2. oxygen supply and removal of metabolism byproducts) may METHODS
override cardiovascular potential (29). The heart rate
reaches a maximal level and further decreases in stroke
volume translate into a decreased cardiac output, compro- Eight endurance-trained nonacclimated male cyclists
mised heat transfer at the skin, and reduced oxygen transport (19 –29 yr) were recruited from the University of Connect-
to the muscle. This physiological phenomenon is exagger- icut cycling club and the local cycling community. Screen-
ated when an athlete in a dehydrated state exercises in the ing information was obtained to ensure that subjects met the
heat (30). Ultimately, the cardiovascular system prioritizes following criteria: 1) no chronic health problems, 2) no
maintenance of blood pressure over thermoregulation or previous history of heat illness, and 3) no history of cardio-
muscle function (29). Consequently, the thermoregulatory vascular, metabolic, or respiratory disease. The physical
system will lose efficiency and sweat rate will decrease characteristics of the subjects were: age, 24 1 yr; height,
while core temperature rises (32). The ensuing hyperthermia 181 2 cm; weight, 70.1 1.0 kg; body fat, 13.5 0.6%;
and peak oxygen uptake, 61.4 0.8 mL kg 1 min 1. The
can lead to physiological impairment and the possibility of
subjects trained regularly and were actively competing in
a serious heat illness (4,36). The decrease in fluid volume
mountain or road cycling races.
associated with dehydration and the subsequent cardiovas-
Subjects attended a briefing meeting before any experi-
cular and thermoregulatory consequences will cause a dec-
mentation to ensure an understanding of the testing param-
rement in endurance performance at as little dehydration as eters and the benefits/risks of the study and signed an
1% of body weight (3,9,35). informed consent statement. All subjects completed a med-
Previous investigations using intravenous rehydration are ical history questionnaire. The protocol was approved by the
not compatible with the purpose of the present investigation University of Connecticut Institutional Review Board for
for the following reasons: a) intravenous infusion was ad- Studies Involving Human Subjects. Subjects were paid for
ministered during exercise, which may not be realistic for an their participation in the study.
athletic population (11,16,23); b) the amount of fluid, or
carbohydrate content, was not consistent (16); c) an ex- Baseline Testing
tended period of time elapsed between rehydration and
Subjects performed a peak oxygen uptake test and body
exercise (7,27); d) comparisons to oral rehydration were not
composition test (underwater weighing) before the experi-
made (11,23); and e) the ensuing exercise stress was not
mental trials. Percent body fat and lean body mass were
performed in a hot environment (16,23). These factors make
calculated according to Brozek (5). The peak oxygen uptake
assumptions regarding the efficacy of intravenous rehydra-
test involved a progressively graded cycle ergometer pro-
tion difficult when considering the parameters of the present
tocol conducted in a comfortable environment (27°C). Sub-
study. Previous studies conducted by our group (7,8,27) jects exercised until they reached volitional exhaustion,
were the most relevant to the present investigation. They achieved a plateau in oxygen uptake, and an RER 1.15.
reported a decreased cardiovascular strain during exercise ˙
The VO2peak tests were 10.01 0.96 min in duration.
subsequent to intravenous rehydration, lower rating of per-
ceived exertion during exercise after oral rehydration, and Experimental Protocol
no differences in thermoregulatory function between intra-
venous and oral trials. However, these authors employed a Food diaries were kept for 3 d before each experimental
45-min rehydration, followed by a 75-min rest, before the trial and analyzed for total calories, carbohydrate, protein,
onset of exercise. fat, sodium, and potassium (Nutritionist IV, First DataBank,
San Bruno, CA). The subjects were provided with detailed
Thus, the purpose of the present investigation was to
instructions on the recording of food and fluid intake. Anal-
examine further the efficacy of intravenous rehydration
ysis revealed no significant differences (P 0.05) among
when compared with the same fluid being orally ingested.
the three trials in total calories, carbohydrate, protein, fat,
Specifically, we asked the question: Does a brief period of
sodium, and potassium consumption.
intravenous rehydration, compared with oral rehydration,
Before each of the three experimental trials subjects were
enhance physiological responses and performance during a provided a preparation checklist and an investigator con-
subsequent maximal exercise session in athletes who were ducted a brief meeting before each trial to remind the subject
4% dehydrated before the rehydration period? A 0.45% of the proper procedures to follow. Each experimental trial
NaCl fluid was chosen because of its widespread usage by consisted of five discrete components. These included, eu-
the medical community in medical emergencies. A 0.45% hydration baseline (EUHYD), dehydration (DHY), dehydra-
NaCl solution also was ingested during the oral rehydration tion baseline (DHYBASE), rehydration (RHY), and the
trial to control for the composition of the rehydration fluid. exercise test (EXER).
We hypothesized that intravenous rehydration during a 20- Euhydration baseline (EUHYD). On the morning be-
min period, as compared with oral ingestion, would be fore each experimental trial subjects reported to the Human
advantageous to cardiorespiratory, thermoregulatory, and Performance Laboratory between 1000 and 1200 h in an
performance variables during a subsequent strenuous exer- euhydrated state. Hydration status was checked via urine
cise session. specific gravity (USG) and a reading of 1.020 precluded
INTRAVENOUS VERSUS ORAL REHYDRATION Medicine & Science in Sports & Exercise 125
3. TABLE 1. Body weight changes during the dehydration process and quantity of fluid provided during the rehydration process.
Fluid During Dehydration
Pre Body Post Body Weight Rehydration at Onset of
Trial Weight (kg) Weight (kg) Loss (%) (ml) Exercise
Control 71.3 1.3 68.5 1.4 3.9 0.3 0† 3.9 0.3†
Drink 71.3 1.3 68.4 1.2 4.0 0.3 1405 112 2.0 0.3
IV 71.3 1.4 68.6 1.2 3.8 0.3 1341 116 1.9 0.3
Values are mean SEM.
† significantly different (P 0.05) from Drink and IV trial.
data collection on this day. The subjects then sat in the ing DRINK subjects consumed an oral solution consisting
environmental chamber (36°C, Model 2000, Minus-Eleven, of 4 g of a commercial flavoring that was sodium-, sugar-,
Inc., Malden, MA) for 20 min. After a 20-min equilibration and calorie-free (Kool-Aid, Kraft General Foods, White
period, blood was drawn from a superficial forearm vein and Plains, NY) dissolved in 0.45% NaCl (Baxter Healthcare
analyzed for hematocrit (Hct), hemoglobin (Hb), plasma Corp., Deerfield, IL). This solution contained 77 mEq/L
osmolality (OSMp), and total plasma protein (TPP). The Na with an osmolality of 154 mOsm L 1, and was served
subjects were then weighed (Model 700 M, SR Instruments, chilled (10°C) to enhance palatability. Servings were ad-
Tonawanda, NY) and asked to provided a urine sample for ministered in equal amounts every 2 min over the 20-min
USG assessment. The body weight recorded during EUHYD RHY period. During IV an intravenous infusion of 0.45%
was used as the starting point for the dehydration procedure. NaCl (22°C) was administered at a 67.1 mL min 1 flow rate
Dehydration (DHY). Following EUHYD the subjects over the 20-min RHY period. The oral solution and the
were instructed not to consume any fluids until the follow- intravenous solutions were calorie free and identical except
ing day. In addition, they were instructed to consume foods for the calorie-free flavoring in the oral solution.
that were low in fluid content during their lunch and dinner. The amount of fluid given in the DRINK and IV trials
The exercise portion of the DHY was conducted during the was equal to 50% of the body weight lost between the
afternoon or evening. The subjects were instructed to cycle HYDBASE and DHYBASE. Subjects received 2% of their
at a moderate intensity for 2 h. The combination of fluid body weight during the RHY and began the exercise session
restriction and exercise was effective in assuring a body at 2% body weight. Upon completion of the 20-min RHY,
weight loss of 4%. another blood sample was drawn and analyzed for the same
Dehydration baseline (DHYBASE). Subjects, 12 h variables analyzed post DHY.
postprandial, arrived at the Human Performance Laboratory
Exercise test (EXER). Following the RHY procedure,
between 0700 and 0830 h in a dehydrated state. Upon
subjects mounted a cycle ergometer (Model 818E, Monark,
arrival, subjects were weighed, positioned a rectal ther-
Sweden). At this time, resting heart rate (HR), oxygen
mistor and a Teflon (Dupont, Wilmington, DE) catheter was ˙
uptake (VO2), respiratory exchange ratio (R), respiratory
placed in an antecubital vein. The body weight was used to ˙
rate (RR), minute ventilation (VE), carbon dioxide produc-
determine the percent of body weight lost because of the ˙ CO2), mean weighted skin (Tsk) and rectal (Tre)
DHY procedure. If they were less than 3.5% or greater than
temperatures, skin blood flow (SkBF), and blood pressure
4.5%, the trial was terminated for that day (see Table 1 for
final averages). The subjects then sat in a wheelchair for 20 (BP) measures were obtained. The subjects performed the
min in the environmental chamber at 36.6°C and at the test at 74% of their peak oxygen consumption. Seat height
completion of this equilibration period, a blood sample was was the same in each trial, and toe clips were used. Air flow
drawn and analyzed for Hct, Hb, OSMp, TPP, plasma lactate (2.3 m s 1) was generated by a fan and directed at the
(LA), plasma glucose (GLU), and plasma sodium (Na p) subject. All timekeeping devices (e.g., clocks, stop watches)
and a urine sample was collected for USG. were kept out of sight of the subjects during the exercise
Subjects were then wheeled outside the environmental test. The exercise session was terminated if the subject
chamber and sat quietly for an additional 110 min in 22°C. stopped at his/her own volitional exhaustion, had a rectal
The 110 min was necessary to satisfy the requirements of temperature 39.5°C, exhibited signs of heat intolerance,
another study that was occurring in conjunction with the or could not maintain a RPM approximately that of the
present study. No manipulation or physiological measure- starting cadence. One of the 24 trials was stopped because
ments were made during this time. of the rectal temperature reading and all others stopped
Rehydration (RHY). Following the 110 min, subjects because of volitional exhaustion.
were then transported in the wheelchair back into the envi- HR was recorded every 2 min, and Tsk and Tre every 4
ronmental chamber and began the RHY procedure. One of min. Plasma LA, GLU, TPP, OSMp, Na p, Hct, and Hb,
the three randomly ordered experimental trials was initiated, stroke volume (SV), cardiac output (Q), BP, SkBF, RPE,
which included: 1) control (CONTROL), 2) drink (DRINK), ˙ ˙ ˙
VO2, VCO2, R, RR, and VE were measured at 5, 15, 30, and
and 3) intravenous infusion (IV). 45 min. Immediately following exercise, another blood sam-
During CONTROL subjects were not given any fluid ple was taken and analyzed for plasma LA, GLU, TPP,
during the 20-min rehydration period. In this trial the sub- OSMp, Na p, Hct, and Hb. This post exercise (POST-EX)
jects began the exercise session at 4% body weight. Dur- body weight was used in sweat rate calculations. Many
126 Official Journal of the American College of Sports Medicine http://www.msse.org
4. psychological parameters were examined but are described
Thermoregulatory measures. Tre was measured us-
ing a rectal thermistor (Model 401, Yellow Springs Instru-
ments, Yellow Springs, OH) inserted 10 cm past the anal
sphincter. Tsk was measured using thermocouples (Model
409, Yellow Springs Instruments) placed on the chest, tri-
ceps, thigh, and calf and was calculated using four sites
according to Ramanathan (26). Sweat rate (L h 1) was de-
termined from DHYBASE and POSTEXER body weights
corrected for respiratory losses and blood draw volume.
SkBF was measured via a laser Doppler flow meter (Per-
imed, Inc., model Periflux PF2B, Stockholm, Sweden). The
sensor was placed and secured on the anterior forearm
midway between the elbow and wrist. Measurements were
taken for 2 min and the arm was held in a consistent, stable
position by the same investigator. Cutaneous vascular con-
ductance (CVC) was calculated using the SkBF divided by
the mean arterial pressure (MAP).
Cardiovascular measures. HR was measured using a Figure 1—Exercise time to exhaustion (mean SEM). † signifi-
cardiotachometer (Polar Electro, Port Washington, NY) fit- cantly different (P < 0.05) from DRINK and IV groups. DRINK vs IV
˙ ˙ ˙
ted to the chest. VO2, VCO2, R, RR, and VE was measured (P 0.07)
using on-line, breath by breath, open circuit spirometry
(Model CPX-D, Medical Graphics, St. Paul, MN). Q was
measured using a CO2 rebreathing technique (10). The HR Newman-Keuls post-hoc test. The level of significance was
recorded just before the onset of the rebreathing technique chosen as P 0.05. All data are presented as mean SEM.
was used for calculations of SV. BP was attained via an
aneroid sphygmomanometer over the brachial artery using a
sphygmomanometer. MAP was calculated using the equa-
tion by Robinson et al. (28). Values for predehydration body weight, postdehydration
Blood and urine measures. USG and TPP were mea- body weight, percent weight loss, fluid infused or ingested
sured by refractometry (Model A300CL, Spartan, Japan). during rehydration, and the level of dehydration at the onset
Na p was measured in duplicate via ion sensitive electrodes of exercise are presented in Table 1.
(Model 984-S, AVL Scientific Corp., Roswell, GA). Hct Time to exhaustion (Fig. 1) for the DRINK (34.86 4.01
was determined, in triplicate, from whole blood by the min) and IV (29.48 3.50 min) trials was greater (P
microcapillary technique. Hb was measured, in triplicate, by 0.05) than the CONTROL (18.95 2.73 min) trial. A trend
the cyanmetHb method (Kit 525, Sigma Chemical, St. (P 0.07) was found between the DRINK and IV trials
Louis, MO) and a spectrophotometer (Bausch & Lomb (Fig. 1).
Spectronic 88, Rochester, NY). Percent change in plasma Plasma Responses. Plasma lactate during DRINK
volume (%‚PV) was calculated using the equation of Dill was lower (P 0.05) than during CONTROL and IV at
and Costill (12). Baseline plasma volume was the DHY minute 15 and postexercise. CONTROL values were greater
values. OSMp was measured in duplicate via freezing point (P 0.05) than IV values at minute 15 (Table 2).
depression (Model 3DII, Advanced Instruments, Needham Plasma glucose during DRINK was lower (P 0.05)
Heights, MA). Plasma GLU and LA was determined, in than values for CONTROL and IV at minute15 and postex-
triplicate, via enzymatic techniques (Model 2003, Glucose/ ercise (Table 2).
Lactate Analyzer, Yellow Springs Instruments). TPP IV pre-exercise and minute 5 were less (P 0.05)
than the corresponding values for CONTROL and DRINK.
CONTROL values during minute 15 were greater (P
0.05) than DRINK and IV values at minute 15. CONTROL
Blood, cardiovascular, and thermoregulatory variables values during postexercise were greater (P 0.05) than IV
were analyzed using a two-way (condition time) analysis values during postexercise (Fig. 2).
of variance (ANOVA) with repeated measures. Other mea- OSMp was similar (P 0.05) at all time points between
sures (e.g., time to exhaustion) utilized a one-way ANOVA trials. IV values during pre-exercise were different (P
(condition). The three conditions were CONTROL, 0.05) than all preceding and ensuing IV values. CONTROL
DRINK, and IV. Significant F-ratios were analyzed using a and DRINK values during pre-exercise were less (P 0.05)
INTRAVENOUS VERSUS ORAL REHYDRATION Medicine & Science in Sports & Exercise 127
5. TABLE 2. Selected measurements pre-exercise, and 5, 15, and 30 min of exercise.
30 Min (POST-EX for
Variable Trial Pre-exercise 5 Min 15 Min Lactate & Glucose)
VO2 (mL min 1)
Control 456.9 30.0 3164.9 120.4 3125.6 120.1 —
Drink 461.9 34.8 3157.6 101.5 3152.3 107.2 2909.5 98.8 (N 6)
IV 451.4 42.0 3272.6 97.9 3248.1 90.0 3023.4 111.5 (N 5)
VCO2 (mL min 1)
Control 363.9 30.2 3098.6 159.6 3116.6 137.0 —
Drink 376.8 23.8 3086.8 139.4 3037.4 136.0 2718.0 137.6 (N 6)
IV 349.5 42.0 3166.5 144.5 3114.8 104.5 2888.8 128.8 (N 5)
Control 0.79 .04 0.98 .02 1.00 .02 —
Drink 0.83 .04 0.98 .02 0.96 .01 0.94 .02 (N 6)
IV 0.77 .04 0.97 .03 0.96 .01 0.95 .02 (N 5)
VE (L min 1)
Control 14.2 1.2 91.4 5.3 108.0 6.5 —
Drink 14.1 1.2 86.8 4.9 93.8 5.0* 97.0 6.9 (N 6)
IV 13.7 1.5 92.9 5.2 101.4 4.8† 106.1 5.6 (N 5)
Respiratory rate (beats min 1)
Control 17.5 1.9 41.6 3.5 52.7 4.5 —
Drink 16.0 2.1 36.1 2.5* 41.3 3.1* 49.0 5.2 (N 6)
IV 17.5 2.1 41.8 4.1 47.1 3.9† 50.0 5.1 (N 5)
Lactate (mmol L 1)
Control 1.0 .00 5.1 1.0 8.8 2.0 7.9 1.7
Drink 1.1 .10 4.6 .90 6.0 1.3* 5.8 1.1*
IV 0.9 .10 5.3 1.1 7.1 1.6† 7.4 1.5
Glucose (mmol L 1)
Control 4.8 .13 4.9 .16 6.4 .36 8.0 .49
Drink 4.7 .12 4.9 .21 5.4 .40* 6.7 .74*
IV 4.7 .14 5.0 .17 6.1 .37 7.7 .59
Values are means SEM. N, 7 for CONTROL at 15 min. N, 8 for all other measurements (except where noted at 30 minutes). * DRINK is significantly (P 0.05) less than CONTROL
and IV. † IV is significantly (P 0.05) less than CONTROL.
than the within values at minute 5, minute 15, and postex- together an overall significant (P 0.05) decrease was
ercise (Fig. 2). found for stroke volume from minute 5 to minute 15.
Na IV values during pre-exercise, minute 5, minute 15, Thermoregulation. Rectal temperature values for the
and postexercise were less (P 0.05) than the correspond- CONTROL trial were greater (P 0.05) than the DRINK
ing values for CONTROL and DRINK. CONTROL values and IV values from minute 0 through minute 12. DRINK
at minute 15 were greater (P 0.05) than the DRINK values were less (P 0.05) than IV values from minute 0
values at minute 15 (Fig. 2). though minute 24 (Fig. 5).
Plasma volume for IV after rehydration and minute 5 Tsk values for DRINK were less (P 0.05) than the
were significantly (P 0.05) different than the correspond- CONTROL values from minute 0 through minute 12 and
ing CONTROL and DRINK values. IV values at minute 15 were less (P 0.05) than IV from minute 4 through minute
and postexercise were significantly (P 0.05) different 12 (Fig. 6).
than the corresponding CONTROL values. CONTROL val- CVC and SkBF was similar among all trials. For CVC a
ues at minute 5 were significantly (P 0.05) different than significant (P 0.05) increase from resting was noted at
the CONTROL values at minute 15 (Fig. 2). minute 5 in the CONTROL trial, while in the DRINK trial
Cardiorespiratory. Heart rate values for the CON- it was noted at 15. For SkBF a significant (P 0.05)
TROL trial were greater (P 0.05) than values for DRINK increase from resting was noted at minute 5 in the CON-
and IV trials from 0 through 8 min and greater (P 0.05) than TROL trial, while in the DRINK and IV trials it was noted
the DRINK values only from 10 through 14 min (Fig. 3). at 15 min (Fig. 6).
Oxygen uptake, carbon dioxide production, and respira- Sweat rate (L h 1 of exercise) for the CONTROL (2.9
tory exchange ratio were similar among all trials (Table 2). 0.2) trial was greater (P 0.05) than the DRINK (2.4
Respiratory rate values for the DRINK trial at minute 5 0.2) and IV (2.4 0.2) trials.
and minute 15 were less (P 0.05) than the CONTROL and
IV values at minute 5 and minute 15. CONTROL values at
minute 15 were greater (P 0.05) than the IV values at DISCUSSION
minute 15 (Table 2).
Expired ventilation values for the DRINK trial at minute
15 were less (P 0.05) than the CONTROL and IV values The most noteworthy finding of this project was the lack
at minute 15. CONTROL values at minute 15 were greater of a significant difference in exercise time to exhaustion
(P 0.05) than the IV values at minute 15 (Table 2). when comparing IV and oral rehydration, thus our original
Stroke volume and cardiac output (Figs. 3 and 4) were hypothesis was not supported. Of the past studies examining
similar among all trials. When all trials were considered intravenous infusion, oral comparison groups were not
128 Official Journal of the American College of Sports Medicine http://www.msse.org
6. that some yet untested combination of oral and intravenous
rehydration is even more effective than either method alone.
The small but significant differences in physiological func-
tion between oral and intravenous rehydration may shed
light on why the nonsignificant differences in performance
occurred. An exploration of some of the psychological
causes has been described elsewhere (19).
The present study supports previous research that has
shown endurance performance decrements with increasing
degrees of dehydration (25,32). During CONTROL, sub-
jects were 4% dehydrated at the onset of exercise, while in
the two rehydration trials they were 2% dehydrated at the
onset of exercise, and the decrement in performance was
likely a result of this difference in body water loss. When
partially rehydrated, our subjects exercised significantly
longer, 13.2 min (about 40%) on average, than during CON-
TROL. This assessment appears quite practical since ath-
letes who compete in the heat are often dehydrated to some
Many indicators of cardiorespiratory function showed
decreased strain during the DRINK as compared with that
during IV. The significantly lower VE and respiratory rate
coupled with the notable, although not significant, differ-
Figure 2—Plasma Responses (mean SEM). N 7 for CONTROL at
15 min, N 8 for all other values. † IV significantly different (P <
0.05) than DRINK and CONTROL. * CONTROL significantly differ-
ent (P < 0.05) than DRINK. ‡ IV significantly different (P < 0.05) than
CONTROL. § CONTROL significantly different (P < 0.05) than
DRINK and IV.
included (11,23), subjects were rehydrated during exercise
(11,16,23), and subjects did not exercise to exhaustion
(7,16). Further, sometimes the rehydration fluids were not
the same concentration (i.e., some contained glucose or
were of different volumes) (16,21), and sometimes the ex-
ercise sessions did not occur in the heat (16,23), which
limits the circulatory strain. Because of these differences in
methodology, comparison with past studies is difficult when
considering performance parameters. In the present study,
there were no significant performance differences between
intravenous and oral rehydration in subjects previously de-
hydrated to 4% of body weight, who were then rehydrated
to 2% body weight and subsequently exercised in the heat
to exhaustion at 74%VO2peak.
Conversely, an intriguing finding was that subjects did, in
fact, avoid exhaustion for an additional 5 min (or 14% more)
in DRINK as compared with that in the IV trial. The lack of
significance (P 0.07) in this finding limits the strength of
any scientific argument when discussing these differences,
but an effect size of 0.51 indicates a trend of moderate
strength. A 14% difference in performance coupled with the Figure 3—Cardiovascular Responses (mean SEM). N 7 for CON-
physiological differences noted in the ensuing sections may TROL at 10 –15 min, N 7 for DRINK at 16 –26 min, and N 6 for
minute 28 –30. N 6 for IV at 20 –22 min and N 5 for minute 24 –28.
encourage highly trained athletes to rehydrate via an oral N 8 for all other values. † CONTROL higher (P < 0.05) than DRINK
route instead of intravenously. Future research may show and IV. * CONTROL higher (P < 0.05) than DRINK.
INTRAVENOUS VERSUS ORAL REHYDRATION Medicine & Science in Sports & Exercise 129
7. tory rate in the CONTROL trial as compared with both
modes of rehydration indicate that the additional 2% of
dehydration ( 2% vs 4%) was enough to elicit significant
differences in cardiorespiratory function. These differences
likely result from the decrease in venous return, which in
turn negatively influences cooling capacity, stroke volume,
and muscular function (20,29).
Both DRINK and IV experiments demonstrated notewor-
thy differences in body temperatures, and this may have
contributed to the increased cardiorespiratory (29) and met-
abolic strain (31) during intravenous rehydration as com-
pared with that in oral rehydration. Exercise during DRINK
Figure 4 —Individual Stroke Volume Responses (mean SEM). N
8 for DRINK. N 7 for IV.
elicited significantly decreased rectal temperature and skin
temperature response as compared with that during IV. The
response in DRINK was the result of either a decrease in
ences in heart rate and stroke volume at 15 min of exercise
heat accumulation or an enhanced ability to dissipate heat.
may indicate that an oral solution of equal volume and
Given an equivalent exercise intensity and environmental
concentration may have better maintained cardiovascular
strain in both trials, it is likely the differences resulted from
capacity during intense exercise than during IV. For exam-
differences in heat dissipation. These results are consistent
ple, heart rates were 3– 6 beats min 1 lower from minutes
with Montain and Coyle (21) who reported lower core
10 –28 of exercise in DRINK compared with those in IV. In
temperatures during exercise with oral as compared with
addition, the improved maintenance of cardiovascular function
intravenous rehydration. Since the concentration of the two
is best exemplified by stroke volume during DRINK which
rehydration fluids used in that study were not identical, it
was 12.1 mL higher at minute 15 of exercise compared with
leaves open to question whether the differences resulted
that in IV. To complement this finding, a 6 beats min 1 higher
heart rate was noted in IV (vs DRINK) at 15 min. This provides
evidence that, during IV, the cardiovascular system compen-
sated for the decrease in stroke volume by increasing heart rate
in an attempt to maintain cardiac output. Although differences
in heart rate and stroke volume between modes of rehydration
were not significant, they are theoretically consistent and rel-
evant for elite athletes.
These findings are in contrast to Castellani et al. (7,8) and
Hamilton et al. (16) who reported a diminished cardiovas-
cular strain (lower heart rate) during exercise, following
intravenous rehydration as compared with oral rehydration.
The results of Hamilton et al. are confounded by the fact that
the intravenous fluid contained glucose while the oral so-
lution did not. The rehydration fluid used by Castellani et al.
(7,8) was the same used in the present study (0.45% NaCl)
which further belies the differences between the two studies.
One possible explanation involves the passing of 75 min
between rehydration and the onset of exercise, while only a
few minutes passed in the present study. A more likely
explanation for these between-study differences was the
influence of exercise intensity on thermoregulatory strain.
The present study found an exaggerated thermoregulatory
strain in the intravenous trial, likely indicating differences in
heat dissipation between the two modes of rehydration.
Because of similarity of other variables (i.e., plasma vol-
ume, osmolality, hormone levels), the differences in rectal
Figure 5—Thermoregulatory Responses (mean SEM). N 7 for
temperature provide a more likely cause for the cardiovas- CONTROL at minute 12 and N 3 for CONTROL at minute 16. N
cular differences noted in the present study (29). 7 for DRINK at minutes 16 –24 and N 6 for DRINK at minute 28.
The present study provided further evidence that increas- N 6 for IV at minutes 20 –24 and N 5 for IV minute 28. N 8 for
all other values. † CONTROL higher (P < 0.05) than DRINK and IV.
ing levels of dehydration increase cardiorespiratory strain * DRINK lower (P < 0.05) than IV. ‡ DRINK lower (P < 0.05) than
(7,32). The significantly higher heart rate, VE, and respira- CONTROL. § DRINK lower (P < 0.05)than CONTROL and IV.
130 Official Journal of the American College of Sports Medicine http://www.msse.org
8. production and osmotic influences (24,34), and/or d)a gas-
trointestinal mechanism, such as a neural, hormonal, and/or
peptide response (14). All of these should be expected
physiologically, because it is natural to receive fluids
through the mouth as compared with through a vein in the
arm. When one is considering the temperature of the rehy-
dration fluids, the specific heat calculations predicted a
0.28°C lower rectal temperature pre-exercise in DRINK as
compared with that in IV. The specific heat calculations
reflect the amount of influence the temperatures of the fluid
have on changes in body temperature. This may explain the
0.3°C lower rectal temperature reported pre-exercise in the
present study. Additionally, the lower baseline is a likely
reason for the small but consistently lower rectal tempera-
tures during exercise. Regardless, oral rehydration limited
the accumulation of heat (as evidenced by a decreased rectal
temperature) and enhanced the dissipation of heat (because
of an improved temperature gradient because drink trial had
lower skin temperatures). The lower rectal temperatures and
skin temperatures may have directly mediated the noted dif-
ferences in cardiorespiratory (respiratory rate, VE, heart rate,
stroke volume) and metabolic (lactate) responses because de-
creased thermoregulatory strain enhances the maintenance of
venous return and muscle dynamics (18,29,30,31,36). Consid-
Figure 6 —Skin Blood Flow Responses (mean SEM). N 7 for ering the rectal and skin temperature findings, it was surprising
DRINK at 0, 5, and 15 and N 5 for 30. N 7 for CONTROL at 15. the skin blood flow responses were not significantly different
N 3 for IV 30. N 8 for all other values. † CONTROL 5 is greater
(P < 0.05) than CONTROL 0. * DRINK 15 is greater (P < 0.05) than between modes of rehydration at minute 5 and minute 15,
DRINK 0. ‡ DRINK and IV 15 is greater (P < 0.05) than DRINK and although it is still possible that there was an earlier onset of
IV 0, respectively. sweating and dilation of the peripheral vasculature in the in-
travenous trial. In fact, with the higher rectal temperature noted
from the route of rehydration or the differential composition in IV, one would expect an earlier onset of events associated
of the rehydration fluid. with cooling (30).
In contrast, Castellani et al. (7) found no differences in
thermoregulatory responses during exercise following oral
and intravenous rehydration. Again, the contrasting results
of the present study and those of Castellani et al. were likely The onset of hypovolemic-hyperosmolality during in-
caused by differences in the intensity of the activity and the tense exercise in the heat has often been postulated as a
period of time between the end of rehydration and the onset major cause of the physiological responses that help the
of exercise. The higher intensity of exercise in the present body cope with heat stress (22,33). No statistical differences
study exacerbated the thermoregulatory strain, and the time were found in osmolality when comparing the two modes of
permitted to respond to the intense exercise in a hot envi- rehydration. The use of identical fluids during both modes
ronment was decreased. Also, the delivery of fluid via oral of rehydration helped control this variable. But, even though
rehydration helped to maintain a lower rectal temperature. no significance was noted, the consistent trend was a lower
Conversely, although IV delivery more rapidly restored IV osmolality, with the difference reaching 5.5 mOsm pre-
plasma volume as compared with drink, it may not have exercise. Thus, the degree of hyperosmolality may have
influenced the temperature of the body core as dramatically. been a cause for physiological differences between the two
The evidence for this stems from a lower rectal temperature modes of rehydration. Sodium, the principal determinant of
following the rehydration period even before the onset of extracellular fluid osmolality (33), was significantly lower
exercise. This was somewhat surprising since minimal time throughout exercise in the intravenous trial. Not surprising,
expired to allow for complete intestinal absorption. total plasma protein, another major factor in body fluid
The lower pre-exercise rectal temperatures coupled with homeostasis (15), was lower in the intravenous trial follow-
the consistently lower values during exercise suggest that ing rehydration and at minute 5 of exercise. Specifically, the
the mode of fluid entry into the body differentially influ- exercise and immediate post exercise osmolality were
ences the body’s capacity to cool itself during intense ex- within 2 mOsm kg 1 for the two modes of rehydration, and
ercise in the heat. Possible reasons for this may arise from the changes within both trials did not fluctuate more than 1
a) the anatomical location of the intestines to the body’s mOsm kg 1 during and immediately following exercise.
core, b) the differences in the temperature of the rehydration Additionally, the plasma volume responses to exercise
fluids, c) an oropharyngeal reflex that mediates hormone between the two modes of rehydration were similar at
INTRAVENOUS VERSUS ORAL REHYDRATION Medicine & Science in Sports & Exercise 131
9. minute 15 and postexercise. The percent change in plasma Plasma glucose was lower in DRINK compared with that
volume values following rehydration and minute 5 showed in IV, at minute 15 and postexercise. This signals a differ-
an enhanced plasma volume restoration for the intravenous ence in metabolic turnover between DRINK and IV. Sur-
trial, but because the measurements were similar by minute prisingly the blood glucose levels were not impacted by
15 of exercise, the differences resulting from the mode of hydration state, but rather by mode of rehydration. This is
rehydration had been negated between minutes 5 and 15. evident in the fact that IV had similar plasma glucose to
The results indicate that the plasma volume changes CONTROL while both were significantly higher than
associated with intense exercise in the heat while dehy- DRINK at minute 15 and postexercise. In fact, the increase
drated was not different (DRINK vs IV) during the sec- in plasma glucose for IV and CONTROL were approxi-
ond half of exercise and at exhaustion. The responses mately 63 and 65%, respectively, as compared with only a
may have been different if the dehydration was “acute” 44% rise in the oral rehydration trial. Since the amount of
instead of “chronic” (from the day before) since the fluid was not a factor, differences in the route of fluid
intracellular fluid stores are likely more influenced when entering the system is a likely contributor to these differ-
chronically dehydrated. ences. The passing of fluid through the typical pathways,
Metabolism. The plasma lactate and plasma glucose oral-pharyngeal-stomach-intestine, may have a role in this
findings indicate that there were metabolic differences when domain by mediating factors (gastrointestinal peptide, insu-
comparing mode of rehydration and the amount of dehy- lin, glucagon, epinephrine) that influence metabolism. Ad-
dration. The higher plasma lactate levels in IV, as compared ditionally, the fact that no differences were found between
with those in DRINK at minute 15 and postexercise, indi- IV and CONTROL in plasma glucose levels, while many
cate an earlier transition to an increased contribution of other physiological differences (e.g., cardiovascular, ther-
anaerobic sources of energy in IV. This could be a result of moregulatory) were noted between these two groups, pro-
differences in lactate removal from the circulation resulting vides further evidence for the route of fluid entry playing a
from factors associated with reduced splanchnic circulation role in the plasma glucose responses. The absence of insulin
(29). Possible explanations for an increase in anaerobic values leaves much room for speculation.
glycolysis in IV include factors that affect the milieu within In conclusion, partial oral rehydration resulted in less
and surrounding the muscle tissue. A decrease in oxygen physiological strain during exercise in the heat compared
perfusing the muscle is a potential cause for these metabolic with partial intravenous rehydration. Contrary to our initial
changes (17). Hypothetically, the findings of increased car- hypotheses, there was no ergogenic benefit gained by intra-
diorespiratory strain in IV as compared with those in venous rehydration. While performance was not signifi-
DRINK, exemplified by differences in expired ventilation cantly different between modes of rehydration, the consis-
and respiratory rate, could have influenced oxygen delivery tently better physiological responses indicate that athletes
to exercising muscles. Another reason for the differences should rehydrate utilizing typical oral routes.
may be the higher rectal temperatures and higher mean
weighted skin temperatures in IV. An increase in rectal The authors thank Timothy Scheett, James Stoppani, Timothy
temperature may potentiate a shunting of some of the lim- Bilodeau, Dean Aresco, David Blair, Sheri Huckleberry, Jason Suth-
erland, Jeffrey Mace, and Melissa Wehnert for their dedicated tech-
ited blood supply from the muscle to the periphery to aid in nical support. We are also extremely grateful to the subjects for their
cooling. A decrease in blood flow to the muscles can cause participation.
a cascade of events, including changes in heat dissipation, This work was supported, in part, by grants from General Nutri-
tion Corporation, and the University of Connecticut Research
lactate removal (affecting the pH of the intracellular envi- Foundation.
ronment), and buffering capacity of the muscle (6,20). Address for correspondence: Douglas J. Casa, Ph.D., ATC, Di-
These factors can combine to impede the normal recovery rector, Athletic Training Education, Department of Kinesiology,
Neag School of Education, University of Connecticut, 2095 Hillside
process between contractions, ultimately causing changes in Rd., Box U-110, Storrs, CT 06269-1110. E-mail: dcasa@uconnvm.
the metabolic capabilities. uconn.edu.
1. AMERICAN COLLEGE OF SPORTS MEDICINE. Position Stand: Exer- 6. BUSKIRK, E. R. and S. M. PUHL. Effects of acute body weight loss
cise and Fluid Replacement. Med. Sci. Sports Exerc. 28:i-vii, in weight-controlling athletes. In: Body Fluid Balance Exercise
1996. and Sport. E. R. Buskirk and S. M. Puhl (Eds.). New York: CRC
2. AMERICAN COLLEGE OF SPORTS MEDICINE. Position Stand: Heat and Press, 1996, pp. 283–296.
Cold Illnesses During Distance Running. Med. Sci. Sports Exerc. 7. CASTELLANI, J. W., C. M. MARESH, L. E. ARMSTRONG, et al. Intra-
28:i-x, 1996. venous versus oral rehydration: effects on subsequent exercise
3. ARMSTRONG, L. E., D. L. COSTILL, and W. J. FINK. Influence of heat stress. J. Appl. Physiol. 82:799 – 806, 1997.
diuretic-induced dehydration on competitive running perfor- 8. CASTELLANI, J. W., C. M. MARESH, L. E. ARMSTRONG, et al. Endo-
mance. Med. Sci. Sports Exerc. 17:456 – 461, 1985. crine responses during exercise-heat stress: effects of prior iso-
4. ARMSTRONG, L. E. and C. M. MARESH. The exertional heat ill- tonic and hypotonic intravenous rehydration. Eur. J. Appl. Physiol.
nesses: a risk of athletic participation. Med. Exerc. Nutr. Health 77:242–248, 1998.
2:125–134, 1993. 9. CRAIG, F. N. and E. G. CUMMINGS. Dehydration and muscular
5. BROZEK, J., F. GRANDE, J. ANDERSON, and A. KEYS. Densimetric work. J. Appl. Physiol. 21:670 – 674, 1966.
analysis of body composition: revision of some quantitative as- 10. DEFARES, J. G. Determination of PVCO2 from the exponential CO2
sumptions. Ann. N.Y. Acad. Sci. 110:113–140, 1963. rise during rebreathing. J. Appl. Physiol. 13:159 –164, 1958.
132 Official Journal of the American College of Sports Medicine http://www.msse.org
10. 11. DESCHAMPS, A., R. D. LEVY, M. G. COSIO, E. B. MARLISS, and S. Effects of primary hypohydration on physical work capacity. Int.
MAGDER. Effect of saline infusion on body temperature and en- J. Biometeorol. 32:176 –180, 1988.
durance during heavy exercise. J. Appl. Physiol. 66:2799 –2804, 26. RAMANATHAN, N. L. A new weighting system for mean surface
1989. temperature of the human body. J. Appl. Physiol. 19:531–533,
12. DILL, D. B. and D. L. COSTILL. Calculation of percentage changes 1964.
in volumes of blood, plasma, and red cells in dehydration. J. Appl. 27. RIEBE, D., C. M. MARESH, L. E. ARMSTRONG, et al. Effects of oral
Physiol. 37:247–248, 1974. and intravenous rehydration on ratings of perceived exertion and
13. GISOLFI, C. V. and A. J. RYAN. Gastrointestinal physiology during thirst. Med. Sci. Sports Exerc. 29:117–124, 1997.
exercise. In: Body Fluid Balance Exercise and Sport. E. R. Buskirk 28. ROBINSON, T. E., D. Y. SUE, A. HUSZCZUK, D. WEILER-RAVELL, and
and S. M. Puhl (Eds.). New York: CRC Press, 1996, pp. 19 –51. J. E. HANSEN. Intra-arterial and cuff blood pressure responses
14. GREENLEAF, J. E. Problem thirst, drinking behavior, and involun- during incremental cycle ergometry. Med. Sci. Sports Exerc. 20:
tary dehydration. Med. Sci. Sports Exerc. 24:645– 656, 1992. 142–149, 1988.
15. GREENLEAF, J. E. and T. MORIMOTO. Mechanisms controlling fluid 29. ROWELL, L. B. Human Circulation Regulation during Physiolog-
ingestion: thirst and drinking. In: Body Fluid Balance: Exercise ical Stress. New York, Oxford University Press, 1986, pp. 363–
and Sport. E. R. Buskirk and S. M. Puhl (Eds.). New York: CRC 406.
Press, 1996, pp. 1–17. 30. SAWKA, M. N., and K. B. PANDOLF. Effect of body water loss on
16. HAMILTON, M. T., J. GONZALEZ-ALONSO, S. J. MONTAIN, and E. F. physiological function and exercise performance. In: Fluid Ho-
COYLE. Fluid replacement and glucose infusion during exercise meostasis During Exercise. C. V. Gisolfi, and D. R. Lamb (Eds.)
prevent cardiovascular drift. J. Appl. Physiol. 71:871– 877, 1991. Carmel, IN: Brown and Benchmark, 1990, pp. 1–30.
17. HARGREAVES, M. Skeletal muscle carbohydrate metabolism during
31. SAWKA, M. N., and C. B. WENGER. Physiological responses to
exercise. In: Exercise Metabolism. M. Hargreaves (Ed.). Cham-
acute exercise-heat stress. In: Human Performance Physiology
paign, IL: Human Kinetics, 1995, pp. 41–72.
and Environmental Medicine at Terrestrial Extremes. K. B. Pan-
18. HARRISON, M. H. Effects of thermal stress and exercise on blood
dolf, M. N. Sawka, and R. R. Gonzalez (Eds.) Dubuque, IA:
volume in humans. Physiol. Rev. 65:149 –209, 1985.
Brown and Benchmark, 1988, pp. 97–152.
19. HERRERA, J., C. M. MARESH, L. E. ARMSTRONG, et al. Perceptual
responses to exercise in the heat following rapid oral and intra- 32. SAWKA, M. N., A. J. YOUNG, R. P. FRANCESCONI, S. R. MUZA, and
venous rehydration. Med. Sci. Sports Exerc. 30(Suppl.):S6, 1998. K. B. PANDOLF. Thermoregulatory and blood responses during
20. HORSWILL, C. A. Applied physiology of amateur wrestling. Sports exercise at graded hypohydration levels. J. Appl. Physiol. 59:
Med. 14:114 –143, 1992. 1394 –1401, 1985.
21. MONTAIN, S. J. and E. F. COYLE. Fluid ingestion during exercise 33. SZLYK-MODROW, P. C., R. P. FRANCESCONI, and R. W. HUBBARD.
increases skin blood flow independent of increases in blood vol- Integrated control of body fluid balance during exercise. In: Body
ume. J. Appl. Physiol. 73:903–910, 1992. Fluid Balance: Exercise and Sport E. R. Buskirk and S. M. Puhl
22. NADEL, E. R., G. W. MACK, and H. NOSE. Influence of fluid (Eds.). New York: CRC Press, 1996, pp. 117–136.
replacement beverages on body fluid homeostasis during exercise 34. TAKAMATA, A., G. W. MACK, C. M. GILLEN, A. C. JOZSI, and E. R.
and recovery. In: Fluid Homeostasis During Exercise. C. V. NADEL. Osmoregulatory modulation of thermal sweating in hu-
Gisolfi and D. R. Lamb (Eds.). Carmel, IN: Brown and Bench- mans: reflex effects of drinking. Am. J. Physiol. 268:R414 –R422,
mark, 1990, pp. 181–198. 1995.
23. NOSE, H., G. W. MACK, X. SHI, K. MORIMOTO, and E. R. NADEL. 35. WEBSTER, S., R. RUTT, and A. WELTMAN. Physiological effects of
Effect of saline infusion during exercise on thermal and circula- a weight loss regimen practiced by college wrestlers. Med. Sci.
tory regulations. J. Appl. Physiol. 69:609 – 616, 1990. Sports Exerc. 22:229 –234, 1990.
24. NOSE, H. and A. TAKAMATA. Integrative regulations of body tem- 36. WERNER, J. Temperature regulation during exercise: an overview.
perature and body fluid in humans exercising in a hot environ- In: Exercise, Heat, and Thermoregulation. C. V. Gisolfi, D. R.
ment. Int. J. Biometeorol. 40:42– 49, 1997. Lamb, and E. R. Nadel (Eds.). Dubuque, IA: Brown and Bench-
25. PINCHAN, G., R. K. GAUTTAM, O. S. TOMAR, and A. C. BAJAJ. mark, 1993, pp. 49 –77.
INTRAVENOUS VERSUS ORAL REHYDRATION Medicine & Science in Sports & Exercise 133