Learn more about Peter Attia’s digital program
When women experience the onset of menopause, the sudden drop in estrogen levels causes physiologic symptoms such as hot flashes and night sweats, collectively known as vasomotor symptoms (VMS). Some women don’t experience VMS at all, while others will experience severe VMS for more than a decade, greatly interfering with quality of life.
Yet these symptoms may be more than just a bother to those who experience them. A growing body of literature suggests that women who experience VMS are also more likely to have poorer cardiovascular health outcomes and possibly poorer cognitive trajectories. Some studies suggest that estrogen has a protective effect on the brain, and the sudden drop in estrogen levels during menopause thus increases the risk of dementia. Is it possible that the frequency and severity of VMS might therefore tell us something about dementia risk? A recent observational study by investigators Thurston et al. attempted to answer this question.
Challenges of Measurement
Delineating the relationship between VMS and other chronic diseases is challenging due to the common practice of using self-report data to quantify the subjective experience of VMS. However, VMS are physiological phenomena – involving changes in body temperature, heart rate, and sweat output – and as such, newer approaches for measuring and quantifying VMS have been developed based on physiological readouts, such as changes in sternal skin conductance, which increases with sweating. Such measures may be a more rigorous way to evaluate the relationship between menopausal symptoms and cognitive decline.
Measurement of dementia risk as it relates to VMS is also a challenge since menopause and its associated symptoms often occur decades before the onset of dementia. However, some biomarkers may appear before the onset of dementia, which would increase the feasibility of both associating VMS with cognitive decline and conducting preventative research. One such biomarker is a volumetric measure of small brain lesions visualized with magnetic resonance imaging (MRI), known as white matter hyperintensities. White matter hyperintensities are a marker of cerebral small vessel damage and they have been linked to the future onset of cognitive decline, dementia, and mortality. In a small pilot study, nocturnal VMS were found to be associated with greater white matter hyperintensity volumes (WMHVs), and the recent study by Thurston and colleagues set out to replicate the findings in a larger population.
About the Study
In this cross-sectional study of 224 women, VMS events were measured using 24-hour physiologic monitoring of VMS via sternal skin conductance in addition to a VMS diary of self-reported symptoms for a further two days of monitoring. Within two weeks of VMS monitoring, WMHV was assessed using MRI scans for whole-brain, as well as lobe-specific, analysis. Since it is equally likely that a third factor is a common upstream cause of both increased VMS and WMHV, other measurements of sleep, cardiovascular risk factors, and blood-based biomarkers such as estrogen levels, APOE4 status, and metabolic health were also measured to account for potential confounding factors.
The women in this study were on average 59 years of age, of late perimenopausal or postmenopausal status, and not currently taking hormone replacement therapy. They had an average of five VMS events per 24-hour period, which is equivalent to the average rate of events for women with moderate symptoms of menopause. The spectrum of VMS was represented from women who experienced no VMS to women who experienced high VMS – up to a four-fold higher rate of VMS than the average. The study found that the women who had more frequent physiologically measured VMS had an associated greater WMHV.
Since daytime and nocturnal VMS may have different effects on brain health, the investigators also considered these timings separately. Interestingly, they found that the association of increased WMHV over the whole brain held only for VMS that occurred during sleep. Indeed, there was a persistent relationship between sleep VMS and whole-brain WMHV, even after adjusting for wakefulness after sleep onset, age, race, education, and cardiovascular risk factors that could also explain the underlying cerebral small vessel damage. Additional covariates were evaluated in the analysis but did not significantly change the observed relationship between VMS and WMHV. These covariates included estrogen levels (both estradiol and estrone), follicle-stimulating hormone levels, depressive symptoms, high sensitivity C-reactive protein (a marker of systemic inflammation), ultrasound-measured carotid intima-media thickness (a measure of atherosclerosis), and APOE4 status.
The associations of WMHV and daytime or nocturnal VMS were more variable when refined to a particular area of the brain. For instance, there were significant associations of sleep VMS with increased WMHV in the deep brain and periventricular areas of the brain, whereas increased WMHV in the temporal lobe was only associated with total increased VMS over 24 hours. (Of note, the researchers did not specify sub-regions within these broad area definitions. While the temporal lobe is known to function in memory and other cognitive processes, the roles of the deep brain and periventricular areas vary widely.) However, daytime, nocturnal, and 24-hour VMS were all associated with increased WMHV in the frontal lobe, an area of the brain that plays an important role in memory, attention, and executive function.
Study Strengths and Limitations
One of the study’s strengths was in the design of measuring VMS both via physiologic monitoring and self-report. Unlike the physiologic metrics, self-reported VMS was not associated with increased WMHV regardless of the time of day after adjusting for the same confounding factors as in the analysis done for physiologically measured VMS. This discrepancy highlights the limitations and lack of reliability of self-report data when it comes to VMS. Methods of self-reported VMS generally underreport symptoms and are influenced by sleep quality, negative mood, poorer memory, and non-adherence, making the more objective physiologic measure a superior method for quantifying VMS in this and future studies.
Very few biomarkers are currently available that can predict the onset of dementia potentially decades down the line. Although WMHVs are not considered a benign part of the aging process, the overall evidence from larger studies suggests that there is a small effect size of greater WMHV on cognitive performance. WMHVs are likely one factor of many that contribute to the complex process of cognitive decline.
Even after adjusting for confounding factors, the persistent relationship indicates that there is a small effect of increased physiologically measured sleep VMS and greater WMHV on a population level. However, the increase in WMHV is overall a weak trend (a small slope with a large spread of data), meaning that VMS alone will not have a good predictive value of WMHV for the individual. That is to say, just because a woman experiences more severe VMS, it does not automatically mean she will also have higher WMHV. The study’s results also do not determine whether VMS is a causative factor or a symptom of another underlying process.
Remaining Questions on Causality and Treatment
The variability in the data means there are likely other factors that influence WMHV. This is a cross-sectional study of mostly postmenopausal women who were not currently on hormone replacement therapy (HRT) – meaning it only presents a snapshot in time with no additional context from the past. For instance, the authors evidently did not collect data on the amount of time since the onset of menopause or on any potential history of HRT use. This study excluded women currently on HRT; however, if WMHV results were stratified by prior HRT use, would this affect the variability in the data and improve the strength of the correlation? If estrogen replacement via HRT is protective against WMHVs, would women with prior use of HRT have lower WMHV than never-users? If so, it would suggest that either the deprivation of estrogen or the severity of VMS per se (e.g., through effects on sleep) is responsible for the development of WMHVs – not some unknown covariate.
The duration of VMS would be also important to consider if the development of WMHV (and subsequent dementia) increases with increasing exposure to VMS. Much in the same way we think about smoking exposure in pack-years (number of packs per day multiplied by the number of years of smoking), VMS might be quantified by multiplying the number of events per day by the duration experienced, which would capture the severity of symptoms and the effects of accumulated exposure. If HRT alleviates VMS, then the overall “exposure” would potentially decrease for the duration of use. The question for future studies will be whether or not the early alleviation of VMS with HRT slows or postpones the progression of WMHVs.
Many people still describe the threshold for prescribing HRT for VMS when symptoms are “bothersome.” However, this description of VMS downplays how disruptive these symptoms can be for many women and diminishes the potential relationship between health and VMS. If HRT were found through future research to delay WMHV progression, this effect would be the latest of many reasons not to shy away from HRT, which not only is the most effective therapy for alleviating VMS but also provides other health benefits, such as maintaining bone mineral density.
The Bottom Line
While the trend of greater WMHVs with more severe VMS is an interesting observation, it remains to be seen if early interventions can alter the outcome. An important question for future studies to answer is the degree to which HRT initiated at the onset of menopause can help postpone or inhibit the development of WMHVs and if doing so would reduce the overall incidence of dementia. These types of studies would also address lingering hesitations and questions surrounding the risks and benefits of HRT. In the meantime, the growing evidence linking VMS and poorer overall health indicates that vasomotor symptoms experienced by women during menopause are more than “bothersome” and as such should be taken very seriously.
For a list of all previous weekly emails, click here.
