Volume Status in Peritoneal Dialysis

Volume Status and Nutrition

There are important links between hydration and nutrition. Serum albumin is a powerful predictive factor for survival on dialysis, although the determinants of albumin concentration are complex. Protein-energy wasting is common in peritoneal dialysis (PD) (8) and is strongly linked with reduced survival. Serum albumin has been measured as a marker of nutrition and thus low serum albumin may be a marker of protein-energy wasting. Serum albumin is certainly affected by nutritional status, for example, correlation with body cell mass (BCM) determined by total body potassium measurement (9). However, serum albumin also reflects hydration. Cross-sectional data demonstrate an inverse relationship between serum albumin and extracellular water (ECW) (10). Thus, increased mortality associated with hypoalbuminemia could be mediated via the effects of associated volume excess. This relationship is further supported in an interventional study, where prospective reduction in ECW by lowering dry weight in patients on PD resulted in a rise in serum albumin and improved blood pressure control (11). Serum albumin is a negative acute-phase reactant, and one of the most important links between serum albumin and mortality is via the malnutrition–inflammation–atherosclerosis (MIA) syndrome (12). An intriguing hypothesis linking volume status and nutrition is that excess fluid may act as an inflammatory stimulus, lowering serum albumin and stimulating wasting and cardiovascular disease via the MIA syndrome (13,14). Finally, through fixing total body weight to a target weight by controlling dialytic fluid removal, undetected wasting will result in reciprocal increase in ECW volume.

Impact of Hydration on Assessment of Nutrition

Abnormal hydration has an important confounding effect on assessments of nutritional state and body composition in PD. The two-component model of body composition divides body mass into fat and fat-free mass (FFM) or lean tissue compartments. Total body water (TBW) is the dominant component of FFM and is comprised of ECW, which varies with hydration, and intracellular water (ICW), which reflects BCM. The two-compartment model assumes a constant 73% water content of FFM. Changes in hydration will alter TBW and thus measurements of FFM, leading to the erroneous appearance of changes in nutrition. A typical clinical scenario is the PD patient suffering wasting with loss of BCM not reflected by change in body weight or FFM measurement due to simultaneous ECW fluid retention.

Thirst and Volume Status in PD

Volume status depends on the balance of fluid intake and removal. While research has focused on ultrafiltration, peritoneal membrane function, and residual renal function, the importance and regulation of fluid intake in PD has received relatively little attention. We have assessed thirst in dialysis patients using a technique involving repeated measurements using visual analog scales on a handheld palmtop computer (EARS: electronic appetite rating system) to produce profiles of thirst over the day (15). Mean thirst ratings were greater in PD patients than in control subjects (15). Moreover, the normal variability of thirst levels seen in healthy controls and hemodialysis patients was diminished in PD patients, who had persistently high thirst scores throughout the day (15). The mechanisms involved in this finding are not clear. Several factors, including sodium balance, angiotensin levels, retained osmols, and volume status, have been proposed as important mediators of thirst in renal disease. This is an important area worthy of further research. In clinical practice, patients are often told that the continuous fluid removal by PD will allow more liberal fluid intake than hemodialysis. If fluid intake is unlimited in PD and governed by thirst, there is a high risk of developing volume overload. Patients on PD should be advised to monitor their fluid intake and have a maximal fluid intake allowance (and sodium restriction) appropriate to their residual urine output and ultrafiltration volumes.

How Well Does PD Perform in Regulating Volume Status?

Peritoneal dialysis should provide an advantage compared with hemodialysis in allowing fluid removal not limited by acute hemodynamic instability and producing a steady volume status. However, some studies have shown that PD patients are in a state of chronic fluid overload with adverse effects on blood pressure and the cardiovascular system (16,17). This may worsen with increasing time on PD, possibly due to loss of residual renal function and reduction in ultrafiltration due to membrane changes over time (18,19). Is this still the situation in current PD practice? A number of developments in recent years have improved fluid management in PD. These include increasing use of automated PD, icodextrin dialysis fluid, use of diuretics in patients with residual renal function to enhanced urine volume, and increasing awareness of the importance of optimal fluid management. In a comparison of PD patients and healthy controls, although ECW/ICW ratio was greater in PD patients (due primarily to ICW reduction, reflecting wasting), there was no difference between absolute values of ECW in the PD group compared with the control group (9). We have observed in the medium term a relationship between ECW reduction from use of icodextrin and improvement in blood pressure control in patients on automated PD (20). The longer-term benefits of icodextrin on fluid balance and body composition have also been demonstrated in randomized controlled studies (21-23). The ability to achieve excellent control of blood pressure and with beneficial cardiovascular effects in routine clinical practice through rigorous approach to sodium and fluid management has been demonstrated by the group in Izmir, Turkey (24,25).

Clinical Approach to Fluid Management: What is the Correct Target Weight?

The benefits of avoiding fluid overload are clear and we have the therapeutic tools to achieve adequate fluid removal, thus avoiding fluid excess in the majority of PD patients. The approach in hemodialysis has typically been to determine target weight by progressively removing fluid to the point of normotension or development of symptoms of hypovolemia. A danger of this approach in PD is that volume depletion and hypotension are important factors in loss of residual renal function (26). The importance of residual renal function to survival in PD is well established (27,28) and it is unknown in PD whether the adverse effects of loss of kidney function through overaggressive volume control would outweigh the benefits of avoiding fluid excess. Thus, the greatest challenge in fluid management may not be the ability to remove “enough” fluid but to remove the correct amount, avoiding both fluid excess and fluid depletion.

Assessing Volume Status and Body Composition: Bioelectrical Impedance Analysis (BIA)

Clinical assessment of hydration has significant limitations and the need for an accurate, precise, and objective measure of volume status in PD patients is clear. The technique of greatest promise in clinical practice, which has been widely studied in renal disease, is BIA. It is important to be aware when considering the use of BIA in clinical practice that there are a variety of underlying technologies utilized in commercially available analyzers, which have various and differing limitations in performance and interpretation (29). BIA is based on the principle that an AC electrical current passed through the body is conducted by body water but not by fat, which acts as an electrical insulator. Thus, impedance is inversely related to body water content. The original technique of single-frequency 50 kHz BIA allows estimation of TBW; however, it does not distinguish between the ECW (reflecting hydration) and ICW (reflecting BCM) components of TBW. Subsequent developments have led to the important ability to distinguish between ECW and ICW, thus providing information on both hydration and nutritional state. One method is analysis of the resistance (R) and reactance (Xc) components of single-frequency impedance by the RXc graph described by Piccoli et al. (30). The relationship between R and Xc is described by the phase angle, which is predictive of outcomes in PD (31). An alternative approach is measurement of impedance to different multiple-frequency currents (as in bio-impedance spectroscopy). Multiple frequency techniques utilize the property of differential penetration of the intracellular space according to the frequency of the current (32). Very low frequency currents pass through the ECW space alone, with increasing conduction also occurring through ICW with increasing current frequency.

Limitation of the accuracy of BIA for measurement of body composition arises from its dependence on a number of assumptions. BIA assumes that the body acts as a geometrically uniform, homogenous, electrical conductor. In reality this is not the case: impedance is more dependent on the limbs due to their smaller cross-sectional area and less sensitive to composition of the trunk (33). Limb abnormalities (e.g., those arising from arteriovenous fistulas) have a disproportionate effect on total body impedance (34).

Application of BIA to PD in Clinical Practice

Bioelectrical impedance analysis is easy to apply in the clinical setting, being portable, easy to use, acceptable to patients, and providing immediate results (35). How can BIA be utilized in clinical practice for management of the individual patient? Clinicians would like a tool that can accurately diagnose fluid (and nutritional) status of a patient from a single measurement. Important requirements for such an approach are accuracy of individual measurements and high sensitivity and specificity in distinguishing normal from abnormal states. There are significant discrepancies between body water volumes determined by BIA and those obtained using gold standard dilution techniques (36). These inaccuracies are even greater in renal disease (36), with risk of incorrect assessment of fluid status or low sensitivity for detecting abnormality. The range of normal variation between individuals raises the possibility that a change in hydration clinically significant to an individual may not be sufficient to result in BIA measurements falling outside normal population limits. In addition, co-morbid disease leading to altered fluid distribution or hemodynamics may result in the “ideal” state of hydration achieved by dialysis differing from normal body composition, as for example in cardiac failure, autonomic dysfunction, and altered fluid distribution with hypoalbuminemia.

One issue with BIA is how values of ECW should be expressed. ECW needs to be related to some marker of body size in individual patients. Commonly, the ratio ECW/TBW or ECW/ICW is employed. These ratios have the attraction of being produced by the BIA analyzer without the need for input of height or weight parameters. We have demonstrated the utility of this approach in a clinical setting where a “hydration score” was determined by multiple-frequency BIA, defined as standard deviations of ECW/TBW ratio away from age- and gender-matched controls (35). However, a limitation of this approach is that the denominator is affected by changes in ICW, or BCM. Thus, in patients with reduced ICW due to wasting, ECW/TBW and ECW/ICW ratios will be elevated due to this reason rather than reflecting fluid overload with ECW expansion. In comparing a group of PD patients with a matched control group undergoing measurement of body water volumes by dilution techniques, ECW did not differ between the PD and control groups. The ECW/ICW ratio was greater in PD due to a tendency to a lower ICW, but normalization of ECW to other indices of body size, such as height and surface area, demonstrated no differences in hydration between PD and control groups (37). A novel model using BIA that attempts to define hydration by distinguishing between normal hydration of lean tissue and excess or deficiency of fluid in ECW has recently been described by Chamney and colleagues (38).

Possibly the most promising potential application of BIA in clinical practice is longitudinal monitoring of complex changes in body composition. BIA has a very high precision if carefully performed and thus is ideal for reliable objective detection of changes in body composition from baseline in individual patients (39). BIA may allow earlier detection of complex and often simultaneous changes in hydration, BCM, and fat that may otherwise be difficult to identify or diagnose by clinical means alone (39).

Anthropometric Equations for TBW and Measurement of KT/V Urea

Estimation of TBW is required in PD for measurement of dialysis adequacy. Urea clearance is normalized to body water volume (V), which is also the volume of urea distribution in the body, in the expression Kt/V. In PD, rather than being directly measured by anthropometric equations, V is estimated, most commonly by the Watson equation in adults. V may also be simply estimated as a fixed proportion of body weight. These estimates have been validated in normal subjects and may be more prone to inaccuracy in PD patients with abnormal fluid status and body composition. Compared with the gold standard of isotope dilution, anthropometric equations for estimating TBW showed a greater range of discrepancy in PD patients than in healthy controls, and estimating TBW as 58% of body weight had a particularly high degree of variability from the reference values (40). Newer equations, including those derived from measurements in patients on dialysis, did not have any benefit compared with the Watson equation (40). The impact of the discrepancies between estimated and true values for V may lead to variations in estimates of Kt/V that are relatively significant in magnitude compared to the modest range of dialytic Kt/V values that may be achieved with current PD techniques (41). This needs to be considered when trying to achieve minimum targets for small solute clearances and could be a factor in the often poor relationship between Kt/V and uremic symptoms.

Conclusions

Management of volume status has emerged as one of the most critical aspects of PD therapy, through both its impact on cardiovascular disease and its interactions with nutrition. Although older literature has identified volume excess as being common in PD, with currently available techniques, achieving adequate fluid removal and euvolemia in most patients is a realistic goal in routine clinical practice. The major challenge to clinicians is to use PD techniques to their maximum potential by accurately diagnosing fluid status and identifying changes in body composition over time in patients. BIA has attracted much attention and appears to be the most promising potential tool to do this. It is uncertain whether BIA can determine fluid status or dry weight with sufficient accuracy in the absence of other aspects of clinical fluid state assessment. However, the high precision of the technique suggests that the most likely role of BIA in clinical practice is that of longitudinal monitoring of patients, allowing early detection of changes in hydration or nutritional state and thus prompt therapeutic intervention. Further evaluation of the feasibility and benefits of such an approach is an important area for future research.