NF Basics R. Thatcher, J. Lubar

Quantitative EEG and
N e u ro f e e d b a c k i n C h i l d ren
and Adolescents
Anxiety Disorders, Depressive Disorders, Comorbid
Addiction and Attention-deficit/Hyperactivity
Disorder, and Brain Injury
a,b, c
Deborah R. Simkin, MD, DFAACAP *, Robert W. Thatcher, PhD ,
Joel Lubar, PhDd,e,f
KEYWORDS
Quantitative EEG Neurofeedback LORETA Anxiety disorders Depression
Addiction ADHD Brain injury in children and adolescents
KEY POINTS
The waveforms on an electroencephalogram (EEG) result directly from synaptic activity in
brain networks, and neurofeedback offers the opportunity for patients to use operant con-
ditioning to alter their waveforms and brain functioning.
Quantitative EEGs (qEEGs) represent a patient’s waveforms compared with normative
EEG databases and may be corepresented with functional magnetic resonance imaging
(fMRI), single-photon emission computed tomography, positron emission tomography
(PET), and magnetic resonance imaging.
qEEG, modern PET, magnetoencephalography, and fMRI studies concur in showing that
the brain is organized by a small set of modules or hubs that represent clusters of neurons
characterized by high within-cluster connectivity and sparse long-distance connectivity.
During child development, functional brain connectivity is substantially reorganized, but
several large-scale network properties seem to be preserved over time, suggesting that
functional brain networks in children are organized like other complex systems in adults.
Continued
Disclosures: Research support received from Eli Lilly and Company; Pfizer Inc; Novartis Pharma-
ceutical Corporation.
a
Committee on Integrative Medicine, American Academy of Child and Adolescent Psychiatry,
Attention, Memory and Cognition Center, 4641 Gulfstarr Drive, Suite 106, Destin, FL 32541, USA;
b
Department of Psychiatry, Emory University Medical School, Atlanta, Georgia; c Neuroimaging
Laboratory, Applied Neuroscience Research Institute, 7985 113th Street, Suite 210, Seminole,
FL 33772, USA; d University of Tennessee, Knoxville, TN, USA; e Southeastern Neurofeedback
Institute, Inc, 111 North Pompano Beach Boulevard, Suite 1214, Pompano Beach, FL 33062,
USA; f International Society for Neurofeedback and Research
* Corresponding author. 4641 Gulfstarr Drive, Suite 106, Destin, FL 32541.
E-mail address: deb62288@aol.com
Child Adolesc Psychiatric Clin N Am 23 (2014) 427–464

http://dx.doi.org/10.1016/j.chc.2014.03.001 childpsych.theclinics.com
1056-4993/14/$ – see front matter Ó 2014 Elsevier Inc. All rights reserved.
428 Simkin et al
Key Points (Continued)

Circuitry dysconnections (in and between these hubs or networks) have been associated
with symptoms of psychiatric illness, and these dysconnections can be targeted using
neurofeedback.
Quantitative EEG provides an opportunity for new evaluation techniques permitting more
efficient treatment matching, such as determining which patients will be responsive to
certain medications or which areas of the brain should be targeted during neurofeedback
treatment.
Clinicians must be intensively trained in appropriate protocols and technologies in order to
provide appropriate clinical neurofeedback treatment and to develop sound research.
Although little research has been done using neurofeedback in children and adolescents
with anxiety disorders, mood disorders, comorbid addiction and attention-deficit/
hyperactivity disorder, and brain injury, the adult literature and more recent advances in
neurofeedback interventions (eg, low-resolution electromagnetic tomography) hold
particular promise for future research using fewer sessions.
Abbreviations: Quantitative EEG and Neurofeedback
ADHD Attention-deficit/hyperactivity disorder

ApoE Apolipoprotein E
BET Brain electromagnetic tomography
BRIEF-R Behavior Rating Inventory of Executive Function-R
DRD2 Dopamine D2 receptor
DSI Diffuse spectral imaging
EEG Electroencephalogram
EMG Electromyography
fMRI Functional magnetic resonance imaging
GABA Glutamate and gamma-aminobutyric acid
HD High dysregulation
IVA-Plus Integrated Visual and Auditory Continuous Performance Test
JTFA Joint time-frequency analysis
LD Low dysregulation
LFP Local field potentials
LORETA Low-resolution electromagnetic tomography
LPFC Lateral prefrontal cortex
LTP Long-term potentiation
MEG Magnetoencephalography
NF Neurofeedback
NIMH National Institute of Mental Health
PET Positron emission tomography
PFC Prefrontal cortex
PTSD Posttraumatic stress disorder
qEEG Quantitative electroencephalogram
RDoC Research domain criteria
rIFG Right inferior frontal gyrus
ROIs Subregions of interest
SMR Sensory motor rhythms
SPECT Single-photon emission computed tomography
SPM Statistical parametric mapping
TOVA Test of variable assessment
USPSTF United States Preventive Services Task Force
vmPFC Ventromedial prefrontal cortex
WISC-R Weschler Intelligence Scale for Children-Revised
Quantitative EEG and Neurofeedback in Children and Adolescents 429
INTRODUCTION
Neurofeedback (NF) is a treatment method for altering brain functioning by the use of
signals provided to a patient that reflect the moment-to-moment changes in the pa-
tient’s electroencephalogram (EEG). The method typically uses advanced statistical
analysis of quantitative data from the EEG (quantitative EEG [qEEG]) to provide
biofeedback to the patient in real time. This approach allows the operant conditioning
of the patient’s EEG, which, perhaps surprisingly, can have the effect of therapeuti-
cally altering cognition, emotions, and behavior.
This article explores the science surrounding NF and reviews the early research on
the use of NF technology for treating psychiatric disorders in children and adoles-
cents. Although surface NF (using 2–4 electrodes, which did not use qEEG initially)
has long been used, several new NF interventions have been developed. Many of
these new interventions, along with surface NF, can now incorporate the use of quan-
titative electroencephalography to enhance their clinical value.
Three major NF methodologies are real-time z-score surface NF, low-resolution
electromagnetic tomography (LORETA), and functional magnetic resonance imaging
(fMRI) NF.
Real-time z-score surface NF (EEG biofeedback) uses 2 to 4 or more scalp elec-
trodes to monitor the brain’s electrical activity in a particular anatomic location. It
uses continuous real-time computerized calculations based on qEEG data comparing
the way that the patient’s brain is functioning on different variables with an age-
matched normative database, using z scores to measure differences from normal
EEG activity. The z score is the number of standard deviations of an observation
(data on EEG waveform) more than (positive) or less than (negative) the mean. The z
scores or standard deviations relative to an age-matched reference population pro-
vide a real-time indication of abnormal instabilities in brain networks. These z scores
provide a guide to train patients toward quantitatively normal waveforms (z 5 0) in
brain regions associated with particular disorders.
If the surface NF involves the use of 2 or more electrodes, coherence (the measure
of the number of connections and communications between groups of neurons) can
also be trained during NF training. However, if only 1 electrode is used during NF
training, no coherence training can occur. In addition, surface NF involves measuring
the amplitude of neurons directly beneath the electrode where 95% of the neurons
arise from a distance of 6 cm and all frequencies are mixed together at each electrode.
However, LORETA uses three-dimensional source localization applied to human
qEEG in which the mixture of frequencies under each scalp electrode are unscrambled
and linked to three-dimensional sources in the interior of the brain with accuracies of
approximately 1 cm in many situations.
LORETA NF uses a different kind of qEEG NF analysis that provides an estimation
of the location of the deep underlying brain generators, called modules or hubs (eg,
the anterior cingulate, insula, fusiform gyrus) and networks of the patient’s EEG ac-
tivity within a frequency band. It allows the clinician to translate qEEG data into a
three-dimensional figure that corresponds with and looks like the images in fMRI
that are associated with disease states. It requires more labor-intensive preparation,
because an electrode cap with 19 electrodes must be applied in every session, but
it can shorten the length of treatment. Coherence training can include multiple
areas.
fMRI NF’s advantage is that it can examine functioning at deep subcortical areas of
the brain. However, the practical disadvantage of fMRI NF is that it is expensive, with
equipment that costs $1 million or more and is not portable.
430 Simkin et al
NF research has been limited in children and adolescents, especially with regard to
anxiety, mood, addiction, and traumatic brain injury (TBI). Research has shown that
functional hub architecture matures in late childhood and remains stable from adoles-
cence to early adulthood. Thus, LORETA NF, which targets modules and connections
between modules, is expected to work the same in patients from 10 years of age to
adulthood.
New transdiagnostic approaches, proposed by the National Institute on Mental
Health, to defining psychiatric disorders based on dysfunctional connectivity are
particularly significant because NF targets dysfunctional connectivity. Clinicians
want to familiarize themselves with possible new treatment interventions that target
these transdiagnostic symptoms. Other uses of qEEG include determining whether
patients will be responsive to medications. In addition, the initial findings on NF
research for treating psychiatric symptoms in youth will be covered.
NEUROFEEDBACK BASICS
NF is based on operant conditioning. NF is based on measurements of evoked poten-

tials as measured in electroencephalography. An evoked potential can be negative or
positive relative to the average baseline potential. An N100 is a negative evoked po-
tential that is generated by a network of neurons, usually provoked by an unexpected
stimulus, in a brain location depending on the sensory modality. P100 is a positive
evoked potential in the same region.
Suppose an electrode is recording a P100 evoked potential in a particular area of the
brain, and the goal of the NF intervention (to treat a symptom) is to change the P100
positive evoked potential to an N100 negative evoked potential. Following the
methods used in operant conditioning, every time an N100 evoked potential begins
to occur by chance, a positive reinforcement is provided. The positive reinforcement
could be milk delivered into a bowl for a cat. In NF, the patient is provided with a pos-
itive reinforcement through the biofeedback presentation of a signal that is linked to
the desired alteration in brain wave pattern. The brain would eventually learn to search
for this N100 evoked potential in order to receive the positive reinforcement associ-
ated with this evoked potential. This example shows how, through operant condition-
ing, NF can provide a reward for the appearance of a particular electrical event in the
brain. The electrical pattern soon begins to appear before the reward, leading to
increased frequency of appearance of that electrical pattern. In a similar way in
humans, if a dot appears on the screen during an NF session with each occurrence
of a particular targeted EEG pattern or rhythm (eg, delta 0–4 Hz, theta 4–8 Hz, alpha
8–12 Hz, beta 12–20 Hz, gamma 20–300 Hz), the rhythm eventually continue to occur
in anticipation of the dot appearing.
By knowing the anatomic location of a particular electrical pattern associated with a
targeted brain function, NF can be used to change a wide variety of brain functions,
including depressed states, anxiety, addiction, injury-induced abnormalities, and
attention-deficit/hyperactivity disorder (ADHD). Thus, a precise system for mapping
locations in the brain is essential to NF.
To detect specific electrical patterns in the brain, EEG electrodes are placed on the
scalp corresponding with anatomic areas associated with particular Brodmann areas,
with the aim of targeting certain functions associated with these areas. A targeted
rhythm can be pinpointed by an electrode placed on the scalp (for example, corre-
sponding with T3, using the International 10–20 System illustrated in Fig. 1, which cor-
responds with a particular area of the brain). For example, T3 is associated with
Brodmann areas 22/42 on the superior temporal gyrus. Affective prosody is
Fig. 1. The International 10–20 System seen from (A) the left and (B) above the head. A, ear
lobe; C, central; F, frontal; Fp, frontal polar; O, occipital; P, parietal; Pg, nasopharyngeal.
(From Ferreira A, Celeste WC, Cheein FA, et al. Human-machine interfaces based on EMG
and EEG applied to robotic systems. J Neuroeng Rehabil 2008;5:10.)
associated with this Brodmann area on the right and language processing on the left.
Abnormal evoked potentials (representing abnormal brain wavelengths) in this region
may be identified on an EEG. Using NF to target specific Brodmann areas that have
abnormal evoked potentials that are found in an individual to be associated with symp-
toms or specific disorders may change that area of the brain back to a normal pattern
and, it is hoped, decrease the symptoms of the disorder.1
NF is based on the clinician’s ability to link a patient’s symptoms and complaints to
dysregulation or deviation from normal in EEG patterns in particular brain regions
known to be related to specific functions. Fig. 2 shows the various anatomic regions
of the brain, originally delineated by Brodmann2 in 1909 based on their microarchitec-
ture and their links to particular functions, as based on the findings of various tech-
niques across many patients. Brodmann areas are macroscopic brain regions of
common functional cytoarchitecture ranging in size from about 1 cm to 6 cm, and a
qEEG can give a precise delineation of these Brodmann areas for each individual.
Following qEEG assessment of abnormal brain rhythms in precisely defined brain re-
gions in an individual, NF treatment aims to modify dysregulated subsystems and
global linkages toward the normal range of function. Periodic qEEG assessments dur-
ing treatment can be used to monitor treatment efficacy. The assessment is similar to
the use of a blood test to identify deviant constituents of the blood (eg, increased liver
enzymes) that can be linked to the patient’s symptoms and aid in making treatment
decisions and monitoring treatment efficacy.3
HOW NEUROFEEDBACK CHANGES ACTION POTENTIALS
The science and techniques involved in NF are described in more detail in the Hand-
book on Quantitative Electroencephalography and EEG Biofeedback by Thatcher.1
The registration of wavelengths on a raw EEG results directly from synaptic electri-
cal action potentials produced by chemical synapses in neuronal networks. There are
432 Simkin et al
Fig. 2. Various functions associated with particular Brodmann areas based on fMRI, PET,
EEG/magnetoencephalography (MEG), and lesion/tumor studies. (Data from Refs.1–5; and
Courtesy of M. Dubin, PhD, Boulder, CO.)
2 types of chemical synapses that produce the EEG wavelengths. First, fast synapses
involve the neurotransmitters glutamate and gamma-aminobutyric acid (GABA), which
are associated with fast gated ion channels (which occur from 0–80 milliseconds).
Second, slow synapses involve dopamine, serotonin, acetycholine, and norepineph-
rine, which are associated with slow voltage gated ion channels (which occur from
100 milliseconds to 1 second). These excitatory and inhibitory postsynaptic potentials
give rise to local field potentials (LFPs). LFPs influence the firing of action potentials in
pyramidal neurons near the surface of the brain.1,6 These wavelengths can be
changed by operant conditioning that targets these networks. Operant conditioning
of the EEG involves changes in synapses caused by what is referred to as a phase
reset. Operant conditioning begins by reinforcing a frequency (eg, 5 Hz) associated
with a particular wavelength (eg, theta) detected by EEG electrodes near the surface
of the brain. When a desired wavelength occurs, there is a burst of action potentials
that impinges on the dendrites and cell bodies of pyramidal neurons. The time it takes
for this wavelength shift to occur is referred to as phase shift duration and it can be
seen on an EEG when the same wavelengths are not in sync (Fig. 3). When a phase
lock occurs, the desired wavelengths will be synchronized (see Fig. 3).
To describe a phase shift or wavelength shift, Thatcher1 gave the following example.
Imagine a family is at a Thanksgiving dinner, and an unexpected relative suddenly ar-
rives whom no one has seen in years. The family shifts from focusing on the dinner to
focusing on the relative at the door. Over time, as more and more of the family mem-
bers recognize the unexpected relative, more family members shift their attention and
move toward the door. If more time is given (phase shift duration), more family mem-
bers move to the door. This analogy also applies for phase shifts in the brain. The
Fig. 3. Phase lock.
longer the duration of the phase shift, the greater the number of neurons that are
recruited.1 When this burst of neuronal activity synchronously occurs in millions of
neurons, then a detectable change in the EEG frequencies occurs, (eg, from theta
4–8 Hz to alpha 8–12 Hz).7
When all the family members arrive at the door, a phase lock occurs (when all of the
EEG frequencies are the same for a particular rhythm). During operant conditioning, a
wavelength rhythm is reinforced when the duration of phase lock is increased and the
frequency of phase shift is decreased. Likewise, a rhythm is inhibited when the dura-
tion of phase lock is decreased and the frequency of phase shift is increased.1 During
NF, a change in an individual’s EEG activity can be reinforced by a visual and/or audi-
tory stimulus. As NF continues, the reinforcement is more difficult to obtain because
the length of time the individual has to sustain the preferred wavelength is increased.
As the individual masters the level of difficulty in sustaining the preferred wavelength,
the clinician increases the level of difficulty in order to continually challenge the brain
by increasing the length of time the individual must hold the preferred wavelength.
Fast excitatory chemical synapses dominate long-distance corticocortical loops
and fast inhibitory synapses dominate short-distance loops. However, long-duration
neurotransmitters shape and mold the mechanisms of long-term potentiation
(LTP).1,8 Synapse modifications occur during the phase lock. Neurotransmitters,
such as dopamine, are released in anticipation of these rewarding experiences that
occur when particular wavelength patterns are elicited during phase lock. Neurotrans-
mitters released during phase lock can influence structural plasticity within the brain.
Phase lock produces a long-lasting enhancement in signal transmission between 2
neurons that results from their synchronous firing, which is associated with increased
dendrite formation and increase neurotransmitter delivery into the cleft. This process
of synapse modification is called LTP, and it requires synapse growth and the devel-
opment of new synapses during learning. Kandel9 received the Nobel Prize in 2000 for
his work linking LTP to DNA, RNA, and protein production associated with learning
and memory formation. Phase lock results in LTP or neuronal learning at the molecular
level that moves the patient having NF treatment toward normal functioning.
In effect, EEG NF uses operant conditioning to (1) reinforce particular brain rhythms
by reinforcing phase lock and decreasing phase shift, or (2) inhibiting rhythms by
decreasing the frequency of phase lock and increasing the frequency of phase. The
phase shifts and phase locks reinforced by NF are associated with long-term synaptic
modifications characterized by changes in neurotransmitter release and neuronal
functioning. For example, if the symptoms of ADHD are known to be associated
with an abnormal wavelength in a particular area of the brain, then NF can change
the wavelength to those found in normal individuals. By using NF, clinicians may be
able to effectively treat the symptoms of ADHD by selectively enhancing dopamine
transmission in the relevant parts of the brain.
434 Simkin et al
In addition to frequency changes and phase shifts, NF can be used to modify syn-
aptic and network functioning by operating on EEG data reflecting brain coherence.
Coherence is a measure of coupling between groups of neurons, or more precisely
a measure of the number of connections and communications (or frequency of activ-
ity) between groups of neurons with a constant phase relationship. Coherence is pro-
portional to phase lock and inversely proportional to phase shift; when coherence is
working well in desired networks, several areas of the brain are in phase lock.
IS ALL BRAIN BIOFEEDBACK ALIKE?
NF is a subtype of EEG biofeedback in which EEG data, or signals based on EEG data,
are used as the feedback to the patient. Other types of biofeedback include electro-
myography (EMG) biofeedback and thermal (or temperature) biofeedback. EMG
biofeedback measures electrical activity associated with muscle contractions, and it
is often used for relaxation training, stress management, peak performance training,
and pain management. Thermal biofeedback uses a temperature sensor (electronic,
computerized, liquid crystal, or a glass thermometer) to detect temperature changes
in the extremities (usually fingertips or toes). Stress and nervous system excitation/
arousal causes blood vessels in the extremities to constrict, and the reduced blood
flow leads to cooling. Thermal biofeedback is used to train people to quiet the nervous
system arousal mechanisms that produce hand or foot cooling, and this is often used
for relaxation training, stress management, and pain management.10
In contrast, NF specifically uses data based on brain functioning, specifically EEG
data. Surface NF, involving 2 to 4 electrodes, started before the advent of qEEG,
but modern surface NF often uses qEEG data for assessment before NF treatment
is begun. LORETA, which involves many more electrodes and can capture EEG
data from deeper structures, is a newer approach that makes integral use of qEEG
data. Other forms of NF are discussed by Hurt and colleagues elsewhere in this issue.
QUANTITATIVE ELECTROENCEPHALOGRAPHY AND ITS ROLE IN NEUROFEEDBACK
Quantitative electroencephalography involves computerized data analysis to precisely

quantify the electrical potentials or frequency bands from 0 Hz to transform the EEG to
a format that elucidates relevant information, such as highlighting specific waveforms
components.11 The mathematical technique used to decompose and transform the
mixture of waves in the human EEG is Fourier transform or, given the periodic nature
of EEG data, the Fourier series coefficients. Joint time-frequency analysis (JTFA) is
another mathematical procedure developed in the late 1980s that gave rise to precise
time-frequency measures and quantification of phase locks and phase shifts.12
For technically informed readers, some features of qEEG that can be used as the
basis for NF include absolute power (the average amount of power uV2 in each fre-
quency band or wavelength and in the total frequency spectrum recorded from
each electrode site), relative power (the percentage of total power contributed by
each frequency band or wavelength in the spectrum from each electrode site), coher-
ence (the amount of synchronization of electrical events in corresponding brain re-
gions, separately for each frequency band and for the entire frequency spectrum),
and symmetry (the ratio of power in each band between a symmetric pair of
electrodes).
A qEEG can only be validly interpreted after artifacts are removed from the digital
record. If a clinician does not have the training to interpret EEGs, then a skilled and
trained professional consultant should be involved. The artifact-free information is
then compared with normative databases to identify brain regions whose EEG data
are deviant by 2 or more standard deviations more than or less than the mean. These
brain regions are targeted for NF training if the z scores (the number of standard de-
viations off the mean) are thought to be relevant to the patient’s symptoms and if these
symptoms are associated with the area of the brain that is typically responsible for the
patient’s malfunctioning.
The uses of normative databases have been validated in the scientific literature. The
normative databases contain the raw EEG records and features derived from analysis
of data from individuals aged 6 months to about 90 years. The number of subjects
required for reliability at each age point were statistically determined and increased
until consistent split-half replications were obtained. The sampling requirement was
interesting because each age required different numbers of participants. For example,
in the ages from 6 months to 13 years, when brain maturation changes are rapid,
higher numbers of subjects were needed. This use of age-regression normative equa-
tions and z scores, good test-retest reliability, and lack of ethnic bias allows the use of
qEEG that is not only noninvasive but highly sensitive to abnormalities in brain function
found in psychiatric populations.13–15 Replications of normative databases have been
extended to cover the range from 1 to 95 years of age for each of the electrode posi-
tions in the standardized International 10–20 System, and they have been broadened
to include measures of absolute power, relative power, mean frequency, coherence,
and symmetry.14–20
Distinctive patterns of qEEG abnormalities have been described in diverse psychi-
atric disorders in adults and youth. These distinctive diagnostic patterns allow differ-
entiation of these disorders from normal and from each other.8,13 A large body of
peer-reviewed published data from independent laboratories reports the sensitivity
of neurometrics in varied clinical populations, including head injury,13,14,21 schizo-
phrenia,22 depression,23 marijuana abuse,24 and ADHD.25,26
VALIDATION OF QUANTITATIVE ELECTROENCEPHALOGRAPHY BY USE OF

NEUROIMAGING AND OTHER TECHNIQUES
In the early 1990s, efforts were made to identify the three-dimensional location of deep
sources in the interior of the brain of the surface (scalp) EEG data, and then to correlate
these deep sources with MRI tomographic data.27 This coregistration of deep EEG
sources to MRI slices is known as EEG tomography (tEEG), electrical neuroimaging,28
or brain electromagnetic tomography.15 These efforts were expanded to coregister all
imaging modalities, including, positron emission tomography (PET), single-photon
emission computed tomography (SPECT), and fMRI, and to create a common
anatomic atlas. Based on neurosurgical identification of areas pertinent to function,
the Talaraich atlas and later the Montreal Neurological Atlas were subsequently
used to incorporate EEG data into the Human Brain Mapping Project.27,29–32 It was
used to visualize Brodmann areas as they were defined for the Talairach brain33 and
to compare Brodmann areas across subjects.
tEEG is based on the ability to measure the location of three-dimensional sources of
the scalp surface EEG in the interior of the brain and then register the sources to MRI
tomographic slices. The advent of tEEG is important because it provides coregistra-
tion of 2 imaging modalities that have similar spatial localization characteristics, in
which fMRI measures blood flow and the qEEG adds a high temporal resolution of
changes in the electrical sources in the brain that are associated with changes in blood
flow.1
In 1994, Robert Marqui-Pascual4 devised accurate estimates of the deep (lower)
brain sources of the EEG patterns in small regional voxels (approximately 4 mm to
436 Simkin et al
1 cm cubic voxels) coregistered to MRI slices. He transformed these raw EEG sig-
nals into three-dimensional images that were then coregistered on the Talairach
MRI atlas. This new method was called LORETA. The Web site to obtain additional
information on LORETA is http://uzh.ch/keyinst/loreta.htm. LORETA provides better
temporal resolution than can be achieved with either PET or fMRI. This high tempo-
ral resolution is important for studies using event potentials (ie, time-locked events)
and also for investigating brain changes proposed to be associated with psycholog-
ical states, such as depression.34 Using LORETA allows a clinician to translate
qEEG data into a three-dimensional figure that corresponds with and looks like
the images in fMRI that are associated with the same disease state. Furthermore,
during NF treatment, as waveforms are adjusted toward normal, the three-
dimensional images generated by LORETA also become more consistent with a
normal fMRI.
A statistical normalization was later applied to LORETA and was called
sLORETA.35,36 The first normative tEEG databases, using z scores and Gaussian or
normal distributions similar to fMRI, are referred to as statistical parametric mapping
(SPM) and were introduced by Valdez in 200132 and followed by Thatcher and col-
leagues37,38 in 2005.
Modern PET, qEEG, magnetoencephalography (MEG), and fMRI studies all agree
that electrical activity in the brain is organized by a small set of modules or hubs
that represent clusters of neurons with high within-cluster connectivity and sparse
long-distance connectivity (Fig. 4).29,39–41
Quantitative electroencephalography and MEG are the only 2 imaging methods that
have sufficient spatial and temporal resolution to measure the millisecond dynamics of
Fig. 4. Quantitative EEG measures short-distance and long-distance coherence, phase de-
lays, phase locking, and phase shifting of different frequencies. The qEEG reflects top-
down causality at the macro level (scalp surface EEG) coordinating the meso and micro levels
of neural organization. The dysregulation of groups of neurons at the micro and meso
levels, which simultaneously mediate specialized functions, can be measured at the macro
level using quantitative EEG. (From Le Van Quyen M. The brainweb of cross-scale interac-
tions. New Ideas Psychol 2010;29:57–63; with permission.)
hubs and modules. Both use z scores to estimate dysregulation in these brain areas,
and these z scores can be linked to a patient’s symptoms. However, quantitative elec-
troencephalography can better detect deeper cortical sources and is much less
expensive than MEG. For that matter, a qEEG is less expensive than fMRI and PET
scans.29
COREGISTRATION OF QUANTITATIVE ELECTROENCEPHALOGRAPHY WITH DIFFUSE

SPECTRAL IMAGING AND OTHER NEUROIMAGING METHODS
The human cortex has been arranged in 6 basic cluster modules that have been
measured using diffuse spectral imaging (DSI) (Fig. 5). Coregistration of quantitative
electroencephalography to these 6 modules based on DSI can be used to deter-
mine phase dynamics and fine temporal coherence within and between these
modules.38
Hagmann and colleagues39 developed these 6 modules. They used DSI to trace
cortical white matter connections of the human cerebral cortex between 66 cortical
regions, using clear anatomic landmarks, based on Brodmann areas.42 From the
66 cortical regions, 998 subregions of interest (ROIs) were calculated using a
connection matrix of inter-regional cortical connectivity. Network spectral analyses
of nodes and edges of the 998 ROIs were grouped into 6 anatomic modules.38
These 6 anatomic modules include, but are not exclusive to, the posterior cingulate,
the bilateral precuneus, the bilateral paracentral lobule, the unilateral cuneus,
Fig. 5. The locations of the 6 Hagmann modules (MOD1–6). These modules represent the
Hagmann DSI-based modules coregistered to quantitative electroencephalography. (From
Thatcher W, North DM, Biver CJ. Diffusion spectral imaging modules correlate with EEG
LORETA neuroimaging modules. Hum Brain Mapp 2011;33:1062–75; with permission.)
438 Simkin et al
the bilateral isthmus of the cingulate gyrus, and the bilateral superior temporal
sulcus. A replication of the 6 modules described by Hagmann and colleagues39,43
using DSI and coregistered to quantitative electroencephalography is shown in
Fig. 5.
In summary, z scores (based on qEEGs) allow clinicians to determine the location
and extent of dysregulation with respect to a group of age-matched controls. A
tEEG normative database of Brodmann areas and hub and modules when linked to
a patient’s symptoms aids a clinician in making a diagnosis. The goal is to target
weak systems and avoid compensatory systems. The z scores or standard deviations
with respect to an age-matched reference population provides a real-time guide to
train patients toward z 5 0 in brain regions associated with particular disorders.1,44–46
The clinical use of qEEG in neuropsychiatry involves 3 distinct steps: (1) a clinical inter-
view and evaluation of the patient’s symptoms and complaints, (2) linking the patient’s
symptoms to functional specialization in the brain based on the scientific literature
(qEEG/MEG; fMRI; PET; SPECT, and so forth) and, (3) real-time z-score surface or
LORETA to modify deviant or deregulated brain regions associated with the patient’s
symptoms and complaints.1,29
Z scores derived from a normative qEEG database can be used as an aid to diag-
nosis, but they cannot be used as a stand-alone approach to diagnosis (just as PET
scans can help but not make a diagnosis). However, qEEG normative database z
scores, like PET scans or fMRI, can be used to monitor the course of treatment (eg,
of transcranial magnetic stimulation, NF training, or psychopharmacologic interven-
tions) or evaluate the comparative efficacy of treatments.
QUANTITATIVE ELECTROENCEPHALOGRAPHY AND NF FOR PSYCHIATRIC CONDITIONS

Role of the EEG Power Spectrum, Neuromodulators, and Psychiatric Disorders
As discussed earlier, the changes in wavelengths produced during phase lock can
cause the release of neuromodulators, which in turn can cause synaptic modifications.
However, these neuromodulators also have a role in specific systems or circuits in the
brain that influence the expression of clinical psychiatric symptoms.
How Complex Homeostatic Systems Regulate the EEG Power Spectrum

Hughes and John47 provide a more detailed explanation of the brief summary given
here. Large neuronal populations in the brainstem, thalamus, and the cortical areas
mediate EEG power spectrums.47 The thalamus is part of the brainstem through which
all sensory information (except olfactory information) flows. This sensory-driven infor-
mation is sent to the cortex, and the cortex relays information back to the thalamus
(the thalamic-cortical-thalamic pathway); the thalamus then sends the information to
the other parts of the brain.
Changes in the EEG frequency in the thalamus can increase or decrease the flow of
information by way of large pacemaker neurons that are distributed throughout the
thalamus. Inhibition of information flow in the thalamic-cortical-thalamic pathway oc-
curs when the rhythm increases from a theta (4–8 Hz) and low alpha (8–10 Hz) range to
the faster rhythms in the high alpha (11–12 Hz) range and still faster beta (12.5–2.0 Hz)
range. These faster activity ranges are thought to be involved in corticocortical and
corticothalamic interactions during information processing.47 When phase shift and
phase lock are functioning efficiently (eg, in a patient without ADHD), neuromodulators
influence rhythms that are responsible for selectively inhibiting cortical regions during
cognitive processing. Deficiencies or excesses of neurotransmitters or modulators
can therefore affect the efficiency of the relay of information, and thereby the
homeostasis of the brain (and potentially play a role in psychiatric disorders). Thus, NF,
by changing theta waves to high alpha and beta waves, may correct these deficiencies
and bring the brain back to a more normative and efficient way of processing.
Comparison with normative databases shows that raw EEG power spectrums (con-
ventional EEG) of alpha, beta, theta, and delta frequency bands for each electrode in
the standardized International 10–20 System have links to psychiatric disorders.
Normal distributions of power in these frequency bands in normal, healthy individuals
also show high test-retest reliability. However, although the conventional EEG can be
used in the diagnostic work-up of such things as acute confusional states; first presen-
tation of schizophrenia; major mood or mania; and refractory behavioral problems
such as obsessions, panic, or violence, qEEG is particularly well suited to identify sub-
tle changes in the topographic distributions of background activity. qEEG may aid in
identifying difficult differential diagnosis, such as: distinguishing between mood disor-
ders and schizophrenia; assessing cognitive, attentional, or developmental disorders;
distinguishing between environmentally induced and endogenously induced mediated
behavioral disorders; evaluating alcohol and substance abuse; and evaluating post-
concussion syndrome.47 More research using NF in children and adolescents is
needed to determine whether NF may be able to change these disease states as well.
How May This Occur in the Case of ADHD?

ADHD provides an example of how changes in the flow between these pathways can
influence psychiatric symptoms. The major qEEG frequency abnormalities seen in
ADHD involve an excess of theta and, in some cases, low alpha.48–50 Furthermore,
an excess of theta and low alpha waves might result from low dopamine levels that
may be caused by a hypofunctioning prefrontal cortex (PFC) and/or the nigrostriatal
system, via low dopaminergic firing. These qEEG findings are in agreement with the
dopaminergic theory of ADHD expressed by Levy,51 which conceptualizes ADHD of
the polysynaptic dopaminergic circuits between prefrontal and striatal centers of ac-
tivity. These findings are also compatible with the neurophysiologic model of ADHD
proposed by Niedermeyer and Naidu,52 which emphasizes prefrontal, frontal and
striatal, and thalamic interconnections. The Levy51 model is also supported by MRI
and PET imaging studies and by behavioral, pharmacologic, and neuroanatomic
studies on the nature of cortical and subcortical disturbances in function that charac-
terizes children with attention and learning problems.16
In summary, the dysregulation of the thalamic-cortical-striatal system in ADHD is
still not completely understood. However, the high levels of theta and low alpha wave-
lengths registered on qEEG, which are associated with ADHD in specific regions of the
brain, are thought to be changed by NF through operational conditioning, changing
the theta waves (inattentive state) to high alpha and beta waves (attentive state). How-
ever, the effects of NF training are not limited to ADHD alone.
Psychiatry as Related to Circuitry: Moving from Categorical Disorders to

Transdiagnostic Symptoms
As psychiatry prepares to better understand psychiatric disorders from a neurobiolog-
ical perspective, clinicians are developing a new way of describing psychiatric symp-
toms. Core connectivity circuits may be disrupted in the brain, which leads to the
expression of these symptoms. Buckholtz and Myer-Lindenberg53 expanded on this
idea. Populations of neurons are segregated to perform specialized functions (ie, lan-
guage processing in the left inferior frontal gyrus). Simple execution of these functions
requires that specialized output of each of these segregated neuron populations be
integrated or connected. The notion that schizophrenia is a disorder of
440 Simkin et al
dysconnectivity has a long history. Connectivity analysis (using fMRI research) in

healthy individuals has uncovered specific networks associated with cognition, affect,
motivation, and social function.
These networks have been identified as part of an initiative at the National Institute
of Mental Health (NIMH). The Research Domain Criteria (RDoC) at NIMH54 have orga-
nized symptom domains that correspond with neuropsychological junctions based on
connectivity between specialized neuronal populations. There are 4 domains: the
corticolimbic, the corticocortical, the frontoparietal, and the default mode network
(Fig. 6). These circuits underpin core executive, affective, motivational, and social
domains. Heritable variation in the function of these circuits produces deficits in
circuit-specific cognitive domains that manifest as psychiatric symptoms. Using these
domains, these symptoms are seen to be common to multiple disorders rather than
specific to unique categorical disorders.
Buckholtz and Meyer-Lindenberg53 summarized these circuits and their function as
follows (see Fig. 6):
1. The amygdala, medial prefrontal cortex (ventromedial and medial orbital aspects
along with the perigenual cingulated cortex), and lateral prefrontal cortex (LPFC)
Fig. 6. Genetic variation affects risk for mental disorder by disrupting cognition-specific
brain circuits. dACC, dorsal anterior cingulate; DRD2, dopamine D2 receptor (associated
with substance abuse, ADHD, schizophrenia, and antisocial personality disorder); LPFC,
lateral prefrontal cortex; MAOA, monoamine oxidase A (associated with serotonin and
mood disorder and antisocial personality disorder); MPFC, ventromedial prefrontal cortex
(ventromedial and medial orbital cortex, and perigenual cingulate cortex); OFC, orbital
frontal cortex; PCC, posterior cingulate cortex; TPJ, temporal parietal junction; ZNF804a, a
series of genome-wide association with schizotypal and decreased social cognition. (Adapt-
ed from Buckholtz JW, Meyer-Lindenberg A. Psychopathology and the human connection:
toward a transdiagnostic model of risk for mental illness. Neuron 2012;74(6):990–1004.
http://dx.doi.org/10.1016/j.neuron.2012:06.002; with permission.)
make up the corticolimbic circuit and is associated with affective symptoms, such
as anger, anxiety, rumination, and hypervigilance. This circuitry in healthy individ-
uals is engaged during negative emotional arousal and requires regulation of these
negative emotions. The dysregulation of these affective symptoms is associated
with mood, anxiety, schizophrenia, conduct, substance abuse, and personality
disorders.
2. The default mode network is thought to be more commonly engaged when people
think about the thoughts, beliefs, emotions, and intentions of others. It involves the
temporoparietal junction, posterior cingulate, and ventromedial prefrontal cortex
(vmPFC). Dysfunction of this circuitry is associated with poor social cognition
and is associated with psychosis, personality disorders, mood disorders, and
ADHD.
3. The frontostriatal network is made up of the vmPFC, LPFC, orbital frontal cortex,
and the striatum. Dysconnectivity can lead to impairment of motivational and
hedonic responses, cognitive flexibility, value-based learning, and decision mak-
ing. Such impairments cut across many mental disorders, including anhedonia in
schizophrenia and mood disorders; impulsivity in ADHD, substance abuse, schizo-
phrenia, and personality disorders; and compulsivity in obsessive-compulsive dis-
order and substance abuse.
4. In addition, the frontoparietal network involves the dorsal LPFC, dorsal cingulate,
and parietal cortex, which are critical to executive function. Executive function in-
volves tasks that use working memory, goal-directed attention, conflict detection,
and performance monitoring. Deficits in executive function span several disorders,
including schizophrenia, ADHD, major depression, and substance abuse.
These pathways are influenced by multiple small effect risk alleles, like the DRD2
allele, which can affect different circuitry in these 4 networks as shown in Fig. 6. There-
fore, a genetic variant of DRD2 affecting, for instance, the frontoparietal network can
cause a deficit in executive function. This deficit increases susceptibility to multiple
disorders because the resulting deficits are not disorder specific. For instance, the
dopamine DRD2 receptor shows significant effects associated with schizophrenia,
ADHD, substance abuse, and antisocial behavior. The reader is referred to the article
by Buckholtz and Meyer-Lindenberg53 for further explanation of how genetic varia-
tions of particular genes in Fig. 6, as well as other variants of genes, can disrupt cir-
cuitry in these networks.
Hence, certain genes can influence connectivity in networks linked to symptom do-
mains, which implies that connectivity changes seen in mental illness reflect the cause
of the illness and are not consequences of the illness.
These pathways can also be influenced by environmental risk factors, as well as epi-
genetics. Nonetheless, this hypothesis could lead to a new way to organize psychiatric
disorders. Instead of discrete categorical disorders, the influence of genes and envi-
ronment can disrupt system-level circuits for several dimensions of cognition, produc-
ing instead transdiagnostic symptoms or symptoms that overlap several disorders,
which may why comorbidity among diagnoses is so frequently observed. Correcting
dysfunction in circuits in these 4 networks can explain how interventions using trans-
magnetic stimulation, qEEG, and (for deeper structures) LORETA NF may work to
relieve symptoms associated with broad domains of mental disorder. For instance,
LORETA can target every brain area listed in all 4 networks with the exception of
the striatum.
Research that gives credibility to targeting these circuits has already begun. For
instance, the Longitudinal Assessment of Mania Study55 was designed to identify
442 Simkin et al
distinct developmental trajectories of behavioral and emotional dysregulation in youth

with bipolar disorder and high dysregulation (HD) and youth without bipolar who had
low dysregulation (LD). Youth with LD had greater activity in the dorsolateral prefrontal
cortex and greater functional connectivity among the amygdala, dorsolateral and
ventrolateral prefrontal cortex, and anterior cingulate. Therefore, a possible use of
LORETA NF in youth with HD with bipolar disorder is to target the dorsolateral prefron-
tal cortex and, perhaps, improve coherence in the less functional circuits (ventral
LPFC to amygdala).
Targeting specific networks in youths who are already ill is not the only way
LORETA NF may be used. Research using MRI is now beginning to identify dysfunc-
tional areas of the brain that may predict the risk of developing a particular disease.
For instance, in a study by Hajek,56 youths who had a risk of developing bipolar dis-
order (identified by having 1 or 2 parents with a diagnosis of bipolar disorder) were
compared with unaffected youths who did not have one or more parents with a diag-
nosis for bipolar disorder. Of those who were unaffected (and who had a risk of devel-
oping bipolar disorder), 8 of 21 had an abnormally large right inferior frontal gyrus
(rIFG) compared with controls. The individuals were followed 4.5 years after the scan-
ning. Of the 8 who had a larger rIFG, 4 converted to a psychiatric disease, 3 devel-
oped a major depression, 1 an anxiety disorder, and 1 a personality disorder (all
requiring medication). Of the unaffected with a risk for developing a bipolar disorder
who had rIFGs that were comparable with controls, only 2 converted to an axis I dis-
order (1 who developed an adjustment disorder following a motor vehicle accident
and 1 who had a milder depressive disorder). Therefore, it can be postulated from
this study that the development of more severe forms of psychiatric disorders can
be better associated with certain abnormal circuits before the diseases develop.
Also, the identified circuits at risk can develop into a variety of disorders, congruent
with the idea behind transdiagnostic symptoms. Therefore, future studies of these
networks may help to predict who may be at risk and interventions to correct these
abnormalities may help prevent the development of psychiatric disorders. More
research is needed.
In summary, targeting networks in individuals with known symptoms associated with
one or several psychiatric diseases and cross-correlating the research with qEEG may
provide an avenue to target these areas using LORETA NF. It is speculated that, by
using LORETA NF, the symptoms of these diseases may be relieved or decreased in
severity. The new trend to consider psychiatric illnesses as transdiagnostic entities
may be an avenue for many new interventions, not limited to LORETA NF, such as
repetitive Transcranial Magnetic Stimulation (rTMS) or cognitive training programs.
NEUROFEEDBACK TREATMENT IN CHILDREN AND ADOLESCENTS
Two types of NF have been described in case reports in adults and youth. Z-score sur-
face NF training, using 2 to 4 or more scalp electrodes, uses continuous real-time
computerized calculations comparing the way that the patient’s brain is functioning
with a normative database on different variables (eg, power, asymmetries, phase-
lag, and coherence). LORETA uses a different kind of qEEG NF analysis that provides
an estimation of the location of the underlying brain generators (eg, the anterior cingu-
late, insula, fusiform gyrus) and networks of the patient’s EEG activity within a fre-
quency band. There is no current published research on the use of LORETA NF in
youths. It requires more labor-intensive preparation, because an entire electrode
cap with 19 electrodes must be applied in every session, but it can shorten the length
of treatment. fMRI NF’s advantage is that it can examine functioning at deep
subcortical areas of the brain. However, the practical disadvantage of fMRI NF is that it
is very expensive, with equipment that costs $1 million or more and is not portable.
Unless specifically stated, the following research has been done in adults.
Neurofeedback Treatment of Depression

qEEG research in adults with depression has focused on the structures involved in
processing emotions, which are found in the left and right hemispheres, including
the amygdala, orbitofrontal cortex, basal ganglia, and the hippocampus. Most of
the research has been done in adults and involves surface NF. The left hemisphere
is involved in processing positive emotions and the right is involved in processing
negative emotions.57 Depression is associated with hypometabolism in the cingulate
and occasionally in the orbitofrontal cortex, insula, anterior temporal cortices, amyg-
dala, basal ganglia, and thalamus.
The details of the associations between brain wavelength rhythms and behavioral
states are complex and do not need to be understood in order to understand NF; how-
ever, some of these details are mentioned here. Delta (0–4 Hz) and theta (4–8 Hz)
rhythms are associated with hypoactive states, and alpha (8–12 Hz) is associated
with activated states. However, higher alpha in the right frontal cortex and lower alpha
in the left frontal cortex are associated with positive mood. Beta rhythms (12–30 Hz)
are usually associated with activated states, but can be subdivided: beta 1 (12–
15 Hz) on the right is associated with a calm and observant state, beta 2 (15–18 Hz)
on the left is associated with a fully attentive and less depressed state, and beta 3
(19–30 Hz) on the left and right is associated with the anxious and irritable state. These
known associations between behavioral states and wavelengths can provide targets
for NF treatment of depression.58–62
Quantitative EEG has been used to distinguish between unipolar depression and bi-
polar disorder in adults.59 Increases in alpha power in the left hemisphere and de-
creases in left parietal-occipital beta and hypercoherence in the right anterior region
were found in adults with unipolar depression, whereas decreased alpha on the left
and increased left parietal-occipital beta were found in adults with bipolar
disorder.16,59–61
The hypercoherence in depression can be related to networks being oversaturated
in order to attempt to overcompensate for the inefficiency in the networks and there-
fore being unable to process emotional information appropriately. Higher theta and
alpha coherence is found primarily in longer-distance connections between the fron-
topolar prefrontal cortex (Brodmann area 10, Fig. 2) and the temporal or parieto-
occipital regions. Higher beta coherence is primarily found in connections within
and between electrodes overlying the dorsolateral prefrontal cortex or the temporal
regions.60,61
Adults with depression show alpha asymmetry on qEEG, and this alpha asymmetry
has been used as a primary target in the treatment of depression in adults and also as
a target in some reports in youth: in alpha asymmetry, the frontal lobes show higher
alpha on the left and lower alpha on the right; in contrast, posterior (parietal-temporal)
regions show lower alpha on the left and higher alpha on the right.59,60 Individuals
showing the frontal alpha asymmetry mentioned earlier typically are prone to experi-
encing negative withdrawal states. Those who experience increased alpha activity on
the right relative to the left tend to experience more positive affective states.61
In female adolescents with depression, one study showed that the right posterior
lobe had findings typically found in adult depression.62 However, the same study
showed that, in those female adolescents with depression and anxiety disorders,
the posterior asymmetry was reduced. These findings highlight the importance of
444 Simkin et al
looking for comorbid conditions that may offset the effects of depression and account
for inconsistencies found in posterior asymmetry in adolescents with depression. In
the same study, female adolescents with depression alone did not show the same
anterior alpha asymmetry found in adults with depression. These findings may be
caused by adolescent frontal regions developing later in life and the findings therefore
not correlating with findings is seen in adults with depression. These facts may indi-
cate that, in female adolescents, posterior alpha asymmetry may be a more predictive
way of detecting depression than anterior asymmetry. This study highlights the impor-
tance of using qEEG findings that match the symptoms of individual patients based on
their developmental age, especially when attempting to use alpha asymmetry as a
target of NF treatment in adolescents.
qEEG may also be used to estimate the risk for developing a mood disorder.63
Baseline electroencephalographic activity was recorded from adolescents from
12 to 14 years old whose mothers had a history of depression (high-risk group)
and whose mothers were lifetime free of axis I mental disorder (low-risk group).
High-risk adolescents showed the hypothesized pattern of relative left frontal hy-
poactivity on alpha-band measures. These findings may indicate that individuals
with left frontal hypoactivation should be followed closely for the development of
depression.
Another potentially useful qEEG parameter is cordance, which combines absolute
power (the amount of power in a frequency band or wavelength at a given electrode)
and relative power (the percentage of power contained in a frequency band relative to
the total spectrum) normalized across electrode sites and frequency bands. Cordance
has a stronger association with regional cerebral blood flow than other EEG measures.
Hunter and colleagues64 showed that cordance can be used to predict antidepressant
treatment response or remission with 70% or greater accuracy. Changes in prefrontal
theta-band cordance within the first week of medication treatment predicted 8-week
treatment outcome in 27 adult subjects. Early change in theta-band cordance in those
treated with either serotonin selective reuptake inhibitors or mixed serotonin-
norepinephrine agents predicted a good response to medication.64 Similar findings
have been reported for the treatment of bipolar depression in adults.65 There are no
qEEG studies predicting antidepressant treatment effectiveness in children or
adolescents.
Although many studies in adults have effectively treated depression in adults by tar-
geting alpha asymmetry, only 1 published study has been conducted using the alpha
asymmetry model in a depressed adolescent. The adolescent had nonbipolar major
depression and had not responded to psychotherapy. Surface EEG biofeedback
was used until the alpha asymmetry changes occurred, which appeared normal for
a nondepressed individual. This treatment required 67 sessions and the patient did
not require medication.66 More research is needed.
This is the first and, at present, only reported case of NF treatment of depression in
an adolescent, and there are no reported cases in children. There are no case reports
in youth regarding the qEEG prediction of response to treatment with psychiatric med-
ications. However, based on findings on the treatment of adults with depression and of
adults with bipolar disorder, this area needs further development in youth.
Neurofeedback Treatment of Anxiety Disorders

Available qEEG studies suggest a high incidence of abnormalities in adults with anx-
iety, panic, and obsessive-compulsive disorder (OCD).16,67–70
Two subtypes of (OCD) in adults have been described. One subgroup had excess
alpha throughout most of the brain along with excess beta in the frontal, central,
and midtemporal regions. The other group had a theta excess mostly in the frontal
areas and at posterior temporal electrodes.71
Youth (aged 10–14 years) with anxiety who responded to 19 channel surface NF
based on EEG characteristics had significant increases in the ratio of amplitudes of
alpha and theta rhythms, sensorimotor and theta rhythms, as well as the modal fre-
quency of alpha rhythm.72 The study was small (n 5 7 in experimental group and
n 5 10 in control group). The groups were divided into those with high levels of anxiety
versus low levels of anxiety based on the Prikhojan questionnaire, the Spielberger-
Khanin test, and the House-Tree-Person projective test. Between 10 and 12 NF ses-
sions were performed. Feedback was used with acoustic sounds when eyes were
closed and with visual stimuli when eyes were opened. Different feedback protocols
were used. When patients tried to control the loudness of white noise (with eyes
closed) or the intensity of colors in pictures (with eyes opened), the NF sessions
were more effective. In the experimental group, estimates of the level of anxiety
decreased in all scales of the psychological tests but were not significant. Significant
decreases in the experimental group were observed in the scales of feelings of inferi-
ority and frustration. In both groups, before and after reports by parents indicated de-
creases in emotional instability. Also, in the control group, significant decreases on the
scale of feelings of inferiority were observed. The study is small and more research is
needed.
In OCD, adults whose qEEG showed increased alpha relative power responded
positively to serotonergic antidepressants (82% response rate), whereas 80% of
adults with increased theta relative power (especially in the frontal area) failed to
improve. Responders to placebo showed increased prefrontal cordance and medica-
tion responders showed decreased cordance within 48 hours of treatment.73,74 No
studies of response to antidepressant medication using qEEG have been done on chil-
dren and adolescents with OCD or depression.
In summary, qEEGs may help to predict response to medication treatment in
depression and OCD, but none of these studies have been done in children and ado-
lescents. One study using NF in adolescents with anxiety has been published with fair
to good results. Quantitative EEG in adolescents with depression and anxiety may
indicate that those with anxiety may be protective against developing more severe
symptoms of depression.
Neurofeedback Treatment of Substance Abuse

The effects of drug abuse on a qEEG vary depending on the substance and the dura-
tion, acute versus chronic, use or abuse. There seems to be a consensus that chronic
cannabis and cocaine use are associated with an increase in alpha brainwave activ-
ity75 and that chronic alcohol abuse is associated with an increase in beta brainwave
activity.76 In addition, several research studies show that, in both alcoholics and
cocaine addicts, the best predictor of relapse is the excessive amount of fast beta
brainwave activity.75,76 EEG investigations of the children of alcoholics have docu-
mented that they also have an excess of fast beta activity, and this predicts a risk
of developing alcohol abuse in adulthood. Alcoholics and their children frequently
have lower levels of alpha and theta brainwaves, which similarly predict a risk for
developing alcohol abuse. Following the intake of alcohol, alcoholics and/or their chil-
dren feel more relaxed and show an increase in the levels of alpha and theta brain-
waves, suggesting self-medication of symptoms.10,75,76
NF has been used to train adult alcoholics to promote stress reduction and achieve
profoundly relaxed states by increasing alpha and theta brainwaves and decreasing
fast beta brainwaves77 when the patients’ eyes were closed. Increasing theta
446 Simkin et al
brainwaves with eyes closed was felt to produce a relaxed state, which is not to be
confused with theta brainwaves that are produced in the eyes-opened state (indicating
the inattention state) as is seen in ADHD. This same study showed promising potential
as an adjunct to alcoholism treatment. In a 4-year follow-up in which only 20% of the
traditionally treated group of alcoholics remained sober, 80% of the subjects who
had received NF training showed sustained reductions in alcohol use. Furthermore,
these subjects showed improvement in psychological adjustment on 13 scales of
the Millon Clinical Multiaxial Inventory, compared with the traditionally treated alco-
holics who improved on only 2 scales and became worse on 1 scale.10,77 This single
study suggests that NF may have some potential to improve treatment effectiveness
in adults with chronic alcohol use. There have been no studies using alpha/theta
training in adolescents with a history of substance abuse, but, speculatively, NF might
in the future have a role in preventing or treating alcohol abuse in youth.
Neurofeedback Treatment of Substance Use Disorders in the Context of Comorbid

ADHD
Although there have not been any studies using NF for the treatment of substance
abuse disorders in adolescents, qEEG research has discovered a difference in individ-
uals who started using cocaine chronically before age 20 years as opposed to those
who became chronic users after age 20 years.78 For example, there was significantly
more theta excess in adolescents who began abusing cocaine before age 20 years
compared with those whose abuse began after age 20 years. A significantly larger pro-
portion of early cocaine users had ADHD. It has been known for some time that many
youth with ADHD have increased theta.79,80 Therefore, ADHD may be a risk factor for
the development of cocaine abuse during adolescents and using NF to suppress theta
in these youth may decrease the risk of developing chronic cocaine abuse.
This example shows how treating ADHD by suppressing theta may be helpful in
reducing the risk of developing cocaine abuse in youth. However, some protocols
used to treat ADHD and a comorbid substance abuse may increase the risk of using
specific substances of abuse. To understand why the risk may be increased, a brief
summary of surface NF protocols is needed. Lubar and colleagues79,80 were the first
to introduce these protocols. In individuals with combined type ADHD, either
increasing sensory motor rhythms (SMR) and suppressing theta (at electrodes C3
and C4 and linked earlobes) or increasing SMR and suppressing beta 2 (C4 and linked
earlobes followed by the first protocol) has led to positive results. SMR involves
strengthening sensory motor inhibition in the cortex. For the inattentive type of
ADHD, Clarke and colleagues81 described a protocol involving theta suppression
and beta 1 enhancement, which was done at electrode Cz with linked ears, at FCz-
PCz with single ear reference or at Cz-Pz with single ear reference.82
If an adolescent has alcohol abuse (decreased theta and alpha and increased beta)
and ADHD combined type, increasing theta (although helpful in some alcoholics) may
cause the ADHD to worsen. In this case, if the adolescent has increased beta at base-
line, the first half of the Lubar protocol (suppressing theta and increasing SMR) may be
of more benefit because decreasing beta may help reduce alcohol abuse and the symp-
toms of ADHD. Suppressing theta and increasing SMR may also be advantageous in
ADHD youth who are abusing cocaine because both of these conditions increase theta.
It has also been shown that individuals who abuse both cocaine and marijuana have
excess alpha activity.75,83 In this case, any protocol in youth with comorbid ADHD and
marijuana and cocaine abuse that increases alpha activity may not be preferred.
Instead, decreasing theta, which reduces ADHD symptoms and decreases the abuse
of cocaine, may be preferred.
Because increases in beta activity seem to predict the risk of developing alcohol
abuse in children of alcoholics75,76 and can increase the risk of relapse in cocaine
and alcohol abusers, any protocol used in youth engaged in alcohol and cocaine
abuse with comorbid ADHD that increases beta activity (as suggested in Clarke and
colleagues81 NF protocol) should be used with caution. The previously mentioned lim-
itations to surface NF in youth with comorbid substance abuse and ADHD are spec-
ulative and baseline and subsequent qEEGs performed during NF along with
correlation to clinical symptoms and outcomes should guide the clinician in treatment
protocols used in youth with comorbid ADHD and substance abuse. These suggested
protocols are summarized in Table 1.
Table 1
Proposed surface NF treatment of comorbid ADHD and substance abuse using qEEG findings
Findings in Chosen Substance of Abuse ADHD Protocol ADD Protocol

Alcoholics: decreased theta and alpha Decrease beta 2 and Decrease beta 1 and
and increased beta increase SMR and increase alpha
increase alpha
Cocaine users in whom increased theta is Decrease beta 2, Decrease beta 1 and
found and the predictor of cocaine increase SMR and decrease theta
relapse may be caused by increased decrease theta, or
beta and increased theta in those who increase SMR
started using before age 20 y
Cocaine and marijuana users with Avoid increasing alpha Avoid increasing alpha
increased alpha and decrease beta 2 and decrease beta 1
Children of alcoholics with increased beta Decrease beta 2 and Decrease beta 1 and
and decreased theta and alpha increase SMR increase SMR
Abbreviation: ADD, attention deficit disorder.
Surface Neurofeedback Treatment of Posttraumatic Stress Disorder

The first study of posttraumatic stress disorder (PTSD) in adults examined surface NF
as an adjunct to conventional treatment of Vietnam combat veterans in a Veterans
Administration hospital.84 To target PTSD symptoms, a sequence of thirty 30-minute
sessions of alpha-theta NF training was provided to 15 veterans, and a contrast group
of 14 veterans received treatment as usual. No attempt was made to provide a sham
or alternative adjunctive treatment of the contrast group. At the end of the study,
among the patients receiving traditional treatment and medication (n 5 14), only 1 pa-
tient decreased medication needs, 2 reported no change, and 10 required an increase
in psychiatric medications. On the Minnesota Multiphasic Personality Inventory, pa-
tients receiving NF training improved significantly on all 10 clinical scales, whereas
the traditionally treated patients showed no significant improvements on any of the
scales. At follow-up 30 months after treatment, all 14 patients in traditional treatment
had relapsed and were rehospitalized, whereas only 3 of 15 patients having NF treat-
ment had relapsed. All 14 of the patients having NF treatment receiving medication
were able to be managed on less psychiatric medication at follow-up.
The only surface NF study of PTSD symptoms in youth examined 26 adopted chil-
dren, aged 6 to 15 years, with histories of abuse and/or neglect.85 Most of the subjects
showed symptoms of reactive attachment disorder and were taking a variety of psy-
chiatric medications before and during the NF trial. All of the children had increased
theta waves in at least one frontal site, and most had decreased delta waves in frontal
regions. NF training consisted of thirty 30-minute sessions involving auditory and
448 Simkin et al
visual feedback, with feedback initially contingent on reduction of delta/theta activity.

Before-and-after Child Behavior Checklists were completed by the adoptive parents.
Before-and-after Test of Variable Assessment (TOVA) evaluation was also done. All
but 3 subjects completed a posttreatment TOVA evaluation. Total syndrome scale
scores decreased an average of 23.05 points (standard deviation 5 21.44) with a
95% CI of 12.73 to 33.39, and were significant at t(18) 5 4.69, P<.001. Six of the 8
CBCL syndrome scale scores significantly improved (medium effect sizes [ESs]
d>0.55), although only nonsignificant improvements were observed in somatic com-
plaints and withdrawn scale scores. The TOVA scores showed significant improve-
ments on omission errors, commission errors, and total variability (d>0.60).
Limitations of the study include the small and nonrandomized sample, lack of control
group, and failure to control for potential effects of medications and other concurrent
therapies. This single study provides encouragement for further NF studies in abused
or neglected children with PTSD.
NEUROFEEDBACK TREATMENT OF INTRAUTERINE BRAIN DAMAGE, MILD HEAD

INJURY, CONCUSSION, AND TBI
Numerous qEEG studies of severe (Glasgow Coma Scale 4–8) and moderate head
injury (Glasgow Coma Scale 9–12) on patients aged 14 years and older have shown
that such injuries produce persistent qEEG abnormalities. These qEEG abnormalities
may include increased theta and decreased alpha power, decreased coherence, and
increased asymmetry.16,86
Regarding treatment of TBI, a recent research review by Thornton and Carmody87
suggests that qEEG-guided NF is superior to both neurocognitive rehabilitation stra-
tegies and medication treatment in the rehabilitation of TBI in adults, especially for
auditory memory. There have been no studies of this type published with regard to
children and adolescents.
Only 1 study in children was conducted that examined EEG NF versus physical/
occupational therapy in 12 children (aged 7 months to 14 years) who had had a stroke
before birth. The mothers had no history of drug use, hypertension, or viral infection.
All children were evaluated for baseline range of motion, and the older children pro-
vided self-descriptions of mood and completed a short-term memory test (how these
tests were administered was not available for this publication). Six children were
selected randomly for EEG NF, for 30 minutes a week, whereas 6 children received
physical or occupational therapy once a week. The EEG NF targeted brain regions
T4, C4, T3, and C3 to inhibit theta waves (4–7 Hz) and reduce their voltage, and to pro-
duce beta waves (15–18 Hz) for 0.5 seconds at 1 mV. After 3 months of treatment, chil-
dren receiving EEG NF had improved range of motion, improvement in concentration
(or, in young children, eye tracking), improved short-term memory, and fewer mood
swings. The control group showed only improved range of motion, but no improve-
ment in cognitive or emotional status.88 This lone report suggests that NF might be
useful for treating a variety of effects of intrauterine stroke, although further studies
are needed.
With regard to EEG indicators of postconcussion syndrome in adults, there is also a
broad consensus that common EEG indicators include increased focal or diffuse
theta, decreased alpha, decreased coherence, and increased asymmetry, similar to
the changes observed in TBIs in general. Multiple reports suggest that these qEEG
variables also successfully separate patients with a history of mild to moderate
head injury, even years after apparent clinical recovery, from normal individuals.21
Compared with postconcussion syndrome, TBI seems to show increases in slow
frequencies (delta, theta, slow alpha 8–10 Hz), and decreases in fast frequency alpha
(10.5–13.5 Hz), and increases in beta levels.89
A study of patients (aged 18–65 years) with TBI (and who had a Glasgow Coma
Scale of 9 or higher) examined qEEG 1 to 3 days after the TBI. Subjects were screened
to determine whether they were carriers of the apolipoprotein E (ApoE), whose E4
allele is thought to decrease blood cerebral blood flow. ApoE4 carriers had fewer
alpha and beta waves and more delta and theta waves compared with noncarriers
and controls. This finding suggests that ApoE carriers may show more abnormal
qEEG effects in the early stages of TBI, which supports the Buckholtz and Meyer-
Lindenberg53 theories that different alleles of certain genes are more likely to be asso-
ciated with connectivity dysfunction.90 These findings may open opportunities for
research to determine whether using NF to target these abnormalities (in ApoE carriers
after TBI) would be a complementary treatment of those who regain consciousness.
Many cognitive functions are affected by TBI, and studies indicate that some of
these same cognitive functions are impaired in ADHD.91 MRI and qEEG studies sug-
gest that the same brain regions are dysfunctional in both conditions.92 Gerring and
colleagues93 found that children who had thalamic injuries had a 3.6 times higher risk
for developing ADHD. The same logistic regression models used in Gerring and col-
leagues’93 analysis also showed a 3.15 times higher risk for developing ADHD if in-
juries occurred in the basal ganglia. Using variable-resolution electromagnetic
tomography qEEG data of children with ADHD, the same study found that excess
theta seemed to be generated from the septal-hippocampal pathway of the basal
ganglia, whereas excess alpha derived from the thalamus.94 These studies empha-
sized the possible correlation between TBI-acquired ADHD and the underlying
mechanisms found in ADHD. It may be that some of the NF protocols used in the
treatment of ADHD are similar to protocols used to improve cognitive function in chil-
dren with TBI.
Although Down syndrome is not classified as a brain injury disorder, some of the
cognitive functions found in TBI and ADHD are also found in these children. In a study
of 8 medication-free children (aged 6–14 years) with Down syndrome, qEEGs gener-
ally showed excess delta and theta EEG patterns. All children displayed limited vocab-
ulary (5–10 words), poor attention and concentration, weak memory, impulsivity, and
behavior problems. All 7 children who completed 60 sessions of surface NF training
showed significant (P<.02) improvement in all areas, as evaluated by questionnaire
and parent interviewing, and beneficial changes were found in qEEGs.95
The overlap of qEEG findings in ADHD, TBI, and mental retardation was further
explored in 23 subjects (7–16 years old) with mild to moderate mental retardation.
Most subjects showed increased theta, increased alpha, and coherence abnormal-
ities, and a few showed increased delta over the cortex. Of 23 patients who received
NF training, 22 showed clinical improvement according to the Developmental
Behavior Checklist-Parent scores, and patients showed significant improvement on
the Weschler Intelligence Scale for Children-Revised (WISC-R) and the TOVA. The
WISCV-R was repeated at 6 months’ follow-up after the completion of NF treatment.
The mean verbal intelligence quotient (IQ) score was 49.6 before treatment and 53.05
after treatment (P<.0295). Performance IQ scores were 55.45 before treatment and
63.20 after treatment (P<.0007). Full scale mean was 49.0 before treatment and
54.6 after treatment (P<.0003). The mean and standard deviation of the number of ses-
sions were 134.78 and 7.874 respectively. The range of the sessions was from 80 to
200. There were significant TOVA improvements and significant decreases in impul-
sivity, temper tantrums, problems with sleep, telling lies, attention, and distractibility.
All improvements in behavior and attention that had been observed during the
450 Simkin et al
treatment remained stable in the 2 years of follow-up according to parents’ reports.96

This study indicates that the use of qEEG may have some indication for increasing IQ
and targeting disruptive behaviors.
These studies suggest that there are similarities in some clinical features and qEEG
findings in TBI, postconcussive symptoms, ADHD, Down syndrome, and other forms
of mental retardation. There is some promising evidence that NF may be helpful in
treating these various conditions using generally similar NF protocols. However, the
research is limited, the qEEG for each individual needs to be carefully evaluated for
potential NF targets, and treatment needs to be matched with the symptoms in order
to develop appropriate qEEG protocols.96
Looking to the future, the United States Army is implementing LORETA z-score
biofeedback as a clinical treatment of active duty military personnel involved in an
extensive rehabilitation program for treating PTSD/TBI at the Fort Campbell Warrior
Resiliency and Recovery Center, headed by Dr Marc Zola in collaboration with Drs
Joel Lubar and Robert Thatcher. The areas targeted include symptoms associated
with concussion, posttraumatic stress, depression, chronic pain, headache and sub-
stance abuse disorder. A qEEG assessment is used to rank order the most deviant
nodes and hubs of the networks most commonly associated with these symptoms
and periodic follow-up qEEG analyses are used to assess the progress of treatment.
The US Army program is just getting started and is scheduled to continue for the next
5 years, during which extensive pretreatment versus posttreatment assessments and
statistical analyses will be conducted. Most of the patients will be between 18 and
25 years of age, and preliminary findings suggest promising results in as little as 12
sessions.
Neurofeedback in Reducing Performance Anxiety

Alpha/theta NF training was used in 2 studies involving musical performance. The ef-
fects of alpha/theta training were the same in both studies. In both studies, NF training
occurred over 10 sessions and involved alpha/theta training with eyes closed and
auditory feedback intended to increase theta more than alpha waves. In a replication
study, students were randomized to one of 6 interventions, including 3 different NF
techniques and 3 other techniques: alpha/theta NF training, beta training (intended
to reduce fast waves at the beta frequencies), or SMR NF (which involved strength-
ening sensorimotor inhibition in the cortex and inhibiting alpha frequencies), and
mental skills performance, aerobic fitness, and the Alexander technique (postural
retraining and somatic stress reduction). The results provided confirmation of the
beneficial effects of alpha/theta NF, which was the only intervention that improved mu-
sic performance (overall quality, musical understanding, stylistic quality, and interpre-
tive imagination). All 6 interventions were successful in reducing preperformance
anxiety, so the performance enhancement by alpha-theta training could not be attrib-
uted to anxiety reduction alone.97
Alpha/theta training has been used with university dancers who compete in ball-
room dancing. Twenty-four male/female pairs were randomly assigned to either
alpha/theta training, heart rate coherence (HRC) training (involving heart rate variability
training), or a nonintervention control group before and after performance was judged
by 2 dance experts who were blinded to which groups got the intervention. Both the
HRC and alpha/theta training groups improved on the overall rating of execution more
than the control group. Subscale ratings revealed that HRC improved technique, but
alpha/theta training improved timing.
As noted by the investigators, theta oscillations are speculated to affect meditative
concentration, and reduce anxiety and sympathetic autonomic activation. Theta
oscillations may also affect virtual spatial navigation, focused and sustained attention,
and working and recognition memory.
In summary, it is thought that alpha-theta training targets optimal performance and
enhancement of technical, communication, and artistic domains of performance in the
arts.
Does Neurofeedback Technique Have to be Modified for Developmental Age?

Recent research98,99 has shown that the major brain networks develop early in life and
are essentially the same in adolescents and adults. One study99 used measures of
network topology to investigate the development of functional hubs in 99 normal sub-
jects, aged 10 to 20 years. Hub architecture was evident in late childhood (age 10 years)
and was stable from adolescence to early adulthood, but the strength of connectivity
changed with development. From childhood to adolescence, the strength of connec-
tions increased between frontal hubs and frontal, parietal, temporal, and cerebellar
spoke regions, whereas connections decreased in the posterior part of the brain. It
was hypothesized that this pattern reflects the gradual maturation of the ability of
the frontal lobe to coordinate distributed cortical functions for goal-directed behaviors.
The subsequent developmental increases in hub-spoke connectivity between adoles-
cence and adulthood were fewer and tended to be more posterior, involving connec-
tions between subcortical hubs (the putamen and cerebellum) and frontal, occipital,
and temporal spoke regions. These findings suggest that developmentally stable func-
tional hub architecture provides the foundation of information flow in the brain,
whereas connections between hubs and spokes continue to develop, presumably
supporting the maturation of cognitive functions. Based on these findings, it is ex-
pected that NF would work in a manner that is identical from age 10 years to adulthood.
Controversies
Although there is extensive scientific evidence to validate the use of qEEG, there has
been limited acceptance in the United States. There are several reasons for this. First,
psychiatrists do not often follow the EEG literature and are unaware of developments
in the field. Second, because of the evolving understanding about technology and
appropriate protocols, some falsely negative findings have been reported16,100,101
and adopted by some expert panels102 but rejected by others.103 More recent studies
provide substantial additional support for the validity and clinical use of qEEG in
several areas of psychiatry.16,104 The American Academy of Pediatrics has recently
listed EEG biofeedback as a level 1, best-support (a valid first-line) intervention for
ADHD (www.aap.org/en-us/advocacy.../aap.../CRPsychosocialInterventions.pdf).
Questions continue to be raised concerning the available normative databases,
but current methodological adjustments and mathematical techniques allow cross-
validation among different normative databases that have high intradatabase
similarity.12 Confidence in normative databases is based on widespread independent
replications,16 test-retest reliability confirmed in short-term and long-term follow-up
large samples,105,106 and small gender differences.
Some critics have argued that NF should not be an accepted intervention until
double-blind placebo-controlled studies using sham controls have been conducted.
There has been considerable debate and controversy in the NF literature about the pit-
falls of attempted sham-NF controls, what technical innovations can be introduced to
provide better controls, the validity of proceeding with randomized controls in the
absence of an acceptable sham procedure, and the use of altered qEEG patterns in
an individual as corroboration of the validity of the intervention (eg, as is often used
with other interventions).
452 Simkin et al
This issue of whether a sham is as effective, more effective, or less effective than
contingent reinforcement may involve a serious misconception. Both experimental
and sham groups are operating under the contingencies of classical and operant con-
ditioning. Imagine a child with ADHD who already spends a lot of time playing com-
puter games and is now placed in front of a display that purportedly is going to help
but he is in the sham group. The child becomes engaged by the display, tries to
make it occur more frequently, and in the process is releasing dopamine in both the
dorsal and ventral attention networks and from dopamine-producing regions within
the brain such as the nucleus accumbens and related areas. The display elicits an un-
conditioned response of increasing EEG activation in many areas including the one in
which the sensor has been placed, such as FZ or CZ, resulting in more beta and less
theta. Sometimes when a burst of beta activity associated with decreased theta activ-
ity occurs at the location where the sensor is placed, reinforcement is delivered. Now
the child is experiencing partial noncontingent reinforcement on a variable ratio
schedule. This same type of reinforcement, used in the gambling industry, is the
most powerful reason why casinos make billions of dollars. As a result, the child in
the sham conditioning group often has an activated EEG, tries hard to get the noncon-
tingent reinforcement, which is sometimes contingent, and shows improvements in
several ADHD indices such as rating scales and perhaps even academic performance.
So what is the answer? It has been repeatedly stated that, because an operant con-
ditioning of a particular EEG pattern reinforcement is usually delivered after 0.5 to
1 second of the production of that pattern, the EEG recording can be marked every
time reinforcement is delivered in both the experimental and sham control group. In
the sham group, a simple correlation can be run between the percentage time that
the reinforcement was contingent and the degree to which the measured EEG param-
eters changed in the desired direction along with all of the appropriate before and after
measures. It would not be surprising that, even if there was a 20% contingency in the
sham group, powerful learning effects might occur. As a result it is impossible to
develop a sham in which reinforcement never coincides with the EEG pattern that is
being trained in the experimental group.
Until a generally agreed on placebo can be developed, continuing active random-
ized controlled studies should be encouraged.
Furthermore, the lack of understanding of NF has led to several other problems
regarding research. For instance, some studies that have questioned the effectiveness
of NF have been poorly done for many reasons.
First, some studies have been done that have little to support the role of a
feasible sham. One reason has to do with the use of different NF protocols that
are not recognized as proven and the use of technology that has little proven effec-
tiveness. In a study that was done to show the effectiveness of a double-blind,
sham-controlled, randomized pilot feasibility trial in patients with ADHD, there
were no significant differences between the NF group and the sham group. How-
ever, the investigators appropriately admitted that they did not use NF technology
that is considered effective.107 The equipment used was Smart Brain by Cybert
Learning Technology (www.smarttech.com). This technology allows altering the
reinforcement threshold from minute to minute, adopting the threshold to just
completed performance, and not requiring focus on the NF training. This technique
is referred to as autothresholding. If a patient is not paying attention, the threshold-
ing may be adjusted to a nonlearning event and the reinforcement may not produce
the desired effect.
Second, some studies were done without evaluating pre-qEEG data. When partic-
ipants are selected based on their pre-qEEG data, the protocol used would be better
matched to the participant and the use of this protocol would more likely produce
a larger ES (an ES is considered large when it is 0.8 or greater). An ES of 0.8 means
that the mean of the treated group is at the 79th percentile of the untreated group.
As noted in an article by Sherlin and colleagues,108 when pre-qEEG data are evaluated
in, for instance, a study involving preselection of participants based on abnormal
theta/beta ratios for ADHD, the ES for inattention was 2.22 and for hyperactivity
was 1.2.109 These ESs are substantially larger than the ES calculated from a meta-
analysis of NF by Arns and colleagues,110 in which the ES was 0.81 for inattention
and 0.4/0.69 for impulsivity/hyperactivity. This later meta-analysis had similar ES dis-
cussed in an article by Faraone and Buitelaar,111 which was comparable with medica-
tion for inattention (0.84) but not for the ES for medication for impulsivity/hyperactivity
(1.01). Therefore, NF must be based on pre-qEEG data and, in this case, may improve
therapeutic outcomes. If children are very hyperactive and unable to be still during NF
and a pre-qEEG analysis has occurred, it is suggested that they may require medica-
tion (if they are good responders to stimulant medication) initially in order to have good
control of their disorder. This can be gradually tapered as the NF proceeds. If the
hyperactivity does not subside, but the attention improves, they may still require medi-
cation. More research is needed.
Third, meta-analysis of NF studies have concluded that the ES of NF was small
because studies were included that did not included standard NF protocols, which
again points to the lack of knowledge surrounding NF. In a meta-analysis of non-
pharmacologic interventions for ADHD by Sonuga-Barker and colleagues112 Arns
and Strehl113 pointed out that, when statistics were recalculated using only studies
with standard NF protocols, parent ratings obtained a significant standardized dif-
ference of 0.58 (95% CI 5 0.12–0.94; z 5 3.52; P 5 .0004) and for teacher ratings
a significant standardized mean difference was found of 0.39 (95% CI 5 0.07;
z 5 2.39; P 5 .02). The inclusion of nonstandardized protocols in a meta-
analysis points to a problem in NF: the lack of the standardized protocols that
are critical to future research. In addition, for every disorder, best-practice
approaches should be individualized and the same before and after instruments
should be uniformly used. These instruments should not be limited to rating scales
from multiple sources and may include neuropsychological instruments, such as
the TOVA (http://www.tovatest.com/), Integrated Visual and Auditory Continuous
Performance Test (IVA-Plus) (http://www.braintrain.com/ivaplus/), and the Behavior
Rating Inventory of Executive Function-R (BRIEF-R) (http://www4.parinc.com/
Products/Product.aspx?ProductID5BRIEF#).
Fourth, with regard to lack of knowledge about standardized protocols, some
studies, although well intentioned, do not use well-known facts about learning
curves and operant conditioning. For instance, in randomized placebo-controlled
trial of NF for children with ADHD done by van Dongen-Boomsma and col-
leagues,114 the percentage of reward used was too high (80%) for operant con-
ditioning to occur. If the reward is too easy (rewards are delivered every time)
no learning occurs.
However, some users of NF are not trained professionals. Given the evolution of
NF, only clinicians who are well trained in neurophysiology, neuroanatomy, and
neuroendocrine and metabolic processes should be allowed to do NF. Given the
need to remove artifact appropriately in order to use qEEG, only clinicians who
are trained to do so should be allowed to extract the artifacts, or a requirement
that an untrained clinician must be supervised by someone who is proficient in
reading EEGs may be necessary. NF is most appropriately used as an adjunct
to a multimodal biological, psychological, and social treatment approach. These
454 Simkin et al
points may indicate that the accreditation used in NF may need to be

reconsidered.
New mathematical techniques for qEEG analysis are developing rapidly, and
new technology will promote this development. The advent of LORETA and
z-score NF allows clinicians to selectively target data that are substantial devia-
tions above or below the normative population means, enabling more accurate
matching of those deviations to symptoms and to the relevant brain regions.
LORETA (19 electrodes) allows targeting of deeper structures than can be reached
with surface NF (2–4 electrodes), will likely reduce the number of sessions and
costs required to get effective results, and seems to have long-lasting effects after
treatment is stopped.
Summary of Tables
In evaluating this limited literature on NF treatment in children and adolescents, the
system developed by United States Preventive Services Task Force (USPSTF) can
be used.
Table 2 provides a summary of the quality of evidence that is currently available in
the published literature as well as a summary of the strength of recommendations that
can be offered to clinicians concerning the value of the treatment. Table 3 is similar,
except that the right-hand column provides the authors’ clinical expert opinions rather
than being based purely on publications.
Table 2
Evaluation of surface NF in youth: the evidence base
Quality of Strength of
Indications NF Method Research Recommendation Evidence Base
Depression Alpha asymmetry Fair Recommend 1 case study66
Anxiety Increase alpha/theta Fair Recommend 1 RCT72
ratio or SMR/theta
ratio
Performance Alpha/theta training Good Recommend 1 RCT and 1 open
anxiety trial97,115
Mental NF based on individual Good Recommend 2 clinical trials
retardation qEEG
TBI Decrease theta/increase Fair Recommend 1 RCT comparing
beta physical therapy
with NF
PTSD Decrease delta/theta Fair Recommend Non–randomized
activity controlled study
Alcohol abuse Alpha/theta training ND ND ND
Substance abuse Alpha/theta training ND ND ND
USPSTF quality of evidence grade is a qualitative ranking of the strength of the published evidence
in the medical literature. Limited, indirect evidence; good, consistent benefit in well-conducted
studies in different populations; fair, data show positive effects, but weak, limited, or indirect ev-
idence; poor, cannot show benefit because of data weakness.
USPSTF strength of recommendations: A, recommend strongly (good evidence of benefit and
safety); B, recommend (fair evidence of benefit and of safety); C, neutral (fair evidence for, but
seems risky); D, recommend against (fair evidence of ineffectiveness or harm); I, insufficient
data.
Abbreviations: ND, no data on youth; RCT, randomized controlled trial.
Table 3
Evaluation of surface NF in youth: authors’ recommendations and personal comments
Strength of
Recommendations
Based on Authors’ Clinical
Indications Published Data Recommendations Authors’ Other Comments
Depression Recommend Possible benefit Requires further study given
the variance between
findings in frontal lobes in
alpha asymmetry
adolescents vs adults. qEEG
should always be used
before and during
treatment with regard to
clinical symptoms.
Studies needed using
LORETA NF
Anxiety Recommend Shows promise and May have greater effect if
improvement based on qEEG. Studies
trends needed using LORETA NF
Performance Recommend Shows promise and Promising for decreasing
anxiety improvement performance anxiety and
trends performance skill.
LORETA NF
Mental Recommend Shows promise and Promising for possible
retardation possible benefit increase in IQ and
based on decreasing disruptive
pretreatment and behaviors and improving
ongoing treatment attention. May help to
on qEEGs reduce medications used to
treat these behaviors.
LORETA NF
TBI Recommend Shows promise May help to increase range
of motion and memory in
persons who have had a
stroke. Need to use qEEG
and watch to see whether
increasing beta will make
symptoms worse. Studies
needed using LORETA NF
PTSD Recommend Recommend Helpful in decreasing PTSD
symptoms and improving
attention in abused
children. Studies needed
using LORETA NF
Alcohol and Neutral with See Table 1 for Studies needed using
substance abuse comorbid ADHD recommendations LORETA NF
USPSTF strength of recommendations: A, recommend strongly (good evidence of benefit and

safety); B, recommend (fair evidence of benefit and of safety); C, neutral (fair evidence for,
but seems risky); D, recommend against (fair evidence of ineffectiveness or harm); I, insufficient
data.
456 Simkin et al
SUMMARY
Although most NF research has targeted ADHD, there are some limited studies in chil-
dren and adolescents suggesting possible applications for treatment of anxiety,
depressive disorder, addiction (with or without ADHD), brain injury (including TBI,
mild head injury, concussion, and intrauterine brain damage), Down syndrome, and
perhaps some other forms of mental retardation in children and adolescents. There
are additional data suggesting that NF might also be useful for treating certain intrac-
table seizure disorders in youth,47,116,117 but a neurologist should be involved in man-
aging those cases. Although some of the data remain controversial, the American
Academy of Pediatrics has recently endorsed NF as a valid and possibly first-line
treatment of ADHD in youth, and the United States Army is implementing an innovative
program to examine the role of NF in treating TBI in veterans. As the technology and
statistical analytical methods advance, many of the remaining uncertainties can be ex-
pected to be addressed.
Research has documented that many of the same areas targeted in the NF treat-
ment of ADHD (right ventrolateral prefrontal cortex, right dorsal anterior cingulate,
left thalamus, left caudate nucleus, and left substantia nigra) are involved in mediating
selective attention and response inhibition in children with ADHD, and similar NF treat-
ment is helpful for improving attention in a variety of other psychiatric disorders.118 The
targeting of such transdiagnostic symptoms holds particular promise for the future of
NF treatment.
A small study has suggested that NF and methylphenidate similarly improve several
behavioral and cognitive functions in children with ADHD.119 This raises the question
of whether nondrug treatment might be able to eventually augment or, conceivably,
supplant some of the established current psychopharmacologic interventions. For
many families, NF might be a more acceptable option than psychostimulants for
treating ADHD. Furthermore, NF might be a particularly valid option for treating
ADHD with comorbid substance abuse, especially because NF may be beneficial
for both conditions, without the risks associated with using abusable medications
in this population.
The use of qEEG to characterize and localize brain dysfunction creates new oppor-
tunities. Current data suggest that qEEG might eventually be useful for predicting
medication responders and nonresponders,120 for identifying patients at risk for devel-
oping antidepressant-induced treatment-emergent suicidal ideation,121 and for iden-
tifying patients at particularly high risk for relapse (depression, cocaine addiction,
nicotine craving).122 In addition to identifying patients at risk by qEEG monitoring,
NF can be used to target the dysfunctional circuitry in relevant brain regions and could
conceivably act proactively to reduce the risks.
The newer technologies may eventually be found to be particularly suitable for
providing a more penetrating picture of dysfunctional circuitry and for yielding more
effective treatments with fewer NF sessions and lower costs. There was a recent pre-
sentation of a patient with multiple diagnosis starting in adolescence (including major
depression, ADHDF, bipolar I, social phobia, TBI, OCD, and PTSD), and multiple med-
ications (including sertraline, dextroamphetamine mixed salts, Synthroid, lithium
carbonate, hydroxyzine, alprazolam, and lorazepam), who had not adequately
responded to surface NF and therapy over a 2-year period. He had made 3 previous sui-
cide attempts. Once LORETA NF was used (for 33 sessions over a 6-month period), the
patient was off all medications and functioning well.123 His Global Assessment of
Functioning (GAF) score improved from 15 to 90. A single case cannot be the guide
for clinicians to use LORETA NF in every person with multiple disorders, but this case
presentation highlights the need to do more research using this intervention, especially
in patients who are not responding to conventional interventions and therapy.
Although LORETA seems to offer some promise for advances compared with the
older surface NF methods, the two approaches can currently be viewed as comple-
mentary. Some children are not able to tolerate the many electrodes required for
LORETA and may be better candidates for the more user-friendly surface NF methods
using 2 to 4 electrodes.
At present, NF can be considered when patients or parents prefer not to use psy-
chopharmacologic treatments or when medication treatments are not well tolerated
or adequately effective.
Child and adolescent psychiatrists are well aware of the importance of well-trained
clinicians who are skilled in doing biopsychosocial evaluation and treatment. NF should
never be used as a single intervention in child and adolescent psychiatry, just as med-
ications should never be the only intervention for depression. NF is properly used when
integrated into a multimodal series of interventions for the child and family that pre-
cedes and continues after the time period of active NF treatment. Although this may
be obvious to child psychiatrists, clinicians should be advised that an increasing num-
ber of lay persons are inappropriately obtaining NF equipment and offering to put elec-
trodes on someone’s head and to alter the brain functioning of individuals with serious
medical and psychological conditions. In less drastic cases, some inadequately trained
clinicians have assumed that a patient with ADHD can benefit from a predetermined
protocol to increase alpha or beta activity, without having conducted proper individu-
alized assessments. The risk, for example, in a patient who already has excess beta
and cortical irritability is the induction of tics, anxiety, or seizure activity.124
Psychiatrists who are collaborating with NF specialists are strongly encouraged to
use licensed clinicians with certification in NF (eg, from the Biofeedback Certification
Institute of America) and or clinicians supervised by clinicians with certification and/or
certification in qEEG (eg, from the EEG and Clinical Neuroscience Society, the Quan-
titative Electroencephalography Certification Board, or the Society for the Advance-
ment of Brain Analysis).
In regards to the effectiveness of neurofeedback, the reader is referred to a recent
review by Arns and colleagues125 which addresses this issue with a concise summary
of previous studies. Although double blind placebo controlled studies are clearly the
preferred type of research in regards to proving the effectiveness of treatment
regimens, until a convincing sham is found, randomized controlled studies based on
appropriate protocols and methods should continue. Meanwhile, a collaborative
group has been formed to address the issue of a credible sham.126
REFERENCES
1. Thatcher RW. Introduction. In: Robert Thatcher, editor. Handbook on quantitative

electroencephalography and EEG biofeedback. St Petersburg (FL): Anipublish-
ing; 2012. p. 10–145.
2. Brodmann K. Vergleichende Lokalisationslehre der Grosshirnrinde in ihren Prin-
zipien dargestellt auf Grund des Zellenbaues. Leipzig (Germany): Johann Am-
brosius Barth Verlag; 1909.
3. Nunez P. Neocortical dynamics and human EEG rhythms. New York: Oxford Uni-
versity Press; 1994.
4. Pascual-Marqui RD, Michel CM, Lehmann D. Low resolution electromagnetic to-
mography: a new method for localizing electrical activity in the brain. Int J Psy-
chophysiol 1994;18:49–65.
458 Simkin et al
5. Pascual-Marqui RD, Lehmann D, Koenig T, et al. Low resolution brain electro-

magnetic tomography (LORETA) functional imaging in acute, neuroleptic-
naive, first-episode, productive schizophrenia. Psychiatry Res 1999;90:169–79.
6. Hughes SW, Crunelli V. Just a phase they are going through: the complex inter-
action of intrinsic high-thresholding bursting and gap junctions in the generation
of thalamic a and 0 rhythms. Int J Psychophysiol 2007;64:3–17.
7. Tiesinga PH, Sejnowski TJ. Mechanisms for phase shifting in cortical networks
and their role in communication through coherence. Front Hum Neurosci
2010;4:196.
8. John ER. From synchronous neural discharges to subjective awareness. Prog
Brain Res 2005;150:143–71.
9. Kandel ER. The molecular biology of memory storage: a dialogue between
genes and synapses [review]. Science 2001;294(5544):1030–8.
10. Hammond DC. What is neurofeedback: an update. J Neurotherapy 2011;15:
305–36.
11. Buzsaki G. Rhythms of the brain. New York: Oxford University Press; 2006.
12. Thatcher RW, Lubar JF. History of scientific standards of qEEG normative bases.
In: Budzinsky T, Budzinsky J, Evans J, et al, editors. Introduction to qEEG and
neurofeedback: advanced theory and applications. San Diego (CA): Academic
Press; 2008.
13. Prichep LS. Use of normative databases and statistical methods in demon-
strating clinical utility of qEEG: importance and cautions. Clin EEG Neurosci
2005;36(2):82–127.
14. Thatcher RW, Walker RA, Biver C, et al. Quantitative normative databases: vali-
dation and clinical correlation. J Neurotherapy 2003;7:87–122.
15. Valdes-Sosa P, Valdes -Sosa M, Carballo J, et al. qEEG in a public health sys-
tem. Brain Topogr 1992;4(4):259–66.
16. Chabot RJ, di Michel F, Prichep L. The role of quantitative EEG in child and
adolescent psychiatric disorders. Child Adolesc Psychiatr Clin N Am 2005;14:
21–5.
17. John ER, Prichep LS, Ahn H, et al. Neurometric evaluation of cognitive dys-
functions and neurological disorders in children. Prog Neurobiol 1983;21:
239–90.
18. John ER, Prichep LS, Friedman J, et al. Neurometrics: computer assisted differ-
ential diagnosis of brain dysfunctions. Science 1988;293:162–9.
19. John ER, Prichep LS. Principles of neurometrics and neurometric analysis of
EEG and evoked potentials. In: Niedermeyer E, Lopes Da Silva F, editors.
EEG: basic principles, clinical applications and related fields. Baltimore (MD):
Williams & Wilkins; 1993. p. 989–1003.
20. Prichep LS, John ER. qEEG profiles of psychiatric disorders. Brain Topogr 1992;
4(4):249–57.
21. Thatcher RW, Walker RA, Gerson I, et al. EEG discriminant analyses of mild
head trauma. Electroencephalogr Clin Neurophysiol 1989;73:94–106.
22. Czobor P, Volavka J. Pretreatment EEG predicts short-term response to haloper-
idol treatment. Biol Psychiatry 1991;30:927–42.
23. Roemer RA, Shagass C, Dubin W, et al. Relationship between pretreatment EEG
coherence measures and subsequent response to ECT. Neuropsychobiology
1991;24:121–4.
24. Struve FA, Straumanis JJ, Patrick G. Persistent topographic quantitative EEG
sequelae of chronic marijuana use: a replication study and initial discriminant
function analysis. Clin Electroencephalogr 1994;25(2):63–75.
25. Chabot RJ, Merkin H, Wood LM, et al. Sensitivity and specificity of qEEG in chil-
dren with attention deficit or specific developmental learning disorders. Clin
Electroencephalogr 1996;27(1):26–34.
26. Suffin SC, Emory WH. Neurometric subgroups in attentional and affective disor-
ders and their association with pharmacotherapeutic outcome. Clin Electroen-
cephalogr 1995;26(2):76–83.
27. Lancaster JL, Woldorff MG, Parsons LM, et al. Automatic Talairach atlas labels
for functional brain mapping. Hum Brain Mapp 2000;10:120–31.
28. Michel CM, Koenig T, Brandeis D, et al. Electrical neuroimaging. New York:
Cambridge University Press; 2009.
29. Thatcher RW. Neuropsychiatry and quantitative EEG in the 21st century. Neuro-
psychiatry 2011;1(5):495–514.
30. Thatcher R, Wang B, Toro C, et al. Human neural network dynamics using multi-
modal registration of EEG, PET and MRI. In: Thatcher R, Hallett M, Zeffiro T, et al,
editors. Functional neuroimaging: technical foundations. New York: Academic
Press; 1994. p. 269–78.
31. Wang B, Toro C, Zeffiro T, et al. Integrating electrophysiological data with brain
images: I- Sources of event related potentials, PET and MRI. In: Thatcher R,
Hallett M, Zeffiro T, et al, editors. Functional neuroimaging: technical founda-
tions. New York: Academic Press; 1994. p. 251–6.
32. Thatcher RW, Lyon GR, Rumsey J, et al, editors. Developmental neuroimaging:
mapping the development of brain and behavior. London, UK: Academic Press,
Limited; 1996.
33. Talairach JT. Co-planar stereotaxic atlas of the brain: 3 dimensional proportional
system: an approach to cerebral imaging. New York: Thieme Medical Pub-
lishers; 1998.
34. Cannon RL. Introduction to LORETA. In: Cannon RL, editor. Low resolution brain
electromagnetic tomography (LORETA) basic concepts and clinical applica-
tions. Corpus Christi (TX): BMED Press; 2012. p. 3–6.
35. Pascual-Marqui R. Standardized low-resolution brain electromagnetic tomogra-
phy (sLORETA): technical details. Methods and Find in Exp. Clin Pharmacol
2002;24(Suppl D):5–12.
36. Bosch-Bayard J, Valdes-Sosa P, Virues-Alba T, et al. 3D statistical parametric
mapping of EEG source spectra by means of variable resolution electromag-
netic tomography (VARETA). Clin Electroencephalogr 2006;32(2):47–61.
37. Thatcher R, North D, Biver C. EEG inverse solutions and parametric vs. non-
parametric statistics of low resolution electromagnetic tomography (LORETA).
Clin EEG Neurosci 2005;36(1):1–9.
38. Thatcher RW, North D, Biver C. Evaluation and validity of a LORETA normative
EEG database. Clin EEG Neurosci 2005;36(2):116–22.
39. Hagmann P, Cammou L, Gigandet X, et al. Mapping the structural core of hu-
man cerebral cortex. PLoS Biol 2008;6:e159.
40. Chen ZJ, He Y, Rosa-Neto P, et al. Revealing modular architecture of human
brain structural networks by using cortical thickness from MRI. Cereb Cortex
2008;18:2374–81.
41. He Y, Wang J, Wang L, et al. Uncovering intrinsic modular organization of spon-
taneous brain activity in humans. PLoS One 2009;4(4):e5226. http://dx.doi.org/
10.1371/journal.pone.0005226.
42. Brodmann VK. Localization in the cerebral cortex: the principles of comparative
localization in the cerebral cortex based on cytoarchitectonics. London, UK:
Springer; 1909. Translated by L. J. Garey, 1994.
460 Simkin et al
43. Thatcher W, North DM, Biver CJ. Diffusion spectral imaging modules corre-
late with EEG LORETA neuroimaging modules. Hum Brain Mapp 2011;33:
1062–75.
44. Collura TF, Thatcher RW, Smith M, et al. EEG biofeedback training using live Z-
scores and a normative database. In: Budzinsky T, Budzinsky H, Evans J, et al,
editors. Introduction to qEEG and neurofeedback: advanced theory and appli-
cations. San Diego (CA): Academic Press; 2008. p. 123–33.
45. Thatcher RW. EEG operant conditioning (biofeedback) and traumatic brain
injury. Clin EEG Neurosci 2000;31(1):38–44.
46. Collura TF, Guan J, Tarrant J, et al. EEG biofeedback case studies using live Z-
score training and a normative database. J Neurotherapy 2010;14:22–46.
47. Hughes JR, John ER. Conventional and quantitative electroencephalography in
psychiatry. J Neuropsychiatry Clin Neurosci 1999;11(2):190–206.
48. DeBoer P, Abercrombie E. Physiological release of striatal acetylcholine in vivo:
modulation of D1 and D2 dopamine receptor subtypes. J Pharmacol Exp Ther
1996;277:775–83.
49. Icarashi Y, Takahashi H, Aral T, et al. Suppression of cholinergic activity via the
dopamine D2 receptor in the rat stratum. Neurochem Int 1997;30:191–7.
50. Russel V, de Villiers A, Sagvolden T, et al. Altered dopaminergic function
in the prefrontal cortex, nucleus accumbens and caudate-putamen of an an-
imal model of attention deficit hyperactivity disorder. Brain Res 1995;676:
343–51.
51. Levy F. The dopamine theory of attention deficit hyperactivity disorder. Aust N Z
J Psychiatry 1991;25:277–83.
52. Niedermeyer E, Naidu SB. Attention deficit disorder and frontal motor cortex
disconnection. Clin Electroencephalogr 1998;28:130–6.
53. Buckholtz JW, Meyer-Lindenberg A. Psychopathology and the human connec-
tion: toward a transdiagnostic model of risk for mental illness. Neuron 2012;
74(6):990–1004. http://dx.doi.org/10.1016/j.neuron.2012:06.002.
54. Insel T, Cuthbert B, Garvey M, et al. Research domain criteria (RDoC): toward a
new classification framework for research on mental disorders. Am J Psychiatry
2010;167:748–51.
55. Phillips ML. More severe developmental trajectory of behavioral and emotional
dysregulation in youth is associated with reduced functional integrity prefrontal
cortical-amygdala circuitry during emotional regulation. Presented at the Amer-
ican Academy of Child and Adolescent Psychiatry Meeting. Orlando (FL),
October 25, 2013.
56. Hajek T. Right inferior frontal gyrus volume and future conversion to psychiatric
disorders: Prospective study of subjects at genetic risk for bipolar. Presented at
the American Academy of Child and Adolescent Psychiatry meeting in Orlando
(FL), October 25, 2013.
57. Adolphs RJ, Tranel D. Emotion. In: Rizzo M, Eslinger PJ, editors. Principles and
practice of behavioral neurology and neuropsychology. Philadelphia: Saunders;
2004. p. 451–74.
58. Walke J, Lawson R, Kozlowski G, et al. Current status of qEEG and neurofeed-
back in the treatment of depression. In: Evans JR, editor. Handbook of neuro-
feedback. Binghampton (NY): Haworth Medical Press; 2007. p. 341–51.
59. Prichep LS, John ER, Alper K, et al. Quantitative EEG in depressive disorders.
In: Shagass C, Josiassen RC, Roemer RA, editors. Brain electrical potentials
and psychopathology. Amsterdam: Elsevier; 1986. p. 223–44.
60. Leuchter AF, Coo IA, Hunter AM, et al. Resting state quantitative electroenceph-
alography reveals neurophysiologic connectivity in depression. PLoS One 2012;
7(2):e3250.
61. Wheeler RE, Davidson RJ, Tomarken AJ. Frontal brain asymmetry and emotional
reactivity: a biological substrate of affective style. Psychophysiology 1993;30(1):
82–9.
62. Kentgen LM, Tenke CE, Pine DS, et al. Electroencephalographic asymmetries in
adolescents with major depression: influence of comorbidity with anxiety disor-
ders. J Abnorm Psychol 2000;109:797–802.
63. Tomarken AJ, Dichter GS, Garber J, et al. Resting frontal brain activity: linkages
to maternal depression and socioeconomic status among adolescents. Biol Psy-
chol 2004;67:77–102.
64. Hunter AM, Cook IA, Leuchter AF. The promise of the quantitative electroen-
cephalogram as a predictor of antidepressant treatment outcomes in major
depressive disorder. Psychiatr Clin North Am 2007;30:105–24.
65. Bares M, Nova T, Brunovsky M, et al. The change in qEEG prefrontal cordance
as a predictor to antidepressant intervention in bipolar depression. A pilot study.
J Psychiatr Res 2012;46(2):219–25.
66. Earnest C. Single case study of EEG asymmetry biofeedback for depression
and independent replication in an adolescent. J Neurotherapy 1999;3:
28–35.
67. Small JG. Psychiatric disorders and EEG. In: Niedermeyer E, Lopes Da Silva F,
editors. Electroencephalography: basic principles, clinical applications, and
related fields. Baltimore (MD): Wilkins and Wilkins; 1993. p. 581–96.
68. Jenike MA, Brotman AW. The EEG in obsessive-compulsive disorder. J Clin Psy-
chiatry 1984;45:122–4.
69. Abraham HD. Stimulants, panic, and BEAM EEG abnormalities. Am J Psychiatry
1989;146(7):947–8.
70. Buchsbaum MS, Hazlett E, Sicotte N, et al. Topographic EEG changes with
benzodiazepine administration in generalized anxiety disorder. Biol Psychiatry
1985;20:832–42.
71. Prichep LS, Mas F, Hollander E, et al. Quantitative electroencephalographic
(qEEG) subtyping of obsessive compulsive disorder. Psychiatry Res 1993;
50(1):25–32.
72. Eismont EV, Lutsyu NV, Pavlenko VB. Moderation of increased anxiety in children
and teenagers with the use of neurotherapy: estimation of the efficacy. Neuro-
physiology 2011;43(1):53–61.
73. Cook IA, Leuchter AF, Morgan M, et al. Early changes in prefrontal activity char-
acterize clinical responders to antidepressants. Neuropsychopharmacology
2002;27(1):120–31.
74. Hanse ES, Prichep LS, Bolwig TG, et al. Quantitative electroencephalography
in OCD patients treated with paroxetine. Clin Electroencephalogr 2003;34(2):
70–4.
75. Prichep L, Alper KR, Kowalik SC, et al. qEEG subtypes in crack cocaine depen-
dence and treatment outcome. In: Harris LS, editor. Problems of drug depen-
dence, 1995: proceedings of 57th Annual Scientific Meeting, The College on
Problems of Drug Dependence, Inc. Rockville (MD): National Institute on Drug
Abuse; 1996. p. 142. Research Monograph. No. 162.
76. Baue LO. Predicting relapse to alcohol and drug abuse via quantitative electro-
encephalography. Neuropsychopharmacology 2001;25:332–40.
462 Simkin et al
77. Peniston EG, Kulkosky PJ. Alcoholic personality and alpha-theta brainwave
training. Medical Psychotherapy 1991;2:37–55.
78. Prichep LS, Alper KR, Sverdlov L, et al. Outcome related electrophysiolog-
ical subtypes of cocaine dependence. Clin Electroencephalogr 2002;33(1):
8–20.
79. Lubar JO, Lubar JF. Electroencephalographic biofeedback of SMR and beta for
treatment of attention deficit disorders in a clinical setting. Biofeedback Self
Regul 1994;9:1–23.
80. Lubar JF, Swartwood MO, Swartwood JN, et al. 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 Self Regul 1995;20:83–99.
81. Clarke AR, Barry RJ, McCarthy R, et al. Excess beta activity in children with
attention-deficit hyperactivity disorder: an atypical electrophysiological group.
Psychiatry Res 2001;103:205–18.
82. Monastra VJ. Electroencephalographic biofeedback (neurotherapy) as a treat-
ment for attention deficit hyperactivity disorder: rationale and empirical founda-
tion. Child Adolesc Psychiatr Clin N Am 2005;14:55–82.
83. Struve FA, Patrick G, Straumanis JJ, et al. Possible EEG sequelae of very long
duration marijuana use: pilot findings from topographic quantitative EEG ana-
lyses of subjects with 15 to 24 years of cumulative daily exposure to THC.
Clin Electroencephalogr 1998;29(1):31–6.
84. Peniston EG, Kulkosky PJ. Alpha-theta brainwave neuro-feedback therapy for
Vietnam veterans with combat related post-traumatic stress disorder. Medical
Psychotherapy 1991;4:47–60.
85. Huang-Storms L, Bodenhamer-Davis E, Davis R, et al. qEEG-guided neurofeed-
back for children with histories of abuse and neglect: neurodevelopmental ratio-
nale and pilot study. J Neurotherapy 2006;10(4):3–16.
86. Rimel RW, Giordani B, Barth JT, et al. Moderate head injury: completing the clin-
ical spectrum of trauma. Neurosurgery 1982;11:344–51.
87. Thorton KE, Carmody DP. Efficacy of traumatic brain injury rehabilitation: inter-
ventions of qEEG-guided biofeedback, computers, strategies, and medications.
Appl Psychophysiol Biofeedback 2008;33(2):101–24.
88. Ayers ME. A controlled study of EEG neurofeedback and physical therapy with
pediatric stroke, age seven months to age fifteen, occurring prior to birth.
Biofeedback & Self-Regulation 1995;20(3):318.
89. Thorton KE, Carmody DP. Traumatic brain injury rehabilitation: qEEG
biofeedback treatment protocols. Appl Psychophysiol Biofeedback 2009;
34:59–68.
90. Jian L, Yin X, Cheng Y, et al. Different quantitative EEG alterations induced by
TBI among patients with different APOE genotypes. Neurosci Lett 2011;505:
160–4.
91. Duf J. The usefulness of quantitative EEG and neurotherapy in the assessment
and treatment of postconcussion syndrome. Clin EEG Neurosci 2004;35(4):
198–209.
92. Gerring JP, Brady KD, Chen A, et al. Premorbid prevalence of ADHD and devel-
opment of secondary ADHD after closed head injury. J Am Acad Child Adolesc
Psychiatry 1998;37(6):647–54.
93. Gerring J, Brady K, Chen A, et al. Neuroimaging variables related to develop-
ment of secondary attention deficit hyperactivity disorder after closed head
injury in children and adolescents. Brain Inj 2000;14(3):205–18.
94. Chabot RJ, di Michele F, Prichep L, et al. The clinical role of computerized
EEG in the evaluation and treatment of learning and attention disorders in
children and adolescents. J Neuropsychiatry Clin Neurosci 2001;13(2):
171–86.
95. Sunrmeli T, Ertem A. EEG neurofeedback treatment of patients with Down syn-
drome. J Neurotherapy 2007;11(1):63–8.
96. Surmeli T, Ertem A. Post WISC-R and TOVA improvement with qEEG guided
neurofeedback training in mentally retarded: a clinical case series of behavioral
problems. Clin EEG Neurosci 2010;1:32–41.
97. Egner T, Gruzelier JH. Ecological validity of neurofeedback: modulation of slow
wave EEG enhances musical performance. NeuroReport 2003;14:1221–4.
http://dx.doi.org/10.1097/00001756-200307010-00006.
98. Power JD, Fair DA, Schlaggar BL, et al. The development of human functional
brain networks. Neuron 2010;67:736–48.
99. Hwang K, Hallquist MN, Luna B. The development of hub architecture in the Hu-
man Functional Brain Network cerebral cortex. Cerebral Cortex 2012. http://dx.
doi.org/10.1093/cercor/bhs227. Available at: http://cercor.oxfordjournals.org/
content/early/2012/08/28/cercor.bhs227.long.
100. Coutin-Churchman P, Anez Y, Uzctegui M, et al. Quantitative spectral analysis of
EEG in psychiatry revisited: drawing signs out of numbers in a clinical setting.
Clin Neurophysiol 2003;114:2294–306.
101. Nuwer M. Clinical use of qEEG. Clin Neurophysiol 2003;114:2225.
102. Nuwer M. Assessment of digital EEG, quantitative EEG and brain mapping:
mapping: report of the American Academy of Neurology and the American Clin-
ical Neurophysiology Society. Neurology 1997;49:277–92.
103. Duffy FH, Hughes JR, Miranda F, et al. Status of quantitative EEG (qEEG) in clin-
ical practice, 1994. Clin Electroencephalogr 1994;25:6–22.
104. Monastra VJ, Lynn S, Linden M, et al. Electroencephalographic biofeedback in
the treatment of attention-deficit hyperactivity disorder. Appl Psychophysiol
Biofeedback 2005;30:95–114.
105. Kondacs A, Szabo M. Long-term intra-individual variability of the background
EEG in normals. Clin Neurophysiol 1999;110:1708–16.
106. Levesque J, Beauregard M, Mensour B. Effect of neurofeedback training on the
neural substrates of selective attention in children with attention-deficit/
hyperactivity disorder: a functional magnetic resonance imaging study. Neuro-
sci Lett 2006;394:216–21.
107. Arnold LE, Lofthous N, Hersch S, et al. EEG neurofeedback for ADHD: double
blind sham controlled randomised pilot feasibility trial. J Atten Disord 2012;
17(5):1–10.
108. Sherlin L, Arns M, Lubar J, et al. A position paper on neurofeedback for the
treatment of ADHD. J Neurotherapy 2010;14(2):66–78.
109. Monastra VJ, Monastr DM, George S. The effects of stimulant therapy, EEG
biofeedback, and parenting style on the primary symptoms of attention-
deficit-hyperactivity disorder. Appl Psychophysiol Biofeedback 2002;27:
231–49.
110. Arns M, de Ridde S, Strehl U, et al. Efficacy of neurofeedback treatment in
ADHD: the effects on inattention, impulsivity and hyperactivity: a meta-analysis.
Clin EEG Neurosci 2009;40:180–9.
111. Faraone SV, Buitelaar J. Comparing the efficacy of stimulants for ADHD in chil-
dren and. adolescents using meta-analysis. Eur Child Adolesc Psychiatry 2009;
14(1):353–64.
464 Simkin et al
112. Sonuga-Barker J, Brandeis D, Cortese S, et al, European ADHD Guidelines

Group. Nonpharmacological interventions for ADHD: systematic review and
meta-analyses of randomized controlled trials of dietary and psychological
treatments. Am J Psychiatry 2013;170:275–89.
113. Arns M, Strehl U. Evidence for efficacy of neurofeedback in ADHD? Am J Psy-
chiatry 2013;170:799–800.
114. van Dongen-Boomsma M, Vollebregt MA, Slaats-Willemse D, et al.
A randomized placebo-controlled trial of electroencephalographic (EEG) neuro-
feedback in children with attention-deficit/hyperactivity disorder. Focus on
Childhood and Adolescent Mental Health. J Clin Psychiatry 2013;74(8):821–7.
115. Raymond J, Varney C, Gruzelier JH. The effects of alpha/theta neurofeedback
on personality and mood. Cogn Br Res 2005;23:287–92. http://dx.doi.org/10.
1016/j.cogbrainres.2004.10.023.
116. Walker JE, Kozlowski GP. Neurofeedback treatment of epilepsy. Child Adolesc
Psychiatr Clin N Am 2005;14:163–76.
117. Lubar JF, Shahsin HS, Natelson SE, et al. EEG operant conditioning in intrac-
table epileptics. Arch Neurol 1981;38(11):699–704.
118. Beauregard M, Lévesque JL. Functional magnetic resonance imaging investiga-
tion of the effects of neurofeedback training on the neural bases of selective
attention and response inhibition in children with attention-deficit/hyperactivity
disorder. Appl Psychophysiol Biofeedback 2006;31(1). http://dx.doi.org/10.
1007/s10484-006-9001-y.
119. Nazari MA, Querne L, DeBroca AD, et al. Effectiveness of EEG biofeedback as
compared with methylphenidate in the treatment of attention-deficit/
hyperactivity disorder: a clinical outcome study. Neurosci Med 2011;2:78–86.
http://dx.doi.org/10.4236/nm.2011.22012. Available at: http://www.SciRP.org/
journal/nm Copyright Ó 2011 SciRes. NM.
120. Arns M, Drinkenburg W, Kenemans JL. The effects of qEEG-informed neuro-
feedback in ADHD: an open-label pilot study. Appl Psychophysiol Biofeedback
2012;37:171–80. http://dx.doi.org/10.1007/s10484-012-9191-4.
121. Hunte AM, Leuchter AF, Cook IA, et al. Brain functional changes (qEEG cord-
ance) and worsening suicidal ideation and mood symptoms during antidepres-
sant treatment. Acta Psychiatr Scand 2010;122(6):461–9.
122. Canterberry M, Hanlon C, Hartwell K, et al. Real time neurofeedback: exploring
the role of severity of dependence. Nicotine Tob Res 2013;15:1–5. http://dx.doi.
org/10.1093/ntr/ntt122.
123. Foster D, Center W, Butler B. Integral neurofeedback is superior to conventional
medication in a case of bipolar. Presented at the International Society of Neuro-
feedback and Research Conference. Dallas, Texas, September 18–22, 2013.
124. Hammond DC, First KL. Do no harm: adverse effects and the need for practice
standards in neurofeedback. J Neurotherapy 2008;12(1):79–88.
125. Arns M, Heinrich H, Strehle U. Evaluation of neurofeedback in ADHD: The long
and winding road. Biological Psychology 2014;95:108–15.
126. Arnold LE, Arns M, Conners K, et al. The Collaborative Neurofeedback Group, A
proposed multisite double-blind randomized clinical trial of neurofeedback for
ADHD: Need, rationale, and strategy. Journal of Attention Disorders 2013.
http://dx.doi.org/10.1177/1087054713482580.

NF Basics R. Thatcher, J. Lubar

Uploaded by

Copyright:

Available Formats

NF Basics R. Thatcher, J. Lubar

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

NF Basics R. Thatcher, J. Lubar

Uploaded by

Copyright:

Available Formats

Quantitative EEG and

Child Adolesc Psychiatric Clin N Am 23 (2014) 427–464

Key Points (Continued)

Abbreviations: Quantitative EEG and Neurofeedback

ADHD Attention-deficit/hyperactivity disorder

NF is based on operant conditioning. NF is based on measurements of evoked poten-

HOW NEUROFEEDBACK CHANGES ACTION POTENTIALS

Fig. 3. Phase lock.

IS ALL BRAIN BIOFEEDBACK ALIKE?

QUANTITATIVE ELECTROENCEPHALOGRAPHY AND ITS ROLE IN NEUROFEEDBACK

Quantitative electroencephalography involves computerized data analysis to precisely

VALIDATION OF QUANTITATIVE ELECTROENCEPHALOGRAPHY BY USE OF

COREGISTRATION OF QUANTITATIVE ELECTROENCEPHALOGRAPHY WITH DIFFUSE

QUANTITATIVE ELECTROENCEPHALOGRAPHY AND NF FOR PSYCHIATRIC CONDITIONS

How Complex Homeostatic Systems Regulate the EEG Power Spectrum

How May This Occur in the Case of ADHD?

Psychiatry as Related to Circuitry: Moving from Categorical Disorders to

dysconnectivity has a long history. Connectivity analysis (using fMRI research) in

distinct developmental trajectories of behavioral and emotional dysregulation in youth

NEUROFEEDBACK TREATMENT IN CHILDREN AND ADOLESCENTS

Neurofeedback Treatment of Depression

Neurofeedback Treatment of Anxiety Disorders

Neurofeedback Treatment of Substance Abuse

Neurofeedback Treatment of Substance Use Disorders in the Context of Comorbid

Findings in Chosen Substance of Abuse ADHD Protocol ADD Protocol

Abbreviation: ADD, attention deficit disorder.

Surface Neurofeedback Treatment of Posttraumatic Stress Disorder

visual feedback, with feedback initially contingent on reduction of delta/theta activity.

NEUROFEEDBACK TREATMENT OF INTRAUTERINE BRAIN DAMAGE, MILD HEAD

treatment remained stable in the 2 years of follow-up according to parents’ reports.96

Neurofeedback in Reducing Performance Anxiety

Does Neurofeedback Technique Have to be Modified for Developmental Age?

points may indicate that the accreditation used in NF may need to be

USPSTF strength of recommendations: A, recommend strongly (good evidence of benefit and

1. Thatcher RW. Introduction. In: Robert Thatcher, editor. Handbook on quantitative

5. Pascual-Marqui RD, Lehmann D, Koenig T, et al. Low resolution brain electro-

112. Sonuga-Barker J, Brandeis D, Cortese S, et al, European ADHD Guidelines

You might also like