Exercise-associated hyponatraemia

Abstract

Overview The aims of this research were to prospectively determine the incidence of exercise-associated hyponatraemia at ultra-distance triathlons and to determine the aetiology of this condition. Specifically the research aimed to determine whether hyponatraemia was due primarily to salt loss associated with dehydration, or to fluid overload, and to determine the site of any fluid retained. Furthermore, hormone levels and risk factors were to be investigated. Finally I planned to investigate the diagnosis and prevention of exercise-associated hyponatraemia. A pilot study was performed at the 1996 New Zealand Ironman triathlon comparing weight changes to post race serum sodium concentrations, and documented an inverse relationship between post-race serum sodium concentrations and percent change in body weight, consistent with fluid overload as the cause of hyponatraemia. This investigation was then repeated at a multi-sport (run-bike-kayak) triathlon, again confirming the same relationship. A larger study was then carried out at the 1997 New Zealand Ironman comparing weight changes during the race and post-race plasma sodium concentrations on 330 race finishers. The same inverse relationship between plasma sodium concentrations and percent body weight change was observed. In this study fluid overload was the cause of most cases of severe symptomatic hyponatraemia although mild hyponatraemia was associated with variable body weight changes. Eighteen percent of race finishers were hyponatraemic, but only 31% of these race finishers sought medical care for symptoms of hyponatraemia. Female athletes were significantly more at risk for hyponatraemia than male athletes. Athletes in the same race (1997 Ironman) who were admitted to hospital overnight for management of exercise-associated hyponatraemia were studied during overnight recovery for fluid and sodium balance. These athletes were compared with control athletes who completed the same race but who maintained their serum sodium concentrations. Hyponatraemic athletes excreted a median fluid excess during recovery where as normonatremic controls had a medium fluid deficit. Estimated sodium deficit was the same in both subjects and controls. Median arginine vasopressin (AVP) concentrations were significantly lower in hyponatraemic subjects than in controls. It was concluded from this study that hyponatraemic athletes have abnormal fluid retention with an increased extracellular fluid volume but without evidence for large sodium losses. There was no evidence of elevated plasma AVP concentrations as the cause of this fluid retention. Eighteen athletes in the same race were also studied prospectively to determine their fluid balance during the race. Athletes lost a median of 2.5 kg of weight during the race, most likely from sources other than fluid loss. Fluid intakes during the race were more modest (median 716 mL/h) than those recommended for shorter duration exercise. Five of the 18 athletes studied also developed hyponatraemia during the race. The fluid and sodium balance of two of these hyponatraemic athletes, recorded prospectively, was reported in detail. Neither athlete had large sodium losses, and both athletes gained weight during the race with an increase in plasma volume. AVP levels were not elevated in either athlete. A further case report from the 1997 New Zealand Ironman was also reported on an athlete who collapsed at the end of the race having gained 5% in body weight and having sustained a drop in serum sodium concentration to 116 mmol/L. At a later date six of the athletes who had developed hyponatraemia during the 1997 New Zealand Ironman triathlon were studied in the laboratory and compared with six athletes who completed the same race without hyponatraemia. Athletes were given a 3.4 L fluid load over two hours at rest to determine whether those who had previously developed hyponatraemia had an impaired ability to excrete a large fluid load compared to normonatremic athletes. When evaluated this way at rest there did not appear to be any unique pathophysiological characteristic to explain why some athletes develop hyponatraemia in response to fluid overload during prolonged exercise. Fluid overload appeared to be in all of the body fluid compartments without evidence of fluid retention in the small bowel. In response to the high incidence and morbidity of hyponatraemia at the 1997 New Zealand Ironman triathlon, a model aimed at the prevention of exercise-associated hyponatraemia was developed at the 1998 New Zealand Ironman triathlon. A pre-race education programme on appropriate fluid intake together with a reduction in the availability of fluid at support stations was associated with a significant reduction in the number athletes attending for medical care with hyponatraemia (3.8% of race starters in 1997 compared with 0.6% of race starters in 1998). Pre-race weighing was introduced as a compulsory requirement of the race to assist with the diagnosis of hyponatraemia post-race. The prevention of exercise-associated hyponatraemia was further studied at the 2000 Capetown Ironman by providing athletes with oral sodium supplementation (median 6.3 g) and matching these athletes to controls who were not given oral sodium supplementation. None of the subjects or controls developed hyponatraemia during this race. Sodium supplementation did however attenuate the loss of weight during an Ironman triathlon. In conclusion, exercise-associated hyponatraemia is a common complication of competing in an ultra-distance triathlon and is often asymptomatic. Female athletes are at increased risk of hyponatraemia. There is a consistent inverse relationship between post-race serum sodium concentration and percent body weight change supporting fluid overload as the aetiology of exercise-associated hyponatraemia. Data from overnight recovery from exercise-associated hyponatraemia also confirm that hyponatraemia is associated with fluid overload. Sodium deficits in hyponatraemic runners are not significantly different from normonatremic controls, indicating that symptomatic hyponatraemia is not due primarily to sodium losses in the sweat. There was no evidence for elevated AVP concentrations in the aetiology of exercise-associated hyponatraemia. Athletes typically lose approximately 2.5 kg of weight during an ultra-distance triathlon, most likely from sources other than fluid loss. Fluid intakes during an Ironman triathlon are more modest than those recommended for shorter duration exercise. When examined at rest there does not appear to be any unique pathophysiological characteristic that explains why some athletes develop hyponatraemia in response to fluid overload during ultra-distance exercise. A preventive strategy to decrease the incidence of hyponatraemia involving education on fluid intake and reduction of fluid availability during the race was associated with a decrease in the incidence of symptomatic hyponatraemia. Oral sodium supplementation in an Ironman triathlon was not necessary to prevent hyponatraemia in athletes who lose weight during the race.