(1) Department of Human Nutrition, University of Otago, New Zealand
(2) Sport and Exercise Science, Swansea University, United Kingdom
* Corresponding author Email:
The reported incidence of exercise-associated hyponatraemia (plasma sodium concentration <135 mmol/L) in endurance sport has increased notably in recent decades. Sodium supplements have been suggested to attenuate the decline in plasma sodium concentration and hence, prevent the effects of exercise-associated hyponatraemia on performance. This article reviews intervention studies that have assessed the impact of sodium supplements on plasma sodium concentrations and/or performance during endurance exercise.
Despite significant results in some laboratory studies show benefits of sodium supplementation, more recent field studies suggest that sodium supplementation has little impact on plasma sodium concentration during a racing situation and no effect on performance. These discrepancies are likely because of design differences between the studies. However, a well-controlled crossover field trial in a hot environment is still required in order to develop practical recommendations.
Exercise-associated hyponatraemia (EAH), is a blood electrolyte disorder induced during the 24 hours, following exercise. It is defined as a plasma-sodium concentration (plasma [Na[+]]) less than the normal reference range; for most laboratories it is found to be 135 mmol/L.
Mild hyponatraemia (plasma [Na[+]] 130 – 135 mmol/L) is often asymptomic[2,3,4,5], although mild symptoms can develop. These include bloating, nausea, lethargy, vomiting and headache[1,7,8,9]. Whilst such symptoms are not life-threatening, they likely have some effect on exercise performance. They are also non-specific, and can often be confused with the symptoms of dehydration. As the athlete’s plasma [Na[+]] decreases below 130 mmol/L, the symptoms correspondingly increase in severity. This occurs as a result of cerebral oedema, producing symptoms, such as confusion, disorientation, seizures, respiratory distress, coma and even possible death[1,10,11,12,13].
The incidence rates of EAH for endurance races has been reported to be up to 30% amongst Ironman triathletes[1,13,14,15,16,17], and in marathon runners up to 13%[3,18,19,20]. Such endurance events host large participation numbers; the 2010 London marathon, for example, hosted 36,000 runners. With this in consideration, such incidence rates correspond to a large absolute number of EAH sufferers in the sporting population.
Whilst excessive fluid consumption has been a clear causative factor of EAH, there is evidence to suggest that excessive sweat sodium loss may also contribute[7,8]. It has been speculated that sodium supplementation during endurance exercise could attenuate the reduction in plasma [Na[+]] in these situations, therefore reduce the risk of EAH and potentially improve performance. The research stems from both laboratory and field studies, each with their own advantages and disadvantages. In this review, intervention trials with a high and low sodium trial, which measure plasma sodium during an endurance exercise protocol will be discussed. These studies are summarised in Table 1.
Summary of sodium supplement intervention trials.
One of the first laboratory studies was undertaken in 1991 by Barr, Costill and Fink. Eight participants completed a crossover intervention of three cycling trials, at 55% VO2max for six hours. Participants consumed fluid to match their sweat rate, with either water and/or a 25 mmol/L (29 mmol/h) NaCl saline solution, or they ingested no fluid during the three trials. Barr et al. observed no significant differences in plasma [Na[+]] between the two ‘fluid’ trials (
These results were similar to those seen in another crossover study by Sanders and colleagues. Six endurance trained male cyclists participated in three trials, again cycling at 55% V̇O2max, but for a shorter time for 4 hours as compared to Barr et al. This time the participants ingested 4.6 mmol/L (4 mmol/h), 50 mmol/L (48 mmol/h) and 100 mmol/L (96 mmol/h) of a sodium carbohydrate drink (~0.965 L/h).
Similar to the findings from Barr et al., plasma [Na[+]] was maintained regardless of whether participants were supplemented with sodium. However, Sanders et al. built on these findings to suggest why plasma [Na[+]] was maintained. When salt tablets were consumed, the intracellular fluid (ICF) moved into the extracellular fluid (ECF), expanding plasma volume and preventing any large increase in plasma [Na[+]] levels. In contrast, when salt tablets were not consumed, the ECF was reduced and the ICF was maintained. Thus, plasma volume decreased and plasma [Na[+]] and osmolality were maintained. This phenomenon was further explained by a significant decrease in renal-free water clearance during the sodium capsule trials compared to the solution-only trial (
The results of these laboratory studies suggested that sodium supplementation may have little effect in preventing EAH[23,24]. However, other studies challenged this view. A crossover intervention trial by Vrijens and Rehrer in 1999 demonstrated that sodium could play an important role in the prevention of EAH. Again cycling at 55% VO2max, but only for three hours, and as with the Barr et al. study, ingesting fluid equal to sweat rates. The participants ingested either water, or Gatorade (18 mmol/L Na[+]). In contrast to Barr et al. and Sanders et al., the Gatorade intervention significantly attenuated plasma [Na[+]] reduction compared to water (water vs. Gatorade = -2.48 mmol/L/h vs. -0.86 mmol/L/h). Unlike the previous two studies, this study reported that one participant became hyponatraemic (plasma sodium 128 mmol/L) during the water trial, interestingly with a fluid intake rate lower than the mean fluid intake rate for the group. It should be noted, as the intervention in the Vrijins and Rehrer study consisted of consuming Gatorade, differences other than sodium were present between the trials, for example carbohydrate and other electrolytes. They also observed that time to exhaustion decreased with lower plasma sodium concentrations. This suggested a role for sodium supplementation in performance, although more likely to be because of the consumption of carbohydrates.
Despite these apparent limitations in the Vrijens and Rehrer study, a later study performed by Anastasiou and colleagues agreed with their results. This trial involved 13 untrained men completing a multi-disciplinary intervention of cycling, walking and calf raises for three hours. Again fluid was ingested equal to sweat loss, and four trials were completed, including high sodium (Gatorade thirst Quencher 36.2 mmol/L), a low sodium (Gatorade Endurance 19.9 mmol/L), water or an artificially flavoured placebo. Both the low and high sodium solutions attenuated the decline in plasma [Na[+]] during the exercise intervention, compared to the water and placebo and some of the participants on the water and placebo had plasma sodium concentrations less than 135 mmol/L at the end of the exercise protocol. Again discrepancies in carbohydrate ingestion existed between the two sodium and two non-sodium trials, and as differences were only observed between the sodium and non-sodium trials, and not between the low- and high-sodium trials, it suggested that something other than sodium could be responsible.
These four studies provide conflicting results on the ability of sodium ingestion to influence blood-sodium concentrations, which can be attributed to differences in methodology. Sanders et al. suggested that the steeper declines in plasma [Na[+]] observed in the Vrijens and Rehrer study could also be because of differences in urine output, which were considerably lower in the Vrijens and Rehrer study. This is likely due to anti-diuretic hormone secretion in response to exercise stress to conserve plasma volume. When combined with greater sweat sodium losses and large fluid intakes, it is understandable why Vrijens and Rehrer reported a greater decrease in plasma [Na[+]].
These studies also highlight some limitations with laboratory-based data collection, particularly in terms of exercise prescription. Exercising at a set VO2max may stimulate physiological responses to endurance exercise, but it is not applicable to a racing situation, where the degree of exercise intensity is higher and constantly being modified. Twerenbold et al. attempted to address some of the limitations associated with the laboratory-based studies in a 2003 crossover-intervention study, recruiting 13 well-trained, healthy female runners to participate in three four-hour running time trials around a 400 m track. The participants consumed 1 L/h of a high sodium-carbohydrate solution (25 mmol/L), low sodium-carbohydrate solution (15 mmol/L), or water for each of the respective three trials.
The change in plasma [Na[+]] from pre-run to post-run was significantly smaller in the high-sodium trial compared to the water trial (sodium vs. water, -2.5mmol/L vs. -6.2mmol/L;
Although the Twerenbold et al. study used a race-like exercise prescription to assess sodium supplementation, it has been criticised, alongside Vrijens and Rehrer and Anastasiou et al. of over-hydrating their participants and inducing dilutional EAH. The standardised-fluid intakes of 1 L/h were shown to elicit dilutional hyponatraemia (below 135 mmol/L) in 69% of their participants, despite being within the recommended fluid intake guidelines for athletes at that time. Indeed, voluntary fluid consumption tends to be about half an athlete’s sweat rate for most sports, which highlights the importance of investigating the effects of sodium supplementation when athletes consume fluids ad libitum, such as during field studies.
The first field study was conducted by Speedy et al., which investigated the effects of sodium supplementation during the 2000 Cape Town Ironman Triathlon. Thirty-eight athletes were recruited at race registration, three days prior to the race. These participants were issued with sufficient salt tablets to provide 700 mg/h of sodium, during the race, which lasted for approximately 12.5 hours. Some of the tablets also contained carbohydrate, but this was not the case for all the salt tablets. The control group (
There were no significant differences in the change in plasma [Na[+]], when matched for pre-race plasma [Na[+]], neither was there any difference in performance between the groups. Interestingly, the prevalence of EAH during this race was particularly low; only one athlete developed asymptomatic hyponatraemia out of the entire field. Previous prevalence rates have been much higher; up to 29% of race finishers were reported to have developed EAH by Hiller et al. in the Hawaiian Ironman Triathlon, and Speedy et al. reported 18% of race finishers in the New Zealand Ironman Triathlon, despite similar temperatures and relative humidity. The low prevalence rates of EAH observed in this Ironman Triathlon could therefore reflect the results gathered in this study; plasma [Na[+]] increased and weight decreased in both the control and supplementation groups, which suggested that participants became hypohydrated, and did not over consumed the fluid.
A similar trial was conducted by Hew-Butler et al. a year later during the 2001 Cape Town Ironman, in which 145 triathletes, with a finish time around 12.5 hours were randomised to receive either salt tablets (10.6 mmol Na[+] per tablet, between 10.6–42.4 mmol/h) or a placebo tablet (596 mg starch per tablet). The randomisation and blinding protocol was advantageous compared to the study by Speedy et al., as it reduced selection-bias in analysis, and allowed a much more comparable control group. Despite this, the results of the Hew-Butler et al. study supported the findings of Speedy et al. and showed no significant differences in plasma [Na[+]], body mass change, performance or prevalence of medical care between the intervention and placebo groups.
Hew-Butler et al. observed a number of participants developing mild hyponatraemia (post-race [Na[+]] 130–135 mmol/L), even though they lost weight during Cape Town race. This should be interpreted with caution as pre-race weight was determined three days prior to the start of the race, so it cannot be confirmed that the loss of body mass occurred during the race and not before the race. This highlights that although over consuming fluid is an important causative factor, other factors can play a role in the development of EAH. There was no difference in post-race plasma [Na[+]] between the intervention and control groups in both the field studies[28,31], however this may be reflective of the inherent limitations with these study designs, where dietary intakes before and during the race are not controlled. The Cape Town Ironman provided all athletes with sports drink (Energade, [Na[+]] = 18 mmol/L), every 20 km in the cycling leg and every 2.5 km in the running leg. It is therefore, probable that the control group consumed sodium-containing sports foods, with a similar consistency to those in the study by Vrijens and Rehrer and hence, it is difficult to directly compare the results observed in the intervention group to the control group. Further, the fact that both these races were undertaken in South Africa and the low prevalence of hyponatraemia that is reported could indicate that the ad-libitum food and fluid intakes in this country are different to other races, where much higher prevalence of hyponatraemia have been reported.
Cosgrove and Black tried to address some of these control issues by conducting a blinded, randomised-crossover study of sodium supplementation in a 72 km road cycling time-trial. Nine well-trained cyclists (5 male cyclists, 4 female cyclists) consumed either a 30 mmol/h sodium, or a placebo.
In line, with the research at the Cape Town Ironman, Cosgrove and Black found that sodium supplementation had no effect on plasma [Na[+]] change (relative change pre-race to post-race, salt = 0.56%, placebo = 0.47%,
In this review, the authors have referenced some of their own studies. These referenced studies have been conducted in accordance with the Declaration of Helsinki (1964) and the protocols of these studies have been approved by the relevant ethics committees associated to the institution in which they were performed. All human subjects, in these referenced studies, gave informed consent to participate in the studies.
It is interesting that all of the studies, which have shown a beneficial effect have been between the trials containing carbohydrate-electrolyte beverage and water trials. This raises the question as to whether it is the sodium per-se or another substance within the drink, which is eliciting the effect. Indeed, the ingestion of carbohydrate is likely to attenuate the stress of exercise. However, as these carbohydrate electrolyte beverages contain more than just sodium and carbohydrate, any of the drink components could have influenced the results. However, as they were not separately assessed, the exact reasoning for these results cannot be confirmed.
The obvious criticism that the studies reporting beneficial effects are because of the over-consumption of fluids that may well be true, but both Barr et al. and Sanders et al. replaced sweat losses, but yet did not see any differences in plasma sodium between the water and sodium containing trials. Nearly all of the studies have failed to result in high rates of hyponatraemia, even when large amounts of sodium-free fluid have been ingested, this conflicts the observational data from endurance races, where up to 30% of athletes have been reported to have EAH. Future research is required to concentrate on those with a history of hyponatraemia to determine why some athletes are more susceptible to hyponatraemia. This sub-population may benefit from sodium supplements, but before conclusions can be made further investigations are necessary.
Whilst there is some suggestion from laboratory studies that sodium supplementation could reduce EAH incidence, and improve performance, in particular during exercise in the heat, recent field trials have demonstrated that sodium supplementation has no effect on plasma sodium concentrations during a racing situation.
EAH, exercise-associated hyponatraemia; ECF, extracellular fluid; ICF, intracellular fluid.
All authors contributed to the conception, design, and preparation of the manuscript, as well as read and approved the final manuscript.
All authors abide by the Association for Medical Ethics (AME) ethical rules of disclosure.
Summary of sodium supplement intervention trials.
|Year, Author, Title||Aim, Design||Methods||Results||Comments|
||Randomised crossover design. Effect of water, saline or no fluid on plasma [Na+] and performance.||8 participants. 3 × 6 h cycling at 55% V̇O2max. Saline (25 mmol/L Na), water or no fluid to replace sweat losses.||No difference in plasma [Na+] between water and saline (
||First controlled study to suggest over-hydration increases risk of EAH.|
||Randomised crossover design. Replace sweat losses with either a sodium-free beverage or sodium-containing sports drink during exercise in the heat.||10 trained men (approx. 25 years). 2 × 3 h cycling 55% V̇O2max. Temperature 34°C, 65% RH. Sports drink (S) (18 mmol/L Na) or water (C) given equal to sweat rate. Sports drink trial ingested around 61 mmol Na+.||4 of 10 subjects completed 3 h cycling in both trials.
||One subject diagnosed with hyponatraemia in the water trial. Consumed 1 L/hr. Higher rate of weight loss than the other participants.|
||Should athletes who typically replace only. ~50% of fluid loss are required replace their Na+? Randomised crossover design.||6 cyclists , 3 × 1.5 h rides at 65% VO2 peak
||Water trial resulted in a lower plasma [Na+] compared to no fluid and saline trials. Similar plasma [Na+] was observed between the no fluid and saline trials.||Exercise duration is shorter than events reporting incidence of hyponatraemia.|
||Randomised crossover design. Electrolyte replacement effects on fluid shifts during exercise.||6 cyclists. 3 × 4 h rides 55% V̇O2max.
||Plasma [Na+] maintained in all interventions (
||Similar to Barr et al. (1991). Less rapid declines in plasma [Na+] than Vrijens and Rehrer (1999), less extreme climate and less fluid intakes. First study to ‘blind’ participants using tablets.|
||Non-randomised intervention trial. Influence of oral sodium supplementation on body mass, plasma [Na+] and plasma volume in athletes during the Cape Town Ironman triathlon.||38 athletes (2 female athletes) provided with 700 mg/h salt tablets, all ingested more than 4 g during the race (S). Compared against 133 controls, not given additional sodium (C).
||Matched for body mass: S had a significantly greater decrease in haematocrit, and an almost significantly greater increase than C in plasma [Na+] (
||Not randomised. Did not include athletes who failed to complete the race.
||Randomised crossover design. Different concentrations of electrolyte replacement in exercise performance and physiology.||13 trained women runners. 3 × 4 h runs on 400 m track. Ingested 1 L/h high Na+ (25 mmol/L), low Na+ (15 mmol/L), water.
||Decrease in plasma [Na+] smaller in high Na+ than water (
||Similar results to Vrijens and Rehrer (1999). No observed fluid shifts as seen in Sanders et al. (2001).|
||Randomised trial. Ingestion of additional Na+ during Cape Town Ironman on plasma [Na+], risk of EAH, performance.||145 triathletes competing in the Cape Town Ironman. Control group was randomised to 40 placebo tablets of starch (C). Experimental group given 40 identical looking tablets, with 244 mg Na+ (10.6 mmol) (S). Asked to consume 1–4 (244–1000 mg) tablets every hour.
||No significant differences (
||Food and fluid intakes were allowed ad libitum, meaning athletes could eat other salt products; not considered in this analysis.
||Randomised crossover design. Different levels of Na+ intake on maintaining plasma volume and preventing EAH during prolonged exercise in the heat.||13 untrained men. Cycling/walking for 3 h followed by 5 min rest, calf raisers and steep walking at 5.5 km/h for 45 min. Replace sweat loss during first 3 h, then 150 ml every 15 min during the walking (ACSM guidelines).
||HNa and LNa trials resulted in stable plasma volume. Pl and W trials tended to decrease plasma volume over time. Plasma osmolarity levels were higher in HNa than Pl and W after phase 1 and thereafter. LNa after phase 2, 3 (
||Moderate levels of sodium attenuate the decline in plasma [Na+] and preserve plasma volume.
||Randomised crossover design. Impact of sodium supplementation on performance and plasma [Na+] during a 72 km cycling time-trial.||9 well-trained cyclists (5 male cyclists, 4 female cyclists) cycling 72 km hilly road course three times, separated by 7–14 days. First trial is familiarisation, second and third trials are 700 mg/h sodium supplement or identical placebo. Drink ad libitum.
||No difference in performance (
||Athletes on sodium supplements consumed significantly more fluid ad libitum (
ACSM, American College of Sports Medicine; EAH, exercise-associated hyponatraemia; ECF, extracellular fluid; ICF, intracellular fluid; RH, relative humidity; Temperature, environmental temperature