Multi-modal connectivity methods for the validation of Fetal Alcohol Spectrum Disorder diagnostic criteria

Project Summary: FASD is a serious public health crisis with 80,000 to 120,000 U.S. pregnancies per year at high risk. FASD diagnostic clinics are very sparse, difficult to sustain because of diagnostic complexity, unable to meet demand, and plagued by disagreement about diagnostic criteria. A recent comparison of five FASD diagnostic systems acknowledged that the lack of objective biomarkers for prenatal alcohol exposure (PAE) contributes to disagreement in the field. The goal of the proposed work is to develop and refine neuroimaging methods capable of identifying subtle neurodevelopmental abnormalities in PAE. A critical component of this project will be to apply these refined neuroimaging ?biomarkers? to examine the validity and performance of the most commonly-used FASD diagnostic criteria (alcohol exposure, dysmorphology, growth, and cognition). The proposed project will push the development of imaging methods and will also include an examination of the sensitivity and specificity of current diagnostic factors. The aims are: 1) perform a comprehensive characterization of the FASD ?connectome? using state-of- the-art Human Connectome Project (HCP) methods in children with PAE, 2) evaluate the sensitivity and specificity of FASD diagnostic criteria to provide objective evaluation of FASD criteria (which remain in flux), 3) characterize relationships between connectomics and cognitive functioning, & 4) examine the developmental course of functional connectivity, white matter connectivity, cortical gyrification, and intra-cortical myelin. Participants will undergo longitudinal re-assessment at 15-months to match the new HCP-development paradigm (HCP-D) to allow for comparison to a large normative dataset being collected in Minnesota as part of the HCP-D project. We will capitalize on the developments spawned from the $40 million HCP and will benefit from evaluating the brains of children with PAE against the backdrop of a very large HCP normative dataset (n=1300). The proposed study will examine 45 participants with PAE and 45 matched controls, ages 8-16, all of whom will be evaluated by the CIFASD dysmorphology project. Participants will undergo MRI scans at the University of Minnesota's Center for Magnetic Resonance Research (CMRR) at baseline and again 15 months later. The imaging protocol, which contains very high resolution structural, diffusion- weighted, and fMRI scans will evaluate cortical gyrification, intra-cortical myelin, white matter connectivity, and functional connectivity. Data from the HCP-D project will be used as secondary ?normative? comparison dataset. We expect to find a number of subtle neurodevelopmental abnormalities in PAE that have not previously been observed and we will characterize individuals with these metrics. Individual identification of neurodevelopmental anomalies will then allow us to probe the screening properties of commonly used FASD diagnostic criteria and to test the performance of Dr. Mattson's Decision Tree model for rapidly identifying children with developmental effects of PAE. The study will have direct implications for FASD clinical practice and public policies as it will provide objective data about the relative advantages and disadvantages of individual diagnostic elements and it will allow for comparisons of competing FASD diagnostic systems.