Higher Tonic Alerting and Lower Reactivity to Exogenous Stimuli Predict Social Cognition in Autism: A Pupillometry Marker of Sensory Processing

Background: Aberrant attentional function in autism spectrum disorder (ASD) was proposed to explain impaired social cognition, although empirical evidence is sparse. Previous findings indicated elevated tonic alerting and attenuated orienting as aberrant attentional functions in ASD. Deviant activity of the locus coeruleus – norepinephrine (LC-NE) system was assumed as underlying mechanism, which can be captured by pupillometry. Here, we investigated the effect of attentional function represented in pupil dilation (PD) during sensory processing of social stimuli on social cognition performance.

Objectives: (1) We hypothesize a distinct PD time course in ASD characterized by the metrics of elevated tonic dilation and delayed pupillary reactivity. (2) We assume principal components within the PD time course to represent attentional functions of alerting and orienting, which correlate with the time course metrics of the PD time course. (3) These principal components are hypothesized to predict social cognition performance, with a negative effect of elevated early alerting and a positive effect of intermediate orienting moderated by ASD group affiliation.

Methods: 23 adolescents with ASD and 23 neurotypical controls (TD) matched for age, gender, and verbal IQ were included. Pupillometry was carried out while participants worked on a validated social cognition test (MASC) consisting of dynamic social stimuli. Video stimuli are each followed by a related social cognition question. Between-group variation of PD time course was estimated by linear mixed models (LMMs) controlling for random effects of participant. PD time course was segmented to attentional-function components by principal component analysis. Factors indicating differential attention functions were correlated with time course metrics derived from previous studies and applied to predict social cognition performance.

Results: A distinct pupil-dilation time course in autism was indicated by the best-fit LMM with a significant interaction of group x time course on PD. The distinct time course was characterized by increased relative tonic dilation and elevated dilation latency compared to TD (d = .50 - .54). Three attentional-function components were extracted from the PD time course. An early alerting/pupillary-light-reflex (PLR) component positively correlated with relative tonic dilation, and an intermediate orienting component negatively correlated with dilation latency. We also extracted a late attentional control component. The three attentional-function components and group affiliation substantially predicted social cognition performance with F(15, 30) = 2.32, p = .02, R² adj. = .31. Significant interactions indicated an enhanced negative effect of early alerting/PLR (ASD: β = -.36, TD: β = -.27, partial η² = .17), and a positive effect of intermediate orienting specific to the ASD group (ASD: β = .45, TD: β = -.04, partial η² = .16).

Conclusions: A distinct pupil-dilation time course in ASD predicted impaired social cognition performance. Higher relative tonic dilation suggested increased tonic LC-NE activity as underlying mechanism of elevated tonic alerting. Elevated tonic LC-NE activity may attenuate reactivity to exogenous stimuli as represented by elevated dilation latency in ASD. This likely impairs sensory processing of dynamic and complex environments as indicated by lower social cognition performance. Pupil-dilation time course is outlined as promising biomarker for future ASD research.