: Clinical and Experimental Research, 13, 271–279.
Pilgreen, K. L. (1995). Physiologic, medical, and cognitive correlates of electroencephalo-
graphy. In P. L. Nunez (Ed.), Neocortical dynamics and EEG rhythm (pp. 195–248).
New York: Oxford University Press.
Plotkin, W. B., & Rice, K. M. (1981). Biofeedback as a placebo: Anxiety reduction facilitated
by training in either suppression or enhancement of alpha brainwaves. Journal of Consult-
ing and Clinical Psychology, 49, 590–596.
Posthuma, D., Neale, M. C., Boomsma, D. I., & de Geus, E. J. C. (2001). Are smarter brains
running faster? Heritability of alpha peak frequency, IQ, and their interrelation. Beha-
vioral Genetics, 31, 567–579.
Pulvermueller, F., Mohr, B., Scheichert, H., & Veit, R. (2000). Operant conditioning of
left-hemisphere slow cortical potentials and its effect on word processing. Biological
Psychology, 53, 177–215.
Rasey, H. W., Lubar, J. E., McIntyre, A., Zoffuto, A. C., & Abbott, P. L. (1996). EEG
biofeedback for the enhancement of attentional processing in normal college students.
Journal of Neurotherapy, 1, 15–21.
Raven, J. C. (1960). Guide to the Standard Progressive Matrices. London: H. K. Lewis & Co.
Rice, K. M., Blanchard, E. B., & Purcell, M. (1993). Biofeedback treatments of generalized
anxiety disorder: Preliminary results. Biofeedback and Self-Regulation, 18, 93–105.
Shouse, M. N., & Lubar, J. F. (1979). Operant conditioning of EEG rhythms and Ritalin in
the treatment of hyperkinesis. Applied Psychophysiology and Biofeedback, 4, 299–312.
Spielberger, C. D., Barker, L. R., Russell, S. F., Crane, R. S., Westberry, L. G., Knight, J. et al.
(1979). The preliminary manual for the State-Trait Personality Inventory. Unpublished
manual, University of South Florida.
Sterman, M. B. (1996). Physiological origins and functional correlates of EEG rhythmic activi-
ties: Implications for self-regulation. Biofeedback and Self-Regulation, 21, 3–33.
Sterman, M. B. (2000). Basic concepts and clinical findings in the treatment of seizure disor-
ders with EEG operant conditioning. Clinical Electroencephalography, 31, 45–55.
Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental
Psychology, 18, 643–662.
Sulser, F. (1989). New perspectives on the molecular pharmacology of affective disorders. Eur-
opean Archives of Psychiatry & Neurological Sciences
, 238, 231–239.
Thatcher, R. (1999). EEG database-guided neurotherapy. In J. R. Evans & A. Abarbanel
(Eds.), Introduction to quantitative EEG and neurofeedback (pp. 29–65). San Diego, CA:
Academic Press.
Thatcher, R. W. (2000). EEG operant conditioning (biofeedback) for traumatic brain injury.
Clinical Electroencephalography, 31, 38–44.
Trudeau, D. L. (2000). The treatment of addictive disorders by brain wave biofeedback: A
review and suggestions for future research. Clinical Electroencephalography, 31, 13–22.
Twemlow, S. W., & Bowen, W. T. (1976). EEG biofeedback induced self actualization in alco-
holics. Journal of Biofeedback, 3, 20–25.
128 EFTHYMIOS ANGELAKIS ET AL.
Downloaded By: [Kounios, John] At: 18:08 23 January 2007
Vernon, D., Egner, T., Cooper, N., Compton, T., Neilends, C., Sheri, A. et al. (2003). The
effect of training distinct neurofeedback protocols on aspects of cognitive performance.
International Journal of Psychophysiology, 47, 75–85.
Wechsler, D. (1995). Wechsler Adult Intelligence Scale. San Antonio, TX: Psychological Press.
Wechsler, D. (1997). Wechsler Memory Scale-III. San Antonio, TX: The Psychological
Corporation.
Welsh, K. A., Butters, N., Hughes, J. P., Mohs, R. C., & Heyman, A. (1991). Detection of
abnormal memory decline in mild cases of Alzhemier’s disease using CERAD neuro-
psychological measures. Archives of Neurology, 48, 278–281.
Woodcock, R. W., & Johnson, M. B. (1977). Woodcock-Johnson Psychoeducational Battery.
Hingham, MA: Teaching Resources.
EEG NEUROFEEDBACK 129
