Background: Although autism's social communicative symptoms are amongst its most obvious, most diagnostic, and most debilitating aspects, differences in non-social cognitive domains -- both deficits and superiorities -- also appear. Dimensional traits present in probands and relatives, and in fact are continuously distributed beyond the autism spectrum and into the normal population. Relating social and non-social traits at the levels of behaviour and physiology in typically developing individuals, individuals with autism spectrum conditions (ASC) and siblings without any clinical diagnosis will help elucidate behavioural and neural phenotypes related to autism, and to cognitive diversity in general.
Objectives: (1) Relate psychometric, behavioural and cognitive electrophysiological aspects of autistic traits within and beyond the autism spectrum. (2) Relate these aspects of autistic traits in social and non-social cognitive domains.
Methods: Three age-matched groups of children ages 10 to 15 years -- those with autism spectrum conditions (ASC), their clinically unaffected siblings, and unrelated controls -- were contrasted using event-related potentials, behavioural performance, and psychometric measures of autistic traits. EEG was recrded whilst children played a video game in which were embedded several cognitive tasks, including a go/no-go task and a task of motion coherence (moving dots) detection. Diagnostic group and psychometric measures were factors in an analysis of variance with ERPs or ERSPs as the dependent variable.
Results: The ASC and sib groups were marked by an early, long-lasting positivity (50-250 ms) and a decreased P3 (300-500 ms), at right centro-parietal sites during no-go trials, and this pattern was graded by groups: for the early positivity the siblings manifested the greatest effect, followed by the autism group, whereas this component was not present in the controls. For the P3, amplitude was greatest in the ASC group and least in the control group, with siblings between the two.
Performance on the Benton Facial Recognition test predicted P3 amplitude in the no-go minus go ERP difference wave. There was a significant Group x Benton interaction, wherein Benton was negatively correlated with mean amplitude for controls (p<0.05). For all groups, Benton score was correlated with decreased frontal alpha power (8-12 hz) during go trials (300-2000 ms post-stimulus; p<0.05). There were no group differences for go/no-go accuracy or reaction-time.
In the motion coherence task, the ASC group was differentiated by a longer latency of alpha suppression in response to motion onset, in an independent component sourced in ventral occipital cortex. Psychophysical thresholds for coherent motion in the ASC group were slightly elevated, with much greater variance than the controls.
Conclusions: In typically developing people, performance on a social task of face-recognition is related to electrophysiological measures of a non-social target detection task. Such a relationship suggests overlapping cortical networks for each of these task domains, social and non-social. Similar go/no-go behavioural performance in all groups suggests that the physiological differences may be the more fundamental ones, and these physiological results are consistent with fMRI results of early posterior and delayed frontal activations during visual attention in ASC probands and siblings.
See more of: Neurophysiology
See more of: Brain Structure & Function