Abnormal EEG lateralization in boys with autism
Introduction
Autism spectrum disorders represent complex developmental disabilities with manifestations that change with age. The abnormal early increase in brain growth and changes in grey and white matter volume indicate an early onset of this neurodevelopmental disorder (Courchesne et al., 2001, Aylward et al., 2002). Studies in very young children are of particular interest for understanding the pathogenesis of autism, but the possibility of functional neuroimaging is limited in investigations of very young children. In contrast, electroencephalogram (EEG) can be recorded even in infants. Therefore, this method is of potential interest for both exploratory purposes and early differential diagnosis of autism.
To date there is no agreement on the EEG features of autism. Although clinical EEG studies generally agree on the high prevalence of epileptiform abnormalities in patients with autism (Rossi et al., 1995, Kawasaki et al., 1997, Gabis et al., 2005, Chez et al., 2006), the results of a few available quantitative EEG studies are strikingly controversial. Stevens and Milstein (1970) claimed that in some children with autism a ‘hyper-mature’ EEG pattern characterized by unusually persistent alpha rhythm can be observed during attention to visual and auditory stimuli. This pattern, however, disappeared when the subjects were involved in stereotyped behavior. Cantor and co-authors (1986), on the other hand, recorded EEG in mentally retarded children with autism under an undefined behavioral condition that required only that the child remained ‘fairly still’, and concluded that the high amount of slow rhythms in their EEGs indicated a developmental lag. Dawson and colleagues (1995) recorded EEG in children with autism during visual attention and found abnormally decreased EEG spectral power over frontal and temporal areas in the delta, theta and alpha frequency ranges, but normal power in the beta range. In contrast, Bashina et al. (1994) observed decreased spectral power in alpha1/2 bands (7.5–11 Hz), but increased spectral power in delta, alpha3 (11.5–13 Hz) and beta bands, ‘at rest’ in children with autism. In addition, an abnormal EEG asymmetry was reported in a few studies (Cantor et al., 1986, Dawson et al., 1995, Bolduc et al., 2002, Lazarev et al., 2004).
The data on EEG asymmetry are of particular interest, because abnormal brain lateralization is considered an important feature of autism in neuroimaging (Chiron et al., 1995, Muller et al., 1999, De Fosse et al., 2004, Chandana et al., 2005, Herbert et al., 2005) and neuropsychological (Klin et al., 1995, Sabbagh, 1999, Gunter et al., 2002) studies. Again, however, available EEG findings disagree even in respect of the directions of the altered interhemispheric asymmetry in autism spectrum disorders (Ogawa et al., 1982, Dawson et al., 1995, Bolduc et al., 2002, Daoust et al., 2004).
The ambiguity of the published EEG results in autism may stem from several sources. The first is differences in experimental conditions during EEG registration across previous studies, which often did not provide an exact description of the children’s behavioral state. In healthy awake children, the EEG pattern is strongly dependent on current behavioral state (Stroganova and Posikera, 1993, Orekhova et al., 2006). The same is certainly true for children with autism. Hutt et al., 1964, Hutt et al., 1965 reported that a ‘flat’ EEG of children with autism tended to synchronize when they were placed in a simple environment and less stressed. We have observed (unpublished data) that stereotyped behaviors, typical of children with autism, may provoke slow-wave EEG oscillations. These considerations suggest that reported quantitative EEG differences between autism spectrum disorder and control groups may be both trait and state dependent. Thus, without precise control for behavior, comparisons of EEG parameters between experimental groups may lead to inconclusive and incompatible results.
The second probable source of the contradictions is the wide age range of investigated subjects. In autism, both the absolute values of studied parameters in different structural and functional domains and the pattern of their age-related changes differ from normal (Kemper and Bauman, 1998, Chugani et al., 1999, Courchesne et al., 2003, Ferri et al., 2003, Carper and Courchesne, 2005, Hazlett et al., 2005). Therefore, EEG abnormalities in autism may be age-specific and averaging over a wide age range may blur or even cancel out true effects.
A third source may be the high variability of autism etiologies and their different prevalence in different studies. Diverse brain disturbances may lead to the common behavioral phenotype of autism but may still be associated with different EEG abnormalities. Therefore, large experimental groups and reproduction of results in independent experimental samples are important.
In the present study, we analyzed spectral power and asymmetry measures of ongoing EEG in 3–8-year-old boys with autism (BWA) and compared these with corresponding measures in typically developing boys (TDB) of the same age. The EEG was recorded under a controlled behavioral condition (sustained attention to moving visual stimuli). This experimental condition is often used in EEG research of infants and young children because it enables relatively prolonged artefact-free EEG recording and produces a consistent EEG pattern characterized by presence of central sensorimotor (mu) rhythm and relatively low amount of theta and posterior alpha rhythms (Stroganova et al., 1999, Marshall et al., 2002, Orekhova et al., 2006).
We performed combined analysis of data obtained from two independent samples in Moscow and Gothenburg, and assessed the reproducibility of results across samples. Given the inconsistency of previous EEG findings in autism, this two-sample approach was considered a necessary precaution. Both EEG spectral power (SP) and asymmetry scores within delta, theta and alpha frequency bands were analyzed. Prior to statistical analysis, the data were checked for normality of distributions and presence of outliers. The necessity of this procedure was dictated by the statistical assumptions of parametric analysis. Besides, the outliers could differ from the rest of the sample by having some specific etiology of the disorder or some unrevealed co-morbid condition. We hypothesized that EEG parameters of ongoing EEG would differ between BWA and TDB and that an abnormal direction of EEG asymmetry would be a more stable feature of autism than abnormally increased or decreased EEG SP.
Section snippets
Participants
The BWA group included 42 boys with autism and 2 boys with other autism spectrum disorders. The inclusion criteria for BWA were (1) age between 3 and 8 years; (2) birth after at least 36 weeks of gestational age and (3) absence of neurological disorders of known etiology (Fragile-X, epilepsy, etc.). We did not include girls and prematurely born children because sex and prematurity interact with brain lateralization (O’Callaghan et al., 1993, Kovalev et al., 2003) and may influence EEG asymmetry.
Results
For the BWA of the Gothenburg sample the distribution of theta2 power deviated from normal at 8 of 16 lateral electrode positions and at all midline positions. Case analysis revealed that four boys were outliers, and two of them even extremes at some of these electrode positions, having much higher theta2 power than the rest of the group. When two extremes and one child who was an outlier at several electrode positions were excluded from the analysis, the distributions became closer to normal
Discussion
The main results of the present study can be summarized as follows: (1) unlike typically developing children, children with autism constitute a heterogeneous group which includes subjects with distinctly different EEG patterns; (2) when a few (cf below) highly deviant EEG variants were not considered, there were no clear differences between typically developing children and children with autism in mean SP within theta or alpha bands; (3) the main distinctive feature of spontaneous EEG
Acknowledgements
We thank Ulla Sandblom for her skilled assistance and Tomas Karlsson, Johan Kling and Andreas Gingsjö for their valuable technical support. We are grateful to Natalia Rimashevskaya for helping with collection of the Moscow autistic sample. This study was supported by the Swedish Research Council (proj. 12170, proj. 2003-4581), the Gothenburg Medical Faculty, and the Bial Foundation Bursaries for Scientific Research (Grant 87/04).
References (77)
- et al.
Superior temporal sulcus anatomical abnormalities in childhood autism: a voxel-based morphometry MRI study
Neuroimage
(2004) - et al.
Localized enlargement of the frontal cortex in early autism
Biol Psychiatry
(2005) - et al.
Significance of abnormalities in developmental trajectory and asymmetry of cortical serotonin synthesis in autism
Int J Dev Neurosci
(2005) - et al.
Frequency of epileptiform EEG abnormalities in a sequential screening of autistic patients with no known clinical epilepsy from 1996 to 2005
Epilepsy Behav
(2006) - et al.
EEG spectral analysis of wakefulness and REM sleep in high functioning autistic spectrum disorders
Clin Neurophysiol
(2004) - et al.
Theta rhythms in the EEG: a genetic trait in childhood epilepsy
Brain Dev
(1988) - et al.
The mismatch negativity and the P3a components of the auditory event-related potentials in autistic low-functioning subjects
Clin Neurophysiol
(2003) - et al.
Autism and epilepsy: cause, consequence, comorbidity, or coincidence?
Epilepsy Behav
(2005) Autistic children’s hand preferences: results from an epidemiological study of infantile autism
Psychiatry Res
(1983)- et al.
Striatal volume on magnetic resonance imaging and repetitive behaviors in autism
Biol Psychiatry
(2005)