Exercise associated hyponatremia was originally described as a common complication of competi-tive ultradistance running and triathlon events (7–9,17,20,23,26,28,30). But recently it has become more common even in female marathon runners (4) and in military personnel (13).
We (28) have previously shown that this condition occurred in 23% of athletes who sought medical care after an ultradistance triathlon; others report incidences varying between 10 and 44%(9,10,23,28). Excessive sodium loss in the sweat has been postulated as the cause of exercise associated hyponatremia (7), although there are no published data on the actual magnitude of sweat sodium loss during an Ironman triathlon. Furthermore, the only two studies that have investigated sodium losses in hyponatremic athletes during ultradistance exercise have reported no significant differences between the sodium losses of hyponatremic athletes and normonatremic controls (11,31). Both of these studies have estimated sodium losses indirectly by measuring sodium balance and deficit in athletes during overnight recovery after an ultradistance race. The mean durations of the postrace recovery periods in these studies were 12.0 h (11) and 12.4 h (13). Irving et al. (11) reported a median sodium loss of 153 mmol in hyponatremic athletes; Speedy et al. (31) reported median sodium losses of 88 mmol. Speedy et al. (29) have also reported sodium losses of 20 and 24 mmol in two hyponatremic triathletes, using similar methodology of studying sodium balance during overnight recovery from hyponatremia. These losses are much less than the sodium deficits (260–1500 mmol) postulated necessary to cause hyponatremia (7,17,20,29). Although there is no evidence for increased sodium losses in hyponatremic athletes, mild hyponatremia has been reported in the presence of dehydration (28), raising the possibility that in some athletes at least, sodium losses (or insufficient sodium intake) may be involved in the etiology of hyponatremia. The bulk of published evidence demonstrates that athletes who develop symptomatic hyponatremia have evidence of fluid overload (11,17,19,27–29,31) without abnormally large sodium losses (from measurement of sodium deficit in recovery as discussed above) (11,29,31). It was initially postulated that fluid excess causes hyponatremia by inducing movement of sodium from the extracellular space into a volume of unabsorbed fluid in a physiological third space in the intestinal lumen (1,15). This would explain the finding that (i) plasma volume was reduced in 90 km ultramarathon runners with hyponatremia (11) and (ii) why the renal response to this fluid excess is clearly inappropriate as an effective diuresis begins only many hours after the cessation of exercise (11).
However, we have recently reported a significant correlation between postrace serum sodium concentrations and postrace hematocrit in athletes completing an ultradistance triathlon, suggesting that the excess fluid in this condition is retained in the extracellular fluid (ECF) rather than in the intestinal lumen (28). If this is correct, the absence of an adequate diuresis in the face of such marked whole body fluid overload remains perplexing.
Most reports indicate that athletes who develop hyponatremia ingest either excessively large (>1200 mL·h−1) (1,6,11,19,26) or what might be considered more modest (<800 mL·h−1) amounts of fluid leading to fluid overload (19,29). It remains to be established why the renal response to fluid retention and expansion of the ECF is inappropriate in these athletes. An inappropriate arginine vasopressin (AVP) response has been postulated by some (2,11,32,33). However, there is only one report of elevated plasma AVP concentrations in a subject who developed exercise-associated hyponatremia while drinking 1900 mL·h−1 over 4 h (1); this study was completed in a laboratory. There have only been two other reports of AVP concentrations in athletes with hyponatremia: Speedy et al. (28) measured normal plasma AVP concentrations in a large group of triathletes who developed hyponatremia during an Ironman triathlon. (Fluid intakes were not measured in this study.) Speedy et al. (29) have also reported AVP concentrations in the low normal range in a case report of two ultradistance triathletes competing in an Ironman. (Fluid intake during the race was only 733 and 764 mL·h−1 in these two athletes, although they had evidence of fluid overload despite this seemingly modest fluid intake.) Normal AVP concentrations would not explain why diuresis was delayed in these athletes. Impaired renal excretion of this fluid load from other causes, such as exercise-induced renal ischemia or in response to other circulating neurochemicals, remains a possibility. Accordingly, the following questions remain to be answered if the pathophysiology of this condition is to be better understood.
First, what mechanisms explain why athletes who develop fluid overload (defined as retention of fluid during exercise with features of weight gain and a postexercise diuresis) during exercise are unable to excrete that fluid excess? Second, is this a unique pathophysiological abnormality specific only to athletes who develop exercise-associated hyponatremia, or are all athletes equally at risk of developing hyponatremia during exercise if they ingest large fluid volumes at sufficiently rapid rates? That is, will the hyponatremia of exercise develop in any athlete provided he or she ingests fluid at a sufficiently high rate for sufficiently long?
To further investigate these possibilities, we have evaluated a laboratory model for developing hyponatremia under resting conditions. The main aim of the present study was to determine whether only athletes who had previously developed exercise-associated hyponatremia during an ultradistance triathlon uniquely show an impaired ability to excrete a large fluid load compared with a normal response in subjects who had completed the same race without developing hyponatremia. We also used indirect measures to give an indication whether there was any evidence for delayed fluid absorption in the small bowel causing hyponatremia as the result of a possible reverse sodium movement between the extracellular space and unabsorbed fluid in the small bowel (16).
Six athletes who developed symptomatic hyponatremia (postrace serum sodium concentration [Na] 128–134 mmol·L−1) during the 1997 New Zealand Ironman Triathlon (3.8-km swim, 180-km cycle, 42.2-km run) were identified from our previous research (28) and invited to participate in this study (study “cases”). Another six athletes who completed the same race without developing hyponatremia ([Na] 135–141 mmol·L−1) (28) served as study “controls.” Athletes gave written informed consent for participation in this study, which was approved by the North Health Ethics Committee. Comparative details of the two groups are provided in Table 1. Athletes were instructed to avoid heavy training on the day immediately before testing. They were instructed to fast from midnight until 0700 h on the day of the test when they consumed a standardized breakfast with a standard fluid intake (0.20 L water, 213 mg Na, 357 kcal). Despite this advice, one control subject consumed a breakfast with 0.25 L of fluid and a sodium content of 190 mg; three other control subjects each consumed only 0.11 L of water during the standardized breakfast. These variations were, however, considered to be inconsequential. All subjects presented for laboratory testing at 0900 h.
Athletes had an indwelling catheter inserted into a large vein in the antecubital fossa. The catheter was kept patent with a flush of 2 mL of 0.9% saline solution every 30 min. Blood specimens were taken from this catheter after withdrawing and discarding 2 mL of fluid to minimize the effect of contamination from the saline flush. Blood specimens were drawn with the athletes in the supine position and whenever possible without the use of a tourniquet. Athletes lay supine for sufficient time for a venipuncture to be performed (this was not formally timed, although estimated to be approximately 4 min, although there were several difficult venipunctures for which the athlete was supine for approximately ten min). Beginning at 0930 h, athletes consumed 400 mL of cold water and then 250 mL every 10 min for 2 h (total of 3.4 L). This intake of 1500 mL·h−1 is less than that described in some case reports of hyponatremia during racing (17,27). However, fluid losses at rest are much less than during exercise; hence, we postulate that our model of fluid intake approximates typical net fluid balance (fluid intake minus fluid loss) in athletes who develop exercise-associated hyponatremia.
Immediately before consuming the initial fluid bolus and at 30-min intervals thereafter, athletes emptied their bladder, were weighed, and had a blood specimen drawn. Athletes were weighed on digital chair scales (AVD FV-D, Atrax, Tokyo, Japan) at 30-min intervals. Urine volume was estimated by weight using a standard top-weigh laboratory balance (Sartorius 2357, Göttingen, Germany).
Blood was analyzed for hemoglobin, hematocrit (Gen-S, Beckman-Coulter, Miami, FL), osmolality (Model 3D3 freezing-point depression osmometer, Advanced Instruments, Norwood, MA), sodium, albumin, urea, creatinine, (Boehringer-Hitachi 717, Mannheim, Germany) and chloride (ABL 625, Radiometer, Copenhagen, Denmark).
Plasma volume changes were calculated using the formula derived from Dill and Costill (5). Urine volume was measured every 30 min and analyzed for sodium and chloride concentration (ABL 625) and osmolality (freezing-point depression osmometer). Symptoms of hyponatremia were documented as they developed. Athletes did not eat until 2.5 h after the beginning of testing at which time they were allowed to eat food ad libitum.
Intracompartment fluid shifts were calculated according to the following method: surface area (AD) in m2 was predicted from the subject’s body mass in kg and height in cm with the equation of Du Bois and Du Bois (19) : AD = 0.00718 cm0.725 × kg0.425. Plasma volume (PV) was estimated from the subject’s surface area with the equation of Retzlaff et al (24) : PV = 1.63.AD. Changes in plasma volume were calculated as described above. Shifts in interstitial fluid (ΔISF) and intracellular fluid (ΔICF) volumes were calculated from Cl− movements (ΔCl−), falls in plasma volume (ΔPV), and total body water losses (ΔTW), as shown below:
In these equations, it is assumed that (a) Cl− losses in sweat and urine come only from the ECF, (b) the Donnan equilibrium of Cl− between the ISF and the plasma remains at 0.95, and (c) the loss of Cl− from the ECF is proportional to the water loss. These equations have been validated in studies of rehydration at rest by Nose et al. (21,22) and were therefore considered appropriate for this study which was also conducted under resting conditions. However, these indirect methods were used merely to provide an estimate of any likely difference between groups and were not considered a substitute for more direct methods of measurement of intracompartmental fluid shifts in response to fluid overload.
The two groups were compared using the Mann-Whitney U-test for nonparametric data.
The six hyponatremic cases comprised three male and three female athletes, compared with two female and four male athletes in the control group. All cases and controls were in good health. Current fitness levels were not formally assessed, although all cases and controls reported that they were either currently competing in triathlons or still fit enough to do so. There was no statistically significant difference between the two groups in median body weight (Table 1) or in their finishing times in the 1997 New Zealand Ironman Triathlon. However, postrace serum sodium concentrations were significantly lower in the experimental cases (P = 0.005), as dictated by the research design. No case or control had suffered from exercise associated hyponatremia before the 1997 Ironman Triathlon. Median previous Ironman competition experience before the 1997 race was 0 races (range 0–4) for cases and 1.5 races (range 0–4) for controls. Although the controls had a higher median age, there was no significant difference in age between cases and controls. The two groups were therefore considered similar: it is unlikely that any difference in previous Ironman experience would alter an athlete’s response to a fluid load. Certainly ultradistance racing experience has never been investigated as a risk factor for hyponatremia. In a multivariate analysis, a young age has been shown to place an athlete at increased risk of hyponatremia (28), although the difference in ages between the groups in our study was not statistically significant.
Figure 1 shows the rates of fluid intake and fluid excretion in all subjects. Rates of fluid intake were not different between cases and controls. Nor was there a difference in cumulative urine volume between cases and controls. The median maximal rate of urine production was 1043 mL·h−1 (range 370–1334) in cases compared with 878 mL·h−1 (range 714–1106) in controls. The time to reach the maximal rate of urine production was 2 h in all cases, and the median was 2 h (range 1.5–2.5) in controls. These values were not statistically different. The maximum rates of fluid excretion were substantially behind the rate of fluid intake (1500 mL·h−1). As a result, body weight increased progressively and equally in both groups (Fig. 2; bottom panel) due to a progressive fluid retention that peaked at 2 h, falling progressively thereafter. Body weight peaked in all cases and controls at 2 h after the commencement of fluid consumption, the median maximum weight gain in cases was 2.1 kg (range 1–2.9) and in controls 2.0 kg (range 1.4–2.5). As before, these values were not statistically different.
Consequent to the progressive fluid retention, serum sodium concentrations fell progressively and equally in both groups (Fig. 2; middle panel). During this study, five of the cases and four of the controls developed hyponatremia (serum sodium concentration < 135 mmol·L−1). The commonest reported symptom of hyponatremia in study cases and controls was light-headedness (N = 4), yawning (N = 3), headache (N = 2), and a feeling similar to mild alcohol intoxication (N = 2). The frequencies of the various reported symptoms in relation to the change in serum sodium concentration were presented in Table 2 but are considered too small for statistical analysis.
The median nadir in serum sodium concentration was 133 mmol·L−1 (range 129–135) for cases and 132 mmol·L−1 (range 129–136) for controls. These represented median changes of −5.5 mmol·L−1 (range −3 to 7) for cases and −6.5 mmol·L−1 (range −5 to 9) for controls. The median time to reach the nadir in serum sodium concentration was 2.5 h in both groups (ranges 2–3.5 h and 1.5–2.5 h). None of the above measurements showed any statistically significant difference between the two groups.
Median percent change in plasma volume for subjects compared with controls is shown in Figure 2 (upper panel). There were no significant differences in the response at any time for cases versus controls. There was an inverse correlation between weight change and change in serum sodium concentrations for data from all study cases and controls, (r = −0.68) (Fig. 3). Data on plasma and urine osmolality are presented in Figure 4. Urine and plasma osmolality fell progressively with increasing cumulative fluid intake with time. There was no significant difference between cases and controls.
Table 3 lists the calculated changes in the volumes of fluid in the different body fluid compartments. In both cases and controls, there was a progressive increase in fluid retention in all three body fluid compartments (ISF, ICF, ECF) as calculated by indirect measures (Table 3). The values were not different between cases and controls for these three body fluid compartments. Table 3 also compares the changes in median body weight with calculated changes in the volumes of the ISF, ICF, and ECF. Any discrepancy in the calculated accumulated fluid volumes and the changes in body mass would likely indicate unabsorbed fluid remaining in the intestine.
The model of oral fluid overload used in this study effectively produced hyponatremia at rest in ultradistance triathletes, some of whom had developed symptomatic hyponatremia in the 1997 226-km Ironman Triathlon in New Zealand. More importantly, hyponatremia developed to an equal extent in triathletes who had previously developed hyponatremia during the New Zealand Ironman Triathlon and those who had completed the same race in an equal time but who maintained their serum sodium concentrations within the normal range.
Furthermore, rates of urine production (Fig. 1) and of change in plasma volume, serum sodium concentrations, and in weight (Fig. 2) were not different between cases and controls when they ingested 1500 mL·h−1 for a total of 3.4 L during 2 h.
Thus, the first conclusion of this is that the physiological response to a large fluid overload at rest did not differ between athletes who finished the 1997 Ironman Triathlon with normal serum sodium concentrations and those who developed hyponatremia in the same race.
Accordingly, at least when evaluated at rest, there does not appear to be any unique pathophysiological characteristic that would explain why the study cases developed hyponatremia during the 1997 Ironman Triathlon. Rather, it would seem that both the study cases and the study controls are at risk of developing hyponatremia whenever their rates of fluid intake substantially exceed their rates of fluid loss (Fig. 1). This does not exclude the possibility, however, that exercise alone may either increase or reduce the risk of developing hyponatremia in athletes who ingest excessive volumes of fluid during exercise. It is possible that only during exercise do additional unique characteristics become apparent, such as a delayed and inadequate diuresis, or higher than normal rates of sweat sodium losses, for example, in those who will subsequently develop exercise-induced hyponatremia. Other risk factors such as nonsteroidal antiinflammatory medication use (11), female gender (28), and a slow running speed (17,19,28) may also predispose an athlete to developing exercise associated hyponatremia.
The second expected finding was that the hyponatremia that developed in these subjects at rest clearly resulted from a progressive fluid overload as there was a direct relationship between the extent of the fluid retention and the fall in the serum sodium concentration (Fig. 3). Sodium losses would not have been involved in the etiology of hyponatremia in this study as these athletes were investigated in the resting state with negligible sodium losses, compared with the larger sodium losses that occur with exercise.
It has been postulated that the hyponatremia resulting from fluid overload could develop either as the result of (i) an inability of the kidneys to excrete fluid as fast as it is absorbed from the intestine when fluid is ingested at high rates, or (ii) as a result of reverse sodium movement into unabsorbed fluid in the small intestine (16,17). The latter would occur if the rate of fluid absorption by the intestine was less than the rate of fluid ingestion: sodium would then move from the ECF into the pooled fluid in the intestine. This postulate (18) comes from calculations based on data collected from hyponatremic athletes during the recovery phase after an ultradistance race (11). Noakes (18) has postulated that 4.1 L of fluid and 520 mmol sodium would need to pool in the intestine to explain a drop in serum sodium concentration to 120 mmol·L−1. However, subsequent studies have reported an increase in plasma volume in hyponatremic athletes (28,29) and do not support this postulate that reverse sodium movement into pooled fluid in the gut is an essential factor in the genesis of the hyponatremia of exercise.
The indirect evidence from this study suggests that all the ingested fluid could be accounted for by distribution within the different body pools so that little if any remained in the small bowel (Table 3). If this is indeed correct, then reverse sodium movement did not explain the hyponatremia that developed secondarily to fluid overload at rest in these subjects. Rather, the inability to excrete fluid as rapidly as it was ingested suggests that a limiting maximum rate of diuresis calculated to be 12–15 mL·min−1 (720–900 mL·h−1) (12) was exceeded by the very high rate of fluid ingestion in this study. Indeed, the range of maximum rates of diuresis measured in these subjects (714–1106 mL·h−1) closely matches these calculated peak rates (12).
Renal ischemia has been postulated as a cause for the failure of the diuresis seen in exercise associated hyponatremia (11). However, this mechanism would not have played a role in these healthy athletes who were studied at rest. Rather, it is possible that, during exercise, hormonal influences could play a permissive role in the development of hyponatremia in those who ingest large fluid volumes. An inappropriate AVP response (from the stresses of exercise) has been postulated as a cause of hyponatremia as discussed above, although the bulk of published evidence does not support this mechanism. Aldosterone also does not appear to play a role as normal or low concentrations have been reported in hyponatremic athletes (28). It is possible that some other hormone (as yet unstudied or unidentified) may play a role in the failure of diuresis in hyponatremic athletes.
Thus, a third, albeit more speculative finding of this study, was that the hyponatremia resulted principally from dilution of sodium in the ECF and not as a result of pooled fluid in the intestine. If this is correct, it indicates that under resting conditions, the capacity of the intestine to absorb fluid can exceed the maximum renal diuretic response calculated as 720–900 mL·h−1(12) and found by us to be between 700 and 1100 mL·h−1. This finding is compatible with our previous finding (28) that plasma sodium concentrations after an Ironman triathlon are significantly related to postrace hematocrit, suggesting that hyponatremia results from an expansion of the ECF with dilution of the sodium content in this space. Furthermore, a case report of two hyponatremic athletes (9) reported a large increase (16–25%) in plasma volume during an Ironman triathlon. All these findings are consistent with the postulated mechanism of serum sodium dilution as a result of expansion of the ECF volume, consequent to fluid retention resulting from an inadequate renal response to whole body fluid overload. Several case reports of exercise associated hyponatremia have described the phenomenon of an athlete with hyponatremia clinically deteriorating in the first hour of two after completing an ultradistance race (1,2,6). This has been explained on the basis of reverse sodium movement into pooled fluid in the bowel during exercise, with subsequent reabsorption of the hypotonic fluid pooled in the bowel when exercise ceases (1,2). However, this postulate is unlikely on the basis of the published evidence, as discussed above. An alternative possibility is that an athlete may continue to drink postexercise exacerbating the hyponatremia. Elevated AVP concentrations are not responsible for this phenomenon as these are not elevated postexercise in hyponatremic athletes (28,29,31), although other as yet unidentified hormones may be involved.
Finally, we found that the symptoms reported by the cases and controls who developed hyponatremia were similar to those described in the literature (28) and included light-headedness and headache. In addition, two subjects reported symptoms similar to those experienced with mild alcohol intoxication. Yawning was noted in three subjects with hyponatremia: this is not a recognized symptom of hyponatremia. The frequency of symptoms and number of subjects in this study were too few to perform a statistical analysis on symptoms. It is interesting that the symptoms developed at relatively mild levels of hyponatremia although the rate of change of serum sodium concentration was high (Fig. 2). In cases of hyponatremia that are not exercise related, the rate of onset of the hyponatremia determines the severity of symptoms, with chronic hyponatremia producing fewer symptoms than acute hyponatremia (14). Similarly, we speculate that the rate of fall in serum sodium concentration may also be an important determinant of the extent to which those with exercise-related hyponatremia develop symptoms. Symptoms of severe hyponatremia are thought to be caused by cerebral edema (3,14,17). Although it is unlikely that any athletes in our study had significant cerebral edema, we speculate that the symptoms reported may relate either to subtle fluid shifts between intracellular and extracellular fluid compartments (Table 3) or to changes in neurological function secondary to the low serum sodium concentration.
We observed that athletes who had developed hyponatremia during a previous ultradistance triathlon and other athletes who had remained normonatremic in the same race have a similar incidence of hyponatremia when drinking large volumes of fluid (3.4 L) while resting for 2 h. Both groups excreted the large fluid overload in precisely the same way and developed an equivalent hyponatremia because their maximum rates of renal fluid excretion were less than their rates of fluid ingestion and, probably, less than their peak rates of intestinal fluid absorption.
Hence, there does not appear to be any unique pathophysiological characteristic, at least when evaluated at rest, that explains why the study cases developed hyponatremia during ultradistance exercise. Rather, the findings suggest that hyponatremia will develop whenever there is a progressive fluid overload, as concluded from our previous studies of this condition (25,28). However, this does not exclude the possibility that, during exercise, there may be additional factors, such as an inappropriate inhibition of diuresis, unusually high rates of sweat sodium losses or low fluid losses in sweat, inappropriate hormonal responses, or preexisting illness that would increase the probability that hyponatremia will develop in those who ingest large fluid volumes during prolonged exercise.
We gratefully acknowledge the support of Sports Science New Zealand and Middlemore Hospital Laboratory, and the participation of the athletes and volunteers in this study. This research has no relationship with any companies or manufacturers who would benefit from this work. The results of the present study do not constitute endorsement of any product by the authors or ACSM.
Address for correspondence: Dr. Dale Speedy, 179A Hill Road, Manurewa, New Zealand; E-mail: [email protected]
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 heat. Med. Sci. Sports Exerc. 25: 543–549, 1993.
2. Clark, J. M., and F. J. Gennari. Encephalopathy due to severe hyponatremia in an ultramarathon runner. West. J. Med. 159: 188–189, 1993.
3. Cluitmans, F. H. M., and A. E. Meinders. Management of severe hyponatremia: rapid or slow correction? Am. J. Med. 88: 161–166, 1990.
4. Davis, D., A. Marino, G. Vilke, J. Dunford, and J. Videen. Hyponatremia in marathon runners: experience with the inaugural Rock ‘n Roll Marathon. Ann. Emerg. Med. 34: S40, 1999.
5. Dill, D. B., and D. L. Costill. Calculation of percentage changes in volumes of blood, plasma, and red cells in dehydration. J. Appl. Physiol. 37: 247–248, 1974.
6. Frizzell, R. T., G. H. Lang, D. C. Lowance, and R. Lathan. Hyponatremia and ultramarathon running. JAMA 255: 772–774, 1986.
7. Hiller, W. D. B. Dehydration and hyponatremia during triathlons. Med. Sci. Sports Exerc. 21: (Suppl.) 219–221, 1989.
8. Hiller, W. D. B. The United States Triathlon Series. Medical Coverage of Endurance Athletic Events. Columbus, OH: Ross Laboratories, 1988, pp. 80–81.
9. Hiller, W. D. B., M. L. O’Toole, F. Massimino, R. E. Hiller, and R. H. Laird. Plasma electrolyte and glucose changes during the Hawaiian Ironman Triathlon. Med. Sci. Sports Exerc. 17: (Suppl.) 219, 1985.
10. Holtzhausen, L. M., T. D. Noakes, B. Kroning, M. De Klerk, M. Roberts, and R. Emsley. Clinical and biochemical characteristics of collapsed ultramarathon runners. Med. Sci. Sports Exerc. 26: 1095–1101, 1994.
11. Irving, R. A., T. D. Noakes, R. Buck, et al. Evaluation of renal function and fluid homeostasis during recovery from exercise-induced hyponatremia. J. Appl. Physiol. 70: 342–348, 1991.
12. Lote, C. J. Principles of Renal Physiology, 3rd Ed. London: Chapman and Hall, 1994, p. 101.
13. Montain, S. J., W. A. Latzka, and M. N. Sawka. Fluid replacement recommendations for training in hot weather. Mil. Med. 164: 502–508, 1999.
14. Mulloy, A. L., and R. J. Caruana. Hyponatremic emergencies. Med. Clin. North Am. 79: 155–168, 1995.
15. Noakes, T. D. Dehydration during exercise: what are the real dangers? Clin. J. Sport Med. 5: 123–128, 1995.
16. Noakes, T. D. Hyponatremia during endurance running: a physiological and clinical interpretation. Med. Sci. Sports Exerc. 24: 403–405, 1992.
17. Noakes, T. D. The hyponatremia of exercise. Int. J. Sport Nutr. 2: 205–228, 1992.
18. Noakes, T. D. Hyponatremia of exercise. New on Sports Nutrition Insider 3: 1–4, 1995.
19. Noakes, T. D., N. Goodwin, B. L. Rayner, T. Branken, and R. K. N. Taylor. Water intoxication: a possible complication during endurance exercise. Med. Sci. Sports Exerc. 7: 370–375, 1985.
20. Noakes, T. D., R. J. Norman, R. H. Buck, J. Godlonton, K. Stevenson, and D. Pittaway. The incidence of hyponatremia during prolonged ultraendurance exercise. Med. Sci. Sports Exerc. 22: 165–170, 1990.
21. 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.
22. Nose, H., G. W. Mack, X. Shi, and E. R. Nadel. Shift in the body fluid compartments after dehydration in humans. J. Appl. Physiol. 65: 318–324, 1988b.
23. O’Toole, M. L., P. S. Douglas, R. H. Laird, and D. B. Hiller. Fluid and electrolyte status in athletes receiving medical care at an ultradistance triathlon. Clin. J. Sport Med. 5: 116–122, 1995.
24. Retzlaff, J. A., W. N. Tause, and J. M. Kielly. Erythrocyte volume, plasma volume and lean body mass in adult men and women. Blood 33: 649–661, 1969.
25. Speedy, D. B., R. G. D. Campbell, G. Mulligan, et al. Weight changes and serum sodium
concentrations after an ultradistance multisport triathlon. Clin. J. Sport Med. 7: 100–103, 1997.
26. Speedy, D. B., J. G. Faris, M. Hamlin, P. G. Gallagher, and R. G. D. Campbell. Hyponatremia and weight changes in an ultradistance triathlon. Clin. J. Sport Med. 7: 180–184, 1997.
27. Speedy, D. B., and T. D. Noakes. Exercise-associated hyponatremia: a review. Sportmedizin 50: 368–374, 1999.
28. Speedy, D. B., T. D. Noakes, I. R. Rogers, et al. Hyponatremia in ultradistance triathletes. Med. Sci. Sports Exerc. 31: 809–815, 1999.
29. Speedy, D. B., T. D. Noakes, I. R. Rogers, et al. A prospective study of exercise-associated hyponatremia in two ultradistance triathletes. Clin. J. Sport Med. 10: 136–141, 2000.
30. Speedy, D. B., I. R. Rogers, T. D. Noakes, et al. Diagnosis and prevention of hyponatremia at an ultradistance triathlon. Clin. J. Sport Med. 10: 52–58, 2000.
31. Speedy, D. B., I. R. Rogers, T. D. Noakes, et al. Exercise-induced hyponatremia is caused by inappropriate fluid retention. Clin. J. Sport Med. 10: 272–278, 2000.
32. Wolfson, A. B. Acute hyponatremia in ultradistance-endurance athletes. Am. J. Emerg. Med. 13: 116–117, 1995.
33. Young, M., F. Sciurba, and J. Rinaldo. Delirium and pulmonary edema after completing a marathon. Am. Rev. Respir. Disease 136: 737–739, 1987.