Welcome to the Science of Sport, where we bring you the second, third, and fourth level of analysis you will not find anywhere else.

Be it doping in sport, hot topics like Caster Semenya or Oscar Pistorius, or the dehydration myth, we try to translate the science behind sports and sports performance.

Consider a donation if you like what you see here!

Did you know?
We published The Runner's Body in May 2009. With an average 4.4/5 stars on Amazon.com, it has been receiving positive reviews from runners and non-runners alike.

Available for the Kindle and also in the traditional paper back. It will make a great gift for the runners you know, and helps support our work here on The Science of Sport.

Tuesday, November 27, 2007

Sports drinks, sweat and electrolytes

A look at sweating, drinking and the role of electrolytes in sports drinks

Today is a pseudo-Part V of our series on Muscle Cramps - I was tempted to call it Part V, but it's a little bit of a departure from what we've been talking about. In our next post, which we will be calling Part V, we'll wrap up this really challenging series and try to summarize all your comments and our posting into one post.

But today, we build on the arguments from yesterday. In yesterday's post, we looked at how hypothesis creation in science must be evaluated with observation, and if the measurement or observation contradicts or fails to agree with the hypothesis, then the theory requires revision!

Well, today, we want to apply that analogy to the issue of fluid and electrolyte replacements. It's an issue that we've touched on before, and even dedicated a post to, but it comes up so often in comments and questions, that we thought we should dedicate another post to it, and throw out some real figures, to prove a point.

So today's post looks at two key concepts:

  1. Are low sodium levels the result of a LOSS OF SODIUM in sweat, or is it the result of excessive fluid intake?
  2. How much difference would it make to drink sports drinks that contain electrolytes? After all, they tell you that they "replace the vital salt your body loses during exercise, which can contribute to hyponatremia and cramp" (ala the marketing and promotional material of those sports drink companies!)
Let's tackle that issue, with the aid of a regular runner - we'll call him Randy.

Some assumptions we must make about Randy

In order to do these calculations, we have to make some assumptions about our imaginary runner:

  • The first, and most important assumption, is that we will have to treat Randy somewhat 'non-homeostatically'. I cannot stress just how important it is that we recognize that during exercise, the body is a finely tuned system, and is able to call on a number of different strategies for regulating its internal situation (this is called homeostasis). This is why we often measure things that are not predicted by equations - there are simply too many variables to factor in. For example, we often find that people do over drink, but they manage to regulate their body's sodium levels - this happens because they have other sources of sodium - hormonal regulation, renal function, and maybe even release of sodium from bone!
The point is that we have to look at the body a little bit "coldly" in order to evaluate these arguments - when we consider what is actually happening, then it becomes EVEN MORE APPARENT that the supposed failures and problems like dehydration and cramps are unlikely to happen! In other words, we will present a number of scenarios where Randy would run into trouble. In reality, that would not even happen, because the body is so well regulated.

  • We must assume body weight - we have assumed that Randy weighs in at about 70 kg. This means that his total EXTRACELLULAR FLUID VOLUME is 17 L.
  • We also assume that Randy is what Gatorade would call a "salty sweater" - this is someone who has higher than normal sodium levels in their sweat. Gatorade themselves can't define exactly what this means, but we've assumed that Randy's sweat has a sodium concentration of 100mM. Note that a fit, well-acclimatized person probably has sodium concentrations of about 20 mM, so Randy really is a "salty sweater"!
  • We also assume that Randy runs for two hours, during which time he sweats at a rate of 1 L per hour (so Randy is running at moderate speed, hence the high sweat rate).
Now, let's take a look at some of the possible outcomes....

Scenario 1 - Randy does not drink at all

In this scenario, shown in the table below, we assume that Randy does not drink at all during his two-hour session. He therefore loses both sodium and water - 2 L of water to be exact, and given his very salty sweat, a total of 200mmoles of sodium (a mmol is a unit for measurement of amount, by the way - for ease of comparison, I've included the mass of sodium in grams as well)

As you will see, Randy's sweat, despite the fact that it's much saltier than most people's, is still very low on sodium compared to his starting sodium concentration (which was 140mM). As a result, his sweat is HYPOTONIC, and the end result is that his sodium concentration GOES UP - it ends at 145mM. Note that it is normal, though on the high side. In fact, when we analysed Comrades runners during our research studies in 2005 and 2006, we found that most runners finished the race with a sodium level between 140 and 145mM, and they were fine. So this scenario is actually pretty good!

Note here that in a "homeostatic system", Randy would simply become thirsty, and therefore drink - this would bring his sodium levels back down, even closer to the starting value. But that's what we look at in our next scenario...

Scenario 2 - Randy drinks WATER, and replaces 50% of his sweat losses

The table below shows the predictions for what would happen if Randy drank 500ml per hour - this is a reasonable rate for someone who drinks to thirst, though it does depend on the conditions.

What you see from this table is that now, the sodium concentration will still be in the normal range of 135 to 145mM. Note that this scenario is again IDEAL - his sodium level is in the normal range, he's been drinking to thirst (approximately) and his physiological situation is 100% normal. In fact, the ideal and most common scenario in reality is probably somewhere between Scenario 1 and Scenario 2 - about 30% to 50% of sweat losses replaced would see you finishing in a "normal" range.

Scenario 3 - Randy drinks WATER, and replaces 100% of his sweat losses

The next scenario we consider is that Randy follows Gatorade's advice to the milliliter and drinks TO REPLACE HIS SWEAT LOSSES EXACTLY. In order to do this, he weighs himself dilligently, works out that he is going to sweat at 1L per hour, and so he drinks 1 L per hour too. The table below shows the predicted changes:

You will notice from the last line of the table that Randy is now HYPONATRAEMIC - his sodium levels have fallen below 130 mM. This is usually going to land Randy up in the medical tent and could even have fatal consequences - people have differing sensitivities to a fall in sodium, but a value of 129 mM would be consider dangerously low.

Quite apart from this, Randy has managed to drink 2 L in two hours, and so depending on how quickly he was running, he might feel very ill just from the strain of drinking that much - I would challenge you to drink this much, unless you are running fairly slowly and it's a hot day!

Therefore, two things can be noted from these examples:
  1. Sweating causes some sodium loss, but because the sodium content of sweat is so low relative to body fluids, your sodium concentration will rise. And it is the concentration that is important, since this controls fluid shifts in the body. It is NOT possible, even for a "salty sweater" like Randy in these scenarios, to LOSE sodium through sweat.
  2. Rather, we have demonstrated that the only possibly theoretical way that the sodium level can fall is if Randy drinks too much. If he drinks more than he sweats, then he will really be in a lot of troube. For example, if Randy took Gatorade's earlier advice and drank 40 Oz per hour (1.2 L), then his sodium levels would fall to 125mM - OVERDRINKING can be FATAL.
Now, some of you may be wondering what would happen if Randy drank SPORTS DRINKS instead of plain water? The theory (and the marketing by the sports drink companies) will tell you that the sports drinks are critical because they contain sodium, which replaces what you lose. Let's look at these scenarios....

Scenario 4 - Randy drinks A SPORTS DRINK, and replaces 50% of his sweat losses

The first point we have to make is that Gatorade has a sodium concentration of 18mM - this is only 13% of what the body fluid concentration is, and only 18% of the sodium concentration of the sweat in our examples. It does not take a mathematician to work out that you cannot replace the losses if your replacement fluid is different from the original! But let's play with the numbers and see.

First, we consider what might happen (in theory, remember), if Randy replaces 50% of his sweat losses with Gatorade:

So, what this table shows is that Randy sweats out 2L in his run, and loses 200mmol of sodium. However, because he is putting some of those electrolytes back, his overall sodium loss is 182 mmol - this equates to about 4.2g of sodium.

You may recall from scenario 2 that if he drank just water, his sodium loss would have been 4.6g - the difference is therefore 0.4 grams! When you consider the volume of 17L that this is going into, then you begin to realise that the sports drinks are making a tiny, tiny difference to the body's sodium concentration! In fact, the difference is only about 1mM - Randy's sodium level would have been 136.4 mM with plain water, and it's predicted as 137.4mM on sports drinks!

That difference is absolutely minimal - in fact, it's negligible. A few people wrote in wondering whether a 1 or 2mM difference might be important - I can assure you that it's not. If it were, then in the time you have taken to read this, you would have seen your sodium levels shift by this "critical" amount! So the bottom line - sports drinks simply DO NOT help prevent the fall in sodium levels caused by potential overdrinking.

Scenario 5 - Randy drinks A SPORTS DRINK, and replaces 100% of his sweat losses

For our final scenario, consider what happens if Randy now drinks to replace all his sweat losses, but he drinks sports drinks instead of water. You may recall from scenario 3 that if he drank only water, his sodium level would be predicted to fall to 128.6mM - a level associated with hyponatremia.

The table above shows that if he drinks sports drinks, then his sodium level will fall to 130.6 mM - this is a difference of 2mM compared to drinking only water.

Again, this is a minimal, and negligible difference - the fact of the matter is that whether Randy drinks water or sports drinks, he still becomes hyponatraemic and could be running the risk of death as a result. Again, I have to emphasize - this difference of 2mM is insignificant - both final values are dangerously low, regardless of what he drank during exercise - and this was only in a two-hour run!

To illustrate this, if Randy ran a full marathon and took 4 hours, following this same pattern of drinking to replace all his sweat losses with a sports drink, then his final sodium concentration would be 121 mM - he would almost certainly end up in hospital. Why? Because he drinks too much of a sports drink that is hypotonic and unable to defend the body's sodium concentration.

So what is Randy to do? What is the best strategy?

So, having run through these five scenarios, I hope it's quite apparent that if your sodium level drops, it is NOT because you are losing salt, it's because you are over-drinking! I hope it is equally apparent that drinking a sports drink does nothing to change this situation.

So what to do? For our imaginary runner, the answer would seem quite obvious - don't overdrink! That only seems obvious because we get ourselves into terrible states of confusion by talking about what overdrinking means, and should we also supplement with salt?

The reality is that most people will finish exercise with a normal or slightly elevated sodium concentration - this is in fact the normal, and desired response. In otherwords, of the scenarios we have looked at here, the "ideal" probably lies somewhere between not drinking, and drinking about 50% of fluid losses. If you do this when you exercise, your sodium level will not change substantially - it will be neither too high or too low.

But let's simplify this as much as possible - ALL YOU NEED TO DO IS LISTEN TO YOUR BODY AND YOU WILL BE FINE. In otherwords:
  • If you are thirsty, then drink!
  • If you are not thirsty, then don't drink!
  • Don't waste money on salt tablets - you don't need them!
Because we're too stupid to think for ourselves

The problem that we (as a sports community) find ourselves in today, with this huge confusion, is the result of allowing ourselves to be treated as stupid. Yes, that's right - the companies that sell sports drinks think that you are too stupid to drink when you should, and that your body is too stupid to realise when it needs salt! Our species learned how to make fire, put a man on the moon, and created a communication system that allows you in Colorado (or wherever you are in the world) to read the words I am typing in Cape Town! But when it comes to drinking - we're just not capable! Surely that doesn't make sense!

The reality is that your body is designed perfectly to regulate its internal environment unless you have some disease or potentially critical problem (being lost in the desert without food or water for a week is one that springs to mind!). Yet for some reason, every animal in the animal kingdom EXCEPT for humans is able to drink when they need to, eat the right foods and get the right nutrients when they need to!

So forget the salt tablets, forget trying to work out what you need to eat and drink when you exercise, and just listen! Your body knows better!


In this post, we've tried to illustrate some of the theoretically possible outcomes when you drink plain water during exercise compared to drinking sports drinks, which contain electrolytes.

What we have shown is that the electrolytes in sports drinks make a very tiny difference - not enough to prevent hyponatremia from developing and certainly not enough to back up the marketing claims made by the companies who tell you that you NEED the electrolytes.

We have used these mathematical scenarios to illustrate the theoretical point - I just want to point out again that in reality, when you do drink water and sports drinks, your body makes adjustments that prevent your sodium level from falling (up to a point - eventually, your poor physiology would be over-stressed and you would start to run into trouble!). So the reality is that even if you do make the mistake of drinking too much, your body does the best possible job of protecting you anyway.

One example of such a mechanism is that when your sodium level begins to fall, you are no longer thirsty, and the body is actually even able to release less sodium into the sweat, just so that you conserve it! It's just as well that our bodies are so 'smart', because if we did not have these regulatory mechanisms, can you imagine how many people would be in serious trouble because they did not simply listen to the body and drink to thirst?

So bottom line - drink to thirst, don't worry too much about what you drink, but just make sure you are getting enough ENERGY in - the one thing that the sports drinks provide that is required is glucose, and so plain water is unlikely to be sufficient for longer than about 2 hours of exercise.

Looking ahead - a wrap of cramping tomorrow

So that's it for today's little 'diversion' from our series. Tomorrow we will wrap up the cramp series, and put our minds to what lies ahead!



Ryan said...

In all of your scenarios where Randy consumes a beverage, his electrolyte concentration drops. Sports drinks do not have enough electrolyte in them to replace the drop, so should runners be taking salt tablets?

Ryan said...

Salt tablets such as these?
They have 360 mg of sodium in them.

Ross Tucker and Jonathan Dugas said...

Hi Ryan

You'll see I have actually added to the original post in response to your comments - the answer is NO, a huge, resounding NO. You don't need salt tablets, all you need to do is listen to your body and drink when you're thirsty!

The body knows when you need fluid, and it knows when you need salt - the whole thirst mechanism is in fact designed for this very reason, and so all you need to do is listen to it!

The industry has gone mad - first they said you could lose too much water. Then when that was shown to be wrong, they said you're losing too much salt! Now that is also being shown to be untrue, so something else will take its place - it's a joke, quite frankly.

Just drink when you're thirsty, and you'll be fine.


Unknown said...


Also note at the end of scenario #2:

In fact, the ideal and most common scenario in reality is probably somewhere between Scenario 1 and Scenario 2 - about 30% to 50% of sweat losses replaced would see you finishing in a "normal" range.

So in reality, you can stay in good shape by just drinking less than 50% of sweat loss, which I assume will correlate nicely with what your thirst is telling you.


Unknown said...


I've enjoyed the series on muscle cramps. It confirmed my suspicions that they were more related to fatigue than anything else.

I hope when you wrap it up you will give us some good advice. It sounds like the science isn't yet there to tell us something as simple as "drink to thirst", but it would be nice to have something to go by other than "don't run too hard".


Anonymous said...

Somewhat off topic but could you address the hydration situation when going to high altitude? Part of the acclimatization process is a quick ridding of fluid for a few days to induce an artificially high hematocrit. Is the bodies thirst mechanism sufficient in this situation? Or should the common advice of "drink more than you think you need" still be followed by mountaineers? Don't think electrolytes have ever been suspected to affect mountain sickness or other altitude maladies but not sure it's been tested either.

Ross Tucker and Jonathan Dugas said...

Hello Stan and Clyde

Thanks for the questions.

Stan, first, you're quite right, we need to give more constructive advice than simply "don't run too hard!"

And so yes, we do plan to wrap up with some constructive advice, though that can be a little tricky. It's actually not difficult to see how the whole electrolyte and fluid concept takes root and becomes so "popular" - it speaks to what seems to be a basic need of people, namely the ability to blame the problem of cramp/fatigue on one single problem, which can be easily solved!

The same goes for the whole lactate causing fatigue theory - it's almost comforting to pin the blame on one molecule or electrolyte. The reality is unfortunately a little trickier than this. And it does highlight just how complex exercise physiology is - I know that Amby Burfoot has said that exercise physiology is a lifetime from figuring out what really happens during exercise - I think he's right. But we'll try when we wrap up the series!

And then Clyde, good point. One of the first responses to arrival at altitude is a diuresis - this is partly driven also by changes in the pH of the body fluids caused by changes in ventilation. I must confess that I've never come across studies, but the topic is filed away for future reference! A series on altitude was in the pipeline anyway. As for fluids and altitude, we'll cover that in a post next week some time!


Ryan said...

Your example is someone running for 2 hours. What about the situation where someone runs for 5 hours in heat, such as a slow runner in the Chicago Marathon 2007? You would think they would get severely hyponatremic because they are drinking excess water to stay cool. Another situation is the ultra-events such as Iron Mans and Ultra marathons. What about salt tablets in those situations?

Ross Tucker and Jonathan Dugas said...

Hi Ryan

Nope, not even then. Here's the numbers to prove it:

If you take a very salty sweater (and I must just emphasize that this value we have used in our scenarios is ridiculously high - we only used it to prove the point - in a "normal" person with a salt content of say 50mM (on the high side of normal) it's even more the case, as I'll show you in a moment.)

Anyway, let's say our guy is losing 100mM sodium in his sweat, and he runs for 5 hours. That's pretty slow, so we will assume that he sweats at about 800ml per hour (it's hot and there's no wind cooling), for a total of 4L.

If he DRINKS TO THIRST (which is what we're saying he should), then he'll probably drink about 400ml per hour - that's a whole bottle of fluid every hour, just about. Then we calculate that his final sodium level will be 135 mM - this is still normal.

However, if he drinks more than this (because he's been told by Gatorade to drink more), then yes, his sodium levels would fall.

NOW, SHOULD HE TAKE SALT TABLETS? The answer is still no, because two wrongs do not make a right! In theory, the use of salt tablets would prevent this fall, but then again, if he'd just listened to his body to begin with, he would never need them! It seems to me that this option, of just listening to your body, is a far easier and cheaper, and ultimately safer one!

Because what we are creating here is a situation where we are going to see people who have ingested too much water, and then to compensate for their mistake (because that's what it is), they take salt - as a result, you'll have what the medical fraternity will call HYPERVOLEMIC NORMONATREMIA - that means high fluid volume, normal sodium.

I hope it's clear to you that this is still bad for you - you do not wish to be high volume - there are a number of medical complications that may result. So rather save the trouble and just drink to thirst.

Just finally, if we consider what would happen when someone has a normal salt concenctration in their sweat, then it's even more obvious.


Now, we assume that our runner is doig the marathon in 5 hours, but his sweat contains only 50mM sodium. Now, if he drinks 2 L and loses 4 L during his marathon, his final sodium concentration is 147 mM! In fact, in this "normal" person, he could even drink 3 L (or 600ml per hour, which is a lot!), and his sodium level would still be about 140mM.

The point is, he'd be fine, if he drank to thirst, just like everyone else. If you now gave this guy salt tablets, you could very well make him very very ill

So if you want our advice, forget the salt tablets - you're just making a select group of businessmen and scientists very wealthy as you buy their product...


Anonymous said...


If an endurance athlete keeps exercising and replacing ~50% of fluid lost with plain water, presumably their dehydration continues to increase and their salt concentration continues to decrease.

At some point surely this would become unhealthy?


Nancy Toby said...

Take your example out to 17 hours. Substantial sodium *is* being lost - and without appropriate and plentiful replenishment, an Ironman triathlete can get in serious trouble.

Nancy Toby said...

Actually, even an average female marathoner can get into serious trouble without appropriate sodium replenishment, since median finishing times for female marathoners exceed five hours.

Ryan said...

Hi Ross,
Please see these recommendations from the Western States 100 medical director:
They recommend to include 300 mg to 1000 mg per hour of salt during this endurance event. And they do warn about HYPERVOLEMIC NORMONATREMIA. Basically, they say you need to practice and maintain a balance.

Nancy Toby said...

I should qualify that - in the USA the median finishing time of nearly 200,000 female marathoners is 5:06.


Average body weight for American women is about 65 kg, btw, not too far from your example. By that time the woman has lost probably over 7% of her body weight without fluid replenishment.

Close to that finish line at five-plus hours is the time we're most concerned about both muscle cramping and hyponatremia, not at two hours.

Ryan said...

I'd like to request that you rerun your numbers with Randy from your spreadsheet with a 5 hour example and a 24 hour example and post them, having Randy drinking only water, and having Randy drinking Gatorade. Keep in mind that it is generally accepted that an athlete should lose no more than 3-5% of body weight. What happens?

Meg & Dave said...

Great post guys. I really appretiate the "let your body have what it wants" mentality.

Salt tablets and all this nonsense seem rediculous.

People have very engrained ideas about this subject, and a lot of people really do feel like drinking like gatorade tells them to is a good thing, let them have their placebo. You don't really need to keep throwin evidence out there.

Good work, I can't wait to read more.


Ross Tucker and Jonathan Dugas said...

In response to Mark's comment up there, we must keep in mind that on average people replace 50% of their weight losses when drinking to thirst. In fact, it will be a bit different for everyone depending on a variety of factors.

The bigger picture is that weight losses are not part of the equation, and we mention weight losses since 1) currently they are accepted as an estimate and therefore all the guidelines and advice speak about about it; and 2) all prior research has used it as a primary variable they measured, and so many result are expressed in this manner. But in fact it is just an association and weight losses do not cause you to be thirsty.

In fact your thirst will stimulate you to drink as much or as little as you need to keep your osmolality within a desirable range. When it is hot outside, this means you are thirstier and ingest more. When it is cold outside (and you hardly sweat) you are not thirsty at all and do not drink much.

As you slow down from fatigue your metabolic rate also slows, and therefore your heat production, and therefore your sweat rate. This then means that you then require less fluid, and will be less thirsty on an hourly basis.

It is not a constant and fixed dilutional effect. Your thirst will only stimulate you to drink when the osmolality of the plasma rises above your thirst threshold. When it sits below your thirst threshold, you will not be thirsty.

With regards to ultra-endurance exercise, if you drink to thirst then what will probably happen over many many hours is that your sodium appetite will kick in, stimulating you to ingest sodium. This is the evidence from the rat studies---first they replenish volume, and then they go for the salt.

So as long as you are drinking to thirst it is not unhealthy, because you will not keep diluting the plasma. It rises, you get thirsty, you drink, the plasma is diluted a bit, thirst goes away for a time until it rises again. If sodium losses were part of the problem, and somehow caused the sodium concentration to fall, then you would not be thirsty anymore as the osmolality of the plasma would be falling (and therefore would be below the thirst threshold), and you would stop wanting to drink.

Sorry, that was bit long, but I hope that clarified the issue a bit. Again and again, drinking to thirst is healthy and will keep you out of trouble---on the too little side and the too much side of things!

Thanks for adding to the discussion, Mark!

Kind Regards,

Nancy Toby said...

Is there any evidence that a salt appetite in humans would a) act in a time frame of 30-60 minutes (because that's how long a hyponatremic person would have) and b) that it operates in human under conditions of intense exertion; and c) how would most runners or other athletes restore losses in the absence of salt capsules (which wouldn't be tasted much on the way down)?

http://www.ajcn.org/cgi/reprint/65/2/692S.pdf from 2007 says "a sodium-specific appetite has not been documented in humans".

Nancy Toby said...

Correction: That article was from 1997.

Nancy Toby said...

And correct my math if I'm wrong, but using your figures for sodium loss, proportional to a 65 kg woman, by 5 hours into a marathon (or cycling in hot weather, etc.) she would be down 10.6 g of sodium, equivalent to 4.4 teaspoons of salt, plus 4.6 liters of water (or 7.1% of body weight).

bk said...

Im sure you heard:

JACKSONVILLE, Fla. -- Dr. Robert Cade, who invented Gatorade and sparked the multimillion dollar sports drink industry, died Tuesday of kidney failure. He was 80.

His death was announced by the University of Florida, where he and other researchers created Gatorade in 1965 to help the school's football players replace carbohydrates and electrolytes lost through sweat while playing in swamp-like heat.

A question from former Gator Coach Dwayne Douglas sparked their research, Cade said in a 2005 interview with The Associated Press. He asked, "Doctor, why don't football players wee-wee after a game."

"That question changed our lives," Cade said.

Cade's researchers determined a football player could lose up to 18 pounds during the three hours it takes to play a game. They also determined 90 to 95 percent of the weight loss was water. Plasma volume decreased by 7 percent and blood volume about 5 percent. Sodium and chloride were excreted in the sweat.

Using their research, and about $43 in supplies, they concocted a brew for players to drink while playing football.

Ryan said...

There are two arguments being made here- Lowered electrolytes causing muscle cramps, and hyponatremia. You've convinced me of your fatigue theory regarding cramps, but you haven't convinced me that runners don't need to drink Gatorade (or other forms of electrolyte replacement) I ran your numbers for the Randy example, if he replaces 50% with water, after 5 hours he is hyponatremic with a concentration of 130 mM. If he consumes 50%-18mM Gatorade, his final concentration is 136 mM and is fine. When you extend the numbers out to 24 hours, the numbers get ridiculous and this model really doesn't work anymore. You need more energy replacement than Gatorade can offer. But, for sake of sharing: Keeping you model the same, Randy would need to drink about 0.9 liters per hour to avoid losing too much weight and would need to consume 1925mg/hr of sodium to stay above the limit.

Unknown said...


I'd like to take a crack at your question and then Ross and Jonathan can tell me if I'm right.

For a 5 hour run, Randy goes at a much slower pace than he would for a 2 hour run. Thus, his sweat rate is greatly reduced. So if Randy relies on his thirst, he will drink less per hour than during his shorter run.

Over LOTS of time, a runner who replaces 50% of sweat loss with pure water will eventually lose too much salt. But the time needed to get in danger is much, much longer than 5 hours.

Am I right?


Anonymous said...

All your calculations are wrong - way off - because you calculate serum sodium changes based on ECF, not TBW. Yes, sodium is mainly in ECF, but it "behaves" as if in total body water (TBW). Why? Because "water chases osmoles." For sake of time, to pick one of your more egregious claims, when Randy drinks 2 L water to match his sweat loss, his serum sodium ends up a "normal" 135.5, not the "dangerously low" 128.5 you claim. See Montain or Weschler to get you math straight; then try to get your head straight on salt and water. Right now, you're in left field.

Ross Tucker and Jonathan Dugas said...

Hello everyone

Thank you so much for the wonderful debate about this issue - we really appreciate all your comments. Unfortunately, because of lack of time, we can't respond directly to all of them but we do promise that over time, we'll do our best to get the answers out.

We particularly enjoyed the last one about the fact that we were way out in left field! We really do enjoy the debate, but unfortunately, there is always one guy who is not mature enough to have a sensible debate and has to throw out accusations like "get your head on straight". What a shame...

Anyway, to respond to that one - we know that the TBW is different. But unluckily for Gatorade advocates (of which I'm sure you're one), the argument becomes even more obvious when you use TBW, according to the latest theories. We were using this example to illustrate a key point or two - perhaps if you read the post properly you will pick up that no fewer than three times, we mentioned that the physiology is far more advanced than the mathematical assumptions we have made - this is effectively what you are referring to - fluid shifts in the body will correct the initial drop that is 'theoretically' caused by the ingestion of fluid and salt.

So yes, I hear your point, and yes, we acknowledged that point on numerous occasions when we referred to the 'homeostatic system' in our post. What we will do is request our friend Lulu Welscher (whose work I'm sure you are familiar with - it suggests that you DON'T need salt and that you can overdrink) to comment and submit a post.

But just to repeat, our intention from this post was to evaluate the two arguments we stated very clearly upfront. It was not to put forward the complex mathematical model for body water regulation. I thought that this was emphasized very clearly in the post - apparently I was wrong.

It also seems to that some are very convinced that salt tablets are necessary and that is fine. We disagree, we believe that the body is quite capable of handling exercise without it, just as the body used to do for many years before "clever" scientists came along and told us we were too dumb to know any better when it came to fluid and salt. It does not seem that we will win the argument in this forum, but hopefully we have stimulated enough debate to get some people thinking.

One last response to Ryan, thanks for all the subsequent questions - again, apologies for not replying to each one, but unfortunately my other deadlines could cost me a job! But just to say, Stan is quite correct in his answer - the sweat rate will also go down, so the translation of this model to longer distances doesn't happen quite so simply.

Then also, just bear in mind that the body is a homeostatic system - it's not a jug of water with some spoons of salt in it. And in that case, it's not as simple as just saying "if you extend to 17 hours". The body is clever enough to know that you'll be exercising for that long, and I can guarantee you that it would simply lose less salt in the sweat and kick in all the other mechanisms to conserve salt that are required. And if for any reason your salt levels did drop, then you would also crave salt.

I simply cannot stress enough, that all you need to do is listen to your body and you will be fine - you don't need to go running with a fluid belt and a dozen salt tablets - 20 years ago, runners would have laughed at us!

Thanks again!

Ross Tucker and Jonathan Dugas said...

One last thing I would like to refer people to is our series on Fluid Intake and Dehydration.

This whole series on Muscle Cramps was actually a follow on from that series, where we first raised some of the issues. I guess one drawback to running these "series" is that each article on its own needs to be considered in context. So the Fluid Intake series may actually resolve a number the questions raised most recently. Check it out here:


Alternatively, you can find the links to article in the far right hand side column of the page!

And join us in left field...!

Anonymous said...

Hi Guy's

The layman here again! Using your analogy of the chess board again, and given that electrolytes have nothing to do with cramping, is there anything else (other than muscle fatigue) that could result in cramping? I'm thinking of possibly something like acids (maybe uric)etc.
One last question........could a drop in carbohydrate stores (not sufficient feeding ; ie. bonking) lead to cramping?


Nancy Toby said...
This comment has been removed by the author.
Nancy Toby said...

In a marathon runner, the sweat rate would only be severely curtailed over the duration of the marathon only in the case of heat stroke when sweating ceases.

Sweat concentration of sodium would continue roughly proportional to the sodium concentration in the blood while sodium losses continue during the event and hyponatremia ensues.

Almond et al. (2005) found that 13 percent of Boston Marathon finishers had hyponatremia, a proportion that is reflected in other studies of marathoners as well.

I'm sorry, but your statements are not supported for marathon runners nor other endurance athletes who go even longer than the median 4 to 5 hours of marathon runners, even at an elite pre-selected event with qualification standards such as the Boston Marathon.

People *DIE* from hyponatremia. It's simply not responsible to tell people that "you will be fine" without substantial sodium replenishment during endurance exercise lasting over 2 hours.

Anonymous said...

Hey Dudes,
Another excellent post! I'm currently a PhD student focusing on neurophysiology, but you're making me want to switch to exercise physiology.
I absolutely agree with your advice. The body will largely take care of itself. As the eminent doctor and writer Lewis Thomas noted:
"The great secret of doctors, known only to their wives, but still hidden from the public, is that most things get better by themselves; most things, in fact, are better in the morning."

Keep up the good work!

Ryan said...

Hi Again, I am enjoying this debate, its a very good subject. I am not an advocate for Gatorade or salt tablets, Ive never used salt tablets before and I hate Gatorade. Im just trying to point out to you guys that you are presenting a simplified model, running some numbers through it and then drawing a conclusion that isn't shown by those numbers. You are trying to be advocates for a scientific approach but aren't doing so with this subject in my opinion. You are very biased towards drinking to thirst and not worrying about replacing lost electrolyte. I believe that this is true for events <2 hours, but for >2 hours it is a different story.

Ross Tucker and Jonathan Dugas said...

hi William

Thanks a lot - that's a great quote! You'll make a great doctor, I'm sure!

We're extremely flattered you would consider a career change based on this, but don't be too hasty! The field of neurophysiology needs guys like you with an integrated view - the field is moving that way anyway, so you'll be a powerful force if you can take elements from other areas of science. Same goes for exercise science, actually - we're getting to the point now where the field will be advanced forward by specialists in other fields, who apply the expertise to specific problems - who knows? Maybe it'll be you!

Good luck with the studies, and keep reading!


Ross Tucker and Jonathan Dugas said...

Hey Ryan

No problem, debate welcomed! We hear your concerns - I wish we could do this as a podcast because I do feel a little like we're talking passed one another! We recognize the limitations of this format (the blog) as opposed to a scientific journal, and to be honest, we're not out here trying to win the scientific arguments - we both have day jobs where we're supposed to do it!

So the arguments have been had, and ideas bounced back and forth in many scientific journals. Our intention is to provide a view of sports science that makes the sciences accessible, interesting and news-relevant (so we're primarily a news site, but news has been slow lately!).

However, we really value our readership who recommend the scientific articles, so here is a list of papers we suggest you read - they deal with the issues we've discussed. We do admit to providing a basic overview of the topic - but then our 'market' is not a PhD thesis-reading one!

Check out these articles:

Noakes et al. (2005). Three independent biological mechanisms cause exercise-associated hyponatremia: Evidence from 2,135 weighed competitive athletic performances. PNAS vol 102, 18550 - 18555

Hew-Butler et al. (2006) Updated Fluid Recommendation: Position Statement From the International Marathon Medical Directors Association (IMMDA). Clin J Sports Med vol 16, 283 - 292

Dugas et al. (2005) Hyponatraemic encephalopathy despite a modest rate of fluid intake during a 109 km cycle race. Br J Sports Med vol 39, epub ahead of print

Hew-Butler et al (2006). Sodium supplementation is not required to maintain serum sodium concentrations during an Ironman
triathlon. Br J Sports Med, vol 40, 255 - 259.

Schwellnus et al. (2004) Serum electrolyte concentrations and hydration status are not associated with exercise associated muscle cramping (EAMC) in distance runners. Br J Sports Med, vol 38, 488-492

Noakes & Speedy (2006) Case proven: exercise associated hyponatraemia is due to overdrinking. So why did it take 20 years before the original evidence was accepted? Br J Sports Med, vol 40, 567 - 572

Noakes et al (2004). The dipsomania of great distance: water intoxication in an Ironman triathlete. Br J Sports Med, vol 38, 16 - 20.

These are our "evidence". We try as much as possible, for 'strategic' reasons to avoid getting drawn into the heavy scientific discussions - as I said, that is something we do from 8 to 5, and the purpose of the blog is to stimulate thought, debate and discussion. So I guess mission accomplished...

Hey, if you can't access any of these and they seem really intriguing, drop as an email with your email address and we'll forward them onto you!


Nancy Toby said...

Thank you for posting those references. I'm still looking for *ANY* studies whatsoever which examine adequate replenishment of sodium (50% or more of estimated losses) during long-duration (4 hours plus) endurance exercise. This aspect seems to be completely lacking in the scientific literature, including the references you posted. Only one (Hew-Butler's 2006 study) looked at sodium replenishment and had some serious flaws - primarily in not measuring sodium intake and providing extremely inadequate sodium replenishment (probably 1/4 or less of losses).

1. Noakes 2004 - "dipsomania" - case study of one athlete in placebo group of sodium study developing clinical hyponatremia - provided 4 grams of sodium post-race for correction
2. Noakes et al. 2005 - no data on sodium intake during exercise
3. Butler et. al. 2006 - recommendations do not include any analysis of adequate sodium replenishment during exercise bouts
4. Dugas and Noakes 2005 - case study of one cyclist who developed hyponatremia - no examination of sodium intake during event
5. Hew-Butler 2006 - Sodium supplementation study (finally!): provided only a mean of 3.6 grams sodium for average exercise duration of nearly 13 hours (estimated sodium losses in sweat probably exceeding 13 grams of sodium). No measurement of electrolyte ingestion from all sources during exercise. Even so, sodium-supplemented group had higher post-race sodium serum levels than placebo.
6 Schwellnus et al, 2004 - Muscle cramping in marathon runners - no measurement of sodium intake during exercise, no electrolyte replenishment protocol. Serum sodium levels significantly (P<.05) lower in cramping group.
7. Noakes and Speedy review/essay, 2006 - no new data, no serious examination of adequate sodium replenishment

Ryan said...

Thanks for the thorough rebuttal Nancy, I didn't want to have to read through all those to form my own response.

Ross and Jonathan have shown their bias on this subject and no matter how much we dispute their hypothesis, they are going to continue to rebut with the same answer.

Anonymous said...

Hmmm, sounds more like Nancy can't let go of her indoctrination than anything else. She's insisting on data but not providing anything substantive to back up her contentions either. For example she cites Almond but that's more of an example of excess fluid consumption than anything else.

Nancy Toby said...

I'd love to see ANY well-designed studies which actually examine the question of adequate sodium replenishment during exercise. I've yet to see any. Unfortunately like many of these flawed studies Almond et al. (2005) did not measure sodium intake, sports drink intake, or water intake during the marathon, but rather relied on retrospective reports on fluid and sports drink intake after the finish line (and made no mention whatsoever of the possibility of oral sodium replenishment).

It's difficult to draw accurate conclusions when the independent variables (water and sodium intake) aren't even measured.

Anonymous said...

Thanks for bringing the science to the masses. I am a long time advocate of no-salt, and destroying that salt-causes -cramps myth. And you know Nancy et al, I don'tadd salt to my food, eat non-processed food etc so all the salt comes from natural sources, yet my sodium conc is normal at rest, I have run many many marathons and Ultras over 30 yrs, I never cramp --slight cramping for first time at an ultra (56km) early this year where I was unable to run at all for 3 wks prior (fits the fatigue/conditioning theory), and I never end a long race (note marathon raced still at sub 3hr)where I feel dehydrated or discomfort after moderate drinking.

One question to teh boys from my alma mater. Liberated metabolic water would certainly bolster the argument for less drinking. How does that affect the final concentration though?

Ross Tucker and Jonathan Dugas said...

Hi Everyone, and we have just been bowled over by the debate on this post and series. It is really fantastic and we are thrilled to see so much reader participation.

We just wish we had more time to answer each and every comment!

In response to Russel's comment up there on 28 November. . .

In many ways we do not know what causes a cramp. The available evidence suggests that crampers do not have different electrolyte concentrations and are not more or less dehydrated. I stated in Part II of this series that until someone develops an experimental protocol that can reproduce cramps during endurance exercise, there will be no controlled studies that examine the effects of fluid balance, substrate availability, temperature, fill-in-your-variable of-choice-here, etc.

Upon cursory examination on Pubmed, endurance exercise does not appear to cause uric acid concentrations to rise, but rather it is very high intensity (120% VO2max) exercise that has an effect on this variable.

Re bonking, hypoglycemia normally is occurs with fatigue, or fatigue precedes hypoglycemia. In any case, these two states normally occur together or in short course of each other.

Therefore any cramping that occurs in the presence of hypoglycemia could be merely an associative relationship and a causal relationship might not exist. However, I must say again that until someone develops a protocol that can reproduce cramping in a reliable and predictable manner, we cannot answer that question with much authority.

Thanks for joining the discussion and great questions!

Kind Regards,

Ross Tucker and Jonathan Dugas said...

Hi Oliver, and thanks for chiming in here with the question about "liberated metabolic water."

While that might sound like something from a WWII documentary, what Oliver refers to is a concept that for each gram of glycogen stored in muscle, three grams of water are stored with it. Therefore when we break down that glycogen this water is thought to be released. However to date no one we know of has measured accurately the fate of this water.

The only paper that has really attempted to analyze this was published in 1982:

Sherman WM, Plyley MJ, Sharp RL, Van Handel PJ, McAllister RM, Fink WJ, Costill DL.

"Muscle glycogen storage and its relationship with water."

Int J Sports Med. 1982, Feb; 3(1):22-4

Regardless, the presence of this "liberated" water might help explain why we lose some weight during exercise yet appear to be in fluid balance.

During marathon one might oxidize 500 g of carbohydrate. If we were to take our weight losses (say, 2.25 kg or 3% of a 75 kg runner) and subtract the oxidized fat and carbohydrate, the person (drinking to thirst, of course!) would have an "actual" fluid (weight) loss of 1.75 kg:

2.25 kg - 0.5 kg of CHO = 1.75 kg

And this is what the body is happy with, and if drinking to thirst then in all likelihood the plasma osmolality would be maintained. Therefore the body is happy with that situation.

So the question is how can the body be happy if it has apparently lost 1.75 L (kg) of water? Well the answer might be with the liberated water.

[Disclaimer: this is an assumption and we are working with gross estimates here]

If we have access to the three g stored with every one gram of CHO, then suddenly we have 1500 g of water that become available to us during exercise:

500 g CHO X 3 g H2O per g = 1500

And that leaves us pretty close to fluid balance even though we have lost some weight. In fact it would mean that although we have lost a bit of weight we are only 0.25 kg (250 g) down.

Note: this is a theoretical statement and not backed up by any available evidence. It is based on assumptions, speculation, and my insight

So I am not saying that this is the reason why we are in fluid balance in spite of losing some weight. I am just trying to elaborate on Oliver's question about this concept, and speculate on how it might contribute to fluid balance during endurance exercise.

In fact, as I have said no one has yet to actually measure this water and where it goes when we break down glycogen.

Thanks again, Oliver, for the relevant question, and for joining the discussion here!

Kind Regards,

Anonymous said...

Ryan and Nancy,

I'm surprised, but you seem determined that at some point salt replenishment is necessary.

Although it may deserve attention, and merit specific studies, I can't help but feel you are arguing a secondary point.

The point of the example of Randy, an unusually salty sweater, was to show with numbers, why drinking Gatorade will not replace electrolytes as is claimed, but will achieve the opposite effect. On the surface it seems counter-intuitive, so the example just illustrates with real numbers, what's going on, because a lot of people weren't getting it.

But let's say, for the sake of argument, that there is a threshold where salt replacement becomes necessary. One way to view it is that, by drinking to thirst, rather than drinking to schedule or weight loss, this threshold will be delayed, if it is ever reached at all.


Anonymous said...

Hey everyone

I agree, Nancy, you and Ryan are determined to bang the salt drum. And like Ray says, and someone else, you're yet to provide any evidence for your own argument.

Instead, you prefer to shoot down what has been quite a thorough overview of the science on the topic, and even that, you're doing badly.

As for the pre-occupation on the Randy example, I agree with Ray above, who has pointed out to you that this is a hypothetical example, designed to demonstrate a point. i love this blog, it's great because it brings the science to everyone in an entertaining and relevant form. Yet I'm discerning enough to realise that I'm not reading a scientific journal - it's a website, and it gives me my starting point, and I've often followed up these great posts by reading through the scientific articles that they mention.

Yet you guys are not doing that - Nancy will claim to, but I am highly sceptical of her ability to read the literature - she has to date demonstrated little appreciation for what those studies are proving - she quotes Almond and the high rate of hyponatremia at the Boston Marathon as a reason to take salt. In fact, that study is precisely the reason you should NOT DRINK TOO MUCH WATER.

If I didn't know any better, I'd guess her and Ryan had shares in an electrolyte replacement company! Of course, I say that in jest, but really, this is a very one-sided debate, because you have these two guys on one side doing their best to translate the science (and of course they have bias, that's why I read it!), against people who clearly adopt the antagonistic position without any evidence at all. It's like watching a third grader teasing the high school valedictorian!

Ross and Jonathan, you guys keep up the great work, great blog!


Anonymous said...

Very interesting post. I think your results are somewhat skewed by hypothesizing a very salty sweater.

You say: "a fit, well-acclimatized person probably has sodium concentrations of about 20 mM". If you redo the analysis using that fit person you get much more reasonable results, and hardly a discouragement of using sports drinks.

You point out that "Gatorade has a sodium concentration of 18mM". If the Sports drink indeed has a 90% concentration of what is in the fluid sweated out, then we'd still get a reduction in sodium, but it would be marginal. For instance, if our runner (sweating only 20mM concentration) drank sufficient 18mM concentration sports drink to replace all lost fluid, he'd end up with a sodium concentration of 139.77, hardly different from the starting concentration of 140! And on the plus side, he'd avoid dehydration. Clearly hyponatraemia is bad and potentially deadly, but surely dehydration is also bad?

If we follow your analysis, and we wish to avoid dehydration, then the logical approach would be to drink more concentrated sports drink (of the kind you can mix yourself), or take salt tablets when we drink. You could then end up without plasma volume loss and without a changed sodium concentration. What would be wrong with that?

Anonymous said...

How excellently worded! i live in Tucsn AZ, and everyone here is obsessed with sports drinks! i keep saying, "thank goodness they were around 500 years ago, right?"
I strongly believe (having grown up in Texas and lived in numerous hot places) that if you drink only what youre thirsty for, you wont have to worry about electrolyte losses unless youre sweating for an extremely extended period of time (i myself have hiked in the desert for 3 or 4 days straight drinking water ONLY and eating regular foods and have NEVER had fatigue or anything while hiking). i will now show this post to the people that tell me i should be dead (since being alive after living in areas like this isnt enough proof that sports drinks are not very useful).

Anonymous said...

Interesting article, but not very valid during very long periods of exercise. Try running a 100 miler with no electrolytes and you'll find things tend to rapidly go south.

Lack of sodium not only causes cramping but the sour stomach that plagues many ultrarunners after hours and hours on the move.

Jeff said...

This is a great discussion. Looks like I'm 18 months late to the party, but I would definitely be interested in a follow-on.

As others have mentioned, it would be great to see analysis of longer duration events. I have competed (survived) ultrarunning and ironman distance events that have gone 8-13 hours. I'm curious if your model holds.

A few observations: 1) it seems to me that the athlete that drinks to thirst and eats to satisfy salt needs is on a vector to dehydration (assuming 50% replenishment for example). So your premise holds for someone exercising for 3-5 hours but would seem to unravel in a 100 mile ultra run. 2) other studies have suggested that the impulse for thirst, salt and fuel are lagging indicators during a physically-challenging event, and that by the time you're feeling them you're potentially too late to address them adequately without rest.

Finally, to be devil's advocate, there were a number of people that suggest that we don't need salt tablets or sports drinks because we didn't have them 500 years ago. That's a specious argument IMO. Likely those that undertook 50 mile+ runs stopped to rest when they felt like crap, rather than being driven to the artificial deadline that we're faced with today in races (a PR time, cutoffs, etc). I would also point out that 500 years ago we didn't have a lot of things that we do now, such as antibiotics, and that average life expectancy was probably closer to 35-40 than to 70-80.


Ross Tucker and Jonathan Dugas said...

Hello Jeff

Never too late to join an argument.

To answer your points:

1) Yes it would hold. In fact, the whole theory behind the salt supplementation comes from Iron Man studies that we did at the SA ironman. As you'll know, there are guys who finish the race in just under 17 hours, and they're in the groups of people we've studied.

Same thing holds - they finish the race at about 3 to 5% dehydration (measured as change in body weight, which is flawed, but it's what everyone uses as the standard)

They also have very normal sodium concentrations. In fact, if people are dehydrated, the salt concentration will rise, as we discussed in this post. So we have this ultra-marathon here in SA - Comrades, which is 90km long and the average finishing time is about 11 hours. We've done a lot of research there, and it turns out most people in the medical tent have very salt concentrations. THey're hypernatremic. If these same people took in more salt, they might be in very serious trouble, so the issue around salt certainly does apply to ultra exercise

This leads onto your point about "voluntary dehydration", and that if your body allowed you to lose 3 kg in 4 hours, you would lost 9 kg in 12 hours. That doesn't happen, which is itself amazing. What we find is that an average weight loss for a marathon is about 2 to 5 kg (in an 80kg man - it's about 3 to 5% body weight). In a half marathon, it's also between 2% and 5%. And during COmrades, the average weight loss is also 3 to 5%. So amazingly, the body doesn't simply allow more and more weight loss - the amount that you replace to thirst seems to vary depending on the length of exercise. Therefore, you can't extrapolate what happens in 1 hour or 2 hours (which is where this voluntary dehydration theory comes from) to ultra-exercise. The body is too smart for that, and would prevent excessive dehydration.

I must emphasize, however, that all of this is done by regulating the osmolality, NOT the body weight. Therefore, your body's thirst mechanism is super sensitive to your osmolality, and in ultra exercise, it defends that. This is why you'd just drink more if you got dehydrated - it's regulated very well indeed.

2) That brings me to point 2. There is little evidence that thirst 'lags' behind. That's only true if you believe that any level of dehydration is dangerous or aversive. That's what the fluid research has told you, and it's what we are saying is incorrect. Fact is, your body's thirst mechanism is spectacularly sensitive to changes in salt concentration and it stimulates thirst even during exercise. Again, that evidence comes from field studies, I'm not aware of any that show that thirst lags behind during exercise.

3) Finally, I don't buy into the argument about not having salt tablets 500 years ago either, so I'm with you there.

What I will say is that 500 years ago, we didn't have companies telling us to drink, we simply drank as desired (unless of course we got caught without water!) and managed to regulate sodium just fine, letting our body tell us what to do.


P.S. Not sure if you have, but there is a whole series on dehydration, not just this post, where we've explained these concepts more, so you may want to browse for them. They're listed under "Featured series" above.


Anonymous said...

You make the comment in your article: "It is NOT possible, even for a "salty sweater" like Randy in these scenarios, to LOSE sodium through sweat."

I am a salty sweater who is covered by salt when I finish a marathon. It is all over my face and my singlet. If I don't use a headband, the salt burns my eyes. I can also taste the very salty sweat on my upper lip. Does this not disprove your statement?

Also, I participated in the Two Oceans ultra some years ago and was saved by a woman who was dispensing table salt about the 29 mile mark. I had become somewhat dizzy and disoriented at that point and thought that I would have to pull out of the event. I imbibed some of the table salt and was immediately able to continue strongly to the end.

I am willing to be convinced by your article. However, the reality makes me nervous in adopting your strategy.