Journal of Human Hypertension (2001) 15, 313–321  2001 Nature Publishing Group All rights reserved 0950-9240/01 $15.00 www.nature.com/jhh ORIGINAL ARTICLE Dose response effect of cyclical medroxyprogesterone on blood pressure in postmenopausal women

PJ Harvey1, D Molloy2, J Upton2 and LM Wing1 Departments of 1Clinical Pharmacology and 2Medicine, Flinders University of South Australia, Bedford Park, Adelaide, South Australia, Australia 5042

Objective: This study was designed to compare with mean values of weeks 3 and 4 of each phase used for placebo the dose-response effect of cyclical doses of analysis. Ambulatory BP was performed in the final the C21 progestogen, medroxyprogesterone acetate week of each phase. (MPA) on blood pressure (BP) when administered to Results: Compared with the placebo phase, end of normotensive postmenopausal women receiving a fixed phase clinic BP was unchanged by any of the proges- mid-range daily dose of conjugated equine oestrogen togen treatments. There was a dose-dependent (CEE). decrease in ambulatory daytime diastolic and mean Materials and methods: Twenty normotensive post- arterial BP with the progestogen treatments compared menopausal women (median age 53 years) participated with placebo (P Ͻ 0.05). in the study which used a double-blind crossover Conclusion: In a regimen of postmenopausal hormone design. There were four randomised treatment phases, replacement therapy with a fixed mid-range daily dose each of 4 weeks duration. The four blinded treatments of CEE combined with a cyclical regimen of a C21 pro- were MPA 2.5 mg, MPA 5 mg, MPA 10 mg and matching gestogen spanning the current clinical dose range, the placebo, taken for the last 14 days of each 28 day treat- progestogen has either no effect or a small dose-depen- ment cycle. CEE 0.625 mg was also administered once dent reduction in clinic and ambulatory BPs over one daily as open labelled tablets to all subjects throughout treatment cycle. the study. Clinic BP was measured weekly with the Journal of Human Hypertension (2001) 15, 313–321

Keywords: menopause; progestogen; oestrogen; blood pressure; ambulatory blood pressure; arterial pulse waveform

Introduction The one notable exception is the Postmenopausal Estrogen/Progestin Interventions (PEPI) Trial.5 This Reliable data regarding the effects of ‘hormone study in healthy, normotensive postmenopausal replacement therapy’ (HRT) on blood pressure women provided the first definitive evidence using a remain deficient. Collectively, previous studies have randomised, placebo-controlled design, that a single variably shown no change,1–12 a decrease1,2,13–29 or 4,9,11,21,30–35 mid-range dose of conjugated equine oestrogen an increase in blood pressure with vari- administered with randomised progestogen treat- ous combinations of HRT. Unfortunately, few of the ments or placebo had no significant effect on blood previous studies have been designed to differentiate pressure.5 In addition to the PEPI study, limited data between the effects on blood pressure of the individ- on the effects of progestogens on blood pressure are ual oestrogen (‘natural’, conjugated equine or available from studies of progestogen-only contra- semisynthetic) and progestogen (C21 or C19 deriva- ceptive formulations administered to normotensive tives administered as cyclical or continuous premenopausal women. From these studies there is therapy) components of the regimen when these little evidence to suggest that contraceptive doses of preparations have been administered in combi- progestogens given alone to normotensive premeno- nation. pausal women are associated with a pressor effect on blood pressure.31,36–38 However, results of some studies have suggested that, while progestogen Correspondence: Dr Paula Harvey, Division of Cardiology, administered alone does not have a demonstrable Toronto General Hospital, 12 Eaton South, Room 414, 200 Eliza- beth Street, Toronto, Ontario, M5G 2C4, Canada pressor effect, when administered in combination E-mail: paula.harveyȰutoronto.ca with oestrogen in combined contraceptive formu- Received 1 September 2000; revised 27 October 2000; accepted lations the progestogen may have a synergistic 4 December 2000 effect, thereby enhancing the pressor effect of the oestrogen.37,39 Postmenopausal progestogen and blood pressure PJ Harvey et al 314 From the currently available data a number of performed to assess endometrial thickness and to questions remain unanswered. Although the results exclude the presence of endometrial hyperplasia in of the PEPI study were reassuring, the study did not women with an intact uterus. Women over the age provide information regarding the different types, of 50 years who had not had a mammogram within doses and regimens of progestogen ‘replacement’ the last 2 years were required to have a mammogram therapy currently used clinically. The effect on to exclude breast carcinoma prior to study entry. blood pressure and other cardiovascular parameters Subjects were excluded if they had any chronic of a particular oestrogen-progestogen combination medical illness, a history of breast cancer, clinical may depend not only on the type and dose of the evidence of pelvic disease, evidence of endometrial oestrogen administered, but also on the type and hyperplasia (defined as endometrial thickness dose of the progestogen.37 The magnitude of any Ͼ5 mm),46 previous intolerance of oestrogen or pro- progestogen-related effect may also be influenced by gestogen in ‘replacement’ doses, the need for any the relative androgenicity of the progestogen prep- chronic medication likely to impact on blood press- aration used.40–44 ure, 24-h urinary sodium excretion Ͻ70 mmol/day In the present study in normotensive postmeno- or Ͼ250 mmol/day and a body mass index of pausal women, a randomised, double-blind cross- Ͼ30 kg/m2. Subjects were also excluded if they had over design was used to compare with placebo the a disturbance of heart rhythm (atrial fibrillation, fre- effects on blood pressure and other cardiovascular quent (Ͼ10/min) atrial or ventricular extrasystoles) parameters of cyclical doses spanning the clinical which rendered blood pressure measurement by dose range of the C21 progestogen, medroxyproges- oscillometry inaccurate. terone acetate, in normotensive women receiving a The study was approved by the Committee on fixed mid-range daily dose of conjugated equine oes- Clinical Investigation and the Drug and Thera- trogen. The study tested the null hypothesis that peutics Advisory Committee at Flinders Medical there would be no significant effect on blood press- Centre. Prior to entry, all subjects were informed ure from the addition of ‘replacement’ doses of about the study both verbally and by an information exogenous semi-synthetic progestogen as cyclical sheet and were asked to provide written consent therapy to daily oestrogen supplementation in post- to participation. menopausal normotensive women. Study design Materials and methods The study was conducted using a randomised, double- Subjects blind placebo controlled crossover design. Each phase was of 4 weeks duration. For each subject the Subjects included in this study were recruited by order of the treatments was allocated according to a public advertisements in the Consulting Clinics at 4 × 4 balanced Latin square. Each subject was allo- Flinders Medical Centre, a general teaching hospital cated a sequence of the treatments denoted by a row of Flinders University of South Australia, and by of the Latin square with individual treatments occu- advertisement in local news media. They were to be pying the cells of the row. Subjects were allocated healthy postmenopausal women aged 60 years or randomly (by random numbers) in blocks of four to less as determined by medical history, physical and the rows of the square and thus a treatment gynaecological examination and screening haemato- sequence. The columns of the square represented logical and biochemical testing. Postmenopausal the treatment phases in chronological order. (A bal- status was defined as either a history of at least 12 anced Latin square is one in which each treatment months of amenorrhoea prior to study entry in occurs only once in each cell in the square and in women with an intact uterus, or a history of pre- addition occurs only once after any other treatment.) vious hysterectomy with or without oophorectomy, There was no washout period between phases, but combined with biochemical evidence of menopause only the measurements recorded in either the last 2 defined as a serum follicle-stimulating hormone weeks (blood pressure and heart rate) or at the end (FSH) concentration у20 IU/L.45 Subjects were of each phase (ambulatory blood pressure, arterial required to be normotensive, defined as having a waveform and biochemical parameters) were used mean clinic sitting blood pressure of р140 mm Hg for comparison, giving an effective washout period systolic and р90 mm Hg diastolic measured using a of at least 2 weeks in each phase. The active phases mercury sphygmomanometer on at least two separ- of the study were preceded by an open 4-week run- ate occasions at least 1 week apart. They were not in phase in which screening examinations were con- to be currently receiving any hormone replacement ducted, exclusion criteria were assessed and patient therapy and any previous hormone replacement consent obtained. therapy was required to have been ceased at least 1 month prior to entry into the active phases of the Treatments and dosage regimens study. Each subject was required to undergo a pelvic examination to exclude any significant pelvic abnor- Each subject received continuous once daily treat- mality. Transvaginal ultrasound examination was ment with conjugated equine oestrogen (CEE)

Journal of Human Hypertension Postmenopausal progestogen and blood pressure PJ Harvey et al 315 0.625 mg throughout the four 1-month randomised Protocol treatment phases of the study. In addition they were randomised according to a balanced Latin square Subjects were reviewed weekly throughout the design to a particular sequence of the four treatment study and measurements were made prior to the phases which were medroxyprogesterone acetate next dose of study medication approximately 24 h 2.5 mg (MPA 2.5 mg), medroxyprogesterone acetate after the previous dose. Blood pressure was meas- 5 mg (MPA 5 mg), medroxyprogesterone acetate ured in the clinic using a semi-automatic sphygmo- 10 mg (MPA 10 mg) and matching placebo (P). manometer (Dinamap Vital Signs Monitor 8100, Each treatment was taken orally at approximately Critikon Inc., Johnson & Johnson, FL, USA) which 48 the same time each day with the medroxyproges- operates by the oscillometric principle. A cuff of terone and matching placebo taken for the last 14 appropriate size for each subject was used by ensur- days of each 28-day treatment cycle, which corre- ing that the bladder length was at least 80% and the bladder width at least 40% of the circumference of sponded to the final 2 weeks of each 4-week ran- 49 domised phase. Each treatment cycle commenced at the subject’s mid-upper arm. On each occasion the beginning of each randomised phase. The three measurements were taken in the sitting pos- oestrogen treatments were administered as open- ition after at least 5 min quiet rest and then again in labelled tablets. The blinded, randomised treat- the standing position after 2 min standing. In each position the average of the 2nd and 3rd readings ments (medroxyprogesterone and placebo) were were taken as the blood pressure for that position. encapsulated in identical-appearing opaque gelatine Blood pressure measurements were made consist- capsules. The placebo preparation consisted of lac- ently on the same arm throughout the study. Heart tose BP powder. The study medications were pur- rate was recorded using the Dinamap and the aver- chased from Ayerst Laboratories Pty Ltd (conjugated age of the 2nd and 3rd values was recorded as the equine oestrogen, ‘Premarin’) and Upjohn Pty Ltd heart rate in each position. Body weight was also (medroxyprogesterone acetate, ‘Provera’). measured in light indoor clothing without shoes. At Study medication encapsulation, randomisation each visit adverse events, both spontaneously and packaging in individually labelled bottles were reported and in response to open questioning, were performed by the Pharmacy Department, Royal recorded. Compliance was assessed by tablet counts Adelaide Hospital, Adelaide, South Australia. Qual- at each visit and expressed as a percentage of num- ity control testing of the study drug formulations, ber of capsules taken/number of capsules that which included identification, disintegration and should have been taken in the particular period. uniformity of weight analysis were conducted at the Ambulatory blood pressure over a 26-h period Royal Adelaide Hospital. Disintegration studies was measured during the last week of each of the comparing the disintegration rate of encapsulated four randomised treatment phases using the Space- and non-encapsulated tablets showed that disinte- Labs 90207 automatic blood pressure recording sys- gration was unaffected by encapsulation and there- tem (SpaceLabs Inc, Redmond, WA, USA). Readings fore no further pharmaceutical testing was perfor- commenced at the time of daily drug administration med. and continued for 26 h. A further dose of tablets was Conjugated equine oestrogen 0.625 mg was administered at the end of 24 h (next scheduled dose chosen for use in the study because it was the oes- period) and the recording continued for a further 2 h trogen used in the PEPI study, and because it rep- to complete the 26-h recording period. The first 2 h resents the mid-range dose of the oral oestrogen of the recording period were discarded leaving most commonly administered as hormone replace- hours 2–26 for analysis.50 Readings were taken at 15- ment therapy.5,47 Medroxyprogesterone acetate was min intervals during the day and 60-min intervals chosen because it is the most commonly prescribed overnight. Each subject was asked to keep a diary C21 progestogen as orally administered HRT.5,47 The of activities during the recording period. For each three doses of MPA were selected to represent the subject, ‘awake’ (‘daytime’) and ‘asleep’ (‘night usual clinical dose range for administration of this time’) periods were identified from the subject’s progestogen in ‘cyclical’ hormone replacement regi- diary, and average blood pressure and heart rate mens. were calculated separately for each of these periods. Subjects suffering an intercurrent illness requiring Average 24-h blood pressure and heart rate were treatment during the study were allowed to continue also calculated for each recording. in the study provided that they did not require Peripheral pulse waveform was recorded by applanation tonometry (SphygmoCor BPAS-1, PVW medications which were known to influence blood 51 pressure or to alter the metabolism of the oestrogen Medical, Sydney, Australia) from the radial artery during the last week of each phase. The time of day and/or progestogen preparations. The prescription at which subjects took their medication and the time of any medication during the study was recorded of the pulse wave measurements in relation to dose and the reason for prescription, dose and route of were maintained constant (22–26 h post dose) in all administration, date of onset and cessation of treat- subjects throughout the studies. The radial wave- ment were noted. forms were used in a computerised algorithm to

Journal of Human Hypertension Postmenopausal progestogen and blood pressure PJ Harvey et al 316 calculate central aortic pressures and systolic aug- regression model was used to test for a relationship mentation, as well as aortic pulse waveforms and between progestogen dose and clinic, ambulatory additional parameters of left ventricular function.52 and calculated aortic blood pressure parameters.57 At the end of each phase a fasting venous blood Friedman’s test for repeated measures relating to sample was collected after 5 min of quiet rest in the non-parametric data was used for sex hormone bind- supine position, for measurement of plasma renin ing globulin and oestradiol.58 Median values for activity (PRA),53 and plasma concentrations of elec- each treatment group were substituted for missing trolytes, urea and creatinine (standard autoanalyser values.59 techniques), and aldosterone.54 Pre-dose serum oes- tradiol (DiaSorin IRMA kit; DiaSorin s.r.l., Saluggia, Italy), follicle stimulating hormone (FSH) (Bioclone Results IRMA kit; Bioclone Australia Pty Ltd, Marrickville, Subjects New South Wales, Australia), and sex hormone binding globulin (SHBG) (Orion IRMA kit; Orion Twenty subjects entered and completed the random- Diagnostica, Orion Corporation, Espoo, Finland) ised treatment phases of the study. Their median age concentrations were also measured. At the end of was 53 years (range: 46–60 years). At randomisation, ± each phase subjects were required to perform a 24- average clinic sitting blood pressure was 122 ± ± h urine collection for measurement of 24-h urinary 3/65 3 mm Hg (mean s.e.), average weight was ± sodium excretion. In addition at the end of each 75 3 kg and average 24-h urinary sodium excretion ± phase subjects had a physical examination to assess was 133 12 mmol/day. Overall median compliance any changes in physical status. was 100% (range: 90–102%).

Data analysis Clinic pre-dose blood pressure, heart rate and weight The average values for parameters in each phase of the study were compared by repeated measures Compared with placebo, there were no significant analysis of variance (BMDP Windows New System, changes in clinic systolic, diastolic or mean arterial Professional Edition, 5V, UCLA, Los Angeles, CA, blood pressures in any of the phases with cyclical USA, 1993) with treatments, position of measure- progestogen (Table 1). There was a small increase in ments (clinic blood pressure values) and time of day clinic heart rate with MPA 5 mg when compared (ambulatory blood pressure values) as main within- with MPA 2.5 mg (P Ͻ 0.01) and MPA 10 mg subject factors and baseline (pre-randomisation) (P Ͻ 0.05) treatments, but there were no changes in values as covariate. For clinic blood pressure, heart clinic heart rate with any of the progestogen treat- rate and body weight, the means of the values in the ments when compared with placebo. There was a last 2 weeks of each phase were used in the analysis. significant effect of change in position (P Ͻ 0.01) in For the other parameters, the values at the end of all phases with an increase in diastolic blood press- each phase were compared. For each ambulatory ure and heart rate between the sitting and standing blood pressure record ‘daytime’, ‘night time’ positions. There was no significant change in body (corresponding to ‘awake’ and ‘asleep’ periods weight with any of the treatments. extracted from the subject’s diary) and ‘average 24 h’ values were calculated and then comparisons between phases were made for each of these values. Ambulatory blood pressure and heart rate The methods inherent to the BMDP 5 V statistical Average ambulatory daytime diastolic blood press- package with appropriate reduction of degrees of ure was reduced with the progestogen treatments, freedom were used to impute values for the missing with the reduction being significant in the MPA data from subjects in whom values from only one or two phases were missing.55 Subjects in whom values from more than two phases were missing were excluded from the analyses. Table 1 Pre-dose clinic blood pressure, heart rate and weight Pairwise comparisons among individual treat- ments for each parameter were made using model Placebo MPA MPA MPA 2.5 mg 5 mg 10 mg predicted mean values and their covariance to gen- erate estimate differences standardized by the esti- SBP (mm Hg) 120 ± 3 121 ± 3 121 ± 3 121 ± 3 mated error of the difference as ‘z-scores’. These ‘z- DBP (mm Hg) 69 ± 269± 269± 268± 2 scores’ were assumed to be approximated by a stan- Heart rate (beats/min) 74 ± 173± 176± 1*† 73 ± 1 dard normal distribution. To allow for the multiple Weight (kg) 75.0 ± 0.3 75.0 ± 0.3 75.0 ± 0.3 75.0 ± 0.3 comparisons the Dunn-S¯ ida´k procedure [P* = 1−(1−P)k] was used.56 A probability of P Ͻ 0.05 The values are model-predicted means (± s.e.) from the BMDP ¯ analysis, and were obtained assuming a mean baseline value for (after Dunn-Sida´k correction where appropriate) all subjects. To allow for multiple comparisons the P values were was regarded as statistically significant. adjusted by the Dunn-S¯ ida´k procedure. *P Ͻ 0.01 vs MPA 2.5 mg; A random-effects generalised least squares †P Ͻ 0.05 vs MPA 10 mg.

Journal of Human Hypertension Postmenopausal progestogen and blood pressure PJ Harvey et al 317 Table 2 Ambulatory blood pressure and heart rate Table 4 Plasma renin activity and plasma aldosterone

Placebo MPA MPA MPA Placebo MPA MPA MPA 2.5 mg 5 mg 10 mg 2.5 mg 5 mg 10 mg

Daytime PRA (ngAl/ml/h) 2.3 ± 0.8 3.0 ± 0.8 2.0 ± 0.8 1.9 ± 0.8 SBP (mm Hg) 130 ± 3 130 ± 3 129 ± 3 128 ± 3 PA (ng/100 ml) 11.8 ± 2.0 11.1 ± 2.0 11.5 ± 2.0 12.2 ± 2.0 DBP (mm Hg)‡ 81 ± 280± 279± 278± 2* MAP (mm Hg)‡ 98 ± 297± 296± 295± 2 ± ± ± ± ± The values are model-predicted means ( s.e.) from the BMDP HR (beats/min) 80 2792782782 analysis, and were obtained assuming a mean baseline value for all subjects. Night time SBP (mm Hg) 114 ± 2 116 ± 2 115 ± 2 116 ± 2 DBP (mm Hg) 65 ± 267± 266± 266± 2 Table 5 Biochemical markers of oestrogen effect MAP (mm Hg) 82 ± 283± 283± 283± 2 ± ± ± ± HR (beats/min) 64 2642662642 Placebo MPA MPA MPA 2.5 mg 5 mg 10 mg The values are model-predicted means (± s.e.) from the BMDP analysis. To allow for multiple comparisons the P values were ¯ SHBG (nmol/l) adjusted by the Dunn-Sida´k procedure. A random-effects general- Median 110 96 93* 98* ised least squares model was used to test for a relationship (range) (42–180) (45–170) (39–160) (30–155) between progestogen dose and change in blood pressure. Ͻ Ͻ *P 0.05 vs placebo; ‡P 0.05 significant regression for dose. FSH (IU/L) Mean ± s.e. 29 ± 329± 326± 323± 3* E2 (pmol/l) Median 140 118 118 135 10 mg phase (P Ͻ 0.05) (Table 2). A significant dose- (range) (55– (60–835) (40–200) (50–215) response relationship across the range of proges- 1735) togen doses used was shown for daytime diastolic ± and daytime mean arterial blood pressures The values for FSH are model-predicted means ( s.e.) from the Ͻ BMDP analysis. Analysis of SHBG and E2 by Friedman’s non- (P 0.05). Average ambulatory daytime systolic parametric test for multiple comparisons. *P Ͻ 0.05 vs placebo. blood pressure and night time systolic, diastolic and mean blood pressures were unchanged in the proge- stogen phases compared with placebo. No change in progestogen treatments compared with placebo average 24-h, daytime or night time ambulatory (Table 4). heart rate was observed with the progestogen treat- ments. Markers of oestrogen action A reduction in serum SHBG was observed with the Calculated aortic pressures and pulse waveform MPA 5 mg and MPA 10 mg treatments (P Ͻ 0.05) and for serum FSH in the MPA 10 mg phase No changes in calculated aortic systolic or diastolic (P Ͻ 0.05) (Table 5). Serum oestradiol was blood pressures were observed during the random- unchanged between treatments. ised treatment phases (Table 3). There were no changes in calculated aortic systolic augmentation indices in any phase. Plasma electrolytes and 24-h urinary sodium excretion Renin-angiotensin system No clinically significant changes in plasma sodium, potassium, chloride, urea or creatinine concen- There were no changes in plasma renin activity or trations were observed with the progestogen treat- plasma aldosterone concentration with any of the ments compared with placebo. There was no sig- nificant change in 24-h urinary sodium excretion in any phase. Table 3 Calculated aortic pressures and augmentation index Adverse effects Placebo MPA MPA MPA 2.5 mg 5 mg 10 mg The most commonly reported adverse effects which included headache, hot flushes, mastalgia and Calculated aortic SBP 124 ± 4 127 ± 4 125 ± 4 123 ± 4 fatigue were reported in all four treatment phases (mm Hg) and did not appear to be dose-related. Calculated aortic DBP 76 ± 279± 277± 275± 2 (mm Hg) Augmentation index (%) 145 ± 4 144 ± 4 146 ± 4 146 ± 4 Discussion

The values are model-predicted means (± s.e.) from the BMDP In this study in healthy normotensive postmeno- analysis. pausal women receiving a fixed mid-range dose of

Journal of Human Hypertension Postmenopausal progestogen and blood pressure PJ Harvey et al 318 conjugated equine oestrogen, the addition of med- togenic component of the hormone replacement roxyprogesterone acetate, administered in a cyclical regimen may be of equal importance and complexity regimen in three different doses spanning the usual with the effects on vascular reactivity dependent on clinical dose range, produced no change in clinic the combination of dose, type, route and pattern of blood pressure, and no change or a small dose- administration of the progestogen.40–41,44 Studies dependent reduction in ambulatory blood pressure have shown that synthetic progestogens may exert a parameters when compared with placebo. Calcu- relative vasoconstrictor effect on the vasculature, the lated aortic pressures, augmentation indices and magnitude of the effect being proportional to the pulse waveform did not differ between phases. relative ‘androgenicity’ of the individual prep- There was no change in body weight in any phase. aration.40–42,44 In contrast to the vasoconstrictor No change in plasma renin activity or plasma aldos- effects observed with the ‘androgenic’ progestogens, terone concentration was observed with the proges- the ‘non-androgenic’ progestogens including pro- togen treatments. Suppression of serum FSH con- gesterone and medroxyprogesterone (which was centration occurred with the highest dose of used in the current study) have been shown to have progestogen and a small reduction in serum SHBG a net vasodilator effect.38,41,44 The observations in concentration was observed in the progestogen the present study examining the dose-response of phases compared with placebo. There was a wide cyclical MPA in the presence of a fixed replacement variation in measured serum oestradiol concen- dose of oestrogen are consistent with MPA having trations across the group but there were no signifi- either no net effect on vascular tone or a small net cant changes in serum concentration between treat- vasodilator effect. The mechanisms of the effect of ment phases. the different progestogens on the vasculature remain The majority of previous studies of the effects of unclear and warrant further investigation. postmenopausal hormone replacement therapy on In addition to the proposed direct adverse effects blood pressure have failed to differentiate between of ‘androgenic’ progestogens on vascular reactivity, the effects of the individual oestrogen and proges- it has been suggested that these preparations may togen components on blood pressure when these also exert an additional unfavorable influence on preparations have been administered in combi- blood pressure through a tendency to induce nation.5 Very few data are available relating to the sodium and water retention, with associated oedema specific effects on blood pressure of different types, and weight gain.40,43 In the present study no change doses and regimens of progestogen. The current in body weight, 24-h urinary sodium excretion or study used a double-blind, randomised, placebo- activity of the circulating renin-angiotensin system controlled design with adequate power to detect was observed with any of the ‘non-androgenic’ small changes in clinic and ambulatory blood press- MPA treatments. ure produced by different doses of a C21 proges- Systolic augmentation of the aortic arterial pulse togen administered in a cyclical regimen in doses is observed with normal aging and contributes to the spanning the usual clinical dose range combined disproportionate increase in systolic blood pressure with a fixed daily dose of oral oestrogen. The results observed with increasing age.61 It may also contrib- of the current study are consistent with those of pre- ute to adverse effects on cardiovascular function.62 vious studies, including the PEPI study, which have Changes in aortic systolic and diastolic pressures found no change2–10,12 or a reduction24–26,29 in blood and in aortic systolic augmentation may occur with- pressure associated with the addition of a proges- out measurable change in peripheral arterial blood togen to ‘replacement’ doses of oestrogen. pressure.61,62 In the present study no significant The results of the present study are in contrast to change was noted to occur in any of the calculated studies which have shown a synergistic pressor aortic pressure indices including augmentation effect when progestogen is added to oestrogen for index with the different doses of cyclical MPA. administration to premenopausal women for the Although the present study used a fixed dose of purpose of oral contraception.31,36–38 Contraceptive oestrogen combined with variable doses of proges- preparations tend to differ from ‘postmenopausal togen, a small but dose-dependent reduction in FSH replacement’ regimens in that they consist of combi- was noted to occur with the progestogen treatments. nations of the ‘potent’ semisynthetic oestrogens ethi- This is consistent with previous studies which have nyloestradiol or mestranol, often with the relatively demonstrated suppression of FSH concentration fol- androgenic C19 progestogens, both administered in lowing administration of progestogen, independent doses in excess of those used in HRT.60 The rela- of oestrogen.40,63 The suppressive effect is believed tively higher doses and potency of the preparations to be due to a direct effect on the gonadotropin- used in oral contraceptive formulations may there- producing cells of the anterior pituitary.40 A small fore influence the effects on the vasculature and reduction in SHBG concentration was also demon- blood pressure. strated with the progestogen treatments compared We have previously shown that the effect of post- with placebo. The reduction in serum FSH and menopausal oestrogen replacement therapy on SHBG64 shown in the present study confirm that the blood pressure is influenced by the type and dose biological effects of medroxyprogesterone acetate of oestrogen administered.1 The effect of the proges- were demonstrable despite the relative lack of effect

Journal of Human Hypertension Postmenopausal progestogen and blood pressure PJ Harvey et al 319 on blood pressure. Although there was wide inter- 7 Prough S, Aksel S, Wiebe R, Shepherd J. Continuous individual variation in plasma concentrations of estrogen/progestin therapy in menopause. Am J Obstet measured oestradiol, there was no significant differ- Gynecol 1987; 157: 1449–1453. ence in concentrations between treatment phases to 8 Hargrove J, Maxson W, Wentz A, Burnett L. Meno- contribute to the change in SHBG concentration. pausal hormone replacement therapy with continuous daily oral micronized estradiol and progesterone. The wide variability in measured circulating serum Obstet Gynecol 1989; 73: 606–612. oestradiol concentrations is likely to reflect the com- 9 Gow S, MacGillivray I. Metabolic, hormonal, and vas- plex composition of the conjugated equine oestro- cular changes after synthetic oestrogen therapy in gens used in the study, together with variable back oophorectomized women. Br Med J 1971; 2: 73–77. conversion of oestrone to oestradiol.65 10 Medical Research Council’s General Practice Frame- In conclusion, this study has shown that in a regi- work. Randomized comparison of oestrogen versus men of postmenopausal hormone replacement ther- oestrogen plus progestogen hormone replacement ther- apy with a fixed daily dose of oestrogen combined apy in women with hysterectomy. Br Med J 1996; 312: with a cyclical regimen of a C21 progestogen span- 473–478. ning the current clinical dose range, the progestogen 11 Akkad AA, Halligan AWF, Abrams K, Al-Azzawi F. has either no effect or produces a small dose-depen- Differing responses in blood pressure over 24 hours in normotensive women receiving oral or transdermal dent reduction in blood pressure over one cycle of estrogen replacement therapy. Obstet Gynecol 1997; treatment. The effects of different progestogen prep- 89: 97–103. arations on the vasculature are complex and more 12 Pripp U et al. A randomized trial on effects of hormone information is required regarding the effects of the therapy on ambulatory blood pressure and lipoprotein different types, doses, regimens and methods of levels in women with coronary artery disease. J Hyper- administration of progestogens currently used in tens 1999; 17: 1379–1386. postmenopausal hormone replacement therapy to 13 Mercuro G et al. Does the menopause play a role as a determine whether there are differences between the risk factor for essential hypertension? Circulation preparations in effect on blood pressure and related 1998; 78: 861–864. parameters or whether the effects are consistent 14 Alfie J et al. Hemodynamic effects of transdermal estra- across the different preparations. diol alone and combined with norethisterone acetate. Maturitas 1997; 27: 163–169. 15 Sudhir K, Esler MD, Jennings GL, Komesaroff PA. Acknowledgements Estrogen supplementation decreases norepinephrine- This study was supported by a grant from the Flin- induced vasoconstriction and total body norepi- ders Medical Centre Research Foundation and a nephrine spillover in perimenopausal women. Hyper- National Health and Medical Research Council of tension 1997; 30: 1538–1543. 16 Lind T et al. A prospective, controlled trial of six Australia Medical Postgraduate Research Scholar- forms of hormone replacement therapy given to post- ship (Dr Harvey). menopausal women. Br J Obstet Gynaecol 1979; 86 (Suppl 3): 1–29. References 17 Luotola H. Blood pressure and hemodynamics in post- menopausal women during estradiol-17-beta substi- 1 Harvey PJ, Wing LM, Savage J, Molloy D. The effects tution. Ann Clin Res 1983; 15(Suppl 38): 1–121. of different types and different doses of oestrogen 18 Hassager C et al. The long-term effect of oral and per- replacement therapy on clinic and ambulatory blood pressure and the renin-angiotensin system in normo- cutaneous estradiol on plasma renin substrate and tensive postmenopausal women. J Hypertens 1999; 17: blood pressure. Circulation 1987; 76: 753–758. 405–411. 19 Hassager C, Christiansen C. Blood pressure during 2 Myrup B, Jensen GF, McNair P. Cardiovascular risk oestrogen/progestogen substitution therapy in healthy factors during estrogen-norethindrone and cholecalci- post-menopausal women. Maturitas 1988; 9: 315–323. ferol treatment. Arch Intern Med 1992; 152: 2265– 20 Kable WT, Gallagher JC, Nachtigall L, Goldbar D. Lipid 2268. changes after hormone replacement therapy for meno- 3 Lieberman EH et al. Estrogen improves endothelium- pause. J Reprod Med 1990; 35: 512–518. dependent, flow-mediated vasodilatation in postmeno- 21 Wren B, Routledge AD. The effect of type and dose of pausal women. Ann Intern Med 1994; 121: 936–941. oestrogen on the blood pressure of post-menopausal 4 Brons M, Jacobsen JG, Kristoffersen K. Changes in women. Maturitas 1983; 5: 135–142. plasma renin concentration, activity and substrate 22 Staland B. Continuous treatment with natural oestro- after twenty-eight days’ treatment with micronized gens and progestogens. A method to avoid endometrial estrogen, ethinylestradiol and norethisterone-acetate stimulation. Maturitas 1991; 3: 145–156. in postmenopausal women. Acta Med Scand 1981; 209 23 MacLennan AH et al. Oestrogen and cyclical proges- (suppl 646): 94–97. togen in postmenopausal hormone replacement ther- 5 The Writing Group for the PEPI Trial. Effects of estro- apy. Med J Aust 1992; 157: 167–170. gen or estrogen/progestin regimens on heart disease 24 Seely EW, Walsh BW, Gerhard MD, Williams GH. risk factors in postmenopausal women. JAMA 1995; Estradiol with or without progesterone and ambulat- 273: 199–208. ory blood pressure in postmenopausal women. Hyper- 6 Saleh AA et al. Thrombosis and hormone replacement tension 1999; 33: 1190–1194. therapy in postmenopausal women. Am J Obstet Gyn- 25 van Ittersum FJ et al. Ambulatory-not office-blood ecol 1993; 169: 1554–1557. pressures decline during hormone replacement ther-

Journal of Human Hypertension Postmenopausal progestogen and blood pressure PJ Harvey et al 320 apy in healthy postmenopausal women. Am J Hyper- of atherosclerotic coronary arteries in cynomolgus tens 1998; 11: 1147–1152. monkeys. J Am Coll Cardiol 1994; 24: 1757–1761. 26 Manwaring P, Morfis L, Diamond T, Howes LG. Effects 45 Anon. Advice on contraception around the meno- of hormone replacement therapy on ambulatory blood pause. Drug Ther Bull 1992; 30: 95–96. pressure and vascular responses in normotensive 46 Granberg S et al. Endometrial thickness as measured women. Blood Pressure 2000; 9: 22–27. by endovaginal ultrasonography for identifying endo- 27 Manhem K, Ahlm H, Milsom I, Svensson A. Transder- metrial abnormality. Am J Obstet Gynecol 1991; 164: mal oestrogen reduces daytime blood pressure in 47–52. hypertensive women. J Hum Hypertens 1998; 12: 47 MacLennan AH, Taylor AW, Wilson DH. Changes in 323–327. the use of hormone replacement therapy in South Aus- 28 Mercuro G et al. Estradiol-17beta reduces blood press- tralia. Med J Aust 1995; 162: 420–422. ure and restores the normal amplitude of the circadian 48 Posey JA, Geddes LA, Williams H, Moore AG. The blood pressure rhythm in postmenopausal women. meaning of the point of maximum oscillations in cuff Am J Hypertens 1998; 11: 909–913. pressure in the indirect measurement of blood press- 29 Szekacs B et al. Hormone replacement therapy reduces ure. Part 1. Cardiovasc Res Cent Bull 1969; 8: 15–25. mean 24-hour blood pressure and its variability in 49 Prisant LM, Alpert BS, Robbins CB. American National postmenopausal women with treated hypertension. Standard for non-automated sphygmomanometers: Menopause 2000; 7: 31–35. summary report. Am J Hypertens 1995; 8: 210–213. 30 Crane MG, Harris JJ, Winsor W. Hypertension, oral 50 Prasad N, MacFadyen RJ, Ogston SA, MacDonald TM. contraceptive agents, and conjugated estrogens. Ann Elevated blood pressure during the first two hours of Intern Med 1971; 74: 13–21. ambulatory blood pressure monitoring: a study com- 31 Spellacy WN, Birk SA. The effect of intrauterine paring consecutive twenty-four-hour monitoring per- devices, oral contraceptives, estrogens, and progesto- iods. J Hypertens 1995; 13: 291–295. gens on blood pressure. Am J Obstet Gynecol 1972; 51 Kelly RP et al. Non-invasive registration of the arterial 112: 912–919. pulse waveform using high-fidelity applanation 32 Notelovitz M. Effect of natural oestrogens on blood tonometry. J Vas Med Biol 1989; 1: 142–149. pressure and weight in postmenopausal women. S 52 Kelly RP, Hayward CS, Avolio AP, O’Rouke MF. Non- African Med J 1975; 49: 2251–2254. invasive determination of age-related changes in the 33 Weinstein L, Bewtra C, Gallagher J. Evaluation of a human arterial pulse. Circulation 1989; 80: 1652– continuous combined low-dose regimen of oestrogen- 1659. progestin for treatment of the menopausal patient. Am 53 Bonnin JM, Cain MD, Jose JS. Hypertension due to a J Obstet Gynecol 1990; 162: 1534–1542. renin secreting tumour localised by segmental renal 34 Plunkett ER, Wolfe BM. Prolonged effects of a novel, vein sampling. Aust N Z J Med 1977; 7: 630–635. 54 Ogihara T et al. A non-chromatographic non-extrac- low dosage continuous progestin-cyclic estrogen tion radioimmunoassay for serum aldosterone. J Clin replacement program in postmenopausal women. Am Endocrinol Metab 1977; 45: 726–731. J Obstet Gynecol 1992; 166: 117–121. 55 BMDP statistical software manual volume 2: to 35 Pfeffer RI. Estrogen use, hypertension and stroke in accompany the 7.0 software release (1992). Dixon WJ postmenopausal women. J Chron Dis 1978; 31: 389– (editor): Los Angeles: University of California Press, 398. 1992, pp 1435–1436. 36 Hall WD, Douglas MB, Blumenstein BA, Hatcher RA. 56 Ludbrook J. On making multiple comparisons in clini- Blood pressure and oral progestational agents. A pro- cal and experimental pharmacology and physiology. spective study of 119 black women. Am J Obstet Gyn- Clin Exp Pharmacol Physiol 1991; 18: 379–392. ecol 1980; 136: 344–348. 57 Stata statistical software manual volume 4: to 37 Wilson ESB, Cruickshank J, McMaster M, Weir RJ. A accompany the 6.0 release (1999). Stata Press, Univer- prospective controlled study of the effect on blood sity Drive East, College Station, Texas, USA, 1999, pressure of contraceptive preparations containing dif- p 420. ferent types and dosages of progestogen. Br J Obstet 58 Armitage P, Berry G. Distribution free methods. In: Gynaecol 1984; 91: 1254–1260. Armitage P, Berry G (eds). Statistical Methods in Medi- 38 Rylance PB et al. Natural progesterone and antihyper- cal Research. Third edition. Blackwell Scientific Pub- tensive action. Br Med J 1985; 290: 13–14. lications: Oxford, 1994, pp 448–468. 39 Wren B. The effect of oestrogen on the female cardio- 59 Armitage P, Berry G. The planning of statistical inves- vascular system. Med J Aust 1992; 157: 204–208. tigations. In: Armitage P, Berry G (eds). Statistical 40 Lobo RA, Stanczyk FZ. New knowledge in the physi- Methods in Medical Research Third edition. Oxford: ology of hormonal contraceptives. Am J Obstet Gyn- Blackwell Scientific Publications, 1994, pp 175–206. ecol 1994; 170: 1499–1507. 60 Goodman HM. Progestins. In: Goodman HM, Munson 41 Williams JK, Adams MR. Estrogens, progestins and PL (eds). Principles of Pharmacology. Basic concepts coronary artery reactivity. Nat Med 1997; 3: 273–274. and clinical applications. Chapman and Hall; 1995, pp 42 Rosselli M et al. Circulating nitric oxide (nitrite/ 827–836. nitrate) levels in postmenopausal women substituted 61 Berger DS, Li JK. Concurrent compliance reduction with 17␤-estradiol and norethisterone acetate. A two- and increased peripheral resistance in the manifes- year follow-up study. Hypertension 1995; 25: 848–853. tation of isolated systolic hypertension. Am J Cardiol 43 Lobo RA. The role of progestins in hormone replace- 1990; 65: 67–71. ment therapy. Am J Obstet Gynecol 1992; 166: 1997– 62 O’Rourke M, Kelly RP. Wave reflection in the systemic 2004. circulation and its implications in ventricular func- 44 Williams JK, Honore´ EK, Washburn SA, Clarkson TB. tion. J Hypertens 1993; 11: 327–337. Effects of hormone replacement therapy on reactivity 63 Wise AJ, Gross MA, Schalch DS. Quantitative relation-

Journal of Human Hypertension Postmenopausal progestogen and blood pressure PJ Harvey et al 321 ships of the pituitary-gonadal axis in postmenopausal 65 Stumpf PG. Pharmacokinetics of estrogen. Obstet Gyn- women. J Lab Clin Med 1973; 81: 28–36. ecol 1990; 75 (Suppl): 9s–14s. 64 Kuhl H. Comparative pharmacology of newer proges- togens. Drugs 1999; 51: 188–215.

Journal of Human Hypertension