Effects of focused selective attention tasks on event-related potentials in autistic and normal individuals

Contextual probability evaluation in autistic, receptive developmental language disorder, and control children: event-related brain potential evidence

Compared 8- to 14-year-old children with either autism or receptive developmental language disorder (RDLD) to age- and IQ-matched normal controls in their ability to detect both frequent (p = .70) and infrequent (p = .30) randomly presented auditory stimuli under task and no-task conditions. Event-related brain potentials (ERPs), behavioral reaction times, and target detection accuracy rates were measured. Although the three groups of children performed in a similar manner on behavioral measures, only the children with autism demonstrated an abnormally small amplitude of the P3b, a component of the ERP. This result is interpreted in terms of (a) the consistency of this finding with other ERP studies involving older individuals with autism; and (b) its significance with respect to the difficulty children with autism have in modifying their expectancies to contextually relevant sequences of auditory information.

Midlatency auditory evoked responses: P1 abnormalities in adult autistic subjects

MLR recordings from a group of 11 high-functioning adult autistic subjects were compared with those from a control group of 11 normal subjects. Components selected for analysis were "Pa", the maximum positivity in the 25-40 msec latency range following stimulus onset, "P1", the maximum positivity within the 50-65 msec latency range, and "Nb," the maximum negative deflection in the 40-50 msec latency range. Statistical analyses of amplitude and latency data were conducted using repeated measures analysis of variance and t test group comparisons. The Pa component showed no significant difference between autistic and control groups. However, 2 types of abnormality were noted in the P1 component: (1) the P1 component was significantly smaller in the autistic subjects at slow rates of stimulation, and (2) the autistic P1 did not change as rates of click stimulation increased from 0.5 to 10/sec, in contrast to the normally produced P1 decrement. Data from the P1 model in the cat, and complementary data from the human, closely link the generator substrate of the P1 potential to cholinergic components of the ascending reticular activating system (RAS) and their thalamic target cells. This is the first report of abnormal P1 responses in autism and strongly suggests that the RAS and/or its post-synaptic thalamic targets may be dysfunctional in this syndrome.

P3 responses to prosodic stimuli in adult autistic subjects

Autistic persons are known to have serious abnormalities in speech prosody. The present study attempted to ascertain whether autistic persons could discriminate and/or recognize prosodic contrasts in auditory stimuli. A group of 11 adult autistic subjects with normal IQ and an age-matched group of normal subjects were studied electrophysiologically and behaviorally during presentations of prosodic and phonemic stimuli. The cognitive P3 potential was recorded in response to rare (20%)/frequent (80%) presentations of phonemic stimuli, 'ba/pa,' linguistic-prosodic stimuli, 'Bob.' (statement)/'Bob?' (question), and emotional-prosodic stimuli, 'Bob' (happy)/'Bob' (angry). Behaviorally, auditory discrimination was tested by requiring a button-press response to each presentation of the rare target stimulus and cognitive association was tested by requiring a match between the verbalized stimulus and an appropriate picture/word. Contrary to our hypothesis, the autistic subjects generally showed normal P3 responses to all stimuli and performed at a normal level in all behavioral tests. However, a significant autistic P3 response to the phoneme 'pa' was not demonstrated. This surprising result was reexamined and shown to reflect an unusually large autistic response to 'pa' as the frequent stimulus in the first recording block, this initial hyper-reactivity prevented a 'frequent/rare' differential when 'pa' was presented as the rare stimulus in a later recording block. In the P3 latency window, both the autistic and control groups showed the largest amplitude responses to emotional-prosodic stimuli; neither the N1 nor P2 showed these stimulus effects. Thus, 'emotional sounds' appear to be particularly effective in activating the neural substrate of the P3 generator system. Overall, these data indicate remarkably normal P3 and behavioral processing of prosodic stimuli by the high-functioning autistic subjects of this study.

P300 seasonal variation

The P300 (P3) event-related brain potential (ERP) was elicited with auditory stimuli in four different groups of normal, young adult subjects (n = 20 in each group) whose ERPs were obtained at different times of the year. P3 amplitude was largest during spring and summer and smaller for the subjects who were assessed during fall and winter. P3 latency was not affected reliably by changes in the seasons. Measurement of the daily sunshine present when each subject was recorded indicated that P3 amplitude is influenced by seasonal variation in the available amount of sunshine. These findings suggest that comparison of P3 amplitude across subject groups must take seasonal differences into account.

Brainstem and middle latency auditory evoked potentials in autism and developmental language disorder

Brainstem auditory evoked potentials (BAEP) and middle latency responses (MLR) were studied in 8 nonretarded subjects with infantile autism (mean age = 23.3, SD = 2.8), 8 subjects with receptive developmental language disorder (mean age = 16.3, SD = 1.4), and normal control subjects matched to each group for age, gender, and Performance IQ. Click stimuli were delivered monaurally to the left and the right ear and binaurally for both the BAEPs (70-dB HL, 7/sec) and the MLRs (60-dB HL, 13/sec). Amplitudes and latencies (Waves I to VI), interwave latencies (III-V, I-V, and I-III), and Wave I/V amplitude ratio of the BAEPs were determined for each group. For the MLR study, Wave Na, Pa, and Nb latencies, and Wave Na-Pa and Pa-Nb amplitudes were calculated. There were no consistent differences in the BAEP and MLR characteristics of the control and the experimental groups. These results suggest that the abnormal cognitive processes indexed by the cognitive and attention-related event-related potential components in infantile autism and receptive developmental language disorder are not due to abnormal sensory processing in the brainstem and in areas central to the brainstem whose activity generates the BAEPs and MLRs.