The prevalence of autism spectrum disorders is approximately 1% of the population (6). It occurs more commonly in boys (estimated prevalence: 1.4%) than girls (estimated prevalence: 0.03%). The available evidence suggests that the prevalence of autism spectrum disorders is increasing over time. While it appears that a relatively greater percentage of affected individuals now receive a diagnosis and that diagnosis tends to occur at an earlier age, the possibility of a true increase in risk occurring with time cannot be excluded (6). The diagnostic criteria for autistic disorder are as follows (7).

A total of at least six items from the three diagnostic criteria, with at least two from first, and one each from second and third.

Delays or abnormal functioning in at least one of the following areas, with onset prior to age 3 years: (i) social interaction, (ii) language as used in social communication, or (iii) symbolic or imaginative play.

The disturbance is not better accounted for by Rett disorder or childhood disintergrative disorder.

Asperger disorder, also known as Asperger syndrome, is generally considered an “autism spectrum” disorder. It shares features with autistic disorder, but is without the impairment in communication and delays in cognitive development that typify autistic disorder. The prevalence of Asperger disorder is approximately 0.026% of the population (8). The diagnostic criteria for Asperger disorder are as follows (7):

Rett syndrome is a neurodevelopmental disorder that primarily affects females (9). It has an incidence of 0.01% in girls by age 12. Rett syndrome is characterized by a period of normal development that is followed by loss of hand coordination, impairment in communication ability, and the development of hand stereotypies, although in some cases the period of normal development may be absent (9). The primary causes of this disorder are mutations in the MECP2 gene that codes for methyl-CpG-binding protein. The diagnostic criteria for Rett syndrome are as follows (7):

Onset of all of the following after the period of normal development:

Supportive criteria include breathing dysfunction, seizures, spasticity, scoliosis, and growth retardation. The diagnosis of Rett disorder is considered tentative until 2 to 5 years of age. The differential diagnosis includes other disorders associated with mental retardation, cerebral palsy, and seizure disorders.

Childhood disintegrative disorder is characterized by normal development until at least age 2 followed by a marked decline in a number of areas of function (10). The prevalence of childhood disintegrative disorder is estimated to be 0.002% of the population (10). The diagnostic criteria for this condition are as follows (7):

PDD-NOS is defined by impairments in the development of social and/or communication skills and stereotypies as may occur in other PDDs; however, the diagnostic criteria for one of these other PDDs are not met. The prevalence of PDD-NOS is estimated to be 0.37% of the population. The diagnostic criteria for PDD-NOS are as follows (7).

A severe and pervasive impairment in the development of reciprocal social interaction or verbal and nonverbal communication skills, or when stereotyped behavior, interests, and activities are present, but the criteria are not met for a specific PDD, schizophrenia, schizotypal personality disorder, or avoidant personality disorder. For example, this category includes atypical autism—presentations that do not meet the criteria for autistic disorder because of late age of onset, atypical symptomatology, or subthreshold symptomatology, or all of these.

A number of different types of EEG abnormalities have been reported in patients with autism spectrum disorders. The most common abnormality described in the literature is epileptiform activity. There are several reports of background slowing and alterations in EEG hemispheric symmetry. A group of studies also identify differences from nonautistic controls in mu rhythm activity that is interpreted as indicative of mirror neuron dysfunction. A small number of studies also document deviations from controls in evoked and event-related potentials and in indices derived from EEG spectral analysis. In considering this literature it is important to be aware of the effects of subject selection on the findings. Studies varied in the severity of autism in the samples evaluated and in other subject selection factors that were likely to have influenced the observed findings (12).

Epileptiform Activity. The available studies vary in the frequency with which epileptiform activity is found in the EEG in patients with autistic disorder. One study evaluated a group of 32 patients with autistic disorder who were referred either for seizure evaluation (clinical seizures have been reported in 5% to 46% of those with autistic disorder) (12) or for 24-hour interictal monitoring (13). In this population, 59% were found to have interictal epileptiform abnormalities, which consisted of (in descending order of frequency) generalized spike-wave complexes, focal sharp waves, multifocal sharp waves, and generalized paroxysmal fast activity/polyspikes.

While this relatively small study might seem likely to overestimate the frequency of epileptiform activity, given that the majority of subjects were referred for seizure evaluation, a similar frequency of epileptiform activity was found in a much larger study, where subjects were not referred for seizure evaluation. Of 889 patients with autism spectrum disorders without known genetic conditions, brain malformations, prior medications, or clinical seizures undergoing ambulatory EEG monitoring, 60.7% were found to have epileptiform activity in sleep (14). In these patients the epileptiform abnormalities were most commonly seen in the right temporal area. Of note, approximately half of the patients treated with valproic acid experienced EEG normalization and decreased epileptiform activity was evident in another roughly 20%.

A relatively high frequency of epileptiform activity was also noted in a study of 86 autistic disorder patients, where EEG data were evaluated during sleep. In this study epileptiform activity was observed in 43% of cases (15). The epileptiform activity was most commonly focal spikes (73%), which were most commonly arising from the frontal region.

Several studies have reported a much lower frequency of epileptiform abnormalities. Among these is a study of 316 autism spectrum patients, where 18% undergoing evaluation were found to have EEG epileptiform activity (16). In those with abnormal EEG activity, this was most commonly observed in the temporal regions followed by the central region.

Similarly, among 57 patients with autism spectrum disorders, 25% were found to have interictal epileptiform abnormalities (17).

A relatively low frequency of epileptiform activity (18.9%) was also found in a study of 106 patients with autistic disorder (18). The epileptiform activity was mostly focal and multifocal and in nearly half of the cases was typical of benign childhood partial epilepsy with centrotemporal spikes.

Thus, the available literature does not provide a definitive picture of EEG epileptiform activity in autistic disorder. The reported frequency of epileptiform activity is highly variable (18% to 61%) as is the spatial distribution of the activity noted. This variability in the studies likely reflects that those diagnosed with autistic disorder are a heterogenous group that varies substantially in the nature of neurophysiologic abnormalities and severity (12). However, it is clear that EEG epileptiform activity is not rare in those with autistic disorder.

Slow-Wave Activity. Rates of EEG slowing were reported by few studies of patients with autistic disorder. The limited available data suggest that either focal or diffuse slowing is seen in roughly 15% of patients with this disorder (16).

Mu Activity. Several studies have evaluated mu activity in patients with autistic disorder based on the hypothesis that the failure to suppress mu activity when observing behavior in others reflects dysfunction of the mirror neuron system in this disorder (19). This hypothesis is based on the model that the mirror neuron system is important for the capacity to imitate, which is believed to play a key role in the development of appropriate social behavior, and some postulate that dysfunction in this system might be the basis for the social deficits in those with autistic disorder (19,20).

In one study, 14 relatively high-functioning adults were compared with 15 IQ- and age-matched controls (20). Those with autistic disorder were found to have diminished capacity to imitate, which was correlated with a decreased attenuation of the mu rhythm in the EEG when observing movement.

Decreased mu rhythm suppression was also noted in a study of ten high-functioning children with autistic disorder (ages 8 to 13 years) who were compared to age- and gender-matched controls (21). While the control subjects exhibited suppression of mu activity in response to moving their own hand and observing hand movement, the autistic disorder subjects suppressed mu activity when moving their own hand, but not when observing hand movements.

In contrast to these two studies, one study evaluating a similar population, 20 high-functioning children (ages 8 to 13 years) with autistic disorder who were compared with 20 controls matched for age and intelligence, failed to find a lack of suppression in mu activity in response to observed hand movements in the autistic disorder patients compared with the controls (19). The discrepancy between this finding and studies demonstrating a failure in mu rhythm suppression in response to observed behavior was assumed to reflect the heterogeneity of the population of those with autistic disorder.

Evoked and Event-Related Potential Studies. A series of studies employing different evoked and event-related potential methodologies have been carried out in autistic disorder patients. These studies have been aimed at identifying physiologic correlates of attributes of this disorder. One of these is sensory gating that was studied by assessing the degree of suppression of the EEG response (normally occurs 50 msec after the auditory stimulus and is referred to as the P50) to the second of a pair of clicks (22). Suppression of the P50 in response to the second of the two clicks was present in the autistic disorder subjects with mental retardation but not the high-functioning subjects with autism. The degree of suppression increased with age in both the controls and those with autism. However, a second study failed to find an abnormal P50 response (23). That study examined the response to a pair of clicks while subjects watched a silent movie in 21 children (4 to 8 years) with autistic disorder and age-matched controls. They found no between-group difference in the response to the second click, but noted a decreased amplitude of response to the first stimulus.

Four studies were carried out assessing visual processing with evoked and event-related potential technology. One study included 20 autistic disorder subjects and 20 controls and employed independent component analysis of visual-evoked potential data (24). This study identified that the autistic disorder subjects increased alpha and gamma power to a lesser degree as a function of changes in stimulus spatial frequency, and there was a shorter time to peak alpha frequency band power in those with autistic disorder. Another study assessing visual processing in autistic disorder subjects examined the attention-related frontal event potentials and sustained attention-related centroparietal potentials in a three-stimulus oddball experiment, where 128 channels of EEG data were recorded (25). The subjects were 11 high-functioning children and young adults with autistic disorder and age-matched controls. The subjects with autistic disorder were found to have higher amplitude and longer latency early responses (P100, N100) to novel distracer stimuli and longer latencies of later responses to novel distractor stimuli (P2a, N200, P3a). For both early and late responses, between-group differences were greater in the right hemisphere. The third study evaluated EEG coherence in response to intermittent photic stimulation in 14 relatively high-functioning boys with autistic disorder (ages 6 to 14 years) and 19 controls (26). Stimuli were applied at frequencies between 3 and 27 Hz and outcome was assessed in terms of the amplitude of the photic driving response and the number of pairs of intrahemispheric leads, where the EEG coherence was “>0.6 to 0.8.” Decreased photic driving response was seen in the right hemisphere in the subjects with autistic disorder. There were also more high-coherence pairs of leads in the left hemisphere in the autistic disorder group. One additional study assessed the processing of visual detail based on the hypothesis that the attention to visual detail is increased in autistic disorder (27). In this study data from 13 subjects with autistic disorder (aged 16 to 28 years) were compared with that of 31 controls. They found that the subjects with autism had an increase in occipital activity 225 msec after stimulus presentation, which was interpreted as evidence of a specific neural abnormality in low-level visual processing.

Thus, studies of sensory gating and visual processing in autistic disorder subjects do not provide a coherent picture of neurophysiologic abnormalities in this condition.

Measures Derived from Spectral Analysis of Background EEG Data. Several studies have examined indices derived from spectral analysis of eyes-closed resting EEG data. As in the other types of EEG studies of autistic disorder, a coherent set of findings do not emerge from these studies. In a study of 44 boys with autistic disorder (ages 3 to 5 years) and age-matched controls greater prefrontal delta power, greater gamma (24.4 to 44.0 Hz) activity, and greater left-right asymmetry in the delta, theta, and alpha bands were found in the autism subject group (28). In another study carried out in 20 children with autistic disorder and gender-, age-, and IQ-matched controls (29), the autism group had increased theta power in the right posterior region, decreased frontal delta power, increased midline beta power, decreased intrahemispheric delta and theta coherence, decreased delta and theta coherence frontally, and decreased delta, theta, and beta coherence posteriorly. Finally, in a study of 18 adults with autistic disorder and 18 controls (30), those with autistic disorder were found to have increased power in the 3 to 6 Hz and 13 to 17 Hz bands and less 9 to 10 Hz power as well as elevated theta coherence in the left frontal temporal regions and between frontal and all other regions.

Few studies have addressed EEG findings in patients with the Asperger syndrome. However, some of the studies of autistic disorder where the subjects were “high-functioning” likely included some subjects who met criteria for the Asperger syndrome. In addition, several studies that primarily included subjects with autistic disorder included a small number of subjects with Asperger syndrome; however, the number of such subjects evaluated was too small to draw any useful conclusions (31,32).

A relatively greater number of studies have evaluated the EEG in patients with the Rett syndrome, although most included a small number of subjects. The largest study evaluated EEG data in 52 girls with the Rett syndrome, primarily aged 2 to 7 years (33). Of the 52 girls studied, 43 were found to have sharp waves or spikes that were often asymmetrical and were usually enhanced by light sleep.

Diffuse or bilaterally synchronous spikes or sharp waves were also noted in half of the 44 patients evaluated in another study (34). The most common abnormality noted in the EEG in this study was diffuse rhythmic slowing. A slow spike-wave pattern was also noted in nine subjects.

Similar findings were also noted in a study of 30 girls with Rett syndrome (35). In this study in addition to background slowing, sharp waves, and spikes, some of those more severely affected were noted to have rhythmic delta slowing and sometimes a generalized periodic spike pattern.

The findings of a number of smaller studies carried out in this population are consistent with those of the larger trials in identifying diffuse slowing and spike- or sharp-wave activity (36, 37, 38, 39 and 40). However, several studies noted focal epileptiform abnormalities and centrally predominant spikes (41,42).

No studies have been carried out that document the characteristics of EEG data in childhood disintegrative disorder.

Few reports document EEG findings in patients with PDD-NOS. Those that have, included a group of patients with PDD-NOS, which was small relative to the size of other groups studied, most commonly autistic disorder (17). Because the EEG findings in the subgroup of patients with PDD-NOS were not specifically described, the EEG findings characteristic of those with PDD-NOS remain unknown.

Few studies have been carried out examining epileptiform activity in the background EEG in ADHD patients. One study reported findings in the background EEG in 176 children with ADHD and noted normal studies in 46% of the patients (45). Epileptiform activity was found in 30% of cases that consisted primarily of focal occipital or temporal spike-wave complexes with occasional patients manifesting bilaterally synchronous spike and wave activity. In another study of background eyes-closed resting EEG data in 30 children with ADHD without neurologic symptoms or findings, epileptiform activity was noted in 10% of the patients (46). Thus, the few reports of background EEG activity in ADHD patients suggest that epileptiform activity is not a rare occurrence in this population.

A number of studies employing evoked and event-related potential methodologies have been carried out in order to evaluate cognitive preparatory processes, attention systems, and systems thought to be involved in inhibition of behavior in ADHD patients (46,47). These studies report a complex set of findings that do not provide an electrophysiologic characterization of ADHD. In these studies, ADHD patients were found to differ from control subjects, including in the visual-evoked response, the P300, and the frontal inhibitory response (47).

Several studies evaluated different aspects of EEG activity occurring during attention tasks. In one study of adults, 38 ADHD patients were compared with 42 non-ADHD controls, greater 8 to 10 Hz, 13 to 14 Hz, and 17 to 18 Hz EEG power was found in the ADHD patients in frontal and parietal regions during a sustained attention task (48). A similar study examined EEG power spectral indices after cues that were given to aid in subsequent tests of discrimination and found a relationship between midline theta and alpha activity only in the controls (49). Event-related potential data were also obtained during the continuous performance test (CPT) in 17 boys with ADHD and 20 healthy controls and a significantly lower P3a response was observed in unmedicated ADHD patients (50).

Studies have evaluated response inhibition using a “go-nogo” task. In one such study, a trend was found for adolescents with ADHD (N = 23) to have smaller frontocentral error positivity (51). Another study evaluated 20 children with ADHD and age-matched controls and found a lower P2 and central N2 in the ADHD patients (52).

A much larger number of studies have carried out spectral analysis of background EEG data in ADHD patients. These studies suggest that ADHD patients as a group tend to have elevated relative theta power, reduced relative alpha and beta power (predominantly frontally), and, as a result, elevated ratios of theta/alpha and theta/beta activities (53, 54, 55, 56, 57, 58 and 59). A few relatively small studies have also reported an increase in delta power in ADHD patients compared with controls (46,60).