• Open Access

The Effects of Hypohydration on Central Venous Pressure and Splenic Volume in Adult Horses


  • This work was performed at New Bolton Center, University of Pennsylvania School of Veterinary Medicine. Portions of these data were presented at the ACVIM Forum (2009).

Corresponding author: Rose Nolen-Walston, Department of Clinical Studies, New Bolton Center, University of Pennsylvania, 382 West Street Road, Kennett Square, PA 19348; email: rnolenw@vet.upenn.edu.


Background: Central venous pressure (CVP) is used in many species to monitor right-sided intravascular volume status, especially in critical care medicine.

Hypothesis: That hypohydration in adult horses is associated with a proportional reduction in CVP.

Animals: Ten healthy adult horses from the university teaching herd.

Methods: In this experimental study, horses underwent central venous catheter placement and CVP readings were obtained by water manometry. The horses were then deprived of water and administered furosemide (1 mg/kg IV q6h) for up to 36 hours. Weight, CVP, vital signs, PCV, total protein (TP), and serum lactate were monitored at baseline and every 6 hours until a target of 5% decrease in body weight loss was achieved. The spleen volume was estimated sonographically at baseline and peak volume depletion. Linear regression analysis was used to assess the association of CVP and other clinical parameters with degree of body weight loss over time.

Results: There was a significant association between CVP and decline in body weight (P < .001), with a decrease in CVP of 2.2 cmH2O for every percentage point decrease in body weight. Other significant associations between volume depletion and parameters measured included increased TP (P= .007), increased serum lactate concentration (P= .048), and decreased splenic volume (P= .046). There was no significant association between CVP and vital signs or PCV.

Conclusions and Clinical Importance: These findings suggest that CVP monitoring might be a useful addition to the clinical evaluation of hydration status in adult horses.


central venous pressure


early goal-directed therapy


total protein

Central venous pressure (CVP) is used in many species to monitor right-sided intravascular volume status and optimize cardiac preload, especially in critical care medicine. Measured at either the cranial vena cava or the right atrium, the 2 determinants of CVP are the cardiac function, which is dependent on the Starling forces of the length-tension relation, and cardiac return of blood from the venous reservoir.1 As a result, changes in CVP can be seen associated with changes in contractility, venous tone, and circulating blood volume. In early goal-directed therapy (EGDT), achieving an adequate CVP is one of the earliest steps in resuscitation of human patients presenting with sepsis, and is associated with significantly decreased mortality.2 The effect of various manipulations on CVP in horses has been documented, including increase in CVP after IV hypertonic solution,3 bovine hemoglobin blood substitute,4 exercise,5 and high intraabdominal pressure,6 and decreased after administration of xylazine7,8 and acetylpromazine.9 A decrease in CVP is associated with reduction of intravascular volume after acute blood loss in adult horses.10 However, it is unclear whether compensatory mechanisms that might occur during long-term total body hypohydration could have a sparing effect on right-sided volume. To most closely simulate the losses associated with clinical hypohydration in the horse (such as colitis) versus acute hemorrhagic hypovolemia, we investigate a water deprivation and diuresis model causing mild-to-moderate hypohydration, and evaluate CVP as a function of total body water reduction measured in percentage loss of bodyweight. The splenic volume is highly correlated with blood volume and plasma volume in the horse.11 Because of the size of the splenic reservoir in horses,11–13 its mobilization during dehydration may be quantitatively more relevant in horses than in humans or other domestic species. The spleen volume was estimated sonographicallya before and after dehydration to quantify the effect of water loss on the mobilization of the splenic reservoir. Our hypothesis was that hypohydration in adult horses is associated with a proportional reduction in CVP and splenic volume.

Methods and Materials

Ten healthy adult horses from the university teaching herd were used. Horses were between the ages of 7 and 14 years (median 11 years) and weighed 458–677 kg (median 529 kg). Breeds included Thoroughbreds, Standardbreds, and Warmbloods. All procedures were approved by the Institutional Animal Care and Use Committee at the University of Pennsylvania.

A 70 cm, 16 g polyurethane single lumen catheterb was aseptically placed in the right jugular vein at the junction of the proximal third and middle third of the neck, with a peel-away introducer. The catheter was premeasured visually against the horse's neck to identify the approximate distance needed to reach 10 cm beyond the thoracic inlet. The catheter was then advanced the corresponding distance (∼60 cm), and oscillatory movements of the indicator ball in a water manometerc corresponding to the respiratory rate indicated that the catheter tip was intrathoracic.9,10 An indicator line was shaved over the point of the right shoulder (level with the right atrium)9 as a consistent reference point for the zero mark of the CVP readings. For evaluation of splenic volume, the left flank was closely clipped and sonograms were obtained with a wide bandwidth 2.5–7 MHz curvilinear-array transducer at a variable displayed depth to best fit the depth of the spleen in each horse. Splenic volume was then calculated with previously describeda modification of a standard ellipsoid formula.

The horses were housed in box stalls, were allowed access to ad lib hay and water, and were fed grain twice daily. Before water deprivation, all water and feed sources were removed and the horses were muzzled to ensure they did not consume the bedding. Furosemided (1 mg/kg IV) was administered every 6 hours; to determine percent hypohydration, the horses were weighed at baseline and then every 6 hours before administration of the next dose of furosemide. When horses had reached a 4% decrease in body weight, they were weighed every 2 hours to prevent overdehydration, with an initial goal of 6–7% decrease in weight. Physical parameters (heart rate [HR], respiratory rate, temperature, borborygmi, and manure production), PCV, total protein (TP), and serum lactate were recorded at baseline and then every 12 hours, and splenic volume was estimated sonographically at baseline and again at peak hypohydration. At the completion of the dehydration protocol the horses were allowed access to ad lib hay and were gradually reintroduced to water containing a balanced oral electrolyte solution over the subsequent 2–4 hours. Grain was withheld for an additional 6–12 hours. The amount of water consumed was recorded at 6-hour intervals. Physical parameters, CVP, and basic blood analysis (as described above) were measured every 6 hours until they returned to baseline values. At the completion of the study, when the horses were fully rehydrated, the central venous catheter was removed.

Statistical Analysis

Data were screened by a Wilcoxon signed-rank test and where appropriate (as indicated with Shapiro Wilk's tests) were log-transformed and analyzed by regression analysis to assess the association of CVP with degree of hypohydration over time. The change in the splenic volume was tested by a Wilcoxon signed-rank test, and the relationship between the change in spleen volume and changes in CVP, laboratory and clinical variables was studied by a Spearman's rank correlation coefficient.e The level of significance was set at <0.05.


CVP at baseline was 11.6 cmH2O (range 6–20.7). By 18–24 hours of water restriction and diuresis, the CVP nadir was −0.8 cmH2O (range −5.5 to 4.8). No complications were noted with the CVP catheters either during or after the study. PCV, TP, and blood lactate concentration all increased during the study period (Fig 1). Little change was seen in vital signs during dehydration (Table 1). A significant association was found between CVP and percent hypohydration (P < .001), with a decrease in CVP of 2.2 cmH2O for every percentage point increase in hypohydration (Fig 2). Other associations between hypohydration and parameters measured included increased TP (P= .007) and increased blood lactate concentration (P= .048). There was no significant association between CVP and HR, respiratory rate, rectal temperature, or PCV. Spleen volume decreased with dehydration in 5/6 horses, with a decrease of 4.8 L (range −6.8 to 0.9) at peak hypohydration (P= .046). The changes in spleen size were moderately or strongly correlated with changes in PCV (r= 0.75), changes in TP (r= 0.81), and changes in HR (r=−0.98). Poor correlation was found with other variables, including CVP (r= 0.14).

Figure 1.

 The effect of dehydration on selected variables over time. Each box-and-whisker pair shows the median, 25th and 75th quartiles, 95% confidence interval, and outliers, both at baseline (before) and after dehydration (after). **P < .001, *P < .05. y-Axis units are specified by variable on the x-axis.

Table 1.   Change in parameters between baseline and peak dehydration.
 Baseline Median (range)Peak Dehydration Median (range)ΔMedian
Bodyweight (kg)529 (458–677)506 (436–642)−23
Heart rate (beats/min)40 (32–44)44 (38–50)+4
Respiratory rate (breaths/min)12 (8–16)14 (8–20)+2
Rectal temperature (°C)37.6 (37.3–37.9)37.7 (37.3–37.8)+0.1
Spleen size (L)25.1 (17.7–32.6)21.2 (15.1–26.6)−3.9
Figure 2.

 Regression of mean central venous pressure (CVP) against hypohydration shows a significant negative linear relationship (P < .001), with a decrease in CVP of 2.2 cmH2O for every percentage point increase in hypohydration.

The first 3 horses underwent dehydration to a reduction of 6.5–7% of bodyweight, which took almost 24 hours. Two of the 3 developed severe neurologic deficits including manic behavior in one (running in circles, bucking, kicking out uncontrollably), and recumbency with paddling and nystagmus in the other. There were systemic signs of shock including hyperlactatemia of up to 7.8 mmol/L. Serum blood gas and electrolyte analysis obtained during the incidents did not demonstrate any remarkable abnormalities. Both horses were resuscitated with an IV bolus of crystalloid fluids and recovered quickly with no long-term ill effects. Because of these initial complications, the remaining 7 horses were dehydrated to maximum of 5.5% and for a maximum of 18 hours, and no horse underwent dehydration for the maximum time allowable in the study design (36 hours).


This study demonstrates that in adult horses, dehydration by water deprivation and diuresis results in a reduction of CVP of 2.2 cmH2O for every percentage point decrease in body weight. During dehydration, blood lactate and total solids also increased significantly. The measurement of CVP is technically simple, and can performed under anesthesia or in the standing horses,9,14 by a variety of techniques including commercial catheters of 50–70 cm in length,10 as was done in this study, Swan-Ganz catheters, or long polypropylene catheters fed through a standard 14 G jugular catheter, which allows CVP to be measured without placement of a dedicated central line.15 Values for CVP in euvolemic adult horses are reported to be 7–12 cmH2O, and under anesthesia are higher in lateral compared with dorsal decubitus.9,10,14 In humans, visual inspection of external jugular fill for subjective determination of “high” versus “low” pressure correlated quite well with the categorical variables of “high” versus “low” CVP.16 External jugular pressures are consistent with CVP in humans when the patient is in a supine, but not a lateral, recumbency.17 It is tempting to obtain a proxy of CVP with a standard jugular catheter in the horse, but unfortunately these measurements are not equivalent, with higher pressures consistently obtained from the external jugular and other smaller vessels.18 Whether CVP has sufficient clinical utility in the horse to justify the cost and risk of their placement is beyond the scope of this study, but in the authors' experience, they have been subjectively useful in determining ongoing volume resuscitation needs of severely compromised patients such as horses with severe colitis or postoperative large colon torsion, especially when complicated by polyuric or oliguric renal insufficiency.

EGDT of septic shock relies heavily on the use of CVP to determine volume status and guide fluid replacement19 After optimization of preload by a goal of CVP > 8–12mmHg (10.9–16.3 cmH2O), EGDT resulted in a 14% reduction in vasopressor use during the first 3 days of hospitalization when compared with standard care, and an absolute reduction of mortality of 16%.19 However, the use of CVP to determine volume replacement has come under considerable scrutiny in the human literature. A systematic review of 24 studies examining the interaction of change in CVP after fluid bolus on intravascular volume or cardiovascular function determined that there was a very poor relationship between CVP and blood volume. CVP failed to predict response to IV fluids.20 A review paper titled “Does central venous pressure predict fluid responsiveness? A systematic review of the literature and the tale of seven mares” cites a study in horses10 as being the only example of a study demonstrating a clear relationship between CVP and volume.20 More recent studies have indicated that arterial pulse pressure variation and stroke volume variation derived from pulse contour analysis21 as well as vena caval diameter in patients undergoing positive-pressure ventilation22 are superior to CVP for predicting fluid responsiveness. However, neither of these techniques are likely to be practical in the population of hospitalized adult equids.

Splenic size decreased with dehydration as anticipated. The sympathetic response associated with stress and hypohydration likely caused mobilization of the splenic reserve, which may help counteract the potential decrease in blood volume and contribute to the increase in the PCV. Noteworthy is the very high correlation (r=−0.98) between the change in splenic size and the change in HR, both signs of a sympathetic response. Some investigators have suggested that the smooth muscle of the splenic capsule might respond similarly to the capacitance vessels, potentially allowing assessment of intravascular fluid reservoir status by measuring gross splenic volume.23 A detailed analysis of the influence of the splenic reserve in hemodynamics during dehydration was beyond the scope of this project. However, the data evidence the importance of considering the mobilization of the large equine splenic reserve when assessing hemodynamic changes in horses.

An unexpected result of this study was the profound abnormalities shown by horses that underwent dehydration for 24 hours. Although it is typically reported that hypohydration of 5–7% is mild,24 we found that in 2/3 horses where hypohydration approached 6–7%, profound signs of hypovolemic shock occurred. Although the neurologic effects would be consistent with acid-base or electrolyte derangements, none were noted on analyses performed at the time of the clinical signs. This finding suggests that perhaps our traditional estimation of hypohydration might not be consistent with actual values. In contrast to other studies that find a strong association between PCV and HR with severity of disease and mortality in horses,25–28 these data did not support a significant increase in either parameter with hypohydration. Even if plasma volume in these horses was maintained by mobilization of interstitial fluid into the vascular space thus maintaining PCV at baseline levels, this effect would be expected also to be protective of both CVP and serum TP. Therefore, this would seem to indicate that adjustment of the splenic reserve of erythrocytes may have allowed for a relatively stable PCV over a range of hydration levels despite significant reduction the size of the spleen itself. Evaluation of systemically deranged cases such as horses with septic shock might allow assessment of splenic size in a disease model, where changes caused by endotoxemia and high levels of circulating catecholamines would probably affect PCV and HR more than the relatively benign influence of simple hypohydration.

In conclusion, we find that horses with slow, systemic dehydration, CVP decreases linearly and lactate and serum TP increase, corroborating the finding that in the horse, CVP is a useful predictor of right-sided pressure and thus intravascular volume. This is consistent with data showing similar findings in horses with acute hemorrhage,10 but in contrast to human studies, where systemic review failed to find a correlation between blood volume and CVP.20 Although total body water loss was the variable measured in this study, it is assumed that the change in CVP is because of hypovolemia from contraction of the plasma volume, though changes in venous tone or cardiac contractility cannot be ruled out as contributors to this effect. The use of CVP as an indicator of hypovolemia secondary to chronic dehydration seems valid in the adult horse; however, its utility in reducing mortality, either independently or bundled with other goal-directed parameters, deserves further investigation.


a Navas de Solis C, Foreman JH, Byron CR, Carpenter RE. Ultrasonographic assessment of the splenic size in horses. J Vet Intern Med 2009; 23:784 (abstract)

b PICC Peripherally Inserted Central Catheter Set, Arrow International, Reading, PA

c Central Venous Pressure Manometer, Smiths Medical ASD Inc, Dublin, OH

d Salix 5%, Intervet, Millsboro, DE

e Stata 10.0, StataCorp, College Station, TX


Supported by a grant from the Raymond Firestone Foundation (2007).

Conflict of interest: none.