The severity of vasomotor symptoms and number of menopausal symptoms in postmenopausal women and select clinical health outcomes in the Women’s Health Initiative Calcium and Vitamin D randomized clinical trial

Abstract

Objective:

This study evaluated whether vasomotor symptom (VMS) severity and number of moderate/severe menopausal symptoms (nMS) were associated with health outcomes, and whether calcium and vitamin D (CaD) modified the risks.

Methods:

The Women’s Health Initiative CaD study was a double blind, randomized, placebo-controlled trial, which tested 400 IU of 25-hydroxyvitamin-D and 1,000mg of calcium per day in women aged 50 to 79 years. This study included 20,050 women (median follow-up of 7 y). The outcomes included hip fracture, colorectal cancer, invasive breast cancer, all-cause mortality, coronary heart disease, stroke, cardiovascular death, and total cardiovascular disease (CVD). MS included: hot flashes, night sweats, dizziness, heart racing, tremors, feeling restless, feeling tired, difficulty concentrating, forgetfulness, mood swings, vaginal dryness, breast tenderness, migraine, and waking up several times at night. Associations between VMS severity and nMS with outcomes were tested.

Results:

No association between VMS severity and any outcome were found. In contrast, nMS was associated with higher stroke (hazard ratio [HR] 1.40 95% confidence interval [CI] 1.04–1.89 for ≥ 2 MS vs none; HR 1.20 95% CI 0.89–1.63 for 1 MS vs none, P trend=0.03) and total CVD (HR 1.35, 95% CI, 1.18–1.54 for ≥ 2 MS vs none; HR 0.99, 95% CI, 0.87–1.14 for 1 MS vs none P trend < 0.001). CaD did not modify any association.

Conclusion:

Severity of VMS was not associated with any outcome. Having ≥2 moderate or severe MS was associated with an increased risk for CVD. The number of moderate/severe MS may be a marker for higher CVD risk.

Video Summary: http://links.lww.com/MENO/A669.

Women typically experience many menopausal symptoms (MS), other than the core vasomotor symptoms(VMS),which consist of hot flashes and night sweats. These other symptoms often include breast tenderness, vaginal dryness, aches and pains, as well as dizziness, heart racing or skipping beats, tremors, feeling restless or fidgety, feeling tired, difficulty concentrating, forgetfulness, mood swings, headache or migraine, and waking up several times at night.^1–11 Studies indicate that 63% to 70% of women experience MS at some time point during menopause^12,13 and they may persist for long periods of time. In the Penn Ovarian Aging study, the median duration of MS was 10.2 years and African American women were more likely to have longer durations of symptoms compared to White women.¹⁴

The Women’s Health Initiative (WHI) clinical trials have previously investigated the influence of menopausal hormone therapy (HT), low-fat dietary intervention, and calcium and vitamin D (CaD) supplementation on VMS^7–9 and other MS.^{7–9,15–18} A report from the WHI-CaD trial found that women randomized to CaD supplementation did not have fewer MS over an average of 5.7 years of follow up.¹⁵ In that study, MS included hot flashes, night sweats, dizziness, heart racing or skipping beats, tremors, feeling restless or fidgety, feeling tired, difficulty concentrating, forgetfulness, mood swings, vaginal dryness, breast tenderness, headaches or migraines, waking up several times at night and waking earlier than planned. In a subsample of 520 women from the WHI-CaD trial with baseline 25-hydroxyvitamin D (25[OH]D) levels, there was no association between plasma 25(OH)D concentrations and number of MS.¹⁹ Little is known about associations of VMS and in particular the breadth of MS with health outcomes. Data suggest that VMS during menopause may be associated with a woman’s risk of cardiovascular disease (CVD). WHI studies assessing the associations between MS and CVD are limited to only VMS. An analysis from the WHI observational cohort found that women with VMS only at WHI enrollment (considered to be “late” or many years after menopause onset) were at an increased risk of coronary heart disease (CHD) and all-cause mortality. Interestingly, women who self-reported VMS only during early perimenopause were at decreased risk of stroke, total CVD events, and all-cause mortality.¹³

The results of CaD supplementation on various outcomes in the WHI have been reported previously.^20,21 During the intervention period of the WHI-CaD trial, CaD had a modest reduction in all-cause mortality; however this finding failed to meet statistical significance (Hazard ratio [HR] 0.91 [95% CI 0.83–1.01]).²⁰ Furthermore, CaD supplementation had no effect on CHD or stroke risk.²² A recent study reported the cumulative findings of the WHI-CaD trial intervention period and 5-year postintervention period. There was no difference in risks of hip fracture, colorectal cancer, CVD, breast cancer, or all-cause mortality rates after an average of 11 total years of followup.²²To the best of our knowledge, however, the relationship between VMS severity and various health outcomes, and whether CaD supplementation modifies these associations, have not yet been studied. There could also be an interaction between vitamin D and MS. In animal studies, vitamin D deficiency was found to be associated with decreased levels of serotonin in the brain,²³ a neurotransmitter shown to decline with menopause and associated with hot flashes.²⁴ Also, in the estrogen alone WHI-HT trial, the increased risk of stroke among women randomized to estrogen was attenuated for women also randomized to CaD, suggesting an interaction between vitamin D and estrogen.²⁵ Similarly, it is also important to investigate the combined influence of all MS, beyond just VMS. The WHI-CaD trial offers an ideal opportunity to address these questions. The objectives of this study were (1) to investigate whether the severity of VMS (none, mild, moderate or severe) and number of moderate or severe MS (nMS) were associated with various health outcomes including hip fracture, invasive breast cancer, colorectal cancer, all-cause mortality, CaD global index (composite endpoint of the aforementioned events), CHD, stroke, cardiovascular death, and total CVD. (2) Assess whether CaD supplementation modifies these risks. We hypothesized that women with greater VMS severity and greater nMS, determined at the baseline of the CaD trial, would have a higher risk of all-cause mortality, stroke, cancer and CVD risk and that CaD supplementation would modify these risks in a favorable direction. Given conflicting data on the risks and benefits of CaD supplementation, identifying a population of women who would most benefit from CaD supplementation is important.

METHODS

Trial information and participant selection

The WHI CaD trial was a double-blind, randomized, placebo-controlled study designed to test the effects of calcium (1,000mg of elemental calcium as calcium carbonate) and vitamin D (400IU/day of 25[OH]D) supplementation on multiple health outcomes in postmenopausal women aged 50 to 79 years (n=36,282). In the WHI CaD trial, women included were randomized mostly a year earlier to HT and/or the dietary modification trial. Median follow-up was 7 years. The eligibility criteria in the CaD trial included many safety parameters (eg, no previous hypercalcemia or renal calculi) and no competing risk indicators (eg, no medical condition associated with survival of less than 3 y). Women were able to participate even if they were taking their own supplemental calcium and/or vitamin D, provided their personal vitamin D supplementation did not exceed 600 IU (later changed to 1,000 IU/d). In addition, women were allowed to initiate nonstudy HT through their health care provider during follow-up. A woman was considered postmenopausal if she experienced no vaginal bleeding for at least 6 months, had a hysterectomy, or had ever used postmenopausal hormones. The protocol specified primary outcome for the CaD trial was hip fracture, and colorectal cancer was the designated secondary outcome. For this post hoc analysis, all participants also randomized in the HT trial were excluded (n=16,089) given the influence HT has on MS including VMS. Therefore, 20,193 of the original 36,282 women in the WHI-CaD trial were considered for inclusion; 143 women had missing VMS data. This left 20,050 participants available for analysis (20,050=36,282 – 16,089 – 143); no other exclusions were made.

Menopausal symptom and baseline characteristics collection

Data on VMS and other MS were by self-reported questionnaires at CaD trial baseline. The questionnaire included a checklist of MS based primarily on the Postmenopausal Estrogen/Progestin Interventions symptom tool.²⁶ The following 14 symptoms, typically associated with menopause,^1–4,26 were considered: hot flashes, night sweats, dizziness, heart racing or skipping beats, tremors, feeling restless or fidgety, feeling tired, difficulty concentrating, forgetfulness, mood swings, vaginal dryness, breast tenderness, headache or migraine, and waking up several times at night. Participants were asked if they experienced a MS within the prior 4 weeks. With the sole exception for waking up, symptoms were reported using a 4-item Likert scale with the following selections: (0) no symptoms; (1) mild (symptom did not interfere with usual activities); (2) moderate (symptom interfered somewhat with usual activities); (3) severe (symptom was so bothersome usual activity could not be performed). For the severity variable, we collapsed into the following categories none, mild, and moderate/severe. Moderate and severe were grouped together as both interfered with usual activities. ‘Did you wake up several times?’ was recorded on a 5-item Likert scale: (0) No not in the past 4 weeks; (1) Yes, less than once a week; (2) Yes, 1 or 2 times a week; (3) Yes, 3 or 4 times a week; (4) Yes, 5 or more times a week; but recoded to the predominant 4-item scale by combining sleep responses (2) and (3). Therefore, participants with at least weekly sleep disturbances were classified as moderate/severe. For the nMS variable, we then categorized women as having two or more, one, or no moderate to severe symptoms. Randomization into the CaD trial occurred at year 1 (91%, n=33,070) or year 2 (9%, n=3,212) after enrollment, and symptoms were ascertained at enrollment and year 1. CaD baseline refers to the most proximal self-report of MS at or before the start of the CaD trial. Baseline demographics, medical history, and symptoms were obtained via self-administered questionnaires. Physical measurements including weight, height, and blood pressure were measured using standardized procedures at annual clinic visits.

Health outcomes

The primary outcomes of this post hoc analysis were hip fracture and colorectal cancer, which coincide with the most recent update of the CaD trial.^20,22 Secondary outcomes included invasive breast cancer, all-cause mortality, and the global index (a summary measure of the overall risks and benefits of CaD supplementation; includes hip fracture, invasive breast cancer, colorectal cancer and death). Additional outcomes (other outcomes of interest) included CHD, stroke, cardiovascular death, and total CVD. Total CVD was a composite outcome including CHD, stroke, congestive heart failure, angina, peripheral vascular disease, carotid artery disease, and coronary revascularization. The process for ascertaining and adjudicating the health outcomes in the WHI has been previously described.²⁷ Outcomes were assessed until study termination on March 31, 2005.

Data and statistical analysis plan

HRs summarized the association between VMS and health outcomes, using Cox proportional hazards regression models with time from CaD randomization as the time-scale. Baseline hazard functions were stratified by: age groups, dietary trial randomization status, prior disease history, race/ethnicity and hysterectomy status with covariate adjustment for education, smoking, bilateral oophorectomy, personal menopausal HT use (never/past/current; main effect and interaction term with hysterectomy status), age (linear), BMI (linear), statin use, history of high cholesterol, aspirin use, use of antihypertensive medication, treated diabetes, family history of MI, systolic and diastolic blood pressure (BP). Age and clinical measurements including, BMI and BP, were the most proximal before CaD randomization to account for time elapsed between WHI enrollment and CaD randomization, mean (SD)=1.1 (0.3) years, and the favorable influence of the dietary intervention. Prior disease history included events that occurred during this time and in addition to history reported at enrollment. Remaining covariates were collected at enrollment and were chosen a priori as standard disease risk factors and factors typically associated with VMS. To assess whether CaD supplementation modified the association between VMS and clinical outcomes a VMS x CaD interaction regression term was added to the model with additional stratification by CaD randomization group. Statistical significance for the interaction between VMS overall and CaD was based on a 1 degree of freedom test for trend and interaction. The interaction between nMS and CaD randomization used a similar approach. A sensitivity analysis, excluded ‘waking up’, the sole MS collected on a 5-point Likert scale, from computation of nMS. Also, when analyzing the nMS and the total CVD outcome, the results were stratified by those with no known prior CVD and those with a prior diagnosis of CVD and stratified by prior use of HT (never vs past/current use). Lastly, exploratory analyses examined the associations between each MS individually and the total CVD outcome. P values were not adjusted for multiple comparisons, unless described in the narrative as being Bonferroni-corrected. Therefore, P values should be interpreted cautiously and up to one false positive is expected for the overall effect of VMS and nMS on clinical outcomes (= 2×9×0.05). All statistical analyses were conducted using SAS software version 9.4 (SAS Institute Inc, Cary, NC) and R software version 3.4 (R Foundation for Statistical Computing).

RESULTS

Women with more severe VMS tended to be younger and were more likely Black/African American, obese, and generally have more medical conditions (see Table, Supplemental Digital Content 1, http://links.lww.com/MENO/A663). Similarly, women reporting two or more moderate or severe MS tended to be younger, Black/African American, obese, and generally have more medical conditions (Table 1). Among all participants in this analysis, 7.4% had prior CVD and 54% reported current HT use. Although moderate or severe symptoms tended to be uncommon, mild symptoms were prevalent; as an example, only 3% reported moderate or severe dizziness, and 16% reported mild symptoms (see Table, see Supplemental Digital Content 2, http://links.lww.com/MENO/A664).

TABLE 1.

Participant characteristics^a by no. of moderate or severe menopausal symptoms at CaD baseline (n=19,337^b)

Characteristic	2+ (N=6,760) N%	1 (N=7,158) N%	None (N=5,419) N%	P valuec
Age at CaD randomization, y, mean i SD	63.0±6.9	64.0±6.7	63.0±6.5	0.64
Age at CaD randomization, 10 y groups
50–59	2,380 (35.2)	2,056 (28.7)	1,828 (33.7)	0.007
60–69	3,066 (45.4)	3,517 (49.1)	2,597 (47.9)
70+	1,314 (19.4)	1,585 (22.1)	994 (18.3)
Race/ethnicity, %				<0.001
Non-Hispanic White	5,557 (82.2)	6,264 (87.5)	4,557 (84.1)
Non-Hispanic Black/African American	726 (10.7)	473 (6.6)	460 (8.5)
Hispanic/Latina	247 (3.7)	182 (2.5)	163 (3.0)
American Indian/Alaskan Native	33 (0.5)	20 (0.3)	20 (0.4)
Asian/Pacific Islander	110 (1.6)	150 (2.1)	160 (3.0)
Unknown	87 (1.3)	69 (1.0)	59 (1.1)
Education > high school degree or GED	5,217 (77.6)	5,722 (80.4)	4,424 (82.1)	<0.001
Smoking status				<0.001
Never	3,417 (51.0)	3,753 (53.0)	2,944 (54.9)
Past	2,857 (42.7)	2,961 (41.8)	2,069 (38.6)
Current	423 (6.3)	369 (5.2)	345 (6.4)
Hysterectomy	3,352 (49.6)	3,057 (42.7)	2,147 (39.6)	<0.001
Bilateral oophorectomy	1,603 (24.2)	1,454 (20.7)	1,111 (20.8)	<0.001
HT use status				<0.001
Never used	2,047 (30.3)	2,521 (35.2)	1,990 (36.8)
Past user	890 (13.2)	815 (11.4)	531 (9.8)
Current user	3,815 (56.5)	3,817 (53.4)	2,891 (53.4)
Treated diabetes (pills or shots)	298 (4.4)	220 (3.1)	167 (3.1)	<0.001
Hypertensiond	2,960 (47.7)	2,966 (44.9)	2,085 (42.1)	<0.001
History of high cholesterol requiring pillse	873 (12.9)	786 (11.0)	516 (9.5)	<0.001
Statin use	469 (6.9)	433 (6.0)	296 (5.5)	<0.001
Anti-hypertensive medication	2,364 (35.0)	2,144 (30.0)	1,399 (25.8)	<0.001
Aspirin use	1,306 (19.3)	1,373 (19.2)	901 (16.6)	<0.001
Family history of MI	3,472 (54.4)	3,538 (51.8)	2,550 (49.2)	<0.001
Prior history of CVDf	712 (10.5)	455 (6.4)	247 (4.6)	<0.001
Body-mass index (kg/m2), median [IQR]	28.5 [7.7]	27.7 [7.2]	27.3 [7.0]	<0.001
Waist circumference (cm), mean ± SD	89.7±13.8	88.0±13.4	86.6±13.2	<0.001
Systolic BP (mm Hg), mean±SD	126.9±17.0	127.2±17.1	126.5±16.9	0.23
Diastolic BP (mm Hg), mean±SD	75.9±9.0	75.6±9.0	75.6±9.0	0.09
Recreational physical activity (MET-h/wk), median [IQR]	5.3 [12.3]	7.5 [14.6]	7.5 [14.4]	<0.001

Summary statistics correspond to N (%) of participants, mean ± SD, or median [IQR] for skewed variables (skew > 1)

BP, blood pressure; CVD, cardiovascular disease; GED, General Educational Development; HT, hormone therapy; MI, myocardial infarction; SD, standard deviation; VMS, vasomotor symptom; WHI, Women’s Health Initiative.

^aParticipant characteristics were measured at WHI enrollment exclusively, except for age.

^b713 of 20,050 participants were missing data on symptoms besides VMS.

^cP values corresponds to a trend test for the association between no. of moderate or severe menopausal symptoms and participant characteristic; computed from a cumulative logit model, with no. of moderate or severe menopausal symptoms as the ordinal response, adjusted for age and race/ethnicity.

^dSelf-report of ever taking medication for high blood pressure, or clinic-measured systolic/diastolic blood pressure ≥140/90 mm Hg.

^eSelf-report or use of anti-hyperlipidemic medications at enrollment.

^fIncludes MI, stroke, revascularization, carotid endarterectomy/angioplasty, angina, peripheral vascular disease and congestive heart failure.

Overall, there was no significant association between VMS severity and any clinical health outcome and findings did not significantly vary with respect to CaD randomization status (Figure 1). In contrast, nMS was associated with stroke (1moderate/severe MS vs none; HR 1.20, 95% CI 0.89–1.63, 2+ moderate/severe MS vs none; HR 1.40 95% CI 1.04–1.89, P=0.03) and total CVD (1moderate/severe MS vs. none; HR 0.99, 95%CI 0.87–1.14, 2+ mod/severe MS vs none; HR 1.35 95% CI 1.18–1.54, P<0.001) but not CHD (1moderate/severe MS vs none; HR 0.91, 95% CI 0.71–1.17, 2+ moderate/severe MS vs none; HR 1.10 95 CI 0.86–1.41, P=0.35; Figure 2). Bonferroni-correction for the nine endpoints considered in this report yielded P values of 0.24 and<0.001 for stroke and total CVD, respectively. The nMS was not associated with any other health outcome (Figure 2).

FIG. 1.

Multivariable adjusted hazard ratios associating VMS severity to clinical outcomes during the CaD trial. Summary statistics are from a Cox proportional hazards model where baseline hazard functions were stratified by age group, dietary trial randomization status, prior disease history, race/ethnicity and hysterectomy status with covariate adjustment for education, smoking, bilateral oophorectomy, menopausal HT use (never/past/current; main effect and interaction term with hysterectomy status), age (linear), BMI (linear), statin use, history of high cholesterol, aspirin use, use of anti-hypertensive medication, treated diabetes, family history of MI, systolic BP (linear), and diastolic BP (linear). Nominal P values corresponds to a test of trend. BMI, body mass index; BP, blood pressure; HT, hormone therapy; MI, myocardial infarction; VMS, vasomotor symptom.

FIG. 2.

Multivariable adjusted hazard ratios associating the number of moderate or severe menopausal symptoms to clinical outcomes during the CaD trial. Fourteen menopausal symptoms were considered: hot flashes, night sweats, dizziness, heart racing or skipping beats, tremors, feeling restless or fidgety, feeling tired, difficulty concentrating, forgetfulness, mood swings, vaginal dryness, breast tenderness, headache or migraine, and waking up several times at night. Summary statistics are from a Cox proportional hazards model described in Figure 1. Nominal P values corresponds to a test of trend.

In a sensitivity analysis (see Figure, Supplemental Digital Content 3, http://links.lww.com/MENO/A665) excluding the sleep variable (MS from the same Likert scale) findings were similar. In addition, the association between nMS and total CVD was similar among those with prior history of CVD (1moderate/severe MS vs none; HR 0.96, 95% CI 0.69–1.35, 2+ moderate/severe MS vs none; HR 1.23 95% CI 0.90–1.68) and those without prior CVD (1moderate/severe MS vs none; HR 1.00, 95% CI 0.86–1.16, 2+ moderate/severe MS vs none; HR 1.37 95% CI 1.19–1.59; P interaction=0.65). The association between nMS and total CVD was similar among those that never used HT (1moderate/severe MS vs none; HR 1.09, 95% CI 0.88–1.35, 2+ moderate/severe MS vs. none; HR 1.39 95% CI 1.12–1.72) and those that reported past or current HT use (1moderate/severe MS vs none; HR 0.96, 95% CI 0.80–1.14, 2+ moderate/severe MS vs none; HR 1.31 95% CI 1.101.55; P interaction=0.85). Prior menopausal HT use did not modify the association between nMS and the remaining clinical endpoints (see Figure, Supplemental Digital Content 4, http://links.lww.com/MENO/A666). Randomization to CaD did not modify the association between nMS and total CVD (HR 1.36; 95% CI 1.13–1.64 vs HR 1.34; 95% CI 1.11–1.62), P interaction=0.82) for participants randomized to CaD versus placebo, respectively (see Figure, Supplemental Digital Content 5, http://links.lww.com/MENO/A667). CaD did not modify the association between nMS and any other health outcome, nor did CaD modify the association between VMS and any other outcome (see Figure, Supplemental Digital Content 6, http://links.lww.com/MENO/A668).

Exploratory analyses (Figure 3) examined the association between the severity of each MS component individually and total CVD. The severity of VMS was not associated with any health outcome, and neither was the individual components of hot flashes (P=0.95) or night sweats (P=0.06). However, moderate and severe night sweats trended toward a significant increased risk of total CVD (HR 1.28 95% CI 1.05–1.56), whereas mild night sweats were not associated with an increased risk of total CVD (HR 1.02 95% CI 0.90–1.16). The severity of many individual MS (ie, dizziness, heart racing or skipping beats, feeling tired, forgetfulness) were strongly associated (P<0.001) with total CVD. In addition, the severity of being restless and fidgety, having headaches or migraines and having difficulty concentrating were associated with total CVD. The prevalence of mild symptom severity for several of these significant MS, exceeded 15% with the 95% CI for the mild versus none contrast above unity (see Table, Supplemental Digital Content 2, http://links.lww.com/MENO/A664; Figure 3 left panel).

FIG. 3.

Multivariable adjusted hazard ratios associating the severity of individual menopausal symptoms to total cardiovascular disease (composite of coronary heart disease, stroke, congestive heart failure, angina, peripheral vascular disease, carotid artery disease, and coronary revascularization) during the CaD trial. Summary statistics are from a Cox proportional hazards model described in Figure 1. Nominal P values corresponds to a test of trend.

DISCUSSION

Among postmenopausal women, severity of VMS within 4 weeks of trial baseline was not associated with any clinical event, and CaD randomization status did not modify any associations. Having two or more moderate or severe MS was significantly associated with increased CVD and stroke, and this association was also not modified by CaD randomization status. An exploratory analysis in this cohort yielded positive associations between the severity of individual menopause related symptoms (including being restless and fidgety, difficulty concentrating, dizziness, heart racing or skipping beats, feeling tired, and forgetfulness) and the incidence of total CVD (CHD, stroke, congestive heart failure, angina, peripheral vascular disease, carotid artery disease, and coronary revascularization). For certain MS, having mild symptoms were not associated with an increased risk of total CVD, whereas having moderate or severe symptoms were associated with an increased risk of total CVD. These symptoms included being restless or fidgety, feeling tired, and difficulty concentrating. This data suggests the type of moderate or severe MS rather than absolute nMS is the more accurate marker of CVD and stroke risk.

Our findings are consistent with a recent study consisting of 14,340 symptomatic menopausal women without a prior history of CHD from a Taiwan National Health Insurance Research Database. Compared to asymptomatic women, those with symptoms had a higher risk of developing CHD (17 vs 12 events per 1,000 patient years, HR 1.3 95% CI 1.3–1.4).¹⁰ Pathophysiologic mechanisms exist to explain the link between VMS and other MS and CVD. VMS severity has been associated with vascular stiffness,²⁸ endothelial dysfunction,²⁹ inflammatory markers such as interleukin-8 and tumor necrosis factor alpha.^30,31 Poor sleep and depression have also been linked to proinflammatory cytokines which may lead to the development of CVD.^32,33 Furthermore, VMS have been linked to risk factors for CVD development including high BP,³⁴ diabetes,³⁵ and abnormal lipids including higher low-density lipid cholesterol, apolipoprotein B, and triglyceride concentrations.³⁶ A study with 168 women aged 45 to 60 years suggested that self-reported hot flashes and night sweats were associated with lower cardiac index, stroke volume index, higher overall vascular resistance index, and higher concentrations of the inflammatory cytokine, interleukin-6.³⁷ A recent study of 138 healthy women across all stages of menopause (premenopausal, early or late perimenopausal and late postmenopausal) found that arterial stiffening and endothelial dysfunction were associated with more frequent and severe MS.²⁹ We hypothesize that severe MS may lead to a decrease in activity, exercise, sleep, and hence an increase in risk of CVD and stroke. Alternatively, it is possible that poor vascular flow due to subclinical CVD and atherosclerosis may lead to the development of MS.

Although the WHI has reported on associations between VMS and CVD events,¹³ coronary artery calcification,⁵ thrombosis,⁶ diabetes,⁷ fractures,⁸ and breast cancer,⁹ the previous studies had not evaluated other MS. Previously, an association between moderate/severe VMS and fracture was observed in WHI clinical trial participants, however in the current analysis there was no association between number of moderate/severe MS and fracture. Several WHI studies have investigated whether other symptoms that may have overlap with MS including depressive symptoms,³⁸ sleep disturbances,³⁹ migraine,⁴⁰ panic attacks,⁴¹ and cognitive impairment⁴² are associated with the development of CVD risk factors and CVD outcomes. This analysis adds to prior WHI reports as the severity of VMS was analyzed for associations with various health outcomes. In addition, the association between the number of moderate and severe MS and health outcomes were examined.

Conflicting data exists regarding the associations of VMS with subclinical CVD. Among, 918 women from the WHI estrogen-alone HT trial, a history of any VMS was associated with decreased coronary artery calcium score compared to those with no VMS (odds ratio 0.66, 95% CI 0.45–0.98).⁵ WHI did not have a similar study in the estrogen and progestin trial as comparison. However, an analysis from the Study of Women’s Health Across the Nation found that VMS were associated with lower brachial artery flow-mediated dilation and greater coronary and aortic calcification.⁴³ In the Study of Women’s Health Across the Nation, frequent baseline VMS (≥6 d) compared to no VMS was associated with a higher risk of CVD (HR 1.62, 95% CI 1.10–2.38).⁴⁴ Among 868 recently menopausal women screened for The Kronos Early Estrogen Prevention Study trial, baseline MS (hot flashes, dyspareunia, vaginal dryness, night sweats, palpitations, mood swings, depression, insomnia, irritability) were not associated with coronary artery calcium scores.¹¹

There are limitations of the current analysis that warrant consideration. In this cohort, there was a wide age range of postmenopausal women (50–79 years), therefore there was variability between enrollment in the WHI and age at menopause onset. Also, whereas VMS (hot flashes and night sweats) are considered hallmarks of menopause, it is unclear whether other MS are truly more common during menopause and directly related to menopausal hormone changes. It is always possible that MS may be associated with other characteristics not adequately controlled for and that residual confounding is possible. Therefore, caution should be used when interpreting the results and our conclusions should be considered hypothesis generating. Another limitation worth discussing is the assessment of MS. MS were assessed at one time point (study baseline) and not assessed later in the follow-up. Therefore, we are unable to determine the differences between persistent versus isolated MS and associations with the health outcomes studied.

CONCLUSIONS

We found no association between severity of VMS and any major health outcomes including total CVD, hip fracture, colorectal cancer, invasive breast cancer, all-cause mortality, CaD global index, CHD, stroke, and CVD death among the postmenopausal women we analyzed. However, the number of moderate or severe MS was associated with more CVD and stroke. Our results suggest that women with more moderate or severe MS may be at increased risk of CVD and stroke. It is unknown whether treating CVD will reduce MS or vice versa. Women with more severe MS may be more likely to see a physician for treatment. Hence, given the association of MS with CVD and stroke, this may be an ideal opportunity to assess their CVD risk.

Supplementary Material

Supplemental digital content is available for this article. Direct URL citations are provided in the HTML and PDF versions of this article on the journal’s Website (www.menopause.org).