Subtype Analysis of LD by QEEG pattern analysis, Biofeedback, 34 (3), 2006, 106-114.
Abstract:The article addresses the growing prevalence and expense of learning disabilities. The author re- views the various current diagnostic classication systems based on psycho-educational, DSM IV and other classication schemes. He argues that none of the current classications clearly dictates precise interventions by diagnostic category. He advocates for a functional classication based on the use of modern neurosci- ence techniques. He reviews the emerging research using quantitative EEG, showing clear EEG signatures of several learning disabilities, and advocates for developing a more precise link between classication and therapeutics.
Key words: Learning Disabilities, quantitative EEG, subtype analysis
The Magnitude of the Problem
"Learning disabled is our number one area of growth, and it is inappropriately used as a catch-all di- agnosis" says Troy Justesen, acting assistant secretary of Special Education & Rehabilitation Services for the Education Dept. "We are working very hard to place a stronger emphasis on early and better identication." (Business Week, May 31, 2004)
Various projections have been generated over the years on the costs of IDEA for special education (IDEA, 2004, Public Law 108-446). The federal government has spent between $460 and $500 billion on special education since 1975 (Wood, 1998). The total estimated cost, over a lifetime, of having the ADHD diagnoses (to insurance companies, parents, subject) is $625,133 (Thornton, 2006). Estimates on the lifetime cost of a learning disability have not been addressed separately, from other diagnoses, in the literature.
Beyond the immediate educational cost issue for these students, we encournter the long-term societal effects, especially in our prison systems. Between 28% and 43% of inmates require special education in adult correctional facilities (vs. 5% in the normal population) and 82% of prison inmates in the U.S. are school dropouts. (Leahy, 2001)
The DSM-IV (Diagnostic and Statistical Manual) provides several classications for the special edu- cation child. These include Learning Disability, NOS (DSM-IV code 315.90), retardation (mild to severe, DSM-IV 317 to 319), reading disorder (DSM-IV 315.00), mathematics (DSM-IV 315.1), Attention-Decit/ Hyperactivity Disorder NOS (DSM-IV 314.90), Disorder of Written Expression (DSM-IV 315.2), Expressive Language Disorder (DSM-IV 315.31), Mixed Receptive-Expressive Language Disorder (DSM-IV 315.31), Phonological Disorder (DSM-IV 315.39), and Developmental Coordination Disorder (DSM-IV 315.40)
The IDEA denes specic learning disability in terms of "imperfect ability to listen, think, speak, read, write, spell or do mathematical calculations." (Public Law 108-446, 118, Stat. 2658).
The considerable heterogeneity of the LD population has lead to a general consensus among profes- sionals that many subtypes exist within the broad category of learning disabilities, and that the boundaries between the subtypes often overlap.
About one-half of the 1.6 million elementary school-aged children (ages 6-11) diagnosed with atten- tion-decit/hyperactivity disorder (ADHD), have also been identied as having a learning disability (LD), according to a study (1997-8) by the Centers for Disease Control and Prevention (CDC) - (CDC, 2003).
The reported association between learning disability and ADHD varies from 10 per cent to 92 per cent, depending on how the research is conducted (tests used, subjects selected, selection criteria, etc.) (Biederman et al., 1991). Gerhardstein et al., (2001) reported an overlap of 50% between reading disabilities and ADHD. Recent research has tended to divide students with LD into two main groups: verbal learning disabilities and nonverbal learning disabilities (Harnadek & Rourke, 1994). Individuals in the verbal disability group have verbal/language decits or specic phonological processing decit as well as impaired reading, written lan- guage and/or spelling skills, while those in the non-verbal group may have problems in visual-spatial-organi- zational, tactile-perceptual, psychomotor, and/or nonverbal problem solving skills. In addition, the non-verbal LD student may have problems in computational mathematics and/or writing skills, and is at an increased risk for social and behavioral problems (Harnadek & Rourke, 1994; Torgenson, 1993).
Additional categorical diagnostic entities have been introduced over the years. The central auditory processing disorder (CAPD) is a popular concept which employs the following auditory criteria: difculties in phonological awareness, auditory discrimination, auditory memory, sequencing and blending. (National Center for Learning Disabilities, 2006).
Interventions Based upon Typologies -- Medical Models
Current psycho-educational categorical typologies operate under a medical model assumption. The model asserts that if the correct diagnosis is rendered, then an appropriate (research based) intervention ap- proach is the appropriate intervention. The limitations of this approach is the assumption that there is an appropriate effective intervention approach for the particular problem. Research support for current psycho- educational intervention models is limited in terms of their documented research effectiveness. For example, although phonics has been demonstrated to be an important element in reading research (Wagner & Torgesen,
1987), support for its effectiveness in improving reading comprehension is limited (Clay, 1990).
Implicit in the psychoeducational model is the neuropsychological concept of double dissociation. The concept asserts that lesion A causes cognitive decit B (with no effect on cognitive decit D) and that lesion C causes cognitive decit D, with no effect on decit B. Thus, separate physical problems underlie separate cognitive problems. For example, a mathematical problem is not related to a reading problem or vice versa. In support of this hypothesis is the subject in the movie Rain Man. This individual has shown us that the brain can excel in certain functions and yet be unable to function in society. This type of phenomena would argue for some functional anatomical or frequency specicity. However, apart from this type of extreme example, Teuber's (1955) double-dissociation theory has had difculty in producing clinically useful examples.
The advent of modern neurodiagnostic techniques has opened up an important method to distinguish among the LD groups. Those seeking to understand neural correlates of learning disability can now draw on functional magnetic resonance imaging (fMRI), positron emission tomography (PET); diffusion tensor imag- ing (DTI), quantitative EEG (QEEG), and others. have opened up an important method to distinguish between the LD groups. With this new medical technology, more precise physical denitions of LD subtypes may become possible.
For the dyslexic subject, the physical dysfunction is located in the superior temporal gyrus (T3) and inferior parietal cortex (P3-T5), particularly in the left hemisphere (Rumsey et al., 1997a; Rumsey et al.,
A number of fMRI neuroimaging studies have compared cortical activation patterns under reading related tasks in readers with dyslexia (DYS) and control groups of nonimpaired readers (NI) (Shaywitz et al., 1998; Shaywitz et al., 2002; Shaywitz et al., 2003). This series of studies showed that non-impaired adults increased their activation at T5 - (posterior superior temporal gyrus), P3 (angular gyrus), and above P3 (supra- marginal gyrus) as the task demands increased from orthographic comparisons to phonological comparisons (Shaywitz et al., 1998). In contrast, dyslexic adults showed over-activation in response to increasing task demands in anterior regions including the F7 location (inferior frontal gyrus). While the nonimpaired readers showed activation of a widely distributed system for reading, the dyslexic readers had disrupted activity in the posterior cortex that involves traditional attentional, visual and language areas.
The anatomic correlates of the reduced function in left temporoparietal regions can also be visualized by diffu- sion tensor imaging (DTI), which identies the white matter tracts (Watts et al., 2003). Using DTI, Klingberg et al., (2000) showed that reading ability is directly related to the degree of connection activity among T3, T5, and P3 (water diffusion of the long myelinated neurons connecting different regions of the brain) for both dyslexic and nonimpaired readers.
In conclusion, modern neuroimaging techniques have provided the beginning of a physical denition for reading disability by identifying the left temporal and left posterior locations as critical areas of neural decits.
Quantitative EEG (QEEG) research has provided additional physical perspectives on the learning dis- ability situation. Evans and Park (1996) identied signicant deviations from a normative database in a group of 8 dyslexic children and 2 adults. These deviations were evident in the left posterior region (in particular the P3 position) and revealed reduced coherence values, usually involving the theta bandwidth (4-8 Hertz).
QEEG studies of the ADHD population have shown two distinct types of patterns. The rst is a slow- ing over frontal and central-midline cortical regions in approximately 80-90% of patients with ADHD (Mann et al., 1992; Chabot et al., 1996; Clarke et al., 2001a). Slowing in qEEG terms refers to elevated (compared to a normative reference group) values of relative power of theta, reduced relative power values of alpha and beta, and elevated theta/alpha and theta/beta ratios, primarily over frontal, frontal-midline (Fz), and central- midline (Cz) cortical regions. This is a hypoarousal condition and occurs in about 80% of the subjects (Chabot et al., 2001; Barry et. al., 2003). The other subtype involves hyperarousal of the frontal lobes. Individuals with this pattern also meet the criteria for ADHD but are more likely to exhibit conduct, mood or anxiety disorders (Clarke et. al., 2001b) and may have ADHD symptoms that are secondary to these other DSM-IV diagnoses.
Thus, QEEG variables assessing frequencies, connection patterns and locations have proven useful in the diagnostic situation.
Proposed Sub-Type Analysis
The following approach can be employed to further advance the classication issue. QEEG variables can be grouped according to two general classes of variables across 4 conceptual categories. The general classes of QEEG variables would include: 1) - the absolute levels of the 4 main frequencies (delta, theta, alpha, beta) across different measurement variables (microvolts, relative power, peak frequency, peak ampli- tude, spectral power) and 2) - absolute levels of connectivity values (coherence, phase) across the different frequencies.
The conceptual categorical distinctions would involve: 1) - the cognitive activity; 2) -the relevant loca- tions; 3) - the absolute value of the variable; and 4) - the degree of activation from a relevant cognitive activity. For the degree of activation variable, the appropriate comparison to the reading condition could be a visual at- tention condition. The comparison would separate out the effects of just looking at a page from processing the written words. This type of methodology is employed in PET studies. It would need to be conrmed, however, that the activation pattern is related to success at the task. Presently, PET studies just show us what happens. Table #1 expresses how these possibilities could be presented for the reading disabled subject.
Delta Theta Alpha Beta
Delta Theta Alpha Beta
Degree of Activation
In this example, the subject's theta values are above normal while the beta values are below normal (absolute level and/or degree of activation from a visual attention task) and the subject's connection patterns show an above normal increase in theta connection values and a decrease in beta connection values (absolute level and/or degree of activation).
Additional Classication Problems
There are qualications to this approach. The rst issue concerns developmental issues. The QEEG variables which are related to effective cognitive performance are different for children then they are for adults (Thornton, 2001). Effective cognitive performance in children (under age 14) concern predominantly relative power and microvolts of the beta1 frequency (13-32 Hz) while for adults (over age 14) the effective QEEG patterns concern coherence issues in the alpha and beta frequencies (Thornton, 2001).
Phonic pronunciation offers an additional developmental concern, as for children (under age 14) the T4 beta activity is critical to successful phonemic pronunciation, while beta activity at the T3 and left poste- rior locations are the critical locations for success in adults (over age 14) (Thornton, unpublished).
Relationship to Psycho-educational Typologies
Given the overlapping categorization problem with current classication schemes, it would seem advisable to shift to a more precise physical denition of a disorder and then measure the psycho-educational consequences of that disorder. The author maintains the position that the brain's electrophysiological system is a distributed one which employs multiple locations and connection activity to accomplish a task success- fully (Thornton, 2001). Teuber's (1955) double dissociation hypothesis becomes problematic in a distributed functional system, as long as specic locations are the main focus. There is some very preliminary data indi- cating that math disability may reside in the high frequency range (23-64 Hz) (Thornton, 2001) which would argue for a frequency approach to, at least, math abilities.
The following graphic gures represent one clinical case which present this approach. The example is a 62 year-old business woman who reported a history of difculty in reading but excellent auditory mem- ory skills. Auditory memory is assessed by the examiner reading four stories to the subject and then asking for immediate recall and then 20 minutes later for delayed recall. Each story contains about 20-25 pieces of information, so the maximum score possible is 200. The mean paragraph recall score for a control group (N=15) (immediate and delayed) was 18 (SD=2.45). This subject obtained a total score of 117, which places her 4.57 SD above the norm. Figure 1 and 2 show the subject's values during the input stage for the coher- ence alpha ashlight variable and the relative power gures for the four frequencies. A ashlight variable is conceptualized as a point of origin which "sends" out a "beam" in a particular frequency. When that "beam" reaches a particular location a value can be obtained and compared to a normative value. In the gures the black circle represents the origin of the ashlight. The gures indicate the subject's standard deviation from the norm for the variable under consideration. Relative power values are also expressed in the gure in terms of their deviation from the norm. The subject's values were below norm for the delta and theta values and above the norm for the beta1 (13-32Hz) values (positive results).
Figure 1 -- Auditory Memory CoherenceAlpha (8-13 Hz) Flashlight Standard Deviation Values during input stage
Subject with exceptional auditory memory abilities
Legend:Black Circle represents origin of ashlight
Values in circles represent standard deviation (SD)
difference from normative group
White Circles represent values between -.50 SD and +.50 SD Light Orange represent values between +.50 SD and +1.5- SD
Figure 2 -- Auditory Memory Input Stage
Relative Power Standard Deviation Values for 4 frequencies
Legend:Numeric Values represent SD differences from normative group.
White Circles: Value between -.50 SD & +.50 SD Blue Circles: Value between -.50 SD & -1.50 SD Orange Circle: Value Between +.50 SD & 1.50 SD
Figure 3 -- Reading Task
Posterior Origin Flashlight Standard Deviation values -- Coherence Beta1 (13-32 Hz) & Coherence
Beta2 (32-64 Hz)
Legend:Numeric Values represent SD differences of subject from normative group
Black Circle is origin of ashlight
Dark Blue Circles represent values below -1.50 SD
Light Blue Circles represent values between -.50 SD & -1.50 SD White Circles represent values between -.50 SD & +.50 SD Light Orange represent values between +.50 SD & +1.50 SD Dark Orange circles represent values above +1.50 SD
Figure 4 -- Reading Task
Relative Power Standard Deviation Values for 4 frequencies
Reading Input Stage
Her reading memory score was 21.5 (immediate and delayed recall). Although the story was not the same as employed in the original research, the comparison is of some value. The mean for the original reading task was 36 (SD=20). Employing these values, the subject reading memory score was 0.72 standard deviation below the norm. The results support her subjective problem of difculty in reading as well as the value of the approach is diagnosing specic skill problems (auditory memory vs reading memory abilities). Figure 3 and 4 shows the coherence beta2 (32-64 Hz) values for the subject during the reading task with posterior origins of the ashlight as well as the relative power gures for the four frequencies.
There are several important points in this example. 1) - There was no evidence of problems in delta, theta or the beta frequencies, a common focus for intervention protocols; 2) - the high frequency coherence decits were blatantly apparent with the lowest value some 3.7 SD below the mean and diffusely patterned in the reading situation; 3) -- the parameters identied as relevant to each cognitive task demonstrated clearly where the electrophysiological problem resided. For example, the high coherence alpha patterns predomi- nantly in the left hemisphere were evident in this subject and are empirically related to good memory scores, while posterior beta coherence patterns were below normal in the reading task and related to successful reading.
The value of diagnosing learning disability patterns by QEEG parameters under task conditions has intuitive and empirical appeal as well as the potential for increased intervention effectiveness. This approach bypasses the problem of overlapping diagnoses with present psycho-educational classication schemes and provides precise intervention parameters which can address the physical cause of the learning disability. The availability of specic QEEG decit patterns allows for precise intervention protocols. This precision has been documented to increase auditory memory abilities in the learning disabled/ADHD child by some three standard deviations. This is considerably above more effective than most presently employed intervention models which produce average improvements in the +.50 to +1.00 standard deviation range (Thornton & Carmody, 2005).
Barry, R.J., Clarke, A. R., McCarthy, R., Selikowitz, M. (2006). Age and gender effects in EEG coherence: III.Girls with attention-decit / hyperactivity disorder. Clinical Neurophysiology, 117 (2), 243-51.
Business Week -- May 31, 2004, Special Needs, Crushing Costs http://www.businessweek.com/magazine/content/04_22/b3885139_mz070.htm
Biederman, J., Newcorn, J., & Sprich, S. (1991). Comorbidity of attention decit hyperactivity disorder with conduct, depressive, anxiety, and other disorders. American Journal of Psychiatry, 148, 564-577
CDC, (1997-8). Prevalence of attention-decit disorder and learning disability: United States, is based on data from the CDC's National Health Interview Survey (NHIS).
CDC, 2003, ADHD: A Public Health Perspective Conference, ADHD and other conditions
Retrieved on 5/17/2006 from http://www.medofceinc.com/f-set3.html
Chabot R.J., & Serfontein G. (1996). Quantitative EEG proles of children with attention decit disorder.
Biological Psychiatry, 40, 951-963.
Chabot, R.A., diMichele, F., Prichep, L., & John, E.R. (2001). The clinical role of computerized EEG in the evaluation and treatment of learning and attention disorders in children and adolescents. Journal of Neuropsychiatry and Clinical Neuroscience, 13, 171-186.
Clarke, A.R., Barry, R.J., McCarthy, R., & Selikowitz, M. (2001a). Electroencephalogram differences in two subtypes of attention-decit/hyperactivity disorder. Psychophysiology, 38 (2), 212-221.
Clarke, A.R., Barry, R.J., McCarthy, R., & Selikowitz, M. (2001b). EEG-dened subtypes of children with attention-decit/hyperactivity disorder. Clinical Neurophysiology, 112 (11), 2098-2105.
Clay, M. (1990) Observation survey of early literacy achievement. Hong Kong: Heinemann
American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders - Fourth
Edition (DSM-IV), Washington, D.C., American Psychiatric Association
Evans,J.R., & Park, N.S. (1996). Quantitative EEG abnormalities in a sample of dyslexic persons. Journal of
Neurotherapy, 2(1), 1-5.
Gerhardstein, R., Bacon, K.D., Anthony, J. (2001, April). Relations between the developmental precursors of ADHD and Reading: A longitudinal study. Presented at the biennial meeting of the society for Re search in Child Development, Minneapolis. MN
Retrieved on 5/12/2006 from
Harnadek, M., & Rourke, B. (1994). Principal identifying features of the syndrome of nonverbal learning disabilities in children, Journal of Learning Disabilities, 27 (3), 144-154.
Klingberg, T., Hedehus, M., Temple, E., Salz, T., Gabrieli, J.D., Moseley, M.E., Poldrack, R.A. (2000). Mi crostructure of temporo-parietal white matter as a basis for reading ability: Evidence from diffusion tensor magnetic resonance imaging. Neuron, 25, 492-500.
Leahy, B.T. (2001). Youth alienated from the education system often enter the prison system, November 29,
2001, Coalition for Juvenile Justice Retrieved on 5/17/06 http://www.juvjustice.org/media/1008.html
Lerner, J. (1989). Educational interventions in learning disabilities. Journal of the American Academy of
Child and Adolescent Psychiatry, 28, 326-331.
Lyon, G. R. (1995). Toward a denition of dyslexia. Annals of Dyslexia, 45, 3-27.
Lyon, G.R. Learning disabilities. In Child psychopathology. E. Marsh and R. Barkley, eds. New York: Guil ford Press, 1996, pp. 390-434.
Mann, C.A., Lubar, J.F., Zimmerman, A.W., Miller, C.A., & Muenchen, R.A. (1992). Quantitative analy sis of EEG in boys with Attention-Decit-Hyperactivity Disorder: Controlled study with clinical im plications. Pediatric Neurology, 8 (1), 30-36.
National Center for Learning Disabilities, Retrieved on 5/17/06 http://www.ncld.org/index.php?option=content&task=view&id=472
Ofce of Special Education and Rehabilitative Services, Federal Register: December 17, 1999 (Volume 64, Number 242)pp.70955-70960
Retrieved on 5/17/06 http://www.ed.gov/legislation/FedRegister/proprule/1999-4/121799b.html
Public Law 108-446l (2004), Individuals with Disabilities Education Improvement Act of 2004. -- To reau thorize the Individuals with Disabilities Education Act, and for other purposes. H.R. 1350 -- 108th Congress, 118 STAT. 26
Rumsey, J.M., Horwitz, B., Donohue, B.C., Nace, K., Maisog, J.M., Andreason, P. (1997a). Phonological and orthographic components of word recognition: A PET-rCBF study. Brain, 120 (Pt 5), 739-759.
Rumsey, J.M., Nace, K., Donohue, B., Wise, D., Maisog, J.M., Andreason, P. (1997b). A positron emission tomographic study of impaired word recognition and phonological processing in dyslexic men. Ar chives of Neurology, 54 (5), 562-573.
Shaywitz SE, Shaywitz BA, Fulbright RK, Skudlarski P, Mencl WE, Constable RT, et al. (2003) Neural systems for compensation and persistence: young adult outcome of childhood reading disability. Bio logical Psychiatry, 54 (1), 25-33.
Shaywitz, B.A., Shaywitz, S.E., Pugh, K.R., Mencl, W.E., Fulbright, R.K., Skudlarski, P., et al. (2002). Dis ruption of posterior brain systems for reading in children with developmental dyslexia. Biological Psychiatry, 52 (2), 101-110.
Shaywitz, S.E., Shaywitz, B.A., Pugh, K.R., Fulbright, R.K., Constable ,R.T., Mencl, W.E., et al. (1998).
Functional disruption in the organization of the brain for reading in dyslexia. Proceedings of the Na tional Academies of Sciences of the United States of America, 95 (5), 2636-2641.
The National Center for Learning Disabilities, 381 Park Avenue South, Suite 1420, New York, NY, 10016
Retrieved on 5/17/06 http://www.ncld.org/index.php?option=content&task=view&id=472
Teuber, H.L. (1955). Physiological psychology. Annual Review Psychology. 6, 267-96.
Thornton, K. (2001). Patent # 6309361 B1 Method for Improving Memory by Identifying and Using QEEG Parameters Correlated to Specic Cognitive Functioning - issued 10-30-2001
Thornton, K., & Carmody, D. (2005). EEG Biofeedback for Reading Disabilities and Traumatic Brain Inju ries. Child and Adolescent Psychiatric Clinics of North America, Jan, 2005a, 137-162.
Thornton, K. (2006). No Child Left Behind Goals (and more) are obtainable with the Neurocognitive ap proach. Vol. I, Booksurge publishing, North Charleston, SC, 2006
Torgesen, J.K. (1993) Variations on theory in learning disabilities. In G.R. Lyon, D.B. Gray, J.F. Kavanagh,
& N.A. Krasnegor (Eds.) Better understanding learning disabilities ([pp.153-170). Baltimore, MD: Paul H. Brookes Publishing Co.
US Ofce of Special Education, OSEP data tables, 2002, Part B, Table AA7
Retrieved on 5/17/06 https://www.ideadata.org/arc_toc4.asp
U.S. Ofce of Education (1977). Denition and criteria for dening students as learning disabled. Federal
Register, 42 (250), p 65083, Washington, D.C., U.S. government printing ofce
Wagner, R. K., & Torgesen, J. K. (1987). The nature of phonological processing and its causal role in the ac quisition of reading skills. Psychological Bulletin, 101, 192-212.
Watts, R., Liston, C., Niogi, S., Ulug, A.M. (2003). Fiber tracking using magnetic resonance diffusion ten sor imaging and its applications to human brain development. Mental Retardation and Developmen tal Disabilities Research Reviews, 9 (3), 168-177.
Wood, R. (1998, December). The trillion-dollar sham in federal remedial education. The National Right to Read Foundation.
Retrieved Sept. 3, 2004 from http://www.nrrf.org/essay_Trillion_116.html