INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE HEMODYNAMIC RESPONSE to hemorrhage consists of two phases (for review, see Ref. 19). During the initial "nonhypotensive" phase, arterial pressure is maintained relatively constant despite continued blood loss because of the actions of the baroreceptor reflex. However, if blood loss continues, the abrupt onset of the second "hypotensive" phase is signaled by a rapid and profound decrease in arterial pressure mediated by sympathoinhibition. Pregnancy modifies both phases. Mammals in late pregnancy exhibit a lesser ability to maintain arterial pressure during hemorrhage; in many species, arterial pressure can decrease during the first phase, and the hypotensive phase is triggered with less blood loss (for review, see Ref. 5). In addition, in pregnant conscious rabbits, the decrease in pressure during the hypotensive phase is less profound (4, 6). Whether pregnant women also demonstrate a similar deficiency has not been directly studied. However, because women in the pregnant state are more susceptible to orthostatic hypotension (2, 9), it is likely that they also are less tolerant of hemorrhage. Given that hemorrhage is a common feature of delivery, it is imperative to understand the mechanism of this change.

Studies demonstrating that the gain of baroreflex control of heart rate and sympathetic activity is attenuated during late pregnancy (for reviews, see Refs. 5 and 11) suggest that the lesser ability to maintain pressure during hemorrhage in pregnant mammals may be due to inadequate reflex responses. The baroreflex helps to maintain pressure not only by vasoconstriction but also by near maintenance of cardiac output despite blood loss (17, 19). However, whether pregnancy modifies the changes in either cardiac output or total peripheral resistance (TPR) evoked during hemorrhage has not been studied in conscious animals. Nevertheless, changes in renal and mesenteric conductances were similar in pregnant and nonpregnant conscious rabbits, and decreases in conductance of the terminal aortic vascular bed during hemorrhage were only slightly attenuated in pregnant animals (4). Thus pregnancy appears to have only modest effects on the regulation of regional resistance during hemorrhage, indirectly suggesting that regulation of cardiac output may also be altered during pregnancy. Therefore, one purpose of the present study was to test the following hypothesis: during hemorrhage, pregnant animals exhibit a lesser ability to maintain cardiac output as well as reduced reflex vasoconstriction as measured by increases in TPR. In addition, because the terminal aorta perfuses both the uterus and the hindquarters (8), it is unknown whether the attenuated terminal aortic vasoconstriction previously observed originates in the uterus or the skin and muscle of the hindquarters. Therefore, a second purpose of this study was to determine whether the response to hemorrhage of the femoral vascular bed, which perfuses the hindquarters, was smaller in pregnant rabbits.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

All studies were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee.

Female (n = 10) New Zealand White rabbits (Western Oregon Rabbit; Philomath, OR) weighing 3.8 ± 0.1 kg (nonpregnant) were used for these experiments. The rabbits were received when they were 14-15 wk old, and allowed a minimum 1-wk acclimatization period. During this time, the rabbits were slowly moved from a high-fiber diet (no. 5326, Ralston Purina) to a high-protein diet (no. 5321, Ralston Purina), increasing 10% high protein/day for 10 days. The rabbits were then maintained on 150 g/day of the high-protein diet (0.25% sodium and 16.2% protein) throughout the study period to enhance breeding efficiency; the animals consumed the high-protein diet for at least 2 wk before the first experiment. All animals were allowed free access to distilled water.

Surgical preparation. Surgery was performed to implant nonocclusive abdominal aortic and vena caval catheters as previously described (10), and either a femoral flow probe or an ascending aortic flow probe. Briefly, the animals were initially anesthetized with a ketamine cocktail containing 5:2.5:1 by volume of ketamine (100 mg/ml; 58.8 mg/kg), xylazine (20 mg/ml; 5.9 mg/kg), and acepromazine (10 mg/ml; 1.2 mg/kg) administered in a volume of 1 ml/kg im, and a surgical plane of anesthesia was maintained with 1:10 ketamine/0.9% NaCl solution administered intravenously as needed. A midline abdominal incision was made in all rabbits and indwelling polyethylene catheters with Silastic tips were implanted in the abdominal aorta (one) and vena cava (two). The catheters were subcutaneously tunneled from the abdominal cavity, and were exteriorized at the nape of the neck. During the same surgery, four rabbits were also implanted with an ultrasonic flow probe (model 2SB, Transonic Systems; Ithaca, NY) around the left femoral artery just below the inguinal ligament. The probes were wrapped with sterile silicon sheeting to prevent fat invagination and to lengthen probe life span, and the probe leads were tunneled subcutaneously to exit with the catheters. In six other rabbits, after a minimum 2-wk recovery period, a second surgery was performed to implant an ultrasonic flow probe (model 6SB, Transonic Systems) around the ascending aorta. Rabbits were initially anesthetized with the ketamine cocktail at one-half the normal dose (0.5 ml/kg), intubated, and a surgical plane of anesthesia was maintained with isoflurane (2%). Rabbits were then placed on a respirator, and the flow probe was implanted around the ascending aorta via a right thoracotomy through the second intercostal space. The probe leads were again exteriorized at the nape of the neck, and the incision was closed in layers. All probe leads and catheters were protected in a 3.5-cm plastic pillbox, which was sutured to the rabbits' skin. The rabbits were given an intramuscular injection of penicillin G procaine (60,000 U) just before surgery and the day after surgery. The animals were also injected with buprenorphine hydrochloride (0.09 mg im; Buprenex) 2-3 h after surgery, and again the next day. The neck incision was treated with topical nitrofurazone antibacterial dressing for 1 wk after surgery. The catheters were flushed immediately after surgery, and then 3 times weekly with the use of sterile 0.9% NaCl, and filled with heparin (1,000 U/ml) to maintain patency.

Hemorrhage protocol. All experiments were performed in the morning. The rabbits were first hemorrhaged in the nonpregnant state. Afterward, the animals were bred with noninstrumented proven male breeder rabbits, and this was considered day 1 of pregnancy. The hemorrhage was then repeated in each animal after 28-30 days of pregnancy (term is 31 days).

Data and statistical analysis. For Figs. 1-3, ~30-s averages of arterial pressure, heart rate, and flows were obtained from the polygraph recordings or the computerized data every 2.5 min beginning with the start of the hemorrhage. However, this method of analysis tends to obscure the details of the rapid pressure fall because pressure drops at a different time in each rabbit. Therefore, in Figs. 6-9, hemodynamic measurements were also quantified from the continuous pressure tracing beginning with the lowest pressure point or pressure nadir to determine the between group differences in the hypotensive phase. Differences between pregnant and nonpregnant rabbits in basal values and in changes during the hypotensive phase were determined with the paired t-test (21). The effect of pregnancy on the hemodynamic responses to hemorrhage was determined using two-way analysis of variance for repeated measures (randomized block) and the post hoc Newman-Keuls test (18, 21). Because for most variables, basal values were different between groups, the post hoc analysis was used to determine at which times significant differences from control could be detected within a group. Finally, the difference in the relationship between cardiac output and TPR produced during hemorrhage in pregnant versus nonpregnant rabbits was determined using analysis of covariance (ANCOVA). All statistics were performed using GB-STAT (Dynamic Microsystems; Silver Spring, MD).

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Effect of hemorrhage on mean arterial pressure and heart rate in conscious rabbits during pregnancy and before pregnancy. Hemorrhage began at time 0, and continued until pressure decreased <45 mmHg; the horizontal solid (nonpregnant) and dotted (pregnant) lines at bottom indicate the average period of hemorrhage. * P < 0.05 compared with time 0 within groups.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Effect of hemorrhage on cardiac output and stroke volume in pregnant and nonpregnant conscious rabbits. Hemorrhage began at time 0 and continued until pressure decreased <45 mmHg; the horizontal solid (nonpregnant) and dotted (pregnant) lines at bottom indicate the average period of hemorrhage. * P < 0.05 compared with time 0 within groups.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Effect of hemorrhage on total peripheral resistance (TPR) and total peripheral conductance (TPC) in pregnant and nonpregnant conscious rabbits. Hemorrhage began at time 0 and continued until pressure decreased <45 mmHg; the horizontal solid (nonpregnant) and dotted (pregnant) lines at bottom indicate the average period of hemorrhage. * P < 0.05 compared with time 0 within groups.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

All pregnant rabbits delivered live kits (7 ± 1 kits; range, 2-12 kits).

Basal values. Pregnancy induced significant alterations in the cardiovascular system at rest (Table 1). Arterial pressure and TPR were lower, and heart rate, cardiac output, and total peripheral conductance were higher. Stroke volume tended to be higher during pregnancy, but this change did not achieve statistical significance (P = 0.06). In addition, whereas 3 of 4 pregnant rabbits exhibited increased femoral flow and conductance (and decreased resistance), overall this change was not significant.

First phase of hemorrhage: cardiac output measurements in pregnant and nonpregnant rabbits. In this group during pregnancy, arterial pressure decreased significantly below control values with less blood loss (Fig. 1). In addition, arterial pressure fell <45 mmHg after a smaller percentage of the initial blood volume was removed (43.6 ± 1.7%, nonpregnant; 29.6 ± 2.2%, pregnant; P < 0.005). Despite the more rapid transition to hypotension in the pregnant rabbits, however, the increase in heart rate reached statistical significance at the same time in the rabbits in both reproductive states (Fig. 1). Nevertheless, the most pronounced difference between the pregnant and nonpregnant state was that cardiac output decreased much more rapidly when the rabbits were pregnant (Fig. 2). Stroke volume also decreased significantly with less blood loss during pregnancy (Fig. 2). Moreover, the decrease in stroke volume after 12.5 min of hemorrhage was greater during pregnancy (1.29 ± 0.14 ml/beat) compared with the virgin state (0.82 ± 0.12 ml/beat; P = 0.01). Thus it appears that the more rapid decrease in cardiac output is due in part to a greater fall in stroke volume. In addition, the lack of a greater rise in heart rate despite the more rapid transition to hypotension potentially contributed to the greater fall in cardiac output in pregnant animals.

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Relationship between cardiac output and TPR during hemorrhage in conscious rabbits during and before pregnancy.

First phase of hemorrhage: femoral flow measurements in pregnant and nonpregnant rabbits. The pregnant rabbits of this group also demonstrated deficiencies in arterial pressure maintenance during hemorrhage; during pregnancy, pressure fell <45 mmHg after only 25.0 ± 2.7% of the initial blood volume was withdrawn, but 39.4 ± 2.9% blood loss was required to decrease pressure to this level when the rabbits were studied before pregnancy (P < 0.05). In addition, whereas hemorrhage elicited significant tachycardia in nonpregnant rabbits (from 150 ± 9 to 178 ± 9 beats/min after 20% initial blood volume loss, P < 0.05), heart rate did not change significantly when they were pregnant (from 189 ± 14 to 211 ± 20 beats/min after 20% initial blood volume loss, P > 0.05). Analysis of the changes in femoral conductance or resistance was restricted to the nonhypotensive phase, or the first 10 min of hemorrhage, because the time at which the rabbits entered the hypotensive phase was widely variable and decreased the ability to detect differences between the pregnant and nonpregnant state. When the rabbits were studied before pregnancy, hemorrhage produced a significant decrease in femoral conductance (Fig. 5, inset) without altering arterial pressure [70.6 ± 2.9 to 72.9 ± 2.4 mmHg after 10 min; not significant (NS)]. In contrast, femoral conductance did not decrease significantly when the rabbits were pregnant (Fig. 5, inset) despite a tendency toward hypotension (62.0 ± 1.5 to 56.1 ± 6.3 after 10 min; NS); however, the responses of individual rabbits were quite variable. Two animals demonstrated little to no vasoconstriction, whereas the others responded similarly, as a function of time, in both the pregnant and nonpregnant state (Fig. 5). Importantly, the animals that exhibited vasoconstriction in the pregnant state also entered the hypotensive phase with less blood loss (after loss of 18.7 and 22.7% of initial blood volume, pregnant; after loss of 39.7% and 33.4%, for nonpregnant, respectively) compared with the other two animals (27.2% and 31.2%, pregnant; 47.1% and 37.5%, nonpregnant, respectively), suggesting that the impact of hemorrhage was more profound in the animals experiencing vasoconstriction. Thus it appears that reflex vasoconstriction of the femoral bed is attenuated in pregnant rabbits, at least in the context of the severity of the hemorrhage.

View larger version (23K): [in this window] [in a new window]   Fig. 5.   Effect of pregnancy on the changes in femoral conductance during hemorrhage (Hem). Both mean (inset) and individual responses of pregnant ( and dashed lines) and nonpregnant ( and solid lines) rabbits are shown. * P < 0.05 compared with time 0 within group. Gaps in the record from individual animals indicate that data were not collected because the rabbit was moving at that time.

Second phase of hemorrhage. To emphasize differences between reproductive states during the hypotensive phase, in Figs. 6-9, the data have been realigned so that the pressure nadir of the hypotensive phase is designated as time 0. As illustrated in Fig. 6, the transition to hypotension occurred rapidly when the rabbits were either pregnant or nonpregnant, but the decrease in pressure during pregnancy was less (Fig. 6 and Table 2). The fall in pressure was associated with a reversal of the initial hemorrhage-induced tachycardia toward baseline, which was not different between states (Fig. 6 and Table 2). Decreases in both cardiac output, partly caused by decreases in stroke volume, and in TPR contributed to the hypotension (Figs. 7 and 8). Whereas the decrease in cardiac output was greater, the decrease in TPR (or the increase in total peripheral conductance) was significantly smaller when the rabbits were pregnant (Figs. 7 and 8 and Table 2). Thus the smaller hypotensive response exhibited during pregnancy was due to less vasodilation. This contrast is readily apparent in the responses of the femoral vascular bed, in which significant increases in conductance were observed during the hypotensive phase of hemorrhage before pregnancy, but during pregnancy, all four rabbits actually showed further decreases in conductance (Fig. 9); thus the responses of femoral flow, conductance, and resistance were significantly different between the nonpregnant and pregnant state (Table 2).

View larger version (18K): [in this window] [in a new window]   Fig. 6.   The hypotensive phase of hemorrhage in rabbits in the pregnant and nonpregnant state: changes in mean arterial pressure and heart rate. Time 0 is the pressure nadir; n = 6, cardiac output group.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   The hypotensive phase of hemorrhage in rabbits in the pregnant and nonpregnant state: changes in cardiac output and stroke volume. Time 0 is the pressure nadir; n = 6, cardiac output group.

View larger version (18K): [in this window] [in a new window]   Fig. 8.   The hypotensive phase of hemorrhage in rabbits in the pregnant and nonpregnant state: changes in TPR and TPC. Time 0 is the pressure nadir; n = 6, cardiac output group.

View larger version (15K): [in this window] [in a new window]   Fig. 9.   The hypotensive phase of hemorrhage in rabbits in the pregnant () and nonpregnant () state: changes in femoral conductance. Time 0 is the pressure nadir. n = 4 (femoral flow group). In this group, arterial pressure decreased from 70.6 ± 2.3 mmHg (3 min before nadir) to 33.8 ± 1.1 mmHg (nadir) before pregnancy, and decreased from 60.0 ± 2.7 mmHg (3 min before nadir) to 38.1 ± 2.9 mmHg (nadir) when the rabbits were pregnant.

Plasma protein concentration and hematocrit during hemorrhage. To assess whether the more rapid decrease in cardiac output observed when the rabbits were pregnant was due to smaller relative blood volumes because of less reabsorption of fluid from the interstitium into the plasma compartment, changes in plasma protein concentration and hematocrit were determined. As expected because of the increased plasma volume, basal levels of protein and hematocrit were lower when the rabbits were pregnant (Table 3; P < 0.01). More importantly, in each reproductive state, both variables decreased during hemorrhage, indicating uptake of fluid. A significant decrease in hematocrit was detected at the same time, but plasma protein decreased with less blood loss, when the rabbits were studied before pregnancy compared with during pregnancy (Table 3). However, because of the marked differences in basal values, percent changes in hematocrit and protein were also assessed. No differences were observed in the percent decrease in plasma protein concentration at any time. The percent decrease in hematocrit was smaller when the rabbits were pregnant (P < 0.05) but only between samples taken after the pressure fall (pressure fall: 8.6 ± 1.8% vs. 5.8 ± 0.9%; 5 min after pressure fall: 12.3 ± 0.5% vs. 8.9 ± 0.6%).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study confirms that in conscious rabbits the late pregnant state alters the response of arterial pressure to hemorrhage during both the first nonhypotensive and the second hypotensive phases (4-6). Specifically, compared with virgin animals, pregnant rabbits are less able to maintain arterial pressure during the first phase of hemorrhage and also exhibit a smaller fall in arterial pressure during the second phase. This differential response is not due to time or the consecutive nature of the pregnant versus nonpregnant rabbit experiments, because pregnant rabbits bled at midgestation respond similarly to nonpregnant animals (6) and because others have shown that the response to hemorrhage in conscious rabbits is reproducible over long time periods (16). In addition, major new findings are that in conscious pregnant animals, the failure to adequately maintain arterial pressure during the nonhypotensive phase is due both to a lesser ability to maintain cardiac output, as well as to insufficient reflex vasoconstriction, possibly in part in the femoral vascular bed. The results further reveal that the smaller degree of hypotension during the second phase is due to less vasodilation, in particular, in the femoral bed.

The cardiovascular responses to hemorrhage have been extensively studied (for review, see Ref. 19). In the initial stages of blood loss, arterial pressure is nearly maintained primarily because of reflex increases in sympathetic activity. Sympathetic activation causes vasoconstriction and increases in TPR. In addition, reductions in cardiac output are minimized (17). Because pretreatment of animals with -blockers to eliminate sympathetic influences on the heart does not significantly modify the hemodynamic responses to hemorrhage, it appears that reflex increases in heart rate or cardiac contractility are not required for cardiac output and arterial pressure maintenance (12, 17, 20). Instead, relative maintenance of preload, due primarily to decreases in venous capacity, may limit decreases in cardiac output despite decreases in blood volume (7, 17).

Before pregnancy, the rabbits in the present study exhibited excellent pressure maintenance before succumbing to the hypotensive phase, and as in previous work, the stability of pressure was mediated by both increases in TPR and maintenance of cardiac output. In contrast, when the rabbits were pregnant, cardiac output began to decrease immediately on initiation of blood loss. The rapid fall in cardiac output was associated with similar increases in heart rate, despite the greater challenge to homeostasis, suggesting that inadequate reflex increases in heart rate may have contributed to the more rapid decrease in cardiac output. In addition, immediate and more profound decreases in stroke volume were observed. Given the evidence that cardiac output maintenance is mediated by decreases in venous capacity (7, 17), the greater falls in cardiac output and stroke volume may be due to an attenuation of reflex increases in sympathetic activity to veins and therefore less venoconstriction. Another possible factor is the rate of fluid movement from the interstitium to the vascular compartment. During pregnancy, there was a tendency for the hemorrhage-induced decreases in hematocrit and protein to be smaller, suggesting that less fluid absorption may have occurred. However, the decreases in hematocrit and protein were similar when expressed as a percentage of control, and significant decreases occurred well after the time significant falls in cardiac output were observed when the rabbits were pregnant. Thus less fluid reabsorption is unlikely to be a major contributor to the greater falls in cardiac output, as well as the earlier entry into the hypotensive phase, observed during pregnancy.

As a function of time, increases in TPR in response to hemorrhage were quantitatively similar in both reproductive states. However, relative to the greater decreases in cardiac output, the reflex increase in TPR was smaller and apparently insufficient to maintain arterial pressure in the pregnant animals. Previous work (17) indicates that the degree of vasoconstriction varies among vascular beds and is produced against the background of local autoregulatory and vasodilatory mechanisms. For example, in the mesenteric and renal beds, significant reflex vasoconstriction is nearly completely counteracted by local mechanisms, so that little or no net increase in resistance is observed (4, 17, 19). In contrast, in the hindquarters, significant increases in resistance occur and contribute to the increased TPR (4, 17, 19). We (4) previously reported that responses of the renal and mesenteric beds were not appreciably different between pregnant and nonpregnant rabbits but that the decrease in conductance of the terminal aortic bed during the nonhypotensive phase was significantly smaller when the rabbits were pregnant. Because the terminal aorta perfuses both the hindquarters and the uterus (8), it could not be determined whether the difference originated in the uterus, in skin and muscle of the hindquarter or both. The present results demonstrate that during pregnancy, despite a more rapid transition to hypotension, the femoral vascular conductance decreases observed in two rabbits were the same when the animals were pregnant and nonpregnant and in two others changed little when they were pregnant. Thus it appears that the smaller degree of vasoconstriction observed in the terminal aortic vascular bed may be due in part to a diminution in the vasoconstriction of skin and muscle.

Humphreys and Joels (14) investigated the effect of pregnancy on the hemodynamic response to hemorrhage, but two aspects of their experimental approach markedly reduced the duration, and therefore the study, of the first phase: first, the rabbits were anesthetized, which markedly attenuates reflex increases in sympathetic activity during the initial phase of hemorrhage (19), and second, blood loss proceeded very rapidly. Nevertheless, they noted that after hemorrhage, the pregnant animals exhibited greater hypotension due to a greater fall in TPR and hindquarter resistance; the cardiac output responses were similar (14, 15). Further work (13) revealed that the response of the femoral vascular bed to stimulation of the sympathetic nerves was similar between pregnant and nonpregnant rabbits, indirectly implying that the smaller degree of vasconstriction in the femoral bed observed in the present study was due to a smaller increase in sympathetic activity rather than decreased vascular responsiveness. The premise that the inadequate vasoconstriction was due to smaller increases in sympathetic activity is consistent with considerable previous work indicating that pregnancy attenuates baroreflex responses of sympathetic outflow, in particular during hypotensive challenges (for reviews, see Refs. 5 and 11).

Another difference exhibited by the pregnant rabbits was that the fall in arterial pressure during the hypotensive phase was considerably smaller. Measurements of TPR revealed that this difference was due to less vasodilation. Interestingly, the femoral bed of pregnant animals failed to show any vasodilation, which is in contrast to the similar degree of vasodilation observed in the mesenteric and renal vascular beds of pregnant and nonpregnant animals (4). While the afferent trigger that initiates the hypotensive phase is not known, the vasodilation has been shown to be due to withdrawal of sympathetic activity (19). The degree of vasoconstriction in the femoral bed during the nonhypotensive phase was less when the rabbits were pregnant; therefore, one explanation for the diminished hypotensive response is that there was a smaller increase in sympathetic activity initially and thus a smaller subsequent withdrawal.

In conclusion, the results of this study indicate that the lesser ability of conscious, pregnant rabbits to maintain arterial pressure is due largely to a greater decrease in cardiac output as blood loss persists but also to inadequate reflex increases in resistance in response to the greater fall in cardiac output, possibly in part in the skin and muscle of the hindquarters. The results also indicate that the smaller fall in arterial pressure during the hypotensive phase observed during pregnancy is due to less vasodilation and a failure for conductance to increase in the femoral vascular bed.