Khushboo Verma1, Satwant Kumar2
1Department of Neurology, 2Center for Perceptual Systems, The University of Texas at Austin
Objective:
1. To investigate how cognitive decline affects psychometric representations of global (Mini-Mental State Examination (MMSE) and Clinical Dementia Rating scale (CDR)) and domain-specific neuropsychological measures.
2. To examine the association between cognitive decline and β-amyloid burden using tracer Positron Emission Tomography (PET) imaging.
Background:
One of the central goals of cognitive neuroscience is to understand how anatomical structure relates to function. In dementia, cognitive functions are disrupted and brain regions are selectively involved. It is unclear, however, how the anatomical distribution of pathological changes relates to cognitive functions as measured by clinical assessments.
Design/Methods:
OASIS-3 dataset was used to evaluate the psychometric correlates of cognitive decline in cognitively normal (n=755), and cognitively impaired subjects (n=655) with ages ranging from 42 to 95 years. Anatomical distribution of β-amyloid burden was assessed using florbetapir (AV-45) and Pittsburgh compound B (PIB) PET imaging. Analysis was limited to subjects who underwent cognitive testing and PET imaging within six months.
Results:
Global cognitive decline as measured with MMSE and CDR is associated with brain-wide β-amyloid deposition (randomization test, false detection rate adjusted p-values (q-values) <0.001). Both PIB and AV-45 capture similar neural correlates of global cognitive measures, but show differential selectivity for individual neuropsychological measures (Pearson correlation coefficients, q-values <0.05). Representational similarity analysis of β-amyloid deposition further confirmed these findings, revealing distributed patterns of global cognitive decline across the cortical and subcortical regions (q-values <0.05). There is a continuum in the low-dimensional psychometric and β-amyloid deposition representation (binomial test, q-values <0.001), suggesting behavioral and anatomical heterogeneity within subjects with a diagnosis of cognitive impairment.
Conclusions:
1. Global cognitive assessments are associated with distributed cortical and subcortical representation.
2. There is a distributed rather than a localized system of cortical and subcortical regions that maintain cognitive function.
3. Behavioral and anatomical heterogeneity exists among subjects with cognitive impairment.