Vasomotor symptoms and their links to cardiovascular disease risk

Abstract

Hot flashes and night sweats, also known as vasomotor symptoms (VMS), are common and bothersome symptoms of the menopause transition. In addition to negatively impacting quality of life, VMS have been associated with multiple indicators of cardiovascular disease (CVD) risk, including an unfavorable CVD risk factor profile, increased subclinical CVD, and elevated risk of CVD events. Several facets of VMS have been associated with CVD risk, including the frequency, timing, duration, and severity of VMS. VMS may signify poor or degrading cardiovascular health among midlife women and indicate women who warrant focused CVD prevention efforts.

Introduction

Vasomotor symptoms (VMS), also known as hot flashes or night sweats, are considered the hallmark symptom of the menopause transition. Estimates indicate that approximately 70% of midlife women living in the United States experience VMS¹ and for a third of these women VMS are frequent or severe.² In addition to being common, VMS are associated with impairments in quality of life and increased healthcare utilization.³

Risk factors for experiencing VMS include smoking, occupying a lower socioeconomic status (SES), or being in the late peri or early postmenopause relative to the premenopause.¹ In addition, factors such as greater anxiety or depressive symptoms, greater reported stress, and financial strain may be associated with a higher occurrence of VMS.^1,4 Further, among women living in the United States, Black women have a greater incidence of, prevalence of, and more persistent VMS, while Asian women report VMS less frequently than Hispanic and non-Hispanic white women.¹

VMS occur in the context of the marked fluctuations in and eventual withdrawal of estradiol during the menopause transition. Longstanding models posit a narrowing of the thermoneutral zone, with VMS representing profuse heat dissipation events occurring in the context of this narrowed zone.⁵ Further, recent advances have identified kisspeptin, neurokinin B, and dynorphin neurons (KNDy neurons) in the etiology of VMS. These neurons located in the arcuate nucleus of the hypothalamus play a central role in both regulation of gonadotrophin releasing hormone pulsatility as well as in thermoregulation. The estradiol withdrawal of the menopause transition results in the removal of negative feedback by ovarian estradiol on the KNDy neurons. These neurons hypertrophy and become overactive, giving rise to VMS.⁶

Another major health issue for women is cardiovascular disease (CVD), the leading cause of death in women.⁷ Most women start experiencing CVD events in the 7^th decade of life.⁸ However, in recent years CVD events are accelerating among younger women (below age 55).⁹ CVD in women is notable for the role of reproductive events, such as ovulation, pregnancy, oophorectomy, and menopause in cardiovascular health.¹⁰ In longitudinal studies, the menopause transition is accompanied by accelerated accumulation of atherosclerosis, adverse changes in lipid profiles and body composition, and increased vascular stiffening; these changes are not explained by chronologic aging itself. ^4,11,12 Alterations in sex hormones, such as changes in endogenous estradiol, progesterone, and follicle stimulating hormone, as well as anovulatory cycles, may contribute to these changes;^10,13–15 however, typically these changes are not fully explanatory. Other menopause-related factors emerging as relevant to women’s cardiovascular health are VMS, which have been associated with indicators of cardiovascular disease (CVD) risk.¹⁶

VMS and CVD Risk Factors

Standard CVD Risk Factors

VMS have been associated with adverse CVD risk factor profiles, including increased blood pressure (BP), insulin resistance, dyslipidemia, diabetes, and metabolic syndrome. These CVD risk factors are modifiable and increase risk for future CVD events.⁷ Several cross-sectional studies have found associations of more frequent self-reported and physiologically-assessed VMS with higher BP evaluated in both the clinic and the ambulatory setting.^17–20 Other cross-sectional work has shown VMS associated with increased odds of hypertension,¹⁸ although this association was not universally observed after accounting for other CVD risk factors.^19,21,22 Findings from the prospective, longitudinal Study of Women’s Health Across the Nation (SWAN) has provided more definitive information, showing that relative to no VMS, frequent VMS were associated with higher systolic BP, pulse pressure, mean arterial pressure, and increased odds of hypertension over more than 10 follow-up visits, even after adjusting for multiple other CVD risk factors.^22,23

In addition to BP, both greater self-reported frequency and severity of VMS have been linked to increased homeostatic model assessment (HOMA) index, an indicator of insulin resistance.^24–26 Further, the presence, greater severity, and longer duration of VMS have been associated with increased risk of incident diabetes in the Women’s Health Initiative study, independent of obesity.²⁷ Presence and severity of VMS have also been associated with adverse lipid profiles,^18,28,29 though not in smaller studies.^30,31 Notably, data from SWAN showed that frequency of self-reported VMS was associated with higher low-density lipoprotein (LDL), high-density lipoprotein (HDL), apolipoprotein A-1, apolipoprotein B, and triglycerides relative to no VMS, even after controlling for additional CVD risk factors.³² The positive associations between VMS and HDL can be interpreted in light of the growing body of literature suggesting that HDL function and subpopulation alter with the menopause transition such that HDL loses its cardioprotection.⁴ Finally, VMS have also been linked to increased risk of the metabolic syndrome (MetS), a cluster of CVD risk factors including elevated BP, insulin resistance, central obesity, and dyslipidemia,^33–36 particularly in more rigorously designed studies.^35,36 Thus, findings on balance, particularly from rigorously-designed prospective cohort investigations, indicate that VMS are associated with multiple adverse CVD risk factors such as elevated BP and hypertension, insulin resistance and diabetes, adverse lipid profiles, and MetS.

VMS and Subclinical CVD

VMS have been linked to a range of subclinical CVD indicators associated with increased risk for future CVD events, including measures of endothelial dysfunction, atherosclerosis, and arterial stiffness.⁷

Endothelial injury is an initiating event in the development of atherosclerosis.³⁷ Vascular endothelial function decreases across the menopause transition,³⁸ and the endothelium is sensitive to sex hormones such as estrogens and progesterone.^39,40 Studies have indicated that the presence of VMS, more frequent physiologically-assessed VMS, more severe VMS, or early onset VMS were associated with lower flow mediated dilation (FMD), a measure of endothelial dysfunction.^41–45 A study that did not find an association between VMS and FMD was conducted among a small sample of relatively thin, healthy women,⁴⁶ indicating that some burden of CVD risk may be required before associations are observed.

Several studies have investigated links between VMS and atherosclerosis. Carotid intima media thickness (cIMT) is a measure of the thickness of the intimal and medial layers of the carotid artery that reflects vascular remodeling and can serve as an early marker of atherosclerosis.⁷ Studies have found a greater frequency and severity of VMS, as well as early onset VMS associated with greater cIMT.^46–49 A study utilizing physiologically-assessed VMS found that more frequent physiologically-assessed VMS were associated with higher cIMT and carotid plaque even after adjustment for additional CVD risk factors and endogenous estradiol.⁵⁰ The few studies that have not found associations between VMS and cIMT were among young women at low vascular risk.^45,51

Additional studies have examined associations between VMS and other indicators of subclinical CVD and even cerebrovascular health. Several studies documented associations between the presence of or more persistent VMS and greater aortic calcification, a measure of calcified plaques in the aorta.^41,52 However, several studies that examined VMS and coronary artery calcification (CAC), a measure of calcified plaque in the coronary artery, did not see associations when adjusting for CVD risk factors.^41,51,53 Of note, these studies were limited as very few midlife women have CAC⁵⁴ and studies among older women required extensive time frame of recall for VMS.⁵³ Work that has examined VMS and arterial stiffness (as assessed by pulse wave velocity) has been mixed. A large study of over 500 women showed associations between greater VMS frequency and higher arterial stiffness,⁵⁵ yet two smaller studies did not show these associations.^46,56 Finally, research has begun to examine associations between VMS and cerebrovascular health. More frequent physiologically-assessed VMS during sleep were associated with greater white matter hyperintensities, an indicator of small vessel disease in the brain, after adjusting for multiple confounders.^57,58 Thus, VMS have been associated with endothelial dysfunction, greater subclinical atherosclerosis, and possibly even adverse changes in cerebrovascular health.

Clinical CVD Risk

Clinical CVD events are arguably the most clinically relevant outcome for cardiovascular health. However, studying links between VMS and CVD events represents a methodologic challenge, as VMS commonly occur during the 4^th and 5^th decades of life, whereas clinical CVD events typically accumulate in the 7^th decades of life and beyond in women.⁸ Early studies examining links between VMS and clinical CVD generally asked older women to recall their VMS over decades prior. For example, the Women’s Health Initiative Observational study (WHI-OS) of 60,027 women examined VMS timing based upon a combination of retrospective and baseline reports of VMS, and found that late VMS that developed after menopause onset were associated with increased risk of incident coronary heart disease (CHD), stroke, CVD (combined CHD and stroke), and all-cause mortality compared to no VMS.⁵⁹ The Women’s Ischemia Syndrome Evaluation (WISE) study found that among 254 midlife women, women with early onset VMS had higher CVD mortality than women with later onset symptoms.⁴³ Conversely, the Rancho Bernardo study of 867 women found that a history of ever having VMS was not associated with increased risk for CVD mortality.⁶⁰ Finally, a post hoc analysis of a combined cohort of 10,787 Dutch and Swedish women found that women reporting night sweats had increased risk of CHD over the subsequent 10 years compared with women without VMS, even when controlling for demographics and CVD risk factors.⁶¹ Thus, early work generally pointed to an association between VMS and CVD events, yet findings, particularly pertaining to the timing of VMS, were mixed.

More rigorous tests of associations between VMS and CVD events have assessed both VMS and CVD prospectively over an extended follow up period. In SWAN, an ethnically diverse sample of 3,000 women were followed from the start of the menopause transition through early old age, assessing VMS and CVD repeatedly and prospectively for 22 years.⁶² Women in SWAN with frequent VMS at baseline or frequent VMS that persisted over the menopause transition had increased risk of a CVD event (combined outcome of myocardial infarction, stroke, revascularization, heart failure, or CVD mortality) relative to women without VMS, even when adjusting for demographic and CVD risk factors.⁶² In addition, the InterLACE consortium, comprised of over 23,000 midlife women from 6 prospective studies, found that greater VMS severity and early or late onset VMS were associated with increased risk of CVD (composite CHD or stroke) independent of various demographic and CVD risk factors.¹⁶ However, the Australian Study on Women’s Health produced mixed findings, with one of their reports of 11,725 women showing associations between prospectively-assessed VMS and risk of CHD over the subsequent 14 years,⁶³ and another analysis of 9,000 women examined 5 years later finding no association between VMS and incident CVD (combined CHD, peripheral artery disease, aortic aneurysm, and cerebrovascular disease events).⁶⁴ In summary, research has generally demonstrated associations between VMS and increased risk of clinical CVD events, particularly in rigorously-designed prospective studies, yet ongoing research on this question is warranted.

Potential Mechanisms Linking VMS to CVD Risk

Multiple physiologic pathways may link VMS to CVD risk. First, the menopause transition is characterized by marked fluctuations in sex hormones such as estradiol and follicle stimulating hormone, which in turn can have vascular impact;¹⁵ However, sex hormones do not typically explain relationships between VMS and CVD risk.^28,50,62 Second, as reviewed above, VMS are associated with adverse CVD risk factors, yet relationships between VMS and subclinical or clinical CVD typically persist with adjustment for CVD risk factors.^50,59,61,62 A range of other physiologic processes have been implicated in relationships between VMS and CVD risk. For example, research with physiologic measures of VMS have found VMS linked to altered autonomic nervous system cardiac control, particularly reduced parasympathetic^65–67 and increased sympathetic⁶⁷ control over the heart, a profile in turn associated with CVD risk.⁶⁶ Moreover, VMS presence, frequency, and severity have been linked to altered HPA axis function,^68–72 such as a blunted cortisol awakening response^68–70 and higher urinary cortisol.^71,72 Altered HPA axis function is implicated in the pathophysiology of CVD.⁷³ SWAN findings indicate that VMS may be associated with a more pro-coagulant profile,⁷⁴ whereas links between VMS and systemic inflammation have been mixed,^74–76 with some data underscoring an important role of adiposity in these associations.⁷⁴ Further, a small literature suggests that VMS may be associated with alterations in cellular adhesion molecules.⁷⁶ Lastly, initial research conducted in the Women’s Health Initiative Observational Study found that severe and late-occurring VMS were associated with accelerated epigenetic aging,⁷⁷ which may contribute to worse cardiovascular health.⁷⁸ In summary, a range of physiologic processes may link VMS to the development of CVD. However, more rigorous prospective studies are needed in this area.

Summary, Methodological Considerations and Recommendations

An accumulating body of literature demonstrates associations between VMS and adverse CVD risk factors, subclinical CVD, and increased risk of incident CVD events (Figure 1). VMS have been associated with unfavorable CVD risk factors such as elevated BP and hypertension, insulin resistance, an adverse lipid profile, incident diabetes, and MetS. Further, VMS have been related to endothelial dysfunction and subclinical atherosclerosis. Across several longitudinal cohorts, more frequent or persistent VMS have been associated with increased clinical CVD risk. Lastly, relationships have been observed between VMS and potential mechanisms to CVD risk such as altered HPA axis function, autonomic nervous system control over the heart, and increased clotting factors.

Figure 1.

Associations between vasomotor symptoms, cardiovascular risk factors, subclinical cardiovascular disease, and clinical cardiovascular disease events

While relationships between VMS and CVD are generally observed, several methodological issues warrant consideration, particularly pertaining to the measurement of VMS. Some research examining associations between VMS and CVD risk utilize retrospective self-reports of VMS which are prone to memory and affective biases, particularly when the frame of recall is extended over many years.^1,79 In addition, assessing VMS experience at a single point in time may not adequately capture VMS over the menopause transition. Future studies should utilize prospective, longitudinal reports of VMS assessed over the duration of the menopause transition, which can minimize recall bias and more comprehensively capture VMS experience.⁷⁹ Smaller clinical studies can consider assessing VMS physiologically, which are useful for difficult reporting periods, such as sleep.⁷⁹ Investigators should also be aware that VMS occurring during wake and sleep may have differential associations with CVD risk.^57,65

Additional methodologic considerations pertain to the assessment of CVD. As women tend to develop clinical CVD relatively late in life, studies investigating clinical CVD should include adequate numbers of appropriately-aged women, and ideally follow women from midlife when VMS occur into their later years when CVD events accumulate. Adjudication and documentation of clinical CVD events should be undertaken. Finally, since VMS are linked to a range of demographic factors (e.g., age, race/ethnicity, socioeconomic position) as well as most CVD risk factors, these indicators should be considered when examining associations between VMS and subclinical and clinical CVD.

Several future research directions warrant investigation. Importantly, any causal nature of associations between VMS and later CVD risk requires further investigation. Additional research is also warranted to clarify the mechanisms that may link VMS to CVD risk. Finally, it is important to understand whether treating the VMS impacts women’s cardiovascular health.

Collectively, the literature indicate links between VMS and poorer CVD risk factors, greater subclinical CVD, and increased risk of clinical CVD events. VMS may indicate underlying vascular risk and may identify women who warrant targeted CVD risk reduction. Thus, VMS are not only common and bothersome, but have implications for women’s cardiovascular health.

Nonstandard Abbreviations and Acronyms

BMI

body mass index

BP

blood pressure

CAC

coronary artery calcification

cfPWV

carotid-femoral pulse wave velocity

cIMT

carotid intima media thickness

CVD

cardiovascular disease

FMD

flow-mediated dilation

HDL

high-density lipoprotein

HOMA

homeostatic model assessment

HPA

hypothalamic pituitary adrenal

KNDy

kisspeptin, neurokinin B, and dynorphin

LDL

low-density lipoprotein

MetS

metabolic syndrome

SES

socioeconomic status

VMS

vasomotor symptoms