Does Gender Moderate Core Deficits in ASD? an Investigation into Social-Communication and Play

Background:  Due to the male dominance in ASD (Brugha et al., 2011; Kim et al., 2011), girls are rarely studied independently. Therefore our understanding of the behavioral phenotype of females with ASD is still limited. It is has been suggested that girls with ASD require a greater symptom “threshold” to reach diagnosis possibly leading to later detection (Dworzynski et al., 2012); however an interaction with intellectual disability potentially makes girls more likely to fall at the severe end of the spectrum (Carter et al., 2007; Fombonne, 2003). In terms of phenotypic profiles, research surrounding girls with ASD girls is extremely varied with some studies suggestive of superior social abilities in early childhood (Kopp and Gillberg, 1992) and others indicative of matched abilities relative to boys (Andersson et al., 2013). 

Core deficits in social-communication and play behaviors are extensively reported in children with ASD and frequently targeted through intervention (e.g., Kasari et al., 2006). In typical development, girls are known to display early advantages in social-communication and play behaviors, however we do not know whether girls and boys with ASD differ on the specific behaviors of initiating joint attention (IJA), initiating behavioral requesting (IBR) or play in early childhood. 

Objectives:  We were interested in whether girls demonstrated superior social-communication skills in early childhood compared to boys. We also examined whether gender moderated the effect of developmental variables on social-communication and play diversity. 

Methods:   40 girls with ASD (mean CA: 40m) were individually matched to 40 boys based on ADOS module and severity score. The two groups were not purposefully matched on MSEL DQ or CA, but did not differ on these variables. Their play complexity was also matched. All children completed measures of early social-communication (ESCS: Mundy et al, 2003) and play (SPA: Ungerer & Sigman, 1981). 

Results:  Girls and boys did not differ in the number of novel acts produced (t (78)= -1.19; p = .24). While girls did initiate joint attention and behavioral requests more than boys, these differences were not significant.

We ran separate ANCOVA models with gender as a predictor while co-varying for various developmental variables. Gender produced significant main effects on IBR when developmental variables were controlled (CA, expressive and receptive language) but not NVIQ. A significant interaction was found between gender and CA on IBR (F (3,1) = 4.70, p = .03). These effects were not found for IJA or play. 

Conclusions:  Overall girls and boys in our sample were more similar than dissimilar supporting recent findings (Andersson et al., 2013). Gender interacted with both chronological age and language abilities for IBR acts. This interaction was moderated by the boys; girls did not improve in their IBR with age or development. Interactions with developmental variables were not found for IJA or play, suggestive of a unique gender role in requesting. Whilst our data is cross-sectional, it is to our knowledge the largest sample of girls studied behaviorally to date and raises interesting questions as to whether one should approach requesting within intervention differently for girls.