Spatial covariance analysis for non-demented adults with Down Syndrome

Healthy aging is associated with a regionally spread network pattern of gray matter (GM) reductions on magnetic resonance imaging (MRI) that preferentially involves frontal and selected temporal brain regions. Down syndrome (DS) provides a model of abnormal aging in which there is increased beta amyloid deposition and risk for Alzheimer´s dementia (AD) by the over expression of genes on triplicated regions of human chromosome 21. To identify the age-related network pattern of MRI GM in non-demented adults with DS, we used a multivariate spatial covariance model, the Scaled Subprofile Model (SSM). High resolution T1-weighted brain volumetric MRI scans from 36 DS adults (mean age = 42.3 ± 8.3 years; M/F = 16/20), 29-62 years of age and clinically screened to exclude dementia were included. Statistical parametric mapping (SPM8) Diffeomorphic anatomical registration using exponentiated Lie algebra (Dartel) voxel-based morphometry (VBM) was used to segment brain images into GM, white matter (WM), and cerebrospinal fluid (CSF) partitions, standardize all the images to the template stereotactic space using linear affine transformation and non-linear warping, modulate and smooth the GM maps. SSM analysis was performed on the GM maps to determine the regional network associated with age in the DS group. Greater subject expression of the first two SSM component patterns were correlated with increasing age (R² = 0.31, p ≤ 0.001) in the DS subjects and this association remained significant after we controlled for gender, total intracranial volume (eTIV), and general intellectual ability on the Peabody Picture Vocabulary Test (PPVT-R). The age-related pattern was characterized mainly by extensive reductions in bilateral parietal, precuneus, perisylvian, temporal regions. After PPVT-R has been controlled, higher expression of the age pattern was associated with poorer cognitive performance. These findings indicate that aging in DS is characterized by - regionally distributed pattern of GM reductions in brain regions that have been associated with a greater extent the progressive effects of AD type pathology. Spatial covariance modeling may help to distinguish the effects of normal aging from pathological aging; and may potentially assist in the evaluation of interventions for age-related cognitive decline.