Gender-Specific Autism Clusters presented at INSAR 2025

It was lovely to be in Seattle for the International Society for Autism Research (INSAR) Annual Meeting in May 2025. The city welcomed us with pleasant weather and the conference venue in the surprisingly calm city center provided an excellent backdrop for meeting our friends. During our time there, we presented our latest findings on gender differences in autism presentation (and fell in love with Pike Place Market!).

Our Poster Presentation

• Gender-Specific Clustering in Autism: Implications for Female Identification and Assessment

  Uğurcan Vurgun, Sunghye Cho, Riccardo Fusaroli, Maggie Rose Pellela, Azia Knox, Aili Hauptmann, Maxine Covello, Monica Madya, Meg Lyons, Amanda Lee, Lizzy Fulop, Alison Russell, Grace Hicks, Juhi Pandey, James Fiumara, Birkan Tunc, Evangelos Sariyanidi, Mark Liberman, Robert T. Schultz, & Julia Parish-Morris. Poster presentation at INSAR 2025.

Key Findings

Our research identified behaviorally-defined autism subtypes, revealing a high-functioning, male-skewed phenotype (AUT_B) that current diagnostic practices may ignore in girls. Through unsupervised learning techniques (k-means for clustering and UMAP for visualization) applied to behavioral assessment data from 99 adolescents, we found:

• Two distinct clusters within the autistic participants: AUT_A (showing broader difficulties) and AUT_B (with a near-typical behavioral profile)
• A significant sex skew: females constituted 51% of the AUT_A group but only 10% of the AUT_B group
• 16 behavioral features where AUT_B participants were indistinguishable from neurotypical participants despite meeting diagnostic criteria

These findings support the hypothesis of a camouflaged, under-diagnosed female phenotype in autism and highlight the need for sex-sensitive screening instruments that target subtle internalizing and self-esteem markers.

Our wonderful PI, Julia Parish-Morris, and me

Implications

Our research has significant implications for improving autism identification and assessment practices, particularly for females whose presentation may not match traditional diagnostic expectations. By recognizing these gender-specific differences, we can work toward ensuring timely identification and appropriate support for all individuals on the autism spectrum.

We are grateful for the opportunity to contribute to the important discussions on autism research and to learn from fellow researchers at INSAR 2025! There was so much to learn overall, I need some weeks (or months?) to digest everything I saw.