Sex Differences in Voluntary Fluid Intake by Older Adults during Exercise : Medicine & Science in Sports & Exercise

Journal Logo

Basic Sciences: Original Investigations

Sex Differences in Voluntary Fluid Intake by Older Adults during Exercise


Author Information
Medicine & Science in Sports & Exercise: May 2005 - Volume 37 - Issue 5 - p 789-796
doi: 10.1249/01.MSS.0000162622.78487.9C
  • Free


During exercise in the heat, body water is lost (dehydration) as a consequence of sweating. Fluid replacement is critical to ameliorate the deterioration in physiological function and performance that accompanies dehydration (7). However, overconsumption of fluids (especially sodium-free fluids such as water) in excess of sweat loss can lead to electrolyte imbalances (1).

Older adults are at an increased risk for fluid and electrolyte imbalances due to an impaired thirst sensitivity (13,16,25) and renal sodium- and water-conserving ability (14,18,19). In addition, anecdotal evidence suggests that hyponatremia (low blood sodium), which usually occurs due to overconsumption of water, most commonly occurs in women (2,3,10,24); however, little data exist on the actual sex-related differences in voluntary fluid intake behavior of older adults during exercise-heat stress.

Numerous studies looking at fluid replacement in young adults have shown that, due primarily to the presence of sodium, carbohydrate-electrolyte solutions (CES) are more effective than water in stimulating voluntary fluid intake (7,15,17), restoring plasma volume (6,8,12), maintaining fluid/electrolyte balance (9), and attenuating increases in core temperature (9) during exercise-heat stress. In addition, Bar-Or and Wilk (4) have shown that ad libitum consumption of water leads to voluntary dehydration in children; however, the addition of flavor, carbohydrates, and sodium chloride to water prevents dehydration when children are allowed to drink ad libitum. For these reasons, CES is the recommended fluid replacement beverage for children and young adults during prolonged activity (4,7). The efficacy of CES as a fluid replacement beverage for older active men and women during or after exercise-heat stress has not been previously studied.

The purpose of the present study was to compare the voluntary fluid intake behavior of older (defined here as 54–70 yr of age) active men and women when provided ample opportunity to drink during rest periods of a moderate-intensity interval cycling protocol in the heat. A second goal was to compare the efficacy of two replacement fluids, CES and water, in maintaining fluid/electrolyte balance, restoring plasma volume, and minimizing thermoregulatory strain associated with exercise-heat stress in older men and women.



Twenty-seven healthy older recreational exercisers (13 men, 14 women, 54–70 yr) volunteered to participate in this study. Participants were informed of the experimental procedures and associated risks before providing written informed consent. This study was approved by the Institutional Review Board for the Protection of Human Subjects of The Pennsylvania State University. Preliminary screening included a resting 12-lead electrocardiogram, skinfold measurements to determine adiposity, blood analysis (CHEM-24), a graded exercise test on a cycle ergometer to determine peak oxygen uptake (V̇O2peak), and a physical exam by a physician. Criteria for exclusion included an abnormal resting or exercise electrocardiogram, smoking, diabetes, coronary artery disease, inactive lifestyle, elite athlete, or the taking of medications that may influence thermoregulatory or cardiovascular variables of interest. Volunteers with medical conditions or on medications not mentioned in the exclusion criteria and that did not pose a safety risk were included in the study. Subject characteristics are presented in Table 1.

Subject characteristics.

Experimental procedure.

All subjects completed two experimental trials in which they consumed either a commercially available CES (6% carbohydrate and 18.0 mmol·L−1 NaCl) or distilled water. During an additional trial, 14 subjects repeated one of the two fluids to determine the reproducibility of fluid intake. Experimental trials were scheduled at least 1 wk apart, and trials were assigned in random order.

Subjects reported to the laboratory on the morning of test days after having fasted overnight. Immediately upon arrival, they were asked to complete Survey 1, a visual-analog rating scale with questions pertaining to subjective feelings of physical and psychological well-being (described later). After collecting a urine sample, emptying their bladder, and being weighed, subjects had an 18-gauge Teflon catheter placed in an antecubital vein in one arm. Next, the subject inserted a rectal thermistor 8 cm past the anal sphincter and then entered an environmental chamber set at 30°C and 50% relative humidity. The subject, wearing shorts and a t-shirt, then mounted a cycle ergometer (Monark Ergomedic 818E) and relaxed in a seated position for approximately 15 min while being instrumented with four skin thermocouples (chest, upper arm, thigh, calf), a Polar® heart rate monitor, and blood pressure cuff. A modification was made to the bike seat, which included a backrest and retractable padded seat that slid under the subject for the rest periods and then pulled out of the way during the exercise periods. This padded seat allowed for improved comfort during the baseline, rest, and recovery periods without having to change body posture by getting off the bike to sit on a chair.

After instrumentation, the subject sat at rest for a 15-min baseline period. Next, the subject cycled at an intensity of 65 ± 1% of V̇O2peak for 15 min, followed by 15 min of rest during which subjects were allowed to drink ad libitum. This interval cycling protocol continued until the subject completed four 15-min bouts of cycling, separated by 15-min rest periods. After completing the fourth exercise bout, the subject remained in a seated position on the bike and was permitted to drink ad libitum during a 30-min recovery period. The subject was asked to complete Survey 2 (which included questions about beverage satisfaction and feelings of physical well-being) at the end of the second rest period (minute 70) and then again during the recovery period (minute 130).

At the end of the 2.5-h interval cycling protocol, the subject exited the chamber, and postexperiment body mass (BM) and then a urine sample were obtained. Finally, the subject was asked to again complete the first survey before leaving the laboratory. A schematic of the study protocol is diagrammed in Figure 1.

FIGURE 1— Schematic of study protocol for each trial. During exercise periods 1–4, subjects cycled at 65% V̇O2peak. Subjects sat upright on the bike seat for baseline, rest 1–3, and recovery. Base, baseline; BP, blood pressure; BS, blood sample; BM, body mass; Exer, exercise; HR, heart rate; RPE, rating of perceived exertion; US, urine sample; V̇O2, oxygen consumption.

Ad libitum fluid intake.

During the rest periods and recovery, a bottle containing 700 mL of 15°C fluid (CES or water) randomly assigned to that given trial was provided to the subjects while they sat quietly at rest. Subjects were not encouraged to drink but were told that more fluid was readily available if needed. Volume of fluid intake was measured to the nearest milliliter after rest periods 1–3 and recovery. The subjects were unaware that their fluid consumption was being measured.


Rectal and skin temperature (Tre and Tsk, respectively) were measured continuously throughout the protocol and recorded as 1-min averages. HR was measured using a Polar® heart rate monitor, and blood pressure was measured by brachial auscultation (sphygmomanometry). Ratings of perceived exertion (RPE) were assessed using the Borg scale (5). Oxygen uptake was measured (ParvoMedics TrueOne 2400 Metabolic Measurement System) 5 min into the second exercise bout to verify that the subject was exercising at the target workload of 65% of V̇O2peak. Pre- and postexperiment BM was measured to the nearest 0.1 kg on a Toledo scale.

Blood and urine analysis.

Venous blood samples (11 mL each) were drawn without stasis. A 2.5-mL aliquot was transferred into an EDTA-treated test tube and immediately analyzed for hematocrit (Hct) and hemoglobin (Hb) in duplicate using a Beckman Coulter Microdiff 16. The remaining 8.5-mL aliquot was transferred into a serum separator tube, allowed 30–60 min to clot, and then centrifuged at 4°C for 15 min. Serum was analyzed for glucose concentration (Sgluc; hexokinase UV method, Olympus Model AU5200), sodium concentration (SNa; ion specific electrode method, Olympus Model AU5200), total protein concentration (Sprot; biuret method, Olympus Model AU5200), and osmolality (Sosmol; freezing point depression, Advanced DigiMatic Osmometer Model 3D2) in duplicate. Pre- and postexperiment urine osmolality (Uosmol) and specific gravity (Usg; Refractometer, Atago A300CL) were also determined in duplicate.


Tsk was calculated as the weighted sum of four sites: chest (Tch), upper arm (Ta), thigh (Tth), and calf (Tleg) using the following equation (20):

Mean arterial pressure (MAP) was calculated as MAP = [1/3] pulse pressure + diastolic BP. Total body sweat loss was calculated from the net ΔBM corrected for fluid consumed and urine excreted. The percent change in plasma volume from the baseline period (ΔPV) was calculated from hematocrit and hemoglobin (11):

Subjective ratings.

Survey 1 was administered to the subject before and after the experiment. Subjective feelings of physical and psychological well-being were assessed by visual analog rating scales. Subjects responded to the following questions, “How alert do you feel?,” “How sad do you feel?,” “How tense do you feel?,” “How much of an effort is it to do anything?,” “How happy do you feel?,” “How weary do you feel?,” “How calm do you feel?,”, and “How sleepy do you feel?” on 100-point scales with responses ranging from “very little” (0) to “very much” (100).

Survey 2 was completed at minutes 70 and 130 to determine subjects’ overall perception of the beverage consumed during the trial. The subject was instructed to mark an “X” in the box that best described their “overall acceptance” of the beverage and their opinion of its “flavor” and “aftertaste.” The questions were on a nine-point scale in which possible responses ranged from “dislike extremely” (1) to “like extremely” (9). This questionnaire also included visual analog scales that measured perceived intensities of overall flavor (“weak overall flavor” to “strong overall flavor”), sweetness (“not at all sweet” to “very sweet”), saltiness (“not at all salty” to “very salty”), off-flavor (“no off-flavor” to “strong off-flavor”), aftertaste (“weak aftertaste” to “strong aftertaste”), and whether the beverage was thirst quenching (“not thirst quenching” to “very thirst quenching”). In addition, Survey 2 included visual analog scales that rated feelings of hunger (“not hungry” to “very hungry”), hotness (“not feeling hot/overheated” to “feeling very hot/overheated”), muscle fatigue (“no muscle fatigue” to “severe muscle fatigue”), difficulty of exercise (“exercise very easy” to “exercise very difficult”), burping (“no burping” to “severe burping”), stomach fullness (“no stomach fullness” to “severe stomach fullness”), and stomach upset (“no stomach upset” to “severe stomach upset”). The subject answered these questions by placing a mark on a 100-point scale between the extreme answers at the opposite ends of the line.

Statistical analysis.

An intraclass correlation coefficient was used to test the reproducibility of fluid intake (22). A Student’s t-test was used to determine any significant differences between sexes in the subjects’ physical characteristics. Variables that were measured through time (e.g., Tre, Tsk, HR, BP, RPE, blood variables, and responses to subjective questionnaires) were analyzed using two-way analysis of variance (ANOVA) (fluid vs time) with repeated measures. A one-way ANOVA was used to compare variables measured once per test (e.g., total fluid intake, ΔBM, ΔUosmol, ΔUsg). To detect any association of a subject’s sex with the physiological and subjective variables measured throughout the protocol, a two-way (fluid vs sex) ANOVA or three-way (fluid vs time vs sex) repeated measures ANOVA was performed where appropriate. The Tukey post hoc test was performed when main effects were found. The significance level for all statistical tests were set to alpha = 0.05. All data are presented as means ± SE.


Physical characteristics of the participants are shown in Table 1. The men had a significantly greater BM, height, and V̇O2peak than the women, whereas women had a significantly higher adiposity than the men.

In the subjective evaluation of the fluids (Fig. 2) men rated the “overall acceptance” and “flavor” of CES higher than water. By contrast, women rated the “overall acceptance,” “flavor,” and “aftertaste” of water higher than that of CES. CES was rated higher by the men than women for “overall acceptance” and “flavor,” whereas the “flavor” of water was rated higher by the women compared to the men. Also, women gave higher ratings than the men for “thirst quenching” when consuming water. The men felt significantly more “hot/overheated” and perceived “exercise difficulty” to be significantly higher than did the women during the water trials.

FIGURE 2— Responses to subjective questionnaires on 9-point (A) and 100-point (B) scales. *:
P < 0.05, CES vs water; † P < 0.05, men vs women.

Across all trials, the reproducibility of total fluid intake was very good (intraclass correlation coefficient = 0.75) (21). Figure 3 presents the total fluid intake, sweat loss, and net fluid balance during the CES and water trials for men, women, and all subjects. Total fluid intake was significantly higher during the CES trials (all subjects). Men drank significantly more CES than water, whereas total intake did not differ significantly between the two fluids in women. Women were in positive fluid balance (increased BM from baseline) during both the CES (P < 0.05) and water trials, whereas men were only in positive fluid balance with CES. When total fluid intake was expressed relative to BM, women drank significantly more water than the men.

FIGURE 3— Fluid intake, net fluid balance, and sweat loss expressed in absolute units (A–C) and relative to BM (D–F). *:
P < 0.05, CES vs water; † P < 0.05, vs baseline; ‡ P < 0.05, men vs women.

Figure 4 illustrates changes in ΔPV (panel A), Sprot (panel B), and SNa (panel C) during the CES and water trials for men, women, and all subjects. During the CES trials plasma volume (PV) was at or above baseline from the third rest period to the end of recovery (all subjects). However, during the water trials, PV was significantly below baseline at this time (all subjects). PV decreased significantly less in the women compared with the men near the end of the protocol in both fluid trials (panel A). During the second half of the protocol Sprot was significantly greater during the water trials versus the CES trials for all subjects (panel B). SNa was significantly below baseline near the end of the protocol in both fluid trials (women and all subjects). Women had a significantly lower SNa than the men during the final exercise bout of the CES trials and from the final exercise bout to the end of the recovery period of the water trials (panel C).

FIGURE 4— Hematological variables in men, women, and all subjects. A. Percent change in plasma volume from baseline (ΔPV); B. serum protein concentration (Sprot); C. serum sodium concentration (SNa). *:
P < 0.05, CES vs water; † P < 0.05, vs baseline; ‡ P < 0.05, men vs women. Pre, preexperiment (supine, thermoneutral); Base, baseline (sitting upright, 30°C, 50% RH); Ex, exercise; Rec, recovery.

Fluid intake per kilogram of BM, Tre, Tsk, Sgluc, Sosmol, BP, HR, and RPE are presented in Table 2. Subjects (all) consumed significantly more CES than water during the first, second, and third rest periods but not during recovery. Women drank significantly more water than the men during the first and second rest periods. There were no differences between fluids in the Tre responses; however, during the water trial the women had a significantly lower Tre than the men from the second exercise period to the end of recovery. Sgluc was significantly higher during the CES trials from the second bout of exercise to the end of recovery (all subjects). There were no significant differences between fluids or sexes in Tsk, Sosmol, BP, HR, RPE, Uosmol (CESpre = 655 ± 38, CESpost = 608 ± 36, waterpre = 604 ± 45, waterpost = 546 ± 38 mosmol·kg−1), or Usg (CESpre = 1.018 ± 0.001, CESpost = 1.018 ± 0.001, waterpre = 1.017 ± 0.001, waterpost = 1.016 ± 0.001 UG) throughout the experiment. Sosmol, Uosmol, and Usg did not change from baseline in men, women, or all subjects.

Fluid intake.

Figure 5 presents the SNa of one woman (65 yr, 45.7 kg) who became hyponatremic (SNa = 126 mmol·L−1) with symptoms (e.g., headache, extreme fatigue, and gastrointestinal distress) during the water trial. Her SNa was 126 mmol·L−1 after consuming 2.8 L of water, but she did not experience symptomatic hyponatremia during the CES trial (SNa = 131 mmol·L−1) in which she drank 2.7 L of CES.

FIGURE 5— Serum sodium concentration in a 65-yr-old, 45.7-kg woman during the CES and water trials. She was hyponatremic (SNa = 126 mmol·L−1) with symptoms by the end of the water trial, but maintained SNa (131 mmol·L−1) above symptomatic hyponatremia when consuming CES. The:
dashed line at 130 mmol·L−1 represents the upper limit for clinical hyponatremia. Volume of total fluid intake for each trial is shown in the legend. Pre, preexperiment (supine, thermoneutral); Base, baseline (sitting upright, 30°C, 50% RH); Exer, exercise; Recov, recovery.


The major findings from this study were: 1) when cool palatable fluids were readily available, active adults aged 54–70 yr drank enough to match sweating rates and maintain their body mass; 2) their fluid intake behavior was repeatable; 3) CES promoted greater voluntary fluid intake and restored PV losses faster than water; and 4) there were sex differences in the fluid intake behavior of older active adults, with women drinking more water per kilogram of BM than men.

Combined data from all subjects.

The present investigation indicates that older adults drink enough fluid to replace sweat losses incurred during moderate-intensity interval exercise in the heat. Furthermore, this drinking behavior was repeatable. The subjects did not experience voluntary dehydration most likely because they had a cool palatable fluid readily available and had ample opportunity to drink between exercise bouts. The lack of voluntary dehydration might also be explained by the low average sweat rate (370 g·h−1 over the entire protocol). In 1947, Rothstein et al. (21) reported that when sweat losses are low (defined as <400 g·h−1) young men working in the heat have little difficulty consuming sufficient fluid to replace periodic losses. The current study suggests that older men and women also drink enough fluid to match sweat losses when sweat rate is <400 g·h−1. Although subjects drank enough to maintain BM during both fluid trials, CES was more effective at stimulating voluntary fluid intake than water, especially in men. As a result of the significantly greater consumption of CES, PV losses were restored faster during the CES trials compared with the water trials. In addition, Sprot was lower in the CES versus the water trials during the second half of the protocol. This indicates a greater hemodilution when CES was consumed, which supports the conclusion that CES more effectively restored extracellular fluid volume. These data are consistent with other studies showing that CES is more effective than plain water in promoting fluid intake and restoring PV losses in younger exercising subjects (6,12,17).

Sex differences.

Both men and women consumed more CES than water (but not significantly more by the women). However, significant sex-related differences were detected for the fluids’ subjective values. Men rated CES higher than water for “overall acceptance” and “flavor.” Conversely, women rated water higher than CES for these categories. In addition, women perceived water to be more “thirst quenching” compared with the men. These differences in the subjective evaluation of the fluids explain why women consumed more water per kilogram of BM compared with the men. The greater fluid intake by the women resulted in a lower Tre than that of men during the final 90 min of the water trials. The sex difference in Tre corresponded with a sex difference in the ratings of how “hot/overheated” the subjects felt and their perception of “exercise difficulty,” that is, the women’s ratings were lower than that of the men’s in both categories during the water trials.

During the water trials, the SNa of the women was lower than that of the men during the final 30 min of the protocol because the women drank significantly more water relative to BM compared with the men. However, the average SNa of the women at the end of recovery (138 ± 1 mmol·L−1) was still within the normal range of 136–142 mmol·L−1. Hyponatremia is defined as SNa less than 135 mmol·L−1 and can be categorized as asymptomatic (usually 130–134 mmol·L−1) and symptomatic (usually <130 mmol·L−1) hyponatremia (24). Common symptoms include headache, light-headedness, confusion, fatigue, gastrointestinal distress, nausea, and malaise (23). In the present study, one woman became clinically hyponatremic (SNa = 126 mmol·L−1) with symptoms by the end of the protocol during the water trial. This subject (65 yr, 45.7 kg) drank 2.8 L of water and gained 2.4 kg of weight during that experiment. She complained of a headache and extreme tiredness near the end of the water trial. In addition, at the end of the recovery period, she rated the following assessments of physical well-being much higher during the water trial in comparison to the CES trial: “How much of an effort is it to do anything?” (51.7 vs 11.7), “How weary do you feel?” (81.4 vs 14.5), and “How sleepy do you feel?” (73.1 vs 15.2). She also rated her degree of “stomach upset” (22.4 vs 4.2), “stomach fullness” (45.5 vs 6.3), and “burping” (25.2 vs 6.3) higher during the water trial, which indicates that she may have experienced gastrointestinal distress at this time.

Anecdotal evidence suggests that hyponatremia most commonly occurs in women, especially those with low body weight and who drink large volumes of sodium free fluids (i.e., water) before, during, and after physical activity (2,3,10,24). A retrospective analysis of symptomatic hyponatremia in marathon runners by Davis et al. (10) showed that this condition develops in older as well as younger women. Six of 23 hyponatremic female marathon runners in the analysis were 50 yr or older. The present study provides supporting evidence that the overconsumption of water may be a health concern among smaller women in this age range. Furthermore, this woman’s data support the notion that a CES is superior to water in limiting reductions in SNa during exercise-heat stress. During the CES trial, this female subject consumed 2.7 L and had a final SNa of 131 mmol·L−1. Therefore, although she consumed similar amounts of CES and water, SNa was maintained above that of symptomatic hyponatremia during the CES trial.

Fluid intake differences between men and women are most likely due to differences in behavior and not in thirst perception. There is no evidence to suggest the existence of a physiological mechanism by which thirst perception differs between older men and women; therefore, it may be that women are more health and safety conscious than men and thus make a greater effort to drink, and in some cases overdrink, the fluid that is available.

In summary, voluntary dehydration does not occur in older adults when cool palatable fluid (either CES or water) is readily available between repeated bouts of moderate-intensity exercise in the heat. In addition, CES promotes greater voluntary fluid intake and restores PV losses faster than water during interval cycling, suggesting that CES is the more effective fluid replacement beverage for older active adults. Furthermore, older women drink more CES and water than is lost through sweating. Overconsumption of water may put smaller women at an increased risk for developing hyponatremia.


1. Armstrong, L. E., W. C. Curtis, R. W. Hubbard, R. P. Francesconi, R. Moore, and E. W. Askew. Symptomatic hyponatremia during prolonged exercise in the heat. Med. Sci. Sports Exerc. 25:543–549, 1993.
2. Ayus, J. C., J. Varon, and A. I. Arieff. Hyponatremia, cerebral edema, and non-cardiogenic pulmonary edema in marathon runners. Ann. Intern. Med. 132:711–714, 2000.
3. Backer, H. D., E. Shopes, and S. L. Collins. Hyponatremia in recreational hikers in Grand Canyon National Park. J. Wilderness Med. 4:391–406, 1993.
4. Bar-Or, O., and B. Wilk. Water and electrolyte replenishment in the exercising child. Int. J. Sport Nutr. 6:93–99, 1996.
5. Borg, G. Perceived exertion: a note on “history” and methods. Med. Sci. Sports Exerc. 5:90, 1973.
6. Carter, J. E., and C. V. Gisolfi. Fluid replacement during and after exercise in the heat. Med. Sci. Sports Exerc. 21:532–539, 1989.
7. Convertino, V. A., L. E. Armstrong, E. F. Coyle, et al. American College of Sports Medicine position stand: exercise and fluid replacement. Med. Sci. Sports Exerc. 28:I–vii, 1996.
8. Costill, D. L., and K. E. Sparks. Rapid fluid replacement following thermal dehydration. J. Appl. Physiol. 34:299–303, 1973.
9. Costill, D. L., W. F. Kammer, and A. Fisher. Fluid ingestion during distance running. Arch. Environ. Health 21:520–525, 1970.
10. Davis, D. P., J. S. Videen, A. Marino, et al. Exercise-associated hyponatremia in marathon runners: a two-year experience. J. Emerg. Med. 21:47–57, 2001.
11. Dill, B. D., and D. L. Costill. Calculations of percentage changes in volumes of blood, plasma, and red cells in dehydration. J. Appl. Physiol. 37:247–248, 1974.
12. Gonzalez-Alonso, J., C. L. Heaps, and E. F. Coyle. Rehydration after exercise with common beverages and water. Int. J. Sports Med. 13:399–406, 1992.
13. Kenney, W. L., and P. Chiu. Influence of age on thirst and fluid intake. Med. Sci. Sports Exerc. 33:1524–1532, 2001.
14. Mack, G. W., C. A. Weseman, G. W. Langhans, H. Scherzer, C. M. Gillen, and E. R. Nadel. Body fluid balance in dehydrated healthy older men: thirst and renal osmoregulation. J. Appl. Physiol. 76:1615–1623, 1994.
15. Murray, R. Rehydration strategies–balancing substrate, fluid, and electrolyte provision. Int. J. Sports Med. 19:(Suppl. 2):S133–S135, 1998.
16. Nadel, E. R., S. M. Fortney, and C. B. Wenger. Effect of hydration state on circulatory and thermal regulations. J. Appl. Physiol. 49:715–721, 1980.
17. Nose, H., G. W. Mack, X. Shi, and E. R. Nadel. Role of osmolality and plasma volume during rehydration in humans. J. Appl. Physiol. 65:325–331, 1988.
18. Phillips, P. A., M. Bretherton, C. I. Johnston, and L. Gray. Reduced osmotic thirst in healthy elderly men. Am. J. Physiol. 261:R166–R171, 1991.
19. Phillips, P. A., B. J. Rolls, J. G. Ledingham, et al. Reduced thirst after water deprivation in healthy elderly men. N. Engl. J. Med. 311:753–759, 1984.
20. Ramanathan, N. L. A new weighting system for mean surface temperature of the human body. J. Appl. Physiol. 19:531–533, 1964.
21. Rothstein, A., E. F. Adolph, and J. H. Wills. Voluntary dehydration. In: Phsyiology of Man in the Desert. Adolph and Associates (Eds.). New York: Interscience Publishers, 1947, pp. 254–270.
22. Shoukri, M. M., and C. A. Pause. Statistical analysis of reliability measurements. In: Statistical Methods for Health Sciences, 2nd Ed. Boca Raton: CRC Press, 1999, pp. 20–28.
23. Speedy, D. B., T. D. Noakes, and C. Schneider. Exercise-associated hyponatremia: a review. Emerg. Med. 13:17–27, 2001.
24. Speedy, D. B., T. D. Noakes, I. R. Rogers, et al. Hyponatremia in ultradistance triathletes. Med. Sci. Sports Exerc. 31:809–815, 1999.
25. Stachenfeld, N. S., L. DiPietro, E. R. Nadel, and G. W. Mack. Mechanism of attenuated thirst in aging: role of central volume receptors. Am. J. Physiol. 272:R148–157, 1997.


©2005The American College of Sports Medicine