Relative Influence of Menopausal Status, Age, and Body Mass Index on Blood Pressure
Abstract The objective of this study was to assess the influence of menopausal status on blood pressure levels in a cross-sectional population study in the setting of a preventive health care program clinically conducted in Ferrara, Italy. The patients were 2397 healthy women 35 to 65 years old (489 in premenopause, 847 in perimenopause, 887 in spontaneous menopause, and 174 in surgical menopause); subgroups were also obtained with increasing duration of menopause (1 to >5 years) matched with premenopausal and perimenopausal women by chronological age at onset of menopause. Measurements of sphygmomanometric blood pressure, age, age at menopause, duration of menopause, and body mass index were made. Postmenopausal women had higher blood pressure than premenopausal and perimenopausal subjects. After adjustment for body mass index, the blood pressure changes with menopausal status were still significant, but not after correction by age. The increased risk of hypertension of postmenopausal women also lost statistical significance after adjustment for age and body mass index. At multivariate analysis, blood pressure showed no significant relation with the duration of menopause, whereas age was a significant covariant for systolic blood pressure, and body mass index for systolic and diastolic blood pressure. A transient initial rise in blood pressure and body mass index was detected in surgical but not in spontaneous menopause. Blood pressure rises after menopause appear to be due more to increased body mass index and aging than ovarian failure per se.
The influence of menopausal status on arterial blood pressure is still unclear. In the previous reports on this issue available in the Medline database, systolic blood pressure levels have been found to show either negative,1 2 3 positive,4 5 or no correlation6 7 8 9 10 11 12 13 with menopause. Contrasting evidence has also been reported for diastolic blood pressure, which has been found to be either decreased,2 increased,4 5 6 7 12 or unaffected1 3 8 9 10 11 13 by menopause. Even if we consider only the available prospective studies,2 3 7 10 13 the results are still contradictory.
This report focuses on the relative influence of menopausal status, age, and body mass index on the sphygmomanometric blood pressure levels of healthy women. This cross-sectional study is part of an ongoing, population-based, preventive healthcare program conducted in Ferrara, a small town of Northern Italy.
We screened 3394 women who volunteered in the preventive care program of the University of Ferrara between November 1990 and November 1994. The age range (35 to 60 years) was chosen to include the age span of women reporting the beginning of menopause plus 5 years necessary to assess the influence of the time from menopause on blood pressure levels. Those who satisfied the following criteria were selected: no tobacco consumption; a daily alcohol intake of less than either 100 mL of wine, 330 mL of beer, or 25 mL of distilled spirits; no history, symptoms, or signs of diabetes mellitus, secondary hypertension, familial dyslipidemia, or cardiovascular, cerebrovascular, or renal disease; normal fasting values of blood glucose, creatinine, sodium, and potassium; and normal complete blood cell count and urinalysis. Women taking antihypertensive medications were classified as hypertensive and were included in the study only for the analysis of the prevalence of hypertension but were excluded from the analysis of the blood pressure level. Women in current need of any other drug that might interfere with blood pressure levels or lipid metabolism were excluded from the study.
As a result of this selection and after exclusion of the subjects in whom data collection was incomplete (47 subjects), the following groups were included in this study: 489 premenopausal women with regular menstrual periods and serum follicle-stimulating hormone <50 IU/L; 847 perimenopausal women with irregular periods or absence of menstruation for not longer than 11 months and follicle-stimulating hormone <50 IU/L; 887 women in spontaneous menopause with last menstrual period at least 12 months before they entered the study and follicle-stimulating hormone ≥50 IU/L; and 174 women in surgical menopause, who had undergone hysterectomy with bilateral ovariectomy for uterine leiomyomatosis and were menstruating with follicle-stimulating hormone <50 IU/L at the time of surgery.
The study was approved by the local ethics committee and followed procedures in accordance with institutional guidelines. Informed consent was obtained from each subject. Complete medical history, physical examination, time from last menstruation, and body mass index (calculated as weight in kilograms divided by the square of height in meters) were recorded. Systolic and diastolic blood pressures were measured by an experienced nurse using a mercury sphygmomanometer and following the recommendations of the American Heart Association.14 Caffeine ingestion in the 120 minutes before measurement was excluded. An ECG tracing was recorded. The first determination was averaged from three sitting readings (beginning after a 5-minute rest) separated by at least 2 minutes. Two other determinations were obtained with the same procedure on separate occasions within the following 4 weeks. Subjects were classified as hypertensive if they were taking antihypertensive medications or if blood pressure was ≥140 mm Hg systolic or 90 mm Hg diastolic as the mean value of these separate determinations. A chest roentgenogram was always obtained, as well as a venous blood sample for estrone, estradiol, and follicle-stimulating hormone determination. In premenopausal and perimenopausal women, blood was drawn on days 5 through 10 of the cycle. The subject was then referred to the central laboratory of the hospital for urinalysis, complete blood cell count, and automated blood chemistry.
Blood samples were drawn into glass tubes without additives and were allowed to clot at room temperature. They were centrifuged at 1500g for 30 minutes at 4°C. Serum concentrations of estrone, estradiol, and follicle-stimulating hormone were immediately determined in duplicate by conventional radioimmunoassays using commercial kits (FSH MAIAClone kit, Ares Serono Diagnostics; Estradiol Coat-A-Count Kit, Diagnostic Products Corp; Estrone Kit, Diagnostic Systems Laboratories). The minimal detectable dose was 0.5 IU/L for follicle-stimulating hormone, 30 pmol/L for estradiol, and 37 fmol/L for estrone. The intra-assay and interassay coefficients of variation were 3% and 4% for follicle-stimulating hormone, 6% and 7% for estradiol, and 7% and 8% for estrone, respectively.
The quality control analysis of the determinations performed by the central laboratory of our hospital during the study period revealed intra-assay and interassay coefficients of variation always <5% and <8%, respectively.
To minimize other possible confounding factors, only nonsmoking, untreated, healthy women with low alcohol consumption were included in the analysis of blood pressure levels. Subjects on antihypertensive medications were classified as hypertensive and were considered only for the analysis of the prevalence of hypertension. Data were stored on a dBase III Plus data base (Borland) and then exported to the statistical package Systat (Systat Inc) for elaboration. After positive testing for gaussian distribution, mean and SDs of the various parameters were calculated for each menopausal category. To appreciate the influence of the duration of menopause on blood pressure, we also selected subgroups of increasing duration of menopause (1 to >5 years) matched with premenopausal and perimenopausal women by chronological age at onset of menopause. The differences among means were compared by ANOVA (both before and after adjustment for the significant covariates, chronological age and/or body mass index) and by Scheffé’s test (as a post hoc procedure when needed). Then relative risks of systolic/diastolic hypertension (blood pressure >140/90 mm Hg) were calculated by logistic regression analysis in spontaneous and surgical menopause compared with premenopause both before and after adjustment for age and body mass index. Logistic regression coefficients and their standard errors were estimated by the method of maximum likelihood,15 from which adjusted estimates of relative odds with 95% CIs were calculated to approximate relative risks.
The characteristics of the study population are given in Table 1⇓ according to the menopausal status. Both systolic and diastolic blood pressures changed significantly with menopausal status. Systolic blood pressure was significantly higher in both spontaneous and surgical menopause than premenopause and perimenopause. Diastolic blood pressure was significantly higher only in surgical menopause. After adjustment for chronological age and body mass index, however, the menopausal status per se showed no significant effect on blood pressure levels (F values of 0.739 and 0.872 for systolic and diastolic blood pressure, respectively, P=NS in both cases). The relative influence on blood pressure variation was higher for chronological age than body mass index, as demonstrated by the more pronounced reduction of statistical significance after adjustment for age (F values of 2.365 and 1.968, both P=NS) than body mass index (F values of 5.886 and 2.989, P=.001 and .030).
The increased hypertensive risk of postmenopausal women (Table 2⇓) also lost statistical significance after adjustment for age and body mass index. Again, the highest relative influence was attributable to chronological age.
ANCOVA performed in subgroups of increasing duration of menopause (1 to >5 years) matched with premenopausal and perimenopausal women by chronological age at onset of menopause (Table 3⇓) demonstrated no significant influence of duration of menopause on blood pressure levels. In both spontaneous and surgical menopause, chronological age was a significant covariant of time after onset of menopause for systolic (but not diastolic) blood pressure, whereas body mass index covaried highly significantly for both systolic and diastolic blood pressures. A transient initial rise in blood pressure and body mass index was apparent in surgical menopause but absent in spontaneous menopause.
The main finding of this study is that sphygmomanometric blood pressure is increased in postmenopausal women as a result of chronological aging and higher body mass index. Menopause per se does not appear to be a risk factor for hypertension.
Previous studies on this subject have led to contradictory results.1 2 3 4 5 6 7 8 9 10 11 12 13 Insufficient sample size in relation to the great number of possible determinants of blood pressure changes, lack of control (premenopausal) groups, and different subject selection criteria may explain these inconsistencies. To minimize the effect of possible confounding factors, we studied only healthy women with no association of other cardiovascular risk factors or hypertensive target-organ damage. Racial influence was also absent in our study, since the whole Ferrara population is white. We avoided misclassification of hypertensive cases by adopting the standard procedure recommended by the official guidelines to define presence of hypertension.
Conversely, a possible concern about our findings comes from the cross-sectional nature of its design, which has known disadvantages compared with longitudinal studies. To partially overcome this limitation, we assessed the influence of menopause duration on blood pressure by comparing subgroups of increasing duration of menopause matched with premenopausal and perimenopausal women by chronological age at onset of menopause. Although not conclusive, this analysis further indicates that no significant influence of duration of menopause was present on the blood pressure levels of our population.
The separate analysis of a large group of women ovariectomized for therapeutic reasons led to the observation of a transient rise in blood pressure after occurrence of surgical menopause, which was related primarily to an increased body mass index. Our data permit no explanation of such a finding. We can only hypothesize that abrupt withdrawal of ovarian function in these women might create a transient hormonal and psychological imbalance that might be reflected by the observed blood pressure and weight changes.
Conversely, the changes in vascular physiology that may lead to increasing blood pressure, particularly systolic blood pressure, may occur during the perimenopausal to early postmenopausal period but may be noted only as an increase in blood pressure later in menopause, thus becoming statistically indistinguishable from an aging effect. Cross-sectional observations indicate that systolic blood pressure may increase faster in older women than men, at least in industrial societies.16 17 18 Were this confirmed by longitudinal studies, a causal effect or at least a hormonal influence of menopause on blood pressure could still be possible. Against this hypothesis, however, is our finding that duration of menopause showed no significant relation with blood pressure levels when the effect of chronological age was excluded by matching the age at onset of menopause (Table 3⇑). Again, the cross-sectional type of our study limits the interpretation of these data. Further longitudinal studies specifically designed to address this issue are needed.
In conclusion, this study suggests that ovarian failure and duration of menopause are not direct determinants of blood pressure rise in middle-aged women. However, weight gain could be an important and potentially preventable determinant of increased blood pressure and risk of hypertension in these women.
This study was supported by grants from the Italian Ministry of University and Scientific and Technological Research (60%).
Reprint requests to Francesco Portaluppi, MD, Hypertension Unit, First Institute of Internal Medicine, University of Ferrara, via Savonarola 9, I-44100 Ferrara, Italy.
- Received August 26, 1996.
- Revision received September 12, 1996.
- Accepted October 10, 1996.
van Beresteyn EC, van t’Hof MA, De Waard H. Contributions of ovarian failure and aging to blood pressure in normotensive perimenopausal women: a mixed longitudinal study. Am J Epidemiol. 1989;129:947-955.
Van Beresteijn EC, Riedstra M, van der Wel A, Schouten EG, Burema J, Kok FJ. Habitual dietary calcium intake and blood pressure change around the menopause: a longitudinal study. Int J Epidemiol. 1992;21:683-689.
Hassager C, Riis BJ, Strom V, Guyene TT, Christiansen C. The long-term effect of oral and percutaneous estradiol on plasma renin substrate and blood pressure. Circulation. 1987;76:753-758.
Grobbee DE, van Hemert AM, Vandenbroucke JP, Hofman A, Valkenburg HA. Importance of body weight in determining rise and level of blood pressure in postmenopausal women. J Hypertens. 1988;6(suppl 4):S614-S616.
Wu ZY, Wu XK, Zhang YW. Relationship of menopausal status and sex hormones to serum lipids and blood pressure. Int J Epidemiol. 1990;19:297-302.
Owens JF, Stoney CM, Matthews KA. Menopausal status influences ambulatory blood pressure levels and blood pressure changes during mental stress. Circulation. 1993;88:2794-2802.
Frohlich ED, Grim C, Labarthe DR. Recommendations for human blood pressure determination by sphygmomanometers: report of a special task force appointed by the Steering Committee, American Heart Association. Hypertension. 1988;11:209A-222A.
Schlesselman JJ. Case-Control Studies: Design, Conduct, Analysis. Oxford, UK: Oxford University Press, 1982.
Gordon T. Blood Pressure of Adults by Age and Sex: United States 1960-62. Washington, DC: National Center for Health Statistics; PHS publication 1000, series 11, No. 4: 1964.
Miller HW. National Center for Health Statistics. Plan and Operation of the National Health and Nutrition Examination Survey. United States 1971-1973. Vital and Health Statistics, series 1. Nos. 10a and 10b. Washington, DC: US Dept of Health, Education, and Welfare publication 73-1310; US Government Printing Office: 1973:46.
Roberts J, Maurer K. Blood Pressure Levels of Persons 60-74 Years. United States 1971-1974. Washington, DC: National Center for Health Statistics; 1977:1.