Age of First Concerns, Early Milestones and Eventual ASD Diagnosis: Are There Sex Differences?

Poster Presentation
Thursday, May 2, 2019: 11:30 AM-1:30 PM
Room: 710 (Palais des congres de Montreal)
C. Harrop1, E. J. Libsack2, R. Bernier3, M. Dapretto4, A. Jack5, J. McPartland6, J. Van Horn7, S. J. Webb3 and K. A. Pelphrey8, (1)University of North Carolina at Chapel Hill, Chapel Hill, NC, (2)Stony Brook University, Stony Brook, NY, (3)Psychiatry and Behavioral Sciences, University of Washington, Seattle, WA, (4)Dept of Psychiatry and Biobehavioral Sciences, University of California, Los Angeles, Los Angeles, CA, (5)The George Washington University, Washington, DC, (6)Child Study Center, Yale University School of Medicine, New Haven, CT, (7)University of Southern California, Los Angeles, CA, (8)University of Virginia, Charlottesville, VA
Background: Despite advances in early detection, the average age of ASD diagnosis often exceeds 4 years, with a significant lag following first concerns (Daniels & Mandell, 2014). Previous research suggests that due to the sex imbalance and resultant male conceptualization of ASD, girls are particularly likely to experience diagnostic delays (Salomone et al., 2015).

In typical development, child sex predicts inter-individual variation across multiple developmental milestones, with females often exhibiting earlier progress. Parental perceptions of what is normative for males and females may engender differential expectations and influence timing of initial concerns in ASD. Early developmental milestones are predictive of later diagnosis and clinical outcomes in ASD (Bedford et al., 2016; Kower et al., 2016); however, the role of sex on early milestones has not been studied in-depth.

Objectives: Explore sex differences in (1) developmental milestones and (2) their contribution to timing of initial concerns and diagnosis.

Methods: 195 participants (105 males) aged 8-17 years with ADOS-2 and ADI-R confirmed ASD diagnosis were recruited across four sites. Early developmental milestones, age of first concerns and age of ASD diagnosis were obtained from the ACE Medical History Form. Univariate ANOVAs tested for sex differences in timing of early milestones, first concerns, ASD diagnosis, and diagnostic lag (time between initial concerns and eventual ASD diagnosis), after controlling for demographic variables (child IQ, maternal education, race, ethnicity, and site). Stepwise linear regressions were used to investigate the relative contributions of demographic variables, sex, IQ, and early developmental milestones to age of first concerns, diagnosis and diagnosis lag.


Between-sex differences: There were trends toward earlier first concerns and diagnosis in males (p=.08). There were no sex differences in diagnostic lag, nor age of first walking. Parents reported earlier first words and phrases in females (p = .04 and .06).

Factors predicting age of first concerns, diagnosis, and lag (Table 1): Demographic variables alone did not significantly predict age of first concerns (R2=.04), age of diagnosis (R2=.03), nor diagnostic lag (R2=.01). All three models reached significance with inclusion of current IQ: first concerns (R2=.08), diagnosis (R2=.11), lag (R2=.05).The inclusion of sex in the model predicting age of first concerns was marginally significant (R2=.98, p=.07), but was non-significant for age of diagnosis, nor lag. There were no sex by interactions. The amount of variance accounted for by inclusion of early milestones was not significant for any models: first concerns (R2=.12), diagnosis (R2=.14) lag (R2=.07). Across models, IQ remained the strongest individual predictor.

Conclusions: Results show sex differences in early developmental milestones related to language in an ASD sample, reflecting what is observed in typical development and potentially suggestive of a female bias in parent’s perceptions of development. Early milestones did not predict the age of initial concerns, diagnosis nor lag. IQ was the most significant predictor, suggesting children with lower IQ, regardless of sex, are identified and diagnosed earlier. Overall, our variables did not account for a large proportion of variance, suggesting other factors are more influential in predicting timing of initial concerns and diagnosis.