Abstract

Converging evidence suggests that the pathophysiologic processes in the brains of Alzheimer's disease (AD) patients begin decades before symptoms occur1–3. Individuals in the preclinical stages of AD often report self-perceived decline in cognitive functions. The self-perception of cognitive decline can be assessed via the cognitive change index (CCI)4,5, a recently proposed measure of perceived neuropsychological function focusing on the memory, executive, and language domains from both self and informant perspectives. In this study, we examine the association of CCI with a novel diffusion metric for axonal density. The axonal density metric is derived via a three-compartmental model, Neurite Orientation Distribution and Density Imaging (NODDI)6. Sixty-one participants with cognitively normal (CN, N=21), subjective cognitive decline (SCD, N=23) and mild cognitive impairment (MCI, N=17) underwent Hybrid Diffusion Imaging (HYDI)7,8 that consists of five b-value shells. Diffusion images were corrected for motion, eddy and geometric distortion9 prior to nonlinear transformation to standard MNI space using ANTS registration10. Median values of diffusion metrics including DTI, NODDI, and q-space analysis were extracted from skeletonized 48 white matter (WM) ROIs provided in the JHU atlas11. General linear regression analyses were performed to test associations between diffusion metrics and CCI scores adjusted for age, sex, and education. For comparisons of demographic and cognitive variables, ANOVAs were employed with Tukey's tests followed by post-hoc t-tests. The three groups did not differ significantly in age, sex, and education distribution (Table 1). Strong associations (p<0.001) between CCI and diffusion metrics were found in ROIs belonging to the limbic system. Specifically, fractional anisotropy (FA) of DTI in the right stria terminalis had negative correlation with CCI-12 (i.e., episodic memory) (p<0.001, Figure 1). Axonal density in the right uncinate fasciculus negatively correlated with both CCI-12 and CCI-TOT (20 item) (p<0.001, Figure 2). Other ROIs had no significant correlations. These results suggest that lower axonal density and fiber coherence are risk factors for self-perceived memory decline. The two most vulnerable white matter tracts - the right stria terminalis and uncinate fasciculus, connect between the amygdala and hippocampus - two of the areas that show the earliest disease-associated changes. ROI map and scatter plot of CCI score and fractional anisotropy (FA). A. The ROI in red denotes the right stria terminalis. B. FA of diffusion tensor imaging (DTI) negatively correlated with CCI-12 (episodic memory) with p=0.0007. Age, sex and education were adjusted in the linear regression analysis. Green dots denote cognitively normal (CN); blue dots denote subjective cognitive decline (SCD); red dots denote mild cognitive impairment (MCI); and the black solid line denotes the regression line. ROI map and scatter plot of CCI score and axonal density. A. The ROI in red denotes the right uncinate fasciculus. B. Axonal density negatively correlated with both CCI-12 (episodic memory) with p=0.00009 and CCI-TOT (20 items) with p=0.0002. Age, sex and education were adjusted in the linear regression analysis. Green dots denote cognitively normal (CN); blue dots denote subjective cognitive decline (SCD); red dots denote mild cognitive impairment (MCI); and the black solid line denotes the regression line.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call