Martha R. Herbert - US grants

Autism is a behavioral syndrome characterized by impairments in social interactions and communication and by repetitive, stereotyped behavior. To date there is no clear and consistently replicated neuroanatomic correlate. While specific, focal core neuroanatomic or cognitive deficits have been sought, this study will examine the syndrome's more pervasive features. The purpose of this study is to understand these pervasive abnormalities at the levels of brain structure and of structure-function relationships. The finding of large brains in autism suggest a pervasive abnormality of the brain and abnormal regulation of brain development whose distributed consequences cannot be fully characterized by focal studies. Moreover, the higher-level cognitive functions disordered in autism are likely to have distributed rather than strictly localized or modular neuroanatomical correlates. By using a novel, detailed neuroanatomically-based whole-brain morphometric method, we will be able for the first time to characterize pervasive, distributed brain volumetric abnormalities. The study will be performed on subjects from two NIH-funded projects, the Autism and Language Disorders Nosology Project, NINDS 20489; and the Language in Autism: Clinical and Basic Studies Project, NIDCD PO1 DC 03610. The Specific Aims will be 1) to test the morphometric hypothesis that dysregulated brain development in autism will manifest in abnormalities in morphometric quantifiers of size, scaling, and profiles of variance, and 2) to test the developmental cognitive neuroscience hypothesis that key abnormal cognitive and linguistic functions in autism will correlate better with distributed morphometric abnormalities such as deviations in neuroanatomic scaling and profiles of variance than with localized deviations in size of neuroanatomic modular units. The system-oriented whole brain morphometric profiles should provide new classes of neuroanatomical correlates that are more appropriate for the widespread higher-level cognitive functions disordered in autism. The nature of developmental dysregulation and neural systems abnormalities in this disorder should thus become clearer.

Autism is a behaviorally defined but biologically heterogeneous disorder, involving a cluster of impairments in language, social interaction and behavior. Large brains are common in autism, but their underlying tissue microstructure and their impact on the autistic phenotype are not understood. This proposal is based on the model that increased white matter volume is related to impaired connectivity and to underlying core processing abnormalities in autism. To test this model we will pursue our prior findings that white matter enlargement drives large brains and is localized to the later-myelinating radiate zone (the white matter closest to the cerebral cortex). We will use multispectral MRI to gain systematic information about the tissue microstructure and patterns of structural connectivity of enlarged white matter in autism, and we will relate these findings to cognitive and behavioral phenotype and to genome data. Forty autistic and forty typically developing boys ages 6-10 will receive a cognitive and behavioral battery and genetic studies. Specific Aim 1: We will construct a convergent multimodal tissue microstructure profile, coregistering boundaries from whole-brain segmentation and parcellation with estimated tissue parameters (T1, T2*, PD) magnetic resonance and diffusion tensor (DTI) data. Data will be analyzed for patterns discernable in multiple measures that may illuminate altered tissue microstructure. Specific Aim 2: We will investigate structural connectivity utilizing DTI tract segmentation and multimodal tissue profiling of segmented tracts. Specific Aim 3: We will test brain-connectivity-processing-endophenotype correlations, including genome data. We predict 1) that perturbations of volume and tissue integrity will be correlated with each other, and will be distributed by zone-more in later-myelinating radiate than deep zone white matter rather than by tract or neural system; 2) that measures of impaired complex processing (e.g., central coherence) will a) correlate with measures of widespread anatomical changes (e.g., white matter increase), b) be associated with autism even if regional brain changes are absent, and c) be associated in severity with severity of specific behavioral endophenotypes as well as of widespread anatomical abnormalities, and 3) that common anatomical and behavioral findings will be associated with varied genome changes. Our study design will also detect alternate outcomes and will yield data relevant for multiple levels of translational research.