Rates of change in structural and functional brain connectivity are associated with longitudinal changes in episodic memory in non‐demented older adults
Year: 2023
Type: article
Abstract: Background Individual differences in brain connectivity have been associated with cognitive variability in older adults, making them ideal markers of early changes that may predict subsequent cognitive decline. Using MRI measures, we previously demonstrated that cross‐sectionally, both structural and functional connectivity each uniquely contributed to variability in episodic memory performance (Alm et al., 2022). The present investigation expanded on these findings to examine whether longitudinal changes in brain connectivity were related to changes in episodic memory. Method Analyses included 152 non‐demented BIOCARD study participants (at first scan, 135 cognitively normal and 17 with Mild Cognitive Impairment). Structural connectivity was measured using radial diffusivity (RD) and mean diffusivity (MD) of two limbic pathways (fornix and hippocampal cingulum) derived from diffusion‐weighted MRI. Based on our previous findings, the salience network was chosen as the functional connectivity network of interest, as measured by resting‐state fMRI. Delayed episodic memory was evaluated with a composite score. Separate linear mixed effects (LME) models with multivariate outcomes tested whether longitudinal changes in episodic memory were related to longitudinal changes in each measure of brain connectivity (mean follow‐up = 3.11 (0.96) years), covarying age, sex, education, and diagnosis. LME models were fit simultaneously, allowing for the random effects in each model to be correlated. Significant correlations between the random slopes indicated a relationship between longitudinal changes in outcome variables. Result Longitudinal memory decline was associated with increases in fornix RD and MD (p’s < .01) and increases in salience network functional connectivity (p < .01), but not hippocampal cingulum measures (p’s > .05). When only cognitively normal participants were included, the fornix findings remained significant (p’s < .01); however, the salience network model failed to converge. Conclusion Extending prior cross‐sectional results, these findings show that rate of memory decline over time is associated with rate of change in both structural (fornix) and functional (salience network) connectivity markers, providing further evidence that both connectivity modalities uniquely contribute to episodic memory function. These may, therefore, be important brain markers in predicting person‐specific disease trajectories, considering structural connectivity at the earliest stages of aging and functional connectivity with disease progression.
Source: Alzheimer s & Dementia
DOI: 10.1002/alz.079875
Institutions: Johns Hopkins University, Johns Hopkins Medicine
Language: English
Subfield: Psychiatry and Mental health
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W4390194367