Covariation Between Brain Function (MEG) and Structure (DTI) Differentiates Adolescents with Fetal Alcohol Spectrum Disorder from Typically Developing Controls

• Published in: Neuroscience 2020-11, Vol.449, p.74-87
• Main Authors: Pinner, John F.L., Coffman, Brian A., Stephen, Julia M.
• Format: Article
• Language: English
• Subjects: Adolescent; Adult; Anisotropy; Child; Cognition; Diffusion Tensor Imaging; diffusion tensor imaging (DTI); Female; Fetal Alcohol Spectrum Disorders; fractional anisotropy (FA); Humans; joint independent component analysis (jICA); Magnetoencephalography; magnetoencephalography (MEG); Pregnancy; prenatal alcohol exposure (PAE); Young Adult
• ISSN: 0306-4522; 1873-7544
• DOI: 10.1016/j.neuroscience.2020.09.053

•Covariation of brain function (MEG) and structure (FA) differentiates adolescents with FASD from controls.•jICA assessed differences in function/structure associations were in a hub of sensory integration, the cerebellum.•Relationship between MEG and FA in cerebellum is associated with verbal ability and impulsivity.•These findings suggest that multimodal integration of MEG and FA provides novel associations between structure and function. The behavioral, cognitive, and sensory difficulties experienced by individuals exposed to alcohol prenatally currently fail to provide early identification for fetal alcohol spectrum disorder (FASD). Attempting to advance this pursuit through a multivariate analysis, we collected magnetoencephalography (MEG) data during auditory, somatosensory, visual paradigms, DTI, and behavior in adolescents ages 12–21 years (FASD: N = 13; HC: N = 20). We assessed the relationship between brain function (MEG) and structure (fractional anisotropy (FA)) utilizing joint independent component analysis (jICA), and examined how this measure relates to behavior. We identified 5 components that reveal group differences in co-variation between MEG and FA. For example, component 5 (t = 3.162, p = 0.003, Hedges’ g = 1.13) contained MEG activity corresponding to all three sensory modalities, most robustly in occipital lobes, and DTI-derived cerebellar FA, underlying the role of the cerebellum in sensory processing. Further, in HCs component 5’s loading factor was positively correlated with verbal ability (r = 0.646, p = 0.002), indicating higher covariation was associated with better verbal performance. Interestingly, this relationship is lacking in FASD (r = 0.009, p = 0.979). Also, component 5 loading factor negatively correlated with impulsivity (r = −0.527, p = 0.002), indicating that stronger function-structure associations were associated with individuals with lower impulsivity. These findings suggest that multimodal integration of MEG and FA provides novel associations between structure and function that may help differentiate adolescents with FASD from HC.