Exploring the interaction of APOE-ε4and PICALM rs3851179 with dynamic functional connectivity in healthy middle-aged adults at risk for Alzheimer's disease.

Abstract

<i>Objective.</i>This study investigates whether dynamic functional connectivity (dFC) dwell-time patterns derived from resting-state functional magnetic resonance imaging (rs-fMRI) can distinguish Alzheimer's disease (AD) genetic risk profiles, specifically the APOE-ε4(A+) and phosphatidylinositol-binding clathrin assembly protein rs3851179 (P+) variants, in cognitively healthy, middle-aged adults.<i>Approach.</i>We estimated recurring dFC clusters from rs-fMRI data and quantified the dwell-time (total duration spent in specific connectivity states) for three cohorts: not-at-risk, A+P-, and A+P+. To evaluate the utility of these temporal features, group differences in dwell-time profiles were assessed, and logistic regression with permutation testing was employed to classify genotypes based on dFC patterns.<i>Main results.</i>Individuals in at-risk groups (A+P- and A+P+) exhibited significantly reduced dwell-time in left-hemisphere hubs compared to the not-at-risk group, aligning with known left-hemisphere vulnerability in early AD progression. The logistic regression models achieved above-chance discrimination of genotypes, with permutation tests confirming a significant trend when distinguishing not-at-risk individuals from the combined at-risk cohorts.<i>Significance.</i>These findings suggest that the temporal dFC features are sensitive to subtle functional brain alterations linked to AD genetic risk before clinical symptoms appear. Dwell-time features represent a promising physiological marker for early risk stratification and warrant further validation in larger longitudinal datasets. Our code is available athttps://github.com/Shyamal-Dharia/APOE-PICALM-dFC-dwell-time.git.