Salience network segregation mediates the effect of tau pathology on mild behavioral impairment

Abstract

Introduction: A recently developed mild behavioral impairment (MBI) diagnostic framework standardizes the early characterization of neuropsychiatric symptoms in older adults. However, the joint contributions of Alzheimer's disease (AD) pathology and brain function to MBI remain unclear.

Methods: We test a novel model assessing direct relationships between AD biomarker status and MBI symptoms, as well as mediated effects through segregation of the salience and default-mode networks, using data from 128 participants with diagnosis of amnestic mild cognitive impairment or mild dementia-AD type.

Results: We identified a mediated effect of tau positivity on MBI through functional segregation of the salience network from the other high-level, association networks. There were no direct effects of AD biomarkers status on MBI.

Discussion: Our findings suggest that tau pathology contributes to MBI primarily by disrupting salience network function and emphasize the role of the salience network in mediating relationships between neuropathological changes and behavioral manifestations.

Highlights: Network segregation mediates Alzheimer's disease (AD) pathology impact on mild behavioral impairment (MBI). The salience network is pivotal in linking tau pathology and MBI. This study used path analysis with AD biomarkers and network integrity. The study evaluated the roles of salience, default mode, and frontoparietal networks. This is the first study to integrate MBI with AD biomarkers and network functionality.

Keywords: biomarkers; brain connectivity; functional magnetic resonance imaging; network analysis; neuropsychiatric symptoms; positron emission tomography; resting state.

FIGURE 1

Main model. The overall segregation of the SaN from the other association networks mediates the effect of Aβ positive/tau positive (Aβ+/tau+) biomarker status on MBI (indirect effect: β = 0.07, P = 0.044). There is no direct path from AD biomarkers to MBI. Solid and dashed arrows indicate significant and non‐significant relationships, respectively (see Table 2). AD, Alzheimer's disease; Aβ–/tau–, AD biomarker negative; Aβ+/tau–, amyloid beta positive/tau negative; Aβ+/tau+, amyloid beta positive/tau positive; DMN, default‐mode network; GM, gray matter; MBI, mild behavioral impairment; MoCA, Montreal Cognitive Assessment; SaN, salience network.

FIGURE 2

Alternative model. Similar to our main model, only the overall segregation of the SaN from the other association networks mediates the effect of Aβ and tau positivity (Aβ+/tau+) on MBI (indirect effect: β = 0.08, p = 0.039). There is no direct path from AD biomarkers to MBI. Solid and dashed arrows indicate significant and non‐significant relationships, respectively (see Table 3). AD, Alzheimer's disease; Aβ–/tau−, AD biomarker negative; Aβ+/tau−, amyloid beta positive/tau negative; Aβ+/tau+, amyloid beta positive/tau positive; DMN, default‐mode network; FPN, frontoparietal network; GM, gray matter; MBI, mild behavioral impairment; MoCA, Montreal Cognitive Assessment; SaN, salience network.