Koenigsberg, Peter S., Kitrian K. Martin, Holly R. Hlava, and Marvin L. Riedesel. 1995.

Sus tained hyperhydration with glycerol

ingestion.

i.ifq Sciences 57 (7): 64s-6s3

0024-3205(9t0031.6-9

SUSTAII\ED HYPERHYDRATION

WITII GLYCEROL INGESTION

Peter S. Koenigsberg, Kitrian K. Martin, Holly R. Hlava, and Marvin L. Riedesel Department of Biology

The University of New Mexico Albuquerque, New Mexico 87131-1091 USA

(Received in final form May 17, 195)

Strmmr*
Heavy exercise lasting more than three hours tends to result in dehydration, as the fluid intake is less than fluid loss by sweat and urine. Dehydration as small as one percent ofbody weight has been reported to decrease work capacity. In present and previous studies insensible water loss and sweat are assummed to be the same in both control and experimental conditions. Fluid intake less urine volume is utilized as an indicator ofeuhydration, hypohydration, or hyperhydration. Previous studies involving glycerol intake describe hyperhydration for 4.5 to 8 hours. The objective of this study was to keep subjects hyperhydrated (retention of water) for 32 or 49 hours. The experimental protocol involved ingestion of a large volume of fluid (39.2 or 51.1 mVkg/d) with glycerol Q.9 tD 3.I g/kg/d) and without glycerol. In both Series I (49 h) and Series II (32 h) experiments, the intake ofglycerol resulted in smaller urine volumes. This study demonshates it is possible to keep human subjects hyperhydrated for extended periods of time and thereby reduce the amount of fluid consumption necessary just prior to or during bouts of negative fluid balance situations.
Key Words: water retention, hyperhydration, glycerol-induced hyperhydration, glycerol

Preventing dehydration of athletes, astronauts, and other active people represents a complex
problem. The renal, endocrine, gasfio-intestinal, central nervous and cardiovascular systems interact in maintaining a physiological state of euhydration. This interaction of systems provides a challenge

to applied physiologists trying to prevent negative water balance during conditions and activities
which result in dehydration. Prior to competition and intense training, athletes tend to avoid drinking for fear of gasEo-intestinal discomfort or inconvenience of voiding urine. As a result, dehydration, decreased sweat rate, increased core temperature, and reduced athletic performance may occur (10,

7). One

approach to preventing hypohydration is to hyperhydrate prior to heavy exercise. Hyperhydration has been demonstrated to provide an advantage for subjects exercising in the heat (6). Ingestion of glycerol (1 g/kg) can induce a state of hyperhydration (4, 5, 9), which increased Author Responsible for Correspondence: Marvin L. Riedesel, Department of Biology, The University of New Mexico, Albuquerque, New Mexico 87131-1091 USA; 505/277 -2824; Fax: 505 1277 03M; E-mail: riedesel@mail.unm.edu.

ffi

Fluid Intake, Glycerol and Urine Volume

Vol.57, No.7,

1995

sweating and decreased the extent of dehydration, in addition to reducing the elevation of heart rate and body temperature of subjects exercising in a hot environment. These studies involved maintain-

ing glycerol induced hyperhydration (GIH) for 2.5 to 4.5 h.

Multiple variables are involved in extending GIH. Glycerol rapidly moves from one fluid compartment to another as evidenced by serum glycerol values reaching a peak within 15 to 90 min after a single oral ingestion (8, 12). The catabolism and renal excretion of glycerol varies with serum concentration (i, tf). The rate of glycerol uptake and retention differs among tissues (4). The effects of glycerol on volume and osmotic receptors have not been described. These variables make the timing and amount of glycerol intake needed to obtain fluid retention difficult to predict. One advantage of using the osmotic action of glycerol to expand total body water is the glycerol ,p"." r.pt"r"its 65% oi totut body mass. Widespread application of GIH is going to require many differeni research approaches. The present study is an attmpt to describe timing and volume intakes of various fluids which can be effective in extending the hyperhydraiton beyond the 4.5 hours described in previous research (4, 9). Prior studies have involved retention of 500 to 900 ml of water and the aim of this study was to have fluid intake exceed urine output by similar amounts for longer periods of time.
Methcxls and Materials The male subjects were students who were limited to sedentary activities such as attending classes during all experiements. This limitation accounts for the 32- and 49-hour duration of experiments. Subjects had no health problems and had signed consent forms approved by the University of New Mexico Human Research Review Committee. A two-day pilot study was conducted prior to all Series I and Series II experiments. This pilot study had two purposes (i) to ensure subjects were euhydrated prior to control and experimental protocols, and (ii) to familiarize subjects with the .outine of ingesting specific volumes of fluids at specific times and recording the time and volume of urine voidi. Ouring ttre pilot studies subjects ingested a minimum of 2500 m1/70 kg body weight per day. Subjects had meals at a student cafeteria and od libitum fluid intake at all times, but were required to record all fluid intake. Statistical analyses were paired t test, linear regression or analysis

ofvariance for repeated measures.

Series

8.7 kg, 20 to 26 years, participated in a 49-h control (water and orange Seven subjects, 75.8 juice) period and a 49-h experimental (water and orange juice with glycerol added) period @able t;. nui"r and volumes of AuiO ingestion were based on semm glycerol and urine data collected in previous 4.5 h studies (4). Subjects reported to the laboratory at 0700, 1200, 1600 and 2000 h each bay for meals, blood draws, and collection of urine voided. The blood draws were conducted after thi subjects had been seated for 15 min. Three subjects completed the control period first and four subjecis completed the experimental period first. There was a one-week interval between the
experimental and control protocols.
Blood was analyzed for hematocrit (Hct), hemoglobin (Hgb), plasma osmolality, serum glycerol and and serum creatinine. Urinesamples were pooled at 2000 h and 0800 h (urine voided between 0700 h of Day 1 represented one pooled sample, and urine voided between 2000 h of Day 2 and 2000 pooled sample, etc.) for determination of volume, specifrc 0700 h of Day 3 represented

"noih"t

Vol.57, No. 7, 1995

Fluid Intake, Glycerol and Urine Volume

&7

gravity' glycerol and creatinine. Serum samples for glycerol and osmolality analyses were frozen at -72"c within 15 min of the blood draw, and all analyses were made on the same day.
TABLE I. The Timing and Volumes of Fluid Intake During the 49-h Control and Experimental Periods

TIME
0700 0730 0800 1000 1130 1200

CONTROI.

FXPFRTMENTAI

ORANGEJUICE. WATER

(g/ke)
4.29 4.33 3.87

GLYCEROL ORANGEJUICE WATER

(e/ks)

(ml/ke)

(e/ks)
3.29

(ml/ke)

,t: *t
2.50

l.2t
1.48

1.00 0.76 0.30 0.30

21.40 3.57 3.57

0.91 1.10

1400 1430 1630

t*
0.38

*t
2.50

1700 2000

'j

a1.43 34.43

1.

l0

^^ 1.43
3.12

TOTAL/DAY 16.6

t3.54

34.43

Glycerol analyses were made by use of a test kit (Stat Pak Enzymatic Triglyceride Glycerol from Behring Diagnostics). Plasma osmolality values were obained by freezing point depression (Advanced Instruments, Precision Model 3R osmometer). Fluid retention was derived by subtracting the urine volume from the fluid intake. Respiratory and sweat fluid losses were assumed to be the same during experimental and control protocols. Serum and urine creatinine were determined by photometric analyses.
Series

II

Six male subjects, 81.6+ 1.0 kg, 20 to 46 years, participated in a crossover experimental design similar to Series I with the following exceptions: (i) The timing and volumes of fluid intake were as described in Tbble II. (ii) The glycerol was administered as a 20% solution (3.12 g/kg/d) nd included an artificial strawberry flavor (provided by Gatorade) to mask the sweet taste of the glycerol. The placebo was the same volume as the glycerol solution and included the strawberry
flavor plus asparytame (Nutrasweet o) and acesulfame K (Sunette
to the glycerol solution.
@)

to provide sweetness similar

The methods and timing of blood and urine collection and analyses were the same as in Series I. Because there were no significant changes in Hgb, Hct, or plasma osmolality in Series I (p > 0. 1), these analyses were not conducted in Series II. Automated chemical analyses of serum samples (Hitachi, Model 747) were conducted. These analyses included glucose, nitrogen, uric acid, calcium, phosphorus, cholesterol, triglyceride, total bilirubin, direct bilirubin, indirect bilirubin, sodium, potassium, chloride, total protein, albumin, globulin, and lactic acid dehydrogenase.

6rA

Fluid Intake, Glycerol and Urine Volume

Vol.57, No.7,

1995

TABLE II. The Timing and Volume of Fluid Intake During the 32-h Control and Experimental periods PLACEBO or
20% GLYCEROL WATER

(ml/ke)
0730
0800 0830
1030 1100 1130 1400

5.0 0.5
1.5
1.5

,y
*

2.0

1430
1630 1700

;
2.0

2.5

*
4.6 37.6

2000

2030

TUIAL
Day 2: 0730
0800 0830
1030

14.5

5.0 21.4 0.5

1.5
1.5

I 100
1130

*
8.5

TOTAL

25.9

Resrrlts

Series

in urine volumes in the control and experimental conditions remaini'significant throughout the 49-h periods (Fig. 1). The mean urine volume two hours after fluid inges=tion was &4 gZ ml with glycerol, and 1289 1 14 ml without glycerol a 73.3% larger voluire with the placebo (p < 0'01). At the end of the 49-h experiments, the mean accumulated urine volume was 4431 + 331 ml and 5177 + 583 ml in glycerol and control experiments, respectively, indicating a GIH of 746 ml assuming the sweat and respiratory fluid losses were the same in both the experimental and control protocols.
difference

The mean volume of urine voided during the first hour after drinking the large volume of water plus the glycerol soluiton was less than when drinking the large volume of water without glycerol. hhe

No significant changes were observed in hematocrit, hemoglobin or plasma osmolality (p

due to the ingestion of glycerol (Table

III). Specific gravity of urine

> 0. 1)

and creatinine clearance were

Vol.57, No.7, 195

Fluid Intake, Glycerol and Urine Volume

&9

H H

URINE, NOGLYCEROL URINE. GLYCEROL

Fig.

1.

Accumulated volume of fluid intake and urine voided with and without glycerol, Series I.

similar for ttre experimental and control days (p > 0.1). Creatinine clearance ranged from 79 to 148 mVmin during connol and 85 to 125 mVmin with glycerol ingestion. Because there were no changes in these data, it is assumed the plasma volume, renal blood floq and glomerular filtration were not changed as a result of glycerol ingestions. Following glycerol ingestion, the mean serum glycerol values 4.5 h after the initial ingestion were near 100 mg/dl on both days one and two (Ihble IV). However, 8.5 h after the initial ingestion, the serum glycerol had decreased by 507o on both days one and two. The mean 24-h urine volume, 2.6 liter, had a mean glycerol content of 39 g. The mean glycerol intake was 235 gl24 h. Thus renal excretion accounts for 17% of the glycerol ingested.

650

Ftuid Intake, Glycerol and Urine Volume

Vol. 57, No.7, 195

TABLE

III.

Hematological and Plasma Osmolality Data, Series

Ilemoglobin

I (t
SE)

Hematocrit

(tSE)
Dav

(tSE)
Dav 2

Plasma Osmolality

Dav

(g'
Dav

100m1-1)

(milliosmoles/kg) Dav Day 2

No Glycerol
1200

45.3 (0.e)

,14.8 (0.9)

15.9 (0.2) 16.0 (0.2)

15.7 (0.3)

299.2 (3.7) 300.6 (3.1) 298.4 Q.3)

298.2 (3.6) 301.4 (3.5) 301.2 Q.0)

1600 h

4s.2 (0.7)
44.7 (0.7)

43.8 (1.2) 43.5 (0.6)

rs.3 (0.4)

2000 h

r5.7 (0.2)
Glycerol

ts.2 (0.2)

1200

43.4 (0.8) 43.3 (0.6) 43.5 (0.7)

43.0 (0.3)
43.2 (O.4) 43.2 (0.4)

ls.0 (0.2)
15.4 (0.3) 15.5 (0.3)

15.4 (0.3) 15.2 (0.3) 15.5 (0.3)

303.4 (1.9)

307.4 (1.7) 30/..4 (2.6) 2e9.8 (3.2)

1600 h

3M.2 (t.s) 3m.6 (2.s)

2m0 h

TABLE IV. Serum Glycerol from Series I, Experimental Protocol

Serum Glycerol (mg/dl)
1200

Day
1600

2000 h 27 4 23 20
25

1200

Day 2
1600

2000 h 24

Subject

2
3

4 5

6
7

Mean
SE

103 7r 101 36 t20 45 81552 130 76 97 59 132 57 109 57 755

l7

r23 68 84388 72306 60312 126 67 47 75 93 51 1474

4 20 24
13

Series

II

Glycerol ingestion resulted in a significant decrease in urine volume from 4 to 24 h (p < 0.01) and from 28 to 32 h (p < 0.02) as determined by linear regression analysis (R values 0.90 and 0.99) (Fig. 2). Folluving initial intake, elevation of serum glycerol values declined to near baseline values after 24 h (Ihble V). The automated chemical analyses values were not changed by the glycerol ingestion and all values were within normal ranges.

Vol.57, No.7, 1995

Fluid Intake, Glycerol and Urine Volume

651

Dfl a{

GLYCEROL NO GLYCEROL

N=6

L!

z
f J

cc @ U)
IU uJ

z
o :) J
tL
1

Fig.2
Fluid retention, assuming insensible water loss was the same with and without glycerol, Series II.

Discussion

The difference between the urine volumes with and without glycerol ingestion remained near 700 ml in both Series I and Series II, even when serum glycerol values were very low. The mean serum glycerol values were 17 mgldlby 2000 h in Series I and were near zero in both Series I and II 24 h after the initial ingestion. The retention of water during the hours when serum glycerol values were low suggests the time required to go from hyperhydration to euhydration is similar to the time required to go from dehydration to euhydration. Although maintainence of fluid balance is very vital, the mechanisms involved are not necessarily rapid (3). Over fifty years ago, Dill reported 12 to 18 h being required for subjects to reestablish euhydration after becoming dehydrated (2). Thus, it should not be surprising that a water load (700 ml in this study) distributed evenly throughout

652

Fluid Intale, Glycerol and Urine Volume

Vol.57, No.7, 195

TABLE V. Serum Glycerol from Series II, Experimental Protocol

Serum Glycerol (mg/dl)

Day
0730 h
Subject
1

1230

1430 h
69 116

Day 2 0730 h t230 h

2
3

4
5

6 Mean SE

0.2 0 0 0 0 0 0

43 128 r28 85 lr4 93 98 t4

r.4
0.4
0 0

111 115

76
109

t29
106

93 94 93 39

2.r
0.1

t2r
63 108 10

0.7 0.4

t*e 12 to 18 h to be lost. Future applications of GIH should identify the mechanisms by which glycerol ingestion modihes fluid balance. Our original hypothesis was that the high aqueous and lipid solubility of glycerol combined with the osmotic action of glycerol would cause expansion of the intra-cellular space. Recent research reports that glycerol induced hyperhydration involves proportional increases in all fluid compartments (11).
body tissues should
The osmolality data in the present study is difficult to explain, as all values are similarly high. In previous studies with the serum glycerol near 100 mg/dl resulted in 5 milliosmoles/kg increase in osmolality and fluid intake without glycerol reduced plasma osmolality (9). Having conducted all osmolality data analyses on the same day may indicate that an error in instrument calibration and/or in sample storage was a factor in causing the high osmolality data. The reliability of the subjects and the assumption that sweat and insensible water loss was similar in the experimental and control periods needs to be considered. The subjects were reimbursed for participating in these studies and may have guessed when they were drinking the glycerol solution. Nevertheless, it is difhcult to imagine that the subjects would have been able to vary their activities and volume of urine to result in fluid intake less urine volume, to be similar in Series I and II. After objections to the flavor of the orange juice solution in Series I, we switched to a strawberry-flavored solution in Series II. The serum glycerol data for Series I and II are similar, suggesting the change in flavor of the solution did not affect the absorption of glycerol. One subject in Series II complained of slight nausea after drinking both the placebo and the glycerol solution (5 ml\kg) at 0730 h. There were no other complaints of discomfort from ingesting either the large volume of water or the glycerol solution.

The data presented support the hypothesis that GIH can be extended for many hours. The best method, however, for maintaining water balance prior to or during situations which result in negative water balance is very complex. Cunently, ingestion of glycerol continues to receive support as a method for alleviating these situations. Montner et al. (5) demonstrated an increase in endurance cycling following hyperhydration with a glycerol solution. These authors also noted less of an
increase in core temperature and reduced heart rate of subjects cycling in a neutral environmental temperature. A one-percent glycerol solution added to a carbohydrate-electrolyte drink was beneficial in conserving fluids in a 6Gh simulated desert exposure (10). Obviously, the volume, concen-

Vol.57, No.7, 195

Fluid Intake, Glycerol and Urine Volume

653

tration, and timing of glycerol and other fluid intake requires additional exploration and modification to meet the requirements of specific situations. Pre-loading with fluid could be particulady important for situations in which subjects must perform heavy work or be in a hot environment. For example, astronauts during exhavehicular activities, workers in chemical warfare suits, firemen in protective gear, and athletes participating in 2- to 3-h events without breaks and with limi0ed access to fluids could have reduced work capacities if they are unable to pre{oad with glycerol solutions.

Acknowledgements

This study was supported in part by NASA grants NAG 9-453/Basic, NAG 9-401 and NIH grant
5M01-RR00997.

References

1. 2. 3. 4. 5'
6.

w.M. BoRTz,P.PAIJL, A.c. HAF4

(1982).

and

w.L. HOLMS, J. clin. Invest. 5rt537-1546

D.B. DILL, Iife, He:f, and Alrirrrde- 211p, Harvard Univ. press, cambridge, Mass.
(1938). J.E. GREENLEAR Med. Sci. Sporrs pas1s. )4((\) &5-656 (Igg2).
T.P. LYoNs, M.L. RIEDESEL,L.E. MEULI and T.w. 22 477-483 (1990).

cHIcK, Med. Sci. Sporrs Exerc.

and D.

P. MONTNER, T.

cHIcK, M.L. RIEDESEL, M. TIMMS

srARK, Med. Sci.

Sports Exerc. ?4(.s5) S152 (1992). S.V. MOROFR D.E. BASS, J. Appl. physiot.2n267-270 (1965).

7.
8.

H.M. NEISLER, WG. LOIZ, K.D. JOHNSON, L.G. MEYER and D.J. HORRIGAN,
J. FASEB 6(4) A1198 (1992).

9.

R. PELKONEN, E.Z. NIKKILA, and M. KEKKI, Diabetologia, 3 1_g (1967). M.L. RIEDESEL, D.Y. ALLEN, c.T. PEAKE and K. AL-eATTAN, J. Appl. physiol.
632262-2268 (1987).

10. 11. 12, 13.

M.N. SAIVKA, Med. Sci. Sports Exerc.24 657-670 (1992). M.N. SAWKA, Md. Sci. Sports Exerc. 25S S35 (1993). S.L. WALD and R.L. MCLAURIN, J. Neur. Surg. 56 323-331(19g2). B. WTNKLER, R. srEELE, and N. ALTSZULER, Am. J. physiol. )16 tgt-tg6 (1969).