Rutgers scientists address under-diagnosis of female autism using new tools


Why are American boys four times more likely to be diagnosed with autism than girls? Are autistic spectrum disorders more prevalent in males, or is there a problem with our diagnostic tools? Researchers at Rutgers University have recently published findings that suggest the latter.

The team — led by Elizabeth Torres, an associate professor in the Department of Psychology at Rutgers University — analyzed functional magnetic resonance (fMRI) images collected from individuals with autism and Asperger’s. They were able to link increased involuntary head movements during fMRI scans to a diagnosis of autism in females.

The group’s findings were published in "Frontiers in Integrative Neuroscience" in June 2017.

Torres and her associates were driven by the desire to identify physiological markers for autism. The current evaluation tools rely on sociological observations — how the children play with toys and interact with others. Critics of this model point out socially acceptable behaviors vary by culture, making such an approach subjective.

“If someone asks you ‘what do you think of autism?’ most people would say ‘difficulties with socialization’, etc. Even though it’s talked about as a neurodevelopmental disorder, it’s been characterized as this cognitive or social thing,” explained Caroline Whyatt, a postdoctoral researcher in the Department of Psychology who co-authored the paper.

The Centers for Disease Control and Prevention released a study in 2014 estimating 1 in 42 boys are on the autistic spectrum, compared to only 1 in 189 girls. Many researchers believe this is due to under-diagnosis of females, based upon society’s perception of acceptable feminine behaviors.

“A girl playing with a doll in the corner which may be a somewhat isolated activity is, arguably, something we think wouldn’t look socially odd. Whereas a boy sitting in the corner playing quietly with nothing would maybe look a little different,” Whyatt said.

There are also different observational criteria based on your scientific discipline — psychology and psychiatry use separate manuals to diagnose autism. The American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders (DSM-5) places Asperger’s on the autism spectrum, the International Classification of Diseases (ICD) characterizes them as separate disorders.

Motor skills are an under-used component of autism diagnosis, Torres said. The worldwide ICD specifies that “poor motor control” is linked to Asperger’s, but the American DSM-5 only lists “restrictive, repetitive behaviors” such as hand flapping as an indicator of autism. Torres considers this a significant omission.

The team focused in on what researchers typically excise as junk data from their fMRIs — the scans where the subjects move their heads and disrupt the image. Torres hypothesized that a neurological condition such as autism would result in increased involuntary head movements, and thus a heightened number of these twitches could correlate to a diagnosis of autism.

“fMRI is one of the subsets of data that are readily available because there are lots of data-sharing initiatives,” said Sejal Mistry, a biomathematics graduate who also worked on the project. The group chose to use the Autism Brain Imaging Data Exchange – an open-access, international database of fMRI scans.

Mistry, who was part of Torres's lab for most of her undergraduate studies, figured out how to extract the head motions encoded in the fMRI data stored in this database. She then set up a system of cataloging the downloaded data for her colleagues to access.

The team analyzed the fMRI scans of 309 females, 63 of them without an autism diagnosis belonging to a control group. They also examined 1,890 male participants, of whom 300 were in a control group. The study also compared individuals diagnosed with Asperger’s to autism.

Whyatt and Carla Caballero — also a postdoctoral researcher in the Department of Psychology — analyzed the data independently, checking the correlations were statistically real.

As a result, the authors could, for the first time, propose a quantifiable way to identify autism in girls. The study found that the number and severity of involuntary head movements were greater in autistic individuals, regardless of gender or age.

Looking at movement profile data, the researchers can predict with high probability if an undiagnosed patient had autism or Asperger’s.

“We live in a world where there is data everywhere. We have all these technological advances, but sometimes it seems that medicine is lagging, specifically with young children,” Mistry said.

The most surprising finding came when comparing females diagnosed with Asperger’s and autism. 

“Asperger’s is a type of autism called high-functioning: people who are socially inappropriate,” Torres said. 

They were expecting that girls with autism would make more involuntary movements, but found the girls with Asperger’s moved more.

This suggests some shortcomings with how Asperger’s is defined, particularly in girls. If it was only a mild form of autism, then the accompanying involuntary head movements should also be milder.

Torres and coworkers have patents pending on their research, hoping it will form the basis of an improved, quantifiable diagnostic detection test for autism in children and adults. They are continuing to use the principles of involuntary movement detection to understand other nervous system disorders.

“The uniqueness of our method is its standardization, you can use it in a 3-year-old, in a 1-month-old or an 80-year-old. You can use it in Parkinson’s, autism or schizophrenia,” Torres said.


Claire Jarvis

Comments powered by Disqus

Please note All comments are eligible for publication in The Daily Targum.