Age, Not Menopause Stage, Drives Cortical Volume Reduction in Midlife Women, New Study Suggests

The working hypothesis that menopause accelerates brain volume loss beyond normal aging is being challenged by research presented this week at The Menopause Society 2025 Annual Meeting, October 21-25, in Orlando.

Coauthor Katrina Wugalter, a PhD candidate in the department of psychology at the University of Illinois Chicago, and colleagues analyzed cortical volumes in 229 cognitively normal midlife women from the Human Connectome Project in Aging and found no significant differences between premenopausal, perimenopausal, and postmenopausal participants after controlling for age.

They did find, however, that age itself strongly predicted cortical volume reductions across 121 brain regions spanning multiple cognitive networks.

Current Knowledge Gaps

Despite growing interest in the effects of menopause stage on the brain, longitudinal neuroimaging studies examining cortical changes across menopause remain scarce. Existing cross-sectional studies examining brain outcomes, eg volumes and glucose metabolism, by menopause stage face substantial methodological limitations, according to Wugalter et al. Small sample sizes (n = 12–15 per stage in some studies), inclusion of menopausal hormone therapy users, confounding of stage with Alzheimer disease (AD) risk factors, use of non-standard menopause stage definitions, and samples with higher prevalence of genetic AD risk factors than found in the general population generally have complicated interpretation of stage effects.

Wugalter and colleagues designed this study to better characterize potential effects of menopause on brain structure during midlife using a larger cohort with more rigorous methodology.

Study Design and Population

The research team examined whether cortical volumes differed between premenopausal, perimenopausal, and postmenopausal women in 229 cognitively normal midlife women enrolled in the Human Connectome Project in Aging (HCP-A), a study of typical aging across the lifespan. The sample included women aged 40–60 years (baseline mean age 49.7±5.8 years) who were not surgically menopausal or menopausal hormone therapy users. The cohort was racially diverse (56% White, 23% Black) with 24% APOE4 carriers. Of these, 108 returned for a second study visit 1 to 2 years later.

Researchers used the Glasser atlas to define cortical regions (180 per hemisphere) and processed them using the HCP Minimal Processing Pipeline. Women were categorized as premenopausal (n=87), perimenopausal (n=69), or postmenopausal (n=73) by STRAW+10 criteria, the standardized staging system for reproductive aging.

While too few HCP-A participants transitioned across menopause stages throughout the study duration to perform an ideal longitudinal analysis, the research team used a mixed effects model to examine group differences, leveraging data from multiple timepoints. The model included a random intercept of participant and controlled for age, race (White versus non-White), education, study visit, and estimated intracranial volume. A false discovery rate correction was applied to all models to control for multiple comparisons.

Findings: Menopause Stage Doesn't Matter

At baseline, perimenopausal women were significantly older than premenopausal women (b = 4.23, P <.001) and significantly younger than postmenopausal women (b = -6.44, P <.001). Compared to the pre- and perimenopausal women, postmenopausal women had significantly less education (Pre vs Post, b = -1.15, P <.001; Peri vs Post, b = -0.87, P <.05) and had a higher proportion of White participants to non-White participants (Χ2 =9.77, P <.01).

The critical finding emerged from the menopause stage analysis: after controlling for age and other covariates, cortical volumes did not significantly differ by menopause stage across any brain region ( pFDR >.05). However, results revealed that volumes of 121 cortical regions spanning multiple cognitive networks were significantly negatively associated with age ( pFDR <.05). These age-related reductions included 26 regions within the default mode network—a brain system critical for memory and self-referential processing that shows particular vulnerability in AD.

Contributions to the Literature

This study addresses several limitations that have hampered previous investigations. The sample size of 229 women substantially exceeds most prior studies examining menopause stage and brain structure. Excluding surgical menopause and menopausal hormone therapy users eliminates two major confounds. Using STRAW+10 criteria provides standardized menopause staging. The racially diverse sample with APOE4 carrier prevalence (24%) representative of the general US population enhances generalizability. The HCP-A's rigorous neuroimaging protocols and validated processing pipelines ensure high-quality brain measurements across 360 cortical regions.

The authors emphasize that in this large sample of cognitively normal midlife women, cortical volumes did not significantly differ by menopause stage, but many declined with age. "These findings, using a large sample and multiple visits, counter previous reports that menopause accelerates the loss of brain volume at midlife beyond normal aging. A more definitive conclusion awaits longitudinal findings," they concluded.

Source: Wugalter K, Wood SC, van doorn J, et al. Reductions in cortical volumes associate with age, not with menopause stage in midlife women. Abstract presented at: The Menopause Society’s 2025 Annual Meeting; October 21-25, 2025; Orlando, FL. Accessed October 21, 2025.