INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

IT HAS PREVIOUSLY BEEN SHOWN that pregnancies complicated by fetal growth restriction and/or preeclampsia are characterized by subnormal expansion of the plasma volume (PV) compartment during pregnancy (11, 17). On the basis of this observation, it is generally assumed that PV expansion is important for normal pregnancy development. Serial measurement of a number of hemodynamic and volume-related variables in human (12, 26), baboon (23), and rat pregnancy (27) has provided indirect evidence for the concept that the first adaptive change in hemodynamics in pregnancy is generalized vascular relaxation, giving rise to the institution of a high-flow, low-resistance circulation and PV expansion. The concomitant rise in plasma renin activity (7, 31) and the decreased threshold for vasopressin release (9, 10) indicate that the consequence of the latter is a relatively underfilled vascular bed (24a), which would normally trigger volume retention. Support for this concept comes from Hytten et al. (18), who demonstrated an increase in TBW in normal pregnant women on a standard salt diet, and also from studies reporting a lack of decrease in hematocrit (Hct) in pregnant rats on a sodium-restricted diet (14, 19, 24). In contrast, Baylis and Munger (5) reported a normal increase in PV in early pregnant rats subjected to complete sodium deprivation. The latter finding suggests that PV expansion in early pregnancy may be achieved by a fluid shift from the interstitium to the intravascular compartment. Because the evidence for both concepts is based on quantification of PV or Hct only, the experimental evidence for either concept of PV expansion is still lacking. The objective of the present study was to determine which one of these two mechanisms is operative in early pregnancy. We therefore measured simultaneously and serially total body water (TBW), extracellular volume (ECFV), and PV together with Hct and plasma osmolality in chronically instrumented pregnant and nonpregnant rats.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Animal preparation. Thirteen Wistar rats (Charles River, Sulzfeld, Germany) were used at the age of 3-4 mo. In addition, a separate group of 10 Wistar rats, at the age of 1 yr, were used for blood donation. All facilities and procedures were approved by the Animal Care and Use Committee of the Universiteit Maastricht. The animals were allowed 1 wk of acclimatization to the centralized experimental animals facilities. This included a 12:12-h light-dark cycle and free access to standard rat laboratory food (salt intake ~70 mg/day; Hope Farms, Woerden, The Netherlands) and acidified water. After 1 wk, 13 animals were mated with a fertile male. The presence of a sperm plug in the cage, which was confirmed in eight of these animals, was defined as day 1 of pregnancy. The animals without a sperm plug in their cage or in the vagina had not mated and were allocated to the nonpregnant group (NP, n = 5). All 13 animals underwent surgery using aseptic techniques as described previously (27). In the pregnant animals (P), surgery was always performed on day 1 of pregnancy. Before surgery, general anesthesia was induced using ketamine (50 mg/kg im) and xylazine (5 mg/kg im). A polyethylene catheter (0.61-mm OD, 0.28-mm ID, heat-sealed into a piece of tubing 0.96-mm OD, 0.58-mm ID), filled with heparinized saline (5 IU/ml 0.9% NaCl), was inserted into a femoral artery and advanced ~4 cm, resulting in the positioning of the catheter tip within the abdominal aorta just below the renal arteries. Another polyethylene catheter (0.96-mm OD, 0.58-mm ID, melted with ether to a piece of silicon) was inserted into the ipsilateral femoral vein and advanced ~4 cm into the inferior vena cava. The catheters were closed with a metal pin, tunneled subcutaneously to the neck, and fixed between the shoulder blades.

Experimental design. TBW, ECFV, and PV were determined with deuterium oxide (D₂O; Ref. 30), sodium bromide (NaBr; Ref. 29), and Evans blue (EB; Ref. 15), respectively, on the basis of their degree of dilution in their distribution spaces. These compounds tend to accumulate when used repeatedly in one animal, particularly when the adjacent intervals are <48 h. This was noted in a preceding pilot study in three NP rats. On the basis of the results of this pilot study, we concluded that a washout period of at least 4 days should be adopted to obtain reliable consecutive data points and also that the measurement error increased progressively when more than three consecutive measurements were performed in one animal. To describe the entire interval of early pregnancy (from day 4 to day 12), we decided to use an interval of 4 days between measurements and to limit the number of measurements to three in each animal. Therefore the animals were measured 4, 8, and 12 days postconception (P) or postsurgery (NP). At the time of each measurement, the animal was placed in an experimental cage. The catheter ends between the shoulder blades were connected to a polyethylene extension catheter (0.61-mm OD, 0.28-mm ID), which allowed sampling and infusion without manipulation of the awake animal. One hour later, the experiments were started by blood sampling (~0.7 ml) for later measurement of the baseline concentrations of D₂O, NaBr, EB, Hct (microcapillary method), and plasma osmolality (Osmomat 030-D, Gonotec). When sampling was completed, a weighed cocktail (~0.25 g) of D₂O (~0.1 ml), NaBr (~1.54 mg), and EB (~0.1 ml, 2%) was injected intravenously as follows. The syringe with the cocktail was connected to the extended venous catheter, ~0.75 ml of blood was withdrawn and mixed with the cocktail in the syringe, and then the ~1 ml of blood-cocktail mixture was injected into the animal. To ensure that the entire cocktail was administered, the syringe was slowly refilled with blood and reinjected a total of four times. The entire injection procedure lasted ~1.5 min. The catheter was flushed with saline. The amount of flushing was kept to a minimum (dead space + 0.1 ml) to avoid interference with the existing distribution of fluid over the various compartments in the animal. Ten minutes after first blood-cocktail injection was completed, the serum concentration of D₂O, NaBr, and EB can be considered to have reached steady state in their distribution compartments. This is based on repeated observations in the three NP animals evaluated during the preceding pilot study. Therefore, at that time, a second heparinized blood sample of 0.7 ml was collected for later measurement of the D₂O, NaBr, and EB concentrations. After completion of this second sampling, the animals received a transfusion of 1.4 ml of heparinized blood from a NP donor rat obtained by cardiac puncture. Ten minutes later, the catheter extensions were disconnected, the catheters were flushed and closed with a metal pin, and the animals were returned to their permanent cages. The experiments were always performed between 9:00 AM and 2:00 PM. After we completed the experiment on day 12, the animals were killed by CO₂ inhalation. In the P animals, litter size, fetal and placental weights, and fetal viability were determined to exclude interference with normal pregnancy development. In the NP animals, the uterus was palpated before every measurement session, and the absence of pregnancy was verified by abduction after the animals were killed. TBW, ECFV, and PV were calculated from the degree of dilution of D₂O, NaBr, and EB, respectively, after correction for baseline concentrations.

Statistics. Data are presented as medians with range. Within groups, each observation was compared with experimental day 4 (Wilcoxon rank-sum test, P < 0.05). Consistency in changes between days 4 and 12 was determined using Friedman's two-way ANOVA by ranks on the consecutive observations within the two groups. P < 0.05 (2 sided) was considered significant.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Maternal weight (MW) at day 4 was 250 and 208 g in the NP and P groups, respectively (Table 1), and only increased in the P group in the course of the experimental period. Litter size and viability as well as fetal and placental weights were comparable to previous observations in P rats (25). In the NP group, neither TBW nor ECFV and PV (presented as absolute and relative value) had changed consistently on days 8 and 12 relative to day 4. In contrast, in the P group, TBW, ECFV, and PV tended to increase in the course of early pregnancy, an effect that was significant by day 12. In the NP group, the median Hct had dropped slightly but consistently only on day 12 (from 39 to 38 vol%). In the P group, the Hct on days 8 and 12 was ~10% lower than that on day 4. In neither group did plasma osmolality change consistently in the course of the experimental period.

To evaluate whether PV had increased in conjunction with an overall rise in TBW or as a results of a fluid shift at the cost of the interstitial volume, we calculated the relative size of each fluid compartment at consecutive measurement sessions (Table 1). In the NP group, TBW (%MW), ECFV (%TBW), and PV (%ECFV) had not changed consistently at the time of the consecutive measurement sessions. In the P animals, TBW (%MW) had slightly increased by day 12 compared with day 4. Neither ECFV (%TBW) nor PV (%ECFV) had changed significantly in the course of the experimental period. Interstitial fluid volume (ISFV; = ECFV  PV) was calculated for each group and together with PV and TBW is depicted in Fig. 1. In the NP animals, ISFV had changed inconsistently. In contrast, in the P group, ISFV had increased in concert with the increase in PV and TBW.

View larger version (24K): [in this window] [in a new window]   Fig. 1.   Plasma volume (PV), interstitial fluid volume (ISFV), and total body water (TBW) on days 4, 8, and 12 postsurgery (A; n = 5) or postmating (B; n = 8). Data are presented as medians together with 75th and 25th percentiles. * Statistically significant difference, P < 0.05 (Friedman's 2-way ANOVA by ranks).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The objective of the present study was to determine whether the increase in PV during pregnancy is established by fluid retention or by a fluid shift within the ECFV from the interstitium toward the intravascular compartment. To this end, we simultaneously measured TBW, ECFV, and PV together with Hct and plasma osmolality, in chronically instrumented pregnant and nonpregnant rats. In the nonpregnant group, the pregnancy-related weight gain and the decrease in Hct was absent. This not only confirmed their nonpregnant status but also excluded pseudopregnancy. The indicators used in the present study, D₂O, NaBr, and EB, have a measurement error of 2, 5, and 5%, respectively (28-30, 32). When ISFV is calculated by subtracting PV from the ECFV, the error in this parameter will be equal to the sum of the two contributors, i.e., ~7%. In the present study, the baseline values observed for TBW, ECFV, and PV were within the normal range previously described for rats (1, 2, 8, 13, 16, 20, 21, 25). To ensure that the sensitivity of the applied method is sufficient to distinguish between the two possible mechanisms of PV expansion, we estimated the hypothetical change in the ECFV, PV, and TBW for both options. If the rise in PV in early pregnancy (~25%) (1, 2, 4, 5, 16) would result entirely from a fluid shift at the cost of the ISFV, the 25% increase in PV would have led to an ~4% decrease in ISFV. In that case, the change in ISFV would remain undetected, since the magnitude of the change in this parameter is smaller than the measurement error of 7%. Discrimination between either a fluid shift or fluid retention to establish PV expansion should therefore be based on whether ECFV and TBW increase in concert with PV. In this study, we observed a consistent increase in both ECFV (+24%) and TBW (+16%) by day 12 of pregnancy. Plasma osmolality did not fall consistently in the course of the experimental period, an observation also reported by others on day 12 (1, 3, 6). On the other hand, these authors found a consistent ~4% fall in plasma osmolality by gestational day 13. The small fall in plasma osmolality together with the inaccuracy in this measurement is likely to have led to this discrepancy. Hct decreased rather abruptly by day 8, probably before the gradual increase in PV. The lack of a simultaneous inverse change in Hct and PV is most likely a consequence of our methodology, in which both PV and Hct are measured intermittently by techniques known to have a limited precision.

The concomitant increase in TBW, ECFV, and ISFV suggests that, in normal rat pregnancy, PV expansion is achieved by fluid retention rather than by a shift within the ECFV. However, because our methodology is associated with a fairly large measurement error, we cannot exclude the possibility that, together with the increase in ISFV, some fluid is also displaced from the ISFV toward the PV. Our results are in agreement with those obtained by others (14, 19, 24) who did not observe an increase in blood volume after sodium deprivation. However, these results are in conflict with those reported by Baylis and Munger (5), who observed normal PV expansion in pregnant rats subjected to zero sodium intake established by a fluid shift from the ISFV toward PV. It should be emphasized, however, that sodium deprivation is a stressful condition which might alter the physiological adaptation to pregnancy. It cannot therefore be excluded that the pregnancy-related increase in PV is accomplished by a different mechanism under such circumstances. In addition, the latter observation does not exclude the possibility that TBW and ECFV may have increased, e.g., by water retention triggered by nonosmotic vasopressin release, to compensate for the inability to raise sodium retention. Because plasma sodium concentration and plasma osmolality were not determined in that study, it is not possible to confirm such an effect.

Water exchange in the capillary bed depends on the capillary and interstitial hydrostatic pressures on the one hand and their counterparts, capillary and interstitial oncotic pressures, on the other hand. These so-called Starling forces preserve a continuous balance of fluid movement within the capillary bed. Under normal physiological conditions, the sum of ISFV and PV, i.e., ECFV, is maintained within narrow limits that are controlled by baroreceptors and volume receptors located over the entire cardiovascular system (22). In the presence of a high-flow, low-resistance circulation, a higher circulating volume is required. After activation of the volume sensors, the ECFV expands rapidly by renal sodium retention and, in a number of conditions, also by dilution (osmoresetting). When sodium intake is zero, ECFV expansion can be achieved by hemodilution in response to nonosmotic vasopressin release.

We conclude from the present study that, in normal rat pregnancy, PV expansion is primarily achieved by fluid retention rather than by a shift within the ECFV.