Could menopause be an early window for protecting the brain against Alzheimer's?

Alzheimer’s disease (AD) is a progressive neurodegenerative disorder that affects memory, thinking and other cognitive functions. It is the most common cause of dementia and disproportionately affects women. Women generally live longer than men, and increasing age is the strongest known risk factor for dementia.

Women also undergo substantial hormonal changes across their reproductive lifespan. During the menopausal transition, ovarian function changes and circulating levels of reproductive hormones, particularly estradiol, decline. Menopause is associated with a range of symptoms, and some women experience temporary changes in cognitive performance. Although many of these changes may resolve or stabilise after the menopausal transition, menopause may also represent a period of increased vulnerability to biological changes associated with later-life brain ageing. This possibility is supported by evidence that earlier and surgically induced menopause are associated with increased risk of Alzheimer’s disease.

This raises an important question: could identifying molecular changes that occur during the menopausal transition provide clues about biological pathways that are relevant to the development of Alzheimer’s disease later in life?

In a fascinating study published in Nature Medicine, researchers investigated proteomic changes associated with menopause in healthy midlife women and examined whether these changes were related to brain aging and cognitive outcomes later in life.

The discovery cohort consisted of 80 women from the spontaneous menopause cohort (UCSB-HAS), including premenopausal, perimenopausal and postmenopausal women, together with 36 age-matched men. The women were rigorously classified according to the STRAW+10 reproductive staging system and were largely free from exogenous hormone exposure. Serum samples were analysed using the NULISAseq CNS panel, while reproductive hormones were measured using liquid chromatography–mass spectrometry (LC-MS) and immunoassays.

NULISAseq CNS is an ultrasensitive, multiplexed proteomic platform that allows many proteins to be quantified simultaneously from a small volume of biological sample. In this study, 118 inflammatory and neurospecific proteins were measured. The researchers then compared protein levels between premenopausal and postmenopausal women, adjusting for age.

Sixteen proteins were significantly higher in postmenopausal women. These menopause-associated proteins could broadly be grouped into several biological categories:

• Inflammatory: CCL2, CCL11, CCL13, CCL26, CXCL1, IL-12p70, TEK and CD63
• Synaptic/neuronal: ACHE, SNAP25, CNTN2 and BDNF
• Metabolic: IGFBP7 and IGF1R
• Alzheimer’s disease-related: BACE1 and p-tau231

Developing a menopause proteomic score

The researchers then used principal component analysis (PCA) to summarise the combined pattern of these 16 proteins. The first principal component (PC1), which captured the largest proportion of the overall variation in the protein measurements, was used to construct a menopause proteomic score. PC1 explained 28.7% of the variance across the 16 proteins.

A higher score represented a more postmenopausal-like protein profile. The score increased progressively across the menopausal transition, from premenopause to perimenopause and then postmenopause.

Interestingly, the proteomic pattern appeared to be more strongly related to menopause stage than to chronological age. Although the menopause proteomic score was correlated with age before adjustment, the association with age was no longer significant after accounting for menopause stage. This suggests that the biological changes captured by the score were not simply a reflection of getting older.

The researchers also examined menopausal symptoms. Night sweats were associated with a higher menopause proteomic score and with higher levels of several inflammatory proteins, including CCL13, CXCL1, CCL26 and CCL2. In contrast, the overall menopause proteomic score was not significantly associated with hot flashes.

Relationship with reproductive hormones

The researchers next investigated how the menopause-associated protein pattern related to reproductive hormones.

The menopause proteomic score was inversely associated with estradiol and positively associated with FSH. When reproductive hormones and age were considered together, FSH remained significantly associated with the score, whereas age did not.

These findings suggest that the proteomic changes identified in the study were closely related to the hormonal changes accompanying the menopausal transition. However, because the study was observational, these associations do not establish that changes in FSH or estradiol directly cause the observed protein changes.

Validation in the UK Biobank

The researchers then tested whether the menopause-associated proteomic signature could be reproduced in a much larger independent cohort: the UK Biobank.

The UK Biobank analysis included 2,814 women and used a different proteomic technology, the Olink proximity extension assay, to measure approximately 2,923 proteins. Overall, 1,300 proteins showed differential abundance according to menopause status after correction for multiple testing (FDR < 0.05), with most showing higher levels in postmenopausal women.

The broader analysis identified changes in pathways involving cytokine signalling, complement activation, extracellular vesicles and catabolic processes.

Of the menopause-associated proteins identified in the UCSB-HAS discovery cohort, 13 could be compared between the two platforms and 9 showed clear replication in UK Biobank. This cross-platform replication provides evidence that at least part of the menopause-associated proteomic signature was reproducible in an independent population.

The UK Biobank data also provided evidence that postmenopausal women had older proteomic estimates across multiple organs and cell types. Postmenopausal women had older proteomic ages for 11 of 13 organs, including the brain, arteries and intestine, and for 36 of 38 cell types, including oligodendrocyte precursor cells.

Are these proteomic changes relevant to cognitive aging?

The researchers then asked whether the menopause proteomic signature identified in midlife was associated with cognitive outcomes later in life.

They examined data from four independent ageing cohorts:

• Alzheimer’s Disease Neuroimaging Initiative (ADNI): a longitudinal study incorporating neuroimaging, biomarkers and clinical assessments to study cognitive decline and Alzheimer’s disease progression.
• BrANCH: a longitudinal community-based study involving repeated neuropsychological assessments.
• WRAP: a cohort with cognitive assessments and information on symptoms and family history, enriched for individuals with a parental history of Alzheimer’s disease.
• UK Biobank: a large prospective cohort used to examine incident dementia over approximately 15.7 years.

Across these cohorts, a higher menopause proteomic score was associated with steeper cognitive decline in ADNI and BrANCH, and with lower global cognition in WRAP.

In the UK Biobank, a higher menopause proteomic score was also associated with a modestly higher risk of subsequently developing Alzheimer’s disease dementia. However, the score was not significantly associated with frontotemporal dementia, vascular dementia or all-cause dementia.

These findings therefore suggest that the biological changes captured by the menopause proteomic score may be relevant to pathways involved in cognitive aging and Alzheimer’s disease later in life.

Importantly, the study does not demonstrate that menopause causes Alzheimer’s disease, nor does a higher menopause proteomic score establish that an individual will develop dementia. The associations are observational and may reflect a combination of biological, hormonal and other factors.

Overall, the study provides evidence that the menopausal transition is accompanied by coordinated changes in the circulating proteome involving inflammation, metabolism, synaptic and neuronal biology, and Alzheimer’s disease-related pathways. The findings raise the possibility that menopause represents an important biological window during which changes in systemic physiology may intersect with processes involved in later-life brain ageing.

The menopause proteomic score therefore represents a potentially useful research biomarker of the biological changes associated with menopause, rather than a diagnostic test for menopause or a clinical predictor of Alzheimer’s disease. Longitudinal studies following women through the menopausal transition, with repeated measurements of hormones, proteins, brain changes and cognition, will be important for determining whether these proteomic changes are merely associated with later cognitive decline or whether they participate mechanistically in the pathway linking menopause to brain ageing and dementia.

Reference

Wood Alexander, M., Rabin, J.S., Caunca, M. et al. Blood proteomics of menopause map to brain aging and dementia risk. Nat Med (2026). doi: https://doi.org/10.1038/s41591-026-04648-4