Clinical presentation of exercise‐associated hyponatremia in male and female IRONMAN® triathletes over three decades

Exercise‐associated hyponatremia (EAH) is common in ultra‐endurance events and severe cases are more common in females. The purpose of this paper is to compare the clinical presentation of EAH between male and female triathletes in ultra‐endurance competitions.


| INTRODUCTION
Exercise-associated hyponatremia (EAH) is a condition characterized by an acute fall in blood sodium concentration that develops during or immediately following physical activity. 1 Diagnostic criteria of EAH is a serum sodium <135 mEq/L regardless of symptom presence, 1 and is commonly observed in ultra-endurance events including marathons, triathlons, and other distance athletic events. [2][3][4][5] The incidence of EAH presenting in athletes for IRONMAN® competitions has been previously reported as high as 20%, though these studies are clearly limited by several factors, including small sample size and limited variance. [2][3][4][5][6] Cases of EAH are often asymptomatic or subclinical, with symptoms being very mild and/or nonspecific, characterized by headache, nausea, vomiting, dizziness, weakness, malaise, fatigue, and muscle cramps. [7][8][9][10][11] On rarer occasions, EAH can be much more severe and potentially life-threatening with symptoms of encephalopathy, seizures, and respiratory distress. [7][8][9][10][11] Given the frequency of EAH within ultra-endurance-related competition, identifying initial signs and symptoms early is critical to preventing severe complications later. 12 Noakes et al. was the first group to report on EAH and hypothesized that water intoxication from over drinking was the primary cause. 13 As, EAH has been noted to have multiple intersecting etiologies with a complicated pathophysiology which can be classified into hypovolemic, euvolemic, and hypervolemic EAH. 3,6,14,15 This classification is important as each has a different underlying pathophysiology and treatment; hypovolemic EAH requires fluid and sodium replacement to prevent circulatory collapse, while hypervolemic and euvolemic EAH requires fluid restriction and sodium replacement. 5 Athletes that develop hypervolemic and euvolemic EAH have a dilutional hyponatremia from an increased total body water relative to exchangeable body sodium. 3,6,14 This occurs by dilution of total extracellular sodium concentration by the ingestion of hypotonic fluids, that is in excess of fluid losses. 3,14,15 Speedy et al. 16 examined ultraendurance triathletes at the 1997 New Zealand Triathlon and found that nearly three quarters of severe EAH cases were due to fluid overload. However, individuals with normal glomerular filtration rates can excrete 500-1000 mL/ hr of dilute urine before developing water retention. 3 Such high volumes are difficult to drink for most athletes. Thus, other factors must be involved in the formation of hypervolemic EAH. 3 Persistent non-osmotic secretion of vasopressin resulting in impaired urinary water excretion is considered to be a contributing factor. 3,6 The exact cause of this is uncertain but previous studies suggest fluid losses (sweat, urine, and insensible) result in a significant serum volume contraction that stimulates consistent vasopressin secretion to maintain circulation at the expense of serum sodium. 3,14 The importance of hypovolemia in EAH has been debated as early data suggested the majority of athletes were affected by hypervolemic or euvolemic EAH, however, newer data suggest a higher prevalence of hypovolemic EAH, with weight loss in EAH supporting volume depletion in the pathogenesis of EAH. 3,14,15 Other variables specific to athlete and/or event have been shown previously to be associated with EAH risk, including low BMI, long racing times, availability of fluids on the course, extreme hot or cold external temperatures, low competition experience, intake of NSAIDS, and female sex. 1,5,6,11,14,[16][17][18][19][20] Menstruant females have been identified as a particularly high risk population for EAH, 1 with several explanations hypothesized. It is possible that estrogen and progesterone inhibit the function of the Na + -K + -ATPase, which normally has an important function in the extrusion of sodium from cells during the development of hyponatremia. 5,21 This inhibition leads to higher risk for developing cerebral edema. 5 Another hypothesized explanation is females have lower average body mass index and longer racing times, both of which have been identified as risk factors for EAH in the general population. 5 This hypothesis is supported by a study of the 2005 Boston Marathon that found female sex was not a risk factor when BMI and finish times were controlled for. 19 A third hypothesized explanation is represented in the higher fluid intake recommended for female athletes within current hydration guidelines. 5 Recommendations on this have changed substantially since EAH was first reported in the early 1980s, with decreasing emphasis on fluid consumption. 5,6 Several associated risk factors have been identified for EAH; however, it is unclear if, and exactly how it presents differently between the sexes. The current EAH literature consists primarily of studies with limited sample sizes and over short time periods, particularly among ultraendurance athletes. Given this gap in understanding, the objective of this study was to compare the clinical presentation of EAH in male and female IRONMAN® triathletes. We hypothesize that clinical variables including altered mental status, abdominal pain, and hypotension will present more commonly in hyponatremic women than men.

| Study population
This study is retrospective review of 28 years of medical data from the IRONMAN® World Championship. The IRONMAN® World Championship, one of the most prestigious ultra-endurance competitions in the world, takes place annually in Kona, Hawaii and consists of a 3.9-km swim, 180-km bike ride, and a 42-km run. Since the inaugural race (1978), IRONMAN® has grown from a small local event with 15 participants into a major international event with over 2500 participants from all over the world. 22 Starting in 1989, medical records for athletes who sought medical attention during and immediately after competition have been recorded (n = 9482). Investigators at Washington State University (WSU) are curators of these records and have electronically entered and managed these data through the Global Triathlon Safety Database, thereby allowing for future statistical analysis. HIPAA guidelines were utilized for all data handling, storage, and analytic tasks for this project, and all procedures have been approved by the Institutional Review Board at WSU. For the purposes of the current project, study investigators examined the records of athletes from 1989 to 2019 ( Figure 1).

| Data collection
Medical records were recorded by physicians, nurses, other medical volunteers within the medical tent at the IRONMAN® World Championship competitions. On occasion, data for certain athletes were collected on the course. General information recorded for each athlete brought to the medical tent included, times of encounters, vitals, complaints, physical findings, lab values, diagnoses, dispatch information, and personal identifiers. Athletes were weighed prior to the race and upon entering the medical tent. Venous blood was drawn in the medical tent and immediately processed with onsite lab analysis. Variable number of tests were available per year, therefore, varying rates of bloodwork was performed. Of note, acutely symptomatic athletes were always tested. Admission to the medical tent was from athletes who presented themselves to the tent and athletes who were brought to the tent from the course, finish line, or massage tent by medical personnel or volunteers.

| Clinical presentations
All complaints (vomiting, syncope, altered mental status, diarrhea, nausea, abdominal cramps, headache, exhaustion, muscle cramps, and dizziness) and diagnoses (heat cramps, exhaustion, and dehydration) were recorded as binary variables, either present or absent. Complaints were documented by medical personnel based on subjective findings noted by athletes. Diagnoses were made by physicians based on expert clinical judgment. Other variables included were age, sex (as defined as race category, male or female), weight change, vitals (pulse, respiratory rate, blood pressure, and temperature), and lab values (Na + , K + , glucose). Percent weight loss is defined as change in weight divided by pre-race weight. Hypotensive is defined as a systolic blood pressure of <90 mmHg or a diastolic of <60 mmHg and hypertensive is defined as a systolic blood pressure of ≥130 mmHg or a diastolic ≥80 mmHg. Tachycardic is defined as a pulse >100 bpm. Tachypneic is defined as a respiratory rate >20 bpm. Hyponatremic is defined as [Na + ] < 135 mEq/L, hypernatremic is defined as [Na + ] > 145 mEq/L, hyperkalemic is defined as a [K + ] > 5.0 mEq/L, and hypokalemic is defined as a [K + ] < 3.5 mEq/L. All vitals and lab values are reported as F I G U R E 1 Distribution of athletes by sex and sodium concentration.
presenting values, that is, the first value that was recorded during their visit to the medical tent.

| Statistical analysis
All data was analyzed using IBM SPSS statistics, version 27.0. Descriptive statistics and independent samples ttests were used to determine and compare incidence values and frequencies for hyponatremic and eunatremic male and female triathletes. Logistic regression was used to determine sex and sodium concentration's ability to predict presence of each clinical outcome reported as unstandardized regression weights (B). Logistic regression was also used to determine how sodium concentration predicts the presence of clinical outcomes for each sex independently. Curve estimation regression was used to plot probability curves for each clinical variable using sodium concentration. For all regression analyses, hypernatremic (>145 mEq/L) athletes were excluded. All statistical significance is reported at a level of p ≤ 0.05.

| Relationship between sex and each clinical variable
Clinical variables found to be present more frequently in women include vomiting, diarrhea, nausea, abdominal pain, hypotension, hypokalemia, and weight gain. Clinical variables found more frequently in men include muscle cramps, heat cramps, dehydration, hypertension, hyperkalemia, and a 5% weight loss (Table 1).

| Relationship between sex, hyponatremia and each clinical variable
For all triathletes, clinical variables found to be inversely related with sodium concentration include altered mental status, hypertension, hypokalemia, and weight gain. In contrast, clinical variables found to be directly related to sodium concentration include muscle cramps, exhaustion, dehydration, hypotension, tachycardia, and a 5% weight loss (Table 2). Table 2 is also graphically represented in Figure S1.
When comparing male and female triathletes, clinical variables found to have differing relationships with sodium concentration include altered mental status F I G U R E 2 Relative frequency of male and female triathletes who received bloodwork per sodium value across all years.
(inversely related in males and not related in females), abdominal pain, muscle cramps, hypotension, and tachycardia (directly related in males and not related in females) and vomiting and hypokalemia (not related in males and inversely related in females). Overall, females lost less weight (−0.8 ± 4.0%) than males during competition (−3.5 ± 3.8%), t(777) = −9.642, p = 0.000.

| DISCUSSION
The overall objective of this study was to compare the clinical presentation of EAH in male and female IRONMAN® triathletes overtime. The medical records analyzed as a part of this study represents the largest medical data set collected on ultra-endurance triathletes and spans over 30 years. For the purposes of the current study, this volume of data has allowed for a meticulous analysis of factors related to EAH. This data will also permit future unique examinations of less common clinical presentations observed during ultra-endurance competition, the majority of which have not previously been reported in the medical literature. The main findings of this study are that clinical variables including altered mental status, vomiting, abdominal pain, muscle cramps, hypotension, tachycardia, and hypokalemia appear to present differently in hyponatremic male and female athletes, confirming our hypothesis.
Our findings confirm that females are more likely than males to be hyponatremic during ultra-endurance exercise. We found that 29.7% of females that received bloodwork testing were hyponatremic, compared to only 21.2% of males. Direct qualitative comparisons on incidence of EAH in athletes to previous investigations are difficult as most studies examined a subset of individuals generalizable to all athletes. 9,16,17,23 Our study analyzes only those athletes that reported to the medical tent and received bloodwork. Given this, we can assume that incidence values reported in our work are slight overestimations for all athletes, given athletes that report to the medical tent and receive bloodwork are theoretically more likely to be hyponatremic. Speedy et al. examined ultra-endurance triathletes (n = 330) participating in the New Zealand IRONMAN® Triathlon and reported that 45% of females receiving bloodwork were hyponatremic, compared to 14% of males. 16 The incidence values found in our study appear to be relatively similar for male athletes, 16 especially when considering the differences in sample population previously noted; however, the authors present a much higher incidence value in females with an analytic sample that is notably limited (n = 38) compared to ours (n = 885). Studies on a variety of other ultra-endurance athletic events, including cycling, swimming, running, and triathlon, found more similar incidence values in males and females compared to our findings. [7][8][9][10][11] In the current study, we also reveal that females receiving bloodwork had significantly lower serum sodium values (137.4 mEq/L) compared to males (139.7 mEq/L). These findings are similar to what Speedy et al. found (134 mEq/L in females and 137 mEq/L in males), and also closely align with previous reports for other ultra-endurance athletic events. 9,16,17,22,23 Relationships among clinical variables reported within our work between sex and EAH during ultra-endurance competition ( Table 2) have not been previously reported in the medical literature. Several of these variables have previously been shown to be associated with either sex or hyponatremia, but our work is novel in that it explores how all three are related. Clinical variables that appeared to present differently in male and female athletes include  altered mental status, vomiting, abdominal pain, muscle cramps, hypotension, tachycardia, and hypokalemia. The etiology of these findings is likely multifactorial for each clinical variable. Hypotheses as to the cause of these differences should be explored in future scientific investigations. Several physiologic characteristics have been previously identified for female athletes that potentially explain several of our findings (Tables 1 and 2), including increased water intake during competition, increased water retention, decreased BMI. 5,21 More than likely, these lead to overhydration susceptibility. Our findings support this in that we found weight change to be strongly associated with both sex and hyponatremia. Blood pressure and dehydration were also found to be strongly inversely related to female sex and hyponatremia. Comprehensively, these findings (weight change, blood pressure, and hydration status) paint a clear clinical picture of fluid overload as the most common etiology of EAH during endurance athletic events. 1 Noakes et al. examined the relationship of sodium and weight gain among 2135 athletes in multiple different endurance events over 8 years and found the mean ± SD serum sodium was 136.1 ± 6.4 mmol/L for athletes who gained weight during competition, 140.5 ± 3/7 mmol/L for those with minimal weight gain, and 141.1 ± 3.7 mmol/L for those who lost weight during competition. 23 These clearly align with our findings.
Although this strong association between overhydration and hyponatremia has been demonstrated through previous investigations, our findings indicate that roughly half of all hyponatremic athletes were dehydrated and lost weight. This clearly demonstrates that although weight gain and overhydration provide a clear etiology for hypervolemic and euvolemic hyponatremia, hypovolemic hyponatremia remains a significant cause of hyponatremia in endurance athletes. Given the possible severity of this condition, and the difference in clinical management, more research is needed to understand the clinical and pathophysiological differences between hypervolemic, euvolemic, and hypovolemic hyponatremia. Clinically, medical personnel should consider hyponatremia when signs of weight loss and dehydration are present.

| Limitations
This study has several limitations, largely due to the nature of utilizing medical records for research and analytic purposes. As recorded by medical professionals, minor variation existed within the criteria for many of these symptoms over the years. All vitals and lab values were reported as a presenting value, with inconsistencies likely over the years in the time between race exit time and lab value/vital measurement. Although our overall sample size is large, given the rarity of several clinical variables and smaller number of female participants, sample distribution serves as a limitation. Finally, our findings are limited in that this data originates only from triathletes who both visited the medical tent and received bloodwork. Therefore, this study's findings may not generalize to all ultra-endurance triathletes as well as no prerace sodium values were gathered. Given EAH is most commonly asymptomatic, athletes visiting the medical tent will have often visited for other reasons. Further research is needed to better understand differences in clinical presentations between sexes as well as the relationship between overhydration, hyponatremia, and sex.

| CONCLUSION
From this work, we conclude that females are significantly more likely to have EAH and lower serum sodium concentrations when visiting the medical tent at IRONMAN® triathlons. We also found that altered mental status, vomiting, abdominal pain, muscle cramps, hypotension, tachycardia, and hyperkalemia appear to present differently between sexes when comparing hyponatremic to eunatremic athletes. Although not the primary goal of this study, we confirm a strong correlation between overhydration and hyponatremia as documented in the literature; however, our findings also show that a significant number of athletes had hypovolemic hyponatremia. Further understanding of how EAH presents will allow for more efficient diagnosis and clarity on life-saving treatment.

| Perspective
This manuscript provides updated understanding on differences in clinical presentations between male and female athletes with hyponatremia. The etiology of many of these relationships is not well understood and requires further investigation. The relationship between sodium and hydration status is going to be further explored by the authors in future work.

AUTHOR CONTRIBUTIONS
Kasey B. Johnson developed study idea, conducted data analysis, and drafted manuscript; Christopher P. Connolly assisted developing study idea, wrote IRBapproved application, managed production of database, and assisted in drafting manuscript; Stephanie P. Cho assisted in managing production of database, and assisted in developing study idea; Thomas K. Miller assisted in drafting manuscript and collected data; Robert E. Sallis assisted in drafting manuscript and collected data; and W. Douglas B. Hiller assisted developing study idea developed database idea, assisted in drafting manuscript, and collected data.