Premium
Default mode network dedifferentiation predicts cognitive performance in Alzheimer disease
Author(s) -
Meeker Karin L.,
Ances Beau M.,
Gordon Brian A.,
Morris John C.,
Benzinger Tammie L.S.,
Waring Jill
Publication year - 2020
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.044790
Subject(s) - clinical dementia rating , cognition , default mode network , dementia , cognitive decline , effects of sleep deprivation on cognitive performance , multivariate analysis of variance , psychology , disease , alzheimer's disease , neuroscience , cognitive impairment , medicine , statistics , mathematics
Abstract Background During aging, brain networks become less functionally specialized, known as dedifferentiation. When dedifferentiation occurs, brain networks become less adaptive, leading to failures in cognitive processing. Although global dedifferentiation occurs with age, it is unknown how global and local dedifferentiation change as a function of cognitive impairment (measured by the Clinical Dementia Rating Scale [CDR]). We assessed whether global and/or local dedifferentiation differ as a function of CDR status and predict cognitive performance in cognitively normal individuals (i.e., CDR=0) and those with mild symptomatic Alzheimer disease (AD; CDR>0). Method Neuroimaging and clinical data from 432 participants (CDR0=356; CDR0.5=63; CDR1=13) were analyzed. Graph theory analytic approaches were applied to RS‐FC data for global dedifferentiation and 7 individual association networks. In graph theory, dedifferentiation is measured by computing segregation, which quantifies the presence of specialized and densely interconnected brain regions. Segregation was computed as (within‐network mean minus between‐network mean)/within‐network mean. Cognitive performance was assessed using Preclinical Alzheimer Cognitive Composite (PACC). The association between age and global dedifferentiation and network specific dedifferentiation was assessed using Pearson’s r correlations. Changes in global and association network dedifferentiation as a function of CDR were evaluated using a multivariate analysis of variance (MANOVA). Individual regression models determined if global and/or network specific dedifferentiation predicted cognitive performance. Age was accounted for in all models. Result Global dedifferentiation increased with age (i.e., segregation decreased: r = ‐0.15, p <0.01; see Figure 1), as did dedifferentiation of the default mode network (DMN; r = ‐0.14, p <0.01) and ventral attention network (VAN; r = ‐0.11, p <0.05). No other networks were associated with age ( p s>0.05). Dedifferentiation of DMN differed as a function of CDR with dedifferentiation higher (i.e., lower segregation) in cognitively impaired individuals compared to cognitively normal individuals ( p s£0.01; see Figure 2). Greater DMN dedifferentiation also predicted poorer cognitive performance, i.e., lower PACC score; p =0.001. No other measure of dedifferentiation (i.e., global or network specific) differed by CDR or predicted cognitive performance. Conclusion Results suggest that global and focal dedifferentiation occurs with aging, but only focal dedifferentiation occurs in AD. Furthermore, dedifferentiation of the DMN is sensitive to decreases in cognitive performance.