If no fluid is ingested during prolonged exercise, water and electrolyte losses in sweat increase plasma sodium concentration and osmolality (19,23). Plasma osmolality rises because more water is lost in sweat than Na+ and water losses come from both the extracellular fluid and the intracellular fluid (21). Because rises in plasma osmolality with dehydration reduce skin blood flow (18,22), it is important for athletes to drink enough fluid to maintain their plasma osmolality in events where dehydration and thermoregulation are of primary concern (20). A complete fluid replacement with either water or commercial CHO-electrolyte solutions maintains plasma volume and osmolality (2,19,23) and extends the time to exhaustion in prolonged exercise, especially in the heat (2,15,17–19).
Even low (<2% of body mass) levels of dehydration may impair high-intensity exercise performance in 31–32°C ambient temperatures. Walsh et al. (30) showed that when cyclists fully replaced their ∼1.1 L fluid losses during 1-h rides at 70% of peak oxygen consumption (V̇O2peak) in the heat, they were able to cycle for 34% longer in a subsequent exercise bout at 90% of V̇O2peak. Below et al. (3) reported that subjects who drank 1.3 L of fluid during a 50-min ride at 80% of V̇O2peak in the heat were able to complete a subsequent (∼10 min) work bout 6.5% faster than without fluid ingestion. In contrast, Robinson et al. (25) found that water ingestion did not improve 1-h cycling performances in a moderate ambient temperature of 20°C. Under that condition, an ingestion of 1.5 L of water had no measurable physiological benefit and only produced an uncomfortable abdominal fullness and a 2% reduction in the “distances covered” in the simulated rides.
The latter finding raises a question of whether it is advisable or practical for athletes to try and replace their fluid losses during competitions in thermoneutral conditions. Most endurance athletes voluntarily drink less than 0.5 L of fluid·h−1 and lose up to 1.0–1.5 L of sweat·h−1 in distance races conducted in 20–25°C environments (20). When athletes attempt to drink fluid at rates closer to their 1.0–1.5 L·h−1 sweat rates, they generally experience gastrointestinal discomfort, especially during running (6,7). Therefore, the American College of Sports Medicine position stand on exercise and fluid replacement states that “during prolonged exercise, frequent (every 15–20 min) consumption of moderate (150 mL) to large (350 mL) volumes is possible” but recommends that “individuals learn their tolerance limits for maintaining a high gastric fluid volume for various exercise intensities” (1). In this study, we examined whether 0.4 and 0.9 L·h−1 rates of fluid intake confer any advantage over ad libitum fluid ingestion on 2-h running performances in a thermoneutral environment.
MATERIALS AND METHODS
Eight male endurance-trained runners participated in this study. All of the subjects ran regularly more than 60 km·wk−1 and had completed marathon races in under 2 h:50 min. The study was approved by the Research and Ethics Committee of the Faculty of Medicine of the University of Cape Town and each subject signed an informed consent.
The means and standard deviations of the subjects’ ages, body masses, heights, peak O2 uptakes (V̇O2 peak), and peak treadmill speeds (PTS) were 31 ± 4 yr, 61 ± 9 kg, 171 ± 9 cm, 68 ± 3 mL·min−1·kg body mass−1, and 20 ± 1 km·h−1, respectively. Body masses and heights were determined with a Model 770 Seca precision balance and stadiometer (Seca, Bonn, Germany). V̇O2 peak and PTS were measured during a progressive exercise test to exhaustion on a Powerjog EG 30 treadmill (Sports Engineering Ltd, Birmingham, U.K.), set at a gradient of 1%. Runs were started at a treadmill speed of 13 km·h−1 and increased by 0·5 km·h−1 every 30 s until the subject could no longer maintain the pace.
During the exercise test, subjects wore a nose-clip and breathed through a mouthpiece connected to an Oxycon Alpha automated gas analyzer (Mijnhardt, Netherlands). Before each test, the pneumotach was calibrated with a Hans Rudolph 3-L syringe (Vacumed, Ventura, CA), and the oxygen and carbon dioxide analyzers were calibrated with room air and a 5% CO2:95% N2 gas mixture. Pneumotach and gas analyzer outputs were processed by a computer that calculated breath by breath ventilation, oxygen consumption (V̇O2), and carbon dioxide production (V̇CO2) in L·min−1. V̇O2 peak values were the average of the highest V̇O2 values measured over the final 60 s.
After the determination of V̇O2 peak, the subjects rested for ∼5 min and then performed a familiarization run on the treadmill. At first, they ran at ∼65% of V̇O2 peak for 45 min and then they ran “as far as possible” in 15 min on the treadmill by adjusting their own running speed. These runs were conducted in our environmental chamber (Scientific Technology, Cape Town, South Africa) at an air temperature of 25°C, a relative humidity of 55%, and a wind velocity equal to the 13–15 km·h−1 treadmill running speeds of the subjects in the first 90 min of the experimental trials. In this practice run, no physiological measurements were made and the subjects drank ad libitum. The only purpose of the run was to familiarize the subjects with the conditions to be used in the experimental trails and to minimize any learning effect.
At weekly intervals after the practice run, the subjects returned to the laboratory at the same time of day and repeated, in random order, the three experimental performance runs. Over that period, the subjects continued their usual training and consumed similar diets, but refrained from strenuous physical activity on the day before each trial. Training and dietary records were kept in order to aid the subjects’ compliance with the requests.
On the day of the trial, the subjects arrived in the laboratory, 2 h after drinking 500 mL·70 kg−1 body mass of a commercial 6.9 g·100 mL−1 glucose polymer solution containing 16 mEq Na+·L−1. The consumption of 380–580 mL of fluid 2 h before the trials was designed to ensure equal hydration at the start of exercise. Equal hydration was later confirmed by similar pre-trial body masses, hematocrits, hemoglobin concentrations, and serum osmolalities (Table 1).
At the laboratory, the subjects urinated and weighed themselves in the nude. An 18-gauge cannula was then inserted into an antecubital vein and attached to a three-way stopcock for the withdrawal of blood. Venous blood samples (10 mL) were collected (a) after the subject had stood on the treadmill for 10–15 min, (b) at 15-min intervals during a 90 min run at ∼65% of V̇O2 peak, and (c) at the end of a subsequent 30-min performance run. After the collection of each blood sample, the venous cannula was flushed with 1–2 mL of sterile saline containing heparin (5 IU·L−1) to prevent coagulation.
Once initial venous blood samples had been drawn, the subjects began running at ∼65% of V̇O2 peak for 90 min. At the end of this run, there was a 2-min rest interval before the 30-min performance run. In that interval and at the end of the performance run, the subjects towelled dry, urinated, and reweighed themselves in the nude for later calculations of sweat rates. Sweat rates were calculated from the decreases in body mass plus the volume of fluid ingested minus any urine excreted. No corrections were made for the approximately 150 g of carbon lost in the measured production of CO2 during the trials.
During the runs, the subjects drank the commercial 6.9 g·100 mL−1 glucose polymer solution containing 16 mEq Na+·L−1 either ad libitum or in set volumes of either 150 or 350 mL·70 kg−1 body mass every 15–20 min. Before each drink, the subjects were asked to indicate their ratings of perceived exertion using the 20-point Borg scale (4) and to rank their stomach fullness on a scale of 1 (empty) to 5 (uncomfortably bloated). Between drinks, V̇O2 and V̇CO2 were measured over 6-min intervals and used to calculate rates of carbohydrate (CHO) and fat oxidation using the formulae of Frayn (11), assuming a non-protein respiratory exchange ratio. Rates of CHO oxidation in g·min−1 were converted to mmol·min−1 by dividing the values by the 180 mg·mmol−1 molecular weight of glucose.
Venous blood samples (10 mL), collected during the trials, were divided into three aliquots and stored on ice until the end of the trial. One aliquot (1 mL) was placed into a tube containing lithium heparin for hemoglobin concentration and hematocrit measurements immediately after the trial. Another aliquot (5 mL) was allowed to clot (in a tube containing SST gel and clot activator), spun at 2500 ×g for 12 min in a Sigma 302K centrifuge (Laborzentrifugen, Germany) at 4°C, and the supernatent was stored at −20°C for subsequent determinations of serum Na+ and K+ concentrations and osmolality. The remaining blood (4 mL) was placed into a tube containing sodium fluoride and potassium oxalate, centrifuged at 2500 ×g for 12 min at 4°C, and the supernatant was stored at −20°C for later measurements of plasma glucose and lactate concentrations.
Hematocrits were measured in triplicate by microcentrifugation. Hemoglobin concentrations were determined in triplicate with the cyanomethemoglobin method of Hainline (13) using a Beckman DU 60 spectrophotometer (Beckman Instruments, Fullerton, CA). Changes in hemoglobin concentration and hematocrit were used to estimate the percent changes in plasma volume during exercise, as described by Dill and Costill (9).
Serum osmolalities were determined from freezing point depressions using an Osmette A automatic osmometer (Precision Systems, Newton, MA). Serum Na+ and K+ concentrations were assayed with ion selective electrodes (KNA 1 Radiometer, Copenhagen, Denmark). Plasma glucose concentrations were measured with an automated LM 3 glucose analyzer (Analox Instruments, London, U.K.). Plasma lactate concentrations were determined with a spectrophotometric enzymatic assay using a commercial kit (Lactate Pap, Bio Merieux, Marcy-L Etoile, France).
All results are expressed as means ± standard deviations (SD). Statistical analyses were performed with a two-way analysis of variance (ANOVA) for repeated measures and the Scheffe post hoc test. Differences in ratings of perceived exertion and stomach fullness were tested with a nonparametric Kruskal-Wallis ANOVA. A value of P < 0.05 was regarded as significant for all determinations.
Similar initial body masses, hemoglobin concentrations, hematocrits, and serum osmolalities suggested later that each subject was equally hydrated at the start of the three trials (Table 1). During the ad libitum, 150 mL·70 kg−1, and 350 mL·70 kg−1 trials, total CHO-electrolyte solution intakes were 0.76 ± 0.35 L, 0.78 ± 0.12 L, and 1.83 ± 0.28 L, respectively (Fig. 1). The intake of an extra ∼1 L of fluid in the 350 mL·70 kg−1 trials did not significantly increase the volume of urine produced above that in the other two trials (Fig. 1). Compared with the 0.19 ± 0.24 L and 0.11 ± 0.21 L urine volumes in the ad libitum and 150 mL·70 kg−1 trials, the 0.36 ± 0.44 L urine volume in the 350 mL·70 kg−1 trials was increased by only 0.1–0.2 L (P = 0.15).
Differences in fluid intakes also had little effect on the estimated sweat rates during the 90- and 30-min runs (Fig. 1). When the mass of the retained, ingested fluid was added to the loss of body mass during each run, the calculated sweat rates were found to be relatively constant over time and similar from trial to trial. Average sweat rates in the ad libitum, 150 mL·70 kg−1, and 350 mL·70 kg−1 trials were 1.29 ± 0.41, 1.15 ± 0.44, and 1.21 ± 0.33 L·h−1, respectively. The only effect of the additional fluid intake was to reduce the subjects’ loss of body mass from 1.95 ± 0.62 and 1.59 ± 0.74 kg in the ad libitum and 150 mL·70 kg−1 trials to 0.75 ± 0.44 kg in the 350 mL·70 kg−1 trials (P < 0.001, Fig. 1).
Less (1.2% vs 2.7–3.4%) dehydration in the 350 mL·70 kg−1 trials than in the other trials, however, did not attenuate the estimated decreases in plasma volume during exercise (Fig. 2). In all three trials, plasma volumes fell by ∼6% in the first 15 min of exercise (P < 0.01) and then declined by a further 2% in the remainder of the trial.
Serum K+ and Na+ concentrations and osmolality.
The fluid ingestion regimen also had little influence on serum K+ and Na+ concentrations and osmolality (Fig. 2). Irrespective of whether the average drinking rate was 0.39 ± 0.06 L·h−1 or 0.91 ± 0.14 L·h−1, circulating K+ concentrations rose from 3.6–3.9 mEq·L−1 to 4.6–4.8 mEq·L−1 in the first 15 min of exercise (P < 0.01) and then remained constant for the rest of the trial. In contrast, the ∼7 mEq·L−1 and ∼5 mosmol·L−1 increases in serum Na+ concentrations and osmolalities in the three trials were more gradual and not statistically significant.
Plasma glucose and lactate concentrations.
There was also no effect of the higher (1.05 ± 0.16 vs 0.45 ± 0.07 g·min−1) rates of CHO ingestion in the 350 mL·70 kg−1 trials than in the 150 mL·70 kg−1 trials on plasma glucose concentrations (Fig. 3). In all three runs, plasma glucose concentrations remained between 5.5 and 6.5 mmol·L−1. Plasma lactate concentrations were also similar in the ad libitum, 150 mL·70 kg−1, and 350 mL·70 kg−1 trials (Fig. 3). At the end of the 90-min runs at ∼65% of V̇O2 peak, plasma lactate concentrations were 1.2 ± 0.2, 1.3 ± 0.6, and 1.5 ± 0.6 mmol·L−1, and at the end of the 30 min performance runs, they were increased to 2.9 ± 1.8, 2.9 ± 1.8, and 3.2 ± 2.1 mmol·L−1, respectively (P < 0.01).
Rises in plasma lactate concentrations in the 30-min performance runs were associated with increases in treadmill running speeds (Fig. 4). After the 90-min runs at ∼65% of V̇O2 peak, the subjects accelerated from 14.3 ± 0.9 km·h−1 to 15.8 ± 0.9, 15.6 ± 1.1, and 15.4 ± 1.4 km·h−1 in the ad libitum, 150 mL·70 kg−1, and 350 mL·70 kg−1 trials. Increases in running speeds raised V̇O2 values from 44.4 ± 3.7 mL·min−1·kg−1 to 49.5 ± 6.0, 50.8 ± 5.3, and 51.9 ± 8.3 mL·min−1·kg−1, or from ∼65% to ∼73%, 75%, and 76% of V̇O2 peak, respectively (Fig. 4). Less dehydration in the 350 mL·70 kg−1 trials than in other trials did not measurably increase V̇O2 values at given running speeds. Oxygen costs per kilometer in the ad libitum, 150 mL·70 kg−1, and 350 mL·70 kg−1 treadmill performance runs were 192 ± 26, 196 ± 23, and 203 ± 27 mL· kg−1· km−1, respectively.
Rates of CHO oxidation were also unaffected by the volumes of fluid consumed (Fig. 4). In all three trials, mean rates of CHO oxidation decreased steadily from 17 ± 4 to 13 ± 3 mmol·min−1 during the 90-min runs and then increased to 15 ± 6 mmol·min−1 with the rise in exercise intensity from ∼65% to ∼75% of V̇O2 peak in the 30-min performance runs (both P < 0.05). However, the acceleration of CHO oxidation in the performance runs did not increase the percent contribution to energy production from CHO oxidation. Respiratory exchange ratios indicated that the percent contribution to energy production from CHO oxidation fell from 87 ± 5% to 65 ± 4% during the 90-min runs and then remained at 65 ± 5% during the 30-min performance runs (data not shown).
Ratings of perceived exertion and stomach fullness.
Ratings of perceived exertion (RPE) were also similar in the three trials (Fig. 5). During each trial, RPE rose from ∼9 to 15 units. In contrast, ratings of stomach fullness (RSF) were significantly greater in the latter half of the 350 mL·70 kg−1 trial than in the other trials (Fig. 5). Two of the eight subjects experienced such severe gastrointestinal discomfort in the 350 mL·70 kg−1 trial that they failed to complete their 30-min performance runs. Those subjects stopped their performance runs after 15 and 18 min.
Increases in fluid intake from ∼0.4 L·h−1 in the ad libitum and 150 mL·70 kg−1 trials to ∼0.9 L·h−1 in the 350 mL·70 kg−1 trial had little effect on the subjects’ 0.05–0.20 L·h−1 urine production during exercise and no effect on their ∼1.2 L·h−1 sweat rates (Fig. 1). Others have also found that fluid ingestion does not significantly increase urine production or sweat rates during moderate intensity exercise. In those studies, cyclists rode at 62–70% of V̇O2peak in ambient temperatures of either 22°C (14) or 32–33°C (18,19,30). Under both moderate and hot conditions, the cyclists’ 1.2–1.4 L·h−1 sweat losses were unaffected by their up to1.2 L·h−1 rates of fluid intake.
In contrast, Montain and Coyle (19) found that increasing (0, ∼0.3, ∼0.7, and ∼1.2 L·h−1) rates of fluid ingestion progressively reduced estimated falls in plasma volume from ∼9% to ∼6% and reduced rises in serum Na+ concentration from ∼6 to ∼2 mEq·L−1 during 2-h rides at ∼65% of V̇O2peak in a 33°C environment. However, others have observed that drinking does not measurably attenuate declines in plasma volume in the first 1–2 h of exercise. With or without fluid replacement, plasma volumes fell by 8–12% at 55–70% of V̇O2peak (2,18), by about 15% at ∼85% of V̇O2peak (25), and by 17–18% at 90% of V̇O2peak (30). In the present study, the intake of ∼1.0 L of additional fluid in the 350 mL·70 kg−1 trial also did not reduce the estimated declines in plasma volume (Fig. 2). Irrespective of the amount of fluid ingested, plasma volumes decreased by ∼6% in the transition from rest to exercise and by a further ∼2% in the remainder of the runs. Over the same period, circulating K+ and Na+ concentrations increased by ∼1 and ∼7 mEq·L−1 and serum osmolality rose by ∼5 mOsm·L−1. The observation that serum Na+ concentrations did not fall when fluid was ingested at the highest rates suggests a benefit of consuming electrolyte-containing drinks during prolonged exercise. Vrijens and Rehrer (29) found dangerous falls in plasma Na+ concentration when their subjects drank water rather than a commercial CHO-electrolyte solution at rates of 1.2 L·h−1 during 3 h of exercise in a 34°C environment.
A greater (∼1.0 vs 0.5 g·min−1) rate of CHO ingestion in the 350 mL·70 kg−1 trial than in the other trials also did not effect plasma concentrations of glucose (≥5 mmol·L−1) and lactate (∼3 mmol·L−1) during the 30-min performance runs (Fig. 3). In all three performance runs, increases in running speeds from ∼14 to 15–16 km·h−1 and rises in exercise intensities from ∼65% to 75% of V̇O2 peak elevated plasma lactate concentrations from ∼1.5 to 3 mmol·L−1 and accelerated CHO oxidation from ∼13 to 15 mmol·min−1 (Fig. 4). However, the acceleration of CHO oxidation in the performance runs did not increase its percent contribution to energy production. In all three trials, the percent contribution to energy production from CHO oxidation fell from about 87% to 65% during the 90 min submaximal runs and then remained at around 65% during the 30-min performance runs (data not shown).
A failure of the rise in exercise intensity to influence patterns of fuel utilization is contrary to the prediction of the “crossover” concept (5) and may be due to the ingestion of CHO during exercise. Rauch et al. (24) found that CHO ingestion eliminated 50% versus 65% differences in final contributions to energy production from CHO oxidation in cyclists drinking water during 3-h rides at 55% or 70% of V̇O2 peak. Van Zyl et al. (28) also showed that the contribution to energy production from ingested and endogenous CHO oxidation remained at ∼65% during a simulated 40-km time-trial at nearly 80% of V̇O2 peak after a 2-h ride at 60% of V̇O2 peak.
Only the ratings of stomach fullness were different in the three trials (Fig. 5). Most of the eight subjects felt uncomfortably bloated after the 1st h of submaximal exercise in the 350 mL·70 kg−1 trial, and two of the subjects experienced such severe gastrointestinal discomfort that they failed to complete their performance runs. Those subjects did not have particularly low sweat rates and were not reluctant drinkers in the ad libitum trial. More gastrointestinal discomfort with greater than ad libitum rates of fluid ingestion may have been due to either the athletes unfamiliarity with running with a full stomach or to delays in gastric emptying at high exercise intensities (6). Part of the uncomfortable abdominal fullness could also have been caused by limits to the rates of fluid absorption from the small intestine. Although maximum rates of intestinal fluid absorption are about 0.8 L·h−1 at rest (8), they may be less during exercise. Some believe that intestinal absorptive capacity is unaffected by exercise intensities that can be sustained for ≥ 30 min (12), and others think that high exercise intensities decrease rates of intestinal fluid absorption (10). It is also possible that fluid absorption was slowed by the 6.9% CHO content of the drink. Ryan et al. (26) have shown that an increase in the CHO content of fluid ingested during exercise from 6 to 8% slows the absorption of water by about 50%. Whatever the mechanism, these data suggest that it may not be possible for athletes to absorb sufficient fluid to maintain hydration during competitive running exercise. The additional ingestion of ∼1.0 L of fluid in the 350 mL·70 kg−1 trial had no measurable effects on plasma volume and osmolality and did not improve 2-h running performances in a 25°C environment. Robinson et al. (25) also found that fluid ingestion did not influence plasma volume or improve 1-h cycling performances in a moderate 20°C environment. In that study, an ingestion of 1.5 L of water only produced an uncomfortable abdominal fullness and reduced the “distances covered” in the simulated rides by 2%.
One reason why greater rates of fluid ingestion did not improve running speeds may be that the ad libitum fluid intakes of our runners were higher than perhaps those of less experienced runners. It is also possible that self-paced treadmill runs are too variable to detect subtle changes in exercise performance. Schabort et al. (27) found a 1.8–4.0% 95% confidence interval in the performances of eight subjects who each ran “as far as possible” on a treadmill in 1 h on three occasions. Alternatively, the effect of fluid replacement on performance may be more noticeable in exercise conducted over a longer duration or in a hot environment were adequate hydration and thermoregulation are of greater concern. Our findings only apply to ≤ 2 h of exercise in a thermoneutral environment. In more prolonged exercise leading to greater levels of dehydration, athletes probably should follow the ACSM recommendation and “attempt to consume fluid at a rate sufficient to replace all the water lost through sweating or consume the maximal amount that can be tolerated” (1). Most studies have shown that drinking delays fatigue in prolonged exercise (2,15–19) and improves exercise performance in the heat (3,30).
We thank Gary Wilson and Judy Belonje for their expert technical assistance.
This study was supported by grants from the Medical Research Council and the Foundation for Research and Development of South Africa, the Nellie Atkinson and Harry Crossley Research funds of the University of Cape Town and Bromor Foods Inc.
1. American College of Sports Medicine. Position stand on exercise and fluid replacement. Med. Sci. Sports Exerc. 28: I–vii, 1996.
2. Barr S. I., D. L. Costill, and W. J. Fink. Fluid replacement during prolonged exercise: effects of water, saline or no fluid. Med. Sci. Sports. Exerc. 23: 811–817, 1991.
3. Below, P. R., R. Mora-Rodriguez, J. Gonzalez-Alonso,and E. F. Coyle. Fluid and carbohydrate ingestion independently improve performance during 1 h of intense exercise. Med. Sci. Sports Exerc. 27: 200–210, 1995.
4. Borg, G. Simple rating methods for estimation of perceived exertion. In:Physical Work and Effort
, G. Borg (Ed.). New York: Pergamon, 1975, pp. 39–46.
5. Brooks, G. A., and J. Mercier. Balance of carbohydrate and lipid utilization during exercise: the “crossover” concept. J. Appl. Physiol. 76: 2253–2261, 1994.
6. Brouns, F., W. H. M. Saris, and N. J. Rehrer. Abdominal complaints and gastrointestinal function during long-lasting exercise. Int. J. Sports Med. 8: 175–189, 1987.
7. Costill, D. L., W. F. Kramer, and A. Fisher. Fluid ingestion during distance running. Arch. Environ. Health 21: 520–525, 1970.
8. Davis, G. R., C. A. Santa Ana, S. G. Morawski, and J. S. Fordtran. Development of a lavage solution associated with minimal water and electrolyte absorption or secretion. Gastroenterology
9. Dill, D. B., and D. L. Costill. Calculation of percentage change in volumes of blood, plasma and red cells in dehydration. J. Appl. Physiol. 37: 247–248, 1974.
10. Fordtran, J. S., and B. Saltin. Gastric emptying and intestinal absorption during prolonged severe exercise. J. Appl. Physiol. 23: 331–335, 1967.
11. Frayn, K. N. Calculation of substrate oxidation rates in vivo from gas-exchange. J. Appl. Physiol. 55: 628–634, 1983.
12. Gisolfi, C. V., K. J. Spranger, R. W. Summers, H. P. Schedel, and T. L. Bleiler. Effects of cycle exercise on intestinal absorption in humans. J. Appl. Physiol. 71: 2518–2527, 1991.
13. Hainline, A. Haemoglobin, standard methods. Clin. Chem. 2: 49–60, 1958.
14. Hamilton, M. C., J. Gonzalez-Alonso, S. Montain, and E. F. Coyle. Fluid replacement and glucose infusion during exercise prevent cardiovascular drift. J. Appl. Physiol. 71: 871–877, 1991.
15. Maughan, R. J., C. E. Fenn, M. Gleeson, and J. P. Leiper. Effects of fluid, electrolyte and substrate ingestion on endurance capacity. Eur. J. Appl. Physiol. 58: 481–486, 1989.
16. McConnel, G. K., C. M. Burge, S. L. Skinner, and M. Hargreaves. Influence of ingested fluid volume on physiological responses during prolonged exercise. Acta Physiol. Scand. 160: 149–156, 1997.
17. Millard-Stafford, M., P. B. Sparling, L. B. Rosskopf, and L. J. Dicarlo. Carbohydrate-electrolyte replacement improves distance running performance in the heat. Med. Sci. Sports Exerc. 24: 934–940, 1992.
18. Montain, S. J., and E. F. Coyle. Fluid ingestion during exercise increases skin blood flow independent of increases in blood volume. J. Appl. Physiol. 73: 903–910, 1992a.
19. Montain, S. J., and E. F. Coyle. The influence of graded dehydration on hyperthermia and cardiovascular drift during exercise. J. Appl. Physiol. 73: 1340–1350, 1992b.
20. Noakes, T. D. Fluid replacement during exercise. In:Exercise and Sports Science Reviews,
Vol. 21., J. O. Holloszy (Ed.). Baltimore: Williams and Wilkins, 1993, pp. 297–330.
21. Nose, H., G. W. Mack, X. Shi, and E. R. Nadel. Shift in body fluid compartments after dehydration in humans. J. Appl. Physiol. 65: 318–324, 1988.
22. Nose, H., G. W. Mack, X. Shi, K. Morimoto, and E. R. Nadel. Effect of saline infusion during exercise on thermal and circulatory regulations. J. Appl. Physiol. 69: 609–616, 1990.
23. Powers, S. L., J. Lawler, S. Dodd, R. Tulley, G. Landry, and K. Wheeler. Fluid replacement drinks during high intensity exercise: effects on minimizing exercise induced disturbances in homeostasis. Eur. J. Appl. Physiol. 60: 54–60, 1990.
24. Rauch, L. H. G., A. N. Bosch, S. C. Dennis, T. D. Noakes, and J. A. Hawley. Fuel utilisation during prolonged low-to-moderate intensity exercise when ingesting water or carbohydrate. Pflugers Arch. 430: 971–977, 1995.
25. Robinson, T. A., J. A. Hawley, G. S. Palmer, et al. High rates of fluid intake do not improve one hour cycling performance in a thermoneutral environment. Eur. J. Appl. Physiol. 71: 153–160, 1995.
26. Ryan, A. J., G. P. Lambert, X. Shi, R. T. Chang, R. W. Summers, and C. V. Gisolfi. Effect of hypohydration on gastric emptying and intestinal absorption during exercise. J. Appl. Physiol.
27. Schabort, E. J., W. G. Hopkins, and J. A. Hawley. Reproducibility of self-paced treadmill performance of trained endurance runners. Int. J. Sports Med. 19: 48–51, 1998.
28. Van Zyl, C., E. V. Lambert, J. A. Hawley, T. D. Noakes, and S. C. Dennis. Effects of medium-chain triglyceride ingestion on carbohydrate metabolism and endurance cycling performance. J. Appl. Physiol. 80: 2217–2225, 1996.
29. Vrijens, D. M., and N. J. Rehrer. Sodium free fluid ingestion decreases plasma sodium during exercise in the heat. J. Appl. Physiol., 86: 1847–1851, 1999.
30. Walsh, R. M., T. D. Noakes, S. C. Dennis, and J. A. Hawley. Impaired high intensity cycling performance times at low levels of dehydration. Int. J. Sports Med. 15: 392–398, 1994.