Sex-Differences in 17-30 Months Old Toddlers: An Analysis of the M-CHAT

Background: Research suggests that in general population males tend to be more prone to developmental delay (Boyle et al 2011) and have a higher activity level (Riddoch et al., 2007), while females have better social orientation (MacCoby et al., 1979), gaze responses (Conellan et al., 2003, Lutchmaya et al., 2002), as well as strengths in imitation (Hittelmann & Dickens 1979).  Studies in ASD population also reveal marked sex differences in behavioral presentation. Females are four times less likely to be diagnosed with ASD than males (Baio, 2012) and, when diagnosed are more likely to exhibit co-occurring anxiety and other internalizing symptoms than males with ASD (Hartley & Sikora, 2009; Solomon et al., 2012). They also exhibit higher symptom severity than males (Dworozynski et al., 2012). These differences are apparent already in preschoolers with ASD (Hartley & Sikora, 2009; Zwaigenbaum et al., 2012).

Objectives: To evaluate sex differences in parent-reported autism-related behaviors in a large epidemiologic Norwegian sample of 17 to 30-months-old toddlers with and without ASD.  

Methods: Data were drawn from the Norwegian Mother and Child Cohort Study (MoBa) and its sub-study the Autism Birth Cohort (ABC). The total sample comprised of 53,728 children, including 185 children later diagnosed with ASD. There were 153 (83%) of males and 32 (17%) of females in the ASD sample. Diagnoses were obtained from the ABC clinic or from the National Patient Register. Questionnaires were completed between 17 and 30 months. We conducted ANOVA to compare the proportion of failures on the M-CHAT between males and females in the non-ASD and in the ASD samples. Chi-square risk analyses were conducted to compare which items were most frequently failed by males and females in the non-ASD stratified samples matched on mean to ASD males and females. Logistic regression was used to analyze which items most frequently failed by males and females in the ASD sample, controlling for sex and total number of failed items.

Results: Non-ASD males (M=.85, SD=1.22) failed more (P<.001) than females (M=.74, SD=1.11). ASD males (M=2.68, SD=3.54) failed less (P<.001) than females (M=5.16, SD=5.34) on mean number of failed items. Risk analyses showed that non-ASD sample females presents strengths on imitation and pretend play. Males showed strengths on enjoy climbing on things and functional play with objects. Logistic regression revealed that ASD females were more likely to fail the imitation item, but less likely to fail the follow to point item compared to ASD males.

Conclusions: In this epidemiologically ascertained Norwegian sample, mothers reported a higher degree of symptom severity in female than male toddlers later diagnosed with ASD.  Females in the non-ASD samples show strengths related to imitation compared to males, a strength that ASD females have lost compared to ASD males. They do however preserve the strength of response to joint attention over males.  The study suggests that females later diagnosed, already during the very early stages, may show strengths in joint attention compared to ASD males, but show selective impairments in imitation.