Does Preferential Looking at Social and Non-Social Aspects of Naturalistic Scenes Differ Between Individuals with ASD with and without Intellectual Disabilities?

Background:  The social motivation theory of autism spectrum disorders (Klin et al., 2002; Chevallier et al., 2012) suggests that core social-communicative deficits may stem from diminished social motivation. One indicator of social motivation deficits is reduced spontaneous attention to faces (notably eyes), people and social actions (Falck‑Ytter & von Hofsten, 2010). While abnormalities have been reported across the life span, most studies were conducted with ASD participants with IQ in the normal range. Although IQ is considered the best predictor of outcome (Howlin et al., 2004) little is known about the effect of IQ on social attention, or whether preferential looking patterns differ between individuals with ASD with and without intellectual disabilities.

Objectives:  The present study aimed to 1) examine group differences in looking patterns on social and non-social aspects of naturalistic social scenes, 2) to explore the role of IQ/intellectual disability (ID) in viewing patterns, and 3) the relationship between looking patterns and level of social impairments/adaptive functioning in individuals with high-functioning ASD and those with ASD+ID.

Methods:  231 individuals with ASD and 186 controls with typical development (TD) or mild intellectual disabilities (IQ 40-148), aged 12-30 years were recruited as part of the multi-center EU-AIMS Longitudinal European Autism Project. In all centers, gaze patterns during viewing of six static images were recorded via using Tobii eye-trackers. Areas of interest (AOIs) were defined as face, eyes, mouth, and objects that were relevant (i.e., part of a social interaction). Preliminary analyses focused on % gaze time. IQ was measured with the Wechsler scales, social-communicative impairments using the Social Responsiveness Scales (SRS-2) and social adaptive behaviour using the social-subdomain of the Vineland scales (VABSsoc). Individuals were divided into intellectual ability groups (i.e. above and below IQ 70).

Results:  We found a significant main effect of IQ group on % gaze time on faces (p < .001) as well as an IQ group x diagnosis interaction (p = .017). Whereas those with mild ID (without ASD) looked on average less at faces than the TD group (p < .001), among the ASD group, intellectual ability level did not affect looking times on faces. In line with previous findings, among individuals with IQ >70, the TD group showed looked longer at faces than the ASD group (p < .001; see Fig.1 for an example of viewing times by group during a highly social scene). In the ASD+ID group, higher % gaze time to eyes was significantly correlated with higher symptom severity (SRS-2 scores) (n = 46, r = .41, p = .007). Figure 2 shows the percentage for composite looking times (i.e. all 6 images) for each AOI by group.

Conclusions:  Individuals with ASD with/without ID showed similar rates of social attention to faces, while IQ affected viewing patterns in the ID group. This finding is inconsistent with the notion that social motivation may be less influenced by cognitive development in ASD. However, unexpectedly looking times to social features were not linked to level of social impairment/adaptive behaviour. We plan to further investigate fixation patterns (e.g., first looks, average fixation times) in static and dynamic social scenes.