Transcript EEG Spectrum Science Council: Neurofeedback for Autistic

EEG Operant Conditioning as a
treatment for Autistic disorders
Robert Coben, PhD
Associate Fellow, EEG Biofeedback (BCIA)
Diplomat, qEEG Certification Board
Massapequa Park, New York
Presented at Autism One 2008
CNS changes in early development
• Autistic disorders are viewed as problems of
early childhood Neuroinflammation impacting
multiple systems including the CNS.
• This inflammation interferes with normal
development of neural connectivities in the
developing brains of these children.
White matter anomalies in Autism
Connectivity in Autism
• MRI reductions in white matter (McAlonan et al., 2004)
• fMRI underconnnectivity in anterior-posterior connections
(Cherkassky et al., 2006)
• fMRI hyperconnectivity across middle frontal regions
(Mizuno et al., 2006).
• Theory of hyperconnected frontal cortices along with frontal
to other hypoconnectivity (Courchesne & Pierce, 2005).
• Less white matter concentration in the genu, rostrum,
splenium (Chung et al., 2004)
• Cell columns are more numerous, smaller and less compact in
frontal and temporal regions (Casanova et al, 2002)
• Diminished connectivity in language areas during sentence
comprehension (Just et al., 2004)
fMRI connectivity
EEG Connectivity
What is EEG operant conditioning
(biofeedback)?
Case Example: Mu NF
20
3.5
18
3.4
16
14
3.3
12
10-12
19-23
7-15
10
8
3.2
3.1
6
3
4
2
2.9
0
1 2 3 4 5 6 7 8 9 1
2.8
Case Example: Mu NF
Case Example: Topographical changes
EEG biofeedback: Efficacy in ADHD
EEG biofeedback: Efficacy in ADHD
• DeBeus (2006)
•
•
•
Attention Training with ADHD
Children: A Double-Blind PlaceboControlled Study
N = 60, with crossover
Attention Scores:
•
Effectiveness of Intervention; P = .0004
•
•
Treatment Effect; P = .0004
Parent Ratings:
•
Effectiveness of Intervention; P = .0002
•
Treatment Effect; P < .0001
Monastra, Monastra,
George (2002)
9
8
7
6
5
4
3
2
1
0
EEGBF
Ritalin
PRE
POST
F/U
Empirical Evidence?
• Most of the work in this area has been based on
case studies and case series.
• Two published controlled studies.
• Others in development.
• No Randomized Controlled Trial (yet).
• Must be evaluated in context and compared to
other forms of intervention.
st
1
Pilot Study
• Jarusiewicz (2002) – Only research based on group data.
12 of 20 participants completed at least 20 sessions (20
– 69, mean = 36). Results showed 26%
improvement/reduction in ATEC symptoms vs. 3%
reduction for control group. Protocols began at C4
(57%) and were individualized based on symptoms. C4
with F7 was used for 75% and the others included F3-F4
and/or T3-T4.
Methods
• 37 ASD children compared
with 12 WLC
• Matched for age, gender, race,
handedness, medications,
symptom severity
• EEGBF was assessment
guided based on EEG
connectivity et al.
• EEGBF done twice weekly for
10 weeks
• Pre-post parent judgment,
rating scales, NP and QEEG
25
20
15
10
5
0
Autism
PDD-nos
CDD
Aspergers
Findings
• 89% reported success
• No reports of worsening
• 40% reduction in autistic
symptoms (ATEC)
• Significant changes in
symptoms, NP findings
and EEG connectivity
• Reduced neural
hyperconnectivity
• Tx response not predicted
by age, medications,
severity of symptoms
EEG connectivity changes
The relative efficacy of connectivity guided and symptom based
EEG biofeedback for Autistic disorders
•
•
•
Compared
findings from
Jarusiewicz
(2002) to
Coben &
Padolsky’s
(2007)
approach
Matched
subjects for
severity of
symptoms and
equated the
sample sizes
Both are
effective, but tx
guided by
EEG
connectivity
appears more
effective
50
3
45
40
2.5
35
2
30
Jarusiewicz
Coben/Padolsky
25
Jarusiewicz
1.5
Coben/Padolsky
20
15
1
10
0.5
5
0
Sens/Cogn
Health/Behavior
Total
0
speech
social
cognitive
behavior
Autistic Spectrum
Disorder: A
Controlled Study of
EEG Coherence
Training focused on
Social Skill Deficits
Robert Coben, PhD
Presented at 2007 ISNR Conference
San Diego, California
Facial/Emotional Processing
Facial/Emotional Processing
Facial/Emotional Processing
Facial/Emotional Processing deficits in
Autism
Facial/Emotional Processing deficits in
Autism
Efficacy of Social Skills Training
• Rao, Beidel, & Murray (2007) recently reviewed
research related to social skills training in
autistic disorders and concluded that empirical
support is minimal at this time.
• Bellini et al. (2007) have also reviewed social
skills training programs. They calculated PND
(% of non-overlapping data points). Mean
intervention effects were 70% (questionable)
and generalization 53% (low).
Theory and Hypotheses
• Social skill deficits in ASD are, at least partially, related to the
neural substrate of visual/facial/emotional processing.
• Altering this neural substrate should then lead to improvements
in social skills.
• H1: NF (coherence training) can improve visual processing and
social skills.
• H2: Improvements in visual processing would predict
enhancements in social skills.
• H3: EEG analyses will show associated improvements in the
neural substrate responsible.
Method
• 50 patients diagnosed with Autistic Spectrum
Disorder
• All underwent Neuropsychological (focus on visual
processing) testing, ratings of social skill deficits,
and QEEG Assessment prior to intervention
• All underwent follow-up Neuropsychological, rating
scale and QEEG assessment following intervention
• Two groups were studied
– 25 patients received 20 sessions of QEEG
Connectivity guided EEG coherence NF
– 25 patients wait list controls
Subjects
Experimental
Controls
pvalue
Age
9.5, 2.38
10.13, 2.72
0.39
Gender
21 male
21 male
1.0
Race
23 caucasian
23 caucasian
1.0
Handedness
24 right handed
22 right handed
0.30
Medications
0.16, 0.37
0.16, 0.37
1.0
ATEC
31.00, 9.16
31.16, 6.36
0.94
Social Skills
70.3, 7.39
69.96, 6.33
0.86
Visual-Perception
-1.63, 1.13
-1.52, 1.07
0.73
Skewness = .185
Kurtosis = -1.192
Skewness = .275
Kurtosis = .304
Neurofeedback Protocol Design –
Sites of Coherence training
Neurofeedback Protocol Design
Results
ANOVA
ATECchange
Sum of
Squares
Between Groups 714.420
Within Groups 4403.200
Total
SocSkillschange
df
Mean Square
1
714.420
48
Sig.
.008
4.625
.037
15.967
.000
91.733
5117.620
Between Groups 181.451
49
1
181.451
Within Groups 1883.254
Total
2064.705
48
39.234
VisualProcesschange Between Groups
F
7.788
49
6.690
1
6.690
Within Groups
20.113
48
.419
Total
26.803
49
Results
Results
Results
• Social Skills change Effect
Size = 0.61
• One study found social skills
training to have an ES of
0.33
• ATEC change Effect Size =
0.79
• Medium to Large
• PND (% of non-overlapping
data points) = 88%
• Social skills training has
mean PND = 70% and
generalization = 53%
Results
Results
Results
Do the changes last?