Distance running performance is influenced by numerous host factors including maximal aerobic power (V˙O2max), muscle fiber type, threshold of blood lactate accumulation (12), fractional use of maximal aerobic capacity (% V˙O2max) (20), neuromuscular power output (23), and running economy (RE) (5,28). However, the vast majority of between-athlete variance in distance running performance is accounted for by RE (primary), V˙O2max (secondary), and % V˙O2max (tertiary) (5,10). RE is defined as the steady-state oxygen uptake (V˙O2) required to run at a given speed (10) and is expressed, in controlled laboratory settings, as the oxygen used per unit mass per minute (mL·kg−1·min−1) while running at a specified treadmill speed. Both physiological and biomechanical factors affect RE. These include, but are not limited to, metabolic adaptations within skeletal muscle (i.e., increased mitochondrial density and oxidative enzyme activity), the ability of muscles to store and release elastic energy (i.e., stiffness), stride length, stride frequency, vertical body oscillations, and breaking forces (15,28).
Athletes, coaches, and sport scientists constantly seek ways to improve RE (15). Although evidence suggests that the RE of competitive distance runners may or may not change, depending on training status at the beginning of a periodized training year and the type of training undertaken (4,6), the acute influences of HY, catecholamine levels, and hyperthermia on RE are not known. However, Coyle (11) noted that some marathon runners perform well despite dehydration of −4 to −8% of body mass. He proposed that this may occur because fluid loss (i.e., reduced body mass) lowers the oxygen cost of movement. For example, if dehydration by 5% of body mass could reduce the absolute oxygen cost of running by 5%, the power output per kilogram of body mass would theoretically increase. This theoretical mechanism is paradoxical because HY of −2 to −5% of body mass decreases maximal aerobic power (V˙O2max) (14,22,29) and diminishes endurance performance in a mild environment (1,9,18). These levels of HY decrease stroke volume (SV), cardiac output (Q), and cutaneous blood flow, secondary to reduced plasma volume and venous return to the heart (13). To our knowledge, no data exist to confirm or refute this hypothesis. Coyle (11) also recognized that successful running performance involves more than energy expenditure, in that winning depends on cognitive function and perception of effort, both of which may be affected negatively by HY.
HY may negatively affect RE in two additional ways. First, the decrease of V˙O2max that results from HY likely involves reduced delivery of oxygen to active skeletal muscle tissue, especially as influenced by the sympathetic nervous system. Although the details are unknown, it is possible that an increase of circulating norepinephrine maintains the V˙O2 (i.e., RE) of endurance athletes during HY by inducing constriction of specific vascular beds and offsetting the plasma volume reduction that accompanies HY (30). Thus, our research team measured plasma norepinephrine (NE) concentrations during the present investigation. Secondly, body temperature rises during exercise in proportion to the level of HY (3,13,30), increasing by 0.1-0.4°C (vs EU) for each 1% loss of body mass (29). In turn, heat storage impairs ATP and creatine phosphate resynthesis, increases phosphofructokinase activity via the Q10 effect, and increases muscle lactate concentration (20). These effects of hyperthermia could directly alter V˙O2, contribute to local muscular fatigue, or influence exercise performance.
Therefore, the present investigation was designed to examine the effects of a fluid deficit (−5.5 to −5.7% of body mass) on RE and the associated responses of competitive collegiate distance runners who exercised in a mild (23°C) environment. The measured variables included V˙O2, rectal temperature (Tre), NE concentration, plasma volume change, plasma lactate, plasma glucose, cardiopulmonary factors, and ratings of perceived exertion. In view of the dehydration-induced body mass loss, we hypothesized that the steady-state V˙O2 required to run at two treadmill speeds would decrease (i.e., enhance RE), explaining why some highly trained endurance runners perform well when dehydrated (21). This investigation is relevant to competitive endurance athletes because they encounter water losses of more than 5% of body mass during prolonged training and competition (9,11). This investigation also is relevant to the design and interpretation of RE experiments because hydration state was controlled in few previously published studies (5,10,20).
Ten members of a National Collegiate Athletics Association Division I men's cross-country team volunteered for this study. During a briefing, all subjects were informed of the risks involved, signed a written informed consent statement, completed a medical history questionnaire and were screened for respiratory, metabolic, or cardiovascular disease. This protocol was approved by the local institutional review board for human subjects.
Before experimental testing, subjects reported to the laboratory three times for preliminary testing. Day 1 involved measurements of height, mass, age, hydrostatic weight for body composition determination, and an exercise test to determine maximal aerobic power (V˙O2max). Treadmill accommodation sessions were conducted on days 2 and 3. On four separate days (i.e., not days 1-3), subjects visited the laboratory for 10 min for body mass measurements and urine specific gravity measurements. These visits had two purposes: a) to establish a 4-d mean baseline morning body mass, and b) to educate subjects regarding the relationship between a morning urine specific gravity measurement and their fluid consumption during the previous 24 h.
Maximal aerobic power (V˙O2max) was measured via a 10- to 15-min incremental test to exhaustion on a motorized treadmill. The subjects ran at a consistent, predetermined speed for the duration of the test. The first 4 min of the test were performed at 0% grade; the grade was then raised to 4% from minutes 4-6, and then increased 2% every 2 min until subjects reached volitional exhaustion. The primary criterion for a true V˙O2max test was achievement of a plateau of V˙O2 (Δ < 150 mL·min−1) with an increase of treadmill grade. All subjects demonstrated this plateau. Other criteria included the subject's volitional exhaustion, rating of perceived exertion (RPE) > 18, respiratory exchange ratio (RER) > 1.1, and attainment of age-predicted maximal heart rate (220 − age) (7). V˙O2, RER, and minute ventilation (V˙E) were measured using open-circuit spirometry (Medical Graphics, CPX-D metabolic system, St. Paul, MN) with breath-by-breath analysis. Measurements of V˙O2, V˙E, and RER were taken at 30-s intervals throughout the test; RPE was recorded every 2 min and at the end of the test. Heart rate (HR) was measured using a transmitting cardiotachometer (Vantage, Polar Electo Co., model 8799, Hartland, WI). Residual lung volume was predicted from a vital capacity measurement, via handheld spirometer (D.R.P., Wiesbaden, Germany). The hydrostatic weighing technique was used to determine body density, and body fat was calculated from body density.
A CO2 rebreathing technique was used to measure Q (16). Subjects practiced this technique twice, at the 8-min point of each 10-min accommodation session. On the command of the technician, at the end of an expiration, the subject pushed a plunger to initiate rebreathing from a preinflated air bag containing 4% CO2. The subject breathed at a respiratory rate of 40 breaths per minute; cadence was set by a metronome at 80 bpm. The regression line of best fit (i.e., CO2 vs time) was adjusted after testing to eliminate final breaths that may have been influenced by recirculation of CO2 or that fell outside of the confidence intervals. Corrections always increased the r2 toward 1.0, improving the fit of the exponential regression equation.
Treadmill accommodation sessions.
All subjects reported to the laboratory on two separate occasions for treadmill accommodation sessions consisting of three 10-min bouts of running with a 10-min rest period between each. These accommodation sessions were performed to train subjects to run on a treadmill and to reduce the learning effect of treadmill running. The first day was performed at 70% and the second at 85% of individual V˙O2max. During these sessions, running speeds for the exercise economy experimental sessions were determined. Absolute V˙O2 was measured at the 6-min point of each session, and treadmill speed was adjusted to elicit the exact V˙O2 required for 70 and 85% V˙O2max. Once subjects maintained the appropriate V˙O2, the speed of the treadmill was measured with a handheld tachometer (Model 8204-20, Cole Parmer Inc.) and recorded for use during subsequent experimental sessions. Care was taken to control clothing and running shoes, the ambient temperature, wind velocity (no fan), and time of the day during these experiments.
This investigation consisted of four experimental treatments on separate days, twice in a euhydrated (EU) and twice in a 5% hypohydrated (HY) state. At each hydration level, subjects performed one 10-min treadmill run per day, at either 70% V˙O2max (EU 70% or HY 70%) or 85% V˙O2max (EU 85% or HY 85%), in a mild and dry environment (air temperature 23°C, relative humidity 50%). The four tests were separated by at least 1 d of recovery. Subjects wore the same shorts, running shoes, and socks for all sessions. Subjects were randomly assigned to the treatments to reduce the possibility of an order effect.
Control of energy balance and dehydration.
Subjects recorded all food and fluid consumption for 3 d prior to their initial laboratory experiment (randomly assigned). After the initial RE experiment, subjects were provided with a record of their dietary intake, with instructions to consume meals that matched this record as closely as possible during the 3 d before all subsequent RE experiments.
One day before the two HY experiments, subjects reported to the laboratory in a well-hydrated state for pretest body mass assessment and instructions regarding dehydration. A pretest body mass measurement was taken in running shorts, and the amount of mass loss necessary to reach 5% HY was calculated. Subjects were then instructed to refrain from any fluid intake during daily living and performed a normal outdoor training run (≥ 1 h, approximately 75% V˙O2max) to lose the necessary body water. Because these distance runners all were members of a collegiate cross-country team, they were accustomed to running virtually every day. Thus, they ran a similar outdoor training run before all tests (i.e., EU and HY sessions) with the assistance and consent of their cross country coach.
Food intake was similar on the two dehydration days before the HY experiments; it consisted of a typical breakfast at home before weigh-in, plus the following food items: two turkey sandwiches on wheat bread with a teaspoon of mustard, two bagels, and a cup of spaghetti with parmesan cheese for dinner. This diet, after breakfast, provided 1177 calories (61% carbohydrate, 24% protein, 15% fat). The total fluid mass provided by these food components was 301 g of water plus 130 g in solid food. This diet was designed with the help of a nutritionist to minimize water consumption yet ensure necessary caloric intake with a concentration of carbohydrates that athletes typically consume. Subjects ate ad libidum for the remainder of the day after laboratory tests. Subjects also ate ad libidum on the days prior to EU experiments. This is acknowledged as a limitation in the experimental design in that the total amount of carbohydrate may have been different on the days before HY tests versus EU tests.
On the morning of all (EU and HY) experiments, subjects arrived at the laboratory in a 12-h postabsorptive state. They consumed no food or fluid during experiments. Body mass, urine specific gravity (Spartan Refractometer, model A 300 CL, Japan), and urine osmolality (freezing point depression; MicroOsmometer, model 3MO, Advanced Instruments, Needham Heights, MA) were measured to determine hydration level. On the days of the two HY trials only, a minimum loss of 4.5% body mass was required, or the experiment was rescheduled.
Subjects inserted a flexible thermister (YSI, Yellow Springs Instruments, model 400, Yellow Springs, OH) 10 cm past the external anal sphincter for measurement of rectal temperature. HR was measured with the same technique used during the V˙O2max test (above). A 20-gauge Teflon™ cannula was inserted into a superficial forearm vein for blood sampling. The cannula was kept patent with isotonic saline placed in a male luer adapter (Abbott Hospitals, Inc., North Chicago, IL). Next, subjects stood during a 20-min fluid equilibration period. A 12-mL preexercise blood sample was drawn for measurement of hematocrit, hemoglobin, lactate (HLA), glucose, and NE concentration. The subjects then began 10 min of running at the appropriate intensity (i.e., either 70 or 85% V˙O2max, in random order), using the treadmill speeds determined during accommodation sessions (see above). V˙O2, VE, and RER were measured during minutes 6-8 of exercise, and CO2 rebreathing (described above) (16) was performed during minute 8 to determine Q (mL·min−1). HR, Tre, and RPE were recorded every 2 min during the exercise bout. SV (mL·beat−1) was calculated from Q and HR values (SV = Q / HR). At the 10-min point of running, the treadmill was stopped and an immediate postexercise blood sample (12 mL) was taken for analysis of hematocrit and the concentrations of hemoglobin, glucose, HLA, and NE.
All blood samples were collected in a plastic syringe, and 6 mL were immediately transferred to a prechilled vacutainer, inverted several times to mix the blood with the preservative ethylenediaminetetraacetic acid, and then stored on ice for 20 min. The remainder of the blood was pipetted into a lithium heparin vacutainer. Separate aliquots of blood were analyzed in triplicate for hematocrit, via the microcapillary technique, following centrifugation (4 min at 9500 × g). Hemoglobin concentration was measured in triplicate using the cyanmethemoglobin method (Sigma Chemical, St. Louis, MO). A 20-μL aliquot of blood was transferred to a borsilicate tube containing 5 mL of cyanmethemoglobin reagent and allowed to stand for 20 min. The tube was then inverted several times and placed in a spectrophotometer (Bausch & Lomb, Spectronic 88), for comparison with a standard curve. Plasma volume changes were calculated from hematocrit and hemoglobin concentration values, following the method of Dill and Costill (17). The remaining heparinized blood was spun in a refrigerated (4°C) centrifuge at 2000 × g for 10 min. Aliquots of this plasma were used for HLA and glucose measurements in triplicate, using an enzymatic technique (Glucose L-Lactate Analyzer, Model 2300, Yellow Springs Instruments, Yellow Springs, OH).
The blood preserved in ethylenediaminetetraacetic acid was spun in a refrigerated (4°C) centrifuge at 2000 × g for 10 min. The separated plasma was transferred to a plastic container and frozen at −80°C for later analysis of NE using high-pressure liquid chromatography (Waters Chromatography, model 712, Milford, MA). NE plasma samples were thawed, and 1 mL was transferred into sample columns (Chromsystems, Munich, Germany). Treated samples were analyzed in triplicate against a standard curve established by the internal standard supplied with the assay kit. All samples and the internal standard passed through the C14 column as a mobile phase and were analyzed by electrochemical detection (Waters Chromatography, model 712, Milford, MA). The intraassay coefficient of variation for this assay was 10.1%, and the sensitivity of the assay was 0.059 nmol·L−1.
Analysis of variance (treatment × time) with repeated measures was used to compare differences among trials. The Tukey's post hoc analysis was used to determine significant differences within and between conditions. Selected statistical correlations were computed for each condition with a Pearson product-moment correlation. The 0.05 level of significance was selected for all comparisons. All data were presented as mean with the associated standard deviation of the mean (± SD). The test subject sample size was determined on the basis of previous studies of RE in our laboratory; we estimated that 10 subjects would allow us to detect a between-treatment difference of 0.3 L·min−1.
The 10 male subjects were highly trained distance runners who averaged 94.9 ± 16.4 km·wk−1 of running during this study. Mean ± SD (range) subject characteristics were age, 20 ± 3 (18-21) yr; height, 178.5 ± 6.3 (168-188.5) cm; body mass, 66.7 ± 5.4 (58.9-76.4) kg; body fat, 9.6 ± 1.9 (6.5-12.9) %; V˙O2max, 4.41 ± 0.25 (4.08-4.73) L·min−1 and 66.5 ± 4.1 (61.2-74.2) mL·kg−1·min−1. Treadmill speeds during experimental sessions were identified as 3.7 ± 0.3 m·min−1 (8.4 ± 2.5 mph) and 4.6 ± 0.3 m·min−1 (10.4 ± 2.8 mph) for the 70% and 85% V˙O2max tests, respectively.
Table 1 presents the hydration indices of the subjects before exercise. The reductions of body water during dehydration, before the 70 and 85% V˙O2max trials, were not significantly different. The preexercise urine specific gravity, urine osmolality, and body mass were significantly greater (P < 0.0001) in the HY conditions versus the EU conditions (Table 1); however, there were no statistical differences between these variables in the two HY sessions or in the two EU sessions.
Table 2 provides the cardiopulmonary and perceptual responses during exercise. RE (either absolute or relative V˙O2) was similar in both EU and HY conditions. Both absolute and relative V˙O2 during the 85% V˙O2max conditions were greater (P < 0.001) than the 70% V˙O2max conditions. HY induced significant decreases of SV and Q at both 70 and 85% V˙O2max; minute ventilation decreased only at the lower exercise intensity. HY did not affect RPE during running trials.
Figure 1 depicts HR responses during the four experimental conditions. Exercise HR were greater during HY conditions versus EU at min 10 of the 70% V˙O2max tests and at minutes 4 and 6 of the 85% V˙O2max tests. As expected, HR was greater (P < 0.05, to P < 0.001) at all time points during the 85% V˙O2max versus the 70% V˙O2max) EU and HY tests.
Before exercise, HY elevated morning Tre versus EU (P < 0.001; Fig. 2). Similarly, Tre was greater (P < 0.001) at all exercising time points during HY, compared with the EU conditions, at the same exercise intensity. The change of Tre (Δ Tre, Fig. 3) from preexercise to minutes 6, 8, and 10 was significantly lower (P < 0.001) during HY 70% than all other trials; this statistically significant difference had little physiological significance and likely resulted from the higher initial resting Tre during HY 70%. Δ Tre (representing heat storage) was similar during EU trials until minute 10, at which point EU 85% became significantly (P < 0.05) greater than EU 70%.
Figure 4 illustrates the response of NE to exercise during the four experimental treatments. The preexercise NE values were similar in all conditions. Both HY (EU vs HY) and exercise intensity (70% V˙O2max vs 85% V˙O2max) elevated postexercise NE values significantly (P < 0.05). Postexercise NE values during 85% V˙O2max sessions were significantly greater (P < 0.001) than during 70% V˙O2max sessions.
Table 3 presents the responses of other blood variables. A significant (P < 0.001) increase of HLA occurred from pre- to postexercise during EU 85% and HY 85% tests; the immediate post HLA, the 20-min post HLA, and Δ HLA were significantly (P < 0.001) greater during the 85% V˙O2max (vs 70%) tests, as anticipated. The decreases of plasma volume were statistically similar in all conditions.
This study examined the effects of a body water loss (5.5-5.7% of body mass) on the RE of highly trained collegiate distance runners (V˙O2max, 66.5 ± 4.1 mL·kg−1·min−1). Measurements of body mass, urine osmolality, and urine specific gravity confirmed that runners were similarly hypohydrated prior to HY 70% and HY 85% experiments and were similarly euhydrated before the EU 70% and EU 85% tests. In view of the reduced body mass (i.e., during the HY 70% and HY 85% conditions; see below), we hypothesized that HY would reduce the steady-state V˙O2 (11) required to run at two treadmill speeds (i.e., enhance RE). However, the absence of any significant difference of V˙O2 for comparisons with identical exercise intensities (i.e., EU 70% vs HY 70% and EU 85% vs HY 85%; Table 2) indicated that a) hydration state had no measurable effect on RE at either 75 or 85% V˙O2max, and b) our initial hypothesis was not supported.
The steady-rate V˙O2 values in Table 2 likely reflect the contribution of multiple influences on RE (5,12,23,24,28). For example, body temperature during exercise increases in proportion to the level of HY (3,13,30), rising by 0.1-0.4°C (vs EU) for each 1% loss of body mass (29). In turn, heat storage has been shown to increase V˙O2 per mole of ATP produced (27), increase the oxygen demand and hydrogen ion concentration (8), and increase the rate of glycolysis and intramuscular glycogen depletion (20). Before this investigation, we recognized that these effects of hyperthermia might directly contribute to local muscular fatigue, alter V˙O2, or influence endurance performance. However, the mean final Tre of 37.1-37.5°C (Fig. 2) were not large enough to induce measurable changes of V˙O2 and HLA (HY ves EU; Tables 2 and 3).
The RER values for EU 70% and HY 70% (Table 2) indicated a shift of substrate use from carbohydrate to lipids during both HY tests (i.e., a reduced RER), an effect that is opposite the response to hyperthermia (29). It is possible that the carbohydrate content of the diet was different on the days before the two HY experiments versus the two EU experiments because subjects consumed a unique, low-water diet before HY experiments. The extent to which dietary levels of carbohydrate influenced the RER values in Table 2 is unknown and is acknowledged as a limitation of this investigation. Given the relative ease with which these highly trained distance runners completed all RE tests, it is unlikely that carbohydrate (i.e., muscle or liver glycogen) limited their exercise performance.
Although V˙O2 was unchanged by HY, the increased HR, increased NE concentration, decreased Q, and decreased SV indicated that the HY experiments involved greater strain than the corresponding EU trials. However, these highly trained distance runners were able to cope with the added stress imposed by HY without a change of RE (i.e., V˙O2) or HLA. The greater plasma NE concentration during HY 70% and HY 85% trials, reflecting increased involvement of the sympathetic nervous system, may have reduced peripheral blood flow via vasoconstriction of peripheral vascular beds (30) (i.e., increased peripheral vascular resistance), thereby maintaining oxygen delivery to active muscle tissue (25). It is known that SV and Q are maintained during low-intensity exercise (< 50% V˙O2max), despite HY, by an increase of circulating norepinephrine and epinephrine (26). SV maintenance is primarily attributable to increased left ventricular ejection fraction, which serves to maintain Q when sweat loss reduces plasma volume and ventricular filling pressure (25). In this case, norepinephrine induces constriction of specific vascular beds and offsets the plasma volume reduction that accompanies HY (30). In fact, the effects of HY and exercise intensity on postexercise NE levels (i.e., an index of sympathetic nervous activity) were additive (Fig. 4). This finding augments previous research from this laboratory (19), in which HY of −5.1% resulted in a significantly greater NE concentration during treadmill walking in the heat (90 min, 33°C, 5.6 km·h−1, 5% gradient) versus three controlled experiments that involved less HY (−1.0, −1.4, and −3.0% of body mass).
Two other findings deserve mention. First, the mean Δ PV% that occurred during the HY 70% experiment, although not significantly different, was considerably lower than in the EU 70% experiment (−2.5 vs −7.1%; Table 3). Because HR (Fig. 1) was similar, the smaller plasma volume shift during HY 70% may have resulted from the decreased SV and Q during EU 70% (Table 2). This Δ PV% difference did not exist at the higher exercise intensity (i.e., EU 85% vs HY 85%; Table 3) despite a smaller SV and Q during the HY 85% test (Table 2). Because numerous factors affect plasma volume shifts (3,17,29,30), the present data base suggests no apparent mechanism for this Δ PV% difference, other than the general concept that HY was involved. Secondly, the change of Tre during the HY 70% experiment was smaller (i.e., during minutes 6-8; Fig. 3) than in all other trials. Similarly, no mechanism for this difference can be discerned from the present experiments, which were conducted in a 23°C environment. These two counterintuitive findings suggest that future studies should verify and examine the existence of a previously unrecognized synergistic effect between 5% HY, reduced thermal strain, and decreased Δ PV% that occurs at a moderate (70% V˙O2max) but not a strenuous (85% V˙O2max) exercise intensity.
In summary, this investigation reported a change of body mass, Q, and Tre (i.e., factors that affect RE) in response to 5% HY with no change of RE at 75 and 85% V˙O2max. The increased strain during HY experiments versus EU was evidently not sensed by these distance runners; RPE was similar (Table 2). In view of the 5% body mass loss, our initial hypothesis stated that HY would cause steady-state V˙O2 to decrease (i.e., enhance RE) at two treadmill speeds. This hypothesis was not supported by our observations, and the question of why some endurance runners perform well when dehydrated (9,11) remains unanswered. Finally, because Tre did not increase appreciably during running in a mild environment (Figs. 2 and 3), and because the present data do not disprove a possible effect of hyperthermia on V˙O2, future experiments in a hot environment (i.e., with Tre > 39°C) are warranted because endurance athletes routinely experience Tre during competition that are considerably greater (39.4-43.2°C) (2). Such experiments also could evaluate the effects of HY during prolonged exercise (≥ 1 h) on RE or RPE by replacing fluid loss (i.e., preventing HY) during EU tests.
This research was funded, in part, by the University of Connecticut Research Foundation. The authors gratefully acknowledge the cooperation of the men's cross-country team members and Coach Gregory Roy during this investigation, as well as the technical assistance of Karen Fish.
1. Armstrong, L. E., D. L. Costill, and W. J. Fink. Influence of diuretic-induced dehydration on competitive running performance. Med. Sci. Sport Exerc.
2. Armstrong, L. E., A. E. Crago, R. A. Adams, W. O. Roberts, and C. M. Maresh. Whole-body cooling of hyperthermic runners: comparison of two field therapies. Am. J. Emerg. Med.
3. Armstrong, L. E., C. M. Maresh, C. V. Gabaree, et al. Thermal and circulatory responses during exercise: effects of hypohydration, dehydration, and water intake. J. Appl. Physiol.
4. Armstrong, L. E., C. M. Maresh, M. Whittlesey, M. F.Bergeron, C. Gabaree, and J. R. Hoffman. Longitudinal evaluation of the exercise-heat tolerance and running economy in collegiate distance runners. J. Strength Cond. Res.
5. Bassett, D. R., and E. T. Howley. Limiting factors for maximum oxygen
uptake and determinants of endurance performance. Med. Sci. Sport Exerc.
6. Berg, K. Endurance training and performance in runners. Sports Med.
7. Borg, G. Perceived exertion in relation to physical work load and pulse rate. Kungliga Fysiografiska Sällskapets i Lund Förhandlingar
8. Brooks, G. A., K. J. Hittelman, J. A. Faulkner, and R. E. Beyer. Temperature, liver mitochondrial respiratory functions, and oxygen
debt. J. Appl. Physiol.
9. Cheuvront, S. N., R. Carter, and M. N. Sawka. Fluid balance and endurance exercise performance. Curr. Sports Med. Rep.
10. Conley, D. L., and G. S. Krahenbuhl. Running economy and distance running performance of highly trained athletes. Med. Sci. Sports Exerc.
11. Coyle, E. F. Fluid and fuel intake during exercise. J. Sports Sci.
12. Coyle, E. F., M. E. Feltner, S. A. Kautz, et al. Physiological and biomechanical factors associated with elite endurance cycling performance. Med. Sci. Sports Exerc.
13. Coyle, E. F., and J. González-Alonso. Cardiovascular drift during prolonged exercise: new perspectives. Exerc. Sport Sci. Rev.
14. Craig, F. N., and E. G. Cummings. Dehydration and muscular work. J. Appl. Physiol.
15. Daniels, J. A physiologist's view of running economy. Med. Sci. Sports Exerc.
16. Defares, J. G. Determination of PVCO2
from the exponential CO2
rise during rebreathing. J. Appl. Physiol.
17. Dill, D. B., and D. L. Costill. Calculation of percentage changes in volume of blood, plasma, and red cells in dehydration. J. Appl. Physiol.
18. Fallowfield, J. L., C. Williams, J. Booth, B. H. Choo, and S. Growns. Effect of water ingestion on endurance capacity during prolonged running. J. Sport Sci.
19. Hoffman, J. R., C. M. Maresh, L. E. Armstrong, et al. Effects of hydration state on plasma testosterone, cortisol, and catecholamine concentrations before and during mild exercise at elevated temperature. Eur. J. Appl. Physiol.
20. Kozlowski, S., Z. Brzeziñska, B. Kruk, H. Kaciuba-Uœcilko, J. E. Greenleaf, and K. Nazar. Exercise hyperthermia as a factor limiting physical performance: temperature effect on muscle metabolism
. J. Appl. Physiol.
21. Noakes, T. D. Overconsumption of fluids by athletes. Brit. Med. J.
22. Nybo, L., T. Jensen, B. Nielsen, and J. Gonzalez-Alonso. Effects of marked hyperthermia with and without dehydration on V˙O2
kinetics during intense exercise. J. Appl. Physiol.
23. O'Toole, M. L., and P. S. Douglas. Applied physiology of triathlon. Sports Med.
24. Paavolainen, L., K. Hakkinen, I. Hamalainen, A. Nummela, and H. Rusko. Explosive-strength training improves 5-km running time by improving running economy and muscle power. J. Appl. Physiol.
25. Rowell, L. B. Human Circulation Regulation During Physiological Stress
. New York: Oxford University Press, 1986, pp. 28-29, 224-225.
26. Saltin, B. Circulatory response to submaximal and maximal exercise after thermal dehydration. J. Appl. Physiol.
27. Saltin, B., and L. Hermansen. Esophageal, rectal and muscle temperatures during exercise. J. Appl. Physiol.
28. Saunders, P. U., D. B. Pyne, R. D. Telford, and J. A. Hawley. Factors affecting running economy in trained distance runners. Sports Med.
29. Sawka, M. N., and K. B. Pandolf. Effects of body water loss on physiological function and exercise performance. In: Perspectives in Exercise Science and Sports Medicine, Volume 3: Fluid Homeostasis During Exercise
, C. V. Gisolfi and D. R. Lamb (Eds.). Carmel, IN: Benchmark Press, 1990, pp. 1-38.
30. Sawka, M. N., A. J. Young, R. P. Francesconi, S. R. Muza, and K. B. Pandolf. Thermoregulatory and blood responses during exercise at graded hypohydration levels. J. Appl. Physiol.