Gender differences in autism spectrum disorders: Divergence among specific core symptoms

Area of Science:

• Neurodevelopmental Disorders
• Genetics and Genomics
• Psychiatry

Background:

• Autism spectrum disorder (ASD) shows a significant male predominance in community-based studies.
• Underrecognition of ASD in females may stem from how they score on diagnostic tools.
• Existing diagnostic algorithms, like the Autism Diagnosis Interview-Revised (ADI-R), have not been updated for potential sex-based differences.

Purpose of the Study:

• To identify specific items within the ADI-R that discriminate between males and females with ASD.
• To assess the impact of these discriminating items on the final ASD diagnosis.
• To investigate potential gender bias in ASD diagnosis using the ADI-R.

Main Methods:

• Discriminant analysis (DA) was performed on two independent cohorts: the PARIS Study (n=594) and the Autism Genetics Resource Exchange (AGRE) program (n=1716).
• The DA utilized raw scores from all ADI-R items as independent variables to classify participants by sex.
• Replication analysis was conducted to validate findings across different ASD cohorts.

Main Results:

• The DA successfully classified participants by sex with significant accuracy in both cohorts (PARIS: 78.9% males, 72.9% females; AGRE: 72.2% males, 68.3% females).
• Stepwise DA identified specific ADI-R items that significantly differentiate between males and females.
• Four of the identified discriminating items are integral to the current ADI-R algorithm for ASD diagnosis.

Conclusions:

• Certain ADI-R items demonstrably differentiate between males and females with ASD.
• The current ADI-R algorithm may contain a gender bias, contributing to the underestimation of ASD prevalence in females.
• Revising the ADI-R algorithm could improve diagnostic accuracy and reduce gender disparities in ASD identification.