Gender Differences in Autism Spectrum Disorder Characterization within a State-Wide Community-Based Sample

Background: Since first described by Kanner, autism spectrum disorder (ASD) has been conceptualized as a predominately male condition. However, the relationship between ASD and gender is more nuanced than previously assumed. Males are more likely to receive an ASD diagnosis than females with equal symptomatology, and females are diagnosed later in life than males (Mandy et al., 2011). This female under-identification may result from gender-specific behavioral profiles in ASD, and a lack of understanding of the female behavioral profile (Dworzynski et al., 2012).

Objectives: To evaluate potential differences in behavioral characteristics and diagnostic outcomes across gender in a community-based sample.

Methods: This study analyzed data from males (n = 1128) and females (n = 335) ages 2-69 years enrolled in the Rhode Island Consortium of Autism Research and Treatment (RI-CART), a state-wide community-based sample. Enrollees had or suspected an ASD diagnosis, came from varied racial and socioeconomic backgrounds, and lived in southern New England. To join RI-CART, enrollees completed behavioral and cognitive measures including the ADOS-2, VABS-II, and KBIT-2.

Results: Female RI-CART enrollees had significantly higher KBIT-2 scores F(1, 274) = 4.552 p = .034, η² = .016, VABS-II overall scores F(1, 1020) = 6.297, p = .012, η² = .006 and social skills subdomain scores F(1, 1028) = 8.003, p = .005, η² = .008. VABS-II daily living skills domain scores and SRS-2 overall scores did not significantly differ by gender. Amongst participants with a community diagnosis, females had a significantly higher age of diagnosis F(1, 560) = 7.47, p = .006, η² = .013. RI-CART enrollees of both genders with a community diagnosis of ASD had higher ADOS comparison scores than enrollees without a community diagnosis F(1, 1024) = 196.97 p < .001, η² = .005. The interaction between community diagnosis and gender was significant. Male enrollees without a community diagnosis had significantly higher ADOS comparison scores than their female counterparts—in contrast, the ADOS comparison scores of enrollees with a community diagnosis did not significantly differ across gender F(1, 1204) = 5.479 p = .019, η² = .019.

Conclusions: Female RI-CART enrollees had significantly higher age of diagnosis than male enrollees, as well as higher KBIT-2, VABS-II social skills subdomain, and overall scores. These findings support the idea of a female-specific ASD behavioral profile. This profile may also contribute to the significantly lower ADOS-2 comparison scores in females questioning an ASD diagnosis. Better verbal, social, and adaptive skills could cause parents and clinicians to under-identify females with ASD. The female profile seen in this study does not mirror the profile identified by some prior studies; this may be due to heterogeneity of the RI-CART sample (Fraizer et al., 2014). Future research should clarify the characteristics of the female behavioral profile, as well as the breadth and magnitude of the gender discrepancy in ASD diagnoses.