Reply to Perrykkad and Hohwy: When big data are the answer

Perrykkad and Hohwy (1) argue that autism was historically diagnosed predominantly in males (4 males to 1 female) (2) and thus the defining characteristics of autism are male biased. They conclude that because the Autism Spectrum Quotient (AQ) was developed within this historical framework, ( i ) the AQ may not capture the phenotypic profile of autistic females and ( ii ) the AQ has a male bias. We agree that autism has been historically underdiagnosed in females but disagree with their conclusions for at least 5 reasons. First, our study (3) shows that regardless of sex, autistic individuals have, on average, high AQ scores. Multiple studies (4, 5) show that autistic females score … [↵][1]2To whom correspondence may be addressed. Email: dmg39{at}cam.ac.uk or sb205{at}cam.ac.uk. [1]: #xref-corresp-1-1