Socio-Dramatic Affective-Relational Intervention (SDARI) Changes the N100 ERP to Prosodic Voices

Background: Several group social skills interventions have been developed to target social communication skills, a core deficit in autism spectrum disorder (ASD; McMahon et al., 2013). One such intervention, Socio-Dramatic Affective-Relational Intervention (SDARI), uses naturalistic performance-based techniques to develop social skills (Lerner et al., 2011). While participation in SDARI is related to social gains on subjective measures and behavioral (including vocal) emotion identification tasks, it is difficult to objectively quantify social communication improvements. Event-related potentials (ERPs) recorded via electroencephalogram (EEG) can be used to index the neural correlates of social processing and thus may be a useful tool for quantifying social improvement. The N100 and N300 are two ERP components related to processing vocal prosody that are aberrant in amplitude in individuals with social deficits (Pinheiro et al., 2011). While the N100 reflects early perceptual coding of auditory stimuli and is modulated by attention, the N300 is related to cognitive evaluation of emotion (Paulmann et al., 2008). Auditory ERPs to emotional prosody may be a useful objective technique for quantifying treatment outcomes in ASD.

Objectives: Examine whether participation in SDARI versus an active control is related to changes in N100 and N300 ERPs.

Methods: Thirty-three children (25 male; M_age=12.77, SD_age=3.01) with ADOS-2 confirmed ASD diagnosis and IQ ≥ 70 (M_IQ=107.94, SD_IQ=15.20) were randomly assigned to participate in 10 weeks of SDARI or a recreational games control group. Pre- and post-intervention, participants completed a measure of vocal emotion recognition (DANVA-2; Nowicki, 2004) while EEG was recorded. N100 and N300 to auditory stimuli of neutral sentences with emotional prosody were extracted from EEG data. We used Generalized Estimating Equations (GEE) to examine change in ERP amplitude and intensity as a function of group membership. We modeled the ERP (N100 or N300) at post-test as the dependent variable, group membership (SDARI vs. CONTROL) as the nested between-subjects variable, and ERP at pre-test as a covariate.

Results: Participation in SDARI resulted in significantly larger (more negative) N100 amplitude at post test (Wald’s c2=4.43, B=2.78, p<.05). To probe this effect we ran two additional GEE models examining N100 amplitude to low and high intensity emotional voices. No significant effects were found for low intensity voices but participation in SDARI was associated with increases in N100 amplitude to high intensity emotional prosody (Wald’s c2=6.75, B=2.25, p<.01). GEE models of change in N100 latency, N300 amplitude, and N300 latency were non-significant (all p>.05).

Conclusions: Results indicate that participation in SDARI, a naturalistic social skills intervention, is associated with increased magnitude of N100 response to highly salient prosodic language in youth with ASD. This suggests that engagement in this performance-based intervention leads to greater attention to and engagement in perceptual coding of emotional voices in comparison to an attention-control intervention. Thus, performance-based interventions such as SDARI may improve sensitivity to salient social cues like prosody. Not only is the N100 malleable in response to intervention, but future intervention studies might consider implementing the N100 as an objective index of social gains resultant from participation in intervention tasks.