Segmental bioelectrical impedance in patients with chronic renal failure

Clinical Nutrition (1996) 15:275-279 © Pearson Professional Ltd 1996

Segmental bioelectrical impedance in patients with chronic renal failure G. WOODROW, B. OLDROYD*, J. H. TURNEY, and M. A. SMITH* Renal Unit, Leeds General Infirmary and *Centre for Bone and Body Composition Research, University of Leeds, Leeds, UK (Reprint requests and correspondence to G. W., Renal Unit, Leeds General Infirmary, Great George Street, Leeds LS1 3EX, UK) ABSTRACT--We studied changes in hydration by whole body and segmental (arm, leg and trunk)

bioelectrical impedance analysis (BIA) in patients with chronic renal failure (CRF) undergoing haemodialysis and continuous ambulatory peritoneal dialysis (CAPD). Mean (SD) fluid removal by haemodialysis of 1.38 (0.81) kg was overestimated by whole body BIA at 1.83 (1.13)I, P< 0.005. Peritoneal fluid drained from the CAPD patients of 1.88 (0.36)kg was underestimated by whole body BIA at 0.59 (0.35)I, P< 0.0001. Resistance and reactance significantly increased for the whole body and all segments (except trunk reactance) after haemodialysis. Drainage of CAPD fluid resulted in smaller increases in trunk resistance and whole body resistance. The increase in trunk resistance was less in CAPD than haemodialysis patients, even though the volume of fluid drained from the peritoneum in CAPD patients exceeded that removed from the whole body during haemodialysis. We conclude that whole body BIA does not estimate changes in body fluid with sufficient accuracy to be of use in clinical practice. Segmental impedance may be a potentially useful method for investigation of regional changes in body fluid, though is insensitive to changes within the peritoneal cavity.

contribution, despite representing a large fraction of the total body mass (in this study mean trunk resistance was less than 14% of whole body resistance). Thus changes or abnormalities in the composition in the limbs may result in marked changes in total body impedance, whereas the technique may be insensitive to changes in the trunk. This has resulted in studies investigating the use of BIA measurement of individual body segments to try to improve the sensitivity of the technique and to allow the study of regional body composition and water content (3-6). There has been great interest in the potential application of BIA in the study and clinical assessment of patients with chronic renal failure (CRF). Disordered fluid balance in these patients is common and may have important consequences such as the development of hypertension and cardiovascular disease in patients with fluid overload. Clinical assessment of fluid balance is perceived as being an inexact procedure, with significant abnormalities often being difficult to detect by clinical examination. The effect of changes in hydration may hinder nutritional assessment by masking the effects of changes in body protein and fat stores, e.g. body weight being maintained by increasing fluid retention in the face of developing malnutrition. In patients on haemodialysis, overestimation of hydration by clinical assessment may result in acute hypotension due to excessive fluid removal by dialysis. BIA is reliant on the distribution of body water between intra- and extracellular compartments (7) and it has been suggested that abnormalities of water distribution in patients with CRF may result in decreased accuracy of BIA in these patients (8). The reactance component (Xc) of impedance may provide additional information about the

Introduction Bioelectrical impedance analysis (BIA) allows assessment of body composition by measurement of the impedance of the human body to the passage of a small AC electric current (1). It allows measurement of total body water, and fat and fat-free compartments. The development of commercially available BIA systems over recent years has stimulated great interest in its use in a wide variety of situations. The technique has the advantage of allowing quick, safe, non-invasive measurement of body composition and is thus suitable for routine clinical use. Its portable nature makes it suitable for use in critically ill patients not suitable for transfer to specialist departments or centres for measurement. The underlying theory of BIA applies the principles of electrical conduction to the passage of an AC current through the human body between electrodes at the wrists and ankles (2). It assumes that the human body consists of a uniform cylindrical conductor whose length is proportional to body height. Estimates of body water or soft tissue composition are then calculated by regression equations derived from the comparison of BIA with reference techniques (such as isotope dilution for the measurement of body water and densitometry for estimation" of soft-tissue composition) in groups of healthy subjects. In reality, the human body is more closely represented by a combination of 5 cylinders (trunk and 4 limbs) and even then, the head and extremities of the limbs are excluded from the path of the current. Impedance or resistance is inversely related to the cross-sectional area of a conductor, and thus the arm (in particular) and legs account for the majority of total body impedance, with the trunk making only a small 275

276

SEGMENTAL BIOELECTRICAL IMPEDANCE

distribution of water in the body, being inversely related to the extracellular fluid compartment (9). Reactance has been shown to be related to extracellular fluid in patients with malnutrition (10) and critically ill patients on an intensive care unit (11). Infusion of saline decreases reactance (12), whereas a reduction in extracellular fluid with diuretic therapy in patients with liver disease resulted in an increase in reactance (13). The ratio of reactance to resistance (Xc/R) has also been shown to distinguish between healthy subjects and critically ill patients (11). Information regarding hydration can also be obtained from the bivariate analysis of resistance and reactance by the 'RXc' graph (14). The aims of this study were to investigate the ability of segmental BIA to detect changes in body hydration in patients with CRF undergoing haemodialysis and continuous ambulatory peritoneal dialysis (CAPD).

Materials and methods

Subject Studies were performed on patients with CRF receiving renal replacement therapy with haemodialysis or CAPD. The haemodialysis patients (21 subjects, 9 male and 12 female) received dialysis in 4-h sessions 2 or 3 times per week using bicarbonate-buffered dialysate and cuprophane dialysers. CAPD was by standard methods, most patients performing 4 exchanges per day (22 subjects, 10 male and 12 female).

Methods Whole body and segmental impedance were performed before and immediately (within 30 rain) after a haemodialysis treatment in the haemodialysis patients and before and after drainage of peritoneal dialysate in the CAPD patients. Fluid loss was assessed by weighing patients before and after dialysis or drainage of peritoneal fluid (and in the CAPD patients this was checked by also measuring the volume of fluid drained). Electrodes were removed after the first set of readings after marking their positions on the skin, and a new set placed for the second set of readings. Impedance was measured using the RJL 101A system. This employed a standard tetrapolar technique with a current of 800 gA at 50 kHz. Electrode positions were as recommended by the manufacturers, with the sensing electrodes placed on the dorsum of the wrist on a line between the radial and ulnar styloid processes and on the dorsum of the ipsilateral foot on a line between the medial and lateral malleoli. Source electrodes were placed overlying head of the third metacarpal on the dorsum of the wrist and the third metatarsal on the dorsum of the foot on the same side as the sensing electrodes. Measurements were made on the left side of the body except for patients with left-sided arteriovenous haemodialysis fistulae in which case the right side was used. Measurements were performed with the patients supine with their arms and legs abducted so that their thighs were not in contact with each

other and their arms were not touching the sides of their bodies. Total body water (TBW) was determined according to the standard manufacturer's supplied equations for the system (the equations for the RJL system being derived from the study of Kushner and Schoeller (15)): Males

TBW = (0.397*height2/R) + (0.143*weight) + 8.40 Females TBW = (0.382*height2/R) + (0.105*weight) + 8.31 (where TBW is in kg, height is in cm, R is resistance in ohms and weight is in kg)

Segmental bioelectrical impedance was measured using the method of Chumlea et al (3, 4). Measurements were performed with the same RJL 101A system used for whole body impedance with measurements made at the same time and with the patients in the same positions as for whole body impedance. Measurements of the leg, arm and trunk segments were performed on the same side as for whole body measurements. For measurement of the leg segment, one pair of electrodes was sited at the ankle, at the same position as for whole body impedance. The other pair was placed in the midline on the anterior surface of the thigh with the sensing electrode at the level of the gluteal crease and the source electrode 5-cm proximal. For the arm, one pair of electrodes was placed at the wrist, at the same position as for whole body impedance measurements. The second sensing electrode was placed on the anterior surface of the shoulder at the midpoint of a line between the acromial process and the axilla, with the source electrode 5-cm medial to it. For measurement of trunk impedance, sensing electrodes were placed in the midline of the anterior surface of the thigh at the level of the gluteal crease and at the suprasternal notch. Source electrodes were placed 5-cm distal to the thigh electrode and 5-cm superior to the suprasternal sensing electrode. Measurements prior to haemodialysis and drainage of peritoneal dialysate, as well as precision measurements, were performed as for whole body impedance measurements with the RJL 101A system described above. Precision of measurement was assessed by performing two sets of readings on subjects not undergoing dialysis, with the electrode positions being marked after the first set of readings, then the electrodes being removed and a second set being placed at the marked positions. The error standard deviation was derived from the equation: s = ~] [Y~(xi-Yi)z/2n] (where s is the error standard deviation and xi and Yi are paired measurements for i = 1 to n). From this, the 95% limit of the error, i.e. the range of values between which the difference between two measurements will lie with a probability of 0.95, was derived from the term 1.96 ~/(2s2). The coefficient of variation was derived from the error standard deviation by the equation: C.V. = 100 s/mean All mean values are quoted as mean (SD). The study was

C L N C A L NUTRITION

approved by the hospital ethics committee and all subjects gave informed consent.

277

Table 3 Changes in whole body and segmental resistance (R) and reactance (Xc) values (ohms) after drainage of peritoneal dialysate in CAPD patients (volume drained 1.88 SD 0.361). Comparisons by paired t-test Before drainage

Results

After drainage

Ratio post-/pre-

Precision values for whole body and segmental measurements resistance and reactance are shown in Table 1. The mean weight change after a haemodialysis session in the haemodialysis patients was 1.38 (0.81)kg. The mean fluid loss estimated by whole body BIA was significantly greater at 1.83 (1.13) 1, P < 0.005. The correlation between actual measured weight loss and fluid loss estimated by BIA was r = 0.86, P < 0.0001. Whole body and segmental resistance and reactance before and after drainage of dialysate are shown in Table 2. Highly significant changes were observed in resistance and reactance of the whole body and whole body segments. The mean weight change after drainage of peritoneal fluid in the CAPD patients was 1.88 (0.36) kg. The mean fluid loss estimated by BIA was significantly less at 0.59 (0.35) 1, P < 0.0001. The correlation between actual measured weight loss and fluid loss estimated by BIA was r = 0.45, P < 0.05. Whole body and segmental resistance and reactance before and after drainage of dialysate are shown in Table 3. Trunk resistance significantly increased by an average of 7.6% after drainage of peritoneal dialysate. Significant increases also occurred in resistance of the whole body and (curiously) the arm, but these were on average less than 2%. Despite the fluid change being greater

Whole body R

Table 1 Precision measurements for resistance (R) and reactance (Xc) measurements (ohms) measured by BIA (electrodes replaced after marking positions)

Table 4 Correlations of changes in resistance and reactance with volume of fluid removed by haemodialysis or drained in CAPD patients (Pearson correlation coefficient)

Error SD

95% limits

C.V. (%)

Xc

564.5 (82.8) 56.7 (13.3)

575.2 (82.4) P < 0.0001 57.7 (12.2) ns

1.02 1.02

Arm

R Xc

289.5 (50.8) 33.9 (12.5)

294.7 (51.9) P < 0.01 32.8 (8.5) ns

1.02 0.97

Leg

R Xc

252.8 (40.7) 32.2 (11.4)

251.6 (38.5) ns 31.6 (10.3) ns

1.00 0.98

Trunk

R Xc

72.4 (12.8) 17.9 (14.2)

77.9 (12.5) P < 0.0001 17.8 (9.0) ns

1.08 1.00

than in the haemodialysis patients, the magnitude of changes in resistance and reactance were far smaller. Correlations between change in weight and changes in whole body and segmental resistance and reactance are shown in Table 4. In the haemodialysis patients, changes in all resistance and reactance measurements (except trunk reactance) were significantly correlated with weight change. In the CAPD patients, there were no significant correlations between changes in any of the resistance and reactance measurements and the change in body weight. Changes in the Xc/R ratio are shown in Table 5. Removal of fluid by haemodialysis resulted in significant increases in the Xc/R ratios for the whole body, arms and legs, but not for the trunk. The Xc/R ratios for the whole body and

Haemodialysis

CAPD

Whole body

R Xc

1.69 0.65

4.67 1,81

0.34 1.28

Whole body

R Xc

r = 0.74 P = 0.0001 r = 0.57 P < 0.01

r = 0.23 ns r = -0.14 ns

Arm

R Xc

1.63 0.82

4.52 2.28

0.68 3.18

Arm

R Xc

r = 0.69 P = 0.0005 r = 0.47 P < 0.05

r = 0.11 ns r = -0.19 ns

Leg

R Xc

1.71 0.73

4.73 2,03

0.75 2.61

Leg

R Xc

r = 0.59 P < 0.005 r = 0.52 P < 0.05

r = 0.00 ns r = ~).04 ns

Trunk

R Xc

1.03 0.58

2,85 1,55

1.61 4.79

Trunk

R Xc

r = 0.71 P < 0.0005 r = 0.37 ns

r = 0.18 ns r = ~).13 ns

Table 2 Changes in whole body and segmental resistance (R) and reactance (Xc) values (ohms) after haemodialysis (weight loss on dialysis 1.38 SD 0.81 1). Comparisons by paired t-test Before dialysis Whole body R Xc

556.0 (102.2) 50.2 (11.5)

After dialysis 606.1 (106.4) P < 0.0001 59.7 (13.1) P < 0.0001

Table 5 Effects of drainage of peritoneal dialysate and haemodialysis on the Xc/R ratio (comparisons by the paired t-test)

Ratio post-/pre1.09 1.19

Arm

R Xc

276.0 (56.9) 27.4 (6.4)

303.4 (60.8) P < 0.0001 32.1 (7.4) P < 0.0001

1.10 1.17

Leg

R Xc

248.5 (47.2) 26.1 (9.5)

267.8 (48.3) P < 0.0001 29.8 (10.3) P < 0.0001

1.08 1.14

Trunk

R Xc

70.7 (16.2) 12.6 (6.3)

80.2 (17.8) P < 0.0001 14.1 (6.0) P < 0.005

1.14 1.12

Haemodialysis Whole body Arm

Leg Trunk CAPD

W h o l e body Arm

Leg Trunk

Before dialysis or drainage of dialysate

After dialysis or drainage of dialysate

0.091 0.101 0.104 0.175

(0.016) (0.021) (0.031) (0.047)

0.099 (0.017) P < 0.0001 0.107(0.021) P < 0.005 0.110 (0.031) P < 0.0005 0.175 (0.046) ns

0.101 0.119 0.128 0.239

(0.025) (0.047) (0.044) (0.133)

0.101 0.114 0.t26 0.229

(0.024) (0.036) (0.040) (0.116)

ns ns ns ns

278

SEGMENTAL BIOELECTRICAL IMPEDANCE

segments were unaffected by the drainage of CAPD dialysate.

Discussion

Assessment of hydration and nutrition are of great clinical importance in the management of CRF. It is thus not surprising that numerous studies have been performed using BIA to assess body water content and body composition in CRF patients. A common problem addressed by these studies is the ability of BIA to measure changes in hydration induced by dialysis. This could be of particular clinical value in situations where it is not possible to assess fluid change easily by weighing, e.g. in ventilated patients. These studies have shown conflicting results, with some suggesting that fluid changes measured by whole body BIA were in close agreement with actual changes (16-18), but the majority showing significant discrepancies (8, 19-21) with a tendency to overestimate the volume change. Studies of the measurement of changes in peritoneal fluid by whole body BIA have uniformly shown that BIA is insensitive to the presence of fluid in the peritoneal cavity (5, 8, 19) and this has been attributed to the relatively small contribution of the trunk to whole body impedance. Because of the disproportionate contribution of the limbs to whole body impedance, some studies have investigated the use of segmental BIA. Impedance of the arm may allow almost as good estimates of body composition as measurement of whole body impedance (4, 22). This is perhaps not surprising in view of the important contribution of the arm to whole body impedance and probably does not offer any major benefits for clinical practice. The main potential application of segmental BIA is to measure the composition and changes in composition in different body regions. We have demonstrated significant changes in whole body and segmental resistance and reactance after fluid removal by haemodialysis. The changes from different segments were proportionally similar, suggesting, as would be expected, that fluid removal is proportionally similar from all body regions. This is in agreement with previous work (23). It should be noted that the limbs account for the majority of the change in whole body impedance. Although estimates by BIA of fluid removal by haemodialysis were strongly correlated with actual weight change, BIA significantly overestimated the actual changes that occured. Although marked changes in biochemistry also occur during a haemodialysis treatment, observations in patients undergoing haemodialysis without fluid removal suggest that the effect of this on impedance is unimportant, with changes in fluid being the major factor causing impedance changes during haemodialysis (24). In contrast, whole body BIA showed far smaller changes after drainage of CAPD fluid, despite the mean fluid loss being greater than in the haemodialysis patients. Although segmental impedance of the trunk showed a proportionally greater change, the actual magnitude was still relatively small. The minor but statistically significant difference in

arm impedance after drainage of CAPD fluid is not easily explained as no fluid changes affecting the limbs are likely to have occurred during the time between the initial and repeat measurements of impedance. Correlation of estimates by BIA of the volume of fluid loss with actual changes was much weaker than for haemodialysis and BIA greatly underestimated the actual volume of fluid drained. Changes in resistance and reactance were significantly correlated with fluid removal during haemodialysis but not drainage of peritoneal dialysate. The changes in resistance and reactance in the CAPD patients were relatively small compared with the precision values for repeated measurement, especially as the technique is likely to be less precise with measurements repeated after much longer periods of time without marking of electrode positions (which is more likely in clinical use for a measure of peritoneal fluid, e.g. in patients with ascites than measurement of changes over a period of hours). Xc/R ratios increased after haemodialysis for the whole body and limbs but not the trunk. As reactance may be inversely related to extracellular fluid, this indicates that in the limbs proportionally more fluid is removed from the extracellular than intracellular space (in keeping with the clinical observation that fluid retention may manifest as oedema), but in the trunk the ratio is unchanged. The lack of change in Xc/R ratios after drainage of CAPD dialysate probably cannot be interpreted any further than indicating that both Xc and R change by relatively small amounts and so it is not surprising that there is no change in the ratio. The changes of fluid resulting from haemodialysis are relatively small but clinically important and for BIA to be of value in the clinical setting it should be able to provide a more accurate assessment of hydration than is possible by clinical examination. Despite a good correlation, BIA significantly overestimates fluid loss by haemodialysis. It is particularly poor at detecting changes in peritoneal fluid. Although after drainage of CAPD fluid, the most marked changes of BIA were in trunk resistance, the magnitude of change was small, especially allowing for the precision of measurement. Although the insensitivity of whole body BIA to peritoneal fluid has been explained as being due to the relatively small contribution of the trunk to whole body impedance, we have shown that greater changes in trunk impedance occurred after haemodialysis, when the change in fluid in the whole body was less than that related to the peritoneal cavity alone in the CAPD patients. This suggests that the problem is not so much the small impedance of the whole trunk, but a particular problem with detection of fluid in the peritoneal cavity. BIA theory assumes that the body (or segments) are of uniform resistivity. However, the anatomical composition of the body means that this is unlikely to be true and it is possible that during the measurement of impedance there is little passage of the current through the peritoneal cavity, due to electrical insulation by surrounding tissues. It is possible that the multiple frequency BIA (MFBIA) may be more sensitive to the detection of intraperitoneal fluid than the single frequency method used in this study as

CLINICALNUTRITION 279

one study found MFBIA to be accurate in the measurement of fluid injected into the peritoneal cavity of rats (25). However, this was not confirmed by a subsequent publication (26). We believe that, despite great enthusiasm in the literature for the use of single frequency BIA for the measurement of acute changes in body fluid (27, 28), it has nothing to offer above methods of assessment currently used in clinical practice, although MFBIA may be more accurate and deserves further evaluation (29, 30). Segmental impedance warrants further investigation and development as a potentially useful method for investigation of regional changes in body fluid, though it appears insensitive to changes within the peritoneal cavity. One of the major problems to be overcome if segmental BIA is to become a useful technique is the standardization of methodology, particularly electrode positioning. The sum of the resistances of the three segments with the electrode placings used in this study is greater than whole body resistance. This is likely to be due to the use of different current electrode configurations for each segmental measurement, resulting in different current distribution from whole body measurements (6).

12.

13.

14.

15. 16. 17.

18.

19.

20.

21.

References 22. 1. Lukaski H C, Johnson P E, Bolonchuk W W, Lykken G I. Assessment of fat-free mass using bioelectrical impedance measurements of the human body. Am J Clin Nutr 1985; 41:810-817 2. Kushner R F. Bioelectrical impedance analysis: a review of principles and applications. J Am Coil Nutr 1992; 11:199-209 3. Chumlea W C, Baumgartner R N, Roche A F. Specific resistivity used to estimate fat-free mass from segmental body measures of bioelectrical impedance. Am J Clin Nutr 1988; 48:7-15 4. Baumgartner R N, Chumlea W C, Roche A F. Estimation of body composition from bioelectric impedance of body segments. Am J Clin Nutr 1989; 50:221-226 5. Rallison L R, Kushner R F, Penn D, Schoeller D A. Errors in estimating peritoneal fluid by bioelectrical impedance analysis and total body electrical conductivity. J Am Coll Nutr 1993; 12:66-72 6. Organ L W, Bradham G B, Gore D T, Lozier S L. Segmental bioelectrical impedance analysis: theory and application of a new technique. J Appl Physiol 1994; 77:98-112 7. Deurenberg P, van der Kooy K, Leenen R, Schouten F J M. Body impedance is largely dependent on the intra- and extra-cellular water distribution. Eur J Clin Nutr 1989; 43:845-853 8. Kong C H, Thompson C M, Lewis C A, Hill P D, Thompson F D. Detemaination of total body water in uraemie patients by bioelectrical impedance. Nephrol Dial Transplant 1993; 8:716-719 9. Segal K R, Kral J G, Wang J, Pierson R N, van Itallie T B. Estimation of body water distribution by bioelectrical impedance. Fed Proc 1987; 46:1334 10. McDougall D, Shizgal H M. Body composition measurements from whole body resistance and reactance. Surg Forum 1986; 37:42-45 11. Scheltinga M R, Jacobs D O, Kimbrough T D, Wilmore D W.

Submission date: 31 January 1996; Accepted date: 4 June 1996

23.

24.

25.

26.

27.

28.

29.

30.

Identifying body fluid distribution by measuring electrical impedance. J Trauma 1991; 33:665-670 Barnett A. Electrical method for studying water metabolism and translocation in body segments. Proc Soc Exp Biol Med 1940; 44:142-147 Panella C, Guglielmi W, Mastronuzzi T, Francavilla A. Whole-body and segmental bioelectrical parameters in chronic liver disease: effect of gender and disease stages. Hepatology 1995; 21:352-358 Piccoli A, Rossi B, Pillon L, Bucciante G. A new method for monitoring body fluid variation by bioimpedance analysis: The RXc graph. Kidney Int 1994; 46:534-539 Kushner R F, Schoeller D A. Estimation of total body water by bioeleclrical impedance analysis. Am J Clin Nutr 1986; 44:417-424 Tedner B, Lins L E, Asaba H, Wehle B. Evaluation of impedance technique for fluid-volume monitoring during haemodialysis. Int J Clin Monit Comput 1985; 2:3-8 B6hm D, Odaischi M, Beyerlein C, Overbeck W. Total body water: changes during dialysis estimated by bioimpedance analysis. Infusionstherapie 1990; 17(Suppl 3): 75-78 Ljungqvist O, Hedenborg G, Jacobson S H, Lins L-E, Samuelson K, Tedner B, Zetterholm U-B. Whole body impedance measurements reflect total body water changes. A study in hemodialysis patients. Int J Clin Monit Comput 1990; 7:163-169 Kurtin P S, Shapiro A C, Tomita H, Raizman D. Volume status and body composition of chronic dialysis patients: utility of bioelectric impedance plethysmography. Am J Nephrol 1990; 10:363-367 De Lorenzo A, Barra P F A, Sasso G F, Battistini N C, Deurenberg P. Body impedance measurements during dialysis. Eur J Clin Nutr 1991; 45:321-325 Jebb S A, Ella M. Assessment of changes in total body water in palents undergoing renal dialysis using bioelectrical impedance analysis. Clin Nntr 1991; 10:81-84 Stewart S P, Bramley P N, Heighton R, Green J H, Horsman A, Losowsky M S, Smith M A. Estimation of body composition from bioelectrical impedance of body segments: comparison with dual-energy X-ray absorptiometry. Bri J Nutr 1993; 69:645~555 Noordegraaf A V, van der Meer B J M, de Vries J P P M, de Vries P M J M. Determination of the relation between alterations of total body water and thoracic fluid content during ultrafiltration by bioelectrical impedance analysis. Nephrol Dial Transplant 1995; 10:382-385 Bradbury M G, Smye S W, Brocklebank J T. Assessment of the sensitivity of bioimpedance to volume changes in body water. Ped Nephrol 1995; 9:337-340 Cornish B H, Ward L C, Thomas B J. Alteration of the extracellular and total body water volumes measured by multiple frequency bioelectrical impedance analysis (MFBIA). Nutr Res 1994; 14:717-727 Cha K, Hill A G, Rounds J D, Wilmore D W. Multifrequency bioelectrical impedance fails to quantify sequestration of abdominal fluid. J Appl Physiol 1995; 78:736-739 McDonald J J, Chanduvi B, Velarde G et al. Bioimpedance monitoring of rehydration in cholera. Lancet 1993; 341:1049-1051 Meguid M M, Lukaski H C, Tripp M D, Rosenburg J M, Parker F B. Rapid bedside method to assess changes in postoperative fluid status with bioelectrical impedance analysis. Surgery 1992; 112:502-508 Ho L T, Kushner R F, Schoeller D A, Gudivaka R, Spiegal D M. Bioimpedance analysis of total body water in hemodialysis patients. Kidney Int 1994; 46:1438-1442 De Lorenzo A, Deurenberg P, Andreoli A, Sasso G F, Palestini M, Docimo R. Multifreqnency impedance in the assessment of body water losses during dialysis. Renal Physiol 1994; 17:326-332