Sex Differences in Adaptive Skill Trajectories from 12 to 36 Months in Infants at High and Low Risk for ASD

Background:   Children with ASD appear to have a distinct profile of adaptive behaviors even when compared to children with non-ASD developmental delays.  ASD-specific trajectories of adaptive skills have been reported with cross-sectional differences evident as early as 12 months.  Here we investigate ASD-specific sex differences in trajectories of adaptive skill development.  This study seeks to determine whether parent report regarding adaptive behaviors may provide insight into the apparent discrepancy between the high male:female diagnostic ratio in ASD and null results with respect to ASD-specific sex differences on measures of ASD symptomatology (ADOS) and cognitive development (Mullen).

Objectives:  To evaluate ASD-specific sex differences in trajectories of adaptive skill development among high- and low-risk infants from 12-36 months of age.

Methods:  Parents completed the Vineland Adaptive Behavior Scales (VABS) interview or questionnaire when their infants/toddlers at high and low risk for ASD were 12, 18, 24, and 36 months of age.  Diagnostic classifications were made at 36 months.  Children with atypical non-ASD outcomes were excluded from these analyses.  42 children were diagnosed with ASD (gender ratio 5:1).  200 children were judged to be developing typically (gender ratio 1.2:1).  Hierarchical linear models evaluated trajectories of each VABS subscale using age equivalencies.  Predictors were sex, diagnostic outcome (ASD or TD), and sex*diagnosis.  A second set of models included baseline verbal mental age based on observed sex differences in that measure and its association with the development of adaptive skills.

Results:  Linear, but not quadratic, effects of time were observed for the Communication (Com), Daily Living (DLS), Socialization (Soc), and Motor (Mot) scales of the VABS (ps<.001).  Significant sex-by-diagnosis interactions were observed with respect to growth in Com (p=.007), DLS (p=.034), and Soc (p<.001), but not Mot (p=.85).  Parent ratings of adaptive skills in boys and girls with ASD were similar at 12 months and skills of girls with ASD improved more rapidly than skills of boys with ASD between 24-36 months.  Main effects of sex were observed with respect to COM, DLS, and SOC.  Main effects of ASD were observed with respect to DLS and SOC, but not COM.  Baseline language skills were associated with trajectories of Com, DLS, and Soc.  When language skills were included in models, a robust sex-by-diagnosis interaction was observed on Com and Soc, but not DLS.  

Conclusions:  Parent-reported adaptive behavior trajectories during the time frame when overt symptoms of ASD first emerge indicate that girls differ from boys within the group of toddlers who are diagnosed with ASD and between diagnostic groups.  Developmental trajectories reflect an improvement of adaptive behaviors among girls with ASD relative to boys with ASD across infant-toddler development.  Further investigation is necessary to confirm that infant siblings are representative of non-familial autism phenotypes.  Viewed longitudinally, these results are not consistent with the Carter Effect, which suggests that in conditions with uneven sex ratios, genetic loading should be significantly greater in the sex with lower prevalence.  Other factors that may contribute to parents' differential reporting about ASD-related behavioral trajectories between girls and boys will be discussed.