Recognition of emotions conveyed by point-light and full-light display in children with Autism Spectrum Disorder (ASD)

The perception and interpretation of other people’s feeling is essential for successful social interaction. As specified by DSM-V, Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder characterized by “Persistent deficits in social communication and social interaction across multiple contexts”, “Deficits in social-emotional reciprocity” and “Deficits in nonverbal communicative behaviors used for social interaction” (American Psychiatric Association, 2013). This impairment in social interaction has been reported very early in development (Klin et al., 2008; 2009) and it has been proposed it could be associated with deficits in processing biological motion information (Dakin & Frith, 2005, Pavlova 2005). In typically developed (TD) individuals, point-light display of biological motion (PLDs, Johansson 1973) are sufficient for the perception of a series of social information, included the emotional state of the agent observed (Alaerts et al., 2011; Atkinson et al., 2004; 2011; 2007; Chouchourelou et al., 2006; Clarke et al, 2005, Dittrich et al., 1996). On the contrary, in literature there are less consistent findings regarding individuals with ASD: some studies have reported impairment in the recognition of non-emotional PLDs (Annaz at al., 2010; Blake et al., 2003, Kaiser et al., 2008; 2010; Koldewyn et al., 2010) whereas others have not (Herrington at al., 2007, Freitag et al., 2008, Murphy et al., 2009; Saygin et al., 2010); however, deficits in emotion understanding have been systematically reported in young children (Moore et al., 1997), adolescents (Hubert et al., 2007; Parron et al., 2008) and adults with ASD (Atkinson et al., 2009; Nackaerts et al., 2012; Philip et al., 2010). Therefore, it is not clear yet if the difficulties in social interaction in individual with ASD are related to a general impairment in the processing of biological motion, or to a selective deficit in emotion recognition or to both of these difficulties. In the present study, we explore how a group of typical developing (TD) children and a group of ASD children recognize a particular aspect of biological motion: the emotional information conveyed by whole body movements. We used different types of stimuli (Atkinson et al., 2004, 2007) which allow us to explore the contribution of body form and body motion information in the recognition and understanding of neutral actions and actions which convey emotions. Method A group of 10 ASD children (mean=9.7 years old) and a group of 23 typical developing children (TD) (mean=9.05 years old) age took part in this experiment. ASD group was composed by children with different level of autism functioning, diagnosis were made by expert psychologists following DSM-IV criteria and confirmed by ADOS (ADOS-Score mean: 5.5). IQ level was assessed for all participants (ASD mean: 69 ; TD mean : 111) Participants were asked to recognize, in a forced-choice paradigm, the emotional content of actions portrayed by an actor in both Point-Light (PLD) and Full-Light (FLD) short movie clips (Atkinson et al., 2004; 2007; 2012). Participant responded by pressing the designated key on a keyboard, selecting one of three options (Happy, Neutral, Fear). Accuracy and response times (RTs) were collected. Results Accuracy rate and response times (RTs) were analyzed separately. The relation between IQ level and performance was also explored. Accuracy: Between group Wilcoxon Signed Rank Test shows a significant difference between ASD and TD: ASD participants recognize less accurately than TD each of the three emotional content (Fear: p= 0.004; Happy: p= 0.04, Neutral: p<.001), in both display type (FLDs: p<0.002; PLDs: p<0.001). Within group Wilcoxon Signed Rank Test was performed separately for the two group, to compare accuracy rate in the three different emotion condition. For TD group, results show a significant different accuracy between Happy and Neutral (p= 0.0006), while no difference was found between Fear and Happy (W = 4871.5, p-value = 0.05999) or Fear and Neutral (W = 3802.5, p-value = 0.9075). In contrast, in ASD group no significant difference on accuracy rate between the three emotions was found (all p>.05). We also explored the possible effects of IQ on children accuracy. Analysis of Regression shows a significant effect of IQ level on the accuracy rate (p<0.001) both for ASD (p=0.003) and for TD (p= 0.024). Analysis of correlation shows a positive relation between IQ level and accuracy rate, both for TD (r=0.1494) and for ASD group (r=0.2910), suggesting that children with a high IQ were better at recognizing the emotional content of body movement. RT: Analysis on accuracy suggested that ASD recognized worse than TD the emotional content conveyed by biological movements, but how do ASD perform when they recognize the emotional content of the observed actions? To explore this issue, we analysed only response time (RTs) relative to correct responses, separately for each group, by a 2X3 Repeated measures ANOVA with Display and Emotions as within factors. In both group, results show no effects of Display type (TD: p=0.932; ASD: p=0.418) or Emotion (TD: p=0.818; ASD: p=0.401) on RTs and the interaction was no significant (TD: p=0.966; ASD: p= 0.966). Wilcoxon Signed Rank Test between group analysis shows that ASD were slower than the TD group in recognizing Fear (W = 2123, p-value = 0.0056) and Neutral (W = 1833, p-value = 0.0781) while no significant differences were found for Happy (W = 1735, p-value = 0.3487) movements. The possible relation between IQ and RTs were also explored: analysis of Regression with IQ and RTs as factors, show a significant relation between IQ and RTs in both groups (ASD, r= 0.00152; TD, r= 0.00347). Correlation analysis shows a negative relation between IQ and RT in both ASD (r= 0.00152) and TD (r= -0.1928055) group, suggesting that the higher is the children IQ level, the faster they are in recognizing body movement. Conclusion The present study shows that ASD children are impaired in emotion recognition conveyed by biological motion: they were less accurate and slower than TD children. These results extend previous findings, suggesting that impaired recognition of bodily expressed emotions in ASD could be at least partly attributable to a deficit in motion processing. The absence of an effect of Display type on emotion recognition in both TD and ASD suggest that the merely movement (PLDs) could be sufficient to convey emotional content of action, while other visual information (FLDs) seem not to help the emotion to be more recognizable