17421
Emotional Prosody Processing in Behavior and Brain Function: Insights from Autism Spectrum Disorder
Social communication crucially depends on efficient processing of subtle nonverbal cues such as the emotional tone of voice, i.e., prosody. Social emotions convey important information about the speaker’s attitudes and intentions (Burnett et al., 2011), thus understanding social emotional prosody is particularly important for successful social interactions. Although impairments in social communication have been consistently identified as a core feature of Autism Spectrum Disorder (ASD) (e.g., Tager-Flusberg, 1981), specific impairments in emotional prosody processing of individuals with ASD remain an under-researched topic with inconclusive results.
Objectives:
In this study, we investigated emotional prosody processing in adult individuals with ASD and healthy controls with newly developed behavioral and functional magnetic resonance imaging (fMRI) tasks. To increase ecological validity of the tasks and thus their sensitivity to the social impairments of individuals with ASD (e.g. Volkmar et al., 2004), the audio stimuli comprised a variety of social emotions (e.g., jealousy), different speakers, as well as implicit and explicit task conditions.
Methods:
Twenty-seven individuals with ASD and control participants (N=22), matched for age, gender and intelligence level, performed the new behavioral prosody recognition task. In this task participants listened to semantically neutral sentences that contained emotional prosody, whereby the majority of items comprised social emotions. Subsequently, participants were asked to label the emotional prosody from four different options. In a separate session, a sample of 20 ASD participants and 21 matched controls, largely overlapping with the initial behavioral sample, performed the emotional prosody fMRI task. In this task participants were presented with semantically neutral audios spoken with emotional or neutral prosody and had to either indicate the speaker’s gender (implicit condition) or the correct emotion label from two options (explicit condition). Audios with emotional prosody covered six basic and six social emotions, matched for valence and arousal levels as determined by pre-ratings.
Results:
Compared to controls, individuals with ASD were slower and less accurate at recognizing social emotions in the behavioral task (RT: t (1, 47) = -2.23, p = 0.03; accuracy: t (1, 41) = 2.72, p = 0.006). On the brain level, emotional prosody processing recruited a fronto-temporal network including the superior temporal sulcus (STS) and the inferior frontal gyrus (IFG) bilaterally in both groups. Importantly, the functional coupling between the right STS and right IFG was significantly reduced for emotional versus neutral prosody in individuals with ASD. When processing social versus basic emotions, individuals with ASD recruited the amygdala and STS to a lesser extent compared to controls and the functional connectivity between these regions was significantly reduced in the ASD group. Activity of core prosody regions such as the bilateral STS and IFG predicted accurate emotion recognition of social emotions on the behavioral task to a greater extent in controls than in individuals with ASD.
Conclusions:
In sum, these results provide evidence for dysfunctional emotional prosody processing in individuals with ASD on the behavioral and neural level and highlight the crucial role of the relationship between behavior and brain function and for unimpaired social functioning.