The effect of hormone replacement therapy on blood pressure ... - Nature

Journal of Human Hypertension (1997) 11, 405–411  1997 Stockton Press. All rights reserved 0950-9240/97 $12.00

ORIGINAL ARTICLE

The effect of hormone replacement therapy on blood pressure and cardiovascular risk factors in menopausal women with moderate hypertension C Kornhauser, JM Malacara, ME Garay and EL Pe´rez-Luque Instituto de Investigaciones Me´dicas, 20 de Enero 929, Leoń, Gto, 37320, Mexico

Objective: To assess the effect of injectable hormone replacement therapy (HRT) vs a placebo in hypertensive menopausal women. Design and methods: Prospective randomised doubleblind study over 90 days. Fifty-five menopausal women with mild to moderate hypertension, stopped their antihypertensive medication and were studied for 2 weeks. Diastolic BP increased to over 105 mm Hg in five patients who were not included in the study. The remaining patients were randomly allocated to three groups: placebo (PL), estradiol valeranate 10 mg (E), and estradiol valeranate 4 mg plus prasterone enantate 200 mg (E+P). A further five patients were excluded from the study for different reasons. Results: Standing and recumbent BP decreased in the

PL group but did not change in the E and E+P patients. No change was observed in the serum levels of total cholesterol or low density lipoprotein (LDL)-cholesterol at the end of the trial. Plasma renin activity (PRA), aldosterone and insulin levels decreased during the study; PRA fell more significantly in the E+P group, the aldosterone reduction was highly significant in both hormone treated groups, specially the E+P group. Plasma insulin decreased in all groups and FSH levels were lower in the two treatment groups. Conclusions: In a randomised controlled trial no rise in BP was found after 90 days in hypertensive women with two forms of HRT. There was an unexpected fall in BP in those women allocated to placebo injections.

Keywords: hormone replacement therapy; menopause; high blood pressure

Introduction During the menopause, low plasma oestrogen levels are associated with an increased risk of cardiovascular disease.1–4 Hormone replacement therapy (HRT) may modify this risk factor, although this effect is still a matter of debate.5 It is possible that ethnic and environmental factors including lifestyle may interact to explain variation in risk for hypertension in menopausal women from differing populations. We have recently reported that in our population the frequency of hypertension and obesity increased at the time of menopause.6 This phenomenon seems to vary in different ethnic roups, as the Framingham study did not report changes in blood pressure (BP), body weight or blood glucose.7 By contrast, increased rates of hypertension during the menopause, have been reported in the People’s Republic of China.8 The effects of HRT on cardiovascular risk factors including hypertension are not completely understood because of the interaction of several other factors. Sex steroids have a complex interaction with Correspondence: Dr Carlos Kornhauser, Instituto de Investigaciones Me´dicas, Universidad de Guanajuato, 20 de Enero 929, PO Box 874 Leoń, Gto, 37320, Mexico

systemic BP. Oestrogen administration increases plasma renin activity (PRA), plasma aldosterone, and causes sodium retention. These factors increase both extracellular volume and cardiac output. 9 Furthermore, oestrogen treatment induces vasodilation.10 Hypertensive patients may also have insulin resistance, with increased plasma glucose and insulin levels.11,12 The interaction of hypertension with these other risk factors must be considered for the assessment of HRT in menopausal women, including overweight and blood lipid levels.13 Plasma fibrinogen has also been proposed as an independent risk factor for cardiovascular disease.14 In this study we report the results of a randomised double-blind trial, examining the effects of injectable HRT in menopausal women with mild to moderate hypertension.

Materials and methods We carried out a prospective randomised doubleblind study to assess the effect of injectable HRT (oestrogen, and oestrogen prasterone preparations, vs a placebo). Fifty-five women were selected according to the following criteria: menopausal, non-diabetic, with mild to moderate hypertension, without neoplastic, metabolic or infectious disease,

The effect of hormone replacement therapy C Kornhauser et al

406

and without evidence of cardiovascular or uterine disease. The menopause was defined when amenorrhea for more than one year in women aged more than 40 years with previous regular menses. Mild to moderate hypertension was defined according to the Joint National Committee on detection, evaluation, and treatment of high BP (JNC-V) whilst not taking antihypertensive medication, but we lowered the diastolic BP (DBP) criteria upper limit of moderate hypertension to 105 mm Hg for ethical reasons.15 The purpose of the study and its possible risks were explained to the patients, and all 55 accepted inclusion. Twenty-three patients who were taking antihypertensive drugs stopped therapy as requested (four of them on diuretics were randomly assigned, one to the placebo (PL), one to the estradiol (E), and two to the estradiol plus prasterone (E+P) groups). After 15 days washout period, five patients had a DBP over 105 mm Hg, so they were excluded from the study and restarted treatment immediately. The remaining 50 patients were randomly allocated to three groups for treatment with intramuscular injections given every month for three consecutive months: Group 1: 16 patients were given PL (saline solution), seven of whom had been on antihypertensive therapy. Group 2: 16 patients receiving 10 mg estradiol valerianate (E), seven of them previously on antihypertensive therapy. Group three: 18 patients on estradiol valerianate 4 mg plus prasterone enantate 200 mg (E+P). Nine of them had previously stopped antihypertensive medication. The patients were not aware of the type of treatment nor the characteristics of the substance injected. No specific diet nor dietary instructions were given to the patients, and they were followed for five visits at 30, 45, 60, 75 and 90 days of treatment. Discontinuation from the trial planned for persistent endometrial bleeding, increase in DBP to above 105 mm Hg, or intercurrent disease occurring during the study. An investigator, blind to the treatment assignment, made the data collection. At the initial visit we obtained details of smoking habit, alcohol intake, occupation, physical activity, and parity. At the initial, and all subsequent visits, we also measured standing height, weight (body mass index [BMI] kg/m2 was calculated), arterial BP and thorax/hip, waist/hip and abdomen/hip circumference ratios. Standing and recumbent arterial BP was taken as the mean of three measurements at the start, the middle, and at the end of each interview after a minimum of 5 min rest, using a mercury sphygmomanometer with a proper cuff placed on the dominant arm. Phases I and V of the Korotkoff sounds were used to determine systolic and diastolic BP. At initiation and at the end of treatment, a fasting blood sample was taken to measure blood levels of glucose, cholesterol, LDL-cholesterol, HDLcholesterol, PRA, aldosterone and insulin. Creatinine clearance rate was calculated with the formula proposed by Schwartz:16 (height ×0.55/plasma creatinine). Total cholesterol, and HDL cholesterol were measured by the Lieberman–Buchard method. LDL cholesterol was calculated as total cholesterol minus the sum of HDL and VLDL cholesterol. Glu-

cose and urea levels were measured by an enzymatic colorimetric test, and serum creatinine by the Jaffe method. Hormones were measured by radioimmunoassay using reagent kits, and a solid phase separation of antibody-hormone. Aldosterone assay used 125I radiolabelled antibody. (Diagnostic Product Corporation, Los Angeles, CA, USA). PRA with a kit from Renckt Sorin-Biomedica (Diagnostic Division, Saluggia, Italy). Follicle stimulating hormone (FSH) was measured by radioimmunoassay in solid phase based in mono and policlonal 125I antibodies. Intraassay variation coefficients were 2.8% to 5.5% for FSH, 13.1% to 13.4% for insulin, 8.9% to 18.3% for aldosterone, and 4.2% to 8.3% for PRA. Insulin was measured with a kit from Diagnostic Product Corporation. Fibrinogen was measured with a kit from Baxter Diagnostics (Deerfield, IL, USA), using a fibrinometer. Ethical considerations The local ethical committee approved the research protocol, requiring close surveillance and strict exclusion criteria. Patients were to be excluded if DBP exceeded 105 mm Hg on a single recording. All patients signed informed consent forms. The pharmaceutical company marketing the hormonal preparations was unaware of the study. Statistical analysis Differences between the three groups at each point of observation were tested by analysis of variance (ANOVA). Significance between pairs of groups and the differences in changes from baseline values between the treatment and the placebo group were evaluated in post-hoc analysis by the Least Significant Difference test. For each group of treatment the initial and final values were compared by Student’s t-test for dependent samples. Regression analysis was used to examine the relationships between PRA, plasma aldosterone and insulin and the changes in BP. Adjustments were made with covariance analysis for the following factors: PRA, aldosterone, FSH, cholesterol and its fractions HDL and LDL and insulin. In view of the multiple comparisons made, significance was accepted with P , 0.01.

Results No patient missed any follow-up visit in which no more than 3 days delay was permitted. After 1 month of treatment, three out of the 50 patients abandoned the study for personal reasons unrelated to treatment (one from the E group and two from the E+P group). Three patients were excluded because of increased BP (one from the PL group, and two from the E+P group). Our analysis includes all patients followed at each scheduled time point. The remaining 44 patients were allocated as follows: 15 in the PL group, 15 in the E group and 14 in the E+P group. Only five women had transient moderate vaginal bleeding, one (6.6%) on PL, two (13.3%) receiving E, and two (14.3%) from the E+P group.


407

Table 1 Characteristics of patients Group

Age (years) Weight (initial), kg Weight (final), kg Body mass index (initial) Parity Fasting glucose (initial), mg/dl Fasting glucose (final), mg/dl Plasma fibrinogen (initial), mg/dl Plasma fibrinogen (final), mg/dl Triglycerides (initial), mg/dl Triglycerides (final), mg/dl FSH (initial), UI/L FSH (final), UI/L

Placebo

Estrogen

E+P

P

Mean

s.d.

Mean

s.d.

Mean

s.d.

55.1 72.3 71.1 31.9 6.7 98.4 99.9 288 244 211 189 90.9 84.4

54.1 12.3 12.4 5.3 4.0 22.2 12.2 65 60 120 100 9.7 7.4

54.1 73.6 73.9 32.6 7.7 97.3 99.7 290 265 183 171 73.8 67.1

5.9 7.2 7.1 3.4 3.4 19.6 16.2 62 41 90 68 20.2 17.6

58.6 70.3 69.9 30.1 7.8 91.7 93.8 279 247 179 173 83.6 64.7

6.6 13.5 13.7 5.6 4.9 14.7 14.0 54 77 78 69 16.4 18.0

NS NS NS NS NS NS NS NS NS NS NS NS 0.007

NS = not significant.

Table 1 shows the characteristics of the patients in the three treatment groups. The women were predominantly obese, multiparous, with mean ages 55.1 to 58.6 years. Thirty-eight of them were housewives. Only five patients reported that they were smokers, and three had moderate to occasional alcohol consumption. No change was observed between the serum levels of cholesterol or LDL-cholesterol before and during the trial in the three groups of treatment (Figure 1). FSH was lower in the groups with oestrogen and oestrogen-prasterone treatment. Figures 2 and 3 show systolic and diastolic BP obtained during the study. Surprisingly in the PL group, values at the end of the study were lower, as compared to those taken at randomisation, both for standing and recumbent BP. Change for SBP was −13.1 (t = 3.93, P = 0.002) for standing and 16.2 (t = 5.39, P , 0.001) for recumbent measurements. Changes in recumbent DBP in the placebo group was −9.0 (t = 4.29, P , 0.001) for standing, but this trend was not statistically significant. In the treated groups, no changes in systolic or diastolic BP were

observed. Comparing the BP values among the three groups, no difference was found at randomisation or at any time during the observation. At the end of treatment, comparisons of standing and recumbent SBP and DBP were not significantly different. The analysis of hormone changes during treatment, showed a significant decrease of PRA, aldosterone and insulin in the patients receiving E or E+P (Figure 4). PRA fell in all groups, but by a larger amount in the hormone treated groups. In the E+P group, the decrease was from 14.3 to 2.2 ng/ml. Using analysis of covariance (ANCOVA) the adjusted means values were 10.8 and 2.00 respectively (P , 0.01) a P , 0.010. In the E group the adjusted means were 11.5 and 3.5 before and after treatment (P , 0.018). The differences for the PL group were not significant. The reduction in plasma aldosterone in the E group was significant but when ANCOVA was applied, this significance was lost. The change in plasma aldosterone in the E+P group was highly significant, decreasing from 98 to 53 pg/ml. When the means were adjusted by

Figure 1 Mean total cholesterol, LDL-cholesterol and HDL-cholesterol levels in the three groups of menopausal women, before and after 3 months. (j), vertical lines indicate standard errors of the mean.


408

Figure 2 Standing BP in three groups of menopausal women with moderate hypertension: Control (n = 15), oestrogen treated (n = 15), and oestrogen-progestin treated (n = 14). Antihypertensive medication was withdrawn on day −15, and HRT initiated on day 0. Vertical bars indicate standard errors of the mean.

Figure 3 Recumbent BP in three groups of menopausal women with moderate hypertension: Control (n = 15), oestrogen treated (n = 15), and oestrogen-progestin treated (n = 14). Antihypertensive medication was withdrawn on day −15, and HRT initiated in day 0. Vertical bars indicate standard errors of the mean.

ANCOVA the values were 102.6 and 49.4 (P , 0.003). Plasma insulin levels decreased significantly in all three groups, although a more significant change was found for the hormone treated groups. The diminution in the E+P group, was from 18.3 to 7.6 mU/ml. When the ANCOVA was applied for the insulin values, the difference in the placebo group was no longer significant, but it persisted in the two treatment groups (P , 0.002 for the E group and P , 0.0001 for the E+P group). The creatinine clearance rates remained unchanged in all three groups.

al,21 and Barret-Connor et al22 found that women receiving oestrogen therapy had lower BP levels than controls. By contrast, Markovitz et al23 in a cross-sectional study found that menopausal women taking HRT had higher BP levels. Wren24 in a review of more than 100 published papers concluded that oestrogens reduce the cardiovascular risk factors. Lip et al25 in an open prospective study of women who required HRT for amelioration of menopausal symptoms, found no changes in SBP or DBP over 14 months of follow-up. In view of the increased frequency of hypertension in menopausal women found in several populations, we explored the HRT effect in women with mild to moderate hypertension for a limited period of time. Antihypertensive medication was withdrawn to avoid the possible interaction of medication which may confuse results. One problem is that the number of patients in the trial was small, and therefore we may have had insufficient power to detect differences between the three treatment arms; this effect may explain the lack of effect of therapy on variables that are known to be strongly influenced by hormone therapy. Two weeks of wash-out may be too short an interval to reverse the effects of the antihypertensive therapy, although this interval was the same manner as in

Discussion The effects of oestrogen on systemic BP are controversial. 17 Oral contraceptives may induce a marginal increase in BP although contraceptive medication contains higher doses than preparations of HRT for menopausal women.18 It has been suggested that HRT does not have any effect on BP in normotensive women19 and information from epidemiological studies suggests that oestrogen replacement therapy may in fact have the opposite effect. Weiss et al 20 suggest furthermore that oestrogen may protect against the development of hypertension. Lind et


409

Figure 4 Changes in PRA, aldosterone, insulin and creatinine clearance rates, before (h), and after (j) HRT, by intramuscular injection for 3 months. Vertical bars indicate standard errors of the mean.

other studies.26–28 The exclusion criteria for patients who developed a DBP higher than 105 mm Hg was necessary to protect the patients’ safety, but this may be a source of bias for the analysis of BP, and other variables. We found that untreated patients showed a progressive decline in BP during follow-up. We interpret this finding by two factors: firstly a bias caused by the exclusion (for ethical reasons) of patients who sustained a marked rise in BP after 2 weeks of discontinuing antihypertensive treatment, and secondly a placebo effect, on follow-up of mild to moderately hypertensive subjects with the patients becoming accustomed to participating in the study, or because they developed increased motivation to comply with the general recommendations on lifestyle. The fall in BP was less in women who received either of the two HRT regimes compared with placebo strongly suggests that HRT may have some pressor effects that did not become fully apparent in the limited 90-day duration of the study. The fall in PRA and aldosterone levels during the study parallelled the decrease of systemic BP. This was more significant in patients receiving active HRT, and is more evident when we examined the change in PRA values between the PL and active HRT groups; the change was significant when we compared the PL and the E+P groups. These find-

ings, may be due to the withdrawal of antihypertensive therapy with a decrease in activity of the reninangiotensin system, as most antihypertensive drugs increase renin levels. HRT seems therefore to enhance this reduction. This is important in view of the opposing effect of larger oestrogen doses in oral contraceptive therapy, which increases angiotensinogen synthesis.29–33 PRA increases in parallel with plasma estradiol levels during pregnancy.34 There were no significant changes observed in the insulin values or change in insulin levels between the PL and the two HRT groups. Fasting plasma insulin levels decreased in all three groups, but blood glucose remained unchanged. We interpret this finding to mean that the withdrawal of antihypertensive therapy causes a reduction of peripheral resistance to insulin. Antihypertensive treatment with thiazides and beta-blockers may increase insulin resistance.35,36 Our results show a recovery of insulin sensitivity after withdrawal of antihypertensives and that HRT enhances this recovery. Further studies should examine the effects of HRT on insulin sensitivity in patients receiving antihypertensive drugs. Fibrinogen, a recently recognized independent predictor for coronary heart disease, has been found to fall after oestrogen taken alone or in combination with a progestin in menopausal women.4 In the present study fibrinogen remained unchanged in


410

women in all three groups. The effect of ovarian steroids on BP may occur at multiple sites, with opposing effects. Furthermore, the route of administration is another source of variation of the effects of HRT. Non-oral administration bypasses the liver, when compared with oral medication where oestrogens enters the portal circulation and the liver. At the end of this study plasma cholesterol and LDL-cholesterol levels were unchanged in the three groups. Some studies have reported lower cholesterol levels in oestrogen treated women,13 and others have reported a diminished cardiovascular risk in oestrogen treated menopausal women.4,34 In our study the HRT treatment did not show any effect on cholesterol and its lipoprotein fractions. The use of HRT, its type, the dosage and the route of administration are influences which further complicate the varied biological state of menopausal women in different populations. We have previously reported that overweight and cigarette smoking, which vary in diverse population, are associated with variations in FSH levels.37 This could mean that different exposure to endogenous oestrogen, and therefore different effects of HRT. Coronary risk factors, and factors related to other problems during perimenopausal years must also be considered. We conclude that after 90 days in a controlled trial of HRT in menopausal hypertensive women there was no rise in systemic BP after withdrawal from antihypertensive therapy, although there was a moderate fall in BP in those women receiving placebo injections. The renin-angiotensin system and insulin resistance decreased during the study and HRT had a slight influence on these changes.

8 9

10 11 12 13

14

15

16

17 18

Acknowledgement This work was supported in part by the Consejo Nacional de Ciencia y Tecnologiá (CONACYT). Grant D0519-M9201. We thank the Hospitals of the Instituto Mexicano del Seguro Social in Leoń, for allowing us to study their patients.

References 1 Matthew KA et al. Menopause and risk factors for coronary heart disease. N Engl J Med 1989; 321: 641–646. 2 Bush TL. The epidemiology of cardiovascular disease in postmenopausal women. Ann NY Acad Sci 1990; 592: 262–271. 3 The writing group for the PEPI trial. Effects of estrogen or estrogen/progestin regimens on heart disease risk factors in menopausal women. JAMA 1995; 273: 199–208. 4 Goldman L, Totetson ANA. Uncertainty about menopausal estrogen, time for action not debate. N Engl J Med 1991; 325: 800–802. 5 Stampfer MJ et al. Postmenopausal estrogen therapy and cardiovascular disease. Ten-year follow-up from the nurses health study. N Engl J Med 1991; 325: 756–762. 6 Kornhauser C, Malacara JM, Donato F, Botello D. Mayor prevalencia de hipertensioń arterial en mujeres mayores de 5 anõs en la ciudad de Leoń, Guanajuato. Rev Invest Clin (Mex) 1994; 46: 287–294. 7 Hortland MC, McNamara PM, Kannel WB. Some

19 20

21

22 23 24 25 26 27 28

atherogenic concomitant of menopause: The Framingham study. Am J Epidem 1976; 103: 304 –311. Wu Z, Wu X, Zhang Y. Relationship of menopausal status and sex hormones to serum lipids and blood pressure. Int J Epidem 1990; 19: 297–302. Helmer OM, Griffith RS. The effect of administration of estrogens in the renin-substrate (hypertensinogen) content of rat plasma. Endocrinology 1952; 57: 421– 426. Grand A. Le risque cardiovasculaire apres la menopause. Ann Cardiol Angeiol 1992; 41: 151–161. Shen DC et al. Resistance to insulin-stimulated glucose uptake in patients with hypertension. J Clin Endocrinol Metab 1988; 66: 580–583. Ferranini E et al. Insulin resistance in essential hypertension. N Engl J Med 1987; 317: 350–357. Tikkanen MJ, Kuusi T, Nikkila EA, Sipinen S. Postmenopausal hormone replacement therapy: effects of progestogens on serum lipids and lipoproteins. A review. Maturitas 1986; 8: 7–17. Landin K, Tengborn L, Smith U. Elevated fibrinogen and plasminogen activator inhibitor (PAI-1) in hypertension are related to metabolic risk factors for cardiovascular disease. J Inter Med 1990; 227: 273–278. Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure. The 1988 report of the joint national committee on detection, evaluation, and treatment of high blood pressure. Arch Intern Med 1988; 148: 1023–1038. Schwartz Gj, Haycock GB, Edelman CM, Spitzer A. A simple estimate of glomerular filtration rate in children derived from body length and plasma creatinine. Pediatrics 1976; 58: 259–264. Stevenson JC, Proudler AJ, Walton C, Godsland IF. HRT mechanisms of action: carbohydrates. Int J Fertil Menopausal Stud 1994; 39 (Suppl 1): 50–55. Prentice RL. On the ability of blood pressure effects to explain the relation between oral contraceptives and cardiovascular disease. Am J Epidemiol 1988; 127: 213–219. Knopp RH. The effects of postmenopausal estrogen therapy on the incidence of atheriosclerotic vascular disease. Obstet Gynecol 1988; 72: 23–30. Weiss NS. Relationship of menopause to serum cholesterol and arterial blood pressure: The United States health examination survey of adults. Am J Epidemiol 1972; 96: 237–241. Lind T, Cameron EC, Hunter WM. A prospective controlled trial of six forms hormone replacement therapy given to postmenopausal women. Br J Obstet Gynaecol 1979; 86 (Suppl 3): 3–29. Barrett-Connor E, Brown WV, Turner J. Heart disease risk factors and hormone use in postmenopausal women. JAMA 1979; 241: 2167–2169. Markovitz JH et al. Psychological, biological and health behavior predictors of blood pressure changes in middle-aged women. J Hypertens 1991; 9: 399– 406. Wren BG. The effect of oestrogen on the female cardiovascular system. Med J Aust 1992; 157: 204 –208. Lip GY, Beevers M, Churchill D, Beevers DG. Hormone replacement therapy and blood pressure in hypertensive women. J Hum Hypertens 1994; 8: 491– 494. Fujita T et al. Factors influencing blood pressure in salt-sensitive patients with hypertension. Am J Med 1980; 69: 334 –344. Parmer RJ, Cervenka JH, Stone RA, O’Connor DT. Autonomic function in hypertension. Circulation 1990; 81: 1305–1311. Nosadini R et al. Impaired insulin-induced glucose uptake by extrahepatic tissue is hallmark of NIDDM


29 30 31

32

33

patients who have or will develop hypertension and microalbuminuria. Diabetes 1994; 43: 491– 499. Saruta T, Saade GA, Kaplan NM. A possible mechanism for hypertension induced by oral contraceptives. Arch Intern Med 1970; 126: 621–626. Helmer OM, Griffith RS. Biological activity of steroids as determined by assay of renin substrate (hypertensinogen). Endocrinology 1951; 49: 154 –161. Newton MA, Sealey JE, Ledingham JG, Laragh JH. High blood pressure and oral contraceptives. Changes in plasma renin and renin substrate and in aldosterone excretion. Am J Obstet Gynecol 1968; 101: 1037–1043. Oelkers W, Blumel A, Schoneshofer M. Effects of ethinylestradiol on the renin-angiotensin aldosterone system and on plasma transcortin in women and men. J Clin Endocrinol Metab 1976; 43: 1036–1042. Wilson M, Morganti A, Zervoudakis I. Blood pressure,

34 35

36

37

the renin-aldosterone system and sex steroids through normal pregnancy. Am J Med 1980; 68: 97–104. Barret-Connor E, Bush T. Estrogen and coronary heart disease in women. JAMA 1991; 265: 1861–1867. Pollare T, Lithell HO, Berne C. A comparison of the effect of hydrochlorotiazide and captopril on glucose and lipid metabolism in patients with hypertension. N Engl J Med 1989; 321: 868–873. Lithell HO, Pollare T, Berne C. Insulin sensivity in newly detected hypertensive patients: Influence of captopril and other antihypertensive agents on insulin sensitivity and related biological parameters. J Cardiovasc Pharmacol 1990; 15 (Suppl 5): S45–S52. Velasco E et al. Gonadotrophins and prolactin serum levels during the perimenopausal period: correlation with diverse factors. Fertil Steril 1990; 53: 56–60.

411