Vasomotor Symptoms of Menopause, Sympathetic Activity, and Blood Pressure in Postmenopausal Females

Abstract

Objectives:

While evidence suggests an association between vasomotor symptoms (VMS; hot flushes and night sweats) and elevated blood pressure (BP), it remains unknown whether females who experience VMS have elevated muscle sympathetic nerve activity (MSNA), a major modulator of BP. We hypothesized that postmenopausal females with VMS would have elevated BP and MSNA at rest and during stress compared to age-matched females without VMS.

Methods:

Participants were grouped based on whether they currently or previously experienced VMS (n=43) or never experienced VMS (non-VMS; n=26). Heart rate (HR: electrocardiography), BP (finger plethysmography) and MSNA (microneurography), were recorded during a 10-min rest, a two-min cold pressor test (CPT) and a five-min recovery.

Results:

While there were no group differences in resting mean arterial pressure (MAP) or MSNA burst frequency (p>0.05), MSNA burst incidence (53±13 vs. 44±9 bursts/100heartbeats, p<0.01) was greater in the VMS group. Resting HR was lower in the VMS group (58±10 vs. 62±9bpm; p=0.04), and the standard deviation of the RR-interval HR variability (HRV) (52±21 vs. 38±16ms, p<0.01), total cardiovagal baroreflex (cBRS;12±6 vs. 9±3ms/mmHg, p=0.02) and cBRS down (cBRSdown; 12±6 vs. 9±4ms/mmHg, p=0.01) calculated via the sequence method were greater than the non-VMS group. HR, MAP, and MSNA responsiveness to CPT were not different between groups (p>0.05).

Conclusion:

Contrary to our initial hypothesis, postmenopausal females with current or prior VMS demonstrated similar BP and MSNA burst frequency, with greater cardiovagal modulation and baroreflex sensitivity compared to the non-VMS group. Future work into this paradoxical finding appears warranted.

New and Noteworthy:

Postmenopausal females, who are otherwise healthy and experience VMS of menopause, did not exhibit heightened MSNA and BP, but demonstrated elevated HRV, total cBRS, cBRSdown and a lower HR compared to postmenopausal females with no current or previous history of VMS. These findings suggest that healthy females with mild to moderate VMS do not exhibit autonomic dysregulation when compared with females who have never experienced VMS.

INTRODUCTION

Cardiovascular disease (CVD) is the leading cause of death in aging females after menopause in the United States (1). Hypertension, defined as resting systolic blood pressure (SBP) ≥ 130 mmHg or resting diastolic blood pressure (DBP) ≥ 80 mmHg, is a major risk factor for CVD (2). While premenopausal females have lower blood pressure (BP) than age-matched males, females experience a higher incidence of hypertension than males after the typical age of menopause (~51 yrs) (3). It is possible that the greater prevalence of hypertension in older females may be linked to vasomotor symptoms (VMS; hot flushes and night sweats) of menopause as some (4–6), but not all (7), studies report that the presence of VMS is correlated with higher SBP and DBP. In a longitudinal study, the occurrence of VMS was demonstrated to be one of the most predictive risk factors for the BP rise over the menopause transition in a sample of 3,302 multi-ethnic/-racial females (8). Indeed, evidence suggests a strong association with VMS and greater CVD risk (9–11). Despite these associations, the mechanistic pathways linking VMS to hypertension and CVD development remain equivocal (12), but are critical to identify, as up to 80% of females experience VMS (13).

The autonomic nervous system, which includes the sympathetic and parasympathetic branches, is an important regulator of heart rate (HR) and BP via baroreflex mechanisms (14). Autonomic dysfunction often presents as elevated sympathetic activity, parasympathetic withdrawal and an attenuated baroreflex sensitivity that can contribute to the development of CVD (15). Postmenopausal females have demonstrated autonomic dysfunction when compared with premenopausal females at rest (16) and in response to stressors such as the cold pressor test (17), and VMS may exacerbate this dysregulation. For example, heart rate variability (HRV), a reliable measure of parasympathetic (but not sympathetic) function (18, 19), is reduced in females during a hot flush (20–22). There is mixed evidence of altered parasympathetic function at rest, in the absence of a hot flush, among females who experience VMS, with some studies finding reduced baseline parasympathetic function compared to females who do not experience VMS (23) and others reporting no difference (24). Other studies that measured catecholamine levels during a hot flush have also produced equivocal results. While one study found a 100% increase in plasma norepinephrine with unaltered epinephrine levels during a hot flush (25), another study found a 40% decline in plasma norepinephrine levels and a 150% increase in epinephrine levels during a hot flush (26). However, these approaches have limitations in their ability to quantify sympathetic function. For example, plasma levels of norepinephrine and epinephrine can lack intraindividual reliability due to clearance rates (27). Therefore, a more rigorous technique is necessary to measure sympathetic activity in females with VMS to help identify autonomic mechanisms linking VMS to CVD risk.

Microneurography, a technique to quantify the activity of postganglionic nerves supplying vascular beds, provides a real-time dynamic assessment of the sympathetic nervous system known as muscle sympathetic nerve activity (MSNA), and it is considered the gold-standard measurement of peripheral sympathetic function in humans (28, 29). While sympathetic activity rises with age in both sexes, the age-related increases are steeper in females compared with males (30). Further, MSNA increases more rapidly with a stressor, the cold pressor test, in postmenopausal females compared with age-matched males and younger females and males (17). These studies, along with others (31), suggest that the increase in resting BP with age may be more autonomically mediated in females than in males. It remains to be seen, however, if there is a difference in autonomic support of BP in postmenopausal females with VMS compared to postmenopausal females without VMS. The purpose of this study was to investigate if postmenopausal females who previously experienced or currently experience VMS demonstrate altered hemodynamic and sympathetic function during rest and in response to a cold pressor test. The cold pressor test was utilized for this study protocol because it is a reliable and reproducible method to increase sympathetic activation and has been shown to detect sex- and age-related differences in a previous study (17). We hypothesized that BP, MSNA and blood pressure variability (BPV, an indirect indicator of sympathetic control of BP) will be elevated and baroreflex sensitivity, an important regulator of BP, as well as HRV will be reduced in postmenopausal females who have a history of VMS compared to females who have never experienced VMS at rest. In addition, we expect that BP and sympathetic reactivity will be exaggerated during a cold pressor test in females with a history of VMS compared with females who have never experienced VMS.

METHODS

Participants

This study was approved by the University of Minnesota’s Institutional Review Board (IRB-STUDY00004979), was registered with ClinicalTrials.gov (NCT04439370) and was conducted in accordance with the Declaration of Helsinki. All participants provided written informed consent prior to engaging in study procedures. Postmenopausal females between the ages of 47 and 69 years old participated in this study. Exclusion criteria included: a body mass index (BMI) ≥ 35kg/m²; history of CVD, apart from a diagnosis of hypertension; autonomic dysfunction; diagnosed carotid stenosis; respiratory disease; nicotine/tobacco use within the past six months; diabetes or non-well-controlled asthma; severe neurologic conditions such as stroke or traumatic brain injury; use of heart or BP medications; sleep apnea; or use of medications determined to alter autonomic function such as hormone replacement therapy; antihypertensive medications; selective-serotonin reuptake inhibitors or selective-norepinephrine reuptake inhibitors. All participants were ≥ 12 months without menses by self-report, indicating that they were postmenopausal. Further, to confirm menopause, estradiol levels were required to be <30pg/mL and follicle-stimulating hormone (FSH) levels to be ≥30pg/mL (32).

Experimental Protocol

Sixty-nine total participants were recruited using a combination of social media advertisements, fliers and the University of Minnesota’s StudyFinder and were screened for eligibility prior to enrollment. Participants completed two study visits. In visit one, participants provided written informed consent and completed a medical questionnaire to confirm study eligibility and report their level of education, history of nicotine/tobacco and alcohol use, a self-identification questionnaire to determine race/ethnicity, the State-Trait Anxiety Inventory (STAI; State and Trait components) to evaluate levels of anxiety (33), the Minnesota Leisure Time Physical Activity Questionnaire (MLTPAQ) to quantify levels of physical activity (34), the Patient Health Questionnaire-9 (PHQ-9) to measure symptoms of depression (35) and the Pittsburgh Sleep Quality Index (PSQI) and Insomnia Severity Index (ISI) to measure symptoms of sleep difficulty (36, 37). Participants also completed the Menopause-Specific Quality of Life Questionnaire (MENQOL), a survey in which participants rank how bothered they have been in the last month by menopausal symptoms in four domains: psychosocial, vasomotor, physical and sexual. Responses are scored on a scale of 1–8, with 1 indicating the absence of a symptom, 2 indicating the symptom is present but the individual is ‘not at all bothered’ and 8 indicating that a symptom is present and the individual is ‘extremely bothered’ (38). Participants who reported in the medical questionnaire either previously (n=8) or currently (n=35) experiencing VMS were placed into the VMS group (total: n=43), while participants without current or past VMS were placed into the non-VMS group (n=26).

Before the COVID-19 pandemic, visit one was conducted in person and a blood draw was completed at the conclusion of this visit (n=18). During and post-pandemic, visit one occurred via Zoom, a virtual meeting platform, and a blood draw was completed at the conclusion of the second visit (n=51). Two participants opted not to complete the blood draw; thus, the blood draw was completed on a total of 67 participants. Levels of FSH, estrogens (estradiol, estrone and total estrogens), progesterone and testosterone were measured from the blood.

For visit two, participants reported to the laboratory after an overnight fast and abstinence from caffeine, exercise and alcohol for 12 hours. The MENQOL and STAI-State questionnaires were completed an additional time on the morning of visit two to determine their current levels of menopausal symptoms and anxiety. Height and weight were measured at the beginning of each visit to calculate BMI. Participants were then instrumented with a three-lead electrocardiogram (ECG) (ADInstruments, Colorado Springs, CO) to measure HR, microneurography to measure MSNA and a non-invasive blood pressure (NIBP) wrist unit (ADInstruments) to measure BP via finger plethysmography. All participants then completed a 10-min rest, a two-min cold pressor test where the participant’s hand was placed in a bucket of ice water (1°-3° C) and a five-min recovery after the cold pressor test (17). Throughout the cold pressor test, participants reported their pain scores on a scale of 0–10 every 15 seconds (0=no pain and 10=worst pain imaginable). Blood samples were collected from the antecubital vein by trained study staff and processed through the Fairview Hospital Acute Laboratory. Follicle-stimulating hormone and progesterone were measured using a chemiluminescent enzyme immunoassay technique (39), while estrogens and testosterone were measured using a liquid chromatography technique (40). Only a subset of participants had detectable levels of certain sex hormones (Table 1).

Table 1.

Participant characteristics.

	VMS	Non-VMS	p-value
Age (yrs)	60±6	60±4	0.91
Age of menopause (yrs)	49±7	47±8	0.29
BMI (kg/m2)	27±4	26±5	0.44
Self-Identification (a.u.)	35 EA, 7 O, 1AA	21 EA, 3 O, 2 AA	N/A
Highest level of education (a.u.)	4 ≤ high school, 39 >high school	1 ≤ high school, 25 >high school	N/A
Past tobacco/nicotine use	1 Yes, 42 No	3 Yes, 22 No	0.12
Current alcohol use	31 Yes, 12 No	16 Yes, 10 No	0.43
STAI-State (a.u.)	29±8	30±6	0.53
STAI-Trait (a.u.)	30±7	28±7	0.44
MLTPAQ (kcal/day)	236±131	257±251	0.78
PHQ-9 (a.u.)	2±2	2±2	0.84
MENQOL – Hot Flush score (a.u.)	3±2	1±0	<0.01
PSQI (a.u.)	8±3	6±3	0.08
ISI (a.u.)	7±4	5±3	0.053
FSH (IU/L)	84±25	93±24	0.10
Estrone (pg/mL)	22±8	32±23	0.10
Estradiol (pg/mL)	5±2	7±5	0.19
Total estrogen (pg/mL)	26±11	40±28	0.09
Progesterone (ng/mL; HF)	0.3±0.2	0.4±0.5	0.35
Testosterone (ng/dL; HF)	21±19	24±13	0.47

Definitions: VMS: vasomotor symptoms; BMI: Body Mass Index; STAI: state-trait anxiety index; EA: European-American; O: Other; AA: Asian-American; MLTPAQ: Minnesota leisure time physical activity questionnaire; PHQ: patient health questionnaire; MENQOL: menopause quality of life questionnaire; PSQI: Pittsburgh sleep quality index; ISI: insomnia severity index; FSH: follicle-stimulating hormone; SBP: systolic blood pressure; DBP: diastolic blood pressure; MAP: mean arterial pressure; HR: heart rate; MSNA: muscle sympathetic nerve activity; hb: heart beats. Sample sizes: Past tobacco/nicotine use: VMS: 43, non-VMS: 25; MLTPAQ: VMS: 35, non-VMS: 14; PSQI: VMS: n=20; non-VMS: n=19; ISI: VMS: n=20; non-VMS: n=20; FSH, testosterone; VMS: n=42; non-VMS: n=23; Estrone: VMS: n=42; non-VMS: n=24; Estradiol and total estrogen: VMS: n=38; non-VMS: n=22; Progesterone: VMS: n=16; non-VMS: n=19.

Measurements

Heart rate and BP were measured continuously in a supine position throughout the study. Two baseline manual BP measurements were taken via a manual sphygmomanometer to calibrate the NIBP before the 10-min rest and prior to the cold pressor test. If the two manual BPs (SBP and/or DBP) were not within 5mmHg of each other, then a third BP was taken.

Multi-fiber MSNA was obtained and recorded via microneurography (ADInstruments, Colorado Springs, CO) using recent guidelines (41). Briefly, the location of the right fibular nerve was located by either external stimulation to evoke involuntary muscle activation resulting in dorsiflexion and/or eversion (ADInstruments) or with the use of ultrasound (Fujifilm Sonosite, Bothell, WA) at the location of the fibular head and/or popliteal fossa. Next, a tungsten recording electrode was inserted percutaneously into the fibular nerve. A reference electrode was inserted 2–3cm from the recording electrode and a ground electrode strap was placed distal to the knee. All electrodes were connected to a neural amplifier (Neuro Amp EX, ADInstruments) and then attached to the PowerLab 16/35 data acquisition unit (ADInstruments). The nerve signal was amplified (total gain 80,000), band-pass filtered (700–2,000 Hz) and integrated (time constant 0.1). The MSNA signal was confirmed by assessing the nerve response to end-inspiratory or end-expiratory apnea as well as the absence of a response to startle (i.e., shout/clap) or stroking of the skin on the dorsum of the foot (42).

Data analysis

Data was recorded with LabChart 8 software (ADInstruments). R-wave peaks of the ECG were automatically detected, confirmed and used to quantify HR. Systolic BP, DBP and mean arterial pressure (MAP) were calculated from BP waveforms using LabChart 8 software.

Cardiovagal baroreflex sensitivity (cBRS) was determined over the 10-min rest period using the sequence method (43), where the average slope of the correlation between SBP and RR-interval was calculated in sequences of at least three consecutive beats of SBP and the RR-interval increasing or decreasing by 1mmHg or 5ms, respectively. Sequences of three or more beats changing in the same direction were only included in the calculation if the correlation coefficient was ≥ 0.7 (i.e., r ≥ 0.7) (44). Sequences were manually inspected to exclude artifacts. Baroreflex sensitivities to rising BP (BRSup) and falling BP (BRSdown) were quantified, as they correlate with the phenylephrine pressor (rising) and nitroprusside depressor (falling) stimuli, respectively (45, 46). Thus, BRSup is the average of the slopes where SBP and RRI increase during at least three consecutive cardiac cycles and BRSdown is the average of the slopes where SBP and RRI decrease during at least three consecutive cardiac cycles.

The HRV analysis software module within LabChart (ADInstruments) was used to calculate both time domain and frequency domain variables over the 10-min rest period. The interbeat interval (between successive R waves of the QRS complex) was measured for both the time and frequency domain analyses. The time domain variables, standard deviation of normal inter-beat intervals (SDNN) and the root mean square of successive R-R Intervals (RMSSD), are variance measures for short-term HRV. A power spectral analysis was used to calculate the high-frequency (HF; 0.15Hz to 0.4Hz) HRV components of both absolute (HF ms²) and relative (HF-%) power, which can be dependent on respiration but are associated with parasympathetic dominance (47, 48). Low-frequency components were not used as they are not a reliable measurement of sympathetic activity (49). All beats were visually examined to make sure they were not associated with ectopic beats.

Blood pressure variability (BPV), a measure predominately controlled by the baroreflex (50–52), was analyzed over the 10-min rest period. The BPV variables of standard deviation (SD), coefficient of variation (SD/mean x 100; CV%) and the average real variability (ARV) (50) were calculated for SBP and DBP using previously utilized formulas (51).

Muscle sympathetic nerve activity data was obtained in a sub-sample of participants during rest (VMS: n=31; non-VMS: n=15) and the cold pressor test (VMS: n=25; non-VMS: n=10). Bursts of MSNA were automatically detected using the Ensemble software program (Elucimed Ltd., Wellington, New Zealand) based on a signal-to-noise ratio of 3:1 and a burst peak latency of ~1.2 s from the previous R wave. Bursts were then independently verified and edited by a trained observer (MLKR). Resting MSNA was expressed as burst frequency (bursts/min) and burst incidence (bursts/100 heartbeats) and was averaged across the 10-min rest period. Additionally, total MSNA (a.u.) and MSNA burst area (a.u.), which are sensitive to changes in response to a stressor (53, 54), were calculated throughout the cold pressor test.

Sympathetic baroreflex sensitivity (sBRS) was calculated using a method outlined by Hamner et al. (55). Diastolic BP for each cardiac cycle in the 10-min rest period was grouped into 3-mmHg bins and plotted against the respective burst incidence of the cardiac cycle. Each data point was weighted according to the number of cardiac cycles because the bins at the highest and lowest diastolic pressures contain fewer cardiac cycles (56). Sympathetic baroreflex sensitivity was calculated as the average slope between DBP and the 3-mmHg bins. Any slopes with an r-value of less than 0.4 were excluded from the analysis (57).

To determine the hemodynamic and sympathetic responses to the cold pressor test, SBP, DBP, MAP, HR, MSNA total activity, burst frequency, burst incidence, and burst area were analyzed at five different time points: the ten-min rest and every 30 sec (0–30, 30–60, 60–90, and 90–120 sec) from the start to the end of the cold pressor test. To evaluate the recovery from the cold pressor test, SBP, DBP, MAP, HR, MSNA total activity, burst frequency, burst incidence and burst area were analyzed at six different time points: 90–120 sec at the end of the cold pressor test and an average of every minute during the five-min recovery (0–60, 60–120, 120–180, 180–240, and 240–300 sec). Sympathetic transduction, a measurement of how sympathetic outflow is transduced into a vascular or BP response, was calculated for the cold pressor test using the delta-delta method by dividing the relative change of MAP from baseline to the last 30 seconds of the cold pressor test by the relative change of MSNA burst frequency (∆MAP/∆MSNA BF) (58).

Statistical analysis

All statistical analyses were performed using SPSS Statistics 28.0.1 (IBM, Chicago, IL). A Shapiro-Wilk test was used to ensure normal distribution of the data. Independent-samples t-tests were then utilized to compare questionnaire scores and resting data between the VMS and non-VMS groups. A chi-squared test was used to analyze smoking history and current alcohol use between VMS and non-VMS groups. A repeated measures Analysis of Variance (ANOVA) test was conducted to investigate main effects of time and group and interactions between time and group (time × group interaction) during the two-min cold pressor test and during the five-min recovery. A Bonferroni adjustment was used to identify pairwise comparisons within the repeated measures ANOVA. To determine if the severity of VMS was related to changes in hemodynamic or sympathetic variables, a bivariate Pearson product coefficient was calculated for the MENQOL visit two VMS score and all hemodynamic and sympathetic variables. Only participants currently experiencing VMS symptoms were included in this analysis. Results are expressed as mean ± SD. All analysis was conducted with a significance level of p<0.05.

RESULTS

Baseline characteristics

Participant demographic information is presented in Table 1. The STAI-State scores increased from visit one to visit two (non-VMS: Δ2±1a.u, VMS: Δ2±1a.u, p=0.01) and this increase was similar in both groups (time × group interaction, p=0.73). Questionnaire data from the STAI-State and MENQOL obtained at visit two was used due to its temporal proximity to the main outcome measures. We confirmed there were no differences between participants previously and currently experiencing VMS in anthropometric, demographic and menopause-related variables (p>0.05). The VMS and non-VMS groups were similar in age (p=0.91), age of menopause (p=0.29) and BMI (p=0.44). As expected, the reported VMS severity in the MENQOL was higher in the VMS group than the non-VMS group (p<0.01). Vasomotor symptom and non-VMS groups were similar in past tobacco/nicotine use (p=0.12), current alcohol use (p=0.43) and STAI-State (p=0.53), STAI-Trait (p=0.44), PHQ-9 (p=0.84) and MLTPAQ (p=0.78) scores with a trend for greater PSQI (p=0.08) and ISI (p=0.053) scores in those with VMS. There were no group differences in FSH (p=0.10), estrone (p=0.10), estradiol (p=0.19), total estrogens (p=0.09), progesterone (p=0.35) and testosterone (p=0.47) levels.

Resting hemodynamic and autonomic data

There were no group differences in resting SBP (p=0.19), DBP (p=0.37), MAP (p=0.17, Table 2) or MSNA burst frequency (p=0.41; Figure 1a). Heart rate was lower in the VMS group than the non-VMS group (p=0.04, Table 2) while MSNA burst incidence was higher in the VMS group than the non-VMS group (p<0.01; Figure 1b). There were no group differences in sBRS (p=0.40), but a steeper total cBRS slope was observed (p=0.02, Table 2). The steeper cBRS slope was likely due to a steeper cBRSdown slope in the non-VMS group compared with the VMS group (p=0.03), and there was a trend of a steeper cBRSup slope in the non-VMS group compared with the VMS group (p=0.07). Both the SDRR (p<0.01, Table 2) and RMSSD (p<0.01, Table 2) HRV measures were higher in the VMS group than the non-VMS group, but high-frequency HRV percentage was not different between groups (p=0.34). Finally, there was no difference in BP SD, CV% or ARV for SBP (p=0.25; p=0.44; p=0.86) or DBP (p=0.73, p=0.63, p=0.97, respectively, Table 2).

Table 2.

Participant baseline hemodynamic and autonomic data.

	VMS	Non-VMS	p-value
SBP (mmHg)	124±14	129±13	0.19
DBP (mmHg)	80±11	82±11	0.37
MAP (mmHg)	94±11	98±10	0.17
HR (BPM)	58±10	62±9	0.04
MSNA BF (bursts/min)	30±8	28±7	0.41
MSNA BI (bursts/100hb)	53±13	44±9	<0.01
sBRS ((bursts/100hb)/mmHg)	-2.1±2	-1.7±1	0.40
Total cBRS (ms/mmHg)	12±6	9±3	0.02
cBRS up (ms/mmHg)	11±6	9±3	0.07
cBRS down (ms/mmHg)	12±6	9±4	0.03
SDRR (ms)	52±21	38±16	<0.01
RMSSD (ms)	39±21	25±13	<0.01
HFpower (ms2)	752±990	400±614	0.11
HFpower (%)	22±15	19±11	0.34
SBP standard deviation (a.u.)	7±3	7±5	0.25
SBP coefficient of variation (%)	5±2	6±3	0.44
SBP average real variability (a.u.)	2±0.4	2±0.6	0.86
DBP standard deviation (a.u.)	4±2	5±2	0.73
DBP coefficient of variation (%)	5±2	6±3	0.63
DBP average real variability (a.u.)	1±0.9	1±0.4	0.97

Definitions: VMS: vasomotor symptoms; SBP: systolic blood pressure; DBP: diastolic blood pressure; MAP: mean arterial pressure; HR: heart rate; MSNA: muscle sympathetic nerve activity; hb: heart beats; BF: burst frequency; BI: burst incidence; sBRS: sympathetic baroreflex sensitivity: cBRS: cardiovagal baroreflex sensitivity; standard deviation of RR intervals: SDRR; root mean square of successive RR interval differences: RMSSD; absolute power of the high-frequency band: HFpower (ms²); relative power of the high-frequency band: HFpower (%). Sample sizes: MSNA BF, MSNA BI, sBRS: VMS: n=31; non-VMS: n=15.

Figure 1:

Baseline differences in muscle sympathetic nerve activity at rest. Muscle sympathetic nerve activity burst frequency (MSNA BF; A) was similar between females with vasomotor symptoms (VMS; 30±8 mean±SD; bursts/min) and those without VMS (non-VMS; 28±7bursts/min). Muscle sympathetic nerve activity burst incidence (MSNA BI; B) was greater in females with VMS (53±13 bursts/ 100 heart beats (hb)) compared with females without VMS (44±9 bursts/ 100hb). *indicates a significant difference between groups.

Severity of VMS, hemodynamic and autonomic data

We conducted an exploratory analysis to determine if VMS were associated with hemodynamic or autonomic variables. We observed an inverse relationship between VMS severity and SBP SD (r=-0.45, p<0.01) and SBP CV (r=-0.42, p=0.01). No other significant relationships were observed (p>0.05).

Blood pressure and HR reactivity

Systolic BP, DBP, MAP and HR increased during the cold pressor test (time effect, p<0.01; Figure 2A, 2B, 2C and 2D, respectively), and these responses were not different between groups (time × group interaction, SBP: p=0.86; DBP: p=0.83; MAP: p=0.58; HR: p=0.52) with no main group effects in SBP (p=0.54), DBP (p=0.89) or MAP (p=0.69). Heart rate was lower throughout the cold pressor test in the VMS group compared with the non-VMS group (group effect, p<0.01). During recovery from the cold pressor test, SBP, DBP, MAP and HR all decreased (time effect, p<0.01; Figure 3A, 3B, 3C and 3D, respectively) similarly for both the VMS and non-VMS groups (time × group interaction, SBP: p=0.32; DBP: p=0.14; MAP: p=0.13; HR: p=0.39) with no main effect of group (SBP: p=0.89; DBP: p=0.27; MAP: p=0.47; HR: p=0.11).

Figure 2:

Changes in hemodynamic variables during the cold pressor test. Systolic BP (SBP; A), diastolic BP (DBP; B), mean arterial BP (MAP; C), and heart rate (HR; D) all increased significantly from baseline (BL) in the vasomotor symptom (VMS) and non-VMS groups (time effect; p<0.01). Heart rate was significantly higher in the non-VMS group compared to the VMS group throughout the cold pressor test (group effect; p<0.01). *indicates a significant increase from baseline in both groups as determined with a Bonferroni pairwise comparison.

Figure 3:

Changes in hemodynamic variables during recovery from the cold pressor test. Systolic BP (SBP; A), diastolic BP (DBP; B), mean arterial BP (MAP; C), and heart rate (HR; D) all decreased significantly from the end of the cold pressor test (CPT) in the vasomotor symptom (VMS) and non-VMS groups (time effect; p<0.01). *indicates a significant reduction from the last 30 sec of the CPT in both groups as determined with a Bonferroni pairwise comparison.

Muscle sympathetic nerve activity reactivity

Muscle sympathetic nerve activity burst frequency, burst incidence, total activity and burst area increased during the cold pressor test (time effect, p<0.01; Figure 4A, 4B, 4C and 4D respectively) similarly for both the VMS and non-VMS groups (time × group interaction, MSNA burst frequency: p=0.50; MSNA burst incidence: p=0.34; total MSNA: p=0.49; MSNA burst area: p=0.62), with no group effect (MSNA burst frequency: p=0.77; MSNA burst incidence: p=0.30; total MSNA: p=0.15; MSNA burst area: p=0.76). In the recovery from the cold pressor test, MSNA burst frequency, burst incidence, total MSNA and burst area decreased (time effect, p<0.01; Figure 5A, 5B, 5C and 5D, respectively) similarly for both the VMS and non-VMS groups (time × group interaction, MSNA burst frequency: p=0.58; MSNA burst incidence: p=0.51; total MSNA: p=0.69; MSNA burst area: p=0.25) with no group effect (MSNA burst frequency: p=0.81; MSNA burst incidence: p=0.26; total MSNA: p=0.45; MSNA burst area: p=0.82). There was no difference in the Δ MAP/ Δ burst frequency from baseline to the last epoch of the cold pressor test (non-VMS 0–120: Δ1±2a.u., VMS 0–120: Δ1±3a.u., p=0.98). Pain increased throughout the cold pressor test (non-VMS 0–120sec: Δ6±2a.u., VMS 0–120: Δ5±2a.u., time effect, p<0.01) similarly in both groups (time × group interaction, p=0.31).

Figure 4:

Changes in muscle sympathetic nerve activity variables during the cold pressor test. Muscle sympathetic nerve activity burst frequency (MSNA BF; A), burst incidence (MSNA BI; B), total MSNA and burst area (MSNA BA) all increased significantly from baseline (BL) in the vasomotor symptom (VMS) and non-VMS groups (time effect; p<0.01). *indicates a significant increase from baseline in both groups as determined with a Bonferroni pairwise comparison.

Figure 5:

Changes in muscle sympathetic nerve activity (MSNA) variables during recovery from the cold pressor test. Muscle sympathetic nerve activity burst frequency (MSNA BF; A), burst incidence (MSNA BI; B), total MSNA and burst area (MSNA BA) all decreased significantly during recovery from the cold pressor test (CPT) in the vasomotor symptom (VMS) and non-VMS groups (time effect; p<0.01). *indicates a significant reduction from the last 30 sec of the CPT in both groups as determined with a Bonferroni pairwise comparison.

Discussion

Given the previous literature suggesting greater CVD risk in postmenopausal females with VMS (6, 8), we expected to find elevated BP and sympathetic activity at rest and in response to the cold pressor test in participants with VMS compared with those who have never experienced VMS. Contrary to our hypothesis, we found no differences in BP between groups at rest or during the cold pressor test. Moreover, although the VMS group demonstrated elevated MSNA burst incidence at rest compared to the non-VMS group, this difference was likely due to the lower HR in this group. Indeed, MSNA burst frequency was not different between groups. Finally, postmenopausal females with VMS demonstrated 1) a lower HR at rest and throughout the cold pressor test than females who never experienced VMS, 2) elevated cardiovagal tone, evidenced by elevated HRV and 3) elevated cBRS, suggesting that autonomic dysfunction is not present in otherwise healthy females who currently experience or have previously experienced VMS.

Vasomotor symptoms of menopause and blood pressure

Vasomotor symptoms of menopause are often associated with elevated resting BP. In a recent review, we found that 66% of studies investigating BP changes associated with hot flushes demonstrated higher resting BP in females who experience hot flushes (12). For example, Gerber et al. observed that average waking and sleeping systolic BP values were elevated in females experiencing hot flushes and that hot flushes independently predicted BP after controlling for age, race/ethnicity, BMI and menopausal status (5). This study is consistent with a large-scale epidemiological analysis from the Study of Women Across the Nation (SWAN), which observed that the presence of VMS was the main predictor for the steeper rise in BP over the menopause transition (8). Further, Jackson et al. from SWAN reported a positive association between the frequency of self-reported hot flushes and BP (6). Thus, the majority of literature, but not all, suggests elevated resting BP in females with VMS. In contrast to these aforementioned studies, however, and consistent with our findings, several studies did not observe significant differences in BP between females with and without VMS (59–63). Our study not only supports those prior studies (59–63), but also extends them by demonstrating that even when exposed to a stressor, such as with a cold pressor test, the BP response was not exaggerated in females who experience or have previously experienced VMS. The discrepant findings across various studies are likely due, in part, to smaller sample sizes and more stringent eligibility criteria in the latter studies as well as the current study. All but one (5) of the studies with smaller sample sizes did not demonstrate a difference in BP between groups. Further, several studies had similar eligibility criteria to our own (excluding participants with high BMI, use of hypertension medication or diabetes (59–63)), while studies that reported a relationship between VMS and BP included these participants and covaried for those factors (5, 6, 10). Collectively, our study and other smaller-scale studies investigating VMS of menopause and BP may be excluding the participants most at risk for hypertension with more stringent exclusion criteria. Nevertheless, it remains unclear if VMS is a major contributing factor to CVD or if other confounding factors are playing a key role.

Vasomotor symptoms of menopause and autonomic function

Autonomic dysregulation can often lead to hypertension and may precede CVD development (15, 64). The imbalance between sympathetic and parasympathetic function has repeatedly been demonstrated in postmenopausal females when compared with premenopausal females (17, 30, 31, 65). For example, an infusion of norepinephrine causes an exaggerated BP response in postmenopausal females compared with young females (65) and MSNA explains a larger degree of variance in BP in postmenopausal females compared with young females (30). This suggests that autonomic support of BP is greater for females after menopause. The decline of estradiol over the menopause transition likely plays a role in the greater autonomic support of BP, as estradiol reduces the sympathetic control over BP both directly (66) and via increases in nitric oxide synthase activity and availability (16). The loss of estradiol with menopause is also implicated in hot flushes and night sweats, but whether there is a link between menopause-associated VMS and autonomic dysregulation is understudied.

Although hot flushes are considered sympathoexcitatory (25, 67, 68), there is a dearth of research available that fully characterizes the mechanism of and physiological response to VMS. A study by Low et al. demonstrated an increase in skin sympathetic nerve activity and skin blood flow, predominantly neurally mediated via sympathetic cholinergic nerves, during a hot flush (69). Studies have also found that during hot flushes there are increases in indirect measures of autonomic activity, such as plasma levels of either epinephrine or norepinephrine (25, 26) and a decrease in high-frequency HRV, a reliable measure of parasympathetic function (70), which indicate a heighted sympathetic drive and attenuated vagal tone (68), respectively. Previous research has also demonstrated that higher levels of hot flush bother are associated with increased resting levels of plasma norepinephrine (67). Collectively, this evidence suggests that hot flushes are sympathoexcitatory, but to date, it was unknown if otherwise healthy females with hot flushes demonstrate elevated resting sympathetic activity and/or greater sympathetic reactivity with stress compared with females who have never experienced a hot flush.

The current study employed the gold-standard measure of sympathetic function in humans, MSNA (28, 71), and found no differences at rest between healthy postmenopausal females who previously experienced or currently experience VMS and those who have never experienced VMS. While we observed that females with VMS had higher MSNA burst incidence compared to those without VMS, this result is likely a function of the lower HR in females with VMS as noted by the lack of group differences in MSNA burst frequency. Alternatively, the increased burst incidence in the VMS group despite the lower HR could be the result of greater sympathetic outflow, which did not result in greater sympathetic transduction or changes in BP (72). Sympathetic reactivity during the cold pressor test, a reliable stressor known to raise BP in part by increasing MSNA (73), was also unaltered in females with VMS. Further, there was no difference in sympathetic transduction from baseline to the cold pressor test in females with and without VMS. Previously, our work demonstrated exaggerated sympathetic reactivity to a stressor in postmenopausal females compared with age-matched males and younger males and females (17). Our current findings demonstrate that the greater increase in sympathetic reactivity experienced by postmenopausal females is not further exacerbated in females with a history of VMS. Consistent with our results, others have also demonstrated that females with VMS have similar sympathetic activity, estimated via impedance cardiography, to females without VMS (74).

Unexpected findings in our study were the lower HR, greater total cBRS and cBRSdown, and elevated HRV in the time domain at rest and/or during the stressor in the VMS group compared with the non-VMS group. The baroreflex is a neural reflex by which the autonomic nervous system modulates BP. A greater total cBRS, driven by cBRSdown, in the VMS group indicates that the baroreflex is more sensitive to a falling-BP stimulus than in females without VMS. The greater cBRSdown could be a compensatory response to chronic hot flushes, such that with an acute drop in BP that occurs in some hot flushes (75), the baroreflex would cause a consequential increase in sympathetic activity to return BP to the operating set point. The occurrence of this drop in BP with VMS may depend on the severity of the hot flush, but this response and associated mechanisms have yet to be investigated. Although cBRS was greater in females with VMS, we did not observe differences in BPV between the groups. There were however, significant, albeit weak, inverse relationships between the variability in SBP and VMS in the VMS group, suggesting that those who were more bothered by their symptoms had lower SBP variability. Females with VMS also demonstrated lower HR and greater HRV compared with the non-VMS group. While previous studies have either found no differences in HR between females with and without VMS (59–61) or significantly higher HR in females with VMS than without VMS (62), this is the first study to observe a lower HR in females with VMS in comparison to females without VMS. Although the reason for the lower HR and elevated HRV is unknown, it is not due to the VMS group being more physically active as reported physical activity levels were similar between the groups. The greater HRV in the current study is corroborated by Gibson et al., who found that females with moderate-to-severe hot flushes, compared with less frequent flushers, trended (p=0.06) towards higher respiratory sinus arrhythmia (74), a measure associated with parasympathetic activation. In addition, they demonstrated a positive association between the change in respiratory sinus arrhythmia and change in frequency of moderate-to-severe hot flashes over 12 weeks. The authors hypothesized that the greater vagal tone in the hot flush group may be related to a heightened interoceptive awareness (76) (i.e. greater parasympathetic activity in symptomatic females who have a greater ability to sense the hot flush).

Strengths and limitations

Strengths of the current study include rigorous inclusion/exclusion criteria and that this is the first study to use gold-standard techniques to directly measure sympathetic outflow to the skeletal muscle vasculature (28, 29), which provides strong confidence that in healthy females who experience VMS, neither sympathetic activity nor BP are elevated. Our study is not without limitations, however. First, we still lack an understanding of any potential changes in sympathetic activity during a hot flush episode. Future research should investigate if sympathetic regulation of BP is altered during a hot flush. Second, we did not document the age when participants began to experience VMS or have additional knowledge into the frequency or severity of hot flushes in previous flushers, both of which may provide insight to hot-flushers who might be at greater CVD risk (13, 77–80). While there is evidence that self-reported information on VMS is subject to recall bias (81), future longitudinal studies should explore the relationship between the timing of VMS, BP and sympathetic activity. In addition, our current cohort experienced mild to moderate severity of VMS and thus, it is possible that only more severe VMS may contribute to autonomic dysfunction. Next, while our rigorous inclusion/exclusion provides confidence in our results, our findings may be less generalizable to the overall postmenopausal population. Additional studies should explore if VMS may increase sympathetic activity or BP in females who have pre-existing CVD risk factors, such as obesity or diabetes. Finally, our study population primarily consisted of educated European American/white females. Because the prevalence of VMS varies among different racial and ethnic groups—for example, African-Americans are more likely to have more frequent and severe VMS than other populations (82) and are at a greater risk for CVD (83)—it is imperative that future work include a more racially diverse cohort to fully explore mechanisms linking VMS and disparities in CVD risk.

Beyond methodological limitations within our study, there are associative factors of menopause that may contribute to the relationship between VMS of menopause and CVD. Because ~80% of females experience VMS (13) and VMS are associated with greater CVD risk, it is difficult to determine if this greater risk is directly related to VMS or if VMS are simply associated with CVD as a result of being a prominent menopause symptom. Further, VMS is often comorbid with sleep difficulty (84) and depression (85), both of which are linked to CVD (86). Although both of our groups had similar self-reported sleep scores, anxiety and depression, future research would benefit from isolating the contribution of each of these factors of menopause on CVD development and risk in females.

CONCLUSION

In conclusion, neither sympathetic activity nor BP were elevated at rest or during a stressor in postmenopausal females who previously or currently experience VMS. In addition, postmenopausal females with VMS demonstrated a lower HR, higher total cBRS, cBRSdown and HRV compared to females without VMS. While the physiological significance of this paradoxical finding is unclear, it may be a conditioned response associated with frequent transient BP reductions during VMS episodes or differences in interoceptive awareness; however, these theories need to be tested in future research.