Marathon Runners' Deaths Attributed to Antidiuretic Hormone

BELMONT, Mass., April 30 -- Exercise-associated hyponatremia, a potentially fatal condition of endurance athletes, may be brought on by suppression of renal water excretion exacerbated by excess fluid intake, investigators here reported.

The deaths in 2002 of two runners who had completed the Boston Marathon, and the post-race collapse of several runners in other years, prompted Arthur J. Siegel, M.D., at McLean Hospital, and colleagues, to look into the contributory mechanisms of antidiuresis.

They found that disturbance of arginine vasopressin (antidiuretic hormone) secretion, brought on by release of interleukin-6 associated with excess fluid and electrolyte intake, triggered a cascade leading to hypotonic encephalopathy and the two cases of fatal cerebral edema, the investigators reported in the May issue of the American Journal of Medicine.

"This is a major paradigm shift for those who think that exercise-associated hyponatremia is due primarily to salt loss or over-consumption of fluids," said Dr. Siegel. "It's also an inside job. Avid drinking may be a precondition, but dysregulation of the anti-diuretic hormone or arginine vasopressin (AVP), which governs water balance, emerges as the root cause."

Dr. Siegel, a runner and member of the Boston Marathon medical team, cautioned that marathoners should take extra precautions to avoid risking hyponatremia.

"The message especially for slower runners, such as charity fundraisers, is awareness that over-hydration is more dangerous than dehydration," he said. "Such participants may need to decrease their drinking rate commensurate with their race pace."

He noted that runners who gain rather than lose weight during training or a race are over-hydrating and need to cut back on both water and sports-drink consumption.

The investigators studied pre- and post-race blood samples from 39 runners who took part in the 2001 Boston Marathon, from 308 runners who collapsed and were treated in the medical tent after the 2004 iteration of the race, and stored blood samples from two runners who had died from cerebral edema.

They tested the samples for creatine kinase, interleukin-6, arginine vasopressin, cortisol, prolactin, and C-reactive protein on the day before the race, and within two hours of the finish. Eleven of the volunteers also provided samples the day after the race.

They found that the runners with normal sodium levels had a mean 3% decrease in body weight, and showed a 40-fold increase in serum levels of the inflammatory cytokine interleukin-6. In these runners, IL-6 levels rose from a mean of 1.6 + 0.5 picograms/mL to 66.6 + 11.9 pg/mL, P=0.001), and this increase correlated significantly with increases in creatine kinase, cortisol, and prolactin, but not arginine vasopression.

In contrast, among 22 collapsed runners with exercise-associated hyponatremia, blood urea nitrogen levels were within the normal reference range, with a mean level of less than 15 mg/dL, but 43% of these runners had measurable levels of arginine vasopressin of more than 0.5 pg/mL, indicating that suppression of water excretion rather than excess fluid load might account for the discrepancy.

"The absence of urination in cases of exercise-associated hyponatremia during treatment in the medical tent indicates that vasopressin levels were sufficient to impair maximal urine free-water excretion," the authors wrote. "This clinical feature also was present in cases of exercise-associated hyponatremia without measurable hormone, suggesting ex-vivo proteolysis as the likely explanation for the negative findings in some specimens."

When they looked at the samples from the two marathoners who died from cerebral edema, they found evidence of less than maximally dilute urines (>100 mmol/kg/H₂O) and urinary sodium concentrations greater than 25 mEq/L, which, in combination with elevated cortisol levels and normal thyroid function, suggested the presence of the syndrome of inappropriate antidiuretic hormone secretion, first described in 1967.

The authors noted that in more recent Boston marathons runners who developed hypotonic encephalopathy were treated with intravenous hypertonic (3%) saline, and experienced rapid clinical improvement without adverse effects.

"Similar to the effects on cortisol and prolactin, stimulation of arginine vasopressin by interleukin-6 may interfere with appropriate physiologic suppression during hypo-osmolality," Dr. Siegel and colleagues wrote.

"Rhabdomyolysis may be linked to the pathogenesis of exercise-associated hyponatremia as has been observed in other clinical conditions in which the syndrome of inappropriate antidiuretic hormone secretion may accompany inflammatory stress. This paradigm provides a unified explanation for the pathogenesis and clinical features of exercise-associated hyponatremia, rather than postulating multiple independent mechanisms."

They acknowledged that the study was limited by the small sample size and the observational design.