What’s the ideal running drink? Noakes weighs in

Posted by Dave on August 23, 2011 | 11 Comments

Lore of Running

930 pages of science!

There’s been a fair amount of discussion of hydration and dehydration on Science-Based Running, but it’s been focused on what to do to prevent serious problems, rather than how to optimize running performance. Let’s assume you’re not going to keel over from heat exhaustion or overhydration in a race. What should you consume before and during the race to maximize your performance?

This summer I’ve been reading Tim Noakes’ encyclopedic volume, Lore of Running and reporting what I’ve found back to you. This week I’m reading his chapter on temperature regulation during exercise. Noakes seems quite convinced that few runners are in danger of dehydration or heat exhaustion during races a half-marathon or longer. Indeed, they are in greater danger of overhydration.

For much of last century, runners were encouraged not to drink at all during races, and the rules often prohibited it. I still occasionally run with people who remember their coaches telling them not to drink during practice because that’s a sign of weakness. Modern elite athletes actually don’t appear to drink much during races either — if hydration is so important, why doesn’t Ryan White carry a CamelBak when he races?

The answer, according to Noakes, is that there is little evidence that runners should force themselves to drink if they’re not thirsty. That said, there’s plenty of evidence that some drinking is very important. The question is, how much, and what should you drink?

One piece of advice I’ve heard given to marathoners is that you should pre-hydrate as much as possible before a race: According to this line of reasoning, the more you pee, the better you’ll do, and make sure you’ve had enough that your urine stream is completely clear.

Noakes disagrees, saying that athletes should have their last drink two hours before race time, and even at that point, only drink enough so that their urine is lightly colored. For races longer than 10K, they should then carry about a half-liter of fluid to the starting line and drink it just before the gun. This is because water takes about two hours to get fully absorbed into your system. Any excess consumed in the meantime will simply be urinated out, along with sodium and other electrolytes your body needs.

There are limits to how much water, electrolytes, and carbohydrates your body can absorb during a race — and the harder you run, the slower your body processes additional fluids. In general, if you are racing hard, you will not be able to replenish fluids and nutrients as quickly as your body burns and sweats them away. Finally, what’s in your drink also affects how quickly it can be absorbed by your body. Take a careful look at this set of graphs:

[Updated based on Timli’s comments below] There’s a lot of information here, but basically it’s showing two scenarios depending on how much liquid is in your stomach. Some runners are uncomfortable running with more than a few cups of liquid in their stomachs, while others can handle larger volumes. So the graphs at the top of the image show a runner who maxes out at a volume of 400 ml, while the graphs at the bottom show a runner who can handle 800 ml of liquid, nearly a full liter. Both runners drink to refill their stomach to maximum every 10 minutes. What the graph shows is how quickly the liquid drunk is delivered to the body: Depending on how much carbohydrate is in the liquid, the body can process a different volume of liquid.

Let’s focus in on runner A. She starts by filling up with 400 ml of liquid. If it’s just water, she can absorb 250 ml every ten minutes, and so she needs to drink another 250 ml to refill at that time. If she drinks a 7% glucose solution, her body processes less liquid, and if she drinks an 18% glucose solution, she can process even less, and can only drink 100 ml every 10 minutes to replenish.

Not take a look at the graph at top-right. This indicates how much water and carbs are delivered to the body in each scenario. The three columns to the left of the graph indicate how much water is delivered to the body. If she’s drinking pure water, her body gets 1.6 liters per minute hour. But if she drinks an 18% carb (CHO) solution, the volume of water she gets decreases to just 0.6 liters. Now look at the two columns on the right. They show how much carbs her body absorbs. Obviously if she’s drinking pure water she’s getting no carbs. And as you can see, drinking an 18% carb solution doesn’t actually deliver much more in the way of carbs than drinking a 7% solution. In fact, her body can only deliver about 1 gram of carb per minute to the muscles, so there’s no point in drinking more than 60 grams of carbs per hour.

Runner B gets more of everything, but has to run with a much fuller stomach: Most runners would be very uncomfortable with 800 ml of liquid sloshing around in their stomachs. In practice, drinking as much liquid as in either of these examples is rarely done. Few runners ever drink more than about 800 ml per hour, and runner B would be drinking more like 3 liters per hour!

Noakes can find no evidence that drinking even 800 ml actually enhances performance. Marathon runners are a little dehydrated after races, but in fact the winners tend to be more dehydrated than other racers. He suggests drinking 500 to 800 ml per hour, whatever is comfortable for the runner. Slower runners, especially when running in cool temperatures, may need even less than this.

He also suggests a drink that contains between 7 and 12 percent carbohydrates, ideally a mixture of various types but not fructose. The sodium content should be around 60 mmol per liter.

As of 2001 when this book was published, Noakes knew of no energy drink on the market meeting his specifications. But since then there has been an explosion in the energy drink market. Do any of today’s drinks hold up?

It’s actually a little hard to say, because energy drinks don’t use the same measurements Noakes does. So what I’ll do is make a little tour of a few drinks on the market today and give it my best shot (based on chemistry class 20 years ago) of converting the units appropriately. Please let me know if I’ve made a mistake here and I’ll fix it as quickly as possible:

60 mmol per liter works out to 1380 milligrams of sodium (based on sodium’s atomic weight of 23).
7 percent carbs per liter works out to 70 grams (by weight), and 12 percent is 120 grams.

Gatorade G (original):
110 mg sodium per 8 ounces = 466 mg per liter.
14 grams carbs per 8 ounces = 60 grams per liter.
Definitely a little low on carbs and very low on sodium

Gatorade G (pro):
200 mg sodium per 8 ounces = 847 mg per liter.
14 grams carbs per 8 ounces = 60 grams per liter.
Only about 62% of the sodium Noakes recommends and a little low on carbs

Powerade (original):
100 mg sodium per 8 ounces = 423 mg per liter.
14 grams carbs per 8 ounces = 60 grams per liter.
definitely a little low on carbs and very low on sodium

Powerade (with Boost supplement):
1250 mg sodium per 20 ounces = 2150 mg per liter.
41 grams carbs per 20 ounces = 70 grams per liter.
This is actually more sodium than Noakes suggests, and right at the low end of recommended carbs. However, 19 of those 41 grams of carbs are fructose, which Noakes suggests avoiding

360 mg sodium per 8 ounces = 782 mg per liter.
No carbs
Only about 57% of the sodium Noakes recommends. No carbs (although you could supplement with gel packs)

So it looks like the Powerade with Boost comes pretty close to Noakes’ recommendations, as long as you’re not worried about consuming too much sodium. Sodium slows delivery of water, so you need to drink more fluids when you are consuming more sodium; if you can’t handle that, then this product probably wouldn’t be for you.

With any of these products, however, you could also always mix them in a higher concentration than the package recommends. For example, you could use double the strength of Gatorade G (pro) to get about 1700 mg of sodium and 120 grams of carbs per liter. Or you could dial it back just a bit and get quite close. The question is, would you want to drink such a strong concoction? I leave it up to you to decide!


11 Responses to “What’s the ideal running drink? Noakes weighs in”

  1. Sean
    August 24th, 2011 @ 12:41 am

    I recall reading in a wilderness medicine text that drinking electrolyte drinks that are overly concentrated could cause diarrhea. I can dig the reference up if needed.

  2. Timiji
    August 24th, 2011 @ 3:53 pm

    Hi Dave,

    Are the graphs right out of “The Lore of Running”? I’m a little confused by your explanation. Based on the symbols, the delta gastric volume is highest for the 18% Glucose polymer (at least in the 800ml graph, bottom left). This is in direct contradiction to the bottom right figure, showing the rate of water delivery as highest for plain water.

    What’s up with that? It could be an error in symbols (not too uncommon in graphs), or I could be confused (also not uncommon!).

    Coyle’s early work on hydration/CHO delivery is also great background on the ‘argument’ between optimizing hydration or CHO delivery.

  3. Dave
    August 24th, 2011 @ 4:04 pm


    The reason is that the CHO solutions are not just water. So an 18% CHO solution delivers more total stuff, but less water. Make sense? In other words, the CHO in the solution isn’t delivered at the same rate as the water in the solution (and these rates depend on the proportion of CHO in the solution and the total volume of fluid in the stomach).

  4. Timiji
    August 24th, 2011 @ 7:39 pm

    Thanks, Dave,

    My curiosity about the source of the graphs remains… are they from ‘Lore’… if so, I will look it up myself.

    I’ve done a bit of basic math to illustrate my (continued) confusion. The lower left and lower right figures don’t measure up to my math at all.

    If indeed the symbols line up with the correct corresponding lines, the change in gastric volume per 10 minutes (800-x) is 200ml for water, 400ml for 7% glucose polymer (CHO) and 500ml for 18%CHO. I agree that the CHO fluids aren’t all water… for math’s sake I’ve assumed (perhaps wrongly) that they can be divided into CHO and water… 7%/93% and 18%/82* for the CHO solutions, water is the 100% reference. Still the math doesn’t add up, nor does it correspond to the water delivery on the right side, even if the symbols are in reverse order as what was intended:

    ∆/10min dilution ∆water calc l/h (x6)
    200 Water 200 12000
    400 7%CHO 372 22320
    500 18%CHO 410 24600

  5. Dave
    August 24th, 2011 @ 8:31 pm

    Ah, now I think your initial comment is right — the graph is labeled backwards, based on the explanation in the text. I have now updated the post and corrected the label on the graph. (And yes, the graph comes from the book)

  6. Stu N
    August 27th, 2011 @ 10:42 am

    “If she’s drinking pure water, her body gets 1.6 liters per minute.”

    – Impressive! Typo, I presume?

  7. Dave
    August 27th, 2011 @ 5:35 pm

    Good catch, Stu! It’s fixed now.

  8. Gerard Pinzone
    February 6th, 2012 @ 9:02 pm

    You guys ought to read what Professor Noakes has to say about his book and carbohydrates now. He recommends you tear out the pages about nutrition in his book and has endorsed Gary Taubes’ view. He’s low carb and loving it.

  9. Richard
    February 12th, 2012 @ 4:59 pm

    Yes, I read the same article. Very interesting.

  10. Bob
    July 21st, 2012 @ 5:21 pm

    Dave, you make a major error in recommending the Powerade with boost. Read the nutrition label on the bottle, the second ingredient after water is high fructose corn syrup. Noakes, and the American College of Sports Medicine, recommend avoiding fructose drinks due to the high incident of gastro intestinal distress that fructose while exercising causes.

    An easy way to waste 6 months of training is to ingest a fructose drink during a race, gastro intestinal problems highly likely to ruin ones performance if you do.

  11. Dan
    July 29th, 2012 @ 10:38 pm


    I’ve read numerous times (going back to the 1980s) that one should never try something for the first time in a race. Still good advice.