Age and Cognitive Functioning Moderate Sex Differences in School-Age Children with Autism Spectrum Disorder

Background: Clarifying how the clinical phenotype of Autism Spectrum Disorder (ASD) differs across sexes is necessary to fully evaluate hypotheses proposed to explain the disproportionate number of males relative to females with ASD. Studies directly examining the clinical phenotype have suggested subtle differences. Females with ASD have, on average, lower cognitive abilities. Controlling for cognitive functioning though, many studies report that males and females with ASD have largely similar levels of autistic symptomatology across development, from early childhood through adulthood. Similarly, preliminary eye-tracking research in our lab has suggested that males and females with ASD attend to others’ eyes for similar amounts of time. However, the social adaptive value of attention to the eyes differs based on sex, suggesting differences in developmental mechanisms of social engagement. To help clarify the effect of sex on the clinical phenotype of ASD, the current study sought to identify factors influencing the magnitude of sex differences.

Objectives: The current study aims to (1) determine factors moderating the social adaptive value of attention to the eyes across sexes and (2) examine specific mechanisms of social engagement underlying differences across subgroups.

Methods: Eye-tracking data were collected while 161 children with ASD (114 male, 47 female) and 49 typically-developing peers (27 male, 22 female) viewed social scenes of children and adults engaged in naturalistic, age-appropriate social interaction within everyday settings. The ASD sample represented a broad range of ages (mean=10.1(2.8), range=5 to 17 years), cognitive functioning (Full-Scale IQ: mean=94.5(22.8), range=32 to 149), and level of social disability (ADOS Calibrated Severity Score: mean=7.1(2.3), range=1 to 10). Both across and within diagnostic groups, males and females were matched on chronological age. Within diagnostic groups, the sexes did not significantly differ in cognitive functioning or in level of social disability. Consistent with our preliminary findings, there were also no significant differences in attention to the eyes based on sex within the ASD group.

Results: Within the full ASD sample, sex moderated the relationship between attention to the eyes and level of social disability (B=15.95, p=0.04). However, age and cognitive profile differentially influenced the relationship between attention to the eyes and level of social disability in males and females with ASD. Further analysis indicated that sex was only a significant moderator for older children (age>11.5 years) and children with higher full-scale IQ (>94.0). Within these subgroups, males with ASD showed the pattern expected from past research, in which more attention to the eyes predicted less social disability (r=-0.401, p=0.01). In contrast, for females, more attention to the eyes trended toward association with greater social disability (r=0.57, p=0.10). Ongoing analyses using more time-sensitive measures of social visual engagement are examining how the timing of attention to the eyes may mediate differences across subgroups.

Conclusions: The current study finds that sex differences in ASD are magnified in older children with higher cognitive functioning. These results also indicate that differences in mechanisms of social engagement may be present even in the absence of manifest mean differences across males and females with ASD.