Clinical and Investigative Medicine

 

Effect of progesterone therapy on arginine vasopressin and atrial natriuretic factor in premenstrual syndrome

Hiroko Watanabe, MD, PhD
David C.W. Lau, MD
H. Lindsay Guyn, BSc, BA
Norman L.M. Wong, PhD

Clin Invest Med 1997;20(4):211-23.

[résumé]

Dr. Watanabe is with the Department of Medicine, University of Calgary, Calgary, Alta.; Dr. Lau is with the Department of Medicine, University of Ottawa, Ottawa, Ont.; Mr. Guyn is with the Department of Community Health, University of Calgary, Calgary, Alta.; and Dr. Wong is with the Department of Medicine, University of British Columbia, Vancouver, BC.

Presented in part at the meeting of the Royal College of Physicians and Surgeons of Canada, September 1993, Vancouver, BC.

(Original manuscript submitted Jan. 6, 1997; received in revised form May 12, 1997; accepted May 16, 1997)

Contents

Abstract

Objectives: To explore the possible role of natriuretic peptides and vasopressin in luteal phase fluid retention in premenstrual syndrome (PMS) and to determine the effect of progesterone therapy on these hormones.

Design: Self-controlled prospective study.

Setting: University-based medical research centre.

Patients: Six patients with PMS were studied during the symptomatic luteal and asymptomatic follicular phases. The follicular phase response was used as the control for each subject.

Interventions: An intravenous infusion of 3% saline solution was administered on an early follicular and a late luteal phase day in 2 menstrual cycles. Progesterone was administered orally during the second luteal phase.

Outcome measures: Osmolality, arginine vasopressin (AVP), atrial natriuretic factor (ANF), and brain natriuretic peptide (BNP) levels in plasma, osmolality, sodium, potassium, cyclic adenosine monophosphate (cAMP) and cyclic guanosine 5´-phosphate (cGMP) concentrations in urine, and thirst sensation.

Results: Mean basal plasma ANF and osmolality levels and the threshold for AVP release and thirst were lower, and mean urinary cyclic nucleotide levels and AVP sensitivity (amount of AVP secreted per unit rise in plasma osmolality) were higher, in the luteal phase than in the follicular phase. With saline loading, there was an increase in plasma osmolality, AVP and ANF and in urinary sodium and cyclic nucleotide levels. Plasma ANF and osmolality levels remained lower in the luteal phase compared with the follicular phase, but AVP levels at the end of the saline infusion were higher in the luteal phase than in the follicular phase. Progesterone therapy caused an increase in plasma ANF and osmolality levels and the AVP threshold and a decrease in AVP levels and sensitivity and urinary cyclic nucleotide levels. BNP levels did not change with phase or treatment. The differences in AVP threshold with phase and treatment were statistically significant (p < 0.001). There was a significant phase effect for plasma ANF (p = 0.02) and a significant or near-significant interaction effect of phase and treatment for plasma ANF (p = 0.06) and urinary cAMP (p = 0.047) and cGMP (p = 0.066). The effect of phase and treatment was not significant for the other measurements.

Conclusions: Luteal phase fluid retention may be due to a relative deficiency of ANF and a lower threshold for AVP release. The symptomatic improvement produced by progesterone treatment may be due to its stimulation of ANF and inhibition of AVP release or synthesis.

Résumé

Objectif : Explorer le rôle possible des peptides natriurétiques et de la vasopressine dans la rétention des liquides pendant la phase lutéinique du syndrome prémenstruel (SPM) et déterminer l'effet qu'a sur ces hormones un traitement à la progestérone.

Conception : Étude prospective autocontrôlée.

Contexte : Centre universitaire de recherches médicales.

Patientes : On a étudié 6 patientes atteintes de SPM au cours des phases lutéinique symptomatique et folliculaire asymptomatique. On a utilisé la réaction pendant la phase folliculaire comme témoin dans le cas de chaque sujet.

Interventions : On a administré par infusion intraveineuse une solution saline à 3 % au début de la phase

folliculaire et à la fin de la phase lutéinique, pendant 2 cycles menstruels. La progestérone a été administrée par voie orale au cours de la deuxième phase lutéinique.

Mesures des résultats : Osmolalité plasmatique, concentrations plasmatiques de vasopressine arginine (VPA), de facteur natriurétique auriculaire (FNA) et de peptide natriurétique cervical (PNC), osmolalité urinaire, concentrations urinaires de sodium, de potassium, d'adénosine monophosphate cyclique (cAMP) et de guanosine 5´-phosphate cyclique (cGMP) et sensation de soif.

Résultats : Le taux plasmatique basal moyen de FNA, l'osmolalité, le seuil de libération de VPA et la soif ont été moins élevés, et les taux urinaires moyens de nucléotides cycliques et la sensibilité à la VPA (volume de VPA sécrétée par unité d'augmentation de l'osmolalité plasmatique) ont été plus élevés, au cours de la phase lutéinique que pendant la phase folliculaire. Avec une charge saline, on a constaté une augmentation de l'osmolalité plasmatique et des taux de VPA, de FNA, de sodium urinaire et de nucléotides cycliques. Le taux de FNA plasmatique et l'osmolalité sont demeurés moins élevés au cours de la phase lutéinique que pendant la phase folliculaire, mais les taux de VPA à la fin de l'infusion de solution physiologique ont été plus élevés au cours de la phase lutéinique que pendant la phase folliculaire. Le traitement à la progestérone a fait grimper le taux de FNA plasmatique, l'osmolalité et le seuil de libération de la VPA, et diminuer les taux de VPA et la sensibilité ainsi que les taux urinaires de nucléotides cycliques. Les taux de PNC n'ont pas changé selon la phase ou le traitement. Les différences quant au seuil de libération de la VPA selon la phase ou le traitement étaient significatives sur le plan statistique (p < 0,001). On a constaté un effet important de la phase (p = 0,02) et un effet d'interaction de la phase et du traitement (p = 0,06) pour le taux de FNA plasmatique, ainsi qu'un effet d'interaction important pour le taux de cAMP urinaire (p = 0,047) et celui de cGMP urinaire (p = 0,066). La phase et le traitement n'ont pas eu d'effet significatif sur les autres mesures.

Conclusion : La rétention de liquide au cours de la phase lutéinique peut être attribuable à une déficience relative de FNA et à une baisse du seuil de libération de la VPA. L'atténuation des symptômes que produit le traitement à la progestérone peut être le résultat d'une stimulation de la FNA et d'une inhibition de la libération ou de la synthèse de la VPA.

[ Top of document ]

Introduction

Patients with premenstrual syndrome (PMS) report many symptoms, such as swelling, bloating, thirst and salt craving, which are associated with decreased urination, increased salt and water intake, and weight gain.1,2 Hormones that could be responsible for the apparent premenstrual sodium and water retention include aldosterone (and other mineralocorticoid hormones), estrogen and arginine vasopressin (AVP). In normal women, the secretion rate of aldosterone increases during the luteal phase, perhaps in response to progesterone, as the increase is abolished when ovulation is suppressed.3 Estrogens cause sodium retention and reduce the renal excretion of sodium and water, but estrogen is a much weaker sodium-retaining hormone than aldosterone.4 Plasma vasopressin levels do not appear to be significantly higher in the luteal phase than in the follicular phase of the menstrual cycle in normal women.5,6 In studies of women with PMS, blood levels of aldosterone were not statistically different,7 while estradiol levels were similar to8 or higher than9 those in control subjects. Vasopressin levels have not been reported in PMS.

The relief of premenstrual symptoms with the onset of menstruation is often associated with a concurrent urinary frequency and reduction in thirst, salt and fluid intake, and body weight.2 This apparent diuresis and natriuresis could be promoted by atrial natriuretic factor (ANF) and brain natriuretic peptide (BNP),10 although plasma levels of ANF do not change during the menstrual cycle.11 A beneficial effect is reported by many patients with PMS treated with spironolactone7,12 and progesterone, both mineralocorticoid antagonists,13 but not estrogen.14 In double-blind trials, progesterone has not been proven to be beneficial when administered vaginally. In the case of oral micronized progesterone, there was a positive effect in a study which used 300 mg,15 but not in one which used 1200 mg,16 compared to placebo. Lower,9 similar,8 or higher17 plasma progesterone levels and increased pulse frequency and decreased pulse amplitude of progesterone secretion18 have been described in patients with PMS than in control subjects. Natriuretic peptide levels have not been reported in this condition.

The purpose of this study was to explore the basis of luteal fluid and sodium retention, including the possible role of vasopressin and atrial and brain natriuretic peptides, and the progesterone effect in PMS.

[ Top of document ]

Methods

Subjects

Six patients with PMS were studied. They were healthy nonobese women of reproductive age with no (other) major illness. None had idiopathic edema, i.e., peripheral edema and diurnal or cyclic weight change greater than 3 kg. All patients had regular ovulatory menstrual cycles and were not taking any medications. They all had a history of moderate to severe PMS with mild to moderate premenstrual fluid retention but no objective peripheral edema. Ovulation was timed with a First Reponse Ovulation Predictor Test kit (donated by Carter Horner Inc., Mississauga, Ont.) and was confirmed with a serum progesterone assay.

The diagnosis of PMS was based on evidence of recurrence of symptoms19 in the premenstrual phase and absence of symptoms in the postmenstrual phase for a minimum of 3 consecutive months. The diagnosis was corroborated by prospective daily charting of symptoms over a 3-month period. The charting of symptoms was continued during the period of the study. The severity of the symptoms was rated on a scale from 0 (none) to 3 (severe) and the sum of the symptoms was calculated for each day. A total luteal phase score at least 50% higher than the total follicular phase score was required for diagnosis.

This study was approved by the Conjoint Ethics Committee of the University of Calgary and the Foothlls Hospital, Calgary, Alta. Written informed consent was obtained from all patients.

Study design

Each patient was tested on an early follicular (mean day 6) and a late luteal (mean day 22) day in 2 different menstrual cycles with and without progesterone treatment (Fig. 1). Progesterone (100 mg 4 times daily) was administered orally from ovulation to the first menstrual day in the treatment cycle. On the morning of the second luteal phase test, patients took 200 mg progesterone orally (Women's International Pharmacy, Madison, Wis.) 2 hours before the first blood sample.

The patients followed an ad libitum diet during the study period. For each test, the patients fasted after supper for at least 12 hours prior to testing and abstained from smoking and any medications. They arrived at 7 am at the Heritage Medical Research Centre, University of Calgary, where the studies were conducted. After providing a urine sample and emptying the bladder, the patient was put at bedrest. An indwelling catheter was placed in both antecubital veins. A blood pressure cuff was placed on one thigh. After 2 baseline blood samples were taken, a 3% hypertonic saline solution was infused continuously using an infusion pump (Flo-Gard 6100, Baxter Corporation, Toronto) over 2 hours at a rate equal to 0.05 mL/kg per minute (Fig. 1). Blood pressure and thirst were monitored during the infusion. Blood samples taken from the contralateral arm were placed immediately on ice and centrifuged at 4°C. Blood samples were drawn every 20 minutes during the infusion for plasma osmolality, AVP, ANF and BNP analyses. Urine was collected immediately after the infusion and volume, osmolality, cAMP, cGMP, sodium and potassium measurements were made on both pre- and post-infusion specimens. A serum sample was obtained at zero time (pre-infusion) for progesterone assay.

Assays

Plasma ANF was measured by radioimmunoassay, as previously described.20,21 Ten mL of blood were drawn into chilled EDTA tubes containing aprotinin (Sigma, St. Louis, Mo.), 20 U/mL, and centrifuged at 3000 rpm for 30 minutes at 4°C. Plasma was removed and an equal volume of 0.l normal hydrochloric acid was added. Samples were stored at -70°C. Plasma was extracted using Sep-Pak Cl8 columns. The recovery of the extraction procedure was 83% ± 3%. The ANF standard and antibody were obtained from Peninsula Laboratory (Belmont, Calif.). The sensitivity was 2 pg per tube, and inter- and intra-assay variations were 10% and 6%, respectively. The 50% displacement point of a standard curve was 35 pg/mL. The normal range for females of reproductive age is 20 to 40 pg/mL, mean 35 pg/mL. Assays were done in duplicate and the results averaged.

BNP antibody and standard were purchased from Peninsula Laboratory (Belmont, Calif.). This antibody cross-reacts with human brain natriuretic peptide but does not cross-react with ANF or rat or porcine BNP. BNP was labelled with 125I by the chloramine-T method. Human BNP was used to construct a standard curve (2 to 500 pg per tube). About 15 000 cpm of 125I -labelled brain BNP was added after a 24-hour preincubation period and incubated again overnight at 4°C. Free and unbound fractions were separated by goat-rabbit gamma globulin (100 µL, 1:50) in the presence of normal rabbit serum (100 µL, 1%). After the addition of 750 µL of polyethylene glycol 8000 (5% solution), the tubes were centrifuged at 3000 rpm in 4°C for 20 minutes. The supernatant was aspirated and the radioactivity in the precipitate was measured in a LKB 2750 minigamma counter (Wallec, Finland). The lower limit of detection of the assay IC50 was 12.0 pmol/L (12.0 pg per tube) and the inter- and intra-assay variations were 10% and 4%, respectively.

AVP assays were performed by Dr. Norman Kasting, Department of Physiology, University of British Columbia, by radioimmunoassay.2 Four mL of blood were collected in heparinized vacutainer tubes. After centrifugation, plasma was separated, an aliquot was removed for osmolality analysis, and the remainder was divided into triplicate tubes before freezing at -70°C. For the AVP assay, plasma was extracted using Sep-Pak C18 cartridges. The recovery was at least 90%. The antiserum titer (final dilution) was 1: 210 000. There was 0.2% cross-reactivity with oxytocin and 17.0% cross-reactivity with arginine vasotocin. The sensitivity was 0.3 pg/mL or 0.12 pg per tube and the intra-assay and inter-assay variations were 8.7% and 9.6%, respectively. The volume of plasma assayed was 250 uL. The AVP assay was done in duplicate and the results averaged. The normal range in humans is 2 to 3 pg/mL.

Sodium and potassium concentrations in the urine were measured with an IL 943 flame photometer (Instrumentation Laboratory, Italy). Radioimmunoassay kits (New England Nuclear, Mississauga, Ont.) were used to measure cAMP and cGMP levels in urine. Progesterone was measured in serum by radioimmunoassay (Coat-A-Count, Diagnostic Products Corporation, Los Angeles). This assay has a sensitivity of 0.1 nmol/L, an intra-assay variation of 2.6% to 6.4%, and an inter-assay variation of 5.0% to 10.2%. The normal values for females are 0 to 5 nmol/L for the follicular phase and 13 to 108 nmol/L for the luteal phase. Osmolality was determined immediately (before freezing of samples) by freezing point depression using an Automatic Precision Systems Inc. Osmometer. Plasma and urine samples were assayed in duplicate and averaged.

Statistical analysis

Results were analysed by linear regression analysis and repeated measures analysis of variance. The regression equation was pAVP = m (pOsm - b) where m is the slope of the regression line and represents the sensitivity or gain of the osmostat, pOsm is plasma osmolality and b is the abscissal intercept, representing the theoretical threhold for AVP release. Differences were analysed by Student's paired t-tests adjusted for multiple comparisons where indicated. Differences were considered significant at the 0.05 level unless otherwise specified.

[ Top of document ]

Results

The age of the patients with PMS ranged from 27 to 42 with a mean of 36 years. All patients had a history of regular menstrual cycles with an interval of 27 to 30 (mean 28) days. The weight of the patients ranged from 47.5 to 66 kg (mean 56 kg) during the 2 test cycles. The weight was higher by 0.5 kg in the luteal phase compared with the follicular phase in the untreated cycle, whereas there was no difference in mean weight in the 2 phases in the treated cycle. Patients kept a symptom record during the 2 cycles of their tests as well as during the 3 months prior to the study. The score means were 10 and 30 for the untreated cycle in the follicular and luteal phases, respectively. In the treatment cycle, the score means were 11 in the follicular phase and 33 just before progesterone was started in the luteal phase. All patients reported symptomatic improvement with progesterone therapy.

Serum progesterone levels were measured in all patients. The mean levels were 1.12 and 15.6 nmol/L for the follicular and luteal phases of the untreated cycle, respectively, and 0.63 and 66.5 nmol/L during the same phases in the treated cycle.

In the untreated cycle, mean basal (time = 0 min) plasma osmolality (pOsm) was lower by 13.8 mmol/kg in the luteal phase (LO, mean 263.7, standard deviation [SD] 16.2 mmol/kg) compared with the follicular phase (FO, 277.5, SD 14.1 mmol/kg) (Fig. 2). In the cycle in which patients had been treated with progesterone for several (mean 7) days before the luteal phase test, this difference between the mean basal follicular and luteal pOsm values was smaller, 4.3 mmol/kg. In the progesterone treatment cycle, the mean basal pOsm was 281.2, SD 11.7 mmol/kg in the luteal phase (LP) and 285.5, SD 10.1 mmol/kg in the follicular phase (FP). Progesterone therapy appeared to increase basal plasma tonicity or reset the osmostat so that patients were less hypo-osmolar.

With infusion of hypertonic saline, there was a significant and continuous increase in pOsm in all cycles (Fig. 2). The change in pOsm with time was statistically significant (p = 0.0001). In the untreated cycle, pOsm was consistently lower, over time, in the luteal phase than in the follicular phase. This difference between the phases almost disappeared with progesterone treatment.

Mean basal plasma AVP levels tended to be lower in the luteal phase (1.31, SD 0.45 pg/mL) than in the follicular phase (1.59, SD 0.51 pg/mL) in the untreated cycle. There was a significant increase in plasma AVP levels in response to saline infusion (p = 0.003). More AVP appeared to be secreted in the luteal phase over time (Fig. 2). Thus, at 120 minutes, mean plasma AVP levels were 4.38, SD 2.20 pg/mL in the luteal phase test and 3.54, SD 1.35 pg/mL in the follicular phase test. In contrast, in the progesterone-treated cycle, the opposite effect was seen. Mean basal plasma AVP levels were higher in the luteal phase (1.57, SD 0.37 pg/mL) than in the follicular phase (1.43, SD 0.26 pg/mL) whereas, at 120 minutes, the mean plasma AVP level was higher in the follicular (4.46, SD 0.10 pg/mL) than in the luteal (3.65, SD 1.48 pg/mL) phase test. That is, a divergence of the follicular and luteal curves occurred over the duration of the saline infusion; it appeared that progesterone inhibited the release of AVP in the luteal phase.

Linear regression analysis was applied to each test. The mean (osmolality) threshold for AVP release was lower in the luteal phase (LO, mean 268.1, standard error of the mean [SEM] 20.11 mmol/kg, LP 278.6, SEM 6.62 mmol/kg) than in the follicular phase (F0 279.6, SEM 7.25 mmol/kg, FP 284.9, SEM 38.8 mmol/kg) in each cycle (Fig. 3, Table 1). With progesterone therapy the difference in the AVP threshold in the 2 phases was less marked. Paired Student's t-tests showed that the differences were statistically significant (p < 0.001) for the comparisons FO-LO, FO-FP, FP-LO, FP-LP, LO-LP, but not for the comparison FO-LP (p = 0.068). The mean thirst threshold was similarly lower in the luteal (LO 274.8, SEM 7.13 mmol/kg; LP 288.2, SEM 2.30 mmol/kg) phase than in the follicular (FO 285.3, SEM 4.45 mmol/kg; FP 291.5, SEM 3.20 mmol/kg) phase, especially in the untreated cycle. Paired Student's t-tests showed that the differences were not statistically significant.

Mean AVP sensitivity (measured from the slope of the regression line) was higher in the luteal phase (0.384, SEM 0.073) than in the follicular phase (0.301, SEM 0.025) in the untreated cycle (Table 1). The converse was true in the progesterone treatment cycle (LP 0.265, SEM 0.023, FP 0.356, SEM 0.132). Progesterone therapy in the luteal phase thus lowered AVP sensitivity. The ratio of AVP sensitivities in the luteal over the follicular values dropped from 1.27 to 0.72, or almost 50% under the influence of progesterone.

Mean basal plasma ANF was lower in the luteal (LO 42.4, SD 20.3 pg/mL, LP 36.4, SD 13.6 pg/mL) than in the follicular (FO 50.7, SD 30.8 pg/mL; FP 47.6, SD 26.9 pg/mL) phase. Plasma ANF levels increased with hypertonic saline infusion in all tests (p = 0.0001), but in the untreated luteal phase, the levels were consistently lower, resulting in a large differential between follicular and luteal values with time, which was statistically significant (Fig. 2). During progesterone treatment, the plasma ANF response to saline was increased to approximate the levels seen in the follicular phase.

A comparison of the plasma luteal to follicular (L/F) ratio in the treated cycle to the ratio in the untreated cycle (LP/FP over LO/FO) showed that there was a significant stimulatory effect of progesterone on ANF but not on osmolality or AVP levels in plasma (Fig. 4). At 80 minutes, mean plasma ANF was higher in the luteal phase (LP 123.9, SD 64.0 pg/mL) than the follicular phase (FP 82.4, SD 48.1 pg/mL) with progesterone treatment, whereas the converse was true in the untreated cycle (LO 73.5, SD 25.1 pg/mL; FO 133.1, SD 131.3 pg/mL). There was no significant change in plasma BNP levels in all studies and no effect of progesterone on BNP (Figs. 2, 4).

Fig. 5 shows boxplots of urinary osmolality, sodium, potassium, cAMP and cGMP excretion before (-20 minutes) and at the end (120 minutes) of the infusions. Saline infusion caused an increase in excretion of these electrolytes and cyclic nucleotides in both cycles. Urine osmolalities tended to be higher in the follicular than in the luteal phases of both cycles, but these differences were not statistically significant. Mean urine volume at the end of the infusion was greater in the luteal (LO 166.6 mL, LP 162.5 mL) compared with the follicular (FO 133.5 mL, FP 125.0 mL) phase in both cycles. Progesterone therapy appeared to slightly increase urinary volume and osmolality, but these effects were insignificant (Fig. 6). Osmolar clearances appeared to be higher in the luteal phase than the follicular phase in both cycles (data not shown). These differences were not statistically significant. Free-water clearances were similar in both phases and cycles.

Urinary excretion of sodium and potassium appeared to be increased in the luteal phase compared with the follicular phase after saline infusion in the untreated cycle. Treatment with progesterone appeared to be associated with a decrease in urinary potassium and a slight increase in urinary sodium excretion, but these differences were small and not statistically significant (Figs. 5, 6). In the luteal phase of the untreated cycle, there was a significantly greater excretion of urinary cAMP and cGMP compared to the follicular phase after saline loading (Figs. 5, 6). There was a reduction in urinary cAMP and cGMP levels during treatment with progesterone in the luteal phase. Progesterone treatment produced a significant increase in the difference between follicular and luteal values and a significant decrease in the luteal to follicular ratio due to its suppressive effect on cyclic nucleoide excretion (Fig. 6).

Repeated measures analysis of variance was performed for the dependent variables AVP, osmolality and log (AVP). Independent variables included in the model were phase (follicular or luteal), treatment (none or progesterone) and their interaction. No significant differences due to treatment or phase or their interaction were found. For plasma ANF, there appeared to be a significant (p = 0.02) phase effect with a phase and treatment interaction effect of borderline significance (p = 0.06).

Similar analysis for urine osmolality, sodium and potassium showed no significant difference with phase or treatment or their interaction. For urine cAMP, there was a significant (p = 0.047) interaction effect of phase and treatment. For urinary cGMP, the interaction effect was a little less significant (p = 0.066). For both cyclic nucleotides, the effect of treatment or phase was not significant.

Individual analyses were also performed within treatment and phase subgroups. No significant differences were found. Boxplots and plots for individuals indicated that the data were variable. Many more subjects (about 5 times as many) would be necessary to have enough power to demonstrate statistically significant differences.

[ Top of document ]

Discussion

In this study we showed that (a) there are biochemical differences between the premenstrual and postmenstrual phases of the cycle in patients with PMS (which may contribute to symptoms of fluid excess in the luteal phase), and (b) progesterone treatment during the luteal phase produced hormonal changes in these patients. We studied osmoregulation in patients with PMS in the symptomatic luteal period with and without progesterone therapy, using the asymptomatic or follicular phase response as a control in each cycle. This is more meaningful than comparing patients with PMS with control subjects, since the sensitivity for vasopressin secretion and the osmotic thresholds for AVP and thirst differ widely between subjects but are highly reproducible for any individual.23 The within-subject variation in the length of the menstrual cycle and its luteal phase is also considerably smaller than that between subjects.24

Progesterone was given orally for 7 to 8 days before the luteal phase test. The dose of progesterone administered (400 mg once daily) was the maximum dose taken by most patients with PMS for symptomatic relief. This is a pharmacologic dose, since the progesterone secretory rate has been estimated to be 2 to 25 mg per day during the luteal phase of the normal menstrual cycle25 and rises to 300 to 400 mg per day during the third trimester of pregnancy. A 200-mg dose was given 2 hours before the first blood sample on each luteal test day. It has been shown that peak progesterone levels are achieved at 2 hours with 200 mg oral micronized progesterone in oil and that levels remain significantly elevated for at least 6 hours.26,27 Similar absorption and kinetics studies have not been reported for other doses of this preparation.

In normal women, basal pOsm, thirst and AVP thresholds have been reported to be lower in the luteal phase than in the follicular phase.28,29 Vokes and associates29 found that the pOsm at which plasma AVP and urinary osmolality were maximally suppressed, as well as calculated osmotic thresholds for thirst and vasopressin release, were lower by 5 mosm/kg in the luteal than in the follicular phase. The lowering of the threshold was considered qualitatively similar to that observed in pregnancy.29 Whereas Vokes and associates found no change in sensitivity of osmoregulation, i.e., the amount of AVP secreted per unit rise in pOsm,29 Spruce and associates28 noted that the slope of the regression line relating AVP to osmolality was lower in the luteal phase. In contrast, our studies showed a lower sensitivity in the follicular phase in the untreated cycle, with a reversal in the progesterone treatment cycle.

The participants in our study did not have idiopathic edema, i.e., objective peripheral edema and weight change of more than 1.4 kg from morning to evening, or cyclic edema (peripheral edema and weight change of more than 3 kg within the menstrual cycle). In a study of patients with cyclic edema reported by Thompson and associates,30 basal plasma osmolality and AVP, thirst onset, and theoretical thresholds for AVP release were found to be similar to the values in healthy controls. The authors concluded that osmoregulation is normal in cyclic edema. ANF levels were not measured in these studies of normal women or of women with cyclic edema.

We have found that, in the luteal phase, patients with PMS have a lower level of plasma osmolality and a lower threshold for thirst and AVP secretion, with an increase in AVP sensitivity, compared with the follicular phase. In addition, we found that progesterone therapy increased AVP and thirst thresholds and decreased AVP sensitivity. Aubry and associates31 found that treatment of 8 normal subjects with cortisol caused an increase in the osmotic threshold for AVP and an increase in urinary flow and free water and osmolar clearances. Since the dose of progesterone used in the present study was pharmacologic, and since progesterone can bind to the glucocorticoid receptor at high concentrations,13 it is likely that the demonstrated effect of progesterone is a glucocorticoid effect. The lack of a statistically significant difference in plasma AVP levels with phase and treatment is likely due to the inadequacy of the acute stimulus given. A more rapid and pronounced increase in plasma sodium levels, with infusion of 5% rather than 3% saline, would probably induce a significant change in plasma AVP. A more hypertonic solution was, however, not used, to avoid side effects.

We also measured plasma ANF and BNP and urinary cAMP and cGMP, which are considered the second messsengers for AVP and ANF, respectively.32,33 The lack of change in plasma BNP levels with time (after saline infusion), phase of the menstrual cycle or treatment with progesterone was not unexpected. BNP, in contrast to ANF, does not appear to respond to acute stimulation, perhaps because atrial tissue contains little preformed BNP.10 Lang and associates34 also found no detectable increase in plasma BNP after acute saline load in man.

The significant reduction in plasma ANF in the luteal phase of the patients with PMS, which was demonstrated in this study, may be important in the pathogenesis of the swelling or fluid retention reported by these patients. Clarke and associates11 reported no change in plasma ANF levels during the menstrual cycle in normal women without premenstrual edema or weight gain. The reversal of this reduction by progesterone is consistent with the natriuretic action of progesterone.25 The natriuretic effect of progesterone has been ascribed to competitive inhibition of aldosterone action on the kidney and an acute aldosterone-independent inhibition of sodium reabsorption at proximal and distal sites in the nephron.35

Our studies suggest that it may be mediated by ANF and may be due to the glucocorticoid action of progesterone. Dexamethasone was found to increase plasma ANF levels before and during saline infusion in patients with primary adrenal insufficiency.36 The heart contains glucocorticoid, but not true mineralocorticoid, receptors and a putative glucocorticoid receptor site occurs in the human ANF gene.37 Since the patients with PMS had taken progesterone for a few days prior to the luteal phase test, it is possible that the effect of progesterone was on synthesis as well as release of ANF. The relief of symptoms with diuresis in patients on progesterone therapy may be due to the increase in ANF levels and the effect on AVP.

Urinary cAMP and cGMP excretion was higher in the luteal phase of the untreated cycle in our patients with PMS. This is similar to results reported in normal women38 and is unexplained. Progesterone treatment caused a reversal of this phase effect. Progesterone has been reported to reduce renal cAMP response to AVP in rat renal medullary cell cultures.32 The reduction in cAMP may explain the decreased AVP sensitivity during progesterone treatment observed in our study.

ANF has no effect on urinary cAMP excretion,39 but cGMP has been considered an indirect measure of plasma ANP. The lack of correlation between urinary cGMP and plasma ANF and the lack of a statistically significant effect on urinary electrolyte excretion of phase and treatment may be due to the lack of a sufficiently profound and prolonged stimulus and the time course used in this study.

[ Top of document ]

Acknowledgements

This study was supported by a grant from Health Canada (National Health Research and Development Program project no. 6609-1344-52) and the Foothills Hospital. Excellent technical assistance was provided by Lynda Murch, Kam Ko and Bonnie Brayshaw.

[ Top of document ]

References
  1. Bickers W, Woods M. Premenstrual tension. Its relation to abnormal water storage. N Engl J Med 1951;245:453-6.
  2. Janowsky DS, Berens SC, Davis JM. Correlations between mood, weight, and electrolytes during the menstrual cycle: a renin-angiotensin-aldosterone hypothesis of premenstrual tension. Psychosom Med 1973;35:143-54.
  3. Gray MJ, Strausfeld KS, Watanabe M, Sims EH, Solomon S. Aldosterone secretory rates in the normal menstrual cycle. J Clin Endocrinol 1968;28:1269-75.
  4. Christy NP, Shaver JC. Estrogens and the kidney. Kidney Int 1974;6:366-76.
  5. Forsling ML, Akerlund P, Stromberg P. Variations in plasma concentrations of vasopressin during the menstrual cycle. J Endocrinol 1982;95:147-51.
  6. Punnonen R, Vinamaki O, Multamaki S. Plasma vasopressin during normal menstrual cycle. Horm Res 1983;17:90-2.
  7. O'Brien PMS, Selby C, Symonds EM. Treatment of premenstrual syndrome by spironolactone. Br J Obstet Gynecol 1979;86:142-7.
  8. Rubinow DR, Hoban CH, Grover GN, Galloway DS, Roy-Byrne P, Andersen R, et al. Changes in plasma hormones across the menstrual cycle in patients with menstrually related mood disorder and in control subjects. Am J Obstet Gynecol 1988;158:5-11.
  9. Wang M, Seippel L, Purdy RH, Backstrom T. Relationship between symptom severity and steroid variation in women with premenstrual syndrome: study on serum pregnenolone, pregnenolone sulfate, 5alpha-Pregnane-3,20-dione and 3alpha-Hydroxy-5alpha-Pregnan-20-One. J Clin Endocrinol Metab 1996;81:1076-82.
  10. Espiner EA, Richards AM, Yandle TG, Nicholls MG. Natriuretic hormones. Endocrinol Metab Clin North Am 1995;24:481-509.
  11. Clark BA, Elahi E, Epstein FH. The influence of gender, age and the menstrual cycle on plasma atrial natriuretic peptide. J Clin Endocrinol Metab 1990;70:349-52.
  12. Vellacott ID, Shroff NE, Pearce MY, Stratford ME, Akbar FA. A double-blind, placebo-controlled evaluation of spironolactone in the premenstrual syndrome. Curr Med Res Opin 1987;10:450-6.
  13. Wambach G, Higgins JR. Antimineralocorticoid action of progesterone in the rat: correlation of the effect on electrolyte excretion and interaction with renal mineralocorticoid receptors. Endocrinology 1978;102:1686-93.
  14. Dhar V, Pearson Murphy BE. Double-blind randomized crossover trial of luteal phase estrogens (premarin) in the premenstrual syndrome (PMS). Psychoneuroendocrinology 1990;15:489-93.
  15. Dennerstein L, Spencer-Gardner C, Gotta G, Brown JB, Smith MA, Burrows GD. Progesterone and the premenstrual syndrome: a double-blind cross-over trial. BMJ 1985;290:1617-21.
  16. Freeman EW, Rickels K, Sondheimer SJ, Polansky M. A double-blind trial of oral progesterone, alprazolam, and placebo in treatment of severe premenstrual syndrome. JAMA 1995;274:51-7.
  17. O'Brien PMS, Selby C, Symonds EM. Progesterone, fluid and electrolytes in premenstrual syndrome. BMJ 1980;280:1161-3.
  18. Facchinetti F, Genazzani AD, Martignoni E, Fioroni L, Nappi G, Genazzani AR. Neuroendocrine changes in luteal function in patients with premenstrual syndrome. J Clin Endocrinol Metab 1993;76:1123-7.
  19. American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th ed. Washington: The Association; 1994:714-8.
  20. Wong NLM, Wong EFC, Au GH, Hu DCK. Effect of alpha- and beta-adrenergic stimulation on atrial natriuretic peptide release in vitro. Am J Physiol 1988;255:E260-4.
  21. Wong NLM, Huang D, Guo NS, Wong EFC, Hu DC. Effects of thyroid status on atrial natriuretic peptide release from isolated rat atria. Am J Physiol 1988;256:E64-7.
  22. Kasting NW, Carr DB, Martin JB, Blume H, Bergland R. Changes in CSF and plasma AVP in febrile sheep. Can J Physiol Pharmacol 1983;61:427-31.
  23. Zerbe RL, Miller JZ, Robertson GL. The reproducibility and heritability of individual differences in osmoregulatory function in normal human subjects. J Lab Clin Med 1991;117:51-9.
  24. Landgren B-M, Unden A-L, Diczfalusy E. Hormonal profile of the cycle in 68 normally menstruating women. Acta Endocrinol 1980;94:89-98.
  25. Landau RL, Lugibihl K. Inhibition of the sodium-retaining influence of aldosterone by progesterone. J Clin Endocrinol Metab 1958;18:1237-45.
  26. Maxson WS, Hargrove JT. Bioavailability of oral micronized progesterone. Fertil Steril 1985;44:622-6.
  27. Hargrove JT, Maxson WS, Wentz AC. Absorption of oral progesterone is influenced by vehicle and particle size. Am J Obstet Gynecol 1989;151:948-51.
  28. Spruce BA, Baylis PH, Burd J, Watson MJ. Variation in osmoregulation of arginine vasopressin during the human menstrual cycle. Clin Endocrinol 1985;22:37-42.
  29. Vokes TJ, Weiss NM, Schreiber J, Gaskill MD, Robertson G. Osmoregulation of thirst and vasopressin during normal menstrual cycle. Am J Physiol 1988;254:R641-7.
  30. Thompson CJ, Burd JM, Baylis PH. Osmoregulation of vasopressin secretion and thirst in cyclical oedema. Clin Endocrinol 1988;28:629-35.
  31. Aubry RH, Nankin HR, Moses AM, Streeten DH. Measurement of the osmotic threshold for vasopressin release in human subjects, and its modification by cortisol. J Clin Endocrinol Metab 1965;25:1481-92.
  32. Hatano T, Ogawa K, Kanda K, Seo H, Matsui N. Effect of ovarian steroids on cyclic adenosine 3´:5´-monophosphate production stimulated by arginine vasopressin in rat renal monolayer cultured cells. Endocrinol Jpn 1988;35:267-74.
  33. Gerzer R, Witzgall H, Tremblay J, Gutkowska J, Hamet P. Rapid increase in plasma and urinary cyclic GMP after bolus injection of atrial natriuretic factor in man. J Clin Endocrinol Metab 1985;61:1217-9.
  34. Lang CC, Choy A-MJ, Turner K, Tobin R, Coutie W, Struthers AD. The effect of intravenous saline loading on plasma levels of brain natriuretic peptide in man. J Hypertens 1993;11:737-41.
  35. Oparil SO, Ehrlich EN, Lindheimer MD. Effect of progesterone on renal sodium handling in man: relation to aldosterone excretion and plasma renin activity. Clin Sci Mol Med 1975;49:129-47.
  36. Roden M, Damjancic P, Vierhapper H. Effect of glucocorticoid substitution on the secretion of atrial natriuretic peptide (ANP) in patients with adrenocortical insufficiency. Horm Metab Res 1991;23:298-9.
  37. Seidman CE, Bloch K, Klein KA, Seidman JG. Nucleotide sequences of the human and mouse atrial natriuretic factor genes. Science 1984;226:1206-9.
  38. Turner GA, Ellis RD, Guthrie D, Latner AL, Skillen AW, Ross WM. Levels of adenosine 3´,5´ cyclic monophosphate and guanosine 3´,5´ cyclic monophosphate in single urine specimens collected from a large population of healthy subjects. Ann Clin Biochem 1982;19:77-82.
  39. Hamet P, Tremblay J, Pang SC, Garcia R, Thibault G, Gutkowska J, et al. Effect of native and synthetic atrial natriuretic factor on cyclic GMP. Biochem Biophys Res Commun 1984;123:515-27.

| CIM: August 1997 / MCE : août 1997 |
CMA Webspinners / >