Self-Other Correspondence In Joint Attention and Autism

Background:

It has been argued that impairment in self-other correspondence may constitute core impairment in autism that applies to joint attention (JA) possibly through impaired function of mirror neurons. Deficits in JA are strong diagnostic indicators of autism. An important feature of JA is that 2 or more people maintain an awareness that they have a common focus of attention. This may depend upon self-other correspondence processes that serve to relate perceived action to enacted gaze-direction. We recently developed an fMRI paradigm to explore the neural correlates of correspondence between self and other for gaze direction and hypothesized that they would be affected by autism.

Objectives: To compare participants with and without autism in an fMRI study of joint attention.

Methods:

Stimuli consisted of video clips whereby a dot moved randomly between 4 corners of the screen. Participants were instructed to follow the target dot by looking at it as it moved. In the centre of the screen, was either a person’s head or an arrow. In 3 conditions, the person was seen to a) look at the dot as it moved (congruous condition), b) make similar movements but direct attention to a different corner of the screen (incongruous condition), or stay still (baseline). In 3 further conditions, the head was replaced by a symbolic stimulus in the form of an arrow (for incongruous and incongruous conditions) or a bar (for baseline). 13 adolescent participants with normal IQ and autism, and 13 age and IQ matched controls were participants.

Results:

Clusters in medial frontal cortex and the intra-parietal cortex resulted from a group difference in an interaction between cue-type and congruity. Both clusters were driven by greater activity in the incongruous compared to the congruous condition for faces, contrasted with the opposite for arrows. Therefore, intraparietal and medial frontal cortex were engaged by the arrow in an opposite way to the face, being active according to whether the direction of the arrow matched, or the direction of the face mismatched the direction of action being executed by the observer. This interaction only occurred for controls and the autism group showed no effect of condition at these locations. No group differences for the arrow vs face contrast meant these results cannot be explained as differences in attention to faces.

Conclusions:

The parietal brain area that was more active in the control but not the autism group is particularly concerned with providing control in the presence of competing demands for visual attention. The area of medial frontal cortex that was more active in the control but not the autism group is associated with mental state attribution. The presence or absence of correspondence between action-direction executed by the self and perceived action-direction had marked influence on activity in these areas, suggesting that detection or monitoring of self-other correspondence is very relevant to mental state attribution and control of visual attention. This absence of this in people with autism supports the hypothesis that a deficit in self-other correspondence underlies problems with both social cognition and executive function.