Adolescence is a critical period for socialization, particularly because social expectations increase with age. In addition, the mounting importance of peer interactions and the growing complexity of social relationships appear to be crucial to adolescents outcome (e.g., Zarrett & Eccles, 2006). Indeed, it was observed in the general population that social difficulties can contribute to lower self-esteem and academic difficulties later in life, with social withdrawal being a risk factor for the later emergence of social anxiety (Greco & Morris, 2001). To navigate peacefully in the social sphere and effectively decode the social environment, subtle abilities are required. On this aspect, individuals with neurodevelopmental disorders are penalized. Indeed, social impairments are reported in many of these disorders, including autism spectrum disorders (ASD) and 22q11.2 deletion syndrome (22q11DS). Both individuals with 22q11DS (e.g., Milic et al., 2021; Norkett et al., 2017a) and with ASD (e.g., Fakhoury, 2015) are characterized by social impairments, including difficulties perceiving and interpreting social signals (i.e. social skills; (Gillis & Butler, 2007)), that may interfere with ability to create, maintain and end social interactions (Uljarević et al., 2020). According to recent models (e.g., Beauchamp & Anderson, 2010), social skills also partially depend on executive, cognitive and verbal competences, all of which are frequently impaired in neurodevelopmental disorders (Bausela-Herreras et al., 2019; De Smedt et al., 2007).
In ASD, alterations in social communication and social interactions are observed, in addition to repetitive and restrictive behaviors and interests. Several social skills are impaired from very early on, notably social smiling, looking at faces, responding to one’s name and making eye contact (Boyd et al., 2010). Moreover, individuals with ASD exhibit weaker adaptive behaviors – defined as the skills required to function and be independent in everyday environments (Sparrow et al., 2005) – in the social domain. Specifically, individuals with ASD have lower skills in the area of socialization, resulting in difficulties making friends, acting in an appropriate way with peers, etc. (Kanne et al., 2011; Yang et al., 2016). In 22q11DS, a neurogenetic condition affecting approximately 1 in 2000–4000 births (Olsen et al., 2018), social impairments are typically observed, in additional to the typical physical (e.g., chronic infections, cleft palate, heart defects, hypocalcemia), cognitive (e.g., IQ around 70, executive functions’ deficits) and psychiatric (e.g., psychosis, attention deficit, anxiety and mood disorders) characteristics of the syndrome (McDonald-McGinn et al., 2017; Schneider et al., 2014). Studies have highlighted that individuals with 22q11DS exhibit poorer social functioning compared to typically developing peers (TD). For example, poorer social skills (Kiley-Brabeck & Sobin, 2006) and more problematic social behaviors (Shashi et al., 2012) have been reported in children with 22q11DS compared to their siblings. Moreover, social immaturity and difficulties in initiating social interactions (Van Den Heuvel et al., 2018) as well as socio-communicative impairments reported by parents (Van Den Heuvel et al., 2017) have been highlighted. Emotion processing has also been found to be impaired, with difficulties in emotion recognition and an abnormal visual exploration of faces (e.g., Campbell et al., 2015; McCabe et al., 2013). Finally, individuals with 22q11DS exhibit weaker adaptive skills (Schneider et al., 2014) and are described as socially withdrawn and more socially inhibited and isolated than TD (e.g., Schonherz et al., 2014). Of note, social skills interventions remain scarce in 22q11DS, but some studies have reported positive outcomes (e.g., Glaser et al., 2018; Shashi et al., 2015).
Social skills thus appear to be of critical importance to better characterize the social profiles of each condition, as differences in the social phenotype of 22q11DS and idiopathic ASD have been highlighted. Indeed, differences were observed in socio-emotional reciprocity, idiosyncratic speech and non-verbal interactions (Kates et al., 2007; McCabe et al., 2013; Pontillo et al., 2018). Moreover, higher levels of empathy, sense of humor and other complex social skills were found in individuals with 22q11DS compared to youth with idiopathic ASD (Angkustsiri et al., 2014). In contrast, some authors have suggested that a significant proportion of 22q11DS meet criteria for ASD (e.g., Vorstman et al., 2006) and reported similarities in emotion recognition, conversations initiation and maintenance, and adaptive socialization (Kates et al., 2007; McCabe et al., 2013; Pontillo et al., 2018). Moreover, social anxiety being a frequent comorbidity of both ASD and 22q11DS (Maddox & White, 2015; Schneider et al., 2014), the association between these two constructs needs to be further explored as they were found to be related (Pickard et al., 2017; White et al., 2013). Finally, a better characterization of the social skills profiles of each clinical population is needed to tailor specific social skills training interventions with distinct therapeutic targets, as social skills training programs have shown compelling results (e.g., Laugeson et al., 2012; Moody & Laugeson, 2020; Trudel & Nadig, 2019; Tse et al., 2007).
However, one of the main limitations of the majority of existing studies in the field is that social skills are assessed through questionnaires, which represents the most classical method to examine this construct (Matson et al., 2007). Nevertheless, most of the existing questionnaires have been developed for a typical population. As a result, they are often complex and lengthy, but also dependent on awareness of difficulties that is frequently absent in clinical populations (Norton et al., 2010). Moreover, some questionnaires are completed by caregivers, introducing a hetero-reported perspective that does not necessarily correspond to direct observation (Bellack et al., 2006). Indeed, direct observation is assumed to provide a more valid estimate than more distal measures such as questionnaires (Harvey et al., 2007). To assess social skills in the most ecological way, it is necessary to approximate a person’s behavior in different daily-life situations, which can be done using role-plays. These systematic and direct observations of social behaviors are considered the gold standard of social skills assessment, as they provide access to real-world behaviors through simulated situational settings (Morrison et al., 2017). This method has been successfully used in various clinical population (e.g., Bellack et al., 2006; Morrison et al., 2017; Verhoeven et al., 2013), including youth with ASD (Paul, 2003; Webb et al., 2004). Ratto et al. (2011) even developed a role play measure specifically designed to measure social skills in individuals with ASD and found it to be successful in discriminating between the groups. However, it should be noted that role-playing games, although approximating reality, is performed in a laboratory setting that remains somewhat artificial (Bellack et al., 1979).
1.1. Aims of the study
The goal of the present study was to investigate social skills using semi-standardized role-plays (i.e. the Social Skills performance Assessment (SSPA; Patterson et al., 2001) in two neurodevelopmental disorders often considered to share similar social profiles and to examine the correspondence with a more standard methodology (i.e. caregiver report). We also aimed to investigate the potential association between social skills and social anxiety. First, participants with ASD and 22q11DS were expected to report lower social skills compared to TD, as measured both by direct observation and reported by caregivers. Second, significant associations were expected to be observed between the two measures of social skills (direct observation and caregiver report). Third, different patterns of social skills in the three groups were expected. Fourth, compared to TD, higher levels of social anxiety were expected in both clinical groups, with higher social anxiety being associated with lower social skills (as measured by direct observation but also by caregiver report). Finally, and additionally to what was co-registered (10.17605/OSF.IO/QF6WN), the impact of gender, age and cognitive abilities on social skills were explored, as well as the association with social functioning and ASD symptom severity.