Researchers traced menopause-associated proteins across hormones, symptoms, and biological aging measures before examining what the same molecular profile could mean for cognitive health decades later.
Study: Blood proteomics of menopause map to brain aging and dementia risk. Image Credit: Joel / Shutterstock
A recent study published in the journal Nature Medicine found that blood protein patterns associated with menopause were linked to proteomic estimates of brain aging and later cognitive outcomes. While not indicating causality, these findings warrant longitudinal studies to validate them and examine their clinical relevance.
Background
Menopause is increasingly viewed as an important transition in the aging process. It involves a marked decline in ovarian hormones, particularly estradiol and progesterone, together with a sustained rise in follicle-stimulating hormone (FSH). These hormonal changes are associated with effects across multiple body systems, including the brain.
Cognitive symptoms such as word-finding and attention difficulties are commonly reported during menopause, and observational studies suggest that measurable cognitive changes are generally subtle and within the range expected with normal aging.
Earlier and surgical menopause have been identified as risk factors for AD, while vasomotor symptoms have been linked to poorer memory, white matter hyperintensities, and AD biomarkers. Yet the biological pathways linking menopause to brain aging remain poorly understood.
The study investigated blood-based molecular differences associated with natural menopause in women aged 43-58 years, using rigorous menopausal staging. The aim was to identify whether these differences were associated with cognitive aging and dementia risk.
Study characteristics
The researchers first studied 80 women, 30 premenopausal, 26 perimenopausal, and 24 postmenopausal, along with 36 age-matched men. Serum samples were analyzed using an ultrasensitive nucleic acid-linked immuno-sandwich assay (NULISAseq) proteomic panel measuring 118 proteins relevant to the central nervous system.
Menopause proteomic score
After adjusting for age, 16 proteins were higher in postmenopausal than premenopausal women. These represented inflammatory proteins, including CCL13, IL-12p70, and CXCL1; synaptic and neuronal proteins such as CNTN2 and BDNF; metabolic proteins including IGFBP7 and IGF1R; and AD-related proteins including BACE1 and phosphorylated tau 231 (p-tau231). The researchers combined these 16 proteins into a composite “menopause proteomic score” using principal component analysis. Higher scores represented a more postmenopausal-like protein profile.
The score increased progressively from premenopause through perimenopause to postmenopause. The score was initially associated with age, but this relationship disappeared after accounting for menopause stage, while the association with menopause stage remained.
By comparison, age was not associated with the score in men. These findings suggested that the proteomic differences were more strongly associated with the stage of menopause than with chronological age. Among the individual AD biomarkers examined, p-tau231 was the only canonical AD biomarker that differed significantly between postmenopausal and premenopausal women after age adjustment. The clinical relevance of midlife elevations in p-tau231 and BACE1 remains unclear.
Men had higher scores than women after adjusting for age, when the same 16-protein composite was applied to both groups.
Menopause proteomic score and hormones
The researchers next examined relationships with reproductive hormones. Estradiol, progesterone, and FSH differed significantly across menopause stages, whereas testosterone, sex hormone-binding globulin, and DHEAS did not. After adjusting for age, the menopause proteomic score was inversely associated with estradiol and positively associated with FSH. When all three hormones were considered simultaneously, only FSH remained significantly associated with the proteomic score.
Compared with age, the menopause score was more closely related to hormone levels when considered together. Several individual proteins showed hormone-specific relationships: some inflammatory proteins tracked with lower estradiol, others with higher FSH.
Vasomotor symptoms and proteomics
Vasomotor symptoms were associated with the greatest increases in inflammatory menopause proteins. Among perimenopausal and postmenopausal women, night sweats were associated with a higher menopause score. Hot flashes showed no significant association with the score. Vaginal dryness was associated with lower scores, though the biological implications remain unknown.
Replication in a large independent cohort
The findings were tested in a much larger UK Biobank validation cohort. This included 2,814 age-matched women aged 45–60 years, split equally between a combined pre- and perimenopausal group and a postmenopausal group. The researchers found that 1,300 of the 2,923 Olink proteins measured differed significantly across menopause stages.
That is, 1,146 proteins were higher and 154 lower in postmenopausal women than in pre-/perimenopausal women. Of the 13 menopause proteins directly compared across both proteomics platforms, nine clearly replicated across cohorts, with similar patterns and strong overlap among the largest effects.
Proteins involved in growth-factor and reproductive signaling were reduced, while those involved in cytokine signaling, complement activation and extracellular vesicle-related processes were increased.
Menopause and aging
Compared with pre- or peri-menopausal women of the same age, postmenopausal women had larger age gaps on proteomic estimates of organ and cell aging. Less favorable estimates were found across 11 of 13 organ measures and 36 of 38 cell measures. Brain and artery aging were among the prominent organ-level associations, alongside intestine aging, while top cell-level measures included female reproductive cells and oligodendrocyte precursor cells.
Later-life menopause protein levels and menopause symptoms
To assess whether menopause-associated proteins remained relevant years after menopause, the researchers looked at menopause scores in later life. In 89 women with an average age of about 69 years, higher menopause scores were associated with a history of hot flashes, sleep problems, and a greater total number of menopause symptoms. CCL2 was the only individual protein associated with vasomotor symptoms in both the midlife and older groups, suggesting that this relationship may persist into later life.
Menopause scores and cognitive risk
The menopause score was calculated in four independent cohorts of older women, whose mean baseline ages ranged from 60.7 to 72.1 years. In ADNI and BrANCH, higher scores were associated with less favorable cognitive change over time; in WRAP, they were associated with lower global cognitive performance in a cross-sectional analysis. In UK Biobank, higher scores were associated with greater risk of incident AD dementia, but not with all-cause, vascular or frontotemporal dementia.
Sensitivity analyses were conducted by restricting the analysis to women with age, adjusting for cardiometabolic conditions, excluding women with bilateral oophorectomy, stratifying by hormone-therapy history, removing p-tau231 from the score, and excluding women with baseline mild cognitive impairment or dementia. These analyses generally supported the main findings.
An exploratory analysis found that APOE4 carriage significantly strengthened the association between the menopause proteomic score and cognitive change in the ADNI cohort, but most tests of effect modification were null.
Limitations
The study has important limitations. The discovery cohort was small and cross-sectional, symptoms were self-reported, and proteins measured in blood may originate from multiple tissues.
The composite score was based on only about 120 proteins in the NULISAseq CNS panel, so it does not capture the full range of circulating biology related to menopause. Three proteins identified in the discovery cohort did not replicate in the UK Biobank, potentially due to differences between proteomic platforms or in menopause classification.
The authors state that longitudinal studies following individuals through menopause, measuring hormones, proteins, symptoms, brain health, and cognition, are needed to establish how these factors change over time.
Conclusion
Overall, the study identified reproducible blood protein differences associated with menopause, particularly involving inflammation, synaptic plasticity, metabolism, and AD-related pathways. The menopause score was more closely related to reproductive hormone levels than chronological age and was associated with some vasomotor symptoms. Higher scores were associated with poorer cognitive outcomes in later-life cohorts.
The findings support menopause as a potentially important period for studying early biological signals relevant to later brain health and cognitive aging. But the observational nature of the evidence means it does not prove that menopause-related proteomic changes cause cognitive decline or dementia.
Core muscle activity rapidly regulates blood flow in the brain