• Open Access

Systemic, Renal, and Colonic Effects of Intravenous and Enteral Rehydration in Horses

Authors

  • G.D. Lester,

    Corresponding author
    1. College of Veterinary Medicine, School of Veterinary and Life Sciences, Murdoch University, Murdoch, WA, Australia
    • Department of Large Animal Clinical Sciences, College of Veterinary Medicine, Island Whirl Equine Colic Research Laboratory, University of Florida, Gainesville, FL
    Search for more papers by this author
  • A.M. Merritt,

    1. Department of Large Animal Clinical Sciences, College of Veterinary Medicine, Island Whirl Equine Colic Research Laboratory, University of Florida, Gainesville, FL
    Search for more papers by this author
  • H.V. Kuck,

    1. Department of Large Animal Clinical Sciences, College of Veterinary Medicine, Island Whirl Equine Colic Research Laboratory, University of Florida, Gainesville, FL
    Search for more papers by this author
  • J.A. Burrow

    1. Department of Large Animal Clinical Sciences, College of Veterinary Medicine, Island Whirl Equine Colic Research Laboratory, University of Florida, Gainesville, FL
    Search for more papers by this author

  • Some of these data were presented at the 2002 American College of Veterinary Internal Medicine Forum, Dallas, TX

Corresponding Author: Guy D. Lester, College of Veterinary Medicine, School of Veterinary and Life Sciences, Murdoch University, South Street, Murdoch 6150 Western Australia; e-mail: G.Lester@murdoch.edu.au

Abstract

Background

Intravenous (IV) and intragastric (IG) administration of fluid therapy are commonly used in equine practice, but there are limited data on the systemic, renal, and enteric effects.

Hypothesis

IV fluid administration will increase intestinal and fecal hydration in a rate-dependent manner after hypertonic dehydration, but will be associated with significant urinary water and electrolyte loss. Equivalent volumes of IG plain water will result in comparatively greater intestinal hydration with less renal loss.

Animals

Six Thoroughbred geldings.

Methods

Experimental study. 6 by 6 Latin square design investigating constant rate IV administration at 50, 100, and 150 mL/kg/d over 24 hours in horses dehydrated by water deprivation. Equivalent volumes of IG plain water were administered by 4 bolus doses over 24 hours.

Results

Water deprivation resulted in a significant decrease in the percentage of fecal water, and increases in serum and urine osmolality. IV fluids administered at 100 and 150 mL/kg/d restored fecal hydration, but increasing the rate from 100 to 150 mL/kg/d did not confer any additional intestinal benefit, but did result in significantly greater urine production and sodium loss. Equivalent 24-hour volumes of plain water resulted in greater intestinal water and less urine output.

Conclusions and Clinical Importance

IV polyionic isotonic fluids can be used to hydrate intestinal contents in situations where enteral fluids are impractical. IV fluids administered at three times maintenance are no more efficacious and might be associated with adverse physiological findings after withdrawal. Bolus dosing of IG water can be used to restore intestinal water with minimal adverse effects.

Abbreviations
CK

cecal potassium

CNa

cecal sodium

CO

cecal osmolality

CRI

constant rate infusion

FK

fecal potassium

FNa

fecal sodium

FO

fecal osmolality

FV

fecal volume

FW%

fecal water percentage

IG

intragastric

IV

intravenous

NG

nasogastric

PCV

packed cell volume

SK

serum potassium

SNa

serum sodium

SO

serum osmolality

TPP

total plasma protein

UK

urinary potassium

UNa

urinary sodium

UV

urine volume

The clinical rationale for the use of fluid therapy in horses varies with disease type and severity, but includes replacement of losses, support of the circulatory system, maintenance of hydration, or promoting hydration of intestinal contents. The latter has been cited as the basis for the high volume IV administration of fluids in the management of horses with impaction or nonstrangulating displacement of the large intestine.[1] The premise is that the high volumes of fluid produce a net flux of water movement into the intestinal lumen and that this fluid load softens the impacted material allowing normal motility to propel contents into the small colon. Most commercial IV solutions are salt-rich replacement solutions. In horses with normal renal function these solutions result in a rapid onset of diuresis and natriuresis that could persist well after the withdrawal of IV fluids.[2]

Over the past decade, evidence indicates that IG, in contrast to IV fluid therapy, might not only be efficacious in the management of uncomplicated obstructive disease of the large colon, but could also result in faster resolution of some diseases at a reduced cost to the client.[3-5] Plain water given via a nasogastric (NG) tube does increase the percentage of fecal water experimentally, and has been used with clinical success in the management of spontaneous large colon impaction.[4, 6] However, minimal effects on intestinal or fecal hydration were reported after continuous IG infusion of tap water in euhydrated horses and it is theoretically associated with complications when larger volumes are used.[7, 8]

Despite the widespread use of IV or IG fluids in clinical practice, there are relatively few experimental reports describing their physiological consequences in normal horses. The objective of the current study was to examine systemic, renal, and enteric effects of varying rates of continuous IV fluids and bolus administration of varying volumes of free water to healthy horses that had been deprived of water for 24 hours.

Materials and Methods

Experimental Animals

Six healthy Thoroughbred geldings aged 3–12 years each with a narrow gauge silastic cecal cannula surgically implanted into the cecal apex were used. Placement of the cannulae was as previously described.[9] Mean body weight at the onset of the experiment was 502 kg (range 452–549 kg) and all had been free of illness or treatments for 2 months before the study. Anthelmintic and vaccination history were appropriate for the region. Horses were primarily fed coastal Bermudagrass hay and unimproved pasture. For each experiment, animals were relocated from pasture to raised covered stalls approximately 24 hours before the initial observation. Horses were fed identical coastal Bermudagrass hay and had water ad libitum during the acclimation period and intermittently thereafter. The project was approved by the Institutional Animal Care and Use Committee at the University of Florida.

Experimental Design

The design was Latin square utilizing 6 animals and 6 treatments. After stall acclimatization, all horses underwent 24 hours (T24 to T0) of water deprivation with continued access to hay ad libitum. This was followed by one of the following treatments from T0 to T24: A) Daily maintenance fluid requirements via a constant rate infusion (CRI) of an isotonic polyionic electrolyte solution1 [containing in mmol/L: Na 140, potassium 5, magnesium 1.5, chloride 98, acetate 27, and gluconate 23] over 24 hours (1xM-IV); B) Twice daily maintenance fluid requirements administered IV with a CRI over 24 hours (2xM-IV); C) Three times daily maintenance fluid requirements administered IV with a CRI over 24 hours (3xM-IV); D) Daily maintenance fluid requirements administered by nasogastric tube (NGT) with the calculated daily volume given over 4 treatment periods every 6 hours from T0 (1xM-IG); E) Twice daily maintenance fluid requirements administered by NGT over 4 treatments every 6 hours from time 0 (2XM-IG); F) Three times daily maintenance fluid requirements administered by NGT over 4 treatments every 6 hours from T0 (3XM-IG). Maintenance, twice, and 3 times maintenance total volumes were defined as 50, 100 and 150 mL/kg/day, respectively, with body weight determined immediately before the period of water restriction. During the treatment period horses did not have access to water, but did have access to hay. A 14-ga Teflon catheter was inserted into either the left or right jugular vein at T0. IV fluids were administered as a CRI with a commercial fluid pump by 1 or 2 channels.2 A commercial coiled line was used to allow horses to move freely within the stall.3 For IG treatments, plain water was given via a NG tube with volumes greater than 10 L given over 15 minutes.

At the completion of each 24-hour treatment period horses were offered access to plain water and hay and monitored for an additional 24 hours (T24–T48). The dry weight of hay and volume of water consumed were recorded. For each horse there was a minimum of 21 days between experiments. Control data were collected from each horse under identical housing conditions with ad libitum access to water and hay over 72-hours. These data were not compared statistically with treatment data. Estimates of large intestinal volume were not made during baseline sessions.

Sampling

Horses were fitted with a light harness that incorporated fecal and urine collection bags. These were worn from T24 to T20 and from T4 to T48. The following objective clinical variables were recorded at T24, T4, T0, T4, T8, T12, T16, T20, T24, T28, T32, T36, T40, T44, and T48: heart rate, respiratory rate, rectal temperature, and subjective assessments of hydration and perfusion: mucous membrane color and moisture, capillary refill time, and response to skin tenting.

Samples of blood, urine, cecal contents, and feces were collected at T24. At 4-hourly intervals, beginning at T4 and continuing through T48, blood, total urine output, cecal contents, and total fecal output were obtained. Blood was analyzed for serum osmolality (SO) (mOsm.kg/H2O), serum sodium [SNa] and potassium [SK] concentration (mmol/L), packed cell volume (PCV) (L/L), and total plasma protein [TPP] concentration (gm/L). Osmolality was measured with a freezing point osmometer,4 [SNa] and [SK] with a flame photometer,5 and [TPP] with a hand-held refractometer. Total urine volume (UV) was recorded and a representative sample processed for osmolality (UO) (mOsm.kg/H2O) and [UNa], and [UK] concentration (mmol/L). Urinary Na and K output for the preceding 4-hour period was calculated by multiplying the electrolyte concentration by volume, and expressed as mmol/4 h. Total fecal wet weight was determined over 4-hour periods. Representative samples were weighed and dried in an oven at 115°C for 52 hours. Samples were reweighed after drying and the percent water calculated (FW%). Samples of fecal liquor were collected and osmolality (FO), and Na [FNa] and K [FK] concentration were measured. The volume of fecal water (FV) and electrolyte output for each 4-hourly period were made by multiplying the fecal wet weight by the FW%, and the electrolyte concentration by FV, respectively. Cecal samples were obtained through the silastic cannula and were assessed for osmolality (CO), and Na [CNa] and K [CK]. Fecal and cecal liquour samples were harvested with cheesecloth filtration.

Chromium (Cr) Marker Studies

A bolus of Cr EDTA (138.55 mg in 50 mL) was prepared by the method described by Binnerts[10] and administered into the cecal body at T4. Fecal Cr concentrations were determined by the method described by Williams and others.[11] Briefly, the dried fecal samples used to determine FW% were ground to pass through 1 mm screens before ashing in a furnace at 512°C for a minimum of 5 hours. Ashed samples were digested and read in duplicate on an atomic absorption spectrophotometer6 at 357.9 nm. By simple kinetics the water turnover rate, volume of the pool size, transit time, and time to 50% recovery of marker were calculated from the log concentration curve of fecal Cr appearance.[12, 13] This method assumes a single pool model. A log plot of Cr appearance in feces produces a curve with the declining portion representing the fractional turnover rate (k). Linear regression of that line results in the following equation:

display math

where Ct is the concentration at time t, A is the intercept at t0, and k is the turnover rate. Pool size is then described by:

display math

where Q is the pool size, C0 is the log concentration at time 0, and q is the Cr dose.

Water content of coastal Bermudagrass hay was calculated by similar weigh, dry, and reweigh procedures to that described for feces. The hay had approximate moisture content between 9 and 13%.

Statistical Analysis

The effects of water deprivation were examined by a paired t-test on combined group data at T24 and T0. The effects of time, treatment, and treatment by time interactions were investigated by repeated measures analysis of variance with time as the repeated measure. There were 13 time points included in the analysis, from T0 to T48. If Mauchly's Test of Sphericity was significant then the degrees of freedom was adjusted by the Greenhouse-Geyser correction. Tukey's HSD test was used to test for significance between group means. Twenty-four hour electrolyte output and Cr data were analyzed by univariate analysis of variance and group means compared by Tukey's HSD test. Data were analyzed by means of a commercial statistics package.7 Significance was based on a P-value ≤ .05. Unless stated otherwise, values in text and graphs are expressed as mean±SE, and values in tables are reported as mean±95% confidence interval.

Results

All horses completed the study without incident except for mild colic signs in 2 horses on separate trials. Both episodes were short-lived and did not require medical intervention. The study was conducted from October to April inclusive. The mean average maximum temperature varied between 19.3°C and 28.0°C and minimum temperature between 5.6°C and 14.7°C. The daily average relative humidity ranged from 49% to 61%. There were no effects of water deprivation or treatment on any objective clinical parameters (data not shown).

Water and Hay Intake

Under baseline conditions, spontaneous water consumption was 50.2 ± 5.2 mL/kg/day. There was a significant effect of treatment on spontaneous water intake during the 24-hour recovery period (= .012) (Table 1). Horses in both 1xM groups consumed significantly more water than animals in the 2xM-IG and both 3xM groups. The volume of water consumed in the 4-hour period after reintroduction of fresh water (T28) was positively correlated with SO (Pearson correlation coefficient = 0.72, < .001) (Fig 1) and serum Na concentration at T24 (Pearson correlation coefficient = 0.80, < .001) across all treatment groups. There was no significant difference in hay intake between treatment groups. There was a trend for less hay to be consumed in the lower fluid volume treatment groups.

Table 1. The effect of treatment on urine production; values in columns with the same superscript are significantly different ( .05). Control values and total intake volumes were not included in the statistical analyzes
 Water Intake (L) (24–48 hours)Total Intake (48 hours)Urine Volume (L) (0–24 hours)Urine Volume (L) (24–48 hours)Urine Volume (L) (0–48 hours)
1xM-IV43.63a,b,c (33.72–53.54)71.05 (61.46–80.64)7.00a,e (1.34–12.65)5.58 (3.93–7.23)12.58a (6.54–18.62)
1xM-IG34.60d,e,f (24.68–44.51)61.67 (52.08–71.27)3.68b,f,h (-1.97–9.33)3.98a (2.33–5.62)7.66b,f,g (1.61–13.70)
2xM-IV30.15 (20.24–40.06)83.11 (73.52–92.70)17.39c,h (11.74–23.04)6.35 (4.70–8.00)23.74c,f (17.70–29.78)
2xM-IG25.32a,d (15.44–35.26)78.15 (68.56–87.74)6.49d,g (0.84–12.15)5.49b (3.84–7.14)11.98d (5.94–18.03)
3xM-IV23.82b,e (13.91–33.73)102.64 (93.05–112.23)29.60a,b,c,d (23.95–35.25)8.96a,b,c (7.31–10.61)38.56a,b,c,d,e (32.51–44.60)
3xM-IG23.41c,f (13.50–33.33)101.88 (92.29–111.47)19.04e,f,g (13.38–24.69)5.27c (3.63–6.92)24.31e,g (18.27–30.35)
P-value0.0120.0000.0020.000
Control26.78 (18.62–34.96)53.54 (43.39–63.69)4.84 (2.40–7.09)4.27 (2.64–5.89) 
Figure 1.

Correlation between serum osmolality and the volume of water spontaneously consumed in the initial 4 hours after withdrawal of treatment for all groups.

Body Weight

Changes in body weight (BW) are illustrated in Figure 2. There was a 6.3% decrease in body weight after water deprivation (501.7 ± 13.3 kg to 470.5 ± 13.3 kg, < .001). None of the treatment groups reached their group mean prewater deprivation BW. Although not statistically significant, a reduction in BW was also present for the baseline group where the bodyweight decreased from 512.5 ± 10.3 kg to 502.3 ± 11.9 kg over an equivalent 72 hours, most of which occurred in the initial 24 hours of confinement. After T0 there were significant effects of time (< .001) and time by treatment (< .001), but no significant difference between treatments. Both of the 3xM treatment groups resulted in the closest return to prewater deprivation mean BW, but both decreased after cessation of treatment. Although not statistically significant, the reduction in BW after treatment was greatest in the 3xM-IV group, such that the mean value at the end of the total experimental period (479.9 ± 11.9 kg) was similar to that after 24 hours of water deprivation (477.3 ± 10.7 kg).

Figure 2.

The effect of water deprivation and fluid treatment on bodyweight (kg) ± SEM. A: 1xM groups; B: 2xM groups; and C: 3xM groups. Panel C also includes 72 hours of control data from horses that did not undergo water deprivation.

Systemic Effects

Water deprivation resulted in a significant increase in [TPP], from 63.2 ± 1.3 gm/L to 70.1 ± 1.2 gm/L. Over the 48-hour treatment and posttreatment period there was a significant effect of time and time by treatment, but no significant differences between treatments (= 0.988). There was a trend for the 2xM-IV and 3xM-IV groups to rapidly normalize [TPP] values during the 24-hour treatment period, but then increase during the posttreatment period before reducing over the final 8 hours of the experiment. PCV did not change significantly in response to water deprivation (0.394 ± 0.007–0.381 ± 0.006 L/L). There were significant effects of time (= .001) and time by treatment (< .001), but no significant differences between treatments (= .739). There were rate dependent decreases in PCV in the IV treatment groups within the initial 4 hours of treatment, with the largest reduction occurring in the 3xM-IV group (0.413 ± 0.015–0.307 ± 0.017).

Water deprivation resulted in a significant increase in SO from 283.0 ± 0.7 to 294.1 ± 0.9 mOsm.kg/H2O, with significant effects of time, treatment, and time by treatment (Fig 3). The 2xM-IV and 3xM-IV groups resulted in a slow normalization of SO over the treatment period, whereas the 1xM-IV group showed a persistent elevation. There was sudden and large decrease in SO in the 1xM-IV group at T28, coincident with consumption of 20.4 ± 3.8 L of water. All IG treatments caused a decrease in SO with the 3xM-IG group reaching a mean low of 265.2 ± 2.0 mOsm.kg/H2O. All IG groups normalized soon after reintroduction of water.

Figure 3.

The effect of treatment on mean serum osmolality (mOsm.kgH2O−1) ± SEM. A: Effect of IG treatment; B: Effect of IV treatment.

Water deprivation resulted in a significant increase in [SNa] from 137.5 ± 0.3 to 142.1 ± 0.5 mmol/L. There were significant effects of time, treatment, and time by treatment (Fig 4). Treatment effects were also similar to SO. There was no significant change in [SK] in response to water deprivation or any significant effect of time or treatment.

Figure 4.

The effect of treatment on mean serum sodium concentration (mmol/L) ± SEM. A: Effect of IG treatment; B: Effect of IV treatment.

Urine

Baseline UV was 4.55 ± 0.55 L/d, or adjusted for bodyweight, 8.74 ± 1.07 mL/kg/d. There were significant time, treatment, and time by treatment effects on UV (Fig 5). Cumulative data are presented in Table 1. All IV groups were associated with greater UV than volume-matched IG treatments. This effect was present for the treatment period as well as the 24-hour recovery period. There was a significant effect of time (< .001) and time by treatment interaction (< .001), but no significant effect of treatment (= .076) on UO (Fig 6). There appeared to be rate-dependent decreases in UO during the treatment period.

Figure 5.

The effect of treatment on mean urine volume (L/4 h) ± SEM. A: Effect of IG treatment; B: Effect of IV treatment.

Figure 6.

The effect of treatment on mean urine osmolality (mOsm.kgH2O/h) ± SEM. A: Effect of IG treatment; B: Effect of IV treatment.

There were significant effects of time (< .001), treatment (= .032), and time by treatment (< .001) on [UNa]. Concentrations in all IV groups were significantly greater than in the IG groups. [UK] also varied significantly with time (< .001) and there was a significant time by treatment interaction (< .001), but no effect of treatment. Urinary electrolyte output data are presented in Table 2. In the 24-hour treatment period, Na output was significantly greater in IV groups than any of the IG groups. All IV groups were significantly different from each other, with urinary Na output increasing with increased volume. Outputs decreased in the recovery period, but remained significantly elevated in the 2xM-IV and 3xM-IV groups over all IG treatments. There was a significant effect of treatment on urinary K output. The 3xM-IV group produced significantly more K than all other groups, with the exception of the 2xM-IV group. During the 24-hour recovery period the K output was increased in all groups, but the 2xM- and 3xM-IV groups remained significantly greater than all IG groups.

Table 2. The effect of treatment on urinary electrolyte output (mean and 95% CI). Values within columns with the same superscript are significantly different ( .05). Control values were not included in the statistical analysis
 Urinary Sodium Output (mmol)Urinary Potassium Output (mmol)
Treatment (0–24 hours)Posttreatment (24–48 hours)Total Period (0–48 hours)Treatment (0-24 hours)Posttreatment (24–48 hours)Total Period (0–48 hours)
1xM-IV1620a,f,j,k,l (942–2298)437 (95–780)2057a,f,j,k,l (1408–2707)1778a (1179–2377)3085 (2280–3889)4863 (3476–6250)
1xM-IG257b,g,j (−48–563)51a,d (1–101)308b,g,,j (−41–658)1270b (813–1727)2140a (1287–2993)3411a (2107 –4715)
2xM-IV3974c,f,g,h,i (3116–4830)502d,e,f (160–845)4476c,f,g,h,i (3369–5583)2482 (1941–3024)3866 (2754–4977)6349 (4723–7974)
2xM-IG103d,h,k (43–164)54b,e (16–93)157d,h,k (82–233)1341c (711–1972)2554b (1614–3494)3895b (2331–5460)
3xM-IV5811a,b,c,d,e (4449–7174)857a,b,c (405–1309)6668a,b,c,d,e (5138–8197)3407a,b,c,d (1846–4968)5208a,b,c (2612–7803)8615a,b,c (4491–12739)
3xM-IG123e,i,l (38–209)23c,f (8–37)147e,i,l (55–238)1453d (213–2693)2452c (845–4060)3906c (1078–6733)
P-value.000.000.000.000.003.000
Control71 (27–115)68 (1–136)139 (43–235)1093 (387–1798)844 (211–1477)1938 (341–3060)

Cecum

CO significantly increased with water deprivation (259.8 ± 3.8–285.8 ± 5.4 mOsm./kgH2O, < .001). There were significant effects of time (= .018) and treatment (= .049). There was a large reduction in CO in the first 4 hours of 3xM-IG fluid administration. All IG groups resulted in a reduction in CO to within or below the control range by the end of the treatment period. The IV treatments were less effective at reducing CO. Baseline [CNa] increased from 124.8 ± 2.6 to 143.0 ± 2.7 mmol/L (< .001) with water deprivation. There was a significant effect of time (< .001), but no significant difference between treatments (= .139), although changes in [CNa] tended to mimic CO. Baseline [CK] significantly decreased from 27.5 ± 1.4 to 16.2 ± 1.0 mmol/L (< .001) with water deprivation. There was a significant effect of time (< .001), but no significant difference between treatments (= .187).

Feces

Water deprivation resulted in a significant reduction in FW% from 79.4 ± 0.3 to 74.7 ± 0.5% (< .001). There were significant time and time by treatment effects, but no significant difference between treatments (= .330) (Fig 7). Both 3xM groups resulted in an increase above the upper limit of the 95% control CI by 12 and 16 hours for the IG and IV treatments, respectively. During the posttreatment recovery period the FW% remained high in the 3xM-IG group, but fell in the 3xM-IV group to values that were outside of the control observation 95% CI.

Figure 7.

The effect of treatment on the percentage of fecal water relative to total wet weight (%) ± SEM. A: Effect of IG treatments; B: Effect of IV treatments. Dashed lines represent control data (95% CI).

There was a significant effect of treatment on FV during the treatment period (= 0.050). FV in the 3xM-IG group (14.1 ± 1.8 L) was significantly greater than the 1xM-IV group (7.1 ± 1.8 L). Other group means fell between these two and were not significantly different. There were no significant differences between groups when FV was statistically compared over the combined treatment and posttreatment periods, although there was a trend of increasing FV with increasing treatment volumes (Table 3).

Table 3. The effect of treatment on fecal water (mean and 95% CI). Values within columns with the same superscript are significantly different ( .05). Control values, treatment volumes, and total intake volumes were not included in the statistical analysis
 Treatment (L) (0–24 hours)Water Intake (L) (24–48 hours)Feed moisture (L)Total Intake (48 hours)Fecal Volume (L) (0–24 hours)Fecal Volume (L) (0–48 hours)
1xM-IV24.99 (22.83–27.16)43.63a,b,c (33.72–53.54)2.43 (1.62–3.24)71.05 (61.46–80.64)7.08a (3.65–10.30)21.83 (14.08–29.57)
1xM-IG24.92 (22.75–27.08)34.60d,e,f (24.68–44.51)2.59 (1.71–3.48)61.67 (52.08–71.27)10.34 (6.91–13.78)23.62 (15.87–31.36)
2xM-IV50.47 (48.30–52.63)30.15 (20.24–40.06)2.99 (2.11–3.89)83.11 (73.52–92.70)11.13 (7.69–14.60)27.58 (19.84–35.32)
2xM-IG50.33 (48.17–52.50)25.32a,d (15.44–35.26)2.95 (2.07–3.84)78.15 (68.56–87.74)12.77 (9.34–16.21)27.75 (20.03–35.25)
3xM-IV75.85 (73.69–78.02)23.82b,e (13.91–33.73)2.97 (2.16–3.78)102.64 (93.05–112.23)12.76 (9.33–16.20)27.79 (20.05–35.54)
3xM-IG75.85 (73.69–78.02)23.41c,f (13.50–33.33)3.93 (2.94–4.92)101.88 (92.29–111.47)14.10a (10.66–17.54)30.83 (23.10–38.58)
P-value.012.755.048.598
Control26.78 (18.62–34.96)3.35 (2.54–4.16)53.54 (43.39–63.69)13.35 (9.91–16.79)27.43 (21.31–33.55)

FO increased with water deprivation (189.2 ± 9.0–243.8 ± 8.8 mOsm.kg/H2O = .002). There were significant time and treatment by time effects, but no difference between treatments. There were significant changes in [FNa] and [FK] after water deprivation. [FNa] increased from 20.6 ± 1.9 to 43.2 ± 3.8 mmol/L (< 0.001) and [FK] decreased from 55.0 ± 1.7 to 48.4 ± 2.5 mmol/L (< .001). There were significant effects of time, treatment, and time by treatment effects for both electrolytes, with all IV groups resulting in increases and decreases in [FNa] (Fig 8) and [FK], respectively. Total fecal electrolyte output data are presented in Table 4. There was a significant effect of treatment on total Na output. Over the combined treatment and recovery periods, fecal Na output in the 2xM- and 3xM-IV groups was significantly greater than all IG treatments. There was a significant effect of treatment on fecal K output and loss was greater in the IG treatment groups.

Table 4. The effect of treatment on fecal electrolyte output (mean and 95% CI). Values within columns with the same superscript are significantly different ( .05). Control values, treatment volumes, and total intake volumes were not included in the statistical analysis
 Fecal Sodium Output (mmol)Fecal Potassium Output (mmol)
Treatment (0–24 hours)Posttreatment (24–48 hours)Total Period (0–48 hours)Treatment (0-24 hours)Posttreatment (24-48 hours)Total Period (0–48 hours)
1xM-IV394a (108–680)886d,e,f (626–1146)1280 (850–1710)247a,d (113–380)523 (405–640)769b,d (555–984)
1xM-IG365b (140–590)145a,d,g (62–229)511a,d (211–811)515 (362–669)714 (371–1059)1230 (785–1674)
2xM-IV795 (468–1123)1129g,h,i (539–1719)1925d,e,f (1022–2828)390b,e (255–525)600 (381–819)990 (668–1311)
2xM-IG277c (109–446)139b,e,h (49–229)417b,e (170–664)673a,b,c (427–918)872b (592–1151)1544c,d (1055–2033)
3xM-IV1018a,b,c,d (465–1571)1158a,b,c (579–1736)2176a,b,c (1167–3185)355c,f (253–458)381a,b (155–607)737a,c (425–1048)
3xM-IG363d (97–630)140c,f,i (50–229)503c,f (161–846)699d,e,f (534–864)866a (666–1067)1565a,b (1242–1889)
P-value.020.000.000.026.044.038
Control347 (108–586)190 (50–330)537 (259–815)677 (394–960)796 (484–1106)1472 (932–2012)
Figure 8.

The effect of treatment on fecal sodium concentration (mmol/L) ± SEM. A: Effect of IG treatments, B: Effect of IV treatments.

Chromium Marker Studies

By means of representative samples we accounted for 61.2 ± 5.1% of the total Cr dose. There was a significant effect of treatment on estimated large intestinal volume (= .048) (Fig 9). Post hoc comparison of means revealed a significant difference between the 1xM-IG and 3xM-IG treatment groups. There were no significant effects of treatment on transit (Fig 10) or time to collection of 50% of the total marker given.

Figure 9.

The effect of treatment on estimated large intestinal volume with chromium EDTA marker transit. Mean ± 95% confidence interval. Values labeled with similar superscripts are significantly different (< .05).

Figure 10.

The effect of treatment on mean transit time of chromium EDTA marker transit from cecum to feces. Mean ± 95% confidence interval. No significant difference between groups.

Discussion

The physiologic effects of IV or IG fluids were evaluated in healthy horses that had been deprived of water for 24 hours. Horses are well adapted to deal with periods of water restriction and easily tolerate 72 hours of water deprivation without adverse clinical consequences, despite a reduction of 10–12% in body mass.[14] Water deprivation represents an experimental model of hypertonic dehydration or relative water deficit.[15] The major homeostatic responses to water deprivation include a reduction in renal and fecal water losses and a decrease in metabolism to protect blood volume.[16] In the present study, water deprivation produced significant decreases in BW and increases in [SNa], SO, [TPP], UO, and [UNa] and [UK]. The impact on intestinal contents was confirmed by a significant decrease in FW% and an increase in FO and [FNa], and by significant elevations in CO and [CNa]. These results are consistent with other studies with similar or longer periods of water restriction.[14, 17-20] The reduction in BW was greater than expected after consideration of the relatively short duration of water restriction and continued access to hay. Carlson (1979) reported a 5.8% decrease in BW after 24 hours of feed and water deprivation under conditions of high environmental temperatures.[14] A small group of Standardbred horses deprived of water for 24 hours, but allowed continual access to feed, lost 3% of their BW compared with 6.3% in the current study.[21]

The intent of the model was to mimic a clinical scenario that has been described as a risk factor for the development of large intestinal impactions.[1] Changes in intestinal water content typically occur slowly and are therefore more likely to be seen in a model of water deprivation rather than other models to induce acute fluid shifts, such as IV isotonic or hypertonic solutions, diuretic use, exercise, or blood loss.[15, 16, 22-24] The principal source of fluid deficit during prolonged periods of water deprivation is from the transcellular compartment, which includes the water contained within the GIT. The intestine of the horse stores approximately 20% of the total body water pool in normally hydrated animals.[25, 26] Net movement of fluid from the intestinal lumen to the circulation occurs during dehydration such that the reduction in intestinal fluid is comparatively greater than the reduction in plasma volume, indicating a key role of intestinal water in the preservation of blood volume.[14, 19, 20, 25] Fluid uptake leads to desiccation of colonic contents, which could predispose to luminal obstruction.

A maintenance fluid value of 50 mL/kg/d was used in the present study of fed horses. Maintenance requirements are based on covering expected sensible (urine and feces) and insensible fluid losses (sweat and respiratory tract vapor). Requirements are dependent upon dietary components and feed intake, environmental conditions, pregnancy and lactation status, and level of exercise. Feed contributes a small volume of water to the intestinal tract (~3–4 mL/kg/d in the present study) and an amount of additional water that is derived from the oxidation of carbohydrates, protein, and fat within the diet (~4–5 mL/kg/d in the present study, extrapolated from English [[27]]). Insensible fluid losses in horses have been described as 31–38% of total fluid output, although losses increase in hot environmental conditions.[14, 28, 29] Maintenance volumes prescribed for sedentary hospitalized horses are generally in the range of 40–60 mL/kg/d, with most authors suggesting a value at the upper end of the range for euhydrated adult horses.[30, 31] Earlier recommendations of 54 mL/kg/d were based on spontaneous intake data reported by Tasker ([29]), although these underestimated the contribution of feed to water balance.29

Intravenous fluid administration was effective at increasing large intestinal and fecal water content at higher treatment volumes, but benefits did not appear to be rate dependent. The FW% increased in the 2xM- and 3xM-IV groups at a similar rate and both groups had reached prewater deprivation moisture levels within 16 hours. Despite maintaining a theoretical fluid deficit the 1xM-IV treatment group was also effective at restoring FW%, such that predehydration levels were reached at the completion of the 24-hour treatment period presumably at the expense of systemic rehydration. FW% declined in the 2xM- and 3xM-IV groups during the recovery period such that levels in the 3xM-IV group fell below reference ranges, indicating some negative rebound effects on intestinal hydration. The volume of large intestinal fluid was estimated by administering Cr-EDTA, a poorly absorbed liquid phase marker, into the cecal body and then assaying feces for Cr content. Consistent with the FV findings, the changes in estimated intestinal volume did not indicate any additional benefit of delivering IV fluids at the highest flow rate over the twice maintenance group. In the present study, horses received a low Na diet with an estimated daily intake of 375-440 mmol. This was reflected by a daily loss of Na through feces of approximately 190–350 mmol and minimal losses of Na in urine. Under normal dietary conditions, Na excretion occurs preferentially through the intestinal tract.[28, 29, 32] In the present study, the ratio of fecal Na output to urinary Na output approached 4:1. After IV treatment, fecal Na output values exceeded those of baseline or IG treatment, confirming net secretion of Na into the intestinal lumen with all IV treatments. Further evidence of an IV treatment effect on the intestinal tract was demonstrated through changes in electrolytes and osmolality; [FNa] significantly increased and [FK] decreased independent of rate across all IV groups, reflecting the relative amounts of these electrolytes provided in IV fluids. The increases in [FNa] coincided with changes in FO, both of which persisted well beyond the completion of the 24-hour treatment period.

There is evidence that IG therapy has benefit in the management of nonstrangulating large colon disease and suggestions of enhanced efficacy over IV treatment.[3, 5] There are several studies that have demonstrated improved efficacy of IG electrolyte-supplemented fluid over plain water with respect to hydration.[7, 33, 34] However, in situations where electrolyte supplementation might not be readily available, we have shown that periodic IG administration with plain water is safe and effective at restoring intestinal hydration. We observed a volume-dependent effect of bolus IG plain water on large intestinal contents and FV, with high volume IG fluids causing a large reduction in CO and [CNa] within 4 hours of initial treatment. A direct dilutional effect on cecal contents is not surprising given that fluid is rapidly emptied from a normal stomach; in a study of adult horses 90% of an 8-L electrolyte solution given via a NG tube was emptied from the stomach within 15 minutes.[35] Transit along an unobstructed small intestine is also rapid, with fluid loads arriving in the cecum and large intestine within 1–2 hours after administration.[36] Interestingly, the FW% had increased in all groups within 4 hours of beginning treatment. Although marker transit in the upper GIT is rapid the passage of a fluid marker through the ventral and dorsal large colon in ponies was relatively slow, indicating that the early increases in FW% were unlikely to be because of direct passage of fluid through the GIT.[36] Rather, increases could be caused by absorption in the proximal GIT with subsequent secretion into the hindgut or caused by increased colonic motility induced by a gastrocolonic reflex.[6] We assessed intestinal hydration after bolus dosing of water rather than constant IG infusion. It was suggested by others that low luminal tonicity facilitated absorption of plain water in the small intestine, thereby reducing delivery to the large colon.[7] It is plausible that bolus dosing of water exceeded a maximal rate of small intestinal absorption, resulting in a larger percentage of the treatment volume entering the hindgut.

Diet, temperature and humidity, fluid therapy, exercise, or drugs, such as diuretics or α2 adrenergic agonists, influence urine flow.[37] Increasing dietary protein and potassium will also increase urine volume; feeding a legume diet can result in a 3–4 times increase in urine flow over diets of Timothy hay or wheat straw.[38-40] Experimentally baseline flow rates are reported in the range of 0.92–1.24 mL/kg/h, but most of these animals had been maintained on high protein, high potassium legume diets.[41-43] Tasker ([29]) reported a flow rate of 0.52 mL/kg/h in horses on a mixture of Timothy and alfalfa hay.[29] Urine flow rates in Arabian geldings on Timothy hay-based diets were only 0.29–0.41 mL/kg/h, compared with 0.74–0.96 mL/kg/h for alfalfa diets (data extrapolated from reference).[40] Consistent with these findings horses in the present study had a mean urine flow rate of 0.36 ± 0.04 mL/kg/h under baseline conditions on a relatively low protein grass hay. Horses that received IV treatments produced greater volumes of urine than IG treatments at equivalent volumes. Peak flows of 8.5 mL/kg/h were observed during the 3xM-IV treatment period.

As anticipated there were highly significant, volume-dependent urinary losses of Na with IV treatments. Natriuresis is an expected finding after IV salt loading in adult horses and neonatal foals and has also been described after IV volume loading with glucose-based solutions.[2, 44, 45] Traditionally this response has been attributed to changes in vascular pressure and down-regulation of the renin-angiotensin-aldosterone system, but marked natriuresis can occur after modest Na loading in the absence of any significant change in blood pressure.[46] Elimination of Na loads through the renal system is associated with induced losses of K and other electrolytes, including calcium and magnesium.[2, 44, 45] These losses could contribute to systemic weakness and reduced intestinal motility. In the present study there was a substantial, IV volume-dependent urinary K loss that well exceeded estimated intake from diet and treatment. Kaliuresis occurs in response to increased Na delivery to the distal segment of the distal tubule and has been recognized as a complication of IV Na loading in horses.[2, 47] In contrast to urinary Na output, the loss of K was significantly greater during the recovery period than during the period of treatment with IV treatments. However, the IG treatments had no significant effect on total urinary Na output, but did induce an increase in urinary K loss.

The inability to restore BW during the treatment and recovery periods in the IV treatment groups was preempted by earlier work, where a similar BW response was seen in horses given 20 L of 0.9% sodium chloride over 4 hours after 48 hours of feed and water deprivation.[2] Those horses had an immediate, but small improvement in BW with fluid administration, which then declined over the following hours to dehydrated levels. The authors suggested that the need to excrete Na resulted in urinary electrolyte and water losses well before rehydration had been achieved. Our results support these observations. In our study, the lowest recorded SO measurement during the reintroduction to water period was 268 mOsm.kg/H2O in an animal in the 1xM-IV group; 1 horse had a measurement of 259 mOsm.kg/H2O during the treatment period in the 3xM-IG group. No gross evidence of RBC hemolysis was noted, but the latter value was higher than that reported to induce osmotic destruction of horse erythrocytes.[48] Seizure activity has been described as a potential complication of high volume plain water consumption because of rapid changes in Na concentration.[8] The largest mean 4 hour reduction in [SNa] after an IG water bolus was approximately 5 mmol/L. In contrast, the reduction in [SNa] after spontaneous drinking in the 1xM-IV group was around twice this value. No adverse clinical or neurologic consequences were seen in either group.

Horses that received IG water typically normalized very quickly after the end of the treatment period whereas there were lingering effects of IV fluid administration that persisted well into the recovery period. In our study IV infusion of isotonic polyionic electrolyte solutions at relatively modest rates resulted in normal SO and [SNa] despite inadequate restoration of BW. An increase in cerebral Na concentration is a primary trigger of thirst as well as a number of other functions, including natriuresis, vasopressin release, and inhibition of renin production.[49] Consistent with this general association SO and [SNa] concentration were highly correlated with water consumption after reintroduction of water in the present study. In contrast to horses that received IG plain water boluses, there were increased urine flow and electrolyte losses in animals receiving IV treatments for 4–8 hours into the recovery period. These residual losses occurred in the face of reduced spontaneous water intake and resulted in several changes over the following 24 hours that reflected rebound hydration problems, including an elevated [TPP] and PCV, and reduced BW.

In conclusion, IV isotonic polyionic fluid administration resulted in improved intestinal and fecal hydration in horses that were previously deprived of water. However, the magnitude of this response was no greater following 150 than 100 mL/kg/d in the current model. Commercial replacement solutions fluids are designed to replace losses on an isoequivalent basis and, as such, are relatively salt rich. The question arises if isotonic maintenance fluids with a lower Na concentration would be equally or more effective in this model. It could be argued that the Na content of replacement fluids might be a requirement for the relatively small effect on intestinal water seen in the present study. Intravenous infusion of replacement isotonic polyionic replacement fluids was not only ineffective at adequately restoring BW in horses with experimental free water loss, but was also associated with a number of complications including significant renal water loss, natriuresis and kaliuresis and rebound dehydration after withdrawal. The latter might be related to inadequate spontaneous water intake in the face of ongoing losses. Potential side effects should be considered when replacement solutions are used to maintain hydration or induce intestinal fluid secretion through overhydration. Plain water given IG was effective at improving intestinal and fecal water. On the basis of previous work it might be more efficacious to use bolus administration rather than continuous flow. Plain water can be safely administered to horses via a NGT bolus, but might induce kaliuresis and hyponatremia at higher volumes.

Acknowledgments

Thank you to Drs L. Chris Sanchez and T. Widenhouse for assistance with data collection. Project was supported by the American Quarter Horse Association.

Conflict of Interest: Authors disclose no conflict of interest.

Footnotes

  1. 1

    Normosol-R; Hospira, Inc Lake Forest, IL

  2. 2

    Baxter Travenol 6301 Dual Channel Pump; Baxter Healthcare, Deerfield, IL

  3. 3

    STAT IV 1000; International Win Limited, Kennett Square, PA

  4. 4

    Osmette; Precision Systems, Sudbury MA

  5. 5

    IL343 Flame Photometer; Instrumentation Laboratories, Lexington MA

  6. 6

    Perkin Elmer 5000; Perkin Elmer, Wellesley, MA

  7. 7

    PASW Statistics 18.0; SPSS Inc, IBM Company Headquarters, Chicago, IL

Ancillary