The ADOS-2 and ADI-R: Are There Sex Differences?

Background:  

The research on sex differences in the behavioral phenotype in ASD has found that females demonstrate fewer repetitive behaviors but more communication deficits and greater anxiety and depressive symptoms than males (Lord et al., 1982; Hartley & Sikora, 2009), with fewer differences in early development (Wijngaarden-Cremers et al., 2013; Reinhardt et al., 2014). While DSM-5 ASD criteria has adequate sensitivity in females with ASD, ranging from .88 to .93 (Huerta et al., 2012), further research is needed to determine whether there are sex differences in the diagnostic instruments used in ASD evaluations.

Objectives:  

The goals of this study are to examine the utility of the ADOS-2 and ADI-R in male and female children and adults and to generate recommendations for diagnosis and classification.

Methods:  

Participants were drawn from clinic and research datasets from the Center for Autism and the Developing Brain (CADB) and include children and adults with diagnoses of ASD and non-ASD diagnoses. Only one data point, the earliest full assessment, was used per participant. The resulting sample included 620 female and 2065 male children and adults.  Sensitivity and specificity analyses of the ADOS-2 and ADI-R were conducted along with item-level analyses. For the purposes of taking into account the developmental and item-level differences in the instruments, the sample was stratified by age and language level.

Results:  

The sensitivity and specificity of the new ADOS-2 algorithms was comparable in the male and female sample. In the female participants, sensitivity ranged from .86 (Module 3) to .98 (Module 1), while for the male participants, sensitivity ranged from .86 (Module 2, Under 5 Years) to .97 (Module 1). The specificity of the ADOS-2 algorithms also demonstrated similar patterns in the males and females in our sample. Additional analyses will examine whether there are sex differences in the ADI-R algorithm. Confirmatory factor analyses will also be conducted to examine whether the 2-factor model of ASD symptoms is comparable across the sexes.

Conclusions:

Preliminary study findings indicate that the gold standard instruments used in autism evaluations function similarly across the sexes.