Vol. 89, Issue 2, 663-671, August 2000
Total body water and ECFV measured using
bioelectrical impedance analysis and indicator dilution in
horses
Mariann
Forro,
Scott
Cieslar,
Gayle L.
Ecker,
Angela
Walzak,
Jay
Hahn, and
Michael I.
Lindinger
Department of Human Biology and Nutritional Sciences, University
of Guelph, Guelph, Ontario, Canada N1G 2W1
 |
ABSTRACT |
The purposes of this
study were 1) to determine the compartmentation of body
water in horses by using indicator dilution techniques and
2) to simultaneously measure bioelectrical impedance to
current flow at impulse current frequencies of 5 and 200 kHz to
formulate predictive equations that could be used to estimate total
body water (TBW), extracellular fluid volume (ECFV), and intracellular
fluid volume (ICFV). Eight horses and ponies weighing from 214 to 636 kg had catheters placed into the left and right jugular veins.
Deuterium oxide, sodium thiocyanate, and Evans blue were infused for
the measurement of TBW, ECFV, and plasma volume (PV), respectively.
Bioelectrical impedance was measured by using a tetrapolar electrode
configuration, with electrode pairs secured above the knee and hock.
Measured TBW, ECFV, and PV were 0.677 ± 0.022, 0.253 ± 0.006, and 0.040 ± 0.002 l/kg body mass, respectively. Strong
linear correlations were determined among measured variables that
allowed for the prediction of TBW, ECFV, ICFV, and PV from measures of
horse length or height and impedance. It is concluded that
bioelectrical impedance analysis (BIA) can be used to improve the
predictive accuracy of noninvasive estimates of ECFV and PV in
euhydrated horses at rest.
hydration; equine; deuterium oxide; sodium thiocyanate; plasma
volume; water balance; extracellular fluid volume; intracellular fluid
volume; bioelectrical impedance analysis; body weight
| |
INTRODUCTION |
IN CLINICAL AND FIELD
SITUATIONS, the rapid assessment of body mass and hydration
status are important for performance, health, and treatment evaluation.
In horses, as in humans, dehydration from disease, prolonged exercise,
or transport impairs health, well-being, and physical and cognitive
performance (4, 19, 27). At
present, it is not yet possible to assess noninvasively, rapidly, and
accurately the magnitude of extracellular and intracellular dehydration
in horses and other mammals (6). Indicator dilution techniques have been used to estimate extracellular fluid volume (ECFV)
and total body water (TBW) in the horse (for review, see Ref. 2).
However, these methods are invasive and time consuming and must be
repeated sequentially to follow a time course of change in body water
compartmentation. Because bioelectrical impedance analysis (BIA)
technology is small, portable, noninvasive, and easy to perform (13),
BIA may prove to be an important diagnostic tool if found to be a
reliable method for assessing body fluid status of horses. The present
study provides an initial assessment of the use of BIA for the
empirical determination of body mass and water compartment volume in
euhydrated horses.
Multifrequency BIA has been used for the determination of hydration
state in humans for a number of years and has been the subject of
review (6, 12, 25). The
principle of the technique is based on the impedance to the flow
of a constant, microampere alternating current passed through
conductive fluid cylinders (conductors) that represent trunk and limb
segments (13, 20). The electrical
conductivity of the body is dependent on the amount of water and
electrolytes present in the various body fluid compartments. With the
assumption that the conductors have a uniform cross section, the volume
(V) of the conductor is proportional to its length (L)
squared divided by its impedance (Z), V = 
L2/Z, where is a specific resistivity
term (6). Specific resistivity is an electrical
characteristic of the conductor that is independent of its volume and
shape (6, 35). The resistance to current flow
is due to the specific resistivity and the volume of the conducting
fluid or the fat-free mass of the animal. Cell membranes act as
electrical condensers and are a barrier to current flow at low
frequencies (
50 kHz). Hence, bioelectrical impedance measured at
1-5 kHz has been used to estimate ECFV (34,
35). In contrast, at frequencies >50 kHz, cell membranes
are not a barrier to current flow, allowing bioelectrical impedance
measured at 200 kHz to estimate TBW.
The horse, like humans, may be considered a series of conductive
cylinders for the purposes of BIA. As in humans, the impedance of the
whole body can be measured by means of a tetrapolar electrode configuration, by using the forelimb and hindlimb together with the
length of the horse to calculate the total and extracellular conductive
volumes of the body. Impedance to the flow of electrical current
injected into one cylinder and detected in another cylinder can be
measured at different frequencies, allowing for the estimation of TBW
and ECFV (6, 13, 24,
35).
The primary purposes of this study were 1) to determine the
compartmentation of body water in horses via indicator dilution techniques and 2) to simultaneously measure bioelectrical
impedance to current flow at impulse frequencies of 5 and 200 kHz.
These measures would allow us to determine whether BIA can be used to reliably estimate TBW and ECFV in horses compared with standard indicator dilution techniques and to determine the practicality of
using BIA in horses. Accordingly, we hypothesized that the use of
dual-frequency BIA increases the accuracy with which TBW and ECFV can
be predicted, compared with using body length or height measures alone
without BIA measures.
| |
MATERIALS AND METHODS |
Subjects
The experiments consisted of a pilot study that used three
horses (two Standardbreds and one Thoroughbred cross) and a subsequent full study that used eight horses (Table
1) from the University of Guelph herd.
Horses were fasted overnight but had access to water in their stalls.
The horses were healthy but not physically trained to perform exercise.
Food and water were withheld during the experiment. The experiments
were approved by the Animal Care Committee and the Veterinary Teaching
Hospital of the University of Guelph and were performed in accordance
with the guidelines of the Canadian Council on Animal Care.
Length was measured with a measuring tape on the left side of the horse
as the horizontal distance from the point of the shoulder to the
furthest curve of the rump without wrapping the tape around the curve
of the rump (Fig. 1). Height of the horse
was measured with a height measurement stick for horses as the vertical
distance from the ground to the highest point of the withers when the
horse was standing squarely on a flat surface (Fig. 1). Body mass was measured with a large animal scale (±0.5 kg, KSL Scales, Kitchener, ON, Canada) during the experiment.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 1.
Schematic diagram showing electrode placement and
anatomic landmarks for height and length measurements.
|
|
Pilot Study
Pilot studies were conducted on three horses to 1)
determine suitable sites for electrode placement, 2) compare
stainless steel electrodes with carbon fiber electrodes, and
3) determine the frequencies of current injection required
to obtain impedance measures that yielded a high degree of correlation
to TBW and ECFV. In the pilot study, ECFV was estimated as 0.222 × body mass (2). On the forelimb, electrodes were placed
on clipped and cleaned areas just above and below the left knee, and on
the left hindlimb electrodes were placed just above and below the hock. After application of electrode gel to skin sites and electrodes, electrodes were secured on the limbs with wide elastic bands with Velcro attachments. Stainless steel and carbon fiber electrodes were
used in sequence at the leg site to compare the two electrode types. On
the torso, the carbon fiber electrodes were placed in pairs 15 cm apart
(center to center) on the neck and rump and secured to the skin with
tape (see Fig. 1 for electrode placement). The stainless steel
electrodes were not used on the torso. Impedance measurements were
obtained in triplicate by using a Bodystat 5000 multifrequency
bioelectrical impedance analyzer (Bodystat, Douglas, Isle of Man, UK),
with data collected at frequencies of 5, 50, 200, and 500 kHz.
Full Study
Infusion and sampling.
The deuterium oxide (D2O) dilution volume has recently been
shown to be useful for the measurement of TBW in horses
(1). Similarly, sodium thiocyanate (NaSCN) and Evans blue
indicator dilution techniques are well established for the measurement
of ECFV (3, 10, 22) and plasma
volume (PV; Refs. 17, 21, 22).
The hair coat over the jugular vein, 10-20 cm below the mandible,
was clipped short to the skin on both sides of the neck. Each jugular
vein catheterization site was aseptically prepared for insertion of
catheters. A topical analgesic [EMLA cream (2.5% lidocaine and 2.5%
prilocaine), Astra Pharma, Mississauga, ON, Canada] was applied
30-45 min before insertion of catheters to desensitize the skin.
Local anesthetic (2% Xylocaine, Astra Pharma) was injected
subcutaneously to complete the analgesia. A catheter (14-gauge,
5.25-in. Angiocath, Becton-Dickinson, Mississauga, ON, Canada) was
inserted anterograde into the left and right veins, secured with tape,
and stitched to the skin. Four-way stopcocks with 20-in. extensions
(Medex-Hilliard) were attached to the catheters for ease of infusion
and blood sampling. Patency of the catheters was maintained with
sterile, heparinized 0.9% NaCl (2,000 IU/l NaCl).
The dilution indicator used for measuring TBW was D2O (110 mg/kg; Refs. 1, 13), infused as a 50% vol/vol solution with 0.9%
sterile saline for a total volume of 60-100 ml. ECFV was measured
using NaSCN (22 mg/kg; Refs. 3, 10), infused as a 20% wt/vol solution
in 0.9% sterile saline for a total volume of 40-60 ml. PV was
measured using Evans blue (0.11 mg/kg; Refs. 17, 23, 24) infused as a
0.5% wt/vol solution in sterile saline for a total volume of 6-10
ml. D2O was purchased from Sigma Chemical (St. Louis, MO)
or from Acros (Fisher Scientific, Nepean, ON, Canada), NaSCN from Sigma
Chemical, and Evans blue from Fisher Scientific.
The indicators were injected in sequence (Evans blue, 9 ml;
D2O, 25-50 ml; NaSCN, 25-50 ml) using the right
catheter over a period of 5 min. Sterility of the infusates was ensured
by using a nonpyrogenic, sterile 0.22-µm nylon filter (Millex-Ap/GS
Filter, Millipore S.A. 67, Mosheim, France) placed on the syringe.
Immediately after the final infusion, the catheter was flushed with 50 ml of sterile 0.9% saline. Blood was sampled from the left catheter with 7.5-ml lithium-heparin syringes (Monovette-Sarstedt, Sarstedt, Germany) before infusion of indicators and 10, 20, 30, 45, 60, 90, and 120 min after infusion. Each blood sample was transferred to
five 1.5-ml Eppendorf centrifuge tubes. Four blood samples were
centrifuged for 3 min at 15,000 g, and the other was placed on ice for subsequent whole blood analysis. Plasma (2-3 ml) was stored in 1.5-ml Eppendorf tubes and kept on ice until analyzed for
Evans blue and NaSCN. Remaining plasma was stored in 1.8-ml screw-cap
cryovials at 
20°C until analyzed for D2O.
BIA measurements.
On the basis of the pilot study results, it was decided that the
electrodes should be situated on the legs above the knee and hock, with
bioelectrical impedance measured at 5 and 200 kHz. The hair coat was
clipped short on the lateral surfaces of the left forelimb and hindlimb
above the knee and hock (about 2-mm hair length). These sites were
cleaned well with water and were dried with gauze or towels. BIA
measurements were obtained by using a prototype Equistat 2005 (Equistat, Douglas, Isle of Man, UK), with shielded leads connecting
the measuring unit to carbon fiber electrode pairs (5-cm diameter)
placed on prepared areas. Before electrode placement, some conductive
gel was rubbed into the hair coat directly where the electrode would
sit. A small amount of gel was also applied to the surface of each
electrode. On the forelimb, the two electrodes (10 cm between centers)
were situated below the elbow on the lateral portion of the radius directly over the common digital extensor, ulnaris lateralis, and
radial carpal extensor muscles. On the hindlimb, the electrode pair (10 cm between centers) was placed on the tibia directly over the long
digital extensor and lateral digital extensor muscles. The distal
electrode was used for current injection (800 µA emitted at
frequencies of 5 kHz and 200 kHz), and the proximal electrode was used
for current detection from the other limb. The instrument reported
impedance in ohms at each frequency. Measurements were repeated a
minimum of three times to ensure repeatability and consistency.
Analyses.
The whole blood sample was analyzed, in duplicate, for hematocrit and
plasma ion and metabolite concentrations (Nova Stat Profile 9+, NOVA
Biomedical, Waltham, MA). Plasma protein concentration was determined
by using refractometry (clinical refractometer model SPR-T2, Atago,
Tokyo, Japan). Plasma was analyzed for Evans blue concentration via the
dual-wavelength method (17) by use of a spectrophotometer
(DU-70, Beckman, Mississauga, ON, Canada). Plasma NaSCN concentration
was measured spectrophotometrically by a microvolume modification of
the method described by Chatterjee et al. (5). Analysis of
plasma D2O concentration was performed by Metabolic
Solutions (Nashua, NH), as described previously (1). The
D2O in plasma and water samples was reduced at 490°C to
produce deuterium gas that was measured with an isotope-ratio mass
spectrometer. The data are expressed in delta D/ml (
) relative to
Vienna standard mean ocean water (VSMOW).
Calculations.
TBW was calculated from plasma D2O concentrations by two
different sets of equations: Eqs. M1 to M3 by
Metabolic Solutions (Nashua, NH) and equation Eq. M4 from
Andrews et al. (1)
|
(M1)
|
where W is grams of water used to dilute the dose, A is grams of
dose administered to the subject, a is grams of dose diluted for
analysis, pre and post are the delta
deuterium values determined for the predose and postdose samples,
dose is the measured deuterium content of the diluted
dose, and tap is the measured deuterium content of local
(tap) water. To convert TBW to kilograms
The D2O dilution technique overestimates TBW by 4%
because of binding of deuterium to acidic amino acids and other
nonexchangeable sites. To correct for the nonexchange of deuterium in
the body, a corrected TBW (TBWMS) was obtained by dividing
TBW from Eq. M2 by 1.04
|
(M3)
|
The value calculated from the equation of Andrews et al.
(1) was also divided by 1.04 and is reported as
TBWA
|
(M4)
|
where Dose is dose in grams, APEdose is atom percent
excess of dose (99.9%), MWdose is molecular weight of
D2O = 20.02 g/mol, P is
D2O vs. VSMOW for plateau sample, b is
D2O of baseline sample (time 0), and
Rstd is ratio of deuterium to hydrogen in VSMOW
(standard = 0.00015576).
ECFV was calculated from plasma NaSCN concentration (5)
and PV from plasma Evans blue concentration (16).
Intracellular fluid volume (ICFV) was calculated as the difference
between TBW and ECFV.
Statistics.
Single and multiple stepwise linear regression analyses were used to
determine relationships among measured variables. Statistical significance was accepted at P < 0.05 at a power of
80%.
| |
RESULTS |
Pilot Study
The range in body mass spanned 119 kg (406.5-525.5 kg),
length ranged from 152 to 160 cm, and height ranged from 146.5 to 156.5 cm.
There was no difference in the bioelectrical impedance
measurements obtained using stainless steel vs. carbon fiber
electrodes. At 5 kHz, impedance was 221 ± 1 
and 219 ± 1 for stainless steel and carbon fiber electrodes, respectively. At
50 kHz, impedance values were 184 ± 1 and 183 ± 1 ; at
200 kHz, 157 ± 1 and 156 ± 1 ; and at 500 kHz, 145 ± 2 and 142 ± 1 .
Bioelectrical impedance measured at each frequency was
considerably less at the torso site than at the legs. The variability, as represented by the standard error of triplicate measures on each
horse, was also less with torso impedance measurements than with leg
impedance measurements. There was also less intersubject variability
with torso site measures compared with leg site measures. At 5 kHz,
impedance was 44.1 ± 1.8 and 211 ± 16 for torso and
leg sites, respectively. At 50 kHz, impedance was 33.4 ± 1.3 and
175 ± 12 ; at 200 kHz, 23.0 ± 1.0 and 148 ± 12 ; and at 500 kHz, 24.6 ± 2.8 and 132 ± 12 .
During the course of our observations and measurements (total of 83 data sets) on seven horses, it was found that the stance of the horse
had no effect on bioelectrical impedance measured at the leg or torso sites.
Blood and Plasma Characteristics
All measured plasma and plasma characteristics were normal:
hematocrit 36 ± 2%, plasma protein concentration 61 ± 3 g/l, Na+ concentration 136 ± 1 mmol/l, K+
concentration 4.1 ± 0.2 mmol/l, Ca2+ concentration
1.26 ± 0.02 mmol/l, Cl
concentration 99 ± 2 mmol/l, and glucose concentration 5.9 ± 0.4 mmol/l.
Fluid Compartment Volumes
Measured values for TBW, ECFV, and PV are presented in Table
1. When normalized for body mass, TBWMS was 0.677 ± 0.022 l/kg, ECFV was 0.253 ± 0.006 l/kg, ICFV was 0.356 ± 0.013, and PV was 0.040 ± 0.002 l/kg. There was no difference
between TBWMS compared with TBWA (0.676 ± 0.023 l/kg).
Linear Regression Analysis
TBW was highly correlated to mass, height, and length (Table
2). Length was the single most powerful
predictor of TBW. Using length and height together provided an
excellent estimate of TBW, reducing the standard error of the estimate
(SEE) by 25% compared with using mass alone. Using mass in addition to
length and height provided no additional predictive power. Inclusion of
bioelectrical impedance measured at 200 kHz in the regression analysis
had no effect on the SEE of these euhydrated horses and did not
increase the predictive power of the equations. Figure
2A shows that measured TBWMS agreed closely with calculated TBW (Eq. 2 with BIA, Table 2).
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 2.
Graphical representations of the relationships between
measured and calculated total body water (TBW, A;
r2 = 0.987), extracellular fluid volume
(ECFV, B; r2 = 0.998), and
plasma volume (PV, C; r2 = 0.975), all in liters. Solid line, mean; dashed lines, SE; dotted
lines, 95% confidence intervals.
|
|
ECFV was also highly correlated to mass, height, and length (Table
3). As with TBW, length was the single
most powerful predictor of ECFV. Using bioelectrical impedance measured
at 5 kHz with length (Eq. 7, Table 3) resulted in a threefold decrease
in the SEE. Similar results were obtained when using height and
bioelectrical impedance measured at 5 kHz, with this correlation
yielding the lowest SEE of 1.25 liters (Eq. 8, Table 3).
There was no benefit from including mass into the regression analysis,
nor from using a combination of length and height together. Calculated
ECFV (Eq. 7 with BIA, Table 3) agreed closely with measured
ECFV (Fig. 2B).
View this table:
[in this window]
[in a new window]
|
Table 3.
Regression equations correlating extracellular fluid volume to mass,
height, length, and bioelectrical impedance measured at 5 kHz
|
|
PV, a component of ECFV, was also highly correlated to mass, height,
length, and bioelectrical impedance measured at 5 kHz (Table
4). Height was a stronger predictor of PV
than was length, and the combination of height and bioelectrical
impedance measured at 5 kHz yielded the lowest SEE (Eq. 13,
Table 4). Calculated PV using height with impedance at 5 kHz (Eq.
13 with BIA, Table 4) closely matched measured PV (Fig.
2C). PV was also highly correlated to ECFV according to the
relationship
|
(16)
|
Body mass (kg) was highly correlated to length (in cm; Eq. 17), and slightly more so to the combination of length and height (in cm; Eq. 19). There was no improvement in the correlation
with the addition of impedance measurements.
|
(17)
|
|
(18)
|
|
(19)
|
View this table:
[in this window]
[in a new window]
|
Table 4.
Regression equations correlating plasma volume to mass, height,
length, and bioelectrical impedance measured at 5 kHz
|
|
| |
DISCUSSION |
The present study simultaneously measured TBW, ECFV, and PV in a
group of horses ranging in mass from 212 to 636 kg. Compartment volumes, normalized to the mass of the animal, were remarkably consistent over the range of body mass. Furthermore, the size of these
volume compartments could be estimated with reasonably good accuracy by
using dual-frequency BIA when horse length or height was used in the
predictive equation. For ECFV and PV, using bioelectrical impedance
measured at 5 kHz greatly improved the predictive estimates of these
volumes compared with not using BIA measures.
Pilot Study
There was no difference in impedance readings between the
stainless steel and carbon fiber electrodes. It was thought that that
localized changes in skin temperature (and hence blood flow) may affect
whole body bioelectrical impedance measurements; however, this has
recently been disproved (9). Nonetheless, it is suggested that carbon fiber electrodes be used because 1) avoiding
changes in skin temperature may increase comfort and acceptance with
repeated use and 2) the fact that carbon fiber electrodes
are soft and compliant makes them less likely to cause compression and
discomfort on the legs.
The intrasubject variability of bioelectrical impedance measurements
obtained from the torso were considerably reduced compared with leg
measurements. This finding, along with the fact that the torso
represents the main conductive cylinder in the body, indicates that the
torso may be a preferred site for obtaining measures of bioelectrical
impedance in large mammals. However, at this site, the difficulties
include the need to shave the hair coat and ensuring good contact of
the electrodes to the skin sites, particularly if the horse is moving
around a bit. Therefore, the torso is not a preferred site.
On the leg sites, the choice of electrode placement above and below the
knee or hock, although yielding reasonably good data, was also deemed
not satisfactory for the following reasons. First, the interelectrode
distance over the joint was variable from one time to the next and
between horses. This variance was due to anatomical differences between
horses and to the shape of the leg around the joints, which made it
difficult to maintain the electrodes in the same position. Second,
equine athletes, as do humans, often experience joint swelling and
inflammation; use of a swollen or inflamed joint for BIA measurements
could be a source of error. For ease of measurement and for avoidance
of the knee and hock joints, it was concluded that placement of both electrodes should be a minimum of 5 cm above the knee or hock joint and
that this placement be standardized, as described in MATERIALS
AND METHODS.
Bioelectrical and morphometric data were analyzed by using a series of
single and stepwise multiple linear regression analyses. Statistical
analysis of the pilot study data indicated that length, height, and
bioelectrical impedance at both 5 and 200 kHz were required to obtain
the best prediction of body mass and estimated ECFV. Impedance
measurements made at 50 and 500 kHz provided no additional power or
accuracy to the predictive equations.
Full Study
Methodology.
In contrast to previous studies that have utilized mass as an
independent variable for estimating the volume of body fluid compartments in horses (see Ref. 2), the present study emphasizes the
use of length and height for two important reasons. First, body mass is
a dependent variable that changes in response to feeding and hydration
state, and as such it can change substantially and rapidly with a time
course that can be in the order of minutes. Therefore, the use of body
mass may result in inaccurate estimates of "normal" compartment
volumes, as has been shown in humans with fluid disorders
(34). Second, length and height measures are constant,
independent variables that can be accurately obtained by careful
measurement using a measuring tape (length) and height-measuring stick.
Volume of fluid compartments.
The present measures of TBW, ECFV, and PV agree well with those
obtained in previous studies using various indicator dilution techniques (Table 5). TBW, ECFV, and PV,
when normalized to body mass, were very consistent among horses over
the >400-kg mass range of the present study. It is noteworthy that
ECFVs obtained in the present study are similar to those reported by
Kohn et al. (22), in which the horses were physically fit
and not fasted before determination of ECFV. The higher values obtained
by Spurlock et al. (36) may be due to the fact that NaSCN
was infused in conjunction with antipyrine (to determine TBW), which
interferes with the NaSCN determination and must be extracted before
NaSCN concentrations can be quantified.
There is evidence in the literature that the volume of water
compartments, expressed per kilogram body mass, varies with trained state and breed (Table 5). In longitudinal studies conducted in humans
(8) and horses (28), as well as
cross-sectional studies performed on horses (relatively untrained:
present study and Ref. 23, PV = 39.4 ± 1.4 ml/kg; well
trained: Ref. 24, PV = 46.8 ± 2.9 ml/kg), PV increases after
several weeks of exercise training. It appears that TBW and PV are
greater in hot-blooded horses (Thoroughbreds, Arabians, quarter horses,
Standardbreds) compared with draft horses (21,
26). Furthermore, among horses studied by Marcilese et al.
(26), Thoroughbred English racehorses had a markedly
greater PV than did saddle horses, and both had a PV greater than that
measured in the untrained horses in the present study. The PV data of
the present study agree well with other studies of untrained horses
(7). Some of the discrepancy among studies may also be due
to differences in measurement and sampling techniques (17)
and fasted state of the horse. In our lab, endurance-trained
Thoroughbreds had a resting PV of 0.049 ± 0.003 l/kg
(24), 10% higher than the horses in the present study.
Andrews et al. (1) appear to have been the first to use
and report the D2O dilution space as a measure of TBW in
horses, although earlier studies utilized tritiated water
(11, 14). In contrast to previous studies
that used oral administration of D2O in horses
(1), the present study infused the D2O (mixed 50-50 with 0.9% NaCl) into the jugular vein to increase the rate of equilibration by bypassing gastric emptying and intestinal absorption. Accordingly, steady-state values of TBW were obtained within 120 min of infusion, compared with 3 h or more after oral administration. For unknown reasons, TBW values measured using tritiated water appear greater than those measured using
D2O. The TBW reported in the present study agrees well with
that of Spurlock (36) and Andrews et al. (1),
in good agreement with TBW measured by direct analysis of carcass water
content of ponies (33).
The predictive equations determined in the present study indicate that
the volume of body fluid compartments may not scale simply as a
function of body mass. Using the scaling factors for TBW (0.666 × body mass) and ECFV (0.222 × body mass) provided by Carlson
(2) will result in an increasing probability of error as
the mass of the horse decreases, yet these equations are reasonably
accurate in the 350- to 550-kg range.
BIA.
The results indicate that TBW and body mass can be accurately predicted
in mature, euhydrated horses by using the variables length and height,
without the need for bioelectrical impedance measurements. It stands to
reason, therefore, that if these equations are used to estimate TBW in
a dehydrated horse, then the value obtained will provide an estimate of
what TBW should be in that horse in the euhydrated condition. Such an
estimate will therefore overestimate TBW in the dehydrated horse.
During the course of endurance rides, horses may lose up to 50 liters
of water in 4-6 h (4, 15), yielding a
magnitude of loss in TBW that is much greater than the SEE obtained
with the regression equations of Table 2. Similar losses of TBW occur
in response to the loop diuretic furosemide administration
(18). It is highly likely, therefore, that such decreases
in TBW would be manifest by a change in bioelectrical impedance
measured at 200 kHz, suggesting that BIA could be used in conjunction
with horse length and height to detect dehydration in horses. BIA may
also be useful to assess the magnitude and rate of increase in PV and
ECFV that occurs during exercise training (as noted above) and with
heat acclimation (24). Although further research on horses
needs to be conducted using BIA to determine changes in TBW and ECFV,
this technique has been used successfully to assess hydration status in
humans who have been administered furosemide (31).
Exercise, training, diuresis, and heat acclimation all result in
simultaneous changes in hematocrit, plasma, and extracellular and
intracellular fluid ion concentrations and volumes that may need to be
considered when using BIA for assessing hydration status. Impedance to
current flow is a function of the volume of the various body fluid
compartments and the ion concentrations in those compartments; therefore, simultaneous changes in compartment volume and ion concentration may impair the accuracy of the technique. Because red
blood cells are a tissue compartment and likely behave as the rest of
the body's cellular mass, it is not likely that changes in hematocrit
during exercise or training will have an appreciable effect on BIA measurements.
The "gold standard" technique most used for the determination of
TBW in horses has been the dilution of tritiated water (see Table 5).
Because tritium is radioactive, it has not been used for the clinical
determination of hydration status. The use of D2O produces
similar values of TBW as does tritiated water (Table 5), indicating
that D2O could be used in clinical testing; however, there
are appreciable time constraints and costs associated with D2O analysis. In contrast, a rapid and inexpensive
determination of hydration status and body water compartmentation could
be achieved by using BIA. Length or height, when used in conjunction
with bioelectrical impedance measured at 5 kHz, improved the predictive accuracy of estimates of ECFV (Table 3) and PV (Table 4). The use of
bioelectrical impedance measured at 5 kHz resulted in a three- to
fourfold decrease in SEE for ECFV and PV. Length and height
measurements yielded strong correlations because they reflect the
distances through which current passes in cylindrical segments of the
body. A similar degree of predictive accuracy has been reported in
humans when using height squared (with intentional omission of body
mass) as the sole morphometric variable (34, 35).
Conclusions
BIA is useful for improving the predictive accuracy for rapid,
noninvasive estimation of ECFV and PV in mature, euhydrated, resting
horses. For the clinician and researcher, BIA provides many advantages
over dilution techniques, which are time consuming and invasive.
Because body mass is not required as an input variable, BIA, together
with measures of length and height, can be readily used in field and
farm situations to assess hydration status. Ongoing research will
determine the practical value of the BIA technique for estimating TBW,
ECFV, and PV in dehydrated horses.
| |
ACKNOWLEDGEMENTS |
We acknowledge the assistance of Nelson Cole, Barry Cole, Gerry
Finlay, and staff of the Veterinary Teaching Hospital. We thank
Dr. Leslie Huber for providing veterinary assistance and consultation.
We are grateful to Bodystat for providing the Equistat 2005.
| |
FOOTNOTES |
This research was supported by the Ontario Ministry of Foods and Rural
Affairs, Bodystat, and the Natural Sciences and Engineering Research
Council of Canada.
Address for reprint requests and other correspondence:
M. I. Lindinger, Dept. of Human Biology and Nutritional Sciences,
Univ. of Guelph, Guelph, ON, Canada N1G 2W1 (E-mail:
mlinding{at}uoguelph.ca).
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Received 21 October 1999; accepted in final form 1 April 2000.
| |
REFERENCES |
1.
Andrews, FM,
Nadeau JA,
Saabye L,
and
Saxton AM.
Measurement of total body water content in horses, using deuterium oxide dilution.
Am J Vet Res
58:
1060-1064,
1997[Medline].
2.
Carlson, GP.
Hematology and body fluids in the equine athlete: a review.
In: Equine Exercise Physiology 2, edited by Gillespie JR,
and Robinson NE.. Davis, CA: ICEEP Publications, 1987, p. 393-425.
3.
Carlson, GP,
Harrold D,
and
Rumbaugh GE.
Volume dilution of sodium thiocyanate as a measure of extracellular fluid volume in the horse.
Am J Vet Res
40:
587-589,
1979[ISI][Medline].
4.
Carlson, GP,
Ocen PO,
and
Harrold DR.
Clinicopathologic alterations in normal and exhausted endurance horses.
Theriogenology
6:
93-104,
1976.
5.
Chatterjee, MS,
Abdel-Rahman M,
Bhandal A,
Klein P,
and
Bogden J.
Amniotic fluid and thiocyanate in pregnant women who smoke.
J Reprod Med
33:
417-420,
1988[Medline].
6.
Chumlea, WC,
and
Guo SS.
Bioelectrical impedance and body composition: present status and future directions.
Nutr Rev
52:
123-131,
1994[ISI][Medline].
7.
Collery, L,
and
Keating J.
Blood volume determination in horses using radioactive phosphorus.
Vet Rec
30:
216-218,
1958.
8.
Convertino, VA.
Blood volume: its adaptation to endurance training.
Med Sci Sports Exerc
23:
1338-1348,
1991[Medline].
9.
Cornish, BH,
Thomas BJ,
and
Ward LC.
Effect of temperature and sweating on bioimpedance measurements.
Appl Radiat Isot
48:
475-476,
1998.
10.
Crandall, LA,
and
Anderson MX.
Estimation of the state of hydration of the body by the amount of water available for the solution of sodium thiocyanate.
Am J Dig Dis
1:
126-131,
1934.
11.
Deavers, S,
Rossborough JP,
Garner HE,
Huggins RA,
and
Amend JF.
Blood volumes and total body water in the domestic pony.
J Appl Physiol
34:
341-343,
1973[Free Full Text].
12.
De Lorenzo, A,
Andreoli A,
Matthie J,
and
Withers P.
Predicting body cell mass with bioimpedance by using theoretical methods: a technological review.
J Appl Physiol
82:
1542-1558,
1997[Abstract/Free Full Text].
13.
De Lorenzo, A,
Candeloro N,
Andreoli A,
and
Deurenberg P.
Determination of intracellular water by multifrequency bioelectrical impedance.
Ann Nutr Metab
39:
177-184,
1995[ISI][Medline].
14.
Dunkle, NJ,
Bottoms GD,
Fessler JF,
Knox K,
and
Roesel OF.
Effects of flunixin meglumine on blood pressure and fluid compartment volume changes in ponies given endotoxin.
Am J Vet Res
46:
1540-1544,
1985[Medline].
15.
Ecker, G,
and
Lindinger MI.
Effects of terrain, speed, temperature and distance on water and ion losses in endurance horses.
Equine Vet J Suppl
18:
298-305,
1995.
16.
Evans, JW.
Effect of fasting, gestation, lactation and exercise on glucose turnover in horses.
J Anim Sci
33:
1001-1004,
1971.
17.
Foldager, N,
and
Blomqvist CG.
Repeated plasma volume determination with the Evans blue dye dilution technique: the method and a computer program.
Comput Biol Med
21:
35-41,
1991[ISI][Medline].
18.
Freestone, JF,
Carlson GP,
Harrold DR,
and
Church G.
Influence of furosemide treatment on fluid and electrolyte balance in horses.
Am J Vet Res
49:
1899-1902,
1988[Medline].
19.
Guthrie, AJ,
and
Lund RJ.
Thermoregulation. Base mechanisms and hyperthermia.
In: Veterinary Clinics of North America: Equine Practice, Fluids and Electrolytes in Athletic Horses, edited by Hinchcliff KW. Philadelphia, PA: Saunders, 1998, vol. 14, p. 45-59.
20.
Hoffer, EC,
Meador C,
and
Simpson D.
Correlation of whole-body impedance with total body water volume.
J Appl Physiol
27:
531-534,
1969[Free Full Text].
21.
Julian, LM,
Lawrence JH,
Berlin NI,
and
Hyde GM.
Blood volume, body water and body fat of the horse.
J Appl Physiol
8:
651-653,
1956[Free Full Text].
22.
Kohn, CW,
Muir WW,
and
Sams R.
Plasma volume and extracellular fluid volume in horses at rest and following exercise.
Am J Vet Res
39:
871-874,
1978[ISI][Medline].
23.
Lindinger, MI,
Geor RJ,
Ecker GL,
and
McCutcheon LJ.
Plasma volume and ions during exercise in cool dry, hot dry and hot humid conditions.
Equine Vet J Suppl
20:
133-139,
1995.
24.
Lindinger, MI,
McCutcheon LJ,
Ecker GL,
and
Geor RJ.
Heat acclimation improves regulation of plasma volume and plasma Na+ content during exercise in horses.
J Appl Physiol
88:
1006-1013,
2000[Abstract/Free Full Text].
25.
Lusseveld, EM,
Peters E,
and
Deurenberg P.
Multifrequency bioelectrical impedance as a measure of differences in body water distribution.
Anim Nutr Metab
37:
44-51,
1993.
26.
Marcilese, NA,
Valsecchi RM,
Figueiras HD,
Camberos HR,
and
Varela JE.
Normal blood volumes in the horse.
Am J Physiol
207:
223-227,
1964.
27.
Maughan, RJ,
and
Lindinger MI.
Preparing for and competing in the heat: the human perspective.
Equine Vet J Suppl
20:
8-15,
1995.
28.
McKeever, KH,
Schurg WA,
Jarrett SH,
and
Convertino VA.
Exercise training-induced hypervolemia in the horse.
Med Sci Sports Exerc
19:
21-27,
1987[ISI][Medline].
29.
Muir, WM,
Kohn CW,
and
Sams R.
Effects of furosemide on plasma volume and extracellular fluid volume in horses.
Am J Vet Res
39:
1688-1691,
1978[Medline].
30.
Obara, J,
and
Nakajima H.
Iron metabolism in equine infectious anemia: III. Measurement of blood volume, mean corpuscular hemoglobin and mean corpuscular hemoglobin concentration and estimation of total iron.
Kachiku Eisei Shikenjo Bull
42:
45-55,
1961.
31.
Olde Rikkert, MG,
Deurenberg P,
Jansen RW,
van't Hof MA,
and
Hoefnagels WH.
Validation of multifrequency bioelectrical impedance analysis in monitoring fluid balance in healthy elderly subjects.
J Gerontol A Biol Sci Med Sci
52:
M137-M141,
1997[Abstract].
32.
Perrson, SGB,
Ekman L,
Lydin G,
and
Tufvesson G.
Circulatory effects of splenectomy in the horse. II. Effects on plasma volume and total and circulating red cell volume.
Zbl Vet Med A
20:
456-468,
1973.
33.
Robb, J,
Harper RB,
Hintz HF,
Reid JT,
Lowe JE,
Schryver HF,
and
Rhee MSS
Chemical composition of adipose tissue, and liver and kidney size in the horse.
Anim Prod
14:
25-34,
1972.
34.
Sergi, G,
Bussolotto M,
Perini P,
Calliari I,
Giantin V,
Ceccon A,
Scanferla F,
Bressan M,
Moschini G,
and
Enzi G.
Accuracy of bioelectrical impedance analysis in estimation of extracellular space in healthy subjects and in fluid retention states.
Ann Nutr Metab
38:
158-165,
1994[Medline].
35.
Siconolfi, SF,
Gretebeck RJ,
Wong WW,
Pietrzyk RA,
and
Suire SS.
Assessing total body and extracellular water from bioelectrical response spectroscopy.
J Appl Physiol
82:
704-710,
1997[Abstract/Free Full Text].
36.
Spurlock, GH,
Landry SL,
Sams R,
McGuirk S,
and
Muir WW.
Effect of endotoxin administration on body fluid compartments in the horse.
Am J Vet Res
46:
1117-1120,
1985[Medline].
J APPL PHYSIOL 89(2):663-671
8750-7587/00 $5.00
Copyright © 2000 the American Physiological Society
TOP
ABSTRACT
INTRODUCTION
![×[(&dgr;<SUB>dose</SUB><IT>−&dgr;</IT><SUB>tap</SUB>)<IT>/</IT>(<IT>&dgr;</IT><SUB>post</SUB><IT>−&dgr;</IT><SUB>pre</SUB>)]](/content/vol89/issue2/fulltext/663/img002.gif)
