Sex Differences in Cognitive and Reasoning Abilities Among Preschool and School-Age Autistic Children

Background: The estimated IQ of individuals on the autism spectrum (AS) differs depending on the intelligence test used. The most replicated finding is the discrepancy between Wechsler Intelligence Scales and Raven Progressive Matrices (RPM), in favor of the latter. Some authors also suggested that IQ in AS children differs between girls and boys, girls tending to have more intellectual impairment than boys (Volkmar et al. 1993). However, other authors reported no discrepancy between the IQ of girls versus boys with autism (Mandy et al. 2012). It is therefore unclear whether the Wechsler-RPM discrepancy characterizes the cognitive profile of both AS girls and boys.

Objectives: 1) To examine whether there are sex differences in performance on Wechsler and on RPM within an AS and a typically developing (TD) group. 2) To compare the discrepancy between Wechsler and RPM between sexes and groups.

Methods: 48 AS children (25 males, 23 females) aged from 4 to 12 years old (M=7.54) and 66 TD children (37 males, 29 females) aged from 4 to 12 years old (M=6.15, p=.13) were assessed using the colored version of the RPM and either the WPPSI-IV (n=18 AS and 29 TD) or the WISC-IV (n=32 AS and 37 TD) depending on their age. A “Wechsler IQ” variable was computed by taking the score of either the WPPSI-IV or the WISC-IV depending on which test the child completed. A discrepancy variable was computed by calculating the difference between Wechsler and RPM in percentiles.

Results:

Regarding results on the Wechsler, an ANOVA showed a main effect of group (p<.001), AS children having a significantly lower performance (M=16.66, SD=25.37) than TD children (M=60.80, SD=30.87). There was no main effect of sex (p=.14) and no interaction effect (p=.15). On the RPM, the ANOVA showed no main effect of group (p=.13) or sex (p=.29). There was a significant interaction (p<.05), autistic boys (M=82.00, SD=27.00) performing better than autistic girls (M=66.42, SD=32.37; p<.05), while no difference was found between TD girls and boys.

Furthermore, for the discrepancy variable, there was a main effect of group (p<.001) and sex (p<.01), with no interaction effect (p=.52). The discrepancy between Wechsler and RPM was greater in the AS group (M= 57.42, SD=3.58) than in the TD group (M=19.54, SD=3.07), and that discrepancy was greater in boys (M=44.88, SD=3.21) than in girls (M=32.08, SD=3.46) from both groups.

Conclusions: The present results first replicated Mandy et al. (2012) findings that boys and girls on the AS do not differ in terms of Wechsler IQ in a sample of preschool and school-aged AS children. In addition to the existing literature, the results show that the difference between boys and girls with autism might be different depending on the test used. Finally, the investigation of sex differences in the RPM-Wechsler discrepancy suggests that this discrepancy characterizes the cognitive profile of AS children of both sexes, but might be smaller among girls (both TD and AS).