Empathy Modulates the Reward Value of Mimicry: Implications for Imitation Based Interventions for Autism

Background: A number of behavioural interventions for children with autism use imitation-based methods to teach social skills (e.g. Early Start Denver Model, Reciprocal Imitation Therapy). However, the mechanism of action for these imitation-based methods is not well established. Social psychological studies have shown that people demonstrate greater liking for those who imitate them, suggesting that imitation/mimicry may alter reward value of social targets (Stel et al., 2008; Stel and Vonk 2010; van Baaren et al., 2003). Functional MRI studies have suggested that the motor-act of mimicking as well as being mimicked can lead to activation of brain regions related to reward processing (Kuehn et al, 2009, 2011; Lee et al. 2006; Likowski et al 2012; Vrticka et al. 2013). These studies lead to the hypothesis that mimicry alters the reward value of a face.

Objectives: To directly test the reward value of mimicry, we measured the impact of being mimicked on preferential looking, using gaze duration as a proxy measure of reward consummation (i.e. longer gaze duration is associated with higher reward value). We further tested individual variability in this effect, as a function of trait empathy (as measured by Empathy Quotient, (EQ)).

Methods: 37 neurotypical adults (17 males) underwent a conditioning experiment in which they performed facial expressions (happy, sad, or neutral). Crucially, <1s after they started making the expression, they saw a video of a person making the same expression (Mimicking face) or another person making a different expression (AntiMimicking face). In the test phase, pairs of ‘Mimicking face’ and ‘AntiMimicking face’ were presented side-by-side on a gaze-tracking monitor. Further, subject’s ability to empathise was assessed via the Empathy Quotient (EQ). To ensure that any observed differences were not related to greater response conflict associated with a given face, we conducted a separate experiment (N=19) investigating the effect of spatial congruency on gaze behaviour.

Results: Participants showed longer gaze duration for Mimicking faces than AntiMimicking faces after conditioning, (controlling for any baseline difference in looking times for the two faces) (t=2.99, p=.005). No such pattern was detected when comparing spatially congruent vs. incongruent faces in the control paradigm. Further, the magnitude of the difference in gaze duration between Mimicking and AntiMimicking faces, was positively correlated with EQ (r=.319; p=.04).

Conclusions: These findings suggest that being mimicked by a face changes its reward value, which is indexed by longer gaze durations. This effect is not driven by different levels of response conflict associated with the different faces, as shown by the results of the control task. Further, this effect is greater in more empathic individuals, suggesting that empathy affects the interaction between mimicry and reward system, i.e. this link seems to be stronger in individuals with higher EQ. These results provide direct evidence for a potential reward-based mechanism through which imitation-based interventions in autism may work, and identify key dimensions of individual variation.