Objective:

To determine the contributions of sex, menopause, and sex hormones to functional network modularity to provide new insights into women’s increased vulnerability to Alzheimer’s disease (AD).

Background:

Women are at twice the lifetime risk for AD than men and exhibit faster tau spread, which follows functional connectivity pathways. Alterations in the default mode network (DMN) emerge in preclinical and early-stage AD, echoing sex differences in DMN organization around menopause. Network modularity typically decreases over aging, possibly indicating reduced specialization; however, the influence of specific reproductive variables remains unknown.

Design/Methods:

Structural and functional MRI were acquired from 222 female and 141 male participants in the Human Connectome Project-Aging dataset. Menopause stage was assigned using STRAW-10 criteria. Whole brain (WBM) and DMN modularity were computed using the Louvain algorithm on weighted, signed functional connectivity matrices, generated by applying z-transformed Pearson correlations between regions in the Shen 368 parcellation.

Results:

Females had higher WBM than males in ages 41-45 and a negative relationship between WBM and age; no age effect was observed in males. WBM was higher in pre- and perimenopausal women compared to postmenopausal women. Across all female participants, follicle stimulating hormone (FSH) was negatively associated with WBM, but positively within perimenopause. DMN modularity was positively related to estradiol in premenopause and negatively to age in females aged 36-45 and males aged 51-55.

Conclusions:

The decline in WBM in females suggests reduced large-scale network segregation, which may reflect accelerated age-related network dedifferentiation. The nonlinear FSH-WBM association in women might indicate dynamic network reconfiguration during perimenopause. Sex-specific age effects on DMN modularity highlights menopause as a critical window of network changes that may predispose to later AD risk. Distinct associations of FSH with WBM and estradiol with DMN modularity suggest that different hormones may influence separate network systems and confer risk to AD through independent mechanisms.