Project Abstract
Speech communication impairments, including persistent difficulties in understanding and interpreting verbal
information during conversation, are a hallmark of childhood autism. Speech-based communication unfolds
over time, and speech comprehension relies on: (1) anticipation of incoming speech as a means of predicting
its content and (2) temporal integration of speech so a listener can accumulate information over time to decode
meaning in an extended utterance. Prominent theories of autism spectrum disorder (ASD) posit deficits in
contextual and global information processing, which are germane to the anticipation and integration of
information during communication. Late childhood is a crucial period for increased and more complex social
interactions, including extended discourse between communication partners. Little is known regarding
anticipatory and integrative components of speech processing, and their contribution to social communication
(SC) deficits, in children with ASD. Advances in experimental design and computational analysis of human
brain imaging data provide a unique opportunity to probe dynamic components of speech comprehension
during naturalistic social interactions in children with ASD, which are difficult to ascertain using behavioral
methods alone. Leveraging innovative fMRI experimental designs, we will for the first time investigate
anticipatory and integrative aspects of naturalistic communication in children with ASD. Our overarching goal is
to identify the neurocognitive mechanisms underlying speech comprehension deficits during naturalistic
communication in children with ASD. The proposed studies include both speaker-listener brain coupling and
temporal integration paradigms and build on our high-impact line of voice perception research in children with
ASD. We hypothesize that children with ASD will show deficits in dynamic mechanisms of speech
comprehension including anticipatory, reactive, and integrative processing with dissociable patterns of
dysfunction in the default mode network (DMN), anchored in medial prefrontal cortex (mPFC) and posterior
medial cortex (PMC), and lateral frontoparietal network (LFPN). While the DMN is often considered a “task-
negative” network, evidence shows that the DMN is crucial for processing social information, including
narrative processing, and is closely linked to SC deficits in ASD. We hypothesize a link between these dynamic
mechanisms and comprehension of global, but not local, narrative information, supporting the Weak Central
Coherence model of ASD. Findings will provide new insights into speech comprehension impairments and
advance our understanding of the role of the DMN in SC and ASD. Our studies will provide critical information
regarding the neurobiological origins of communication impairments in ASD and will inform the development of
age-appropriate treatment for older children with ASD. Our aims are in line with the NIH directive on Autism
Research (PA-18-400), emphasizing brain mechanisms and sophisticated measures of social communication.