A small midlife study identified protein changes linked to menopause, while four older women’s cohorts connected the pattern with later cognitive health.
A set of blood protein changes associated with menopause was also linked to cognitive health decades later in older women. The findings suggest that midlife biology may help researchers study later dementia risk, but they don’t show that menopause causes dementia or that the proteins can predict an individual’s future health.
The study, published in Nature Medicine, first compared serum proteins in 80 healthy women aged 43 to 58: 30 were premenopausal, 26 perimenopausal and 24 postmenopausal. The researchers also included 36 men of similar ages for comparison. They identified 16 proteins that were more abundant in postmenopausal than premenopausal women after accounting for age. The proteins were connected to inflammation, synaptic function, metabolism and Alzheimer’s disease processes.
A protein pattern linked to menopause stage
To check whether the initial results held up beyond this small group, the team examined data from nearly 3,000 women in the UK Biobank. Thirteen of the 16 proteins had been measured in both studies; nine showed similar menopause-related differences in the larger cohort. That replication adds support to the pattern, though it doesn’t establish what the changes mean for an individual’s brain or health.
The researchers combined the 16 proteins into a “menopause proteomic score,” intended to capture their shared pattern while reducing the influence of variation in individual measurements. Compared with premenopausal status, perimenopausal status was associated with a 1.36-unit higher score, and postmenopausal status with a 3.25-unit higher score. The analysis indicated that menopause stage, rather than chronological age, was driving these differences.
Inflammation-related proteins made up much of the pattern, in line with earlier research reporting increased systemic inflammation after menopause. Other parts of the signature involved synaptic and Alzheimer’s-related markers. These observations fit with findings from animal studies of changes in synaptic plasticity and the density of connections between brain cells, as well as human brain imaging research reporting menopause-related differences in amyloid-beta and tau markers. Those earlier findings provide context; they don’t prove that the blood proteins cause brain changes.
Hormones and symptoms showed less consistent links
The protein score tracked with some, but not all, reproductive hormone measures. Higher follicle-stimulating hormone (FSH) levels were associated with a higher score, while higher estradiol levels were associated with a lower score. The researchers found no association with progesterone. Looking at individual proteins, some inflammation-related markers were inversely associated with estradiol, while several proteins across the study’s biological categories were positively associated with FSH.
The authors suggested that the hormones may relate to different biological pathways, since the proteins fell into two clusters in their analysis. Prior research has proposed that estrogen can influence immune responses. FSH receptors have also been reported in brain regions vulnerable to Alzheimer’s disease, and experiments in animals have linked FSH to Alzheimer’s-related changes, brain inflammation and cognitive problems. These lines of evidence help frame questions for future research; they don’t show that changing hormone levels would prevent dementia in people.
The study also examined vasomotor symptoms, including hot flashes and night sweats. Earlier research has associated some of these symptoms with vascular inflammation, cardiovascular risk and Alzheimer’s-related brain changes. In the current analysis, results differed between cohorts, although some suggested an association between certain symptoms and the menopause score. That inconsistency makes the symptom findings difficult to interpret as a clear relationship.
What the later-life results suggest
To explore whether the midlife protein pattern was relevant over the longer term, the researchers analyzed four separate cohorts of older women. In these groups, higher baseline menopause scores—scores that more closely resembled the postmenopausal pattern—were consistently associated with poorer later cognitive health and a higher risk of Alzheimer’s disease dementia.
This is a meaningful link across stages of life, but it needs careful interpretation. The results show an association, not that menopause directly causes dementia. Nor do they establish that the score can forecast a particular person’s outcome: that would require further study of how well it predicts risk, and whether it adds useful information beyond other health factors.
Study co-senior author Kaitlin Casaletto, an associate professor at the UCSF Fein Memory and Aging Center, said the team doesn’t think menopause directly causes dementia. She described the possibility that molecules measured around menopause could help researchers estimate later risk. Co-senior author Rowan Saloner, also at the center, characterized the proteins as potential markers of both menopause and cognitive aging. Those are research interpretations, not validated clinical uses.
Madeline Wood Alexander, a University of Toronto graduate student, framed menopause as a normal life stage and a chance to better understand the biology of brain aging in women. Casaletto also described midlife as a period when brain resilience may be shaped. The study supports further investigation of that window, while leaving open whether these protein changes can guide prevention or care.
What readers can take from the findings
For now, the work is best understood as an early step toward mapping biological changes across menopause and studying how they relate to later cognition. The initial protein analysis involved a modest sample, and the larger cohort replicated only nine of the 13 proteins measured in both datasets. The later-life findings strengthen the case for follow-up research, but they don’t establish cause and effect or provide a personal risk test.
Menopause-related cognitive symptoms, such as attention or word-finding difficulties, have been reported in previous research, but this protein study does not explain those experiences or assess how to treat them. Anyone concerned about changes in memory or thinking, bothersome menopausal symptoms, or questions about hormone-related care can discuss them with a doctor or pharmacist. Future studies will need to test whether the protein pattern is reliable in broader groups and whether it offers information that can meaningfully support clinical care.
