Visual Exploration As a Measure of Social Motivation in ASD

Background: We should use precision when discussing social impairment, a cornerstone symptom domain of ASD. Currently, social affect generally is measured in terms of coarse global ratings based on distinct behaviors and summary impressions. Latent to this aggregate measurement, we hypothesize that there are independent though related dimensions, such as social -cognition, -interest, and -anxiety, that may interact across development to create various profiles of social ability or impairment. To evaluate the utility of isolating these social dimensions, we must establish psychometric or psychophysiological methods of measuring them.

Objectives:  To explore a visual exploration eye-tracking (ET) task as a potential measure of social interest by comparing ET data across ASD and typically developing (TYP) cohorts, and assessing its association with psychometric measures of social anxiety, social cognition, and experienced pleasure in social and nonsocial events.

Methods: Our sample included participants with high functioning ASD (HFA; n=71), as well as gender-, age-, and IQ-matched TYP (n=64) controls (HFA: M_age=12.2 years, SD_age=3.0 years; M_IQ=101.7, SD_IQ=17.4; TYP: M_age=12.7 years, SD_age=3.0 years; M_IQ=112.2, SD_IQ=14.9). The protocol included a passive viewing ET paradigm that probes visual salience (Sasson et al., 2008). This “Visual Exploration Task” consists of complex arrays that contain social (e.g., pictures of people) and nonsocial stimuli (e.g., pictures of objects). Fixation and exploration patterns were compared across the diagnostic groups and also were assessed for association with several socially-relevant psychometric measures, including subscales of the Social Responsiveness Scale (SRS; Constantino et al., 2003), Autism Diagnostic Observation Schedule (ADOS; Lord et al., 2000), Multidimensional Anxiety Scale for Children (MASC; March, 1997), Anxiety Depression and Mood Scale (ADAMS; Esbensen et al., 2003), and ratings of anticipatory and consummatory pleasure in social and nonsocial events respectively (Temporal Experience of Pleasure Scale; TEPS, Gard et al., 2006).

Results:  In preliminary analyses of the first 44 HFA and 30 TYP with coded ET data, the ASD group showed reduced exploration (number of images fixated) of social images (HFA M=14.0(4.6), TYP M=19.2(5.0), p<.05) coupled with increased duration of fixation on nonsocial images (HFA M=479 ms (227), TYP M=423 ms (174), p<.05).  Composite severity of social deficits was modestly positively correlated with exploration of nonsocial images within the ASD group (r=0.37, p=.04).  HFA demonstrated significantly greater deficits than TYP on the SRS Social Motivation, ADAMS Social Avoidance, MASC Social Anxiety, TEPS Anticipatory and Consummatory subscales (F’s[1,73] ranging from 17.3-156.1; all p’s<.001). ET data on the rest of the sample has been collected and is currently being processed, such that within- and across-group associations linking ET and psychometric data will be presented for the entire sample.

Conclusions:   Valid measures of social interest in ASD (i.e., those minimally confounded by social anxiety and the social skill deficits that mark this population) will allow us to explore variability of social interest in ASD by subgroups and across development. This also may inform the development of individualized treatment: for some individuals, building social motivation and/or targeting social anxiety may boost effects of targeted interventions on social cognition.