Atypical Interference Effect of Action Observation in Autism Spectrum Conditions

Background: Observing incongruent actions interferes with ongoing action execution. This “Interference Effect” is larger if the observed actor has human rather than robotic form (Kilner et al., 2003) and if the movement proceeds with biological motion rather than constant velocity motion (Kilner et al., 2007). The Mirror Neuron System (MNS) is a possible neural substrate for the Interference Effect (Blakemore & Frith, 2005). Little is known of the biological specificity of Interference Effects in Autism Spectrum Conditions (ASC).

Objectives: The current study used virtual reality to manipulate both actor form and motion and thus enable an investigation of the biological specificity of Interference Effects in ASC.

Methods: High-functioning adults with ASC and age-, gender- and IQ-matched typical controls performed horizontal sinusoidal arm movements whilst observing arm movements conducted by a virtual reality agent with either human or robot form, which moved with either biological motion or at a constant velocity. In another condition, participants made the same arm movements while observing a human. Observed arm movements were either congruent or incongruent with executed arm movements. An Interference Effect was calculated as the average variance in the incongruent action dimension during observation of incongruent compared to congruent movements.

Results: Control participants exhibited an Interference Effect when observing real human and virtual human agent incongruent movements but not when observing virtual robot agent movements. Individuals with ASC differed from controls in that they showed no Interference Effect for real human, virtual human or virtual robot movements. Interference Effects did not significantly vary according to the type of motion (biological or constant velocity) observed.

Conclusions: The current study demonstrates atypical Interference Effects in ASC. Given that the MNS is a plausible neural correlate of the Interference Effect the current results can be considered consistent with the broken MNS hypothesis of ASC (Ramachandran & Oberman, 2006). However, more recent theories suggest that rather than the MNS per se being ‘broken’ in ASC it may be that modulation of the MNS is atypical (Cook, Barbalat, Blakemore, 2012; Hamilton, 2008; Spengler et al., 2010; Kana et al., 2011). For instance, social priming has been demonstrated to elevate imitation in typical adults but not in adults with ASC (Cook & Bird, 2011). It may be that, in the current experiment, the human form present in the human agent acts as a ‘pro-social prime’ for typical individuals but not for individuals with ASC. Atypical social modulation of this sort may result in an elevated Interference Effect for the control group but not for the ASC group. Thus the current results are consistent with both broken MNS and atypical modulation accounts of ASC.