Girls, Boys and the Language of Autism


Cultural norms profoundly influence the ways we believe boys and girls “should” typically play, interact socially, and communicate. For example, cultural norms would predict that from a basket of toys, boys would pick the cars and trucks, while girls would choose the dolls. These preconceived beliefs extend to how boys and girls communicate, with boys commonly seen as more assertive, aggressive, and object-oriented in their speech and interactions, whereas girls might be perceived to be quieter, more empathetic and socially engaged, and more likely to talk about what they think and feel.

Whether it’s nature or nurture that instills these gender-based perceptions, the impact of these expectations on the diagnosis and definition of Autism Spectrum Disorder (ASD) ** may be profound.  A growing body of research suggests that ASD appears differently in girls and women than it does in boys and men. The majority of ASD research has been conducted in males, meaning the diagnostic criteria and our understanding of ASD have been overwhelmingly based on how ASD presents in boys and men. Does the fact that fewer women have been diagnosed with ASD – and therefore underrepresented in research - mean girls are protected from ASD somehow? Or does ASD look different in girls, meaning that they are missed or misdiagnosed - or don’t meet male-referenced diagnostic criteria and thus are excluded from traditional ASD studies?

At the Center for Autism Research (CAR) at the Children’s Hospital of Philadelphia (CHOP), developmental psychologist Julia Parish-Morris, Ph.D., began to raise these questions in her research. She wanted to know why girls were not diagnosed with ASD as frequently as boys and what ASD symptoms look like in girls. In CAR’s Quantitative Linguistics Lab (QL2), Dr. Parish-Morris and her team analyzed thousands of vocal samples from boys and girls, both with and without ASD, looking for differences in the way autistic girls and boys used certain types of words during storytelling. They found that girls on the autism spectrum used cognitive processing words, such as “think” and “know”, much more frequently than autistic boys, even when their ASD symptoms were of a comparable severity. The results, which were recently published in the journal Molecular Autism, could help to explain the camouflaging effect seen in girls with ASD, and how it may lead to  delayed or missed diagnosis and treatment.

I think every time we learn a little bit more about how ASD manifests in girls and women, we increase our chances of identifying them in childhood and being able to quickly provide the personalized supports they deserve,” said Parish-Morris.

This is not the first study from QL2 to find sex differences between boys and girls with ASD. Previous research revealed subtle, but important, differences in the way boys and girls with ASD use filler words like “um” and “uh”. In what may be termed linguistic camouflage, Parish-Morris and her colleagues found that boys with ASD use “um” fillers far less frequently than their male peers without ASD, whereas girls with ASD used “um” nearly as much as their peers. Many small linguistic differences can result in big changes in how girls are perceived. At INSAR 2019, QL2 lab member and research assistant Meredith Cola presented new research showing that girls with ASD make more typical first impressions than boys during a naturalistic conversation with a non-expert. “This evidence of behavioral camouflage by girls during a 5-minute conversation was fascinating in light of deeper social communication challenges that were revealed during clinical interviews,” explained Cola.

Girls with ASD don’t just sound different – they may perceive the world differently too! In a separate study presented at the International Society for Autism Research’s 2018 Annual Meeting, the QL2 team collaborated with Dr. Clare Harrop at UNC to compare whether or not girls and boys on the autism spectrum paid more attention to faces, or to the non-social aspects in a depiction of a social interaction. They found that autistic girls spent more time focused on faces than autistic boys, but less than typically developing girls and about the same as typically developing boys. “This finding highlights the importance of understanding the social challenges of ASD in the correct context, with the appropriate reference group. That is, to understand autism in girls, the behaviors of girls with ASD should be compared with other girls; not with a mixed group of non-ASD girls and boys.”

Taken together, these recent studies from the QL2 team demonstrate how actual gender-based differences, coupled with deeply held beliefs about how boys and girls act socially, may have led to the perception that boys with ASD are more impaired than girls, even when the severity of their symptoms is equivalent. “Researchers are beginning to acknowledge that ASD presents differently in boys and girls (and in non-binary or transgender individuals), and that’s vitally important to our ability to properly understand and serve all people with ASD,” says Dr. Parish-Morris. “As we better understand ASD and what the symptoms look like across many different groups of children, we can better identify causal factors and develop more precise interventions and supports.”