Increased Connectivity of Voice Processing Brain Networks in Females with Autism: A Preliminary Study of Gender Differences in ASD

Background:

Autism spectrum disorders (ASD) are poorly understood in females. The incidence of autism is greater in males, but it is unknown whether this disparity is due to actual occurrence rate or a bias in diagnosis; the criteria for diagnosing ASD has been developed using data almost exclusively from studies of males, and symptoms of ASD often manifest differently in females. Some believe that, due to genetics and environmental factors, females are more likely than males to have characteristics that help them compensate for and mask symptoms of ASD.

Autism is characterized by a lack of engagement with social cues, a primary of which is speech. We have developed a model of human voice processing composed of the core voice processing system, anchored in the superior temporal sulcus (STS), and an extended voice processing network, encompassing reward and affective areas. Previous research in our lab has shown that children with ASD show underconnectivity between core and extended voice processing networks; however, it is unknown whether connectivity of the voice processing system in ASD varies across genders.

Objectives:

The objective of this study is to examine gender differences in connectivity of voice processing brain systems in high-functioning ASD.

Methods:

15 females and 15 males with high-functioning ASD (8-16 y/o) were matched on age, IQ, ADOS Social Affect score, and head motion in the scanner. All participants underwent resting-state fMRI scans. In addition, participants were administered an extensive battery of questionnaires and clinical and cognitive assessments to determine IQ and ASD diagnosis. Functional connectivity analyses were used to examine between-group differences in connectivity for bilateral posterior and anterior STS (i.e., core voice processing network) seeds.

Results:

Compared to males with ASD, females with ASD demonstrated increased connectivity between nodes of the core voice processing network and a wide array of brain systems, including (1) reward and affect processing regions, including the nucleus accumbens, ventral tegmental area, orbitofrontal cortex, and amygdala, (2) canonical language processing regions, instantiated in middle and superior temporal cortex, inferior frontal gyrus, and angular gyrus, and (3) key nodes of the salience network, including dorsal anterior cingulate and anterior insula. No voxels were greater for males compared to females with ASD.

Conclusions:

Preliminary results suggest that compared to males with ASD, females with ASD demonstrate increased connectivity between the core voice processing network and an extended network of brain regions, including regions implicated in reward, affective, language, and salience processing, providing initial indication that females and males with ASD may in fact utilize different brain mechanisms for processing social cues. Further analyses and larger participant samples are needed to properly investigate the interaction between gender and diagnosis and the role this interaction plays in connectivity patterns of voice processing brain networks.