Earlier menopause linked to accelerated brain aging decades later

The age at which menopause occurs can leave a lasting imprint on brain health that becomes visible decades later. In a large longitudinal study of older women, earlier menopause was associated with more rapid cognitive decline, earlier clinical diagnosis of Alzheimer's disease, and adverse changes in brain structure. The study was published on August 25, 2026, in JAMA Network Open.

The most striking neuroimaging finding was the accelerated accumulation of white matter hyperintensities—changes visible on magnetic resonance imaging (MRI) that typically reflect damage to small vessels and surrounding neural tissue. In women with natural menopause, the model showed that menopause occurring five years earlier was associated with approximately a 15% greater accumulation of such changes over the subsequent ten years of life.

However, the study did not demonstrate a similarly compelling increase in classic Alzheimer's pathology in postmortem brain examinations. A link with overall brain volume existed, but it was weak, while the strongest structural signal was associated specifically with white matter damage. This suggests that one of the pathways linking earlier ovarian failure with late cognitive aging may involve brain vascular health, and not just amyloid and tau protein accumulation.

Importantly, this is an observational study. It demonstrates a consistent association, but does not prove that earlier menopause per se directly causes dementia. The authors view age at menopause primarily as a marker of future neurological risk, available already in midlife, which could potentially help identify women who require particularly careful prevention of vascular and cognitive impairment.

Basic data	Result
Women with cognitive abilities	2603
Average age at inclusion	78.3 years
Natural menopause	1,753 women
Surgical menopause	850 women
Median age at menopause in general	49 years old
Natural menopause	50 years
Surgical menopause	43 years old
Serial MRI	774 women
APOE ε4 carriers	22.3%
Median observation	7 years
Earlier menopause	more rapid global cognitive decline
Episodic memory	decreased faster
Alzheimer's disease	was diagnosed earlier
The most powerful MRI result	accumulation of white matter foci
Menopause 5 years earlier	≈15% more such changes per decade
Hormonal therapy	no convincing change in MRI connectivity was detected
DOI	10.1001/jamanetworkopen.2026.30973

Why the age of menopause may matter for the brain

Menopause isn't just the cessation of menstruation. It's accompanied by long-term changes in the levels of estrogen, progesterone, and other sex hormones, which have effects far beyond the reproductive system. Estrogens are involved in the regulation of synaptic plasticity, myelination of nerve fibers, mitochondrial function, and neuroimmune processes. Therefore, the timing of the end of cyclical ovarian hormone production could theoretically influence the subsequent resilience of the brain to aging.

Previous epidemiological studies have linked earlier menopause and a shorter reproductive lifespan to an increased risk of cognitive decline and dementia. However, most neuroimaging studies have assessed the brain only once or examined women directly in midlife. This left an important question unanswered: whether the association with age at menopause persists 20–30 years later and whether it influences the rate of subsequent brain change.

The new study is unique in that the researchers used repeated cognitive assessments, serial MRI scans, and, for some participants, postmortem brain imaging. This allowed them to simultaneously examine three levels: how the woman performed on cognitive tests, how her brain structure changed, and whether typical Alzheimer's disease pathology was detected after death.

This comprehensive approach is particularly important because cognitive aging is rarely explained by a single process. An elderly person may simultaneously exhibit small vessel disease, atrophy, Alzheimer's changes, microscopic infarcts, and other lesions. Therefore, the correlation of several independent indicators provides a more meaningful picture than a single memory test or MRI.

The study observed women up to 18 years of age.

The analysis utilized two well-known American aging cohorts: the Religious Orders Study and the Rush Memory and Aging Project. The former includes members of religious orders, while the latter consists of elderly residents of local communities. Participants were required to be free of known dementia upon enrollment and subsequently underwent annual clinical and neuropsychological assessments. Both programs also require consent for postmortem brain imaging.

The main cognitive analysis included 2,603 women, with an average age of 78.3 years at enrollment. Approximately 67% reported natural menopause, and approximately one-third reported surgical menopause. The median age at menstrual cessation was 50 years after natural menopause and 43 years after surgical menopause.

774 participants had repeat 3-Tesla brain MRIs. The scans were performed approximately every two years, and the researchers calculated gray and white matter volume, total brain volume, cerebrospinal fluid volume, and the volume of white matter hyperintensities. Postmortem autopsy data were also available for some women.

Statistical models accounted for age, education, race, body mass index, smoking, use of menopausal hormone therapy, and the presence of the APOE ε4 variant, one of the strongest known genetic risk factors for late-onset Alzheimer's disease. Statistical weighting was additionally applied for MRI, as women who survived to the neuroimaging stage and consented to it were, on average, cognitively healthier than the entire original cohort.

Cohort characteristics	Meaning
Total women	2603
Average age of inclusion	78.3 ± 7.9 years
Median observation	7 years
Maximum observation	under 18 years old
Natural menopause	1753
Surgical menopause	850
Average body mass index	27.4 kg/m²
APOE ε4 carriers	475 - 22.3%
Women with serial MRI	774
Hormonal therapy was used around menopause	18.8%
Average reproductive period	34.1 years

Earlier menopause was linked to faster memory decline.

Over the course of the observation period, the women completed a battery of 19 neuropsychological tests annually. These were used to calculate a general cognitive score and individual assessments of episodic memory, semantic memory, working memory, perceptual speed, and visuospatial abilities. This allowed us to study the rate of change rather than comparing two groups at a single age point.

Each additional year of earlier menopause was associated with a small but statistically significant acceleration of decline in the overall cognitive score. The coefficient was -0.0009 standard deviations per year of earlier menopause, and the result remained statistically significant after adjusting for multiple comparisons.

A more pronounced association was found with episodic memory—the ability to remember and recall events and new information. Here, the coefficient was -0.0014 standard deviations for each year of earlier menopause. For semantic and working memory, the direction of the effect was similar, but after statistical correction, the results no longer met the significance criterion. No convincing association was found for perceptual speed and visuospatial functions.

Therefore, the study does not show a sharp cognitive decline after early menopause. The effect of a single year is very small. Its significance lies in the fact that small differences in rates can accumulate over decades, especially when menopause occurs 5-10 years earlier than average. This long-term cumulative effect is consistent with the overall logic of the study.

What cognitive functions were associated with age at menopause?

Indicator	Result
General cognitive function	faster decline
Episodic memory	faster decline
Semantic memory	the direction is similar, but after correction it is insignificant
Working memory	the direction is similar, but after correction it is insignificant
Speed of perception	there is no significant relationship
Visual-spatial abilities	there is no significant relationship

Alzheimer's disease was diagnosed somewhat earlier

The authors assessed not only test results but also the time to clinical diagnosis of Alzheimer's disease. In the full analysis, earlier menopause was associated with earlier achievement of the diagnostic threshold for Alzheimer's disease. After adjusting for multiple comparisons, the association remained statistically significant.

The effect size was small for each individual year. The authors calculated that the time-to-diagnosis ratio was 0.998 for each year of earlier menopause. However, when the 10-year difference is accumulated, the effect becomes more significant: according to the researchers, menopause 10 years earlier corresponded to approximately a 2% earlier onset of clinical Alzheimer's disease.

With the average age at Alzheimer's diagnosis in this cohort being approximately 88.4 years, this difference corresponded to approximately 1.8 fewer years of life without an Alzheimer's diagnosis for a woman who experienced menopause ten years earlier. This is a model estimate, not a guarantee of an individual prognosis: a specific woman who experiences menopause at age 40 will not necessarily develop the disease 1.8 years earlier than a woman who experiences menopause at age 50.

Interestingly, the standard analysis of the probability of being diagnosed with Alzheimer's disease by the last visit did not reach statistical significance across the entire cohort. This highlights the difference between two questions: whether the disease will develop at all and when it will be diagnosed. The new study provides more compelling data on the timing of clinical manifestations, rather than on the absolute probability of developing the disease.

Alzheimer's disease rate	Result
Diagnosis at last visit, entire cohort	there is no statistically convincing connection
Time to diagnosis	associated with earlier menopause
Time ratio 1 year earlier	0.998
Menopause 10 years earlier	≈2% acceleration of diagnosis
Translation of the model at the average age of diagnosis	≈1.8 less years without Alzheimer's disease

The strongest result was found not in the gray matter, but in the white matter.

MRI allowed us to determine which structural brain changes were most closely related to the timing of menopause. Overall brain volume did indeed decrease somewhat more rapidly with earlier menstrual cessation, and the result remained statistically significant after adjustment. However, the effect size was very small: f² = 0.013, even below the traditional threshold used to denote a small effect.

For individual gray matter, white matter, and cerebrospinal fluid volumes, the direction of differences was generally consistent with faster aging associated with earlier menopause, but the results did not survive correction for multiple comparisons. Similarly, exploratory signals in the anterior cingulate, entorhinal, and frontal cortex did not retain statistical significance after such correction.

A completely different picture emerged for white matter hyperintensities in women with natural menopause. Here, the effect size reached f²=0.30 with an adjusted p<0.001—this was the largest structural difference among all neuroimaging parameters in the study. The earlier menstruation ceased naturally, the more rapidly these lesions increased in size in the following decades.

Moreover, the difference became more pronounced with age. This means the association did not appear to be a one-time structural shift occurring at menopause. Rather, the data supported a scenario in which early termination of the reproductive period is associated with a more vulnerable trajectory of white matter vascular aging, which becomes increasingly pronounced as a woman ages.

What do white matter hyperintense lesions mean?

In certain MRI modes, damaged areas of white matter appear brighter, hence the term hyperintensity. These lesions are very common with age. They are largely associated with small cerebral vascular disease, chronic poor blood flow, disruption of the vascular barrier, and damage to myelin and axons.

A small amount of such changes may remain asymptomatic for years. However, when significantly accumulated, they are associated with a higher risk of cognitive impairment, problems with executive functions and walking, as well as stroke and dementia. Therefore, researchers have used them as one of the macrostructural indicators of vascular brain damage.

In the analyzed cohort, for women with natural menopause at approximately age 50, the model estimated an increase in the volume of such lesions at approximately age 70 of approximately 9.9% of the baseline volume per year. This does not represent a loss of 9.9% of all white matter; rather, it represents an increase in the already existing, relatively small volume of hyperintense lesions.

The difference between women with earlier and later menopause reached approximately 3.4 percentage points of additional growth per year in certain periods. When extending the model to a ten-year interval, the authors calculated that menopause five years earlier corresponded to approximately a 15% greater accumulated volume of white matter hyperintensities.

The most important MRI findings

Indicator	Association with early menopause
Total brain volume	accelerated reduction, but very little effect
Gray matter	there is no convincing independent connection
White matter in general	there is no convincing independent connection
Cerebrospinal fluid	there is no convincing independent connection
Regional crust	individual signals, but did not withstand the correction
White matter hyperintensities	strong association with natural menopause
Effect size for these foci	f²=0.30
Menopause 5 years earlier	≈15% more savings over 10 years

Why the results are especially pronounced after natural menopause

The researchers analyzed natural and surgical menopause separately. At first glance, one might expect a stronger structural effect after surgical removal of ovarian function, as hormone levels decline much more rapidly. However, the most pronounced accumulation of white matter hyperintense lesions was found among women with natural menopause.

In surgical menopause, there was no convincing association between the age of surgery and the accumulation of these MRI changes. The authors suggest that a gradual, natural decline in ovarian function and an abrupt surgical cessation of hormonal production may have different effects on the brain. In natural menopause, hormonal changes are spread out over time, and residual ovarian function persists longer.

On the other hand, cognitive and clinical associations appeared stronger in some analyses after surgical menopause. For example, before adjusting for multiple comparisons, each additional year of earlier surgical menopause corresponded to a roughly 3% higher odds of Alzheimer's disease at last follow-up. However, after strict adjustment, many subgroup results were no longer statistically significant.

There is also a methodological explanation. There were fewer women with surgical menopause in the MRI subgroup, and the most vulnerable of them may not have survived to the late neuroimaging stage or may not have been healthy enough to participate. Therefore, the lack of a strong MRI association after surgical menopause cannot be considered evidence that the surgery is safer for the brain than early natural menopause.

Postmortem examination showed no more classic Alzheimer's pathology

A detailed postmortem examination of the brain was performed on a large proportion of the deceased participants. Neuropathologists, blinded to the women's clinical data, assessed the prevalence of amyloid plaques and neurofibrillary changes and assigned a standardized grade of Alzheimer's disease.

In the overall cohort, age at menopause was not significantly associated with the severity of Alzheimer's changes after death. For surgical menopause, a certain signal was observed before adjustment for multiple comparisons, but after adjustment, it was no longer statistically significant. For natural menopause, there was virtually no association.

This is one of the most interesting findings of the entire study. Women with earlier menopause demonstrated a faster decline in cognitive function and an earlier clinical diagnosis, but the researchers did not observe a proportional increase in classic Alzheimer's pathology. Therefore, the association may not be entirely due to amyloid and tau proteins.

One alternative explanation is vascular damage. White matter is highly sensitive to the condition of small vessels, and the menopausal transition is accompanied by changes in lipid metabolism, blood pressure, glucose metabolism, and vascular function. The combination of moderate atrophy with a marked accumulation of white matter hyperintense foci is consistent with the model of additional vascular contribution, although the new study does not directly prove this mechanism.

Indicator	Result
Clinical cognitive decline	there is a connection
Earlier diagnosis of Alzheimer's disease	there is a connection
Volume of white matter hyperintensities	expressed connection
General brain atrophy	a small connection
Classic Alzheimer's pathology at autopsy	there is no convincing connection
Possible explanation	contribution of vascular and other non-amyloid damage

A shorter reproductive period gave a similar picture

In addition to age at menopause, the authors calculated the reproductive period—the period between the onset of menstruation and menopause. On average, it was approximately 34 years: approximately 35.9 years for natural menopause and 30.4 years for surgical menopause. Age at menopause and the length of the reproductive period were closely related.

A shorter reproductive period generally showed the same pattern of results as earlier menopause. It was associated with an earlier diagnosis of Alzheimer's disease and a more rapid decline in general cognitive function, episodic memory, semantic memory, and working memory.

The association with white matter changes associated with natural menopause was again particularly strong. For reproductive age, the effect size reached f²=0.37, which is even larger than when using menopause age alone.

However, interpreting the reproductive period as a precise number of years of "estrogen exposure" is not possible. The hormonal environment changes radically during pregnancy, breastfeeding, contraceptive use, various illnesses, and perimenopause. The study lacked systematic data on pregnancy, childbirth, breastfeeding, and oral contraception, which the authors explicitly cite as a limitation.

Did menopausal hormone therapy protect the brain?

This question is particularly important because if earlier cessation of ovarian function is indeed associated with an unfavorable brain trajectory, an obvious hypothesis arises: could timely initiation of hormone therapy offset some of this risk? The authors specifically conducted an additional analysis taking into account systemic estrogen therapy initiated approximately five years before or after menopause.

In the main cohort, approximately 18.8% of women used this therapy. It was more common after surgical menopause, which is logical, as the abrupt cessation of ovarian function was particularly often accompanied by hormonal treatment.

However, the study found no statistically significant change in the association between age at menopause and MRI trajectories depending on hormone therapy. No significant interactions were found for total brain volume or other volumetric measures.

This does not mean that hormone therapy does not protect the brain. The authors emphasize that the proportion of women treated according to modern guidelines and within a suitable time window was small. Furthermore, the study involved generations of women who were prescribed different types of hormones, doses, and routes of administration. Therefore, the study was not a clinical trial of hormone therapy and cannot provide a definitive answer regarding dementia prevention.

What can be said about hormonal therapy from this work?

Question	Answer
Was therapy taken into account statistically?	Yes
Used it around menopause	18.8% women
Did it change the relationship with brain volume?	no convincing effect was found
Does this prove lack of neuroprotection?	No
Can HRT be prescribed for the prevention of dementia in this work?	No
Need separate studies?	Yes

Why Estrogen Is Only Part of the Possible Explanation

The authors discuss the loss of estrogen signaling as a biologically plausible mechanism. Estrogens influence synapse formation and stability, myelin, neuronal energy metabolism, inflammation, and vascular function. Therefore, prolonged exposure to endogenous hormones could theoretically increase the resilience of nervous tissue.

But surgical menopause reminds us that it's not just about estrogen. The ovaries continue to produce androgens even after natural menopause, and when both ovaries are removed, this source of hormones is also eliminated. Some androgens are converted to estrogens in other tissues, so the hormonal changes after surgery are more profound than simply the disappearance of cyclic estradiol.

Moreover, early natural menopause may itself be a marker of accelerated systemic aging. The same biological or social factors that lead to earlier cessation of ovarian function may independently increase the risk of cardiovascular disease, sleep disorders, metabolic disorders, and brain aging.

Therefore, researchers propose considering the age of menopause both as a biological effect and as an integral marker of the overall aging process. Even if part of the association proves non-causal, such a marker could still be useful for risk assessment, as the age of menopause is easily determined decades before the onset of dementia.

Does early menopause mean a woman will definitely develop dementia?

No. Even the most significant study results relate to average differences between large groups, not to the fate of a specific individual. Alzheimer's disease and other types of dementia have a multifactorial origin: age, heredity, vascular disease, blood pressure, diabetes, smoking, physical activity, education, sleep, and many other factors play a significant role.

For example, the presence of the APOE ε4 variant significantly influences the likelihood of Alzheimer's disease, but even this does not determine a person's future with absolute certainty. Similarly, menopause at age 43 does not necessarily mean cognitive decline will begin earlier, and menopause at age 52 does not guarantee protection.

Furthermore, many of the study's numerical effects were small. The association between age at menopause and total brain volume was below the traditional threshold for a small effect, and the differences in the rate of cognitive decline for each individual year of menopause were very modest. The exception in magnitude was the white matter hyperintensity lesion index in natural menopause.

Therefore, the practical significance of this study lies not in the "early menopause = dementia" prognosis, but in potential risk stratification. If further studies confirm these findings, menopause age could be considered alongside blood pressure, diabetes, smoking, and other factors to identify women in whom prevention of cerebrovascular injury is especially important.

What can potentially be changed if the age of menopause can no longer be changed?

The age of natural menopause itself cannot be changed in most cases. Therefore, the clinical value of such a marker lies not in an attempt to "delay menopause," but in using information about it to earlier manage modifiable brain risk factors.

It's especially logical to focus on vascular factors, given their strong association with white matter damage. Controlling blood pressure, diabetes, lipids, smoking, and physical activity already has independent effects on the prevention of stroke, small vessel disease, and dementia. The new study suggests that such prevention may be particularly important for women with very early menopause, although the authors did not directly test this strategy.

Equally important are sleep disturbances, including obstructive sleep apnea, which the authors themselves cite as potential factors linked to both menopause and brain aging. Mood and cardiometabolic health may also be part of the overall chain.

In other words, the study doesn't reveal a new pill, but rather a potential window for midlife prevention. A woman can learn the age of her menopause several decades before the typical age of dementia, theoretically providing enough time to address modifiable factors. The authors consider this to be one of the most practically important findings.

Why the work is stronger than most previous studies

The first advantage is the longitudinal design. Instead of a single MRI, the authors used repeated images and were able to assess the rate of brain change within a single individual. This is significantly more informative than simply comparing the brain volumes of different women at a single point in time.

The second was a multi-year, annual cognitive assessment using 19 standardized tests. This allowed us to determine the trajectory of various cognitive domains and compare it with MRI and clinical diagnostic data for Alzheimer's disease.

Third, the presence of postmortem neuropathology. This allowed the researchers to demonstrate an important discrepancy: clinical and structural consequences were noticeable, but there was no proportional increase in classic Alzheimer's pathology. This result significantly changes the interpretation of the possible mechanism.

Finally, the statistical models accounted for numerous potential confounding factors and adjusted for multiple comparisons. This is especially important in studies with a large number of cognitive, MRI, and pathological measures, where without such adjustment, statistically significant results could easily be obtained by chance.

But the study has serious limitations.

The main limitation is that the age of menopause and the onset of menstruation were determined based on the participants' own recollections, and the women reported these information in old age. Such data are fairly reliable at the population level, but an individual woman may be off by several years. The authors note that such an error is more likely to weaken the true association than to artificially strengthen it.

Second, the researchers did not have complete data on the number of pregnancies, oral contraceptive use, breastfeeding, and a number of other reproductive factors. Therefore, the calculated duration of the reproductive period is only an approximate indicator of overall hormonal exposure.

Third, the category of surgical menopause was heterogeneous. The database could not reliably distinguish between removal of both ovaries, hysterectomy with ovarian preservation, and some other surgical options. Meanwhile, their hormonal consequences vary significantly.

Fourth, the MRI subgroup was healthier than the entire original cohort. Women who survived to the start of the 3-Tesla MRI program and were able to undergo regular screening had better baseline cognitive status and were later diagnosed with Alzheimer's disease. Statistical weighting reduces this selection effect, but does not guarantee its complete elimination.

Finally, 92.2% of the participants were white women, and the average education level was approximately 16 years. This is a rather specific and well-educated American cohort of older adults, so the exact effect sizes cannot be automatically generalized to all countries and social groups.

Limitation	Possible influence
Age of menopause according to recollections	a mistake of several years
There is no complete data on pregnancy and contraception	incomplete assessment of hormonal history
The exact type of operations is unknown	heterogeneous surgical group
Observational design	causality has not been proven
Reverse causality is possible	Early menopause may be a marker of general aging.
Healthier survivors underwent MRI	selection effect
92.2% of participants are white	limited tolerability
Low prevalence of modern HRT	its protective effect cannot be reliably assessed

Could early menopause be a symptom of accelerated aging rather than a cause?

This is one of the central methodological questions of the study. Let's imagine that a certain set of genetic, metabolic, or social factors simultaneously accelerates ovarian and vascular aging. In this case, a woman enters menopause earlier and accumulates brain damage later, but the loss of estrogen itself may be only part of the overall picture, not the root cause.

The authors explicitly identify this reverse causality as a serious problem of interpretation. In this regard, surgical menopause represents a unique natural experiment: the timing of ovarian cessation is determined by the surgery, not just by biological aging.

The authors believe that some clinical and cognitive associations were significantly stronger after surgical menopause to be consistent with a causal role for hormonal cessation. However, even surgery is not truly randomized: women were operated on for specific medical conditions, which themselves can impact subsequent health.

Therefore, the most accurate conclusion is that the data support a biological contribution to the timing of ovarian function loss, but do not completely separate it from the overall process of systemic aging. Proving the mechanism will require other types of research—genetic causal analyses, more precise prospective hormonal observations, and, for specific questions, clinical trials of interventions.

What the study changes in understanding female brain aging

In neuroscience, menopause was previously often viewed as a relatively brief transitional stage during which sleep, mood, or subjective memory may temporarily decline. New research shows that the timing of this transition may be linked to brain trajectories decades later.

It's particularly telling that the strongest structural finding relates to white matter. This takes the discussion beyond classic Alzheimer's disease and highlights the role of cerebrovascular aging. The elderly brain ages not only due to amyloid and tau; the condition of small vessels and white matter also significantly determines cognitive reserve.

The study also shows why age at menopause cannot be assessed in isolation. A shorter reproductive lifespan produced similar outcomes, surgical and natural menopause had different patterns, and hormone therapy in this cohort could not explain all the differences. Biology is likely much more complex than the simple formula "less estrogen = more dementia."

The most promising idea is to use menopause as an early checkpoint for brain health. Clinical dementia is still very far off at 45-55 years of age, but it is during this period that vascular risk, sleep, metabolism, physical activity, and other potentially modifiable factors can be assessed long before irreversible structural changes occur.

The main conclusion

A study of 2,603 women found that earlier menopause was associated with a more rapid decline in general cognitive function and episodic memory, as well as a slightly earlier clinical onset of Alzheimer's disease. These associations persisted after accounting for age, education, body weight, smoking, hormone therapy, and APOE ε4.

Structurally, the most striking finding was the accumulation of white matter hyperintense lesions after natural menopause. According to the authors' model, a woman whose menopause occurred five years earlier would accumulate approximately 15% more of these changes over the course of a decade than a comparable woman with a later menopause.

However, the association with overall brain volume reduction was weak, and no significant increase in classic Alzheimer's pathology was detected at autopsy. Therefore, the results are particularly consistent with the hypothesis that menopause timing may influence vascular and white matter aging of the brain, although this mechanism remains to be confirmed.

The study does not prove that early menopause causes dementia, nor does it demonstrate that hormone therapy prevents it. The most compelling practical conclusion is much more cautious: the age at menopause may be a simple marker of future brain vulnerability, known decades before symptoms appear, and therefore potentially useful for earlier prevention of modifiable risk factors.

News source

Campagna MP, Schneider JA, Barnes LL, Arfanakis K, Levi Dunietz G, Bennett DA, Bove RM. Age at Menopause and Brain Atrophy Among Older Women. JAMA Network Open. 2026;9(8):e2630973. Article published August 25, 2026. DOI: 10.1001/jamanetworkopen.2026.30973.

This is a novel longitudinal cohort study, not a randomized clinical trial. The analysis included 2,603 women with long-term cognitive data, 774 participants with serial 3-Tesla MRI, and a large subgroup with postmortem neuropathological examination. Follow-up in the two original cohorts was up to 18 years.

The study was conducted by researchers from the University of California, San Francisco, Rush University Medical Center, and the University of Michigan. The Religious Orders Study and Rush Memory and Aging Project cohorts are supported by grants from the US National Institutes of Health. The authors disclosed specific research grants and external affiliations; the article does not include any significant commercial disclosures directly related to the menopause age result.