INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

HIGH PLASMA ESTROGEN AND PROGESTERONE levels can lead to fluid retention (9, 28, 32, 33, 36), hypertension (20), and edema (21, 32-34). One proposed mechanism for these changes in body fluid regulation may be due to changes in the osmotic regulation of arginine vasopressin (AVP) and thirst sensation (27, 35, 37). Studies using hypertonic saline infusion (HSI) have demonstrated a lower osmotic threshold for AVP release when plasma estrogen and progesterone are elevated, such as during the luteal phase of the menstrual cycle (27, 37) and late pregnancy (9). In a recent study using dehydration, we demonstrated that this shift in osmotic stimulation of AVP and thirst occurs with no change in body fluid balance, indicating a downward resetting of the osmoreceptors (30).

These changes in body fluid regulation are usually attributed to estrogen (1, 2, 6, 25). Progesterone may also have independent effects on water regulation, however. Water retention and edema occur during early pregnancy (9) and during use of oral contraceptives containing progesterone only (OCP) (30); under both conditions, progesterone is elevated without elevations in estrogen. Progesterone can influence neuronal function in the hypothalamus (10), and progesterone receptors are found in the paraventricular nucleus (3), a primary area involved in AVP synthesis and release. In young women taking OCP, we found a small reduction in the osmotic thresholds for AVP release and thirst onset during exercise-induced dehydration (30). However, this shift may have been influenced by exercise effects (14) or plasma volume (PV) loss that occurs concomitantly with the increases in plasma osmolality (P_osm) during dehydration (23).

The purpose of this study was to determine progesterone effects on osmotic regulation of AVP release and thirst sensation, as well as the impact of estrogen on those effects. To this end, we administered oral contraceptives containing estrogen and progesterone (OCEP) or OCP to young women and evaluated their responses to HSI (3% NaCl). HSI provides a strong osmotic stimulus but without the confounding effects of exercise and PV loss. OCEP agents deliver pharmacological levels of estrogen and progesterone, while OCP pills contain no estrogen. Thus these two oral contraceptive preparations differ significantly in their estrogen and progesterone activity, providing the appropriate conditions in which to isolate the effects of progesterone on body fluid and Na⁺ regulation. We hypothesized that OCP pills would reduce the threshold for osmotic AVP release and thirst sensation in response to HSI, as would OCEP pills. Consideration of PV and the effect of arterial pressure is essential for a complete evaluation of body fluid regulation, so we also determined the effects of our oral contraceptive regimen on Na⁺ regulation and the Na⁺-regulating hormones. Finally, we measured plasma cortisol concentration (P_[Cort]) during the infusion to evaluate potential stress responses to the intense thirst induced by HSI.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Design

The study design employed four HSI tests, which included an infusion conducted in the early follicular phase, 2-4 days after the beginning of menstrual bleeding (low estrogen and progesterone), and an infusion in the midluteal phase, 7-9 days after the luteinizing hormone peak (high estrogen and progesterone). Each woman tested her urine using an ovulation prediction kit (OvuQuick, Quidel, San Diego, CA) to determine the monthly peak in luteinizing hormone, which occurs ~48 h before ovulation. After completing the two baseline tests, the women again underwent HSI after 4 wk of OCEP or OCP treatment (single-blind, random assignment). After a 4-wk "washout" period, the subjects crossed over to the other pill treatment. During OCEP, subjects received 0.035 mg of ethinyl estradiol and 1 mg of the progestin norethindrone daily. During OCP, subjects received 1 mg/day of the progestin norethindrone. To verify phase of the menstrual cycle, plasma levels of estrogen and progesterone were assessed from the baseline blood sample before infusion.

HSI Protocol

On reporting to the laboratory, the subjects voided their bladders and were weighed to the nearest 10 g on a beam balance. The subjects were then seated in a semirecumbent position for a 60-min control period in an environmental chamber (27°C, 30% relative humidity) to ensure a steady state in PV and plasma constituents. During this control period, a 20-gauge Teflon intravenous catheter was placed in an antecubital or forearm vein in each arm with a heparin block (20 U/ml) to maintain catheter patency. A blood pressure cuff was positioned for automatic readings by a sphygmomanometric device (Colin Medical Instruments, Komaki, Japan) to monitor changes in blood pressure. A three-lead electrocardiogram (Colin Medical Instruments) provided continuous heart rate monitoring.

At the end of the control period, a baseline blood sample was taken, thirst perception was assessed, and a urine sample was collected. After these baseline samples, 3.0% NaCl was infused at 0.1 ml · kg body wt¹ · min¹ for 120 min. Blood was sampled at 10, 20, 30, 40, 50, 60, 75, 105, and 120 min during the infusion. After the infusion, the subject rested seated for the next 120 min. This recovery period consisted of a 30-min rest period followed by a 30-min drinking period (15 ml/kg body wt of water) followed by another 60-min rest period. Blood was sampled at 30, 90, and 120 min of recovery, and an extra 5 ml of blood were drawn at 10-min intervals during the final 30 min of recovery for PV determination (Evans blue dye; see Blood Volume). In total, 200 ml of blood were drawn from each subject during each experiment. Urine was collected at baseline, immediately after infusion, and 60 min after the drinking period. Thirst perception was assessed at all blood sampling times. The subjects were weighed before and at the end of the infusion period, after the drinking period, and at the end of the protocol. Blood pressure and heart rate were recorded every 15 min throughout infusion and every 30 min after infusion.

All blood samples were analyzed for hematocrit (Hct), hemoglobin (Hb), total protein (TP), plasma osmolality (P_osm), plasma concentration of creatinine (P_[Cr]), plasma AVP concentration (P_[AVP]), P_[Cort], and serum concentrations of Na⁺ (S_[Na⁺]) and K⁺ (S_[K⁺]). The baseline blood sample was also analyzed for concentrations of 17-estradiol (P_[E2]) and progesterone (P_[P4]). Blood samples at baseline, the end of the infusion, and 60 and 120 min after infusion were analyzed for concentrations of atrial natriuretic peptide (P_[ANP]) and aldosterone (P_[Ald]) and plasma renin activity (PRA). Volume, osmolality (U_osm), and concentrations of Na⁺ (U_[Na⁺]), K⁺ (U_[K⁺]), and creatinine (U_[Cr]) were measured from all urine samples.

Blood sampling. Blood samples were separated into aliquots. One aliquot was immediately analyzed for Hb and Hct. A second aliquot was transferred to a heparinized tube to be analyzed for P_osm, P_[Cr], and P_[Ald]. A third aliquot, for the determination of S_[Na⁺] and S_[K⁺], was placed into a tube without anticoagulant. The remaining aliquots were placed in tubes containing EDTA for analysis of P_[AVP], P_[Cort], P_[ANP], and PRA. The tubes were centrifuged at 4°C, and the plasma was removed. After centrifugation, the EDTA samples were frozen immediately at 80°C until analysis.

1

1

Blood Volume

Thirst Ratings

Calculations

We used the pre-HSI Hb concentration and Hct to estimate the baseline PV from the Evans blue dye measurement at the end of the experiment.

Fractional excretions of water (FE_H2O) and Na⁺ (FE_Na⁺) were calculated from the following equations


where the subscript f is glomerular filtrate, U is urine flow rate, and S_[Na⁺] is S_[Na⁺] in protein-free solution (meq/kgH₂O). Glomerular filtration rate (GFR) was estimated from creatinine clearance.

Data Analysis

Statistics. The variables over time (control tests and hormone intervention tests) were analyzed by conditions (OCEP vs. OCP) using ANOVA for repeated measures. When significant differences were found, orthogonal contrasts tested differences between specific means related to the hypothesis of interest. Values are means ± SE. Differences were considered statistically significant when P < 0.05 (BMPD Statistical Software, Los Angeles, CA).

Sample size calculation. Expected P_[AVP] responses within and between groups are derived from data from our laboratory during HSI (28). In an earlier study during HSI, the P_[AVP]-P_osm threshold decreased by 6 mosmol/kgH₂O, with an estimated pooled standard deviation for the group of 3 mosmol/kgH₂O.

()

()

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Baseline

Baseline P_[E2] and P_[P4] confirm that subjects were tested in the follicular and midluteal phases (Table 1; P < 0.05). Endogenous levels of 17-estradiol and progesterone were at follicular phase levels during oral contraceptive administration, as would be expected during exogenous estrogen-progesterone administration. During OCEP and OCP, ethinyl estradiol and norethindrone would be elevated to pharmacological levels in blood and tissue but were not measured by our assays. Preinfusion Hct was increased and estimated PV was reduced during the luteal phase compared with the follicular phase, OCP, and OCEP (Tables 1 and 2; P < 0.05). P_osm was lower in the luteal phase and during OCEP than in the follicular phase (Fig. 1; P < 0.05). However, P_[AVP] was unaffected by menstrual phase or either oral contraceptive pill (Fig. 1). Baseline P_[Cort] was greater during OCEP than in all other conditions (Fig. 1; P < 0.05).

View larger version (35K): [in this window] [in a new window]   Fig. 1.   Plasma osmolality (Posm) and plasma concentrations of arginine vasopressin (P[AVP]) and cortisol (P[Cort]) at baseline, in response to hypertonic saline infusion and recovery in the follicular and midluteal menstrual phases, and during oral contraceptive treatment with combined estrogen and progesterone (progestin + estrogen) and progesterone only (progestin). *Luteal phase vs. follicular phase. Luteal phase vs. progestin + estrogen. #Follicular phase vs. progestin + estrogen. ¶Progestin + estrogen vs. progestin. Differences were accepted as significant at P < 0.05. Values are means ± SE.

Preinfusion P_[Ald] was increased in the luteal phase relative to the follicular phase, although PRA was greater in the luteal phase, OCEP, and OCP than in the follicular phase (Fig. 2; P < 0.05). U, U_osm, Na⁺ excretion, C_H2O, and C_osm were unaffected by phase or pill treatment, indicating that baseline hydration levels were similar among the treatment days (Table 3). Although mean baseline U_osm is greater than mean P_osm during the follicular phase, OCP, and OCEP, C_H2O is slightly above zero under these three conditions. These apparently inconsistent findings may be the result of variability in the subjects' U_osm within each condition at baseline (82-858 mosmol/kgH₂O) concomitant with only small mean differences in U_osm among the conditions (Table 3). This anomaly led to fairly wide variability in U (0.29-6.00 ml/min), C_H2O (2.02-4.31 ml/min), and C_osm (3.46-0.57 ml/min) across the four conditions. Perhaps because of differences in the subjects' habitual water intake, the subjects' water intake (10 ml/kg body wt) on the morning of the tests did not fully hydrate some women but "overhydrated" others. However, within each woman, the hydration levels were similar at baseline across all four tests.

View larger version (30K): [in this window] [in a new window]   Fig. 2.   Plasma renin activity (PRA) and plasma concentrations of plasma aldosterone (P[Ald]) and atrial natriuretic peptide (P[ANP]) at baseline, in response to hypertonic saline infusion and recovery in the follicular and midluteal menstrual phases, and during oral contraceptive treatment with combined estrogen and progesterone and progesterone only. *Follicular phase vs. luteal phase. #Follicular phase vs. progestin + estrogen. Follicular phase vs. progestin. §Luteal phase vs. progestin. Luteal phase vs. progestin + estrogen. Differences were accepted as significant at P < 0.05. Values are means ± SE.

Heart rate (73 ± 3, 71 ± 2, 70 ± 2, and 69 ± 3 beats/min) and mean blood pressure (82 ± 2, 80 ± 3, 76 ± 2, and 76 ± 3 mmHg) in the follicular and luteal phases, OCP, and OCEP, respectively, were also unaffected by menstrual cycle phase or by oral contraceptive treatment before HSI.

HSI

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Mean P[AVP] and mean thirst responses to increases in Posm during hypertonic saline infusion in the follicular and luteal phases and during oral contraceptive treatment with progesterone only and combined estrogen and progesterone. Means at 30, 60, 90, and 120 min of hypertonic saline infusion are used for clarity. Values are means ± SE.

PRA and P_[Ald] decreased during HSI in all conditions, but only P_[Ald] remained greater in the luteal phase than in the follicular phase (Fig. 2; P < 0.05). During HSI, P_[Cort] fell steadily in all conditions but remained greater with OCEP (Fig. 1; P < 0.05). Renal Na⁺ and K⁺ excretion increased similarly during HSI in all conditions, as did C_osm and C_H2O (Table 3). HSI had no effect on heart rate (69 ± 2, 72 ± 3, 72 ± 2, and 68 ± 2 beats/min) or mean blood pressure (83 ± 4, 83 ± 4, 83 ± 3, and 81 ± 4 mmHg) for follicular and luteal phase and OCP and OCEP, respectively under any of the four conditions. There were no treatment-by-time interactions for any of the blood, renal, or cardiovascular variables during HSI.

Recovery

During the recovery period, neither renal function nor electrolyte excretion was affected by menstrual phase or oral contraceptive administration (Table 3). Moreover, overall fluid balance (i.e., HSI + drinking  urine output) was unaffected by either phase of the menstrual cycle or oral contraceptive administration (19 ± 1, 19 ± 1, 20 ± 1, and 20 ± 1 ml/kg body wt for follicular phase, luteal phase, OCP, and OCEP, respectively). Heart rate (68 ± 3, 71 ± 3, 75 ± 2, and 70 ± 3 beats/min) and mean blood pressure (78 ± 2, 78 ± 3, 78 ± 3, and 76 ± 3 mmHg) for follicular phase, luteal phase, OCP, and OCEP, respectively, during recovery were also not affected by menstrual cycle phase or oral contraceptive treatment. There were no treatment-by-time interactions for any of the blood, renal, or cardiovascular variables during recovery.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our major finding was that administration of oral contraceptive pills containing estrogen with progesterone, but not pills containing progesterone alone, decreased the osmotic threshold for AVP release during HSI in young women. Our findings support previous studies in which AVP secretion persists during HSI at lower P_osm at periods in the menstrual cycle corresponding to high P_[E2], leading to lower renal free water excretion and maintenance of the lower plasma tonicity (27, 35, 37). The present investigation extends these earlier findings demonstrating no independent effect of progesterone and indicates that this shift in osmotic regulation of AVP during OCEP and the luteal phase is primarily an estrogen-mediated effect. The shift in osmotically mediated AVP release occurred without excess fluid retention during HSI or recovery. Thus the estrogen-induced shift in osmotic regulation of AVP represents a shift in body water regulation to a lower P_osm operating point or a resetting of osmoreceptors (30). Moreover, although elevated estrogen and progesterone levels were associated with greater PRA and aldosterone, these increases in the Na⁺-regulating hormones did not drive up blood pressure or Na⁺ retention. Finally, OCEP led to increases in P_[Cort] before, during, and after HSI, although without any measurable metabolic or cardiovascular effects; this P_[Cort] increase was likely due to changes in cortisol-binding protein induced by the synthetic hormones.

The estrogen-related shift in osmotic regulation of AVP release most likely occurs via the central nervous system. Estrogen acts directly on AVP-synthesizing neurons in the hypothalamus (1, 2, 6, 25), and estrogen receptors have been identified in the nuclei of neurophysin- and AVP-producing cells in the mouse supraoptic nucleus (25). There is also evidence of angiotensinergic innervation of vasopressinergic cells in the paraventricular and supraoptic nuclei, which are modulated by estrogen (31). In addition, estrogen may impact AVP release indirectly via catecholamines (7).

Progesterone administration without estrogen seemed to have little effect on osmotic regulation of AVP. In an earlier study, progesterone treatment increased the threshold and decreased the slope of the P_[AVP]-P_osm regression lines in women with premenstrual syndrome (38). In our study, osmotic threshold for AVP release in OCP was similar to that in the luteal phase and OCEP. In addition, during OCP, there was a shift to a lower osmotic threshold for AVP release relative to the follicular phase in some, but not all, subjects. Taken together, these observations suggest an inconsistent effect of OCP on osmotic AVP regulation. There is evidence that progesterone receptor activity in the brain stimulates AVP production from the paraventricular nucleus, although this AVP stimulation by progesterone appears to depend on the presence of estrogen (3). During the midluteal phase and during OCEP, estrogen concentrations in blood and tissue are greatly increased over follicular phase levels. Even though blood and tissue estrogen concentrations remained at follicular phase levels during OCP, follicular phase tissue levels of endogenous estrogen can vary considerably among individual women. Thus we can only speculate that subtle differences in hypothalamic endogenous estrogen levels among the subjects during OCP led to the variability in the P_[AVP] response.

The different P_[AVP] responses to osmotic stimulation during the luteal phase vs. OCP may be due to structural or metabolic differences in endogenous progesterone vs. the synthetic progestin in OCP, norethindrone. Actions of progesterone in the brain may respond only to progesterone, and not to norethindrone, a progestational derivative of testosterone, which differs in structure from endogenous progesterone. Unlike endogenous progesterone, norethindrone does not produce the metabolite 3-hydroxydehydroprogesterone (11), a primary regulator of the GABA_A receptor, which modulates a number of neural pathways (24). Norethindrone metabolites may have similar effects on GABA_A, and there is evidence indicating that oral contraceptives similar in structure to norethindrone increase GABA_A levels in rat brain (8, 26). However, progestin effects on GABA_A have not been demonstrated in humans or in the hypothalamus of animals, so the impact on AVP release needs to be investigated further.

Although the most likely mechanism for the sex hormone-related shift in osmotic regulation of AVP is the central nervous system, peripheral mechanisms, such as changes in PV, also play an important role in the regulation of AVP release (23). However, although changes in PV shift the P_[AVP]-P_osm response curve, in our study PV was decreased in the luteal phase but was increased during OCEP, despite similar reductions in the osmotic thresholds for AVP release. Furthermore, the difference in PV between the follicular phase and the luteal phase and between the follicular phase and OCEP (<10%) would have increased central venous pressure by <1 mmHg (15), an insufficient change in central venous pressure to load cardiopulmonary baroreceptors enough to cause changes in the P_osm-P_[AVP] response curve (23).

Corticotropin-releasing hormone and AVP interact at the level of the anterior pituitary (16), so we determined whether HSI induced a P_[Cort] response. P_[Cort] was greater only during OCEP relative to the other three experimental days, which is consistent with earlier studies and has been attributed to increases in cortisol-binding globulin (CBG) and associated with changes in the binding characteristics of cortisol to CBG during estrogen administration (5). Circulating cortisol is usually ~93-96% bound: ~83% to CBG and ~10% to albumin. However, during estrogen administration, cortisol binding can increase to ~99%, with the greater binding primarily to CBG. We attempted to measure differences in free P_[Cort] across our four conditions using filters (Pall Filtron, Ann Arbor, MI) designed to separate free cortisol from its bound form. Once filtered, we found no differences in the free, biologically active portion of circulating cortisol among the different conditions. Thus the greater circulating cortisol should have had no significant physiological effects on our subjects. The decrease in P_[Cort] over the course of the experiment in all four conditions was likely due to the normal circadian fall in cortisol release from its early morning peak to its later-day nadir.

Our data did not show a strong effect of estrogen or progesterone on the osmotic regulation of thirst. A study by Vokes et al. (37) demonstrated a 5 mosmol/kgH₂O fall in this threshold in the midluteal vs. the early follicular phase. We did not see any statistically significant effect of menstrual cycle or oral contraceptive pills on osmotic thirst regulation; however, we found trends showing that the thirst threshold fell from 286 mosmol/kgH₂O in the follicular phase to 283 and 284 mosmol/kgH₂O in the luteal phase and OCEP, respectively. Moreover, an earlier study also demonstrated a nonstatistically significant 3 mosmol/kgH₂O fall in thirst onset during osmotic stimulation between the follicular and luteal phases of women with normal menstrual cycles (35). Taken together, these studies suggest a small shift in the osmotic regulation of thirst. The difficulty in detecting these differences consistently may be a result of variability in the subjects' subjective interpretation of the thirst scales or subtle differences in the scales themselves (22).

Arterial pressure also plays an important role in body fluid regulation, primarily by modulating Na⁺ excretion. During the luteal phase, a progesterone-mediated inhibition of aldosterone-dependent Na⁺ reabsorption at distal sites in the nephron produces a transient natriuresis and a compensatory stimulation of the renin-aldosterone system (19). The renin and aldosterone stimulation is a component of a system that evolved to maintain blood pressure and plasma water and Na⁺ content during the progesterone peak in the luteal phase. The progestin in OCEP and OCP (norethindrone), in contrast to endogenous progesterone, lacks antimineralocorticoid properties and did not increase P_[Ald], alter Na⁺ excretion, or increase blood pressure.

Our studies found a decrease in the osmotic threshold for AVP release during the administration of estrogen with progesterone, although there were no changes in osmotic regulation of AVP with progesterone administration alone. These data support the hypothesis that the observed changes in osmotic AVP regulation during the luteal phase and combined oral contraceptive administration are mediated by estrogen, most likely acting in the central nervous system. Thus mechanisms for the water retention and edema during periods of increased progesterone need to be addressed. Finally, the progestin in oral contraceptive pills differs structurally from endogenous progesterone; therefore, studies examining changes in body fluid distribution using the endogenous form of progesterone should be considered.

Perspectives