The Impacts of SSRI Medication and Comorbid Anxiety on Brain Function in ASD

Background: Resting state functional connectivity of fMRI data identifies co-activated brain networks and is a promising method for defining biologically similar ASD subtypes and new treatment strategies. However, efforts to characterize alterations in functional connectivity among individuals with ASD are hindered by confounding factors that impact brain function and are variably present across individuals, including medication usage and comorbid psychiatric conditions. Quantifying the impacts of known sources of variability in ASD on functional connectivity will directly inform interpretation of existing findings and guide ASD neurobiological models.

Objectives: Our goal was to quantify functional connectivity characteristics in ASD associated with medication use and psychiatric comorbidities in the context of serotonergic systems.

Methods: Four primary samples were identified from the Autism Brain Imaging and Data Exchange. ASD+SSRI (N=19) were taking an SSRI and no other psychoactive medications, and had no comorbid disorders; ASD-Unmedicated (N=19), an age, sex, IQ, and severity matched ASD comparison group with no medications or comorbidities; ASD+Anxiety (N=19) had a comorbid anxiety disorder and no medications; ASD-NoComorbid (N=19), an age, sex, IQ, and severity matched ASD comparison group with no medications or comorbidities. The four ASD groups will also be compared to typically developing age, sex, and IQ matched controls. Resting state fMRI data was analyzed using the Configurable Pipeline for the Analysis of Connectomes. Whole brain cross-correlation was performed using 116 cortical, subcortical, and cerebellar regions of interest from the Eickhoff-Zilles anatomical template.

Results: SSRI use was associated with reduced cerebellar-cortical and intra-frontal connectivity. In ASD patients taking an SSRI relative to unmedicated ASD Controls, cerebellar-default mode network (DMN) connectivity was increased (cerebellum and posterior cingulate, precuneus), and fronto-cerebellar connectivity was decreased (cerebellum and inferior frontal gyrus, rectal gyrus). In addition, SSRI use was associated with increased connectivity between right ventromedial prefrontal cortex and both left inferior frontal gyrus and right rectal gyrus.

Conclusions: Results indicate pervasive changes in functional connectivity associated with SSRI use and with comorbid anxiety in ASD. Altered cerebellar function has been implicated in ASD, and our findings suggest that SSRIs may mitigate cerebellar-DMN disconnections, an important finding because the DMN may be related to theory of mind and social impairments. We have also shown that comorbid ASD+Anxiety was associated with altered temporal pole connectivity. Anxiety comorbid with ASD includes more atypical symptoms than anxiety without ASD, and the marked changes in social-emotional and fear processing network connectivity provides important markers of ASD+Anxiety that may provide insight into unique neural mechanisms associated with anxiety in ASD.