FRONTAL BILATERAL SYNCHRONOUS THETA WAVES AND THE RESTING EEG COHERENCE

IN CHILDREN AGED 7-8 AND 9-10 WITH LEARNING DIFFICULTIES

R.I. Machinskaya, A.V. Kurgansky

Institute of Developmental Physiology of Russian Academy of Education, Moscow

e-mail: regina_home@inbox.ru

Abstract

The resting state cortical functional connectivity was studied in children of 7-8 (N = 29) and 9-10 (N=23)

years with learning difficulties whose EEG showed the frontal bilateral synchronous theta waves (FTW)

and in the control children of 7-8 (N = 32) and 9-10 (N=16) years who did not experience school

difficulties and whose EEG did not show signs of abnormality. The functional connectivity was estimated

in the θ, α1, α2, and β1 frequency bands via measuring the coherence for the resting EEG that was free

from any abnormal patterns. Compared to control children, there was a reduction in the strength of the

functional coupling between the frontal and anterior temporal cortices found predominantly in the left

hemisphere of FTW children. The relative weakness of the coupling between the frontal cortex and the

other cortical areas was more pronounced in children of 7-8 than in children of 9-10 years. These between-

group differences were unaffected by the frequency band or gender factors.

Key words: frontal theta waves, resting EEG coherence, children with learning difficulties

INTRODUCTION

Bilateral-synchronous bursts of frontal theta waves (FTW) (fig. 1) are EEG patterns typical for

children with learning difficulties who had no history of neurological or psychiatric disorders. These EEG

patterns were detected by visual analysis in children aged 6-7 and 7-8 with low level of academic

performance (60-80 % in different samples) along with no learning difficulties and EEG abnormality in

control groups [1-3]. The results of visual EEG analysis agree with spectral analysis data: there is an

increase of spectral theta power in fronto-central cortices in children with learning difficulties [4, 5].

Complex electroencephalographical and neuropsychological studies [3, 6] showed that the main cause of

poor academic performance of children aged 6-7 and 7-8 with FTW was deficit of programming, voluntary

regulation and control.

Paste fig. 1

Seeking the neurophysiological basis of detected cognitive deficits we turned to the clinical

electroencephalographical data on the etiology of slow waves in frontal areas. These EEG patterns

appeared to be thalamus damage manifestations, especially its medial parts [7, 8]. Damage of thalamic

medial parts, in turn, is accompanied by cognitive and behavior regulation difficulties similar to the lobe

syndrome symptoms [9-11]. The reason of lobe syndrome deficits manifestation in the case of damage to

the mediodorsal thalamic nucleus (MD) is due to its close connectivity with the prefrontal cortex (PFC) via

strong reciprocal links which form together the integrated morpho-functional fronto-thalamic system [12-

14]. The immediate evidence of connection between FTW and thalamic neurons activity was found when

modeling the lobe syndrome in mice by cobalt application to their frontal cortex [15]. These findings state

that slow synchronous θ waves in the frontal cortex are accompanied by generation of θ waves in MD

thalamic neurons and, in addition to that, θ oscillations are coherent in both structures. Comparing the data

above and results of our own electroencephalographical and neuropsychological studies prepared the

ground for treating FTW as a sign of non-optimal functioning of the fronto-thalamic brain system.

In our previous study [16] we compared directed intracortical interactions* during FTW recordings

and periods free of these patterns by vector autoregressive model [17]. The analysis of the mutual influence

of EEG signals showed that the presence of common neuronal activity in the θ frequency range

synchronized due to the connectivity with deep brain structures is the most possible reason for the

emergence of synchronous slow waves in the frontal cortices. It was also shown that cortical functional

interaction during the FTW recording is determined by the increase of diffuse involvement of frontal and

central areas to the interaction both between each other and with other cortices.

These data is in a good agreement with findings of the fronto-thalamic etiology of slow

synchronous oscillations in frontal cortices recorded in the rest condition with eyes closed. The occasional

reduction in the degree of autonomy of frontal cortices as a result of fronto-thalamic system non-optimal

functioning found in the study [16] could be one of the reasons of cognitive deficits in children with FTW.

The significance of the rest state as the initial background for the cognitive activity was shown in studies

conducted by D.A. Farber et al [19, 20], who compared age-related changes of the background EEG and

cognitive processes in children. It could be assumed that one to several seconds long malfunctions in the

frontal brain areas during the FTW recording may impede switchover from the rest state to the activity

state. However, the extent to which such malfunctions influence the brain work during cognitive activity is

still the question in the frame of background EEG analysis.

At the same time, quantitative background EEG analysis allows investigating another question

which is also required to be studied to search the reasons of cognitive deficits in children with FTW. We

are speaking about the possible negative influence of non-optimal fronto-thalamic functioning on the

morpho-functional maturation of intracortical connectivity during developmental course.

Fronto-thalamic system plays an important role in selective modulation of cortical activity during

attentive processing [21, 22], preparation for the relevant information analysis [2] and keeping information

in working memory [23]. The lacking selective modulation influence of fronto-thalamic system on the

brain cortex could have an effect not only on the brain functioning during the cognitive performance but

also cortical neuronal network formation during developmental course. The latter assumption is reasonable

because both local cortical neuronal networks [24] and brain conduction pathways are being matured for a

* Hereinafter terms “intracortical interaction”, “functional interaction” or “functional connectivity” are used to describe linear

statistical relationship between spatially distant neurophysiological events [18], in this case between frequency components of

EEG signals recorded from spatially distant cortices.

long period during developmental course [25]. Moreover, there are findings testifying the dependence of

the brain morpho-functional maturation on the individual experience, particularly those types of activity

which require the involvement of selective voluntary attention [26, 27].

It was shown that maturation of brain neuronal networks during developmental course is reflected

in background EEG characteristics [19, 28]. In this case, if non-optimal fronto-thalamic functioning really

influences the formation of cortical neuronal connections, we can expect this influence to be reflected in

cortico-cortical connectivity estimated via resting-state EEG coherence during the periods free of FTW.

The previous study of resting-state α coherence was conducted in another group of primary schoolchildren

who had EEG signs of immature brain regulatory systems including immature fronto-thalamic system

manifested as FTW. These findings published in [29] point to the specificity of cortical functional

interaction via the α rhythm in children with FTW. In the current study we continued to analyze functional

brain cortex organization in children with FTW presented in our previous publication [16] and our task was

to estimate the parameters of EEG rhythms coherence in four frequency ranges: θ, α 1, α 2 and β 1. We

compared brain electrical activity (EA) recordings in children aged 7-8 and 9-10 with learning difficulties

whose EEG showed frontal θ waves and in children of the same age with no learning difficulties and EEG

abnormalities.

METHODS AND MATERIALS

We analyzed EEG recordings of 100 children aged 7-8 and 9-10, who attended regular primary

schools and had no history of neurological or psychiatric disorders. All the children and their parents gave

informed consent to participate in the study. The distribution of sample by age, sex and the presence of

FTW in children EEG is reflected in table 1. Approximately one half of the sample demonstrated low level

of academic performance (scored 2 or 3 according to the Russian five point grading system) in the main

courses and had bilateral synchronous θ waves in their EEG recordings, the other half of the sample were

of the same age but experienced no learning or behavioral difficulties and had no brain EA abnormalities.

Paste table 1

EEG was recorded in the rest condition with eyes closed from 12 symmetrical leads: F3/F4, C3/C4,

T3/T4, P3/P4, T5/T6, O1/O2 placed according to the international 10-20 system. EEG was referenced to the

numerical average of the right and left mastoid channels.

As our current task was to analyze functional interaction of areas during the periods free of FTW

and other EA abnormalities an expert first of all looked through EEG recordings to mark and cut segments

which contained artifacts. Finally, FTW and artifact-free EEG was not a continuous recording but a

sequence of segments different in size and separated by deleted chunks. In each individual EEG recording

such segments were subdivided into adjacent intervals 0.5 seconds long. Total duration of “pure” (FTW

and artifact-free) recording varied from child to child. To avoid the influence of this circumstance on

analyzed parameters of EEG coherence we used the sequence of first forty segments 0.5 seconds long (the

total duration was 20 seconds) for computing individual parameters of the coherence. The choice of the

number of segments being processed was determined by the shortest “pure” recording in the sample.

Functional cortico-cortical interactions were estimated by 12-channel vector autoregressive model

(hereinafter VAR-model). The VAR model is the simplest model capturing the linear statistical

dependencies in multichannel time series and successfully used in electrophysiology (see [17, 30] for a

review). Due to the fact that 40 segments chosen to analyze were not a continuous recording we used the

method suggested in [31] to calculate VAR coefficients. A VAR-model order of 20 was used as an optimal

order. This choice of order guaranteed the avoidance of serial correlations of autoregressive residuals and

simultaneously provided the spectral resolution of about 1 Hz which is high enough [30]. Using the VAR

coefficients computed for each pair of leads with numbers k and m we calculated the cross-spectrum

)( fSkm and power spectra )( fSkk and )( fSmm to get the coherence function:

)()()()( fSfSfSfC mmkkkmkm .

To analyze the coherence functions we chose four frequency ranges with bounds individually

selected for each subject by the following procedure. First, the fixed frequency margins 4–7 Hz for θ, 7–9

Hz for α1, 9–12 Hz for α2, and 12–15 Hz for β1 were chosen. Then, for each subject we made a common

for all lead pairings histogram of frequency locations (with 0.5 Hz step within 3-16 Hz range) of the local

maxima of coherence function. This histogram was smoothed using 11-point Bohman window. Then all

four a priori chosen frequency ranges were checked for local maxima of the smoothed histogram. If there

was one maximum within the range we applied 2 Hz wide interval with the center corresponding to the

frequency location of this maximum as an individual range. If there were two or more maxima we chose

the biggest one and set the range the way we did in the previous case. If there were no such maxima we

chose the entire a priori frequency range for the individual range. Mean coherence values in each

individual range were calculated as individual coherence function estimates were used in the further group

statistical analysis.

Statistical testing was performed using multivariate analysis of variance (MANOVA). Pairwise

comparisons were accomplished with ANOVA.

RESULTS

Using multivariate analysis of variance (MANOVA) we studied the influence of age (two-level

AGE factor: 7-8 and 9-10 years) and frontal θ waves presence (two-level FTW factor) on EEG coherence

in four frequency ranges (four-level RANGE factor: θ, α 1, α 2 and β) for interhemispheric and

intrahemispheric pairs of leads. To study connectivity specific for the left and right hemispheres we

included the HEMISPHERE factor (two-level) into MANOVA schema and the LOCATION factor to

analyze regional differences. According to the fact that there was different boy-girl ratio in the compared

groups (see table 1) the sex factor was also taken into account when performing MANOVA. To minimize

pairwise comparisons during the statistical analysis of LOCATION factor we applied MANOVA

separately for several subsets of lead pairs: F-X, C-X, P-X, Ta-X and Tp-O, where X corresponds to each

remaining lead in the same hemisphere. One and the same pair of leads was never used in different subsets.

By analyzing the influence of between-subjects factors it was shown that age, sex and the presence

of FTW in EEG had no significant main effect on the EEG coherence value in any subset of lead pairs.

Whenever the between-subject factors affected the coherence they did it via interaction with other factors.

As the main task of this study was to estimate a specific character of cortico-cortical functional

interactions in children with frontal θ waves in EEG we analyzed only the impact of FTW factor or its

interaction with other independent variables mentioned below. Table 2 lists significant effects for all

subsets of leads. Table 3 contains descriptive statistics for coherence in separate pairs of leads.

First of all it should be mentioned that the main effect of FTW was not found in this sample of

children. FTW factor influence was mediated by other conditions most commonly associated with the

location of EEG leads (HEMISPHERE and/or LOCATION factors). In all subsets of lead pairs we found

no connection between FTW influence and either sex or EEG frequency range. Due to this fact the post-

hoc analysis of EEG data was performed regardless of the subjects’ sex and EEG frequency ranges. Thus

below are shown the results of comparison of coherence value averaged over four frequency ranges.

Paste table 2 and 3

The impact of FTW factor on the interaction of frontal areas with other cortices depended on

children’s age in the studied sample. This dependence was reflected in FTW AGE interaction as a

tendency and in significant FTW LOCATION AGE interaction. Post-hoc analysis of FTW AGE

interaction showed that average (over four frequency ranges) coherence value in F-X lead pairs is higher

in the control group of children aged 7-8 (С = 0.456±0.021) than in children with frontal θ waves (C =

0.424±0.022). This difference found in junior age group turned to be significant (F(1, 59) = 4.546, p =

0.037). There were no significant differences for this factor in children aged 9-10 (F(1, 37) = 1.043, p =

0.314), with coherence value being lower in the control group than in group with FTW (С = 0.428±0.029

and С = 0.447±0.024, respectively). As regards the other “participant” of this interaction, which is the AGE

factor, we found no significant differences in both control and FTW groups. Coherence value seemed to

decrease in control group from 7-8 to 9-10 years (С = 0.456±0.02 at 7-8 years, С = 0.428±0.028 at 9-10

years, F(1, 46) = 2.778, p = 0.103) and increased in group with FTW (С = 0.425±0.023 at 7-8 years, С =

0.447±0.026 at 9-10 years, F(1,050) = 1.834, p = 0.182).

Significant FTW LOCATION AGE interaction suggests that FTW influence on coherence in the

F-X lead pairs depended not only on age but LOCATION of another lead in the pair. Therefore in each of

the five lead pairs included in this subset (omitting the HEMISPHERE factor) we compared coherence

over the FTW factor separately in groups of children aged 7-8 and 9-10. Differences with p < 0.01 were

regarded as significant after correction for multiple comparisons (five pairs of leads). Hence, the

difference for the lead pair F-P (F(1, 59) = 5.20, p = 0.02) found in children aged 7-8 considered to be only

a tendency, while, for the lead pair F-Ta, it was accepted as significant (F(1, 59) = 8.62, p = 0.005). In both

lead pairs, the level of functional cortical interaction was lower in children with FTW as compared to the

control group. Children aged 9-10 with FTW also had low coherence in F-Ta lead pair, but the difference

reached a significant level only in the left hemisphere (F(1, 37) = 5.85, p = 0.004). In other F-X lead pairs

we found no FTW factor influence both in 7-8 and 9-10 children. Finally, completing to analyze functional

connectivity of frontal areas it should be noted that the influence of FTW factor was in general (omitting

AGE and lead LOCATION) more pronounced in the left hemisphere. As compared to the control group,

EEG signals recorded from frontal areas of the left hemisphere in children with FTW turned to be less

statistically related to EEG signals from other areas of the same hemisphere (С = 0.475±0.015 in control

group, C = 0.442±0.015 in group with FTW, F(1, 92) = 4.73, p = 0.015). For lead pairs of the right

hemisphere, we found no statistically significant differences in EEG coherence between the control group

and children with FTW. Such a selective FTW influence on functional connectivity of frontal cortices

found in the left hemisphere was reflected in the significant HEMISPHEREFTW interaction (see table

2).

The dependence of FTW factor influence on children’s age was also found for interhemispheric

functional connections (table 2). Descriptive statistics of interhemispheric coherence values is presented in

table 4.

Paste table 4

In younger group of children with FTW, interhemispheric interaction of symmetrical cortices was

weaker than in the control group (in control group С = 0.497±0.018, in group with FTW С = 0.457±0.020;

F(1, 57) = 3.930, p = 0.027). In 7-8 years old children the FTW effect was differently manifested for

different pairs of symmetrical areas (FTW LOCATION, F(5, 53) = 3.36, p = 0.01)). The coherence

decrease in children with FTW as compared to the control group reached the statistical significance

(corrected for 6 comparisons) for interhemispheric connections between frontal areas (F(1, 57) = 6.767, p =

0.002). In the older age group, the picture of interhemispheric coherence relationship was opposite:

children with FTW had a stronger interhemispheric interaction as compared to the control group, however

the difference did not reach the significance (С = 0.494±0.027 vs. С = 0.472±.026; F(1, 35) = 1.618, p =

0.121). In children aged 9-10, there were also no significant regional differences in FTW influence on

interhemispheric coherence.

Since no FTW AGE interaction was found for any other subset of leads all post-hoc comparisons

presented below were performed leaving aside AGE factor. Significant

FTWHEMISPHERELOCATION interaction (see table 2) allows analyzing FTW factor influence on

the coherence in separate lead pairs for each subset of leads.

Analyzing coherence of EEG signals in pairs of leads, included central areas, we again found the

stronger FTW influence on the functional connectivity in the left hemisphere as compared to the right

hemisphere. In left hemisphere lead pairs С3-T3 and С3-T5, the coherence was higher in children of the

control group as compared to children with FTW. Taking into account multiple comparisons (eight lead

pairs of the right and left hemisphere) the significant differences were found only for C3-T3 (F(1, 98) =

7.798, p = 0.006). In the right hemisphere, such differences were observed only for С4-T4 (F(1, 98) =

8.454, p = 0.005), and coherence in the lead pair C4-T6 did not differ even for average values. The analysis

of the coherence in P-X lead pairs showed no differences except a single almost significant difference in

P3 – O1 lead pair (F (1, 98) = 5.705, p = 0.019). It should be noted that, unlike the anterior cortices, the

coherence value for the pairs of posterior locations of the left hemisphere was higher in the children with

FTW than in the control group (see table 3). There was no FTW factor influence on the right hemisphere

coherence. For either hemisphere, the FTW factor had no significant influence on the strength of anterior

temporal to inferior temporal or occipital areas functional links; nor had it any impact onto the connection

strength between inferior temporal and occipital areas.

Thus, the analysis of the resting state EEG coherence showed that the decrease in functional

connectivity of fronto-temporo-central cortices, mainly in the left hemisphere, was a common phenomenon

for children aged 7-8 and 9-10 with FTW as compared to the control group. In the parieto-occipital areas of

the left hemisphere, on the contrary, children with FTW approached a significant increase in the strength of

cortico-cortical interaction.

We found no significant age-related changes of EEG coherence in both FTW and control groups.

Yet, the coherence analysis demonstrated age influence on the degree of between-group differences: a

relative weakness of functional connectivity between frontal and other areas was in general more

pronounced in children with FTW at 7-8 years.

DISCUSSION

Results of the study made it possible to answer affirmatively the question of whether the difference

between spatio-temporal organization of artifact-free background EEG in children with FTW and the

control group exists. A qualitative comparison of current results with our previous results [29] on the

background EEG coherence in children with FTW obtained in other sample with other device and methods

for coherence calculation allows speaking of similarity in these data. Like in the previous study, current

results demonstrated the background EEG coherence decrease mainly in closely spaced lead pairs which

included anterior temporal area of the left hemisphere (T3). The similarity of results of the two independent

studies points to their robustness. This allows considering the “weakening” of connections between the left

anterior temporal area and nearby areas of the same hemisphere as a specificity of the resting state cortico-

cortical interaction in children with learning difficulties and frontal θ waves in EEG. Along with it, the

current study demonstrated the weakening of functional links between the frontal and anterior temporal

areas of the left hemisphere and frontal interhemispheric connection in children with FTW, as compared to

the control group. The previous study did not show such significant between-group differences. The

possible reason for the discrepancy is that in [29] we confined the analysis to the α frequency band only.

We could assume that taking into account θ and β activities, which are observed in frontal brain areas,

made possible to estimate the functional connectivity among these cortices more accurately. In this study,

unlike the previous study, we found between-group differences in coherence in caudal parts of the left

hemisphere which were opposite in sign, i.e. stronger functional interactions between occipital and parietal

areas in children with FTW, as compared to the control group. The possible reason for the divergence

between results of the two studies is the age difference between compared samples: in the previous study

we analyzed FTW influence on the cortico-cortical connectivity only in children aged 7-8 while in this

study we compared the EEG coherence in children aged both 7-8 and 9-10. The latter (i.e. the inclusion of

9-10 year old children) could contribute into higher coherence values for children with FTW, as compared

to the control group. Indeed, although we found no coherence changes associated with age in the control

group there was a coherence increase in children with FTW.

The question of the age-related coherence changes associated with the resting state functional

interaction of cortices is worthy of separate discussion. This study showed no significant age-related

changes of coherence both in children with FTW and the control group whereas the study [29] showed

coherence increase of α oscillations in occipito-parietal and temporo-central leads of both hemispheres in

children with FTW from the age of 7-8 to 9-10 years and vice versa coherence decrease of α oscillations in

fronto-central and parieto-posttemporal lead pairs of the left hemisphere in the control group. In this case,

the divergent results from two studies could be also explained by analysis limiting the α range in the first

study and coherence analysis in four EEG frequency ranges performed in the second study. The

longitudinal analysis of synchronization of different EEG rhythms in children from 6 to 12 with typical

development [32] showed that it is α rhythm coherence that undergoes the most pronounced age-related

changes. In the cited study, researchers have been finding α coherence increase in serial EEG recordings

with time lag of approximately 4 years thus revealing the general trend only. Meanwhile some

electroencephalographical studies [19, 33] demonstrated a nonlinear dependence of EEG coherence on

children’s age. Alpha rhythm coherence decrease in the left hemisphere leads in the control group, found in

[29], agrees with data in [33] which states that locus of the fastest growth of EEG coherence shifts from the

left to the right hemisphere during the period from 8 to 10-11 years. The decrease of α coherence mainly

manifested in left hemisphere leads at the age period from 8 to 10-11 followed by another increase till the

age of 16 is also described in [19]. On the basis of the current and previous studies we can assume the

different developmental courses of cortico-cortical connectivity in children with FTW and control group.

The next question that we should discuss in the framework of this study aimed at seeking the

neurophysiological mechanisms of learning difficulties in children is: what is the relationship between

functional organization of the brain cortex in the resting state, on the one hand, and successful cognitive

performance, on the other? An important role of the functional cortical interaction during the resting state,

viewed as a condition that determines the readiness for information processing and overt activity, was

shown in a number of electrophysiological studies that have been traced back to 90s of the previous

century [20, 28, 33, 34]. Mainly it were the studies of age-related changes in spatio-temporal organization

of the brain EA that lightened the fact that even when a person is just relaxing and does not join any

activity the different brain cortices function as a united and relatively stable system. It was also

demonstrated that resting state functional connectivity is changing in its character and strength during the

developmental course [20, 28, 32, 35, 36]. By analyzing the slow (about 0.1 Hz) oscillating metabolic

processes in neuronal networks (BOLD signals) it was found over the last few years that the functional

connectivity between different brain structures is topographically specific not only for cognitive

performance but also for the resting state. Such a functional coupling of brain structures was called a

resting state network (RSN) [37, 38] or an intrinsic connectivity network (ICN) [39]. In most cases, the

resting state network includes structures which are also being activated during different motor and

cognitive performances [37, 40] and structures which are silent (default mode networks - DMN) when

performing a cognitive task [41]. According to S. Bressler [39] resting state networks, topographically

similar to cognitive neuronal networks, give evidence to the fact that the human brain is initially organized

into relatively stable distributed systems. Resting state networks provide a sort of skeleton for more

dynamic functional connectivities during cognitive performance. EEG and fMRT data comparison [41,43]

showed that the configuration of resting state functional connectivity estimated through the analysis of

slow oscillating metabolic processes is mostly similar to that found with EEG analysis. These data prove

the existence of relationship between the spatio-temporal organization of resting state rhythmical EA and

the functional organization of actual neuronal networks, effectiveness of inner linking of which determines

the effectiveness of different cognitive performances. The above mentioned data allow treating the results

of this study as an evidence for the specificity of the morpho-functional brain organization in children with

FTW.

Regarding the current study results in the framework of the reflection of functional brain

organization in the resting state EEG we can assume that the non-optimal fronto-thalamic functioning

impede the actual interaction among frontal, temporal and central cortices of mostly the left hemisphere.

The age-related specificity of differences between control children and children with FTW point to rather

indirect association between the non-optimal fronto-thalamic functioning and the decrease of the left

hemisphere functional cortical interaction. And since it is so, it is reasonable to assume atypical morpho-

functional maturation of cortical neuronal networks in children with FTW. One of the possible reasons for

the sensitivity of the left hemisphere neuronal networks to the non-optimal fronto-thalamic functioning can

be its neurochemical specificity, e.g. the higher, than in the right hemisphere, activity of dopaminergic

neuronal connectivity [43]. Fronto-thalamic system is, in its turn, linked to the limbic-cortical dopamine

system through MD. A lacking involvement of this connectivity can negatively affect the formation of left

hemisphere neuronal networks. The other possible reason is the poor selective modulation of the cortex

exerted by non-specific thalamic nuclei. This can affect negatively the morpho-functional maturation of the

left hemisphere neuronal networks which are known to be more locally specialized than in the right

hemisphere [44].

In previous complex neurophysiological and electrophysiological studies it was shown that children

with FTW demonstrated deficit of the nominative function of speech in addition to the pronounced lack of

executive control. This deficit reflects in poor vocabulary and semantic errors during both speech (writing

by dictation, retelling) and visuo-spatial (recognition and description of visual objects) performance [6, 45].

The quantitative analysis of resting state EEG presented in this study provides an additional information

about neurophysiological mechanisms of cognitive deficits mentioned above. According to Luria’s clinical

studies [46] the nominative function of speech is associated with the left hemisphere functioning and

specifically disturbed in the case of damage to postfrontal and temporal areas of this hemisphere. It was

these areas in children with FTW that showed a relative weakness of cortico-cortical interaction.

Summarizing the results of electroencephalographical and neuropsychological studies of children with

FTW versus the syndromes of local brain injuries allows to assume that the basis of deficit of the

nominative function of speech in children with immature or dysfunctioning fronto-thalamic system can be

an inappropriate functional connectivity between frontal and temporal left hemisphere structures.

Results of this study suggest that the relative weakness of cortico-cortical interaction in fronto-

temporal region of the left hemisphere can be one of the neurophysiological factors of cognitive deficits

and associated learning difficulties in primary schoolchildren, whose EEG shows frontal bilateral-

synchronous θ waves.

CONCLUSIONS

1. Coherence analysis of artifact-free resting state EEG in θ, α 1, α 2 and β 1 frequency ranges

showed that children aged 7-8 and 9-10 with learning difficulties and frontal θ waves (FTW) have a

decreased, as compared to control children of the same age, functional connectivity of frontal and anterior

temporal cortices, the decrease being more pronounced in the left hemisphere.

2. In children with FTW, the specificity of resting state functional cortical connectivity estimated

via analyzing the coherence of brain rhythmic electrical activity did not depend on children’s sex and EEG

frequency range.

3. The non-optimal fronto-thalamic functioning, reflected in FTW presence, exerts more influence

on the frontal cortex functional connectivity in children aged 7-8 than in children of 9-10.

ACKNOWLEDGMENTS

This study was supported by the Russian Foundation for Humanities Research. Grant no. 10-06-00693а.

REFERENCES

1. Machinskaia, R.I., Lukashevich, I.P., and Fishman, M.N., Dynamics of Brain Electrical Activity

in 5-to 8-year-old Normal Children and Children with Learning Difficulties, Hum. Physiol., 1997, vol. 23,

no. 5, p. 5.

2. Machinskaya, R. I., Functional Maturation of the Brain and Formation of the Neurophysiological

Mechanisms of Selective Voluntary Attention in Young Schoolchildren, Hum. Physiol., 2006, vol. 32, no.

1, p. 20.

3. Semenova, O.A. and Machinskaya, R. I., Neiropsihologicheskii i neirofiziologicheskii analiz

vozrastnyh preobrazovanii poznavatel’nyh funktsii i riskov uchebnoi dezadaptatsii v predshkol’nom

vozraste (Neuropsychological and neurophysiological analysis of age-related changes of cognitive

functions and risks of learning and behavioral difficulties in preschool children), Nov. Issled., 2012, vol.

30, no. 1, p. 45.

4. Lubar, J.F., Bianchini, K.J., Calhoun, W.H., et al., Spectral Analysis of EEG Differences

Between Children With and Without Learning Disabilities, J. Learn. Disabil., 1985, vol. 18, no. 7, p. 403.

5. Chabot, R.J., Michele, F., Prichep, L., and John, R., The Clinical Role of Computerized EEG in

the Evaluation and Treatment of Learning and Attention Disorders in Children and Adolescents, J.

Neuropsychiatry Clin. Neurosci., 2001, vol. 13, no. 2, p. 171.

6. Machinskaya, R. I. and Semenova, O. A., Peculiarities of Formation of the Cognitive Functions

in Junior School Children with Different Maturity of Regulatory Brain Systems, J. Evol. Biochem. Physiol.,

2004, vol. 40, no. 5, p. 528.

7. Mayorchik, V.E., Izmeneniia EEG v zavisimosti ot lokalizatsii opuholi mozga (Changes in EEG

Depending on the Localization of Brain Tumor), in Klinicheskaya elektroentsefalografiya (Clinical

Electroencephalography), Rusinov, V.S., Ed., Moscow: Meditsina, 1973, p. 106.

8. Lukashevich, I.P. and Sazonova, O.B., The Effect of Lesions of Different Parts of the Optic

Thalamus on the Nature of the Bioelectrical Activity of the Human Brain, Zh. Vyssh. Nerv. Deiat. Im. I.P.

Pavlova, 1996, vol. 46, no. 5, p. 866.

9. Terao, Y., Sakurai, Y., Sakuta, M., et al., FDG-PET in an Amnestic and Hypersomnic Patient

with Bilateral Paramedian Thalamic Infarction, Rinsho Shinkeigaku, 1993, vol. 33, no. 9, p. 951.

10. Kalashnikova, L.A., Gulevskaya, T.S., and Kashina, E.M., Disorders of Higher Mental Function

Due to Single Infarctions in the Thalamus and in the Area of the Thalamofrontal Tracts, Neurosci. Behav.

Physiol., 1999, vol. 29, no. 4, p. 397.

11. Zoppelt, D., Koch, B., Schwarz, M., et al., Involvement of the Mediodorsal Thalamic Nucleus

in Mediating Recollection and Familiarity, Neuropsychologia, 2003, vol. 41, no. 9, p. 1160

12.Goldman-Rakic, P.S., Porrino, L.J., The primate mediodorsal (MD) nucleus and its projection to

the frontal lobe, J. Comp. Neurol., 1985, vol. 242, p. 535. doi: 10.1002/cne.902420406

13. Zhang, D., Snyder, A.Z., Shimony, J.S., et al., Noninvasive Functional and Structural

Connectivity Mapping of the Human Thalamocortical System, Cerebr. Cortex., 2010, vol. 20, no. 5, p.

1187.

14. Lujan, J.L., Chaturvedi, A., and McIntyre, C.C., Tracking the Mechanisms of Deep Brain

Stimulation for Neuropsychiatric Disorders, Front. Biosci., 2008, vol. 13, p. 5892.

15. Kim, J., Woo, J., Park, Y-G., et al., Thalamic T-Type Ca 2-Channels Mediate Frontal Lobe

Dysfunctions Caused by a Hypoxia-Like Damage in the Prefrontal Cortex, J. Neurosci., 2011, vol. 31, no.

11, p. 4063. doi: 10.1523/JNEUROSCI.4493-10.2011

16. Kurgansky, A. V. and Machinskaya, R. I., Bilateral Frontal Theta-Waves in EEG of 7–8-Year-

Old Children with Learning Difficulties: Qualitative and Quantitative Analysis, Hum. Physiol., 2012, vol.

38, no. 3, p. 37.

17. Kaminski, M., Determination of Transmission Patterns in Multichannel Data, Phil. Trans. R.

Soc. B., 2005, vol. 360, no. 1457, p. 947. doi: 10.1098/rstb.2005.1636

18. Fingelkurts, Al.A., Fingelkurts, An.A, and Kahkonen, S., Functional Connectivity in the Brain –

is it an Elusive Concept?, Neurosci. Biobehav. Rev., 2005, vol. 28, no. 8, p. 827

19. Alferova, V.V. and Farber, D.A., Otrazhenie vozrastnyh osobennostei funktsional’noi

organizatsii mozga v elektroentsefalogramme pokoia (the Reflection of Age-related Specificity of

Functional Brain Organization in Resting State Electroencephalogram), in Strukturno-funktsional’naia

organizatsiia razvivaushegosia mozga (Principles of Structural and Functional Brain Organization in

Ontogenesis), Adrianov, O.S., Ed., Leningrad: Nauka, 1990, p. 45

20. Kniazeva, M.G. and Farber, D.A., Spatial Structure of Intra- and Inter-hemispheric

Connections: Factor Analysis of the Coherence of the Resting EEG, Fiziol. Cheloveka, 1996, vol. 22, no. 5

p. 37

21. Machinskaia, R.I., Neurophysiological Mechanisms of Voluntary Attention: A Review, Zh.

Vyssh. Nerv. Deiat. Im. I.P. Pavlova, 2003, vol. 53, no. 2, p. 133.

22. Joseph, R., Neuroscience: Neuropsychology, Neuropsychiatry, Behavioral Neurology, Brain &

Mind: Primer, Overview & Introduction, University Press Science Publishers, 2011, 4th ed.

23. Watanabe, Y. and Funahashi, Sh., Thalamic Mediodorsal Nucleus and Working Memory,

Neurosci. Biobehav. Rev., 2012, vol. 36, no. 1, p. 134.

24. Huttenlocher, P.R., Neural Plasticity. The Effects of Environment on the Development of the

Cerebral Cortex, Cambridge, Massachusetts, London: Harvard University Press, 2002

25. Hagmann, P., Sporns, O., Madan N. , et al., White Matter Maturation Reshapes Structural

Connectivity in the Late Developing Human Brain, Proc. Natl. Acad. Sci. U S A, 2010, vol. 107, no. 44, p.

9067. doi: 10.1073/pnas.1009073107

26. Segalowitz, S.J., The Role of Neuroscience in Historical and Contemporary Theories of Human

Development, in Human behavior, learning, and the developing brain, Coch, D., Fischer, K.W., and

Dawson, G., Eds., New York: Guilford, 2007.

27. Galvan, A., Neural Plasticity of Development and Learning, Hum. Brain Mapp., 2010, vol. 31,

no. 6, p. 879. doi: 10.1002/hbm.21029

28. Thatcher, R.W., North, D.M., and Biver, C.J., Development of Cortical Connections as

Measured by EEG Coherence and Phase Delays, Hum. Brain Mapp., 2008, vol. 29, no. 12, p. 1400. doi:

10.1002/hbm.20474

29. Machinskaia, R.I., Sokolova, L.S., and Krupskaia, E.V., Formation of the Functional

Organization of the Cerebral Cortex at Rest in Young Schoolchildren Varying in the Maturity of Cerebral

Regulatory Systems: II. Analysis of EEG alpha-rhythm Coherence, Fiziol Cheloveka, 2007, vol. 33, no. 2,

p. 5.

30. Kurganskii, A.V., Study of Cortico-Cortical Functional Connectivity with Vector

Autoregressive Model of Multichannel EEG, Zh. Vyssh. Nerv. Deiat. Im. I.P. Pavlova, 2010, vol. 60, no. 6,

p. 740.

31. Cui, J., Xu, L., Bressler, S.L., et al., BSMART: A Matlab/C Toolbox for Analysis of

Multichannel Neural Time Series, Neural Netw., 2008, vol. 21, no. 8, p. 1094. doi:

10.1016/j.neunet.2008.05.007

32. Gmehlin, D., Thomas, C., Weisbrod, M., et al., Development of Brain Synchronisation within

School-age – individual Analysis of Resting (α) Coherence in a Longitudinal Data Set, Clin. Neurophysiol.,

2011, vol. 122, no. 10, p. 1973. doi: 10.1016/j.clinph.2011.03.016

33. Thatcher, R.W., Cyclic Cortical Re-Organization: Origins of Human Cognitive Development, in

Human Behavior and the Developing Brain, Dawson, G. and Fischer, K., Eds., New York: Guilford, 1994,

p. 232.

34. Marosi, E., Harmony, T., Sanchez, L., et al., Maturation of the Coherence of EEG Activity in

Normal and Learning-disabled Children, Electroencephalogr. Clin. Neurophysiol., 1992, vol. 83, no. 6, p.

350.

35. Shepoval'nikov, A.N., Tsitseroshin, M.N., and Pogosian, A.A., Role of Various Cortical Areas

and Their Connections in the Formation of Spatial Ordering in the Field of Brain Biopotentials During

Postnatal Ontogenesis, Fiziol Cheloveka, 1997, vol. 23, no. 2, p. 12.

36. Bezrukikh, M.M. and Terebova, N.N., Functional Organization of the Brain Cortical Areas at

Rest Condition in Children 5, 6 and 7 years old, Fiziol Cheloveka, 2010, vol. 36, no. 6, p. 61.

37. Damoiseaux, J.S., Rombouts, S.A.R.B., Barkhof, F., et al., Consistent Resting-State Networks

Across Healthy Subjects, Proc. Natl. Acad. Sci. U S A, 2006, vol. 103, no. 37, p. 13848. doi:

10.1073/pnas.0601417103

38. Van den Heuvel, M.P., Mandl, R.C.W., Kahn, R.S., et al., Functionally Linked Resting-State

Networks Reflect the Underlying Structural Connectivity Architecture of the Human Brain, Hum. Brain

Mapp., 2009, vol. 30, no. 10, p. 3127. doi: 10.3389/fnana.2012.00036

39. Bressler, S. L. and Menon, V., Large-scale Brain Networks in Cognition: Emerging Methods

and Principles, Trends Cognitive Sci., 2010, vol. 14, no. 6, p. 277. doi: 10.1016/j.tics.2010.04.004

40. Britz, J., Van de Ville, D., and Michel, Ch.M., BOLD Correlates of EEG Topography Reveal

Rapid Resting-State Network Dynamics, Neuroimage, 2010, vol. 52, no. 4, p. 1162. doi:

10.1016/j.neuroimage.2010.02.052

41. Greicius, M.D., Kaustubh, S., Menon, V., and Dougherty, R.F., Resting-State Functional

Connectivity Reflects Structural Connectivity in the Default Mode Network, Cereb. Cortex, 2009, vol. 19,

no. 1, p. 72. doi: 10.1093/cercor/bhn059

42. Mantini, D., Perrucci, M.G., Del Gratta, C., et al., Electrophysiological Signatures of Resting

State Networks in the Human Brain, Proc. Natl. Acad. Sci. U S A, 2007, vol. 104, no. 32, p. 13170. doi:

10.1073/pnas.0700668104

43. Toga, A.W. and Thompson, P.M., Mapping Brain Asymmetry, Nat. Rev. Neurosci., 2003, vol.

4, no. 1, p. 37.

44. Hutsler, J., Galuske, R.A.W., Hemispheric Asymmetries in Cerebral Cortical Networks, Trends

Cognitive Sci., 2003, vol. 26, no.8, p. 429. doi: 10.1016/S0166-2236(03)00198-X

45. Sugrobova, G.A., Semenova, O.A., and Machinskaia, R.I., Osobennosti reguliatornyh i

informatsionnyh komponentov poznavatel’noi deiatel’nosti u detei 7-8 let s priznakami SDVG

(Peculiarities of Regulatory and Information-Related Components of Cognitive Activity in 7-8 Year Old

Children with ADHD), J. Hum. Ecol., 2010, no. 11, p. 19.

46. Luria, A.R., Vysshie korkovye funktsii i ih narusheniia pri lokal’nyh porazheniiah mozga

(Higher Cortical Functions in Man and Their Impairment Caused by Local Brain Lesions), Moscow: MSU

Press, 1962.

O2

P4

C4

F4

T4

T6

O1

P3

C3

F3

T3

T5

FIGURES, LEGENDS AND TABLES

Figure 1. Resting state EEG of a 7-year-old child. Bilateral-synchronous frontal theta waves (FTW) are

underlined

Table 1. The distribution of sample by age, sex and the presence of frontal θ waves in EEG.

7-8 years 9 -10 years

boys girls total boys girls total

Controls 18 14 N = 32

mean age =

7.68±0.42

10 6 N = 16

mean age =

9.60±0.57

FTW 23 6 N = 29

mean age =

7.82±0.56

20 3 N = 23

mean age =

9.80±0.56

Total 41 20 N = 61

30 9 N = 39

Table 2. Significant influence of frontal θ waves presence (FTW factor) on resting state EEG coherence.

MANOVA results.

Factors Pairs of leads

F-O, F-P,

F-C, F-Ta,

F-Tp

C-O, C-P,

C-Ta, C-Tp

P-O, P-Ta,

P-Tp

Ta –O,

Ta - Tp

Tp-O

Interhemispheric

FTW AGE F(1, 92) = 3.70,

p = 0.057

F(1, 92) = 3.81,

p = 0.045

FTW HEMISPHERE F(1, 92) = 6.61,

p = 0.012

FTW LOCATION F( 4, 89) = 3.16,

p = 0.018

F(3, 90) = 3.30,

p = 0.024

F(5, 88) = 2.62,

p = 0.03

FTW LOCATION

AGE

F(4, 89) = 2.86,

p = 0.028

F(5, 88) = 2.44,

p = 0.041

FTW

HEMISPHERE

LOCATION

F(3, 90) = 5.98,

p = 0.001

F(2, 91) = 5.97,

p = 0.004

F(1, 92) = 7.73,

p = 0.007

Table 3. Mean values and standard deviations of EEG coherence* over the left (LH) and right (RH) hemispheres in 7-8 and 9-

10-year-old children of the control and FTW groups.

Age Pairs Control group Children with FTW

RH LH mean over

hemispheres

RH LH mean over

hemispheres

7-8 F-C 0.741(0.082) 0.766(0.064) 0.753(0.065) 0.711(0.082) 0.735(0.081) 0.723(0.076)

F-P 0.433(0.104) 0.474(0.074) 0.454(0.084) 0.392(0.094) 0.413(0.104) 0.402(0.093)

F-Ta 0.602(0.082) 0.612(0.077) 0.607(0.072) 0.557(0.085) 0.545(0.099) 0.551(0.077)

F-Tp 0.240(0.086) 0.283(0.075) 0.262(0.067) 0.238(0.071) 0.257(0.070) 0.247(0.057)

F-O 0.205(0.069) 0.205(0.070) 0.205(0.065) 0.196(0.061) 0.197(0.058) 0.196(0.054)

C-P 0.739(0.071) 0.769(0.054) 0.754(0.054) 0.714(0.064) 0.714(0.074) 0.714(0.061)

C-Ta 0.710(0.093) 0.720(0.086) 0.715(0.082) 0.655(0.076) 0.666(0.108) 0.661(0.078)

C-Tp 0.449(0.097) 0.510(0.074) 0.479(0.071) 0.425(0.109) 0.456(0.100) 0.440(0.086)

C-O 0.320(0.085) 0.312(0.084) 0.316(0.078) 0.319(0.085) 0.305(0.089) 0.312(0.076)

P-Ta 0.531(0.105) 0.570(0.089) 0.550(0.087) 0.503(0.089) 0.540(0.113) 0.522(0.086)

P-Tp 0.711(0.066) 0.754(0.050) 0.733(0.047) 0.704(0.095) 0.736(0.063) 0.720(0.067)

P-O 0.695(0.050) 0.648(0.070) 0.671(0.047) 0.706(0.066) 0.681(0.079) 0.693(0.066)

Ta-Tp 0.503(0.103) 0.557(0.092) 0.530(0.084) 0.490(0.090) 0.541(0.122) 0.516(0.082)

Ta-O 0.236(0.078) 0.249(0.070) 0.242(0.065) 0.246(0.075) 0.255(0.088) 0.250(0.060)

Tp-O 0.645(0.082) 0.696(0.070) 0.671(0.057) 0.668(0.076) 0.661(0.085) 0.664(0.069)

9-10 F-C 0.705(0.082) 0.766(0.080) 0.736(0.068) 0.725(0.081) 0.751(0.077) 0.738(0.073)

F-P 0.341(0.098) 0.420(0.108) 0.381(0.099) 0.412(0.104) 0.451(0.100) 0.432(0.096)

F-Ta 0.579(0.102) 0.643(0.064) 0.611(0.078) 0.577(0.131) 0.571(0.135) 0.574(0.119)

F-Tp 0.199(0.050) 0.254(0.090) 0.226(0.059) 0.267(0.085) 0.270(0.082) 0.268(0.074)

F-O 0.184(0.066) 0.186(0.079) 0.185(0.063) 0.217(0.090) 0.230(0.089) 0.223(0.087)

C-P 0.700(0.072) 0.721(0.111) 0.710(0.082) 0.716(0.110) 0.739(0.079) 0.727(0.084)

C-Ta 0.707(0.089) 0.744(0.057) 0.726(0.061) 0.664(0.088) 0.691(0.108) 0.678(0.081)

C-Tp 0.365(0.080) 0.457(0.105) 0.411(0.073) 0.449(0.135) 0.448(0.120) 0.449(0.118)

C-O 0.295(0.084) 0.258(0.094) 0.277(0.070) 0.302(0.107) 0.313(0.088) 0.307(0.090)

P-Ta 0.496(0.080) 0.547(0.110) 0.521(0.088) 0.518(0.096) 0.566(0.097) 0.542(0.083)

P-Tp 0.672(0.076) 0.719(0.106) 0.696(0.084) 0.738(0.084) 0.733(0.059) 0.735(0.066)

P-O 0.686(0.119) 0.622(0.098) 0.654(0.097) 0.686(0.082) 0.664(0.064) 0.675(0.059)

Ta-Tp 0.436(0.080) 0.536(0.107) 0.486(0.081) 0.501(0.113) 0.540(0.102) 0.520(0.079)

Ta-O 0.218(0.042) 0.213(0.098) 0.215(0.056) 0.244(0.102) 0.281(0.109) 0.263(0.097)

Tp-O 0.669(0.091) 0.691(0.063) 0.680(0.062) 0.685(0.088) 0.709(0.091) 0.697(0.077)

* table contains descriptive statistics with the RANGE factor excluded; table cells contain MEAN

VALUES (STANDART DEVIATION); coherence values with significant differences for FTW factor

pairwise comparisons are shown in bold.

Table 4*. Mean values and standard deviations of interhemispheric EEG coherence in 7-8 and 9-10-year-

old children of the control and FTW groups.

age pairs of leads control group children with FTW

7-8 F4-F3 0.764(0.056) 0.723(0.076)

C4-C3 0.601(0.082) 0.565(0.115)

P4-P3 0.592(0.107) 0.539(0.113)

T4-T3 0.192(0.083) 0.202(0.077)

T6-T5 0.224(0.059) 0.220(0.065)

O2-O1 0.600(0.092) 0.616(0.068)

9-10 F4-F3 0.736(0.101) 0.744(0.062)

C4-C3 0.570(0.083) 0.596(0.088)

P4-P3 0.534(0.084) 0.557(0.111)

T4-T3 0.195(0.087) 0.234(0.117)

T6-T5 0.180(0.060) 0.207(0.074)

O2-O1 0.592(0.100) 0.629(0.115)

* The format of this table is the same as that for the table 3.