Journal of Integrative Neuroscience, Vol. 6, No. 1 (2007) 175–190
Imperial College Press
DIFFERENT BRAIN ACTIVATION PATTERNS IN DYSLEXIC
CHILDREN: EVIDENCE FROM EEG POWER AND
COHERENCE PATTERNS FOR THE DOUBLE-DEFICIT
THEORY OF DYSLEXIA
The Brain Resource Company B.V./Brainquiry B.V.
Nijmegen, 6525 EC, The Netherlands
Radboud University Nijmegen, Educational Sciences
Nijmegen, 6525 EC, The Netherlands
Radboud University Nijmegen/EEG Resource Institute
Nijmegen, 6525 EC, The Netherlands
Radboud University Nijme gen/Behavioral Science Institute
Nijmegen, 6525 EC, The Netherlands
Received 23 October 2006
Revised 20 December 2006
Aims: QEEG and neuropsychological tests were used to investigate the underlying neural
processes in dyslexia.
Methods: A group of dyslexic children were compared with a matched control group from
the Brain Resource International Database on measures of cognition and brain function
(EEG and coherence).
Results: The dyslexic group showed increased slow activity (Delta and Theta) in the frontal
and right temporal regions of the brain. Beta-1 was speciﬁcally increased at F7. EEG
coherence was increased in the frontal, central and temporal regions for all frequency bands.
There was a symmetric increase in coherence for the lower frequency bands (Delta and
Theta) and a speciﬁc right-temporocentral increase in coherence for the higher frequency
bands (Alpha and Beta). Signiﬁcant correlations were observed between subtests such as
176 Arns et al.
Rapid Naming Letters, Articulation, Spelling and Phoneme Deletion and EEG coherence
Discussion: The results support the double-deﬁcit theory of dyslexia and demonstrate
that the diﬀerences between the dyslexia and control group might reﬂect compensatory
Integrative Signiﬁcance: These ﬁndings point to a potential compensatory mechanism of
brain function in dyslexia and helps to separate real dysfunction in dyslexia from acquired
Keywords: Dyslexia; EEG; QEEG; coherence; double-deﬁcit theory.
Developmental dyslexia is characterized by diﬃculties with accurate and/or ﬂuent
word recognition, by poor spelling and decoding abilities. These diﬃculties typi-
cally result from a deﬁcit in the phonological component of language that is often
unrelated to other cognitive abilities . Dyslexia is probably the most common
neurobiological disorder aﬀecting children, with prevalence rates ranging from 5 to
10 percent, and is a persistent, chronic condition .
Reading problems manifest themselves mainly in the following areas: diﬃculty
in learning to utilize correspondence regularities between graphemes and phonemes
[12, 37] poor phonological awareness, i.e., awareness of the sound structure of words,
especially phonemic awareness as manifested in the ability to analyze and manipu-
late sounds within a syllable  and poor use of orthographic word reading strate-
gies; and consequently inaccurate and non-ﬂuent word identiﬁcation [22, 26]. As a
result of these diﬃculties, full alphabetic or phonological reading skills are often
not attained. A large body of research has been conducted on the relation between
phonological awareness and learning to read. Strong support has been provided that
lack of phonological awareness can cause diﬃculties with the acquisition of reading
and writing [29, 36]. Being able to distinguish and identify the diﬀerent phonemes
in a word is part of this awareness. Research in the past decades has provided ample
evidence that dyslexic children have problems with phonological awareness and other
aspects of phonological processing. There is a general agreement that this phonolog-
ical processing deﬁcit has to do with problems in phonological encoding . Poor
readers are less precise in phonemic discrimination, they have problems on a variety
of phoneme segmentation and awareness tasks , and they are slower in rapid
naming of objects, digits and letters [42, 43], as well as in producing rhyming words
. It can be hypothesized that dyslexia is fundamentally a linguistic problem
which involves a deﬁcit in phonological encoding. Elbro, Borstrøm and Petersen 
tested this hypothesis by predicting dyslexia from phonological processing abilities
of kindergarteners. It was shown that three language measures contributed indepen-
dently to predict dyslexia: letter naming, phoneme identiﬁcation, and distinctness of
phonological representations. The results further indicated that the quality of phono-
logical representations in the child’s mental lexicon may also be a determinant of the
EEG Power and Coherence in Dyslexia: Double-Deﬁcit 177
development of phonemic awareness. Alternatively, there is the claim that reading
problems originate from a more general temporal processing deﬁcit .
Dyslexia has been attributed to deﬁciencies in visual, linguistic, and low level
sensory functions but most studies have been falsiﬁed empirically and logically
. Most research emphasizes the phonological deﬁcit in children with dyslexia
[1, 20, 33, 34], that is segmenting spoken words into their underlying phonological
elements and linking each letter to its corresponding sound. A number of studies also
support the double-deﬁcit theory. The double-deﬁcit theory proposes that reading
disabilities can be the result of: 1) poor phonological awareness and/or, 2) auto-
matic naming skills. Poor phonological awareness refers to disabilities in identifying
and manipulating sounds in speech, whereas poor automatic naming implies the dis-
ability to translate visual information into a phonological code. The double-deﬁcit
hypothesis proposes that accordingly, subtypes of dyslexia can be distinguished
showing a deﬁcit in either one or both of these skills [4, 43]. It is also claimed
that a deﬁcit in both skills yields the lowest reading performance. The dyslexic sub-
types could be produced by diﬀerential processing deﬁcits in the frontal-cerebellar
phonological system . The unique contribution of each frontal and cerebellar mea-
sure to the classiﬁcation of dyslexic participants and the prediction of phonological
and naming performance support this view.
Previous research has also linked dyslexia and reading disabilities to neurologi-
cal data. There are anatomical studies [7, 13] which show an absence of the usual
left-right hemisphere asymmetry of the planum temporale in dyslexia or suggest a
possible role of the left inferior frontal gyrus in speech perception and rapid audi-
tory processing, as well as in phonological aspects of reading , although no strong
eﬀects have been reported . Eckert et al.  found anatomical anomalies under-
lying the double-deﬁcit subtype of dyslexia. Their ﬁndings suggest that impairments
in a frontal-cerebellar network may play a role in delayed reading development in
To study the neural factors of dyslexia, functional neuroimaging has been used.
However, there is not much evidence with respect to developmental dyslexia since
this research has focused on (young) adults . Only Shaywitz and Shaywitz [32, 33]
used children in their neuroimaging studies in order to examine the neural systems
for reading during the acquisition of literacy. These reports show a failure of left
hemisphere posterior brain systems to function properly during reading [32, 33].
The majority of studies show increased activation in the basal surface of the tempo-
ral lobe, the posterior portion of the superior and middle temporal gyri, extending
into temporoparietal areas and the inferior frontal lobe during tasks requiring read-
ing and phonological processing . Shaywitz et al.  supports these ﬁndings,
however they show evidence of right hemisphere activation in the posterior temporal
parietal regions. This could reﬂect compensatory processes or could indicate that
other nonlinguistic factors are related to reading disability [32, 33, 38].
A few studies have focused on event related EEG changes in tasks directly related
to the reading diﬃculties of dyslexic children. Rippon and Brunswick  found
178 Arns et al.
that dyslexic children showed increased frontal theta activity in a phonological task,
whereas there were no diﬀerences between the dyslexic group and the control group
in a visual task. Furthermore, there was a marked parieto-occipital right greater than
left asymmetry in beta EEG activity in the dyslexic group with respect to the phono-
logical task and the visual task. Klimesch et al.  found that dyslexics have a lack
of attentional control during the encoding of words at left occipital sites and a lack of
a selective topographic activation pattern during the semantic encoding of words.
EEG coherence is a measure which displays functional connectivity between brain
areas, and could hence be an interesting measure to demonstrate deviation in func-
tional connectivity. To date, few EEG studies have considered EEG coherence. Sklar
et al.  found higher intrahemispheric coherence and lower interhemispheric coher-
ence during text processing in dyslexics compared with normals. This was also sup-
ported by Leisman and Ashkenazi . During rest, Shiota, Koeda, and Takeshita
 reported both increased intra- and inter-hemispheric coherence in dyslexic chil-
dren. Furthermore, Marosi et al.  found a frequency-dependent eﬀect on EEG
coherence at rest where diﬀerences between children with poor reading/writing abil-
ities were compared with children with good reading/writing abilities, with the for-
mer showing higher coherence in the delta, theta and beta bands and lower coherence
in the alpha bands during rest .
Weiss and Mueller  proposed that EEG coherence in the diﬀerent frequency
bands played diﬀerent roles: increased coherence in the theta band correlates with
language-related mnemonic processes and theta coherence was increased if task
demands increased and more eﬃcient information processing was required. Alpha
coherence seemed important for sensory processing and higher alpha coherence for
semantic processing. Beta and gamma coherence has been linked with more complex
linguistic sub-processes such as syntax or semantics .
Our aim was to compare brain function of dyslexic children with non-dyslexic
children on diﬀerent neurophysiological and neuropsychological measures. Our ques-
tion focused on whether diﬀerent EEG activation patterns can be found in dyslexia,
and to what extent correlations between reading and spelling abilities and speciﬁc
tasks for rapid naming and phonological awareness, can be found to address the
double-deﬁcit theory of dyslexia . We also assessed neuropsychological function
in these groups in order to exclude further cognitive diﬀerences between the groups
potentially confounding the EEG ﬁndings. Our hypothesis was that the groups will
not show diﬀerences on neuropsychological measures, but that children with dyslexia
will show increased inter- and intrahemispheric coherence.
2. Materials and Methods
Nineteen children with dyslexia (11 males and 8 females; average age = 10.33; range
8.0–15.98) and nineteen control children (matched on age, gender and education;
EEG Power and Coherence in Dyslexia: Double-Deﬁcit 179
11 males and 8 females; average age 10.34; range 8.01–16.03) were used to inves-
tigate the diﬀerences in brain function and neuropsychological performance. All
dyslexic children went to regular schools. They were diagnosed with dyslexia by
their remedial teachers, who worked with a structured protocol for diagnosing chil-
dren with dyslexia on the basis of their reading and spelling development from
grade 1 . The control group was drawn from the Brain Resource International
Database (BRID: www.brainresource.com, for more details also see [9, 10]) and
children were chosen from this database who did not have dyslexia or learning
Exclusion criteria included a personal or family history of mental illness, brain
injury, neurological disorder, serious medical condition, drug/alcohol addiction; and
a family history of genetic disorder. All subjects voluntarily gave written informed
Subjects were seated in a sound and light attenuated room, controlled at an
ambient temperature of 22
F. Electroencephalographic and neuropsychologi-
cal assessments were completed in order.
2.2. Language tests
The group of children with dyslexia was submitted to a range of tests to inves-
tigate correlations between EEG and neuropsychological ﬁndings of dyslexia. The
included tests were measures of tasks related to reading: Rapid Naming of Letters,
Articulation, Phoneme deletion  and Spelling .
2.3. Electroencephalographic data acquisition
Participants were seated in a sound and light attenuated room, controlled at an
ambient temperature of 22
C. EEG data were acquired from 28 channels: Fp1, Fp2,
F7, F3, Fz, F4, F8, FC3, FCz, FC4, T3, C3, Cz, C4, T4, CP3, CPz, CP4, T5,
P3, Pz, P4, T6, O1, Oz and O2 (Quikcap; NuAmps; 10–20 electrode international
system). Data were referenced to averaged mastoids with a ground at Fpz. Horizontal
eye-movements were recorded with electrodes placed 1.5 cm lateral to the outer
canthus of each eye. Vertical eye movements were recorded with electrodes placed
3 mm above the middle of the left eyebrow and 1.5 cm below the middle of the
left bottom eye-lid. Skin resistance was < 5 K Ohms and above 1 K Ohm for all
electrodes. A continuous acquisition system was employed and EEG data were EOG
corrected oﬄine . The sampling rate of all channels was 500 Hz. A low pass
ﬁlter with attenuation of 40 dB per decade above 100 Hz was employed prior to
The EEG data were recorded for two minutes with eyes open (EO). Subjects were
asked to sit quietly. Subjects were asked to ﬁx their eyes on a red dot presented on
a computer screen.
180 Arns et al.
2.4. Electroencephalographic variables
Each two minute epoch was divided into adjacent intervals of four seconds. Power
spectral analysis was performed on each four second interval by ﬁrst applying a
Welch window to the data, and then performing a Fast Fourier Transform (FFT),
next the average power spectra were calculated.
The power was calculated in the following frequency bands delta (1.5–3.5 Hz),
theta (4–7.5 Hz), alpha (8–13 Hz), alpha1 (8–11 Hz), alpha2 (11–13 Hz), SMR
(12–15 Hz), beta (14.5–30 Hz), beta1 (14.5–20 Hz), beta2 (20–25 Hz) and beta3 (25–
30 Hz). The data were then square-root transformed to approximate the normal
distributional assumptions required by parametric statistical methods.
Neuropsychological assessment was done using a touch screen monitor. Besides the
subtests for dyslexia, other neuropsychological tests were included in order to estab-
lish that the children were otherwise completely normal. Measures included: memory
recall and memory recognition (number of correctly reproduced words on trials 1, 5,
6, 7; number or correctly recognized words), verbal interference test — equivalent
to the Stroop test (Number correct text and color condition), tapping test (Number
of taps with the dominant and nondominant hand), timing test (proportional bias)
and switching of Attention test part A and B (equivalent to the WAIS Trails A
and B; time to complete the A and B form) (see [9, 10] for details of these tests).
All tests were fully computerized and subjects’ responses were recorded via touch-
screen presses. Reliability and validity data of these tasks are reported elsewhere
[2, 10, 25].
3. Statistical Analysis
3.1. Missing values
If missing values were present for a given statistical test, those cases were excluded
for that analysis. The number of missing values per group are included in the results
3.2. Statistical analyses
Since we expected quite local eﬀects on some measures due to the localized diﬀer-
ences in brain function for dyslexia, we did not perform the traditional GLM, since
small localized eﬀects could average out in the overall tests. Therefore, we performed
one-way ANOVA’s but used very stringent alpha correction. Signiﬁcance levels were
set as follows: for the EEG power data, the signiﬁcance level was set to p<0.05 and
for the coherence data, signiﬁcance levels were set to p<0.001. For EEG coherence
there were many more data points per frequency band (> 100 coherence values),
hence the lower p value for coherence compared to EEG power.
EEG Power and Coherence in Dyslexia: Double-Deﬁcit 181
The obtained signiﬁcant diﬀerences between the dyslexia and control group were
then submitted to a bivariate correlation analysis together with the severity ques-
tionnaire data, and a correlation matrix was obtained for correlations between vari-
ables within the group of dyslexic children.
4.1. EEG power
An overview of Delta and Theta power for all sites is depicted in Fig. 1 and the
signiﬁcant diﬀerences are indicated.
The following diﬀerences between the two groups were found:
Delta: increased Delta power for the dyslexia group at Fp1 (F = 6.315, df = 1, 33,
p =0.017), Fp2 (F = 4.861, df = 1, 34, p =0.034), F7 (F = 4.806, df = 1, 34,
p =0.035) and T6 (F = 6.193, df = 1, 35, p =0.018).
Fig. 1. Mean EEG power for Delta and Theta for the dyslexic group (in red) and the control
group (in black). All ﬁndings have p<0.05.
182 Arns et al.
Theta: increased Theta at Fp1 (F = 11.072, df = 1, 33, p =0.002), Fp2 (F =
5.074, df = 1, 34, p =0.031) and F7 (F = 8.267, df = 1, 34, p =0.007).
Beta 1: increased beta-1 at F7 (F = 4.450, df = 1, 34, p =0.042).
4.2. EEG coherence data
Figure 2 gives an overview of the signiﬁcant increases in coherence per frequency
band. All increases have p values of p<0.001. Due to the many signiﬁcant ﬁndings,
no detailed statistics are given. All coherence values were increased for the dyslexia
group red connections are from homologous pairs (both right and left hemisphere)
and purple are uniquely right or left hemispheric increased coherences. Note the
speciﬁc patterns which include mainly frontal, central and temporal sites. Also note
Fig. 2. Signiﬁcant increased coherences for the dyslexic vs. the control group for the diﬀerent
frequency bands. Red connections are from homologous pairs (equal right and left) and purple
connections are unique right or left hemispheric increased coherences. All p values’s < 0.001.
EEG Power and Coherence in Dyslexia: Double-Deﬁcit 183
Fig. 3. Signiﬁcant correlations between dyslexia subscales and signiﬁcant EEG coherences in the
diﬀerent EEG bands. Red = Delta Coherence; Orange = Theta Coherence; Light Blue = Alpha
Coherence; Dark Blue = Beta Coherence. Thick lines represent signiﬁcances of p<0.001 and thin-
ner lines of p<0.05. Note the speciﬁc independent patterns for some of the patterns especially
Articulation with a centro-temporal pattern but also the continuous involvement of the right tem-
poral region for all measures. Also note the clear diﬀerences between the slow (Delta and Theta;
Red and Orange) vs. higher (Alpha and Beta; Light and Dark Blue) EEG frequencies and the
similarities between Delta/Theta and Alpha/Beta.
the bi-laterally increased delta coherence fronto-central and the right fronto-central
increased coherence speciﬁcally in the alpha and beta band.
The dyslexia group named fewer words in the word condition of the verbal interfer-
ence test (F = −6.994, df = 36, p =0.012) but there was no diﬀerence on the color
condition of the (F = 2.330, df = 36, p =0.136). The dyslexia group recognized
fewer words as compared to the control group (F = 8.914, df = 36, p =0.005) on
the memory recognition task.
184 Arns et al.
Table 1. Signiﬁcant correlations between coherence values for the
diﬀerent frequency bands vs. the 4 dyslexia subtests. Note that all
correlations are rather high and positive, indicating that increased
coherence between a given electrode pair is related to better per-
formance on that test.
Location vs. Subtest Correlation and Sign
Delta C4-C3 vs. ART r =0.568; df = 17; p =0.017
T4-FC4 vs. ART r =0.508; df = 17; p =0.037
C4-T4 vs. ART r =0.527; df = 17; p =0.030
T3-FC3 vs. ART r =0.541; df = 17; p =0.025
T3-FC3 vs. PD r =0.520; df = 17; p =0.033
C3-F7 vs. RNL r =0.638; df = 17; p =0.006
C3-Fp1 vs. RNL r =0.662; df = 16, p =0.005
CP4-F8 vs. RNL r =0.527; df = 18; p =0.025
FC3-F7 vs. RNL r =0.576; df = 17; p =0.015
T4-F8 vs. SPL r =0.529; df = 17; p =0.029
CP4-T4 vs. SPL r =0.491; df = 17; p =0.045
Theta C3-F7 vs. RNL r =0.598; df = 17; p =0.011
FC3-Fp1 vs. RNL r =0.772; df = 16; p<0.000
T3-FC3 vs. ART r =0.527; df = 17; p =0.030
C3-T3 vs. ART r =0.532; df = 17; p =0.028
Alpha T4-FC4 vs. RNL r =0.576; df = 17; p =0.015
T4-FC4 vs. PD r =0.653; df = 17; p =0.005
C4-T4 vs. RNL r =0.508, df = 17; p =0.038
C4-T4 vs. PD r =0.565; df = 17; p =0.018
Beta C4-T4 vs. RNL r =0.501; df = 17; p =0.041
C4-T4 vs. SPL r =0.617, df = 17; p =0.008
C4-T4 vs. PD r =0.602; df = 17; p =0.011
CP4-T4 vs. RNL r =0.521; df = 17; p =0.032
CP4-T4 vs. SPL r =0.637; df = 17; p =0.006
ART = Articulation
PD = Phoneme Deletion
RNL = Rapid Naming Letters
SPL = Spelling
4.4. Within group correlations
The within group correlations were performed on 18 dyslexic children only (one
subject was removed from the analysis due to his age). This child was 16 years old
whereas the majority of the group was around 10 years of age, and his inclusion may
have lead to spurious age-related correlations.
Figure 3 and Table 1 shows the signiﬁcant correlations between the obtained
signiﬁcant measures reported in the previous section and the sub-tests used to mea-
sure the severity of dyslexia. All signiﬁcant EEG power and EEG coherence mea-
sures (63 measures: 8 EEG and 55 coherence) were submitted to the correlation
analysis with the four dyslexia sub-tests: Rapid Naming Letters — RNL; Phoneme
Deletion — PD, Articulation — ART and Spelling SPL. The results are depicted in
EEG Power and Coherence in Dyslexia: Double-Deﬁcit 185
four diﬀerent colors, each depicting a signiﬁcant correlation between that variable,
between those locations for the given frequency band. The thickness of the line also
depicts the signiﬁcance level (thin p<0.05; thick p ≤ 0.001).
Interestingly, there were no signiﬁcant correlations between the EEG power data
and the EEG coherence data within frequency bands, hence the increased coherence
for dyslexic children cannot be explained by the increased delta and theta frontally.
There was only one signiﬁcant correlation, between EEG power and the severity
of dyslexia: the power of Theta at FP1 and spelling (r =0.510; df = 16; p =0.044).
For coherence the signiﬁcant diﬀerences are depicted in Table 1 and are also
visually depicted in Fig. 3.
This study focused on brain function patterns and neuropsychological ﬁndings in
children with developmental dyslexia and aimed to establish a link between EEG
parameters and dyslexia relevant constructs. EEG ﬁndings showed an increased (left)
frontal and right temporal slow activity in the Delta and Theta bands and increased
Beta 1 power at F7. Since all EEG data have been EOG corrected using Gratton
et al. , it is very unlikely the frontal increased Delta and Theta is due to resid-
ual EOG. EEG Coherence data showed increased coherence in frontal, central and
temporal regions. However, the increased coherences seemed to show a frequency
speciﬁc eﬀect, where the slower frequencies (delta and theta) showed a more sym-
metrical increase in right and left frontal, central and temporal networks, whereas
the higher frequencies (alpha and beta) showed a more speciﬁc right-hemispheric
eﬀect originating at T4 and F8. Correlational analysis showed that these increased
coherences were an eﬀect in itself, since there were no correlations between the
increased delta and theta power on the one hand and the increased coherence in
the according band on the other hand; hence the increased delta and theta power
were not the cause of the increased coherence ﬁndings. High coherence between
two EEG signals suggests high cooperation and synchronization between underlying
brain regions within a certain frequency band . Increased coherence can thus be
interpreted as increased functional connectivity. This could implicate that dyslexic
children have increased activity within frontal, central and temporal networks.
There were also signiﬁcant diﬀerences between the dyslexic and the control group
on the verbal interference tests (similar to the Stroop test) and the memory recog-
nition test. These ﬁndings are directly related to the dyslexic problems experienced
by this group, since dyslexic children have decoding problems. The dyslexic children
named fewer words on the verbal interference test than the control children but
had no impairment when required to name the color of the word relative to normal
children. Dyslexic children also recognized fewer words on a memory recognition
test whereas spontaneous memory recall was not aﬀected at all. These ﬁndings sug-
gest that interpretation of neuropsychological data derived from these speciﬁc tests
should always be treated with caution, and may be that dyslexic status should be
186 Arns et al.
incorporated into the interpretation of neuropsychological data to safeguard false
positive ﬁndings on these tests.
Correlational analyses revealed signiﬁcant correlations between the obtained sig-
niﬁcant EEG ﬁndings and the tests: articulation, rapid naming of letters, spelling
and phoneme deletion. The correlated patterns (as depicted in Fig. 3) showed quite
speciﬁc patterns for all these 4 sub-tests. Interestingly, all these correlations were
positive and high (explaining > 30% of the variance), suggesting that better perfor-
mance on these tests was associated with increased coherence. Given the fact that all
the coherence ﬁndings were increased in comparison to the control group, it might be
concluded that these patterns reﬂect compensatory mechanisms and do not explain
the deﬁcit per se (where negative correlations would be expected). The EEG in this
study was not recorded during the completion of the dyslexia speciﬁc tests, hence
we recorded resting state eyes open EEG which correlated highly with these tests.
This further supports the fact that these patterns can be considered compensatory
patterns since they are also present at rest. Furthermore, this demonstrates that
clear associations can be found between passive brain states and deviant behavior,
demonstrating the utility of integrative approaches.
There seems to be a clear distinction between delta coherence on the one hand
and beta coherence on the other. The increased coherence for dyslexic children
was prominent and symmetric for the delta band; but localized to the left hemi-
sphere for the beta band. The correlations also demonstrate this; a slow (delta and
theta) coherent network over left frontal and central regions, and a faster (alpha and
beta) network originating at T4. Although EEG coherence between diﬀerent cortical
regions is largely established by cortico-cortical and thalamo-cortical interactions
, subcortical brain areas also contribute to both inter- and intra-hemispheric
functional communication . Lower bandwidths such as delta frequency in the EEG
coherence spectrum have particularly been associated with limbic contributions to
cortico-cortical coupling , hence these increased low-frequency coherences could
indicate a limbic contribution.
The core dysfunction in dyslexia seems to consist of increased slow activity at
left frontal and right temporal (T6) regions, and bilateral increased coherence in
the slower frequency bands (delta and theta), as opposed to acquired-compensatory
mechanisms consisting of right-hemispheric increased coherences in the higher fre-
quency bands (alpha and beta) and a left frontal increased coherence in slower bands
originating from C3 and FC3. The increases in coherence in the delta band fronto-
central suggest a strong limbic involvement as part of the core deﬁcit in dyslexia,
although this requires further study.
In this study, children showed delays in both rapid naming and phonological
awareness. These delays correlated with the activation in the frontal-cerebellar
phonological system. The EEG ﬁndings in our study showed an increased activation
pattern in dyslexic children, mainly in frontal and temporal lobes. Furthermore, the
correlation analyses showed signiﬁcant correlations with spelling, phonological skills
and rapid naming with quite diﬀerent topographical representations, suggesting
EEG Power and Coherence in Dyslexia: Double-Deﬁcit 187
involvement of diﬀerent neural mechanisms. It can tentatively be concluded that
the frontal-cerebellar network may be critical to the precise timing of mechanisms
that underlie the double-deﬁcit theory of dyslexia, suggesting the existence of three
subtypes of reading disability: dyslexics with deﬁciencies in phonological skills, poor
rapid naming skills or a combination of both types. Thus, the present study supports
the theory of Eckart et al.  hypothesizing that impairments in a frontal-cerebellar
network may play a role in delayed reading impairment in dyslexia. These authors
reported that anomalies in a cerebellar-frontal circuit are associated with rapid
automatic naming and phonological processing.
Previous EEG studies have shown diﬀerent ﬁndings. Rippon and Brunswick 
found no speciﬁc activation patterns with respect to dyslexic children. Weiss and
Mueller  have proposed several roles for coherence in the diﬀerent frequency
bands (also see introduction), however in this study, we did not use a task-related
protocol, making comparison to this study diﬃcult.
This study contributes to the theory that neurobiological causes underlie
dyslexia. The increased activation patterns of dyslexic children seem to be associ-
ated with the double deﬁcit type of dyslexia. In future research, it will be important
to examine the relation between EEG data and the phonological or orthographic
deﬁcits. Outcomes of these studies might further contribute to the diagnosis of sub-
types of dyslexia.
Finally, this study demonstrated that increased EEG power could not explain
the increased coherence ﬁndings in dyslexia, suggesting these measures reﬂect dif-
ferent neural networks. The positive correlations between coherence and the diﬀer-
ent tests demonstrated that these increased coherences might reﬂect compensatory
mechanisms rather then being part of the real core dysfunction in dyslexia, whereas
the increased slow activity might be part of the core dysfunction in dyslexia. This
should be taken into account in future studies to elucidate dysfunctional networks in
dyslexia. These dysfunctional networks can be dissociated from acquired compen-
satory mechanisms. Also, treatments focused on normalizing brain function (e.g.,
rTMS, EEG Biofeedback or Neurofeedback) will beneﬁt from this given they could
target the deﬁcit rather than target acquired compensatory mechanisms.
Data from The Brain Resource International Database were provided by the Brain
Resource Company (BRC: www.brainresource.com). We would also like to thank
local BRC clinics for data collection of the control group. All scientiﬁc decisions
are made independent of Brain Resource Company’s commercial decisions, via the
independently operated scientiﬁc division BRAINnet, which is overseen by the inde-
pendently funded Brain Dynamics Center and scientist members. We would also
like to acknowledge the contributions from Ine Giepmans and Minnie War for the
data collection and Sabine de Aukje Bootsma and Hanneke Friesen for acquiring
the QEEG and neuropsychological data.
188 Arns et al.
 Bishop DVM, Cerebellar abnormalities in developmental dyslexia: Cause, correlate or
consequence? Cortex 38:491–498, 2000.
 Clark CR, Paul RH, Williams LM, Arns M, Fallahpour K, Handmer C, Gordon E,
Standardized assessment of cognitive functioning during development and ageing using
an automated touchscreen battery, Arch Clin Neuropsychol 21(5):449–467, 2006.
 Davey MP, Victor JD, Schiﬀ ND, Power spectra and coherence in the EEG of a vegeta-
tive patient with severe asymmetric brain damage, Clin Neurophysiol 111:1949–1954,
 Eckert M, Leonard C, Richards TL, Aylward E, Thomson J, Berninger V, Anatomical
correlates of dyslexia: Frontal and cerebellar ﬁndings, Brain 126:482–494, 2003.
 Elbro C, Scarborough HS, Early identiﬁcation, in Bryant P, Nunes T (eds.), Interna-
tional Handbook of Children’s Reading, Dordrecht, Kluwer, pp. 339–359, 2003.
 Elbro C, Borstrøm I, Petersen DK, Predicting dyslexia from kindergarten. The impor-
tance of distinctness of phonological representations of lexical items, Read Res Quart
 Galaburda AM, Sherman GP, Rosen GD, Aboitiz F, Geschwind N, Developmental
dyslexia: Four consecutive patients with cortical anomalies, Ann Neurol 18:222–233,
 Geelhoed J, Reitsma P, PI-dictee, Lisse, Swets and Zeitlinger, 1999.
 Gordon E, Integrative neuroscience in psychiatry: The role of a standardized database,
Aust Psychiat 11(2):156–163, 2003.
 Gordon E, Cooper N, Rennie C, Hermens D, Williams L, Integrative neuroscience: The
role of a standardized database, Clin EEG Neur osci 36:64–75, 2005.
 Gratton G, Coles MG, Donchin E, A new method for oﬀ-line removal of ocular artifact,
 GoughPB,EhriL,TreimanR(eds.),Reading A cquisition, Lawrence Erlbaum Asso-
ciates, Hillsdale, NJ, 1992.
 Habib M, The neurological basis of developmental dyslexia. An overview and working
hypothesis, Brain 123:2373–2399, 2000.
 Horwitz B, Rumsey JM, Donohue BC, Functional connectivity of the angular gyrus in
normal reading and dyslexia, Proc Natl Acad Sci USA 95:8939–8944, 1998.
 Høien T, Lundberg I, Dyslexia: From Theory to Intervention, Dordrecht, 2000.
 Instituut voor Orthopedagogiek & Cito. Normeringsonderzoek Screeningstest voor
Taal- en L eesproblemen basisonderwijs jaargroep 1 tot en met 8.Nijmegen/Arnhem:
Instituut voor orthopedagogiek/Cito, 2004.
 Klimesch W, Doppelmayr M, Wimmer H, Gruber W, R¨ohm D, Schwaiger J, Hutzler F,
Alpha en beta band power changes in normal and dyslexic children, Clin Neurophysiol
 Locatelli T, Cursi M, Liberati D et al., EEG coherence in Alzheimer’s disease, Electroen
Clin Neuro 106:229–237, 1998.
 Leisman G, Ashkenazi M, Aetiological factors in dyslexia: IV, Cerebral hemispheres
are functionally equivalent, Int J Neurosci 11:157–164, 1980.
 Lyon GR, Shaywitz SE, Shaywitz BA, A deﬁnition of dyslexia, Ann Dyslexia 53:1–14,
EEG Power and Coherence in Dyslexia: Double-Deﬁcit 189
 Mann VA, Phoneme awareness and future reading ability, JLearnDisabil26:259–269,
 Manis FR, Doi LM, Bhadha B, Naming speed, phonological awareness, and ortho-
graphic knowledge in second graders, JLearnDisabil33(4):325, 2000.
 Marosi E, Harmony T, Becker J, Reyes A, Bernal J, Fernandez, Rodriguez, Silva, Guer-
rero, Electroencephalographic coherences discriminate between children with diﬀerent
pedagogical evaluation, Int J Psychophysiol 19(1):23–32, 1995.
 Nunez PL, Srinivasan R, Westdorp AF, Wijesinghe RS, Tucker DM, Siberstein RB,
Cadusch PJ, EEG coherency I: Statistics, reference electrode, volume conduction,
Laplacians, cortical imaging, and interpretation at multiple scales, ElectroenClinNeuro
 Paul RH, Lawrence J, Williams LM, Clark RC, Cooper N, Gordon E, The validity of
“IntegNeuro”: A new computerized and standardized battery of neurocognitive tests,
In J Neurosci 115:1549–1567, 2005.
 Plaut DC, McClelland JL, Seidenberg MS, Patterson K, Understanding normal and
impaired word reading: Computational principles in quasi-regular domains, Psych Rev
 Pugh KR, Mencl WE, Jenner AR, Katz L, Frost SJ, Lee JR, Shaywitz SE, Shaywitz
BA, Functional Neuroimaging studies of reading and reading disability (Developmental
Dyslexia), Ment Retard Dev D R 6:207–213, 2000.
 Rippon G, Brunswick N, Trait and stait EEG indices of information processing in
developmental dyslexia, Int J Psychophysiol 36:151–165, 2000.
 Scarborough HS, Children at risk for reading disabilities. Phonological awareness and
some other promising predictors, in Shapiro BK, Accardo PJ, Capute AJ (eds.), Speciﬁc
Reading Disability: A View of the Spectrum, York Press, Timonium, MD, pp. 75–119,
 Shiota M, Koeda T, Takeshita K, Cognitive and neurophysiological evaluation of
Japanese dyslexia, Brain Dev 22:421–426, 2000.
 Sklar B, Hanley J, Simmons WW, An EEG experiment aimed toward identifying
dyslexic children, Nature 240:414–416, 1972.
 Shaywitz BA, Shaywitz SE, Pugh KR, Mencl WE, Fullbright RK, Skudlarski P et al.,
Disruption of posterior brain systems for reading in children with developmental
dyslexia, Biol Psychiat 52:101–110, 2002.
 Shaywitz SE, Shaywitz BA, Dyslexia (Speciﬁc Reading Disability), Biol Psychiat
 Snowling MJ, Language and literacy skills: Who is at risk and why? in Bishop DVM,
Leonard LB (eds.), Speech and Language Impairment in Children: Causes, Character-
istics, Interventions and Outcome, Psychology Press, Hove, UK, pp. 245–260, 2000.
 Tallal P, Experimental studies of language learning impairments: From research to
remediation, in Bishop D, Leonard L (eds.), Speech and Language Impairments in
Children, Psychology Press, Hove, pp. 131–155, 2000.
 Torgesen JK, Wagner RK, Rashotte CA, Prevention and remediation of severe reading
disabilities: Keeping the eye in mind, Sci Stud Read 1:217–234, 1997.
 Van Orden GC, Phonological mediation is fundamental to reading, in Besner D,
Huphreys GW (eds.), Basic Processes in Reading, Lawrence Erlbaum, Hillsdale, NJ,
pp. 77–103, 1991.
190 Arns et al.
 Vellutino FR, Fletcher JM, Snowling MJ, Scanlon DM, Speciﬁc reading disability
(dyslexia): What have we learned in the past four decades? J Child Psychol Psyc
 Weiss S, Mueller HM, The contribution of EEG coherence to the investigation of lan-
guage, Brain Lang 85:325–343, 2003.
 Wagner RK, Torgesen JK, Rashotte CA, Hecht SA, Barker TA, Burgess SR, Donahue J,
Garon T, Changing relations between phonological processing abilities and word-level
reading as children develop from beginning to skilled readers: A 5-year longitudinal
study, Dev Psychol 33(5):468–479, 1997.
 Wolf M, Obregon M, Early naming deﬁcits, developmental dyslexia, and a speciﬁc
deﬁcit hypothesis, Brain Lang 42:219–247, 1992.
 Wolf M, Bowers P, The question of naming speed deﬁcits in developmental reading
disabilities: An introduction to the double-deﬁcit hypothesis, JLearnDisabil33:322–
 Wentink H, Verhoeven L, Protocol leesproblemen en dyslexie [Protocol for scr eening
reading problems and dyslexia], Nijmegen. Expertisecentrum Nederlands, 2001.