Abstract
Background: Alzheimer's disease (AD) manifests differently in males and females. However, the neuro-mechanism and influence factors are still unknown. Objective: To explore sex differences in brain network topology during AD disease progression and its association with cognition and possible influencing factors. Methods: Resting-state functional magnetic resonance imaging (MRI) data and cognitive scores were collected from 82 AD patients (50 females), 56 amnestic mildly cognitive impaired patients (29 females), and 63 healthy controls (38 females). Global and regional topological network metrics and modular architecture were calculated. Two-way ANOVA was performed to explore group and sex interactions and their main effects. Mediation analysis was used to explore the relationship among education, inter/intra-network connectivity, and the Mini-Mental State Examination (MMSE) score. Results: Lower levels of education, lower MMSE scores, and a positive correlation between the level of education and MMSE scores were found in female AD patients (p = 0.024, r = 0.319). Significantly lower connectivity strength within the sensorimotor network, dorsal attention network, ventral attention network (VAN), and between the sensorimotor and VAN were observed in male AD patients (p = 0.006, 0.028, 0.046, and 0.013, respectively). Group and sex interactions were also found in nodal properties, mainly in the frontal lobe, temporal lobe, middle cingulum, precuneus, and postcentral gyrus. Several of the altered brain network properties were associated with cognitive behavior in male AD patients. Education regulated the MMSE score through the mediation of connection strength between the default mode and limbic networks (LN) in the patient group (aMCI and AD combined). Conclusion: Our results demonstrate that sex differences exist at the brain network level in AD. Sex differences in network topology and education are correlated with sex differences in cognition during AD progression.