Share this article on:

Hyponatremia in runners requiring on-site medical treatment at a single marathon


Medicine & Science in Sports & Exercise: February 2002 - Volume 34 - Issue 2 - p 185-189
CLINICAL SCIENCES: Clinical Investigations

HSIEH, M., R. ROTH, D. L. DAVIS, H. LARRABEE, and C. W. CALLAWAY. Hyponatremia in runners requiring on-site medical treatment at a single marathon. Med. Sci. Sports Exerc., Vol. 34, No. 2, pp. 185–189, 2002.

Study objective Literature reports indicate an increasing number of cases of hyponatremia in athletes participating in moderate endurance events such as standard marathons. In this study, we evaluated the incidence of hyponatremia in marathon finishers requiring medical treatment on-site and attempted to assess the contribution of fluid type ingested and nonsteroidal antiinflammatory drug (NSAID) use to the development of hyponatremia.

Methods We examined a prospective, convenience sample of runners requiring intravenous hydration at the final medical tent of a standard marathon course and a comparison group of finishers who did not require intravenous hydration. After giving informed consent, subjects had blood drawn and answered a questionnaire regarding fluid intake on the course and NSAID use before the race. Blood samples were analyzed on-site for serum sodium values as well as other hematologic parameters.

Results Fifty-one subjects requiring intravenous hydration as well as 11 subjects who did not were enrolled. Three subjects (5.6%; 95% CI, 0–11.9%; missing = 8) in the intravenous hydration group had serum sodium less than 130 mEq/L. None of the three runners suffered neurologic or pulmonary consequences and only one required overnight hospital admission for hydration. The small number of hyponatremic subjects precluded the analysis of the role of fluid type or NSAID use in the development of hyponatremia or the development of a model for prediction.

Conclusion This study found a 5.6% incidence of hyponatremia in marathon runners requiring medical treatment.

Department of Emergency Medicine, University of Pittsburgh, Pittsburgh, PA

Submitted for publication November 2000.

Accepted for publication May 2001.

Increasing participation in endurance sports such as marathons in recent years has stimulated research into physiologic changes and possible health complications during these events. Medical attention has focused on the incidence and mechanism of hyponatremia with its attendant pulmonary and neurologic implications in those athletes. Although studies generally show either no change (13,15,20) or an increase in serum sodium (12,21,26,27) after such events, cases of hyponatremia in ultraendurance or triathlon athletes have been reported (1,4,16). Observed incidence of hyponatremia in ultraendurance events that last over 6 h ranges from 0.3% to 27%(5,7,8,17,23). Among athletes requiring medical attention during or after ultraendurance events, the incidence is even higher, ranging from 9% to over 60%(7,19). The frequency of occurrence of hyponatremia has prompted medical directors to set up field laboratories at some ultraendurance events (7,18).

Researchers have investigated possible fluid replacement regimens to prevent such complications (9,14,24). No consensus on an optimal regimen exists because the mechanism of the development of hyponatremia during exercise is unclear. Even the volume status of the hyponatremic athletes is under debate, with arguments for both dehydration (6,22,27) and overhydration (5,16) as contributing factors. Proposed mechanisms include excessive sodium loss from sweating (6), ingestion of only hypotonic fluids (10,25), impairment of renal function from exercise (3,13) or from the use of medications such as nonsteroidal antiinflammatory drugs (NSAID) (2,11), and inappropriate free-water secretion (23,28,30).

Hyponatremia has been assumed to be uncommon in events lasting less than 4 h such as standard marathons (6). However, case reports of hyponatremia after moderate exercise are accumulating (1,2,22,29,30). Increased participation by less well-trained athletes with resultant prolonged finish time and increased environmental exposure may contribute to the growing incidence (29). Athletes seeking medical attention during these events have traditionally been treated symptomatically in the field without the aid of laboratory testing, and transported to a hospital only if they have insufficient improvement with field rehydration. Because of these case reports, some medical directors have proposed routine testing of serum electrolytes in all athletes seeking treatment. In order to improve the assessment of the need for testing in this population, we evaluated the incidence of hyponatremia in finishers of a standard 42-km (26-mile) marathon who sought on-site medical treatment. In addition, we investigated the association of fluid intake and NSAID use with the occurrence of hyponatremia.

Back to Top | Article Outline


The institutional review board (IRB) of our institution approved this study. The design was a prospective observational study using a convenience sample of finishers of a standard marathon held in May 2000. The race started at 8:30 a.m. on a U.S. Track and Field–approved course. Aid stations with options of water and sports drink mixed from powder (premixed drink has a sodium content of 220 mEq/L) were available each mile along the course. A medical tent at the finish line treated runners who finished the race for dehydration or other injuries. In general, runners unable to tolerate oral rehydration were given intravenous normal saline until vital signs normalized and oral intake resumed. At physician discretion, runners with inadequate improvement were transferred to local emergency departments via ambulances. Reasons for transfer included persistent orthostasis and/or persistent vomiting. Subjects requiring intravenous hydration at this medical tent were approached for informed consent. Marathon finishers who presented to the medical tent for oral hydration or non–hydration-related issues were enrolled as control subjects. All enrolled subjects signed informed consent.

Each subject had 5 mL of blood drawn before normal saline infusion for the intravenous group or a large amount of oral fluid intake for the control group. Blood samples were stored in pediatric sodium heparinized Vacutainer tubes labeled with the subject’s bib number and placed in ice until they could be analyzed by an on-site iSTAT portable clinical analyzer owned by our department. Measured hematologic parameters included sodium, potassium, chloride, hematocrit (Hct), glucose, and blood urea nitrogen (BUN). The sodium in the heparinized tube affects measured sodium values by less than 1 mEq/L, which we considered clinically insignificant. The results of blood tests were not available to medical personnel for clinical decision making because on-site testing was considered investigative by our IRB. Subjects reported the frequency and type of fluid intake on the race course as well as any NSAID use before the race. Additional information such as age, sex, and finishing times were obtained by referencing the bib number for the subject.

Continuous variables were reported as mean ± standard deviation and 95% confidence interval and discrete variables were reported as frequencies. Differences in variables between the two treatment groups were tested using chi-square or Fisher’s exact test for discrete variables and t-test for continuous variables. The primary outcome of interest was the proportion of subjects with serum sodium values less than 130 mEq/L. We chose 130 mEq/L as the cutoff because clinical management and/or disposition rarely change for serum sodium between 130 mEq/L and 140 mEq/L. Other outcomes of interest included the differences in the proportion of hyponatremic subjects and the means of the remaining hematologic values between the treatment groups as well as the differences in the frequency of NSAID use, age, sex, and finishing time between hyponatremic and normonatremic subjects.

Back to Top | Article Outline


On the basis of information from previous marathons, we had anticipated the enrollment of 50 subjects requiring intravenous hydration and 50 control subjects receiving oral or no hydration. Weather at this marathon, however, was sunny, hot, and humid (Table 1), qualifying as moderate risk for heat-related illnesses according to the American College of Sports Medicine (ACSM) guideline for distance running. The severity of the heat stress contributed to an increased number of runners requiring intravenous hydration. Despite prerace announcement of the race conditions and warnings not to achieve personal records during this race, 338 runners of the 3200 starters and 2700 finishers required evaluation throughout the marathon course and 241 patients at the final medical tent. Thirty-five patients required transport to the hospital, including nine from the final medical tent. We enrolled 51 subjects requiring intravenous hydration and 11 control subjects (N = 62). These two groups did not differ in age (P [t55 = −0.82] = 0.41, missing = 5), sex (P [chi-square1 = 1.32] = 0.25), finishing time (P [t57 = −1.83] = 0.07, missing = 3), or NSAID use within 24 h of the marathon (P [chi-square1 = 0.42] = 0.52, missing = 4) (Table 2). The proportion of men did not vary in terms of starters (76%), finishers (74%), runners requiring intravenous lines (76%), and subjects enrolled (78%). The average age of subjects requiring intravenous hydration (36.2 ± 9.8 yr) also did not differ significantly from that of all finishers (37.4 ± 10.1 yr). All subjects except for one in the control group drank at every fluid station and 47 (79.7%) of the subjects drank a mixture of water and sports drink throughout the course. Eight runners drank only water.

Table 1

Table 1

Table 2

Table 2

Mean serum sodium, chloride, BUN, glucose, and Hct values for the treatment groups are shown in Table 2. Many samples hemolyzed before measurement with resultant unusable potassium values; therefore, serum potassium was excluded from analysis. The groups did not differ significantly with respect to the remaining parameters (sodium, P [t52 = 1.68] = 0.10, missing = 8; chloride, P [t52 = 1.37] = 0.18, missing = 8; BUN, P [t52 = −1.16] = 0.25, missing = 8; glucose, P [t55 = 1.23] = 0.22, missing = 5; Hct, P [t52 = 0.01] = 1.00, missing = 8). The frequency distribution of serum sodium values is shown in Figure 1.



The incidence of hyponatremia, defined as serum sodium value less than 130 mEq/L, is 5.6% in this study (N = 3; missing = 8; 95% CI, 0–11.9). All three hyponatremic subjects received intravenous hydration and drank at least some sports drink at every aid station. Detailed oral intake can only be estimated. If each runner finished one 100-mL cup of drink at each station, about 2.5 L of fluid would be consumed over the entire course. The amount of intravenous fluid administered was recorded for one of the three hyponatremic runners. Table 3 lists the descriptive data as well as clinical course for those three individuals. No unique sign or symptom identified the hyponatremia runners. In addition to the three hyponatremic runners, 16 normonatremic subjects presented with vomiting, 19 with dizziness, four with headache, and eight with initial systolic blood pressure below 90 mm Hg. None of the three patients suffered any neurologic complication. Interestingly, one subject who did present to the medical tent with new-onset seizure had a normal serum sodium value of 142 mEq/L. His initial temperature at the medical tent was 106°F. The small number of subjects with hyponatremia prohibited the performance of logistic regression analysis to determine the possible contribution of the type of fluid intake or NSAID use on the development of hyponatremia. The subject with sodium of 122 mEq/L did not use any NSAID before the race. This information was not available for the other two subjects.

Table 3

Table 3

Back to Top | Article Outline


As the most comprehensive evaluation to date of the incidence of hyponatremia in marathon runners requiring on-site medical treatment, this study demonstrated a clinically significant proportion (5.6%; 95% CI, 0–11.9) of hyponatremic runners. However, this marathon was conducted during a particularly hot and humid day. It may not have been representative of all marathons. Interestingly, Nelson et al. (12) examined electrolyte changes in 47 runners in the 1987 marathon in the same city and did not report any occurrence of hyponatremia. The more severe heat stress in the 2000 race may have predisposed runners to health complications such as hyponatremia. The actual incidence of hyponatremia for all runners requiring medical treatment in this marathon may have been underestimated because some runners were transported from aid stations on the course before completion of the race. This group who do not reach the finish line may be at higher risk because they could represent the less prepared runners. Others sought medical attention on their own at local emergency departments after the race. There are no data estimating the incidence of hyponatremia in that population. Although the incidence of hyponatremia in the runners who required medical treatment is clinically significant, this entity is still a rare event in all the runners of this race (∼0.1% of all finishers). We chose to focus on the subset of runners who required medical treatment because that was the population of concern to the medical providers. The incidence of hyponatremia is understandably higher in this subgroup of runners than in the general population of marathon runners.

Although none of the hyponatremic runners suffered neurologic or pulmonary complications, the fact that the runner with the lowest sodium value sought further medical attention after discharge from the medical tent suggests that clinical appearance and vital signs of the runners alone may be inadequate for determining readiness for discharge. Symptoms of exertional hyponatremia include nausea, vomiting, headache, and altered mental status. Unfortunately, these symptoms also overlap those of other exercise- and heat-related illnesses. During a follow-up telephone interview, that runner reported that she continued to have nausea and a headache after her discharge from the medical tent, and that her serum sodium was found to be low during an emergency department evaluation that evening. This protracted course of hyponatremia may reflect inadequate medical treatment initially or continuation of the physiologic forces causing this condition.

The design of this study could not determine the relative contribution of salt loss or excessive intake of water to the hyponatremia observed in some runners. Although sodium loss through sweat (6) and replacement with hypotonic solutions (10,25) may contribute to hyponatremia, the problem is fundamentally caused by an inability to secrete free water appropriately (28). Renal damage from exercise (13), medication (2,11) or myoglobin (3), and inappropriate secretion of vasopressin and antidiuretic hormone (1,20,30) have been postulated to play a role. Since these mechanisms take some time to reverse, runners may be at prolonged risk. Furthermore, delayed hyponatremia may develop in initially normonatremic finishers with the relief of splanchnic vasoconstriction and increased gastric absorption of fluid ingested during the race (4). The present data may support the idea that exertional hyponatremia may result from diverse causes. For example, the woman with sodium of 122 mEq/L had a Hct somewhat lower (37%) than observed in other subjects (Tables 1 and 3). At the same time, one other hyponatremic subject had an unusually high Hct (54%). It is possible that the former subject exhibited hemodilution secondary to excessive water intake, whereas the latter subject was hemoconcentrated after excessive fluid loss. The relative contribution of these factors to hyponatremia could be distinguished by a prospective measurement of prerace and postrace sodium levels as well as fluid intake and output and weights of individual runners during a race.

Although clinically significant, the small number of hyponatremic runners in this study precluded the analysis of fluid type ingested or NSAID use as possible contributing factors. For the same reason, one could not build a model of clinical variables to predict hyponatremia on the basis of these data. Further studies could increase the number of hyponatremic cases by obtaining hospital data of the runners transported from the course or combining data from several marathons. These larger data sets may give improved estimates of actual incidence of hyponatremia. Since we did not measure prerace sodium, we do not know whether the hyponatremic runners may have had low sodium to start with; however, other studies that did measure preevent sodium reported no occurrence of preevent hyponatremia (12,15,21,27). Furthermore, this study makes no attempt to evaluate the incidence of hyponatremia in marathon runners not seeking treatment or the incidence of prolonged or delayed hyponatremia. This assessment would require canvassing of local emergency departments for runners presenting for medical evaluation as well as measurement of pretreatment and posttreatment sodium values of runners receiving treatment at the medical tents.

Results of this study challenge the belief that hyponatremia is uncommon in athletes participating in nonultraendurance events. The prevalent medical practice of treatment on-site and disposition on the basis of clinical status may need to be reevaluated in light of the clinically significant incidence of hyponatremia, the current inability to predict hyponatremia on the basis of unique presenting signs or symptoms, and the protracted and delayed nature of the postulated mechanisms for the development of hyponatremia. The fact that all three hyponatremic runners had a finishing time over 4 h may be a helpful triaging criterion. On-site electrolyte testing may be helpful, especially for race conditions that place runners at increased risk of heat-related illnesses according to ACSM guidelines. On the basis of these guidelines, hyponatremia or other metabolic disturbances might have been anticipated in runners becoming ill during the hot and humid conditions of the present marathon. Because of the findings of this study, the upcoming race will be started a 30 min earlier to help alleviate the heat stress, and runners will receive more frequent updates on the weather and race condition. The medical staff of this marathon will be testing on-site the sodium levels of all runners requiring intravenous hydration before and after infusion. A finding of hyponatremia will be added as a criterion for hospital transfer. The cost of this practice will be approximately $10 per test in addition to the purchase of portable analyzers at approximately $6000 each. More extensive testing of runners requiring treatment and increased follow-up data would be helpful in assessing the cost-effectiveness of those measures.

This study was supported by a grant from the Pittsburgh Emergency Medicine Foundation.

We thank Shawn Hicks, Christopher Lightfoot, David Newman, M.D., and Lori Wylie, M.D., for their assistance during the marathon.

Address for correspondence: Margaret Hsieh, M.D., 230 McKee Place, Suite 400, Pittsburgh, PA 15213; E-mail:

Back to Top | Article Outline


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. Ayus, J. C., J. Varon, and A. I. Arieff. Hyponatremia, cerebral edema, and noncardiogenic pulmonary edema in marathon runners. Ann. Intern. Med. 132: 711–714, 2000.
3. Bailey, R. R., E. Dann, A. H. B. Gillies, K. L. Lynn, M. H. Abernethy, and T. J. Neale. What the urine contains following athletic competition. N. Z. Med. J. 83: 309–313, 1976.
4. Clark, J. M., and F. J. Gennari. Encephalopathy due to severe hyponatremia in an ultramarathon runner. West. J. Med. 159: 188–189, 1993.
5. Frizzell, R. T., G. H. Lang, D. C. Lowance, and R. Lathan. Hyponatremia and ultramarathon running. JAMA 255: 772–774, 1986.
6. Hiller, W. D. B. Dehydration and hyponatremia during triathlons. Med. Sci. Sports Exerc. 21: S219–S221, 1989.
7. Hiller, W. D. B., J. L. O’Toole, E. E. Fortess, R. H. Laird, P. C. Imbert, and T. D. Sisk. Medical and physiological considerations in triathlons. Am. J. Sports Med. 15: 164–167, 1987.
8. 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 (Abstract). Med. Sci. Sports Exerc. 17: 219, 1985.
9. Millard-Stafford, M., L. B. Rosskopf, T. K. Snow, and B. T. Hinson. Water versus carbohydrate-electrolyte ingestion before and during a 15-km run in the heat. Int. J. Sport Nutr. 7: 26–38, 1997.
10. Mudambo, K. S. M. T., G. P. Leese, and M. J. Rennie. Dehydration in soldiers during walking/running exercise in the heat and the effects of fluid ingestion during and after exercise. Eur. J. Appl. Physiol. Occup. Physiol. 76: 517–524, 1997.
11. Muller, J. Why marathon runners collapse. S. Afr. Med. J. 74: 247, 1988.
12. Nelson, P. B., D. Ellis, F. Fu, M. D. Bloom, and J. O’Malley. Fluid and electrolyte balance during a cool weather marathon. Am. J. Sports Med. 17: 770–772, 1989.
13. Neviackas, J. A., and J. H. Bauer. Renal function abnormalities induced by marathon running. South. Med. J. 74: 1457–1460, 1981.
14. Nicholas, C. W., K. Tsintzas, L. Boobis, and C. Williams. Carbohydrate-electrolyte ingestion during intermittent high-intensity running. Med. Sci. Sports Exerc. 31: 1280–1286, 1999.
15. Noakes, T. D., and J. W. Carter. Biochemical parameters in athletes before and after having run 160 kilometres. S. Afr. Med. J. 50: 1562–1566, 1976.
16. 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. 17: 370–375, 1985.
17. 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.
18. Norman, R. J., P. B. Coutts, and J. Godlonton. Organisation of a field laboratory at an ultramarathon. Clin. Chem. 34: 643–644, 1988.
19. Norman, R. J., J. Godlonton, R. H. Buck, and P. B. Coutts. Why marathon runners collapse. S. Afr. Med. J. 74: 246, 1988.
20. Pastene, J., M. Germain, A. M. Allevard, C. Gharib, and J.-R. Lacour. Water balance during and after marathon running. Eur. J. Appl. Physiol. Occup. Physiol. 73: 49–55, 1996.
21. Rama, R., J. Ibanez, M. Riera, M. T. Prats, T. Pages, and L. Palacios. Hematological, electrolyte, and biochemical alterations after a 100-km run. Can. J. Appl. Physiol. 19: 411–420, 1994.
22. Schmidt, W., J. Rojas, D. Boning, H. Bernal, S. Garcia, and O. Garcia. Plasma-electrolytes in natives to hypoxia after marathon races at different altitudes. Med. Sci. Sports Exerc. 31: 1406–1413, 1999.
23. Speedy, D. B., T. D. Noakes, I. R. Rogers, et al. Hyponatremia in ultradistance triathletes. Med. Sci. Sports Exerc. 31: 809–815, 1999.
24. Sproule, J. The influence of either no fluid or carbohydrate-electrolyte fluid ingestion and the environment (thermoneutral versus hot and humid) on running economy after prolonged, high-intensity exercise. Eur. J. Appl. Physiol. Occup. Physiol. 77: 536–542, 1998.
25. Vrijens, D. M. J., and N. J. Rehrer. Sodium-free fluid ingestion decreases plasma sodium during exercise in the heat. J. Appl. Physiol. 70: 154–160, 1999.
26. Wells, C. L., T. A. Schrader, J. R. Stern, and G. S. Krahenbuhl. Physiological responses to a 20-mile run under three fluid replacement treatments. Med. Sci. Sports Exerc. 17: 364–369, 1985.
27. Whiting, P. H., R. J. Maughan, and J. D. B. Miller. Dehydration and serum biochemical changes in marathon runners. Eur. J. Appl. Physiol. 52: 183–187, 1984.
28. Wolfson, A. B. Acute hyponatremia in ultra-endurance athletes. Am. J. Emerg. Med. 13: 116–117, 1995.
29. Young, M., F. Sciurba, and J. Rinaldo. Delirium and pulmonary edema after completing a marathon. Am. Rev. Respir. Dis. 136: 737–739, 1987.
30. Zelingher, J., C. Putterman, Y. Ilan, et al. Case series: hyponatremia associated with moderate exercise. Am. J. Med. Sci. 311: 86–91, 1996.


© 2002 Lippincott Williams & Wilkins, Inc.