Alexithymia Mediates Facial Muscle Mimicry of Observed Pain in Autism

Background: Poor empathy and deficits in mimicry of others’ emotions have been reported in autism spectrum disorder (ASD). However, studies have either examined only cognitive aspects of empathy or only measured self-reported empathy. Furthermore, most studies have not adequately controlled for cognitive factors influencing empathy expression such as alexithymia; that is, the inability to express or understand one’s own emotions.

Objectives: This study was the first to measure involuntary facial mimicry of painful expressions as an indicator of empathy in ASD. Muscle reactivity was examined in individuals with varying levels of autism traits. We hypothesised that muscle mimicry and subjective affective empathy are strongly correlated with alexithymia, but are not impaired in ASD once alexithymia has been controlled for. We further hypothesised that the ability to understand others’ emotions, i.e. cognitive empathy, would enhance affective responses.

Methods: The relationship of cognitive empathy and alexithymia to muscle reactivity and subjective affective responses to pain were examined. In total, 98 individuals participated (27 ASD; 79 Male; ages 14 – 46). Participants were shown videos of painful facial expressions lasting 5s each, during which activity in the facial muscles associated with pain (verum; M. orbicularis oculi, M. zygomaticus major) and surprise (control; M. medial frontalis) were recorded using bipolar surface electromyography. Affective state and perceived pain was rated on a 7-point scale after completion of the videos. Autism traits were calculated from observed and self-reported autism behaviours using the Autism Diagnostic Observation Schedule 2 and the Autism Spectrum Quotient. Alexithymia was measured on the Toronto Alexithymia Scale (TAS-20) and cognitive empathy was measured using the Faux Pas task. Affective states, perceived pain and muscle reactivity were predicted from autism traits, alexithymia and cognitive empathy in separate mixed effect models.

Results: Muscle activity increased significantly from the start of the video to the time of greatest pain and was greater in muscles associated with pain than muscles associated with surprise. High trait autism was correlated with lower muscle amplitude, reduced perceived pain intensity and lower subjective affective ratings while observing pain. However, once alexithymia was controlled for, autism traits were no longer correlated with muscle amplitude or affective states. Furthermore, there were no differences in habituation or speed of muscle response between low and high autism trait participants. Cognitive empathy did not predict muscle reactivity or affective responses.

Conclusions: Alexithymia mediates the relationship between autism traits and affective empathy and muscle mimicry. Interestingly, though affective responses to perceived pain are unimpaired in those without alexithymia, participants with high autism traits may report perceived pain as less intense than those with low autism traits. Of note, complex cognitive empathy does not seem necessary for an affective empathic response to perceiving another’s pain. These results support theories of separate cognitive and affective neural pathways for empathy.