Secondary Logo

Journal Logo


Central, Peripheral, and Other Blood Volume Changes During Hemodialysis

Prakash, Suma*; Reddan, Donal; Heidenheim, A. Paul*; Kianfar, Cynthia*; Lindsay, Robert M.*

Author Information
  • Free


Intradialytic hypotension and resulting symptoms such as cramps, fatigue, and so on, are major concerns that contribute to morbidity for patients on conventional hemodialysis. The mechanisms and factors involved in blood pressure changes and fluid shifts remain incompletely understood. Many investigators have examined the relationship between blood pressure and hydration status or volume overload. Katzarski et al.1 found that patients with hypertension before dialysis tended to have higher extracellular fluid volumes. Other investigators have found that ambulatory blood pressure (pressure measured over 24 hours) and postdialysis blood pressure are better related to total body water than is predialysis blood pressure. 2 The compartment from which fluid is removed may also affect blood pressure changes during dialysis. A measurement of central blood volume (CBV), which is defined as the volume of blood in the heart, lungs, and large blood vessels in the vicinity of these organs at any given time (a volume of 0.8–1.6 L), is thought to have some influence on the maintenance of blood pressure. CBV has been found to be a more sensitive indicator of morbid events on dialysis than cardiac output (CO) in hemodialysis patients. 3 Bioelectric impedance, which is usually used for nutritional analysis, can be used in the measurement of extracellular and intracellular volumes (ECF and ICF) and total body water (TBW) during dialysis. 4 The objectives of the current study were to (1) Examine the amount and direction of changes in blood volume and body fluid compartments after ultrafiltration during hemodialysis; and (2) Determine whether any correlations existed among mean arterial pressure (MAP), peripheral vascular resistance (PVR), CO, CBV, delta blood volume (ΔBV), TBW, and ICF + ECF before and after ultrafiltration (UF) during hemodialysis.


Study Design and Measurements

This was an observational study designed to examine the above relationships. Twenty patients, receiving three times weekly hemodialysis for 3 to 4 hours at each session, were studied. These patients all received regular hemodialysis treatments at the Adam Linton Dialysis Unit, London Health Sciences Centre, London, Ontario. The patients were chosen on the basis of having well functioning arteriovenous accesses (necessary for hemodynamic measurements) and known to have hemodynamic stability during dialysis procedures. Amputees and patients considered to be markedly volume overloaded on clinical grounds (elevated jugular venous pressure, gross peripheral edema, pulmonary edema, and so on) were excluded. No other selection or exclusion was used. At the beginning of dialysis, each patient was started on minimum ultrafiltration, which is defined as that amount of fluid removal that resulted in no change in blood volume as measured by online hematocrit monitoring. This amount was, on average, 0.1 L/hour. This strategy was chosen to allow the initial volume and hemodynamic readings to be made (phase I). Thereafter, the appropriate UF rate to approximately achieve target body weight over the next 2 hours was applied. Percentage change in blood volume over the HD session (ΔBV) was monitored using an on-line optical hematocrit monitor (Crit Line, In-Line Diagnostics, Utah). Once patients were at the target weight or at 2 hours into the session, they were again ultrafiltrated at minimum rates to obtain the second set of measurements (phase II). This procedure is shown schematically in Figure 1. Dialysis was then continued with the UF rate set to ensure target weight by termination time. The hemodynamic measurements carried out at phases I and II were CO, CBV, PVR, MAP, and ΔBV.

Figure 1
Figure 1:
Schematic of measurements in phase I and phase II.

MAP was calculated from blood pressure readings obtained before and during hemodialysis. CO was measured using an indicator dilution technique where the indicator was the ultrasound velocity of blood flowing through the dialysis tubing and the dilution agent was normal saline at 37°C (Transonics Monitor HD 01, Transonics Systems, Inc., Ithaca, NY). CBV was calculated from CO and transit time of the dilution agent, whereas PVR was calculated by dividing MAP by CO. CO and CBV measurements using these methods have been described in detail elsewhere. 3 ΔBV (percentage change in blood volume) between phase I and phase II was recorded.

The bioelectric impedance measurements carried out were ECF, ICF, and TBW. A bioelectric impedance analyzer (BIA Quantam Handheld Model, RJL Systems, MI) was used to obtain resistance and reactance values. These values, along with the height in centimeters, weight in kilograms, age, gender, and activity level of the individual were entered into a database program called Cypress 1.0, which calculated ECF, ICF, and TBW values in liters. The ECF:ICF ratio was calculated from these measurements. The measurement of TBW using isotopic dilution methods [e.g., deuterium or 18O- labeled water] represents the “gold standard,”5 but is not practical in the clinical setting. BIA measurement has been shown to be a simple, safe, and inexpensive method to measure the different water compartments of the human body and has been studied in hemodialysis patients. 6 The method is based upon a model describing the human body as a circuit consisting of capacitors and resistors lying in an electric field. 7 BIA, when compared with the criterion standard, has been shown to be a precise and reproducible technique for determining TBW and ECF. 5 The bioelectric impedance analysis measurements were done on the study patients in phase I and phase II, and were also done once on 10 healthy controls for comparison. These control subjects were not matched in any manner (age, weight, gender, and so on) to the study patients.

Data Analysis

All hemodynamic and bioelectric impedance readings were done twice at phase I and phase II for the dialysis patients, and the mean value was recorded. Two readings on the bioelectric impedance analyzer were done on the control group and the resulting mean for each individual was used in the statistical analyses.

Statistical Analysis

Data are presented as mean ± standard deviation (μ ± SD). Student’s t-test was used for between group comparison of means. Pearson correlation coefficients were used to analyze the bivariate relationships between the measures of interest. Multiple regression analysis was used to evaluate the determinants of the change in mean arterial pressure.


Phase I

Before UF, dialysis patients had a mean ECF volume of 19.5 ± 3.15 L, which was significantly higher than the control group with ECF of 18.2 ± 3.15 L (p < 0.001). ICF values were 21.05 ± 5.76 L for the dialysis group pre UF and 23.02 ± 8.17 L for the control group. TBW values for the dialysis group and the controls were 40.57 ± 8.4 L and 39.22 ± 11.80 L, respectively. The ECF:ICF ratio was significantly higher in patients before UF (0.96 ± 0.16) than controls (0.74 ± 0.124) (p < 0.001). This finding is demonstrated graphically in Figure 2. All hemodynamic measurements taken in phase I are listed in Table 1. Predialysis blood pressure (MAP) correlates with CBV (r = 0.46, p = 0.04). MAP also tends to correlate with TBW, and this result almost reaches statistical significance (r = 0.42, p = 0.06). These relationships are shown in line graphs in Figures 3 and 4, respectively.

Figure 2
Figure 2:
Extracellular fluid volume: intracellular fluid volume (ECF:ICF) ratios in dialysis patients and the control group.
Table 1
Table 1:
Hemodynamic Measurements Taken in Phase I
Figure 3
Figure 3:
Phase I (preultrafiltration): Mean arterial pressure (MAP)versus central blood volume (CBV).
Figure 4
Figure 4:
Phase I (preultrafiltration): Mean arterial pressure (MAP)versus total body water (TBW).

Phase II

After UF ECF, ICF, and TBW are 18.15 ± 3.13, 20.51 ± 5.54, and 38.67 ± 8.03, respectively. ECF:ICF ratio after UF is 0.917 ± 0.162, which is significantly higher than that of the control group (p = 0.003) but lower that the predialysis ratio (p < 0.001). Although all of the fluid compartments were smaller after UF, the greatest proportion of fluid was removed from the ECF. As in phase I, all hemodynamic measurements for phase II are listed in Table 1. MAP no longer correlates with CBV (r = 0.37, p = 0.108) or TBW (r = 0.19, p = 0.43); MAP does, however, correlate with PVR (r = 0.46, p = 0.04) and ΔBV correlates inversely with PVR (r = −0.50, p = 0.02). The correlation between MAP and PVR is shown graphically in Figure 5 and that between ΔBV and PVR in Figure 6. Despite a fall in blood volume of 7.11 ± 2.49% on average, CBV is maintained after UF, with an average change of only 0.067L. The average change in body weight between the two phases was 2.43 kg. Multiple regression analysis showed that the change in MAP was determined by changes in the following parameters: CBV, CO, PVR, ECF, ICF, TBW, weight, and ΔBV [R2 = 0.885 for all variables combined (p = 0.001)]. When the independent variables were analyzed, the B values for the factors seeming to be most responsible for change in MAP, as would be expected, were CO (B = 11.23 mm Hg per L/min, p = 0.001) and PVR (B = 2.369 mm Hg per mm Hg 41 L/min, p = 0.001).

Figure 5
Figure 5:
Phase II (postultrafiltration): Mean arterial pressure (MAP)versus peripheral vascular resistance (PVR).
Figure 6
Figure 6:
Phase II (postultrafiltration): Delta blood volume (ΔBV)versus peripheral vascular resistance (PVR).


The present study suggests that dialysis patients are ECF expanded both preultrafiltration and postultrafiltration compared with healthy people. This finding has also been reported by Spiegel and coworkers. 8 Before ultrafiltration, MAP correlated with hydration status. The tendency for ECF, ICF, and TBW to all correlate with MAP predialysis may be partially due to autocorrelation, as these parameters are all calculated from the same resistance and reactance values for a particular individual. Multiple regression analysis showed that mean arterial pressure was dependent upon several factors, the most significant of which were CO, PVR, and weight. The association of MAP with CBV showed statistical significance; that of MAP with TBW was almost significant. The R2 values for MAP and both CBV and TBW were approximately 0.25, suggesting that other factors influence MAP. Some of these factors could include autonomic neuropathy, such as occurs in patients with diabetes and those taking blood pressure medications. When all of the other independent factors, including PVR, CO, and weight were taken into account, as in the regression analysis, much more of the variation in blood pressure could be explained. Although many factors contributing to hypertension in dialysis patients have been studied, 1,2 conditions such as intradialytic hypotension remain incompletely explained and thus warrant further investigation.

In this study, the change in TBW between phases I and II was on average 1.89 L. Zaluska et al.9 showed that total body water measured using bioelectric impedance analysis was less accurate and underestimates actual total body water decrease compared with the amount of fluid removed as read on the dialysis machine. Hence, our data showing a weight reduction of 2.43 kg corresponding to a volume change of 2.43 L reflects this. Full body bioimpedance is believed to underestimate total body water, because it is not as accurate as segmental bioimpedance in detecting fluid changes centrally, as in the trunk. Utilization of segmental bioimpedance in future studies may provide a more accurate assessment of fluid shifts during dialysis. During ultrafiltration, most fluid is removed from the ECF, as expected, given that the ECF volume includes the intravascular volume from which fluid is directly removed.

Measurement of hemodynamic parameters and fluid compartment volumes on dialysis are not in widespread use and, as such, there are no gold standards for comparison. This study shows that when blood volume falls, PVR increases to maintain MAP and CBV, which likely prevents intradialytic hypotension. Failure of this mechanism will result in hypotension. The ability to noninvasively monitor for this complication could reduce intradialytic morbidity. With the ongoing development of modern techniques, it will be possible to more accurately study the hemodynamics and pathophysiology of blood pressure during dialysis.


1. Katzarski KS, Charra B, Luik AJ, et al: Fluid state and blood pressure control in patients treated with long and short hemodialysis. Nephrol Dial Transplant 14: 369–375, 1999.
2. Lins RL, Elseviers M, Rogiers P, Van et al: Importance of volume factors in dialysis related hypertension. Clin Nephrol 48: 29–33, 1997.
3. Krivitski NM, Depner TA: Cardiac output and central blood volume during hemodialysis: methodology. Adv Ren Replace Ther 6: 225–232, 1999.
4. Jabara AE, Mehta RL: Determination of fluid shifts during chronic hemodialysis using bioimpedance spectroscopy and an in-line hematocrit monitor. ASAIO J 41: 682–687, 1995.
5. Van der Ham EC, Kooman JP, Christiaans MH, et al: Body composition in renal transplant patients: Bioimpedance analysis compared to isotope dilution, dual x-ray absorptiometry and anthropometry. J Am Soc Nephrol 10: 1067–1079, 1999.
6. Ho TL, Kushner RF, Schoeller DA, Gudivaka R, Spiegal DM: Bioimpedance analysis of total body water in hemodialysis patients. Kidney Int 46: 1438–1442, 1994.
7. Lukarski HC: Biological indices considered in the derivation of the bioelectrical impedance analysis. Am J Clin Nutr 64 (Suppl): 397–404, 1996.
8. Spiegel DM, Bashir K, Fisch B: Bioimpedance resistance ratios for the evaluation of dry weight in hemodialysis. Clin Nephrol 53: 108–114, 2000.
9. Zaluska WT, Schneditz D, Kaufman AM, Morris AT, Levin NW: Relative underestimation of fluid removal during hemodialysis hypotension measured by whole body bioimpedance. ASAIO J 44: 823–827, 1998.
Copyright © 2002 by the American Society for Artificial Internal Organs