Share this article on:

The Effects of Normal and Hypertonic Saline on Regional Blood Flow and Oxygen Delivery

Wan, Li MD*†‡; Bellomo, Rinaldo MD, FRACP, FJFICM*; May, Clive N. PhD†

Section Editor(s): Takala, Jukka

doi: 10.1213/01.ane.0000266438.90360.62
Critical Care and Trauma: Research Report

BACKGROUND: Boluses of crystalloids are frequently given to patients to optimize vital organ perfusion. However, their effect on regional blood flow and oxygen delivery in the normal mammalian circulation has not been studied. We compared the effects of normal or hypertonic (3%) saline or control on regional blood flow and oxygen delivery in normal animals.

METHODS: We conducted a randomized, controlled animal study in seven merino cross-ewes. We implanted chronic flow probes around aorta, coronary, renal, and mesenteric arteries. We randomized animals to three different interventions: observation (control), normal saline (1 L over 15 min), or 3% saline (300 mL over 15 min). We continuously measured central hemodynamics and organ blood flow for 210 min.

RESULTS: Both fluids increased heart rate, cardiac output, central venous pressure, peripheral conductance, coronary and mesenteric blood flow, and conductance in the first hour (P < 0.05). In the second and third hour, both maintained a greater cardiac output, total peripheral conductance, and mesenteric blood flow (P < 0.05) than control, but the difference decreased. In contrast, renal blood flow was unaffected and, because of hemodilution, renal oxygen delivery was decreased in the first hour (P < 0.05). Simultaneously, urine output and creatinine clearance increased (P < 0.05) in both groups. Finally, 3% saline significantly, but transiently, increased serum sodium and osmolarity.

CONCLUSIONS: Normal and hypertonic saline have similar systemic and regional hemodynamic effects. They also have no effect on renal blood flow and initially decrease renal oxygen delivery while increasing urine output.

IMPLICATIONS: Normal and hypertonic salines have similar hemodynamic effects, but no effect on renal blood flow.

From the *Department of Intensive Care, Austin Health, Heidelberg, Victoria, Australia; †Department of Cardiovascular Physiology, Howard Florey Institute; and ‡Department of Pharmacology, University of Melbourne, Parkville, Melbourne, Australia.

Accepted for publication March 9, 2007.

Supported by a grant from the Austin Hospital Intensive Care Trust Fund.

Address correspondence and reprint requests to Rinaldo Bellomo, MD, FRACP, FJFICM, Department of Intensive Care, Austin Health, Heidelberg, Victoria 3084, Australia. Address e-mail to

Normal saline is perhaps the most commonly used agent for fluid resuscitation in critically ill patients (1,2). Hypertonic saline is also increasingly used as a resuscitation fluid (3–6), especially in the setting of traumatic brain injury (7–9). Given the findings of the SAFE trial (10), there is an increased impetus toward seeking a greater understanding of how and when crystalloid solutions should be used.

The systemic hemodynamic effects of these two fluids seem similar and have been well described (11–13). However, there is little information on their comparative effects on regional blood flows and, in particular, on renal perfusion, oxygen delivery (DO2) and function. This is unfortunate, because the major aim of fluid resuscitation is to increase vital organ DO2 so as to prevent hypoxia and maintain function. Thus, an understanding of the effects of these fluids on such variables is potentially clinically relevant. Accordingly, both the magnitude and duration of the effect of a bolus of normal saline or 3% saline at clinically relevant doses and at equivalent amounts of sodium chloride on vital organ blood flow and, specifically, on renal blood flow (RBF) and renal function should be of interest to clinicians. Greater understanding of these effects in the normal mammalian circulation should provide the necessary physiological foundations to appreciate likely effects on regional circulations under a variety of clinical circumstances.

Similarly, these fluids are likely to affect serum osmolarity and natremia differently (11,12), but there is limited controlled information on the intensity and duration of such changes when equivalent amounts of sodium chloride are given.

Accordingly, we conducted a randomized, cross-over, large animal experiment to investigate the effects of a bolus infusion of 1 L of normal saline or 300 mL of 3% saline on changes in systemic hemodynamics, regional blood flows, regional DO2, serum sodium, and osmolarity.

Back to Top | Article Outline


Animal Preparation

The Animal Experimentation Ethics Committee of the Florey Institute approved the study. Seven merino ewes weighing between 34 and 52 kg were procured for chronic instrumentation. The animals underwent three separate operations before experimentation, as previously described (14,15).

The first surgical procedure consisted of oophorectomy and carotid loop formation. Two weeks later we implanted transit-time flow probes (3 and 20 mm, respectively) (Transonics Systems, Ithaca, NY) around the circumflex coronary artery and the ascending aorta, and after further 2 wk, we implanted transit-time flow probes (6 mm and 4 mm, respectively) around the superior mesenteric and left renal arteries (16,17). The transit-time flow probes on regional blood vessels were connected to flowmeters (Transonics Systems) 2 wk after surgery.

The day before experimentation, we implanted a carotid loop arterial Tygon catheter (inner diameter, 1.0 mm; outer diameter, 1.7 mm) (Cole-Parmers; Boronia, Australia) and two internal jugular venous polythene catheters (Critchley; Silverwater, Australia) (inner diameter, 1.2 mm; outer diameter, 1.7 mm) for the measurement of arterial and central venous pressures (CVP) and for fluid infusion. A urinary catheter was inserted for urine sampling and flow measurement. Analog signals [mean arterial blood pressure (MAP), CVP, cardiac output (CO) and regional blood flows] were collected using a personal computer data acquisition system using custom software written at the Howard Florey Institute. Data were collected at 100 Hz for 10 s at 1-min intervals throughout the experiment protocol. Conductance was calculated for each regional bed as regional blood flow/MAP. Total peripheral conductance was calculated as CO/MAP.

Back to Top | Article Outline

Experimental Protocol

Throughout each experiment all animals were awake, received no sedation or anesthesia, and were spontaneously breathing.

Before randomization, all animals were monitored for 30 min (control period) and then randomized into a control group, a normal saline group, or a 3% saline group. Each animal received either observation (control), a bolus of normal saline (1 L over 15 min), or a bolus of 3% saline (300 mL over 15 min), and monitoring was continued for a 3-h period. The first hour was defined as the Initial Period. The remaining 2-h period was defined as the Delayed Period. To avoid any confounding effect of other therapies, during the duration of the entire experiment, animals received no additional fluids.

Urine and blood samples were collected at 0, 15, 30, 60, 90, 120, 150, and 180 min. Blood gases were measured using a Radiometer (ABL system 625/620 Radiometer, Copenhagen, Denmark). Serum and urinary creatinine and electrolytes were measured using an automated multichannel biochemical analyzer (Hitachi 747, Roche Diagnostics, Sydney, NSW, Australia).

The fractional excretion of sodium (FENa) was calculated according to the following formula: FENa = (UNa × PCr)/(PNa × UCr) × 100 (UNa = urinary sodium concentration, PCr = plasma creatinine concentration, PNa = plasma sodium concentration, UCr = urinary creatinine concentration). In addition, creatinine clearances were calculated every 30 min during the experiment. Osmolarity was measured using the freezing point technique and is reported in mOsm/L.

DO2 was calculated by the following equation:

At the end of the experiment, all catheters were removed and the animals were allowed to recover for 5–7 days before being randomly crossed over to another arm of the study.

Back to Top | Article Outline

Statistical Analysis

Data are presented as means ± sd. Friedman's nonparametric analysis of variance was used to compare the overall differences among three groups over time. The Wilcoxon's signed rank test was performed to compare the study variables between any of the two different interventions. A P < 0.05 was considered statistically significant.

Back to Top | Article Outline


There were no differences among the three study groups in baseline measurements (Figs. 1–4).

Compared with control, during the first hour, resuscitation with both fluids increased heart rate, CO, CVP, and total peripheral conductance (TPC) (P < 0.05; Fig. 1). Both fluids also increased coronary blood flow (CBF) and conductance and mesenteric blood flow (MBF) and conductance (P < 0.05, Fig. 2). During the second and third hour after treatment, both fluids maintained a greater CO, CVP, TPC, and MBF (P < 0.05) than control, but the difference decreased and was small (Figs. 1 and 2). However, because of hemodilution, overall systemic DO2 did not change during the study period (Fig. 3). Coronary and mesenteric DO2 values were also unaffected (Fig. 3).

Despite the clear systemic and regional changes, neither treatment affected RBF or renal conductance (Fig. 4). Furthermore, although systemic DO2 remained constant, because of hemodilution, renal DO2 significantly decreased in the first hour (Fig. 4).

Although renal DO2 decreased, both normal saline and 3% saline significantly increased urine output (UO) and creatinine clearance (Fig. 5).

Infusion of 3% saline significantly increased serum sodium from 143.9 ± 3.8 to 150.0 ± 6.2 mmol/L over the first hour, followed by a rapid decrease to 145.9 ± 5.6 mmol/L 1 h later. In parallel, serum osmolarity also increased transiently (P < 0.05, Fig. 6). Normal saline infusion did not affect serum sodium level and osmolarity (P > 0.05).

Back to Top | Article Outline


We conducted a randomized, controlled, cross-over study of normal saline and hypertonic saline fluid resuscitation to assess and compare their effects on regional blood flow and DO2 in the normal mammalian circulation. We found that, as expected, both fluids increased CO, CVP, TPC, and CBF and MBF by approximately 50%. But we also found that these effects were mostly transient (first hour) and significantly decreased in magnitude over the following 2 h to approximately 10%. Because of hemodilution, both fluids actually failed to increase coronary or mesenteric DO2. More importantly, RBF was unaffected and, because of hemodilution, renal DO2 was decreased by approximately 20% in the first hour in both groups. Simultaneously, UO and creatinine clearance increased significantly during this period. Finally, we observed that 3% saline significantly increased serum sodium and osmolarity compared with control and saline also for the first hour.

The systemic changes reported in our study are expected and consistent with previous reports (18). However, for the first time we were able to assess the effects of these two fluids at equivalent amounts of sodium on CBF, MBF and RBF and DO2 in conscious animals. We found that both fluids increased flow to the coronary and mesenteric circulation and that the increase was associated with regional vasodilatation. Thus, the changes in these two regional beds mirrored systemic hemodynamics. However, because these fluids induced significant hemodilution, calculated DO2 to the heart and gut remained unchanged. The effects of saline on CBF have, to our knowledge, only been described in a model of severe sepsis (19). Similar to the effect in septic animals, in this study, saline administration increased CBF in normal sheep. This observation and the increase in coronary conductance and CO suggest that regional vasodilatation, possibly related to increased metabolic demand or to the effects of fluid itself on afterload through decreased blood viscosity (20), played a part in this change. The effects of hypertonic saline were essentially identical. No other information concerning the effect of fluids on coronary DO2 has been reported.

The effect of normal saline on MBF and conductance has not been assessed in controlled studies in conscious animals, except for a model of severe hyperdynamic sepsis (19). Consistent with our findings in normal animals, saline also increased MBF in septic animals. The effect of hypertonic saline on MBF was identical to that of normal saline. Both fluids, however, were associated with similar levels of hemodilution and no change in mesenteric DO2. These observations, in parallel with the observations made in septic animals for the coronary circulation, suggest that the physiological principles established by studying the normal mammalian circulation may apply to hemodynamic states altered by disease, such as hyperdynamic sepsis or during anesthesia. On the other hand, in a study of ventilated, anesthetized animals with hemorrhagic shock, normal and hypertonic saline have been shown to affect mesenteric DO2 differently, with a greater increment in association with hypertonic saline (12). Unfortunately, this study lacked a control population and only analyzed the effect of these fluids over a 60-min period—an important shortcoming given our findings on the transient effects of fluid administration. In a hypodynamic model of sepsis, other investigators (13) recently demonstrated similar transient effects of hypertonic saline and Ringer's solution on portal blood flow.

An important and clinically relevant finding of our study was that RBF and renal conductance did not change in either treatment group. In fact, because of this lack of response and the effect of hemodilution, renal DO2 decreased during the first hour. Consistent with these observations, in a study of hyperdynamic sepsis, we recently demonstrated that saline resuscitation did not affect RBF (19). In a model of hypodynamic sepsis, other investigators have recently reported only minor and transient changes in RBF with either Ringer's solution or hypertonic saline (13). No data on renal DO2 were included. In a model of hemorrhagic shock, Chiara et al. showed no difference in flow effects of these two fluids but a greater DO2 with hypertonic saline at 60 min. Unfortunately, no data were provided on a control group or on any independent effect beyond the initial first hour because of the addition of blood transfusion to the study protocol (12). Further, Prough et al. found similar and only transient effects of resuscitation with normal or hypertonic saline in a model of hemorrhagic shock (11). No DO2 data were reported. None of these investigations, however, reported functional changes.

We found that, during the decrease in DO2in the first hour, UO and creatinine clearance increased. This observation is similar to observations made in hyperdynamic sepsis (19); namely, that UO and creatinine clearance are dissociated from, and cannot be clinically used as, predictors of changes in RBF. This paper extends such observations to renal DO2. It highlights that, if the intent is to increase DO2 to the kidney with fluid resuscitation, the opposite might happen under circumstances of stable RBF (MAP within the autoregulatory range). This is likely explained by fluid-induced hemodilution and stable RBF when the kidney is operating within its autoregulatory state (20). Moreover, even in models of hemorrhagic shock or hypodynamic shock, the renal effects of fluid resuscitation with either normal or hypertonic saline are transient (11,13). Because the glomerular filtration rate is partly regulated by plasma oncotic pressure, the dilutional effect of fluids on protein concentration can be reasonably expected, in the setting of unchanged blood flow, to increase both UO and creatinine clearance, as previously noted (21,22) and as seen in this experiment. It is important to appreciate that this increased glomerular filtration rate will lead to increased solute presentation to the tubules (esp. the thick ascending loop of Henle) and increased renal medullary oxygen consumption. If this occurs while DO2 is decreased (as seen here with saline or 3% saline loading), the renal medulla is then exposed to an increased risk of ischemia (23). These observations on the limited and temporarily negative effect of crystalloid fluid loading on RBF and DO2 also stand in clear contrast with the marked increase seen when normal animals receive norepinephrine infusion at moderate doses (24)

Finally, we observed a differential effect of hypertonic saline on serum sodium and osmolarity, as expected. The magnitude and duration of these effects has been previously assessed by Cooper et al. using 7.5% saline but in the setting of multiple other interventions (6). We found these effects to be moderate in size and transient in duration. The maintenance of a high osmolarity state for longer than 1 h would likely require repeated administration of hypertonic saline.

Our study has both strengths and limitations. The possible strengths include the presence of a control group (a feature absent in all other relevant studies identified in the literature), the use of an identical dose of sodium chloride, the study of conscious animals, the assessment of renal function (also a feature absent in all other relevant studies), and the measurement of study variables over an extended period. The limitations of our study relate to its assessment of regional flows without simultaneous measurement of regional oxygen extraction. This makes the assessment of the effect of changes in DO2 difficult. However, we wished to study conscious animals, which makes it technically extraordinarily difficult to maintain cannulation of regional veins. Further, in our study, the chronic implantation of such cannulas in the sheep leads to vascular thrombosis. We could only calculate total regional values for conductance (the inverse of resistance). Changes in global organ conductance in any given direction may mask within-organ changes in the opposite direction such that, for example, in the kidney, changes in cortical and medullary conductance may differ. This aspect of our study needs to be considered when interpreting our findings. We did not induce hemorrhagic shock before fluid administration, making our observations less clinically relevant to one possible setting in which such fluids might be given. However, we consider that the first fundamental step in a planned research program dedicated to understanding the effect of such fluids should be to study them in the normal mammalian circulation. Our observation that in normal animals, the findings for normal saline mirror those reported in hyperdynamic sepsis, supports the relevance of such an initial approach, and the findings concerning the effect on the renal circulation are likely to apply to most, if not all, situations where RBF is operating within the limits of autoregulation (MAP between approximately 60 and 120 mm Hg). We now plan to study these fluids further in a variety of simulated disease states. We used 3% hypertonic saline instead of 7.5% hypertonic saline as others have used. However, on the basis of our observations, we consider it unlikely that, provided an equal dose of sodium chloride is given over the same period, the regional flow findings would differ. The increase in serum osmolarity, however, might be greater with 7.5% saline (6). Finally, we did not compare these two crystalloid solutions to a colloid solution. This would be clinically important, and accordingly, such studies are now under way in our laboratory.

In summary, both normal and hypertonic saline significantly, but only transiently, increase CO, CVP, TPC, CBF and MBF. Yet, they do not affect RBF, and they transiently significantly decrease renal DO2 while increasing tubular solute presentation. In addition, although 3% saline significantly increases serum sodium and osmolarity, this effect lasts only for approximately 1 h. Knowledge of the direction, magnitude and nature of these changes may assist clinicians in appreciating the likely physiological consequences of their interventions.

Back to Top | Article Outline


1. Practice parameters for hemodynamic support of sepsis in adult patients in sepsis. Task Force of the Am College of Critical Care Med, Society of Critical Care Med. Crit Care Med 1999;27: 639–60.
2. Hollenberg SM, Ahrens TS, Annane D, Astiz ME, Chalfin DB, Dasta JF, Heard SO, Martin C, Napolitano LM, Susla GM, Totaro R, Vincent JL, Zanotti-Cavazzoni S. Practice parameters for hemodynamic support of sepsis in adult patients: 2004 update. Crit Care Med 2004;32:1928–48.
3. Mattox KL, Maningas PA, Moore EE, Mateer JR, Marx JA, Aprahamian C, Burch JM, Pepe PE. Pre-hospital hypertonic saline/dextran infusion for post-traumatic hypotension. The U.S.A. Multicenter Trial. Ann Surg 1991;213:482–91.
4. Younes RN, Aun F, Accioly CQ, Casale LP, Szajnbok I, Birolini D. Hypertonic solutions in the treatment of hypovolemic shock: a prospective, randomized study in patients admitted to the emergency room. Surgery 1992;111:380–5.
5. Younes RN, Aun F, Ching CT, Goldenberg DC, Franco MH, Miura FK, Santos SS, Sequeiros IM, Rocha e Silva M, Fujimura I, Birolini D. Prognostic factors to predict outcome following the administration of hypertonic/hyperoncotic solution in hypovolemic patients. Shock 1997;7:79–83.
6. Cooper DJ, Myles PS, McDermott FT, Murray LJ, Laidlaw J, Cooper G, Tremayne AB, Bernard SS, Ponsford J. Prehospital hypertonic saline resuscitation of patients with hypotension and severe traumatic brain injury: a randomized controlled trial. JAMA 2004;291:1350–7.
7. Gunnar W, Jonasson O, Merlotti G, Stone J, Barrett J. Head injury and hemorrhagic shock: studies of the blood brain barrier and intracranial pressure after resuscitation with normal saline solution, 3% saline solution, and dextran-40. Surgery 1988; 103:398–407.
8. Gunnar W, Kane J, Barrett J. Cerebral blood flow following hypertonic saline resuscitation in an experimental model of hemorrhagic shock and head injury. Braz J Med Biol Res 1989;22:287–9.
9. Qureshi AI, Suarez JI, Bhardwaj A, Mirski M, Schnitzer MS, Hanley DF, Ulatowski JA. Use of hypertonic (3%) saline/acetate infusion in the treatment of cerebral edema: effect on intracranial pressure and lateral displacement of the brain. Crit Care Med 1998;26:440–6.
10. Finfer S, Bellomo R, Boyce N, French J, Myburgh J, Norton R. A comparison of albumin and saline for fluid resuscitation in the intensive care unit. N Engl J Med 2004;350:2247–56.
11. Prough DS, Whitley JM, Taylor CL, Deal DD, DeWitt DS. Small-volume resuscitation from hemorrhagic shock in dogs: effects on systemic hemodynamics and systemic blood flow. Crit Care Med 1991;19:364–72.
12. Chiara O, Pelosi P, Brazzi L, Bottino N, Taccone P, Cimbanassi S, Segala M, Gattinoni L, Scalea T. Resuscitation from hemorrhagic shock: experimental model comparing normal saline, dextran, and hypertonic saline solutions. Crit Care Med 2003; 31:1915–22.
13. Garrido AP, Cruz RJ Jr, Poli de Figueiredo LF, Rocha e Silva M. Small volume of hypertonic saline as the initial fluid replacement in experimental hypodynamic sepsis. Crit Care 2006; 10:R62.
14. Di Giantomasso D, May CN, Bellomo R. Norepinephrine and vital organ blood flow. Intensive Care Med 2002;28:1804–9.
15. Di Giantomasso D, Morimatsu H, May CN, Bellomo R. Increasing renal blood flow: low-dose dopamine or medium-dose norepinephrine. Chest 2004;125:2260–7.
16. Bednarik JA, May CN. Evaluation of a transit-time system for the chronic measurement of blood flow in conscious sheep. J Appl Physiol 1995;78:524–30.
17. Dean DA, Jia CX, Cabreriza SE, D'Alessandro DA, Dickstein ML, Sardo MJ, Chalik N, Spotnitz HM. Validation study of a new transit time ultrasonic flow probe for continuous great vessel measurements. ASAIO J 1996;42:M671–6.
18. Seki K, Aibiki M, Ogura S. 3.5% hypertonic saline produces sympathetic activation in hemorrhaged rabbits. J Auton Nerv Syst 1997;64:49–56.
19. Wan L, Bellomo R, May CN. The effect of normal saline resuscitation on vital organ blood flow in septic sheep. Intensive Care Med 2006;32:1238–42.
20. Kumar A, Anel R, Bunnell E, Habet K, Neumann A, Wolff D, Rosenson R, Cheang M, Parrillo JE. Effect of large volume infusion on left ventricular volumes, performance and contractility parameters in normal volunteers. Intensive Care Med 2004;30:1361–9.
21. Loutzenhiser R, Griffin K, Williamson G, Bidani A. Renal autoregulation: new perspectives regarding the protective and regulatory roles of the underlying mechanisms. Am J Physiol Regul Integr Comp Physiol 2006;290:R1153–67.
22. Fleming SJ, Dallemagne CR, Endre ZH, Yesberg NE, Cross RB. Acute lowering of plasma oncotic pressure increases filtration fraction and sodium excretion in conscious sheep. Ren Physiol Biochem 1992;15:334–40.
23. Marsh DJ, Sosnovtseva OV, Chon KH, Holstein-Rathlou NH. Nonlinear interactions in renal blood flow regulation. Am J Physiol Regul Integr Comp Physiol 2005;288:R1143–59.
24. Heyman SN, Rosen S, Brezis M. The renal medulla: life at the edge of anoxia. Blood Purif 1997;15:232–42.
© 2007 International Anesthesia Research Society