INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

IT IS GENERALLY ACCEPTED that endurance exercise training increases plasma volume (PV) (3, 6, 7). This exercise-induced hypervolemia is associated with an increase in plasma albumin content (3, 6). The additional plasma albumin acts, via its colloid osmotic properties, to retain water in the vascular compartment. Convertino et al. (3) showed a progressive expansion of plasma albumin content during 3 days of endurance exercise training. Gillen et al. (6) showed that plasma albumin content increased within 1 h after intense exercise and remained elevated for 40 h. Although the importance of albumin in exercise-induced hypervolemia is evident, the mechanism(s) by which plasma albumin content increases after exercise remains unclear. Possible mechanisms that would act to increase plasma albumin content after exercise include 1) increased lymphatic return of albumin, 2) increased albumin synthetic rate, and 3) decreased transcapillary escape rate of albumin.

Lymph flow rate and lymphatic return of albumin are stimulated by exercise and by increased fluid pressure in exercising muscles (1, 11). The increase in plasma albumin content within the 1st h after exercise is likely due to increased lymphatic return of protein, since the time course of increase in albumin synthesis is too slow to account for the immediate increase in plasma albumin content (6, 7). Yang et al. (20) and Nagashima et al. (13) reported that albumin synthetic rate increased 24 h after exercise and would contribute to the increase in plasma albumin content 24 h after exercise. Haskell et al. (8) demonstrated that the transcapillary escape rate of albumin decreased 22 h after intense exercise. Therefore, we suspect that the latter two mechanisms contribute very little to the rapid plasma albumin expansion that occurs within 1 h after intense exercise (6).

We recently demonstrated that the exercise-induced PV and plasma albumin content expansions were attenuated in the supine posture (14). We proposed that posture impacted lymphatic return of albumin into the vascular compartment. Lymph returns to the general blood circulation via the junction of the thoracic duct with the jugular and subclavian veins. Lymph flow generally proceeds against a pressure gradient as the result of intrinsic contractile activity. However, the normal outflow pressure that the lymphatic system must overcome is central venous pressure (CVP) (2). We hypothesized that elevation in CVP (i.e., lymphatic outflow pressure) in the supine posture limits lymphatic albumin return to the vascular compartment. This increase in lymphatic outflow pressure accounts for the lack of PV expansion after exercise in the supine posture.

To examine this hypothesis, we used a nonexercise model involving bolus saline infusion (4, 15) to stimulate lymph flow and lymphatic return of protein. We manipulated lymphatic outflow pressure in the upright and supine posture mechanically through application of lower body positive (LBPP) and negative pressure (LBNP). Specifically, we increased lymphatic outflow pressure in the upright posture by applying LBPP in an attempt to limit lymphatic protein transport into the vascular space and decreased lymphatic outflow pressure in the supine posture with LBNP to promote lymphatic protein transport into the vascular space. We tested the hypothesis that changes in posture act to modify lymphatic albumin return primarily via its mechanical impact on CVP and lymphatic outflow pressure.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects. Twenty young, healthy, nonsmoking, untrained but active volunteers (10 men and 10 women) participated in the study. Female subjects were not using estrogen-based contraceptives 2 mo before and during the study period. We gave each subject an orientation of the laboratory and explained the procedures and risks involved in the study before they participated. Written informed consent to a protocol that had been approved by the Yale University School of Medicine Human Investigation Committee was obtained from each participant. Participants were randomly divided into an upright and a supine group. The subjects in the upright group (n = 10, 5 men and 5 women) had a mean age of 25.4 ± 1.6 yr, a body weight of 64.9 ± 3.3 kg, and a height of 171.7 ± 2.9 cm. The subjects in the supine group (n = 10, 5 men and 5 women) had a mean age of 26.9 ± 1.7 yr, a body weight of 64.9 ± 3.6 kg, and a height of 170.7 ± 3.0 cm.

Protocol. The study consisted of two sets of experiments, and each set included a control and a treatment experiment. In the first set, the control experiment was in the upright seated posture, and the treatment consisted of 40 mmHg of LBPP in the same posture. LBPP was produced with medical antishock trousers covering the legs (MAST III-AM, David Clark). In the second set, the control experiment was supine rest, and the treatment consisted of 20 mmHg LBNP in the supine posture. LBNP was created with the subject's lower body placed within an LBNP box and sealed at the iliac crest with a flexible Neoprene skirt. The order of experiments was randomly assigned for each subject. The interval between experiments was 10-15 days in male subjects and 25-30 days in female subjects. Female subjects performed the experiments only during the first 5 days of their menstrual cycle (early follicular phase) to control hormonal effect on fluid balance.

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Experiment protocol. CV, cardiovascular parameter measurements; Blood, blood sampling; Urine, urine sampling; Base1 and Base2, baseline; Infu, saline infusion; Rec1-Rec4, recovery; PV, plasma volume measurement; LBNP and LBPP, lower body negative and positive pressures. Time is in minutes.

Blood analysis. Blood samples were analyzed in triplicate for hematocrit (Hct, microcentrifuge) and Hb concentration (cyanmethemoglobin method, Sigma Chemical) immediately after the samples were collected. We transferred part of the blood to a lithium-heparin-treated Vacutainer and centrifuged the sample at 4°C for 15 min. The obtained plasma was used to determine plasma osmolality (freezing-point depression; Advanced Digimatic Osmometer model 3DII, Advanced Instrument), plasma albumin concentration (bromcresol green, Sigma Chemical), and total protein concentrations (biuret method, Sigma Chemical). The remaining blood was transferred to a Vacutainer with no additives, allowed to clot, and centrifuged at 4°C for 15 min. The obtained serum was used to determine serum sodium and potassium concentrations (model IL943 Automatic Flame Photometer, Instrumentation Laboratory). Blood samples from Base1, Base2, Rec2, and Rec4 were also transferred to K₃-EDTA-treated Vacutainers and centrifuged at 4°C for 15 min for plasma renin activity and aldosterone analysis and to K₃-EDTA-aprotinin-treated Vacutainers for atrial natriuretic peptide (ANP) analysis. PV was measured with an Evans blue dye (T-1824) dilution method. Percent change in PV from a certain time point (t) during the experiment to baseline (Base1) was calculated from Hb and Hct values with the following formula

Absolute values of PV at each time point were converted by correcting the percent changes in PV with the absolute value made with Evans blue dye.

Urine analysis. Subjects were requested to empty their bladders completely when they urinated. We measured urine volume with a graduated cylinder to the nearest milliliter. Urine osmolality and sodium and potassium concentrations were measured using the methods described above.

Cardiovascular variables. We monitored heart rate (HR), cardiac stroke volume (SV), thoracic impedance (Z₀), and blood pressure 5 min before each blood sampling. HR was monitored from an electrocardiogram (EK-8 Electrocardiograph, Soma Technology). SV and Z₀ were measured using an electrical thoracic impedance method (model 304B, Minnesota impedance plethysmograph). Cardiac output was calculated as HR × SV. We measured systolic (SAP) and diastolic (DAP) arterial pressure on the right arm with an automated cuff system (model STBP-780, Colin). Mean arterial blood pressure was calculated as 1/3 SAP + 2/3 DAP. Acute changes in CVP (CVP) during application of LBNP and LBPP were estimated from changes in Z₀ (Z₀) using the following equation
This equation was obtained from published work comparing directly measured CVP (central line) and Z₀ (17). These calculations represented the estimated changes in CVP from Base1 to Base2. This equation was applied only to data collected during Base1 and Base2, because the conditions were similar to those used to generate the equation. The equation was not used to estimate changes in CVP during saline infusion.

Statistics. Comparisons of variables with a given experimental group were determined by analysis of variance for repeated measures. Differences at specific time points were identified by paired t-test. Values are means ± SE, and changes represent differences between certain time point and Base1 values. The null hypothesis was rejected at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Systemic cardiovascular function during each experiment is shown in Tables 1 and 2. In the upright posture, resting values were similar for the control and LBPP experiments (Base1). Application of LBPP increased SV by 6.3 ± 3.0 ml/beat (P < 0.05) and reduced Z₀ by 0.7 ± 0.2  (P < 0.05) compared with rest without LBPP. The decrease in Z₀ predicted a 3.0 ± 1.4 mmHg increase in CVP with application of LBPP (P < 0.05). The combination of saline infusion and LBPP in the upright posture maintained the elevated SV and reduced Z₀ (P < 0.05) throughout the 2-h recovery period. These observations indicated a fluid shift toward the heart. In the supine posture, resting values were also similar for both experimental trials. Application of LBNP reduced SV by 16.4 ± 5.0 ml/beat (P < 0.05) and increased Z₀ by 1.0 ± 1.5  (P < 0.05) compared with rest without LBNP. The increase in Z₀ predicted a 3.8 ± 1.4 mmHg decrease in CVP with application of LBNP (P < 0.05). Saline infusion tended to return SV toward levels seen before LBNP, but Z₀ values remained lower throughout the 2-h recovery period (P < 0.05). In the supine posture, the combination of saline infusion and LBNP produced a persistent increase in Z₀, indicating a shift of fluid away from the heart.

Figure 2 shows the change in plasma albumin content after bolus saline infusion in the upright and supine postures. Plasma albumin content increased after bolus saline infusion in the upright posture and remained elevated throughout the 2-h recovery period (P < 0.05). Application of LBPP prevented the increase in plasma albumin content after saline infusion in the upright posture. In supine control, bolus saline infusion had no impact on plasma albumin content. In contrast, in the supine posture, application of 20 mmHg LBNP resulted in an increase in plasma albumin content after saline infusion (P < 0.05).

View larger version (26K): [in this window] [in a new window]   Fig. 2.   Changes in plasma albumin content after saline infusion in the upright posture with and without LBPP and in the supine posture with and without LBNP. Values are means ± SE. *P < 0.05 vs. Base1; P < 0.05 vs. control.

Figure 3 shows the ANP response to LBPP and LBNP and bolus saline infusion in the upright and supine postures. In the upright posture, LBPP and bolus saline infusion had little impact on the resting level of plasma ANP. However, application of LBPP prevented a time-dependent decline in ANP during the recovery period (P < 0.05). In the supine posture, plasma ANP decreased during application of LBNP (P < 0.05). In control experiments, ANP increased above Base1 in response to bolus saline infusion (P < 0.05) and fell over time back toward Base1 levels. In contrast, ANP was unchanged by bolus saline infusion in the supine posture during application of LBNP. The magnitude of change in plasma albumin content (mean values or areas under the curve) after saline infusion was not associated with the plasma level of ANP (P > 0.05, r² = 0.02 and 0.07 for mean values and areas under the curve, respectively).

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Atrial natriuretic peptide (ANP) response to application of LBPP and LBNP in upright and supine postures, respectively, and the subsequent response to saline infusion. Values are means ± SE. *P < 0.05 vs. Base1; P < 0.05 vs. control.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

It is generally accepted that the lymphatic system pumps fluid against a pressure gradient and that the normal outflow pressure that must be overcome by this system is CVP (2). In addition, as lymphatic outflow pressure increases, lymph flow decreases (2, 5, 19). In the present study, we hypothesized that lymphatic albumin transport was reduced in the supine posture primarily because of the rise in CVP and its impact on lymphatic outflow pressure. To test this hypothesis, we designed a series of experiments in which we mechanically manipulated CVP, via application of LBPP or LBNP, in an effort to reverse the impact of posture on lymphatic albumin transport. The significant finding of this study is that mechanical manipulation of CVP did reverse the postural impact on lymphatic albumin transport after saline loading. These findings support the hypothesis that the lymphatic return of albumin after bolus saline infusion is limited in the supine posture because of a high CVP. This conclusion is supported by observations of Pilon and Bittar (18) that thoracic duct lymph flow is reduced in anesthetized dogs during elevations in CVP caused by positive end-expiratory pressure ventilation.

Plasma albumin content increased within 30 min after saline infusion in the upright posture. The combination of saline infusion and LBPP resulted in a lower Z₀ and higher SV in the upright posture. These data are consistent with an increase in CVP and an increase in lymphatic outflow pressure. Under these conditions, plasma albumin content was unchanged after saline infusion. One interpretation of these data is that increased outflow pressure retarded the flow through the lymphatic system after saline loading and limited the return of albumin to the vascular compartment. In the supine posture, bolus saline infusion had little impact on plasma albumin content. Again, we assumed that this reflected a high lymphatic outflow pressure due to blood shifted toward the thorax. The combination of saline infusion and LBNP resulted in an increase in Z₀ and reduction in SV, indicating a reduction in CVP. In this condition, plasma albumin content increased, presumably because lymphatic return of albumin increased as lymphatic outflow pressure was reduced. Changes in plasma total protein content were consistent with those observed for plasma albumin content.

Our interpretation is based on the premise that lymphatic outflow pressure is positively related to the CVP. In the case of the large terminal lymph ducts, the pressure in the jugular or subclavian veins represents the outflow pressure. Under normal resting condition, lymphatic outflow pressure is higher in the supine posture and lower in the upright posture. Direct measurements of CVP were not performed in this study because of its invasive nature. However, earlier work from our laboratory indicates that the relationship between SV and Z₀ supports our conclusion that LBPP increased and LBNP decreased CVP (16) (Tables 1 and 2). Thus we are confident that we were able to manipulate CVP sufficiently with LBPP and LBNP. The fact that the change in plasma albumin content after saline loading could be manipulated by application of LBPP and LBNP indicates that the change in CVP was sufficient to impact lymphatic outflow pressure and lymphatic albumin transport.

We stimulated the lymph flow rate with bolus isotonic saline loading. Presumably, lower plasma colloid osmotic pressure (15) and higher plasma hydrostatic pressure caused by saline infusion altered Starling forces and increased capillary filtration into the interstitial compartment. Lymph flow rate must increase to remove the excess amount of fluid and relieve the high interstitial hydrostatic pressure to prevent edema. This assumption is supported by many animal studies where direct measurements of lymph flow were possible (1, 2, 5, 19).

Plasma ANP concentration can influence capillary permeability and, thereby, impact plasma protein content (12). Albumin transcapillary escape rate is dependent on posture and plasma ANP levels (10). Although lower ANP levels could act to reduce capillary permeability and contribute to a greater increase in plasma albumin (9, 12), it is important to note that the whole body albumin transcapillary escape rate is higher in the upright than in the supine posture, despite the lower plasma ANP levels (10). We believe that posture has the primary role in determining albumin transcapillary escape rate, with plasma ANP levels acting to provide some minor modulation of albumin transcapillary escape rate and, therefore, plasma protein content. In the upright posture, application of LBPP and bolus saline infusion had little impact on plasma ANP levels. Therefore, we do not consider the reduction in the increase of plasma albumin content after bolus saline infusion with application of LBPP associated with plasma ANP and its subsequent impact on capillary permeability. In the supine posture, ANP levels decreased with application of 20 mmHg LBNP and remained low during the recovery period after bolus saline infusion. However, we cannot attribute the reversal of the plasma albumin content responses with application of LBNP in the supine posture observed in the present study to the corresponding changes in plasma ANP levels. In support of this conclusion, we could not identify any significant relationship between plasma ANP levels and the increase in plasma albumin content after bolus saline infusion in the present study. Thus we conclude that the impact of LBPP and LBNP on changes in plasma albumin content is primarily converted by its mechanical impact on lymphatic outflow pressure and not by its impact on plasma ANP levels.

In summary, saline loading caused an increase in plasma albumin content presumably via increased lymphatic protein transport. This response is abolished in the supine posture and is attributed to the increase in CVP and lymphatic outflow pressure. The posture-dependent modulation of lymphatic protein transport was reversed by application of LBPP, to increase CVP in the upright posture, and LBNP, to reduce CVP in the supine posture. The interaction of posture and mechanical manipulation with LBPP or LBNP on plasma albumin content cannot be explained by changes in circulating levels of ANP. Thus we conclude that mechanical manipulation of the lymphatic outflow pressure in the upright or supine posture is associated with a significant modulation of lymphatic return of albumin to the vascular space after bolus saline infusion. When we apply these findings to our earlier work (14), we predict that the inability to increase plasma albumin content and PV after intense exercise in the supine posture is likely due to the mechanical influence of posture on CVP and the lymphatic outflow pressure.