Sunday, May 20, 2012

The ESSA/SDA Conference – Tim Noakes and the Hydration Debate - Part 2

In the last post we looked at the first two of Professor Tim Noakes' main arguments in his recent ESSA/SDA Conference presentation. In Part Two we'll examine the remaining arguments. I'll also draw my own conclusions from having read through the research on hydration and performance, and identify the areas which I feel we still don't have a full understanding.

Argument #3 – That Hyponatraemia is due solely to fluid overload, without any contribution from sodium losses in so-called “salty sweaters”

Hyponatraemia (or low blood sodium) is a condition that occurs in some endurance athletes towards the end of a race. It’s more common in females, slower runners and in cooler conditions. Some people have used the term “overhydration” to describe what happens to these athletes, suggesting that overconsumption of fluid is the issue. Tim Noakes put forward the argument that hyponatraemia is caused by overconsumption of fluid, not underconsumption of sodium.
To demonstrate this, in a 2010 article Noakes graphed some data from an existing 2008 study in which runners’ fluid and sodium intake was varied whilst sweat sodium losses, fluid losses and their effect on blood sodium was measured 1. The first graph showed that the athletes who lost the most sodium and potassium through sweat finished the exercise bout with the highest (not lowest) blood sodium levels.
The second graph plotted blood sodium against body weight loss, showing that almost 95% of the change in blood sodium concentration could be explained by the change in weight alone, with weight loss predicting higher blood sodium levels and weight gain predicting a fall in sodium (which if large enough could lead to hyponatraemia). My only issue with this is that Noakes had previously argued that weight loss was a poor predictor of the change in Total Body Water, but then presented body weight data as a surrogate for fluid losses.
Having said that, simple common sense would suggest that fluid intake has a much greater influence on blood sodium levels compared to actual sodium intake. Blood sodium is controlled within a fairly narrow range of around 130-145mmol/L. Hyponatraemia occurs when sodium falls below 125mmol/L. Yet even the saltiest sweaters only produce sweat sodium concentrations of around 90-100mmol/L – this means that you will ALWAYS lose more water relative to sodium. Blood sodium concentrations will increase even in the most salty sweaters unless fluid intake is greater than losses. Likewise the highest sodium sports drinks like Gatorade Endurance only provide around 35mmol/L of sodium. So no matter how salty your sports drink, it will always have a diluting effect on your blood and lower your blood sodium concentration. The reduction in sodium concentration may be slightly slower compared to drinking water, but clearly the amount of fluid is far more important than the sodium content.

Argument #4 – That the most successful athletes are usually those who drink the least, lose the most weight and finish with the highest body temperatures during competitive sport
Tim provided observations from marathon events that the fastest athletes are usually amongst those who lose the most weight (again weight loss used as a surrogate for sweat losses), despite the fact that they’re competing on the course for a shorter time 2. This has been shown in several other studies and on face value appears to present a convincing case.
The relationship between % body weight change and race finish time in 64 competitors at the 2009 Mont Saint-Michel Marathon. r = 0.217, p < 0.0000001. Source: Br J Sports Med. 2011 Nov;45(14):1101-5.

However care needs to be taken when interpreting observational data. Firstly, despite the statistical significance of this finding only 4% of the variation in race finish time was explained by weight change. Secondly, remember that body weight only explains about 60% of the variation in blood sodium levels. And finally, just because there is an association between two factors doesn’t mean that one necessarily causes the other. Is it simply that the fastest athletes simply have less opportunity to drink and therefore consume less fluid?
To get a better picture we instead need to look at intervention studies. There are surprisingly few well designed studies where participants were given different amounts of fluid and asked to perform a time trial – I’ve previously written up these studies so click here if you want more details. But to cut a long story short the only study that’s measured this properly showed that drinking ad libitum (as much or as little as desired) was just as effective as deliberately drinking to prevent a set amount of sweat losses 3.

Other arguments made during the presentation
Sweat sodium is a reflection of sodium intake, not the other way around. Therefore salty sweaters are simply those who consume a high sodium diet leading up to a race.
Noakes’ evidence was based on two papers. In the first in 1998 subjects consumed a standardised high, medium or low sodium diet (348, 174 and 66mmol/day respectively) for 3 days in a environmental chamber at 25 deg C, then the chamber was cranked up to 40 deg C for 5 days 4.  They showed that sodium losses from sweat, urine and faeces over a 12 hour period increased with sodium intake (see table pasted below). However it should be noted that these measurements were taken at rest, not during exercise. Whether the same is true for sweat sodium during exercise I'm not sure, I'm certainly not aware of any published literature that's investigated this. What we also don’t know is whether the amount of sodium consumed DURING exercise will result in a change in sweat sodium later during the same exercise bout.
The other paper described a study conducted during World War II, when the Americans looked at fluid and sodium needs for the war in the Pacific 5. They similarly showed a change in sweat sodium to match intake, however again this was not measured during exercise.

That pre-competition hydration is not important, because athletes who begin exercise more dehydrated will simply become more thirsty and drink more fluid ad libitum to "catch up" and prevent any loss of performance.
This may well be the case, but what if on a hot day the amount needed to quench thirst and reduce blood sodium concentrations was greater than the fluid an athlete has access to or can practically carry? Or more than the rate of that the stomach can empty (~1-1.5L/hour) or what you can tolerate sloshing around in your gut whilst trying to win a race? If I had the choice to begin well hydrated and only need 300mL/hour running an ultramarathon, or start more dehydrated but need to drink 800mL/hr to maintain the same level of performance, I know which one I'd choose.

Applying this knowledge in the real world of sport

Fluid availability
On a practical note, the trouble that we have in some sports is that the availability of fluids for athletes when they feel the need to drink may become an issue. In a nicely controlled lab on a stationary bike, with no race tactics, distractions, communication with teammates, and with as much access to fluids as you want, it's easy to show that drinking to thirst is sufficient to optimise performance. What we still don't know is whether athletes in all competition situations are sufficiently in tune with their perception of thirst to drink enough to optimise performance without any other prompts or specific race fluid plan, and whether they have enough access to fluids to drink ad libitum in all conditions. This is a topic that would benefit greatly from more research into what actually occurs during competition in a variety of sports (eg. triathlon, ultramarathons, road cycling, adventure races, etc.).

Fluid is often an important source of carbohydrate too
With all this talk of fluid and sodium, it’s important to remember that sports drinks also provide carbohydrate, and the benefit of carbs to performance in longer endurance events is well established. Given that research evidence now shows benefits to performance from very high carbohydrate intakes (up to 100 grams an hour in competitive cyclists and triathletes) it’s hard to see practically how athletes would achieve this without significant fluid intake (of at least 400-500ml an hour). This is where the value of individually tailored race nutrition plans becomes important. Sports dietitians can work with athletes to come up with practical strategies to ensure that performance is optimised with adequate carbohydrate, but without resulting in overconsumption of fluid (and the risk of hyponatraemia) in order to achieve this.

What it all means
Drawing conclusions from all this research requires an ability to understand and interpret the data in a practical context. For me, a lot of the hydration research of the past has taken valid scientific data and misinterpreted its meaning, drawing inappropriate conclusions. These interpretations have resulted in subsequent research mostly going in the one direction, which has further reinforced these conclusions rather than looking at the problem from different perspectives.
Tim Noakes however has dared to interpret the data differently, drawing very different conclusions and taking his own research down a different path. He should be congratulated for this, as academic science is all about challenging established thinking by examining problems from different angles. Sadly not many people have looked at hydration science from another perspective in the way Tim has, and as a result we’re still missing some important pieces of the puzzle, especially at the elite performance end of the spectrum.
So what conclusions would I draw from the research?
  • Change in body weight is a rough guide to sweat losses during endurance exercise, but only a rough guide. It’s widely used because it’s very easy to measure, but just because it’s easy to measure doesn’t make it the best tool to use.
  • Research that investigates the effect of fluid intake and hydration on performance really needs to measure dehydration in terms of blood sodium, osmolarity and Total Body Water as well as changes in body weight. I'm not convinced that we fully understand which of these parameters is most important one to regulating exercise performance, so we need to continue to measure more than one.
  • Large sweat losses accompanied by a complete lack of available fluid can have performance and health consequences if untreated. However for the vast majority of athletes this is not an issue because at least some fluid is available to them, or the event is too short for this level of untreated dehydration to occur.
  • Elevated core temperature to a dangerous point during actual outdoor competition is quite rare, as the athlete simply paces themselves to prevent this from occurring. However any downregulation to prevent an increase in temperature is therefore detrimental to performance.
  • Collapse towards (or after the end) of ultraendurance events races is more likely to be due to Exercise Associated Postural Hypotension, hypoglycaemia or another medical reason rather than heat exhaustion or dehydration.
  • The amount of fluid consumed by athletes has a far greater influence on blood sodium levels (and therefore risk of hyponatraemia) than the amount of sodium consumed. Hyponatraemia is therefore caused by overdrinking fluid rather than not consuming enough sodium.
  • Sweat sodium at rest is a reflection of dietary sodium intake, with those on a higher sodium diet producing saltier sweat. However we can't be sure if the same is true for sweat sodium during exercise.
  • The fastest finishers in endurance events are often found to have lost the most weight, but this does not necessarily mean that dehydration improves performance. To properly understand the relationship between fluid losses, consumption and performance requires intervention studies rather than simple observation and correlation data.
  • In self-paced, fixed distance or duration events that mimic real competition (including adequate wind resistance and cooling) there is only one intervention study that’s actually investigated the effect of various levels of fluid intake on performance. The results of this study showed that when athletes drank as much or as little as they liked they optimised performance compared to when no (or inadequate) fluid was available to them. Consuming extra fluid above and beyond what the athletes naturally chose to drink did not further improve performance.
  • Starting hydration for athletes is probably still important (especially in warm environments), because thirst in an already dehydrated athlete may dictate that they need to drink more than they can physically or practically manage. This would likely result in loss of performance due to drinking below ad libitum levels.

The missing pieces of the puzzle
Despite these conclusions there’s still a lot we don’t know about hydration and performance, particularly at the elite end of sport where every second counts. I‘m sure there’s a PhD or two in here for those willing to take it on.

  • As far as I’m aware no one has ever repeated the one study that looks at ad libitum fluid intake and performance to validate that result. This needs to be done in an environment that simulates real-life competition – real levels of wind resistance, fixed distance/duration that reflects real competitive races, in self-paced time trials. The studies also need to measure blood sodium, osmolarity and Total Body Water rather than relying on weight changes alone to predict sweat losses.
  • From my discussion with Tim at the ESSA/SDA Conference thirst was not measured in the above study, so we don’t know exactly how and why the participants chose to drink the amount that they did. Is “drink according to thirst” sufficient advice or are athletes influenced by pre-conceived ideas of how much they should or shouldn’t drink.
  • We don’t know whether advice to simply drink to thirst is sufficient for elite athletes in the midst of competition, or whether other factors going on in a race would distract athletes enough to prevent them from regulating their drinking enough to optimise performance.
  • We still don’t have a clear picture of what’s most important to measure during hydration studies that measure sports performance outcomes – TBW, blood sodium or osmolarity, or a combination. Studies should aim to measure these along with an athlete’s perceived thirst to get a better feel for how this regulatory mechanism works during exercise.

1.  Noakes T (2011). Changes in body mass alone explain almost all of the variance in the serum sodium concentrations during prolonged exercise. Has commercial influence impeded scientific endeavour? Br J Sports Med. 45(6):475-7.

2.  Zouhal H et al (2011). Inverse relationship between percentage body weight change and finishing time in 643 forty-two-kilometre marathon runners. Br J Sports Med. 45(14):1101-5.

3.  Dugas J et al (2009). Rates of fluid ingestion alter pacing but not thermoregulatory responses during prolonged exercise in hot and humid conditions with appropriate convective cooling. Eur J Appl Physiol. 105:69–80.

4.  Allsopp A et al (1998). The effect of sodium balance on sweat sodium secretion and plasma aldosterone concentration. Eur J Appl Physiol. 78: 516-521.

5.  Conn JW (1962). Some Clincal and Climatological Aspects of Aldosteronism in Man. Trans Am Clin Climatol Assoc. 74:61-91.


  1. I think if you read Waterlogged you will find that all these questions have in fact been answered already. I simply did not have time in my presentation to present all the evidence that I collected in the 31 years it took to write Waterlogged. Thanks for highlighting these issues.

  2. I do no know Tim Noakes for a long time. However, I became a strong supporter of him. I do not mean that he is right in all he says, but at least he is able to give different view, he is able to criticize whatever he or other have done in the past.

    Please keep doing as great job as you have been doing so far.

  3. maggie.danhakl@healthline.comApril 18, 2014 at 9:35 AM


    Healthline just published a visualization of your daily value of sodium. In the chart, you can see what half of your DV of salt looks like for 30 foods:

    This is very valuable content as it puts nutrition information into perspective and helps a person understand how much sodium is actually in their food. I thought this would be of interest to your audience and wanted to see if you would include this as a resource on your page:

    If you do not believe this would be a good fit for a resource on your site, even sharing this on your social communities would be a great alternativeto help get the word out.

    Thanks for your time reviewing. Please let me know your thoughts and if there are any questions I can answer.

    All the best,
    Maggie Danhakl • Assistant Marketing Manager
    p: 415-281-3124 f: 415-281-3199

    Healthline • The Power of Intelligent Health
    660 Third Street, San Francisco, CA 94107 | @Healthline | @HealthlineCorp

    About Us: