Logo image
Back
73. Increased Neural Encoding Variability of Dynamic Social Stimuli in Autism Spectrum Disorder
Abstract   Peer reviewed

73. Increased Neural Encoding Variability of Dynamic Social Stimuli in Autism Spectrum Disorder

Avery Van De Water, Lisa Byrge, Daniel Kennedy, Dorit Kliemann and James Traer
Biological psychiatry (1969), Vol.99(10 Supplement), pp.S130-S130
05/15/2026
DOI: 10.1016/j.biopsych.2026.03.307

View Online

Abstract

Background The neural basis of social difficulties in autism spectrum disorder (ASD) remains not well understood. Prior studies typically used static/unimodal stimuli or reduced complex stimuli to only a few dimensions, obscuring real-world complexity. Here, we used videos to engage brain networks and trained encoding models to predict parcelwise brain activity from high-dimensional audio, visual, and socially-relevant (e.g., valence) features. We hypothesized that brain activity in ASD would be less accurately predicted and more heterogeneous within and across participants than in NT participants, particularly for socially-relevant features. Methods Participants (48 ASD [13 female], 90 neurotypical [NT; 25 female]) viewed six videos (5-20 minutes) while brain activity was measured using functional MRI. Cross-validated linear ridge-regression encoding models assessed prediction accuracy variance, and pattern similarity. Results Prediction accuracies were strikingly consistent between groups and significantly above control models across brain networks and videos (p(FDR) <0.05). Auditory and language networks were most accurately predicted. Within-participant predictions were equally variable between groups. However, autistic participants exhibited lower within-group encoding-pattern similarity than NT participants (U=2054, delta-z=0.32, p(FDR) <0.001) and showed greater across-group than within-group similarity (W=129, delta-z=0.11, p(FDR) <0.001). Analyses using only socially-relevant features showed the largest across-group differences in encoding-pattern similarity. Conclusions The novel application of encoding models to multiple dynamic videos suggests greater heterogeneity in neural representations of social information in ASD. Although group-averaged encoding accuracies were comparable across groups, ASD participants had more heterogenous regional encoding patterns. These idiosyncratic patterns underscore the need to incorporate individual variation in brain representations when developing translational neuroimaging approaches in autism.

Details

Metrics

1 Record Views
Logo image