Content uploaded by Robert Coben
Author content
All content in this area was uploaded by Robert Coben
Content may be subject to copyright.
Assessment-Guided Neurofeedback for Autistic Spectrum Disorder
Robert Coben, Ph.D.
Neurorehabilitation & Neuropsychological Services
And
Ilean Padolsky, Ph.D
Neurorehabilitation & Neuropsychological Services
Address for Correspondence to:
Robert Coben, Ph.D.
Director, Neurorehabilitation & Neuropsychological Services
1035 Park Blvd., Suite 2B
Massapequa Park, NY 11762
Telephone: 516-799-8599
Fax: 516-799-4054
Neurofeedback for ASD
1
Abstract
Background. Research reviewing the epidemiology of Autism (Medical Research Council,
2001) indicated that approximately 60 per 10,000 children (1/166) are diagnosed with Autistic
Spectrum Disorder (ASD). Jarusiewicz (2002) published the only controlled study documenting
the effectiveness of neurofeedback for Autism based on one outcome measure. The
present study extended these findings with a larger sample size, broader range of assessments,
and physiological measures of brain functioning.
Methods. Assessment-guided neurofeedback was conducted in 20 sessions for 37 patients
with ASD. The experimental and control groups were matched for age, gender, race,
handedness, other treatments, and severity of ASD.
Results. Improved ratings of ASD symptoms reflected an 89% success rate (p < .0001).
Paired sample t-tests indicated statistically significant improvement in Autistics who
received Neurofeedback compared to the control group. Other major findings included:
40% reduction (p < .0001) in core ASD symptomatology (indicated by ATEC Total Scores),
and 77% (p=.0392) of the experimental group had decreased hyperconnectivity or no change.
Reduced cerebral hyperconnectivity was associated with positive clinical outcomes in this
population. In all cases of reported improvement in ASD symptomatology, positive treatment
outcomes were confirmed by neuropsychological and neurophysiological assessment.
Conclusions. Evidence from multiple measures has demonstrated that neurofeedback
can be an effective treatment for ASD. In this population, a crucial factor in explaining improved
clinical outcomes in the experimental group may be the use of assessment-guided neurofeedback
to reduce cerebral hyperconnectivity. Implications of these findings are discussed.
Neurofeedback for ASD
2
KEYWORDS. Neurofeedback, QEEG analysis, Infrared (IR) imaging, neuropsychological
assessment, Autistic Spectrum Disorder, Autism, Asperger’s Syndrome, executive deficits,
hyperconnectivity, hypoconnectivity.
Introduction
In recent years, Autistic Spectrum Disorder (ASD) has shown a dramatic increase in
prevalence. A review of prevalence survey research for ASD (identified by DSM-IV
criteria for Autism, Asperger’s Syndrome, and Pervasive Developmental Disorder-Not
Otherwise Specified) across the United States and the United Kingdom reported rates of
ASD substantially increased from prior surveys indicating 5 to 10 per 10,000 children to
as high as 50 to 80 per 10,000 (equivalent to a range of 1 in 200 to 1 in 125 children with
ASD) (Blaxill, 2004). Another review of research on the epidemiology of Autism
(Medical Research Council, 2001) indicated that approximately 60 per 10,000 children
(equivalent to a range of 1 in 166 children) are diagnosed with Autistic Spectrum
Disorder.
Autism is defined as a neurodevelopmental disorder characterized by impairment in
social interaction and communication. Historically, Kanner and Asperger introduced the
term Autism (Kanner & Eisenberg, 1956; Asperger, 1944). Further research concluded
that Autism can be categorized as part of a spectrum of heterogeneous disorders. This
continuum of disorders is characterized by a broad range of abilities and levels of
severity. The common feature of Autistic Spectrum Disorder (ASD) is qualitative
impairment in social and communication domains, as well as imaginative development
(Wing & Gould, 1979). More current research indicates that Autism is one of a range of
Neurofeedback for ASD
3
related Pervasive Disorders including: Asperger’s Disorder, Pervasive Developmental
Disorder-Not Otherwise Specified (PDD-NOS), Childhood Disintegrative Disorder
(CDD), and Rett’s Disorder (Medical Research Council, 2001).
The triad of symptoms including impaired communication, social skills, and
imaginative development formed the basis for the current international classification
systems- International Classification of Diseases (ICD-10; WHO, 1993) and Diagnostic
and Statistical Manual, 4
th
edition (DSM-IV; APA, 1994). Both diagnostic systems
characterize ASD as a disorder of early onset (before the age of 3), with impairment
in social interaction, communication and imagination, as well as restricted interests and
activities (Medical Research Council, 2001).
The heterogeneity within the spectrum of Autistic Disorders has led researchers to
propose a division of Autism into subgroups: 1) Low, medium, and high-functioning; and
2) Non-regressive and regressive subtypes differentiated by age of onset. Regressive
Autism occurs in 15-40% of children with ASD. This disorder is characterized by
normal development for 15-19 months followed by loss of vocabulary, reduced social
interaction and responsiveness, and sometimes repetitive play behavior (Medical
Research Council, 2001).
In some cases, children with Autism may never develop patterns of typical speech.
Their speech may be inflexible and unresponsive to the context. Speech may be limited
to echolalia or narrow topics of specialized knowledge. Communicative impairment
includes nonverbal cues such as eye contact, facial expression, and gesture. Social
behaviors are often characterized by lack of interaction; play lacks cooperation and
imagination and is narrowly focused on repetitive activities (Belmonte et al., 2004).
Neurofeedback for ASD
4
Executive deficits associated with Autism have been attributed to frontal lobe
dysfunction resulting in perseveration and the inability to shift attention. Weak central
coherence (a preference for local detail over global processing) has been attributed to
individuals with Autism to explain their superior ability to attend to details. In addition,
weak central coherence also predicts the tendency of people with Autism to have deficits
in understanding global systems or the relation between the parts and the whole (Baron-
Cohen, 2004).
The other subdivisions of Autistic Spectrum Disorder include: Asperger’s Disorder,
Pervasive Developmental Disorder-Not Otherwise Specified, Childhood Disintegrative
Disorder, and Rett’s Disorder. Individuals with Asperger’s Syndrome frequently have
high levels of cognitive function, speech is characterized by literal pedantic
communication, difficulty comprehending implied meaning and fluid motion, as well as
inappropriate social interaction. Pervasive Developmental Disorder-Not Otherwise Specified
(PDD-NOS) reflects deficits in language and social skills which do not meet the criteria of other
disorders. In contrast, Childhood Disintegrative Disorder and Rett’s Disorder both have normal
periods of early development followed by loss of previously acquired skills. Most of the
conditions described involve deficits in communication and social skills, however they
vary considerably in terms of onset and severity of symptomatology included within the
Autistic Spectrum of Disorders (Attwood, 1998; Hamilton, 2000; McCandless, 2005; Sicile-
Kira, 2004; Siegel, 1996).
Current research suggests that Autistic Spectrum Disorders may be associated with
functional disconnectivity between brain regions. There is evidence for anomalies in the
functional connectivity of the medial temporal lobe ( Baron-Cohen, 2004; Belmonte et al., 2004).
Neurofeedback for ASD
5
Abnormalities were found specifically in the functional integration of the
amygdala and parahippocampal gyrus (Welchew et al., 2005). This points to the need for
therapeutic interventions which address ASD as a neurodevelopmental and brain disorder.
Recent survey research reported on the therapies that parents most frequently selected
for their children with Autistic Spectrum Disorder (Green et al., 2006). The majority of
parents reported utilizing as many as seven different treatment modalities to ameliorate
their children’s symptoms of Autism. These include speech therapy (the most common),
visual schedules, sensory integration, applied behavior analysis, medications, special diets, and
vitamin supplements (Green et al., 2006).
The Research Units on Pediatric Psychopharmacology (RUPP) Autism Network
(2005a; 2005b) has conducted two separate studies related to the use of Risperidone and
Methylphenidate. In the first of these studies (RUPP Autism Network, 2005a), Risperidone was
effective in reducing irritability, but with side effects and a significant relapse rate. In the
Methylphenidate study (RUPP Autism Network, 2005b), 49% of the sample was considered
positive responders, but with significant non-responders and an 18% side effect rate.
Behavior therapy is another frequently implemented treatment for children with
Autism. Smith et al. (2000) demonstrated that intensive treatment conducted over two to three
years was successful in improving IQ and language functions. Sallows & Graupner (2005)
observed a significant improvement in 48% of the subjects. Rapid learners were in regular
education by age 7. The best outcomes were associated with the capacity for imitation, social
responsiveness, and language.
Although behavior therapy improves social, cognitive and language skills, years of intensive
training are required before children can attain positive treatment outcomes. Parents who select
Neurofeedback for ASD
6
behavior therapy for their children with Autism appear to be highly motivated and committed to
completion of the program.
Another intervention that has been studied in terms of efficacy is vitamin, mineral, and
enzyme supplementation. Adams & Holloway (2004) conducted a randomized, double-
blind placebo controlled study to investigate the effects of a multivitamin/mineral supplement on
ASD (n=20). The results indicated that 84% of their sample had improved sleep and
gastrointestinal symptoms, but there was a side effect rate of 18%. Chez et al. (2002) found
that L-carnosine supplementation led to improved ratings of behavior, socialization and
communication.
When vitamin and mineral deficiencies are treated, there can be improvement in
certain conditions co-occurring with Autism such as gastrointestinal and sleep disorders.
However, some children with Autism may have allergic reactions to certain forms or
dosages of vitamin and mineral supplementation. Therefore, careful monitoring of
dosage levels and adjustments are required.
Special diets are another biomedical non-drug intervention which were found to
be effective in the treatment of Autism. Reichelt & Knivsberg (2003) found that a gluten-
free/casein-free diet followed over four years led to improvements in cognitive, social, language,
and behavioral domains. The total number of children who improved following the dietary
intervention was not reported in the study. Therefore, percentage of improvement for the group
receiving the intervention could not be calculated.
Based on research reporting the co-occurrence of gastrointestinal conditions
with Autism, secretin (a gastrointestinal hormone) has also been studied as a treatment
for Autism. Roberts et al. (2001) investigated the effects of repeated doses of intravenous
Neurofeedback for ASD
7
secretin on 64 children diagnosed with Autism.in a randomized, placebo controlled
study. Following treatment, receptive and expressive language improved in both groups but the
amount of improvement did not distinguish between groups. However, parents anecdotally
reported the following changes: sleep improvement in 7 children (10.9%), 4 of whom had
diarrhea according to the GI questionnaire ( 6.25%), toilet training in 3 shortly after the injection
(4.68%); and more connectedness in 5 children (7.8%). Twenty-one percent of children
receiving secretin injections had generalized flushing in the neck, face or chest immediately
following the injection (Roberts et al., 2001).
Another condition that can co-occur with Autism is heavy metal toxicity which
involves excessive levels of mercury. Chelation therapy utilizes Di-mercaptosuccinic-
Acid (DMSA) to clear the body of mercury or other toxic metals. Bradstreet et al. (2003)
conducted a case control study of mercury toxicity in children with Autistic Spectrum
Disorders (n=221). Following, an oral chelating agent, urinary mercury concentrations were
significantly higher in 221 children with Autistic Spectrum Disorders than in 18 normal controls
(p < .0002). Vaccinated children with ASD had significantly higher urinary mercury
concentrations then did vaccinated controls (p < .005 ).
Holmes (2001) documented the progress of children with Autism (n=85; 40 aged 1-5
yrs.; 25 aged 6-12 yrs.; 16 aged 13-17 yrs.; and 4 aged >18 yrs.) treated with chelation
(DMSA + lipoic acid) for at least four months. Marked improvement in behavior, language, and
social interaction was noted in 35% of children 1-5 years of age. Moderate improvement was
found in 39% of children aged 1-5, 28% of children aged 6-12 and 6% of children aged 13-17.
However, 52% of children aged 6-12, 68% of children aged 13-17 made only slight
improvement, and 75% of individuals over 18 made no improvement. The results of the Holmes
Neurofeedback for ASD
8
study indicate that chelation therapy was effective for children with Autism under the age of six.
In contrast, the majority of older children and adolescents did not benefit from this treatment
(Kirby, 2005). Holmes (2001) noted that younger patients excreted larger quantities of mercury
than did older patients which may explain this discrepancy in treatment outcomes.
Rimland (2005) in association with the Autism Research Institute collected responses
from 23,700 parents of children with Autism rating the efficacy of biomedical drug and
non-drug interventions. The benefit to harm ratios for several of the therapies discussed
previously are listed below in Table 1.
---------------------
Insert Table 1
---------------------
As shown in Table 1, the most effective treatments are chelation, digestive enzymes, and
gluten-/casein-free diets. These findings are based on parent report only and additional
research is necessary to provide further support for these findings. Special diets can also result
in improved ASD symptoms, however regression in symptoms can occur after discontinuation
of the diet (Reichelt & Knivsberg, 2003). Digestive enzymes must also be continued to maintain
improved treatment outcomes. Vitamin therapy and secretin may also be beneficial, however
some children with Autism may have allergic reactions to secretin and certain forms of vitamin
and mineral supplementation (Adams & Holloway, 2004;Roberts et al. 2001).
The least effective treatments for ASD were Ritalin, Risperidal, Thorazine, and Haldol.
Although neuroleptics (i.e., Thorazine and Haldol) may reduce dysfunctional behaviors
associated with ASD, adverse side effects such as sedation, irritability, and extrapyramidal
dyskinesias limit the use of these medications (Committee on Children with Disabilities, 2001).
In addition, side effects can include weight gain (for Risperidal), decreased appetite and
Neurofeedback for ASD
9
difficulty falling asleep (for Ritalin). There may also be a rebound of aggressive behavior when
medication is discontinued (RUPP Autism Network, 2005a; RUPP Autism Network, 2005b).
In comparison, neurofeedback is a non-invasive therapeutic intervention which has been
shown to enhance neuroregulation and metabolic function (Coben, 2005b, 2005c). In contrast to
behavior therapy, positive treatment outcomes as a result of neurofeedback training are achieved
over the course of several months rather than a year or more of intensive training.
Neurofeedback has no adverse side effects while psychopharmacological interventions,
as well as certain vitamin/mineral supplementation and secretin are associated with side
effects. The therapeutic treatment outcomes of neurofeedback training are maintained
over time and do not reverse after treatment is withdrawn (Linden, Habib & Radojevic,
1995; Lubar et al., 1995; Monastra et al., 2005; Tansey, 1993) as in drug therapy, diet
therapy, and supplementation with vitamins, minerals, and enzymes.
In 1994, Cowan & Markham conducted the first case study of neurofeedback with Autism.
QEEG analysis was performed on an 8 year old girl diagnosed with high functioning Autism
during eyes open and at rest. The findings indicated an abnormally high amount of alpha (8-10
Hz) and theta (4-8 Hz) activity predominately in the parietal and occipital lobes. Based on these
results, a neurofeedback protocol was designed to suppress the ratio of theta and alpha (4-10 Hz)
to beta (16-20) EEG activity at central and parietal sites using a bipolar (sequential) montage
(two scalp electrodes and one ear reference electrode). The findings following 21 neurofeedback
sessions included: increased sustained attention, decreased autistic behaviors (inappropriate
giggling, spinning), improved socialization based on parent and teacher reports. There were
also substantial improvements in the Test of Variables of Attention (TOVA) for measures of
inattention (omission), impulsivity (commission) and variability. A follow-up TOVA was
Neurofeedback for ASD
10
administered two years later. All scores were within normal limits. In addition, the girl
continued to maintain positive social interactions as reflected by engaging in team sports.
Other researchers have also reported positive treatment outcomes or normalizing trends for
children with Autism or Asperger’s Syndrome treated with neurofeedback (Sichel et al., 1995;
Scolnick, 2005). However, these studies utilized only single case or small group designs without
control groups. Thompson & Thompson (1995) conducted research on neurofeedback combined
with metacognitive strategies for a group of boys (n=15; aged 8-14). Nine of the children met
criteria for Asperger’s Syndrome and the others met criteria for Attention Deficit Disorder and
Learning Disabilities. All 15 boys improved as indicated by parent-teacher interviews, academic
function and sustained visual and auditory attention.
Jarusiewicz (2002) published the only group study documenting the efficacy of
neurofeedback for Autistic Disorders. Forty participants responded to a request to
participate in the research. Only 12 of the 20 experimental group children completed
20 or more sessions (range 20-69; mean=36 sessions) necessary for data analysis. Measurement
of treatment outcome was based on the use of only one assessment measure- the Autism
Treatment Evaluation Checklist (ATEC). The initial protocols were reward at site C-4
referenced to the contralateral ear in the 10-13 Hz range or lower depending on each child’s
ATEC score. Inhibits were set at 2-7 Hz and 22-30 Hz. The 2-7 Hz inhibit was selected due to
the significant levels of delta and theta found in the spectrals of all the children in the study. This
protocol was applied to 57% of the children with adjustments as necessary (Jarusiewicz, 2002).
For children that experienced problems with vocalization during training, an F7
electrode placement with a right ear reference was utilized. The protocol included
augmenting 15-18 Hz and inhibiting 2-7 Hz and 22-30 Hz. When children were able
Neurofeedback for ASD
11
to maintain training without demonstrating signs of overarousal, additional five minute
increments were provided until the session reached 30 minutes in duration. This protocol
was administered 15% of the time, and was frequently followed by the C4 electrode
placement and initial protocol for calming effects (Jarusiewicz, 2002).
For children who required assistance in enhancing socialization and communication
skills, a bipolar F3-F4 electrode placement was employed. A 7-10 Hz to 14.5-
17.5 Hz augment and 2-7 Hz and 22-30 Hz inhibit protocol was utilized. This protocol
was employed 12% of the time, and it was discontinued if giggling and inappropriate
laughter occurred (Jarusiewicz, 2002).
For children who experienced emotional instability, a bipolar T3-T4 electrode
placement was implemented, beginning with 9-12 Hz rewards and inhibits at 2-7 Hz/
22-30 Hz. Protocol frequencies were adjusted up or down if further reduction of anxiety,
sadness, and hyperactivity were necessary. The protocol was employed 13% of the time.
Children received one to three training sessions per week, with two sessions per week
as the most common frequency of sessions.
Children with Autistic Spectrum Disorder who completed neurofeedback training
attained a 26% average reduction in the total ATEC rated autism symptoms in contrast to
3% for the control group. Parents reported improvement in socialization, vocalization,
anxiety, schoolwork, tantrums, and sleep while the control group had minimal changes in
these domains (Jarusiewicz, 2002).
Further research on methods of developing effective neurofeedback protocols for
children with Autistic Spectrum Disorders is needed. Autism encompasses a broad
range of symptoms (e.g., anomalies in communication, social behavior, cognitive and
Neurofeedback for ASD
12
motor function, seizure activity, obsessive compulsive behavior, atypical sleeping and
eating patterns), therefore one single assessment measure may not provide sufficient data
to accurately target specific sites associated with dysfunction and disregulation. Coben’s
(2005a, 2005b, 2005c) research has shown that improved outcomes can result from
assessment providing multiple data points to guide the development of individualized
neurofeedback protocols which target specific brain regions to increase activation,
symmetry, and interconnectivity.
In the present study, we seek to extend Jarusiewicz’ findings with a larger sample
size and broader range of measures to evaluate treatment outcome. The assessments
utilized included: neuropsychological tests, ratings of behavior and executive function,
Quantitative EEG (QEEG) analysis, Infrared imaging to accurately target dysfunctional
or disregulated regions in need of remediation, as well as parent rating of treatment
outcome. Treatment protocols were assessment-based and individualized for each child
receiving neurofeedback training.
Method
Participants
Thirty-seven children diagnosed with Autistic Spectrum Disorder (ASD) participated
in the study and served as the experimental group. There were 12 participants in the
wait-list control group similarly diagnosed with ASD. The experimental and control
group were matched based on age, gender, race, handedness, other treatments, and
severity of ASD as indicated by the Autism Treatment Evaluation Checklist (ATEC).
The experimental group received assessment-guided neurofeedback training for at least
Neurofeedback for ASD
13
20 sessions. Of the initial 38 patients that began the study, only one patient dropped out
prior to completion of the study. No new treatments were undertaken by any participants
during the course of the study. Procedures were explained to parents and informed
consent was obtained for their children to participate in the study. Refer to Table 2 for
the demographics of the neurofeedback group and Table 3 for the demographics of the control
group.
-------------------------
Insert Table 2
------------------------
-----------------------
Insert Table 3
------------------------
As shown in Table 4 below, the ASD diagnoses for the experimental group were as follows:
56.8% (n=21) had Pervasive Developmental Disorder-Not Otherwise Specified (PDD-NOS);
18.9% (n=7) Autism; 13.5% (n=5) Asperger’s Disorder; and 10.8% (n=4) Childhood
Disintegrative Disorder. The majority of participants (75.7%) were diagnosed with
PDD-NOS or Autism.
--------------------
Insert Table 4
--------------------
Procedure
A diagnostic interview was conducted with the parents to ascertain core
behavioral, cognitive and social/emotional issues of concern as part of a comprehensive
neurodevelopmental history. Following the interview, neurobehavioral rating scales
were administered which included: the Autism Treatment Evaluation Checklist (ATEC),
Gilliam Asperger’s Disorder Scale (GADS), Gilliam Autism Rating Scale (GARS),
Neurofeedback for ASD
14
Behavior Rating Inventory of Executive Function (BRIEF), and Personality Inventory for
Children (PIC-2). Baseline measures also included neuropsychological evaluation of
executive, attentional, visual-perceptual, and language functioning. All participants also
underwent Quantitative EEG (QEEG) analysis. Another measure of underlying cortical
activity was Infrared (IR) imaging. IR imaging assesses the thermoregulation of specific
brain regions which is associated with metabolic activity and regional Cerebral Blood
Flow (rCBF). IR imaging was conducted prior to and following each training session.
All other assessments were administered prior to and following treatment.
Materials
Assessment Instruments
At the completion of the study, parents rated the effectiveness of assessment-guided
neurofeedback. An index of Parental Judgment of treatment efficacy was computed to
provide a benefit-harm ratio. The index consisted of three categories of Parental
Judgment: 1. Improved; 2. No Change; and 3. Worse. The Parental Judgment
Ratings were compared to those calculated by Rimland (2005) for other therapeutic
approaches to ASD (as previously described).
The Autism Treatment Evaluation Checklist (ATEC; Rimland & Edelson, 2000) was
developed as a valid means of assessing the effectiveness of treatments for Autism. The
ATEC consists of a one-page checklist to evaluate the severity of the core symptoms of
Autism as rated by parents or primary caretakers. The instrument is divided into four
subtests consisting of: 1. Speech/Language/Communication (14 items); 2. Sociability (20
items); 3. Sensory/Cognitive Awareness (18 items); and 4. Health/Physical/Behavior (25
Neurofeedback for ASD
15
items). The Autism Research Institute developed an Internet scoring procedure that
computes the four subscale scores and a total ATEC score. The severity of disorder is
reflected by higher subscale and total scores.
The ATEC was normed on the first 1,358 ATEC forms submitted to the Autism
Research Institute by mail, fax, or Internet. The Pearson split-half coefficients reflecting
internal consistency were: Scale I: Speech .920; Scale II: Sociability .836; Scale III:
Sensory/Cognitive Awareness .875; Scale IV: Health/Physical/Behavior .815; and
ATEC Total: .942. The ATEC was shown to be a reliable measure with strong internal
consistency indicating that items within each scale measure the same domain of
behavior. Therefore, pre-treatment ATEC scores can be reliably compared with post-
treatment scores.
The Gilliam Asperger’s Disorder Scale (GADS; Gilliam, 2001) is a behavioral rating
scale. The GADS consists of 32 items divided into four subscales including: Social
Interaction (10 items); Restricted Patterns of Behavior (8 items); Cognitive Patterns (7
items); and Pragmatic Skills (7 items).
The GADS was normed on a sample of 371 individuals (aged 3-22; males n=314/
Females n=57) diagnosed with Asperger’s Disorder from across 46 states, the District of
Columbia, Canada, Great Britain, Mexico, Australia, and other countries. Internal
consistency reliability coefficients ranged from .87 to .95 for total Asperger’s Disorder
Quotient across samples of children with and without identified disabilities. The test-
retest reliability for the Asperger’s Disorder Quotient is .93 (p < .01). These results indicate
that the GADS has a high level of stability reliability for use as a pre-/post-treatment
measure of individuals with Asperger’s Disorder. Construct validity was indicated by
Neurofeedback for ASD
16
analyses finding that: GADS scores are minimally related to age; items on
the subscales are representative of behaviors associated with Asperger’s Disorder;
persons with other diagnoses score differentially; GADS scores are strongly related to
each other and performance on other tests that screen for serious behavioral disorders;
and the GADS can discriminate among individuals with Asperger’s Disorder and those
with behavioral disorders.
The Gilliam Autism Rating Scale (GARS; Gilliam, 1995) is a behavioral checklist.
The GARS is comprised of four subtests (Stereotyped Behaviors; Communication; Social
Interaction; and Developmental Disturbances) of 14 items each. The scale was normed
on a sample of 1,092 children and young adults (aged 2-28) across 46 states, the District
of Columbia, Puerto Rico, and Canada.
The internal consistency reliability coefficients for all subtests and total Autism
Quotient range from .88 to .96. The stability or test-retest reliability ranges from .81 to
.88 for all subtests and total Autism Quotient. These results indicate high levels of
stability reliability required for pre-/post-treatment assessment of individuals with
Autism. The construct validity was confirmed by analyses finding that: items of the
subscales are representative of the behaviors associated with Autism; GARS scores
strongly related to each other and to performance on other screening tests for Autism;
GARS scores are not related to age; and individuals with other diagnoses score
differentially on the GARS.
The Behavior Rating Inventory of Executive Function (BRIEF; Gioia, Isquith, Guy, &
Kenworthy, 2000) is a questionnaire completed by parents or teachers of children to
assess executive behaviors. The parent and teacher forms of the BRIEF contain 86 items
Neurofeedback for ASD
17
within 8 theoretically and empirically derived clinical scales that measure different
aspects of executive functioning: Inhibit, Shift, Emotional Control, Initiate, Working
Memory, Plan/Organize, Organization of Materials, and Monitor.
For parent and teacher forms of the BRIEF internal consistency was high ranging
from .80 to .98. Test-retest reliability ranged from .80 to .92 across overall indices of
Behavioral Regulation, Metacognition, and Global Executive Composite. These results
indicate high reliability stability needed for pre-/post-treatment assessment. The validity
of the BRIEF is confirmed by factor analysis indicating a two-factor model.
The Personality Inventory for Children, Second Edition (PIC-2; Lachar & Gruber,
2001) is a multidimensional, objective questionnaire developed to evaluate domains of
adjustment in children and adolescents. The PIC-2 was normed on a standardization
group (N=2,306) reflecting a cross-section of children in the United States. Data was
representative of urban, suburban, and rural areas, across socioeconomic status (SES)
(including poor, blue-collar, middle-class, and upper SES status), as well as the major
ethnic groups (Asian, Black, Hispanic, Caucasian, Other).
The PIC-2 contains 275 items completed by parents or parent surrogates to identify
domains of adjustment consisting of: Cognitive Impairment, Impulsivity &
Distractibility, Delinquency, Family Dysfunction, Reality Distortion, Somatic Concern,
Psychological Discomfort, Social Withdrawal, and Social Skill Deficits. A Behavioral
Summary is made up of the first 96 items of the PIC-2 and contains composite scales
(i.e., Externalization, Internalization, Social Adjustment, and Total Score).
The internal consistency ranges from .78 to .95 for the composite scales and the Total
Score. Test-retest stability was .89 for all composite scores including the Total Score for
Neurofeedback for ASD
18
nonclinical and clinically referred samples. These results indicate high reliability
stability necessary for pre-/post-treatment assessment. Validity was confirmed by factor
analytic studies of the PIC-2 Standard Form Adjustment Subscales which yielded a five
factor solution (Externalizing Symptoms; Internalizing Symptoms; Cognitive Status;
Social Adjustment; and Family Dysfunction) and a two factor solution for the Behavioral
Summary Short Adjustment Scales (Externalizing and Internalizing).
Neuropsychological testing has been sufficiently validated as a reliable procedure for
evaluating cognitive functions (Lezak, 1995) and was utilized for this purpose in our study.
Neuropsychological measures constituting composite indices of attention, visual-perceptual,
executive function, and language skills (Delis-Kaplan Executive Function System; NEPSY;
Integrated Visual and Auditory Continuous Performance Tests; and others) were administered to
assess pre-/post-treatment levels of attention, visual-perceptual, language, and executive
function. All Neuropsychological measures used, including the Delis-Kaplan Executive
Function System (D-KEFS; Delis, Kaplan, & Kramer, 2001), Developmental
Neuropsychological Assessment (NEPSY; Korkman, M., Kirk, U., & Kemp, S., 1998),
Comprehensive Test of Visual Functioning (CTVF; Larson, Buethe, &
Vitali, 1990), Rey Complex Figure Test and Recognition Trial (RCFT; Meyers &
Meyers, 1995), Expressive One-Word Picture Vocabulary Test (EOWPVT; Upper-
Extension; Gardner, 1983), Expressive One-Word Picture Vocabulary Test-Revised
(EOWPVT-R; Gardner, 1990), and The Integrated Visual and Auditory Continuous Performance
Test (IVA; Sanford & Turner, 2002), have demonstrated adequate reliability and validity.
Quantitative EEG (QEEG) involved recording and digitizing EEG based on the
International 10/20 System of electrode placement utilizing the Deymed Diagnostic (2004)
Neurofeedback for ASD
19
TruScan 32 Acquisition EEG System. (Refer to Table 5 for specifications).
Data were acquired (during eyes closed/eyes open conditions) using a stretchable electrode
cap embedded with 19 sensors with frontal reference, prefrontal ground, and linked ears;
attached to the scalp by means of electrode paste. The duration of recording was a total of 20
minutes; 10 minutes in each condition. All data was manually artifacted in NeuroRep
(Hudspeth, 1999) and analyzed with the same EEG analysis software including measures of
multivariate coherence or connectivity. The neuroelectric eigen image (NEI) can be defined as a
3-D structure which results from the Principal Components Analysis (PCA) of the multichannel
(i.e., 19) EEG waveforms. PCA results routinely show that EEG waveforms can be explained by
3 orthogonal waveform components that refer to the lateral, anterior-posterior, and dorsoventral
position of recording electrodes. Although every effort is made to situate electrodes at equal
distances on the scalp, it is abundantly clear that PCA results show that functional
interelelectrode distances are not equal and therefore, must be estimated as vector distances:
squareroot (dx
2
+dy
2
+dz
2
). Therefore, it can be seen that a connectivity image (CIM) can be
constructed as the average interelectrode distances that converge on each of the 19 electrodes,
with 3 elements for each edge electrode and 4 elements for internal electrodes. Thus, normative
average and standard deviation reference data were computed for the 19 electrode sites of the
CIM indices based on 30 normal adults. Statistical comparisons are made with effect size
estimates, r = z/squareroot(N), based on methods discussed in Rosenthal &DiMatteo (2001)
(W.J. Hudspeth, personal communication, July 25, 2006). Further analyses included measures
of absolute and relative power, as well as connectivity processed by the Neurometric Analysis
System (NxLink, 2001; John, 1988) and Neuroguide (Thatcher et al., 2003) EEG software (both
FDA approved) with age referenced normative databases. A permanent record was made prior to
Neurofeedback for ASD
20
initiation and at the completion of the study for both the assessment-guided neurofeedback group
and the control group. The reliability and validity of QEEG has been established (Thatcher et
al., 2003).
NeuroCybernetics EEGer Training System (NeuroCybernetics Inc., 2006) was the software
utilized to perform assessment-guided Neurofeedback. Hardware included Thought Technology
encoders. Sensors (Grass Silver Disc 48” Electrodes with SafeLead protected terminals; Grass
SafeLead, 2006) were applied to the patient’s scalp to measure EEG activity. The signal is then
fed back to the patient in visual and aural form based on relative amplitude/threshold values.
The patient learns to inhibit frequencies which are excessively generated and augment
frequencies which are targeted for training.
The aural reward rate is limited to 2 Hz so each individual sound is audible to the patient.
The aural reward is a prerecorded sound file of a short ½ second beep when specified amplitude
conditions are met. The visual feedback consists of simple graphics providing a continuous
display of the ratio of amplitude to threshold for each stream of data. Visual feedback can be
provided in the following game formats: 4mation, Boxlights, Highway, Island, Jumpbox, Mazes,
EEG Chomper, Space Race, Cubes, and Starlight (NeuroCybernetics Inc., 2006) (Refer to
Table 5 for specifications).
-----------------
Insert Table 5
------------------
A ThermoVision A20M camera from FLIR Systems (2006) was used for infrared imaging.
As part of the imaging procedure, the camera (mounted on a tripod) was set up
approximately two feet from the patient and the thermal image was projected onto a
screen (Please Refer to Table 6 for specifications).
Neurofeedback for ASD
21
------------------
Insert Table 6
-----------------
Infrared (IR) imaging of the prefrontal area was performed prior to and following each
neurofeedback training session. IR imaging assesses the levels of thermoregulation. Thermal
output is assigned thermal degrees. The levels of thermal activity are associated with underlying
metabolic activity and regional Cerebral Blood Flow (rCBF). Research indicates that IR
imaging is a valid and reliable measure of brain activity, metabolic processes, and rCBF
(Carmen, 2004; Coben, Carmen, & Falcone, 2005a; Coben, 2005b; Coben, 2005c;
Toomin et al., 2004).
Neurofeedback Protocols
Training protocols were based on the combined use of all assessment information with a
heavy emphasis on initial QEEG which included analysis of absolute, relative power, and
connectivity measures. Protocols included primarily sequential (bipolar) or interhemispheric
montages individualized for each patient. The focus was on reducing hyperconnectivity which
was frequently observed in posterior-frontal to anterior-temporal regions. These protocols
remained constant during the training period of 20 sessions and were conducted twice weekly.
For each patient, the neurofeedback protocols were determined based on regions of maximal
hyperconnectivity. For example, one patient had maximal hyperconnectivity in the right
frontal region primarily in alpha. A protocol was designed for this patient to inhibit alpha (the
frequency range of maximal hyperconnectivity) and reward low beta at F8/F7.
Eighty-nine percent of the 37 patients had sequential (bipolar) versus unipolar montages.
Ninety-four percent of the sequential (bipolar) montages included frontal or temporal electrode
Neurofeedback for ASD
22
sites including F8-F7, Ft8-Ft7, T4-T3, or F7-F8. In one case, F6-F5 was applied and in the other
F4-F3. Reward bands ranged from 5-16 Hz. A delta inhibit protocol as low as 1-2 Hz ranging
to as high as 6 Hz was utilized for 92% of the patients. In 100% of patients, a high
beta inhibit protocol was applied ranging from 18-50 Hz with the greatest overlap at 18-30 Hz.
A third inhibit ranging within a 7-14 Hz range was utilized for 68% of the patients.
Results
The experimental group was composed of 37 patients diagnosed with ASD;
84% were males, 16% female, 97% Caucasian , and 3% Asian-American.
Seventy-three percent were right-handed, 13.5% left-handed, and 13.5% had mixed hand
dominance. Fifty-nine percent of patients did not take medication; 22% were taking
one medication, 14% two medications, and 5% three medications. Of the initial 38 patients that
began the study, only one patient dropped out prior to completion of the study. Please refer to
Table 2 for demographics.
No significant differences were noted between the experimental and control group for age,
gender, race, handedness, number of medications, ATEC score, and other treatments. Eighty-
three percent of controls were males and 17% were females. All controls were Caucasian.
Seventy-five percent were right-handed; 17% left-handed; and 8% had mixed hand dominance.
Sixty-seven percent did not take medication; 17% were taking one medication; 8% were taking
two medications; and 8% were taking three medications. Please refer to Table 3 for
demographics. Over the course of the study, patients in the control group made no significant
changes in: Parental Judgment of Treatment Outcome, parent rating of symptom severity,
neuropsychological, or neurophysiological measures.
Neurofeedback for ASD
23
Parental Judgment of Treatment Outcome
Following treatment, improvement (decrease) in ASD symptoms was reported by parents
for 89% (n=33) of the experimental group [sign statistic = 33, p < .0001]. Eleven percent (n=4)
reported no change. All positive treatment outcomes reported by parents were confirmed by
neuropsychological and neurophysiological assessment. There were no reports of symptoms
worsening. The benefit to harm ratio was calculated at 89:1 exceeding all currently available
therapies or treatments for ASD.
Parent Ratings
Table 7 below for the pre-/post-treatment results of parent ratings of ASD
indicates that patients in our sample had initial ATEC Total Scores primarily
in the mild to moderate ranges of severity. A trend toward positive skewedness and lower initial
Total ATEC Scores associated with milder levels of ASD symptoms was noted (Shapiro-Wilk
Coefficient p= .0330). The majority of initial ATEC Scores (88%) were mild to moderate (0-
79
th
percentile), however there were six participants in the moderate to severe range (59
th
- 89
th
percentile).
Following neurofeedback training, a highly statistically significant reduction in
ASD symptomatology was reported on the ATEC [t (30) = 6.98, p < .0001] representing a 40%
reduction in ASD symptoms. This finding was confirmed by highly significant reductions in
ASD behaviors, executive deficits, and symptomatology associated with ASD following
treatment as reported on the: GADS [t (27)= 6.00, p < .0001], BRIEF [t (30)= 5.04, p < .0001],
and.the PIC-2 [t (32)= 6.28, p < .0001] as shown in Table 7.
Neurofeedback for ASD
24
-------------------
Insert Table 7
-------------------
Neuropsychological Testing
As indicated by Table 8 below, there were highly significant improvements for the
experimental group on composite measures of attention (n=20) [t (20)= -6.30, p < .0001], visual
perceptual functioning (n= 17) [t (17)= -7.79, p < .0001], and executive function (n=26)
[t (26)= -5.34, p < .0001]. Although the sample size for participants completing the language
assessment was small (n=4), improvement in language skills reached statistical significance as
well [t (4)= -3.25, p=.0474].
------------------
Insert Table 8
-----------------
Infrared (IR) Imaging: First Session
As shown in Table 9 below, the experimental group had a statistically significant
enhancement in the minimum or lowest thermal reading [t (34)= -2.25, p=.0313] and a highly
significant decrease in the range of thermal degrees [t (34)= 4.52, p < .0001] in the first session
of assessment. These findings indicate that even in the first session, patients in the
experimental group were able to elevate their metabolic activity and regulate the range or
variability of output.
--------------------
Insert Table 9
----------------------
Neurofeedback for ASD
25
IR Imaging: Last/20
th
Session
By the 20
th
session, there was a statistically significant decrease in the maximum thermal
reading [t (33)= 2.17, p= .0379] as well as a statistically significant decrease in the range of
thermal degrees [t (33)= 2.91, p=.0065] indicating a continuation of self-regulation of metabolic
activity or thermal regulation. Please refer to Table 10.
--------------------
Insert Table 10
--------------------
Evidence of Enduring Change: Comparison of First and 20
th
/ Last Session
As indicated by Table 11 below, throughout the course of treatment, the experimental group
significantly increased the minimum thermal reading [t (34)= -3.31, p=.0022] and significantly
reduced the range of thermal degrees [t (34)= 3.39, p= .0018]. The experimental group enhanced
metabolic activity (i.e., thermal regulation), regulated this output, and maintained these changes
by the 20
th
session of neurofeedback. Change in thermal regulation occurred both within
sessions and across sessions suggesting that change in metabolic regulation was enduring.
----------------------
Insert Table 11
----------------------
QEEG Connectivity
A total of 77% of the experimental group had either a decrease in hyperconnectivity
(n=15) or no change (n=5). Reduced hyperconnectivity patterns were statistically significant
[sign statistic = 15, p = .0392]. In this population, reduction in cerebral hyperconnectivity was
associated with positive clinical outcomes.
Neurofeedback for ASD
26
Predictors of Response to Therapy
As shown in Table 12 below, Kurtosis and Skewedness for the percentage of change in
ATEC Total Scores were not significant indicating an even spread of scores approximating a
normal distribution. Additional regression analyses ruled out confounding variables extraneous
to the effect of treatment (severity of ASD as measured by Pre-ATEC Total [F (1, 28)= .23, p=
.6338]; age [F (1, 28)= 1.83, p= .1868]; and number of medications [F (1, 28)= .46, p=.5014].
------------------------
Insert Table 12
-------------------------
Discussion
The major findings of our study included: A 40% reduction in core ASD symptoms, and 89%
of the experimental group had improved ratings of ASD symptomatology. Highly significant
improvement was noted for the experimental group on measures of attention, executive and
visual perceptual function. A significant increase also occurred in language skills. IR imaging
confirmed elevated metabolic activity even within the initial treatment session. Enduring change
was indicated by enhanced metabolic activity, regulation of output, and maintenance of changes
within and across the 20
th
treatment session. The benefit to harm ratio of 89:1, exceeded all
current treatments for ASD as surveyed by Rimland (2005). Seventy-seven percent of the
experimental group had either a decrease in hyperconnectivity patterns or no change. Reduced
hyperconnectivity as well as enduring change in metabolic activity confirmed neurophysiological
change following neurofeedback.
The experimental and control group were matched for age, gender, race, handedness,
other treatments, and severity of ASD. The variables extraneous to the treatment effect were
Neurofeedback for ASD
27
controlled and did not interact with the effect of assessment-guided neurofeedback. In addition,
regression analyses ruled out the effect of intervening variables (severity of ASD, age, and
number of medications) interacting with the treatment effect. Therefore, it was likely that
assessment-guided neurofeedback was the causative factor in improving ASD symptomatology
as confirmed by neurobehavioral, neuropsychological, and neurophysiological findings.
The purpose of our research was to replicate the previous controlled neurofeedback
study conducted by Jarusiewicz (2002). This is the second controlled study to
demonstrate improvement in the core symptoms of ASD following neurofeedback.
Our study provides support for positive treatment outcomes of neurofeedback for ASD.
The five levels of treatment efficacy which provide guidance for applied psychophysiologic
research have been outlined (Monastra, 2005) as follows:
Level 1
: “Not empirically supported” rating assigned to treatments supported by evidence
from only case studies in non-peer-reviewed journals and anecdotal reports.
Level 2: “Possibly efficacious” rating given to treatments investigated in at least one study with
sufficient statistical power and well-identified outcome measures but lacking randomized control
groups.
Level 3
: “Probably efficacious” rating assigned to treatments which demonstrate beneficial
effects in multiple observational studies, clinical studies, wait list control studies, and within-
subject and between-subject replication studies.
Level 4
: “Efficacious” rating given to treatment studies containing a no-treatment control,
alternative treatment, or placebo control group using randomized assignment proven statistically
superior to the control or equivalent treatment with well-defined procedures facilitating
replication. Positive treatment outcomes are confirmed by at least two independent studies.
Neurofeedback for ASD
28
Level 5: “Efficacious and specific” rating assigned to treatments that demonstrate
statistically superior results compared to a placebo, medication, or other treatment
in at least two independent studies.
Our research- the second controlled study to report a positive treatment outcome of
neurofeedback for ASD- supports neurofeedback as possibly efficacious; the second level of
efficacy rating as defined by the Association for Applied Psychophysiology & Biofeedback
(AAPB, 2006). This rating describes research containing sufficient statistical power, well
identified outcome measures, however lacking a randomized control group.
Our study may be the first step in establishing a Level 3 criteria rating of neurofeedback as
probably efficacious in the treatment of ASD. We replicated another controlled study
(Jarusiewicz, 2002). A broader range of outcome measures confirmed the reduction of ASD
symptomatology following neurofeedback. Further research is necessary utilizing randomized
control groups to establish neurofeedback as an efficacious treatment for ASD.
Our research, in contrast to Jarusiewicz’ (2002) study, demonstrated greater
improvement in clinical outcomes following assessment-guided Neurofeedback
reflected by a 40% compared to 26% reduction of ASD symptoms in fewer sessions
(20 versus an average of 36). This finding indicates a 54% increase in treatment efficacy and a
44% decrease in the number of sessions required for positive treatment outcome.
In contrast to the prior research conducted by Jarusiewicz (2002), the enhanced
treatment outcome of assessment-guided neurofeedback may be explained by the following
factors: 1) a milder degree of ASD in the experimental group; 2) utilizing multiple data points to
target specific brain regions for individualized neurofeedback protocols; 3) sequential (bipolar)
protocols in contrast to mostly unipolar protocols employed by Jarusiewicz (2002).
Neurofeedback for ASD
29
It is likely that the first factor- severity of ASD symptoms- can be excluded;
as previously discussed, regression analyses as well as the use of a control group ruled
out any interaction of this variable with the treatment effect. In addition, the reduction of
ASD symptomatology was also evident for patients (in the experimental group) with the
most severe ASD ratings.
The second factor, pertaining to the use of assessment-(primarily QEEG) guided
neurofeedback, may be a crucial factor in explaining the improved treatment outcomes.
Neurofeedback training protocols were based on the combined use of all assessment information
with a strong emphasis on initial QEEG analysis of absolute, relative power, and connectivity
measures. In contrast, Jarusiewicz (2002) utilized neurofeedback protocols based on symptom
complaints of patients. In our study, improved treatment outcomes resulted from assessment
providing multiple data points guiding the development of individualized neurofeedback
protocols targeting specific brain regions to increase activation and reduce hyperconnectivity.
The use of a sequential (bipolar) montage is another possible factor contributing to improved
treatment outcomes in our study. Sequential montages consisting of one active sensor site and
one reference site located over brain regions can reinforce interhemispheric communication
while reducing hyperconnectivity within and across brain regions. In contrast, Jarusiewicz
(2002) frequently utilized monopolar montages consisting of an active sensor site over a brain
region and a reference sensor on the ear which targets neurofeedback training to only one
brain region. Further research is needed to investigate the impact of sequential
compared to unipolar montages on treatment outcomes for neurofeedback in general
as well as protocols specific to individuals with ASD.
Our research found that decreased hyperconnectivity resulted in improved treatment
Neurofeedback for ASD
30
outcomes in an Autistic population. Individualized neurofeedback treatment protocols may
address patterns of hyperconnectivity as well as the heterogeneity characterizing ASD. Other
researchers investigated the impact of cortical hyperconnectivity on brain anatomy and function.
Belmonte et al.’s (2004) model of Autism is characterized by increased local
connectivity within the neural assemblies of a specific brain region while there is
decreased long-range connectivity with other brain regions. Courchesne & Pierce (2005)
described a pattern of over-connectivity (hyperconnectivity) within the frontal lobe and long-
distance disconnection (hypoconnectivity) between the frontal lobe and other brain regions
associated with ASD. . Reduction of long-distance cortical to cortical reciprocal activity and
coupling disrupts the integration of information from emotional, language, sensory, and
autonomic systems (Courchesne & Pierce, 2005).
Researchers also investigated the impact of mirror neurons on ASD symptomatology.
High functioning individuals with ASD failed to suppress Mu wave activity in the mirror neuron
system (MNS) as hand movement was observed, while, controls were able to suppress Mu wave
activity.(Oberman et al., 2005). Lack of MNS activity in area F5 (pars opercularis) was also
reported in children with Autism during imitation of emotional expression. Lack of MNS
activation during imitation and observation of emotional expression was associated with
dysfunction in social domains in both studies (Oberman et al., 2005; Dapretto et al., 2006).
Dysfunctional integration of the frontal lobes with other brain regions is frequently linked to
deficits in the executive system. The long-term consequences of deviation from patterns of
normal frontal lobe development are atypical patterns of brain connectivity (Hill, 2004).
. In SPECT scans of children with Autism, abnormal regional cerebral blood flow in the
medial prefrontal cortex and anterior cingulate gyrus was related to impaired communication and
Neurofeedback for ASD
31
social interaction. Altered perfusion in the right medial temporal lobe was associated with the
obsessive desire for sameness (Ohnishi et al., 2000). Functional neuroimaging studies have
linked social cognition dysfunction and language deficits in Autism to neural substrates (Just et
al. 2004; McAlonan et al., 2005: Pelphrey, Adolphs, & Morris, 2004; Welchew et al., 2005;.). In
a study utilizing diffusion tensor imaging, disruption of white matter tracts was associated with
social cognition found in the following regions: the fusiform gyrus and the superior temporal
sulcus linked to face and gaze processing and the anterior cingulate, amygdala, as well as the
ventromedial prefrontal cortex associated with awareness of mental states and emotional
processing. These impairments may disrupt neural connectivity required for children with
Autism to develop appropriate social skills (Barnea-Goraly et al., 2004).
The aforementioned research confirms that patterns of cortical connectivity have
a substantial impact on the social, emotional, and cognitive function of individuals with
ASD. Assessment-(primarily QEEG) guided neurofeedback targets brain regions to
reduce cortical hyperconnectivity. Our research findings indicated that significant
improvement in core ASD symptoms was achieved utilizing assessment-guided
neurofeedback.
In regard to the limitations of our study, the subjects consisted of a selected pool
of patients in the experimental group and a wait-list control group. When treatment is
selected by patients (via parents), there is the potential for selection bias to interact
with the treatment effect. Therefore, randomized assignment of treatment and control groups is
needed to confirm that there was no interaction between the treatment effect and subject
selection. In addition, comparison with an alternative treatment group would further establish
the efficacy of neurofeedback. Long-term follow-up would be beneficial to demonstrate that
Neurofeedback for ASD
32
positive treatment outcomes are maintained over time and we plan to include follow-up findings
in future research.
In light of the findings of this study and others regarding the links between
cortical connectivity patterns, reduced cerebral blood flow, and executive, behavioral,
as well as emotional/social functioning, it would be beneficial for future research to
further investigate interhemispheric connectivity (left vs. right hemisphere) comparisons
as well as intrahemispheric connectivity between the frontal, temporal, central, parietal,
and occipital lobes in Autism and other conditions. Further analysis of the QEEG
data will provide information regarding neurophysiological changes that occur as a
result of neurofeedback, and we intend to include these findings in future research.
Coherence is analogous to the squared correlation coefficient between a pair of
EEG waveforms, represented by the temporal voltage oscillations in each waveform.
The signals are normalized over the entire record to minimize the influence of signal
amplitudes and, thereby, emphasize the relationship between the pair of EEG profiles
(Bendat & Piersol, 1980). The exact equation for such a calculation can be found in Bendat
and Piersol (1980) equation 3.43.
Coherence anomalies have been associated with drug resistant epilepsy and mild closed
head injury. QEEG-guided coherence training is a form of neurofeedback that has been
successfully employed to normalize abnormal QEEG coherence in patients with mild
closed head injury and to reduce seizures in refractory epilepsy (Walker, Norman, & Weber,
2002; Walker, 2003).
Treatment goals are based on coherence anomalies identified by QEEG analysis.
Increased focal power in a frequency band or increased coherence between brain regions
Neurofeedback for ASD
33
may be downtrained while deficient focal power or decreased coherence between brain
regions may be uptrained (Walker, Norman,& Weber, 2002; Walker, 2003: Walker &
Kozlowski, 2005). The promising results demonstrated with QEEG-guided coherence training
warrant further research with other populations characterized by coherence anomalies such as
those with ASD. In addition, the specificity of neurofeedback treatment protocols for ASD may
be enhanced by identifying the effect of: unipolar and sequential montages, levels of absolute
and relative power for delta, theta, alpha, and beta activity associated with specific brain regions,
as well as exploring whether neurofeedback can alter activity in the mirror neuron system. It
would also be advantageous to further explore the impact of assessment-guided neurofeedback
on domains of executive, emotional, and behavioral function for groups of individuals with
varying functional levels of ASD (i.e., Severe vs. Moderate or Mild) in studies utilizing
randomized control groups.
Neurofeedback for ASD
34
References
Adams, J.B. & Holloway, C. (2004). Pilot study of a moderate dose
multivitamin/mineral supplement for children with autistic spectrum disorder.
The Journal of Alternative and Complementary Medicine, 10(6), 1033-1039.
American Psychiatric Association (1994). Diagnostic and statistical manual of mental
disorders, fourth edition. Washington, D.C.: Author.
Association for Applied Psychophysiology & Biofeedback (2006). Efficacy: How we rate the
efficacy of our treatments or how to know if our treatments actually work. Retrieved
February 22, 2006 from
http://www.aapb.org/i4a/pages/index.cfm?pageid=3336
Asperger, H. (1991). “Autistic psychopathy” in childhood. In Frith (Ed. and Trans.).
Autism and asperger’s syndrome (pp. 37-92). Cambridge, UK: Cambridge University
Press (Original work published 1944).
Attwood, T. (1998). Asperger’s syndrome: A guide for parents and professionals.
London, England: Jessica Kingsley Publishers.
Barnea-Goraly, N., Kwon, H., Menon, V., Eliez, S., Lotspeich, L., & Reiss, A.L. (2004).
White matter structure in autism: Preliminary evidence from diffusion tensor
imaging. Biological Psychiatry, 55, 323-326.
Baron-Cohen, S. (2004). The cognitive neuroscience of Autism. Journal of Neurology,
Neurosurgery & Psychiatry, 75, 945-948.
Belmonte, M.K., Allen, G., Beckel-Mitchener, A., Boulanger, L.M., Carper, R.A., &
Webb, S.J. (2004). Autism and abnormal development of brain connectivity.
The Journal of Neuroscience, 24(42), 9228-9231.
Bendat, J.S. and Piersol, A.G. (1980). Engineering applications of correlation and
Neurofeedback for ASD
35
spectral analysis. New York: John Wiley and Sons.
Blaxill, M.F. (2004). What’s going on? The question of time trends in autism. Public
Health Reports,119, 536-551.
Bradstreet, J., Geier, B.A., Kartzinel, J.J., Adams, J.B., & Geier, M.R. (2003).
A case-control study of mercury burden in children with autistic spectrum disorders.
Journal of American Physicians and Surgeons, 8(3), 76-79.
Carmen, J.A. (2004). Passive infrared hemoencephalography: Four years and 100
migraines. Journal of Neurotherapy, 8(3), 23-51.
Chez, M.G., Buchanan, C.P., Aimonovitch, M.C., Becker, M., Schaefer, K., Black, C.,
et al. (2002). Double-blind, placebo-controlled study of L-carnosine
supplementation in children with autistic spectrum disorders. Journal of Child
Neurology, 17, 833-837.
Coben, R., Carmen, J., & Falcone, A. (2005a, September). Advances in infrared
imaging. Presented at the 13
th
Annual Conference of the International Society of
Neuronal Regulation, Denver Colorado.
Coben, R. (2005b, September). Assessment-guided neurofeedback for autistic
spectrum disorder. Presented at the 13
th
Annual Conference of the International
Society of Neuronal Regulation, Denver Colorado.
Coben, R. (2005c, September). Passive infrared hemoencephalography for traumatic
brain injury. Presented at the 13
th
Annual Conference of the International Society of
Neuronal Regulation, Denver Colorado.
Committee on Children with Disabilities (2001). Technical report: The pediatrician’s
role in the diagnosis and management of autistic spectrum disorder in children.
Neurofeedback for ASD
36
Retrieved October 24, 2006 from
http://pediatrics.aappublications.org/cgi/content/full/107/5/e85
Courchesne, E. & Pierce, K. (2005). Why the frontal cortex in autism might be talking
only to itself: Local over-connectivity but long-distance disconnection. Current
Opinion in Neurobiology, 15, 225-230.
Cowan, J. & Markham, L. (1994, March). EEG Biofeedback for the attention problems of
autism: A case study. Presented at the Annual Meeting of the Association for Applied
Psychophysiology and Biofeedback.
Dapretto, M., Davies, M.S., Pfeifer, J.H., Scott, A.A., Sigman, M., Bookheimer, S.Y., et al.
(2006). Understanding emotions in others: Mirror neuron dysfunction in children with autism
spectrum disorders. Nature Neuroscience, 9(1), 28-30.
Delis, D.C., Kaplan, E., & Kramer, J.H. (2001). Delis-Kaplan executive function
system. San Antonio, TX: PsychCorp.
Deymed Diagnostic (2004). TruScan 32 Specifications. Retrieved March 24, 2006 from
http://www.deymed.com/truscan32.asp
FLIR Systems (2006). ThermoVision A20M Infrared Camera. Retrieved October 23,
2006 from
http://www.flirthermography.com/cameras/specs_pop.asp?camera_id=1033
Gardner, M.F. (1983). Expressive One-Word Picture Vocabulary Test Manual.
Novato, California :Academic Therapy Publications.
Gardner, M.F. (1990). Expressive One-Word Picture Vocabulary Test-Revised Manual.
Novato, California: Academic Therapy Publications.
Gilliam, J.E. (1995). Gilliam Autism Rating Scale Examiner’s Manual. Austin,
Texas: Pro-Ed
Neurofeedback for ASD
37
Gilliam, J.E. (2001). Gilliam Asperger’s Disorder Scale Examiner’s Manual. Austin, Texas:
Pro-Ed.
Gioia, G.A., Isquith, P.K., Guy, S.C., & Kenworthy, L. (2000). Behavior Rating
Inventory of Executive Function. Lutz, FL:.Psychological Assessment Resources, Inc.
Grass SafeLead (2006). Genuine Grass Precious Metal Recording Electrodes: SafeLead.
Retrieved October 23, 2006 from
http://www.grasstechnologies.com/
Green, V.A., Pituch, K.A., Itchon, J., Choi, A., O’Reilly, M., & Sigafoos, J. (2006).
Internet survey of treatments used by parents of children with autism. Research
in Developmental Disabilities, 27, 70-84.
Hamilton, L. (2000). Facing autism: Giving parents reasons for hope and guidance
for help. Colorado Springs, Colorado: WaterBrook Press.
Hill, E.L. (2004). Executive dysfunction in autism. Trends in Cognitive Sciences, 8(1),
26-32.
Holmes, A. (2001). Autism treatments: Chelation of mercury. Retrieved November 2,
2005 from
http://www.healing-arts.org/children/holmes.htm
Hudspeth, W.J. (1999). NeuroRep QEEG Analysis and Report System (Version 4.0)
[Computer Software]. Los Osos, CA: Grey Matter Inc.
Jarusiewicz, B. (2002). Efficacy of neurofeedback for children in the autistic spectrum:
A pilot study. Journal of Neurotherapy, 6(4), 39-49.
John, E.R., Prichep, L., Fridman, J., & Easton, P. (1988). Neurometrics: Computer-
assisted differential diagnosis of brain dysfunction. Science, 239, 162-169.
Just, M.A., Cherkassy, V.L., Keller, T.A., & Minshew, N.J. (2004). Cortical activation
and synchronization during sentence comprehension in high-functioning autism:
Neurofeedback for ASD
38
Evidence of underconnectivity. Brain, 127(8), 1811-1821.
Kanner, L. & Eisenberg, L. (1956). Early infantile autism 1943-1955. American Journal
of Orthopsychiatry, 26, 55-65.
Kirby, D. (2005). Evidence of harm: Mercury in vaccines and the autism epidemic: A
medical controversy. New York: St. Martin’s Press.
Korkman, M., Kirk, U., & Kemp, S. (1998). A Developmental Neuropsychological
Assessment Manual. San Antonio, TX: The Psychological Corporation.
Lachar, D. & Gruber, C.P. (2001). Personality Inventory for Children, second edition.
Los Angeles, CA:Western Psychological Services.
Larson, S.L., Buethe, E., & Vitali, G. (1990). Comprehensive Test of Visual
Functioning. East Aurora, New York: Slosson Educational Publications, Inc..
Lezak, M.D. (1995). Neuropsychological assessment, third edition. New York: Oxford
University Press.
Linden, M., Habib, T., & Radojevic, V. (1995). A controlled study of the effects of EEG
biofeedback on cognition and behavior of children with attention deficit disorder
and learning disabilities. Biofeedback and Self-Regulation, 21, 35-50.
Lubar, J.F., Swartwood, M.O., Swartwood, J.N., & O’Donnell, P.H. (1995). Evaluation
of the effectiveness of EEG training for ADHD in a clinical setting as measured by TOVA
scores, behavioral ratings, and WISC-R performance. Biofeedback & Self-Regulation, 20(1),
83-99.
McAlonan, G.M., Cheung, V., Cheung, C., Suckling, J., Lam, G.Y., Tai, K.S., et al. (2005).
Mapping the brain in autism: A voxel-based MRI study of volumetric differences and
intercorrelations in autism, Brain, 128 (Pt2), 268- 276.
Neurofeedback for ASD
39
McCandless, J. (2005). Children with starving brains: A medical treatment guide for
autism spectrum disorder. Putney, VT: Bramble Books..
Medical Research Council (2001). Medical Research Council Review of Autism
Research: Epidemiology and causes. Retrieved October 23, 2006 from
http://www.mrc.ac.uk/Utilities/Documentrecord/index.htm?d=MRC002394
Meyers, J.E. & Meyers, K.R. (1995). Rey Complex Figure Test and Recognition Trial
Professional Manual. Odessa, Florida: Psychological Assessment Resources, Inc.
Monastra, V.J. (2005). Electroencephalographic biofeedback (neurotherapy) as a
treatment for attention deficit hyperactivity disorder: Rationale and empirical
foundation. Child and Adolescent Psychiatric Clinics of North America, 14,
55-82.
Monastra, V.J., Lynn, S., Linden, M., Lubar, J.F., Gruzelier, J., & LaVaque, T.J.
(2005). Electroencephalographic biofeedback in the treatment of attention-deficit/
hyperactivity disorder. Applied Psychophysiology and Biofeedback, 30(2), 95-114.
NeuroCybernetics Inc. (2006). Specific EEGer Technical Parameters. Canoga Park, CA:
NeuroCybernetics, Inc.
NxLink (2001). Neurometric analysis system. Richland, WA: NxLink Ltd.
Oberman, L.M., Hubbard, E.M., McCleery, J.P, Altschuler, E.L., Ramachandran, V.S., &
Pineda, J.A. (2005). EEG evidence for mirror neuron dysfunction in autism
spectrum disorders. Cognitive Brain Research, 24(2), 190-198.
Ohnishi, T., Matsuda, H., Hashimoto, T., Kunihiro, T., Nishikawa, M., Uema, T., et al. (2000).
Abnormal regional cerebral blood flow in childhood autism. Brain, 123(9), 1838-1844.
Pelphrey, K., Adolphs, R., & Morris, J.P. (2004). Neuroanatomical substrates of social
Neurofeedback for ASD
40
cognition dysfunction in autism. Mental Retardation and Developmental Disabilities
Research Reviews,10, 259-271.
Reichelt, K. & Knivsberg, A.M. (2003, October). Why use the gluten-free and casein-
free diet in autism and what the results have shown so far: Peptides and autism. Presented
at the Defeat Autism Now Conference, Portland, Oregon.
Research Units on Pediatric Psychopharmacology Autism Network (2005a). Risperidone
treatment of autistic disorder: Longer-term benefits and blinded discontinuation after
6 months. American Journal of Psychiatry, 162(7), 1361-1369.
Research Units on Pediatric Psychopharmacology Autism Network (2005b).
Randomized, controlled, crossover trial of methylphenidate in pervasive
developmental disorders with hyperactivity. Archives of General Psychiatry, 62(11),
1266-1274.
Rimland, B (2005). Parent ratings of behavioral effects of biomedical interventions
Retrieved October 23, 2006 from
http://www.autismwebsite.com/ari/treatment/
Rimland, B. & Edelson, S.M. (2000). Autism treatment evaluation checklist (ATEC).
Retrieved October 23, 2006 from
http://www.autismeval.com/ari-atec/report1.html
Roberts,W., Weaver, L., Brian, J., Bryson, S., Emelianova, S., Griffiths, A.M,
et al. (2001). Repeated doses of porcine secretin in the treatment of autism: A randomized,
placebo-controlled trial. Retrieved October 24, 2006 from
http://pediatrics.aappublications.org/content/vol107/issue5/index.shtml
Rosenthal, R. & DiMatteo, M.R. (2001). Meta-analysis: Recent developments in quantitative
methods for literature reviews. Annual Review of Psychology, 52, 59-82.
Sallows, G.O. & Graupner, T.D. (2005). Intensive behavioral treatment for children with autism:
Neurofeedback for ASD
41
Four-year outcome and predictors. American Journal of Mental Retardation, 110, 417-438.
Sanford, J.A. & Turner, A. (2002). Integrated Visual and Auditory Continuous
Performance Test Manual. Richmond, VA: BrainTrain..
Scolnick, B. (2005). Effects of electroencephalogram biofeedback with asperger’s
syndrome. International Journal of Rehabilitation Research, 28(2), 159-163.
Sichel, A.G., Fehmi, L.G., & Goldstein, D.M. (1995). Positive outcome with
neurofeedback treatment in a case of mild autism. Journal of Neurotherapy,1(1),
60-64.
Sicile-Kira, C. (2004). Autism spectrum disorders: The complete guide to
understanding autism, asperger’s syndrome, pervasive developmental disorder, and
ASDs. New York, New York: The Berkley Publishing Group.
Siegel, B. (1996). The world of the autistic child: Understanding and treating autistic
spectrum disorders. New York, New York: Oxford University Press:.
Smith, T., Groen, A.D., & Wynn, J.W. (2000). Randomized trial of intensive early
intervention for children with pervasive developmental disorder. American Journal on
Mental Retardation, 105, 269-285.
Tansey, M.A. (1993). Ten-year stability study of EEG biofeedback results for a hyperactive
boy who failed a fourth grade perceptually impaired class. Biofeedback & Self-
Regulation, 18(1), 33-44.
Thatcher, R.W., Walker, R.A., Biver, C.J., North, D.N., & Curtin, R. (2003).
Quantitative EEG normative databases: Validation and clinical correlation. In
Lubar, J.F. (Ed.) Quantitative electroencephalographic analysis (QEEG) databases for
neurotherapy: Description, validation, and application (pp. 87-121). New York: Haworth
Neurofeedback for ASD
42
Press.
Thompson, L. & Thompson, M. (1995). Exceptional results with ‘exceptional
children.’ Presented at the Annual Conference of the International Society for
Neuronal Regulation.
Toomim, H., Mize, W., Kwong, P.C., Toomim, M., Marsh, R., Kozlowski, G.P.,
et al. (2004). Intentional increase of cerebral blood oxygenation using hemoencephalography
(HEG): An efficient brain exercise therapy. Journal of Neurotherapy, 8(3), 5-21.
Walker, J.E., Norman, C.A., & Weber, R.K. (2002). Impact of QEEG-guided
coherence training for patients with a mild closed head injury. Journal of
Neurotherapy, 6(2), 31-45.
Walker, J.E. (2003, September). Long term remediation of seizures in
refractory epilepsy with QEEG-guided neurofeedback training. Presented at the
11
th
Annual Conference of the International Society for Neuronal Regulation, Houston, TX.
Walker, J.E. & Kozlowski, G.P. (2005). Neurofeedback treatment of epilepsy.
Child and Adolescent Psychiatric Clinics of North America, 14, 163-176.
Welchew, D.E., Ashwin, C., Berkouk, K., Salvador, R., Suckling, J., Baron-Cohen, S.,
& Bullmore, E. (2005). Functional disconnectivity of the medial temporal lobe in
asperger’s syndrome. Biological Psychiatry, 57, 991-998.
Wing, L. & Gould, J. (1979). Severe impairments of social interaction and associated
abnormalities in children: Epidemiology and classification. Journal of Autism and
Developmental Disorders, 9, 11-29.
World Health Organization (1993). The ICD-10 classification of mental and behavioral
disorders: Diagnostic criteria for research. Geneva: Author.
Neurofeedback for ASD
43
Appendix
____________________________________
Table 1: Benefit to Harm Ratios
__________________________________________
Treatments Ratios
_____________________________________________________
Risperidal 3.0: 1
Ritalin 0.7:1
Haldol 0.9: 1
Thorazine 0.7: 1
B6 with Magnesium 10: 1
Digestive Enzymes 20: 1
Intravenous Secretin 6.7: 1
Gluten-/Casein-Free Diet 20: 1
Chelation 35: 1
____________________________________________________
Note. All benefit to harm ratios listed were reported by Rimland
(2005) based on parent ratings of biomedical interventions.
________________________________________________
Table 2: Demographics of Neurofeedback Group
________________________________________________
.
Total
Age Gender Race Handedness Number of Meds ATEC Score
________________________________________________
Mean 31 Males 36 Caucasian 27 Right 22 None Mean
8.92 years 45.161
6 Females 1 Asian- 5 Left 8 One
Range American Range
3.92- 5 Mixed 5 Two 12-100
14.66 years
2 Three
________________________________________________
Note. Total ATEC Score was computed from the Autism Treatment Evaluation Checklist
(ATEC; Rimland & Edelson, 2000).
Neurofeedback for ASD
44
_______________________________________________________________________
Table 3: Demographics of Control Group.
________________________________________________
Total
Age Gender Race Handedness Number of Meds ATEC Score
________________________________________________
Mean 10 Males 12 Caucasian 9 Right 8 None Mean
8.5 years 44.32
2 Females 2 Left 2 One
Range Range
4.26- 1 Mixed 1 Two 16-92
14.07 years
1 Three
________________________________________________
Note. Total ATEC Score was computed from the Autism Treatment Evaluation Checklist
(ATEC; Rimland & Edelson, 2000).
___________________________________________
Table 4: ASD Diagnoses for the Neurofeedback Group
___________________________________________
Autism PDD-NOS CDD Asperger’s Disorder
___________________________________________
7 21 4 5
________________________________________________________________
Note. ASD=Autistic Spectrum Disorder; PDD-NOS=Pervasive Developmental
Disorder-Not Otherwise Specified; CDD=Childhood Disintegrative Disorder.
Neurofeedback for ASD
45
_______________________________________________
Table 5: Specifications for:
TruScan 32 NeuroCybernetics EEGer
EEG System* Neurofeedback Training System**
_________________________________________________
Number
of Channels 32 2 channels of EEG
data at 256 Hz
Sampling 128, 256, All sampling is done by external
Stored data 512, or 1024 Hz. EEG amplifiers/ converters at 256 Hz.
Analog 4,096 Hz
Sampling per channel.
Frequency
Encoders Thought Technology Encoders
Maximal
Input DC Offset: +
250 mV
Filtering Equivalent Filter coefficients were precomputed
input noise and provided in 1/8 Hz steps
is 1 mVp-p. from 0 to 50 Hz.
0.1 Hz- Lowpass filters input can be independently
100 Hz specified as 0-40, 0-50, 0-30 Hz to
with minimize 50 or 60 Hz interference.
impedance
below 10 K
ohm.
Common
Mode Rejection
Ratio:102 dB. In
Bandwidth 0-60
Hz with all inputs
shorted to ground.
Isolation Mode
Rejection Ratio:
140 dB.
Power
Source: Four AA Batteries
_____________________________________________
Note. Specifications for equipment were obtained from: * Deymed Diagnostic (2004)
and ** NeuroCybernetics Inc. (2006).
Neurofeedback for ASD
46
________________________________________________
Table 6: Specifications for ThermoVision A20M
________________________________________________
Field of View: 25 degrees X 19 degrees/
0.3 m.
Detector
Type: Focal plane array (FPA)
uncooled microbolometer.
Spectral Range: 7.5 to 13 microns
Thermal Sensitivity: At 50/60 Hz : 0.12
degrees C at 30 degrees C.
Accuracy (% of reading): +
2 degrees C or + 2%.
Individual Emissivity Settings: Individually settable.
Measurement Corrections: Reflected ambient,
distance, relative humidity,
external optics. Automatic,
based on user input.
Power
Source: AC operation: AC adapter
110/220 VAC. 50/60 Hz
(included). DC operation:
12/24V nominal , <6W.
________________________________________________
Note. Specifications were obtained from FLIR Systems Inc. (2006).
_____________________________________________________________________
Table 7: Parent Ratings for Neurofeedback Group
______________________________________________________________
Initial Total ATEC %ile Severity
Range= 28.000-56.500 9
th
-39
th
%ile Mild-Moderate
Pre-ATEC Total Post-ATEC Total Significance (p)
Mean=46.100 Mean=27.733 p < .0001
Pre-GADS ADQ Post-GADS ADQ Significance (p)
Mean=83.852 Mean=72.519 p < .0001
Pre-BRIEF GEC Post-BRIEF GEC Significanc (p)
Mean=71.700 Mean=64.767 p < .0001
Pre-PIC-2 TOTC Post-PIC-2 TOTC Significance (p)
Mean=71.250 Mean=64.250 p < .0001
________________________________________________________________________
*Note. ATEC=Autism Treatment Evaluation Checklist; GADS ADQ=Gilliam
Asperger’s Disorder Scale Asperger’s Disorder Quotient; BRIEF GEC=Behavior
Rating Inventory of Executive Function Global Executive Composite; PIC-2 TOTC
=Personality Inventory for Children Second Edition Total Composite
Neurofeedback for ASD
47
______________________________________________
Table 8: Neuropsychological Testing* for Neurofeedback Group
_____________________________________________________________________
Pre-Attention Post-Attention Significance (p)
Mean z= -1.694 Mean z=-0.518 p < .0001
Pre-Visual Perceptual Post-Visual Perceptual Significance (p)
Mean z= -2.445 Mean z= -1.442 p < .0001
Pre-Executive Post-Executive Significance (p)
Mean z= -1.699 Mean z= -0.741 p < .0001
Pre-Language Post-Language Significance (p)
Mean z= -1.588 Mean z= -0.663 p=.0474
_____________________________________________________________________
*Note. All neuropsychological testing consisted of composite scores for indices of
attention, visual perceptual, executive, and language domains.
_________________________________________________________________________
Table 9: Pre-/Post-IR Imaging in the First Session for the Neurofeedback Group
_________________________________________________
1
st
Pre-Min Mean 1
st
Post-Max Mean Significance (p)
93.52 93.90 .0313
1
st
Pre-Max Mean 1
st
Post-Max Mean Significance (p)
97.30 97.19 .5046
1
st
Pre-Range Mean 1
st
Post-Range Mean Significance (p)
3.77 3.29 < .0001
________________________________________________________________________
* Note. Min=Lowest thermal reading; Max=Highest thermal reading.
________________________________________________________________________
Table 10: Pre-/Post-IR Imaging of Last/20
th
Session for the Neurofeedback Group
________________________________________________________________________
20
th
Pre-Min 20
th
Post-Min Significance (p)
Mean= 94.33 Mean= 94.11 .3661
20
th
Pre-Max 20
th
Post-Max Significance (p)
Mean = 97.75 Mean= 97.26 .0379
20
th
Pre-Range 20
th
Post-Range Significance (p)
Mean = 3.41 Mean= 3.16 .0065
______________________________________________________________________
* Note. Min=Lowest thermal reading; Max=Highest thermal reading.
Neurofeedback for ASD
48
_______________________________________________________________________
Table 11: Pre/Post -IR Imaging: Comparison of 1
st
and 20
th
Session
_______________________________________________________________________
1
st
Pre-Min Mean 20
th
Pre-Min Mean Significance (p)
93.52 94.33 .0022
1
st
Pre-Max Mean 20
th
Pre-Max Mean Significance (p)
97.30 97.75 .0654
1
st
Pre-Range Mean 20
th
Pre-Range Mean Significance (p)
3.77 3.41 .0018
___________________________________________________________________
* Note. Min=Lowest thermal reading; Max=Highest thermal reading.
____________________________________________________________________
Table 12: Predictors of Response to Therapy
____________________________________________________________________
ATEC Total Kurtosis Skewedness
Mean= 38.770 p=.4419 p=.4295
Median= 38.750
Range= 20.000-52.543
Pre-ATEC Total R
2
Significance (p)
.01* .6338
Age .06* .1868
Number of Medications .02* .5014
______________________________________________________ ____
Note.. ATEC=Autism Treatment Evaluation Checklist. *R
2
= percentage of total
variance in percentage of change in ATEC Total Score