Athletes: obey your thirst

Tim Noakes has a nice review paper in the Annals of Nutrition & Metabolism on hydration for athletes: Is Drinking to Thirst Optimum? It is open access at least for now.  In short, Noakes makes a convincing case that drinking to ‘stay ahead of thirst’ during exercise is unnecessary and simply drinking according to thirst is best for several reasons.

It has a great history on how it came to be that before 1965, it was commonly accepted that athletes should not consume liquids during competition (even during marathons), yet now are generally told to drink liberally prior to and during competition to prevent dehydration (flashbacks to high school of kids carrying around gallon jugs of water or refilling our Nalgene bottles after every class to “hydrate”- ugh.).  Noakes identifies 4 key factors that changed this thinking:

  1. In 1965, when the sports drink that is now known as Gatorade was created in 1965, a study by the creator (Dr. Robert Cade) demonstrated that the carb solution improved a 7-mile run/walk performance, but did not prevent heat stroke as initially hypothesized.
  2. In 1969, 2 physiologists (Wyndham and Strydom) showed that athletes who were dehydrated by >3% of prerace body weight had body temperatures that were “considered unacceptable.”  Based on this (and that maintaining 0% dehydration maintained body temperature), they suggested marathoners should drink 250 ml of fluid every 15 minutes (1 L per hour).  Noakes notes that previous research had suggested that ~2  liters can be lost through sweat without physiological hindrance, their suggestions should have been halved to about 500 ml per hour- about the amount that Noakes’ research has found that athletes will consume based on thirst (ad libitum).
  3. After 1982, the United States Army Research Institute for Environmental Medicine (USARIEM) developed for the US Military the position that consumption of very large amounts of water in hot environments would give a tactical advantage.  A highly flawed, improperly randomized study was published that concluded heat illness was reduced with greater water intake.  USARIEM scientists are advisors to the bodies such as the American College of Sports Medicine (ACSM) and still influence their current positions on fluid intake which until 2007 included: “as much as tolerable during exercise.”
  4. The sports drink industry- growing partnerships with scientists and industry.  From 1985 to 2003, this industry grew from 217 million to 2.69 billion.

In the rest of the paper, Noakes describes research and concludes these basic tenets:

  • Mild dehydration from ad libitum water consumption is not associated with an increased risk of health consequences.
  • There seems to be a 1-2 liter fluid reserve that does not need immediate replenishment during exercise.
  • No evidence of performance enhancement from very high water intakes.
  • Unlikely that fluids per se will prevent heat illness (unless significant dehydration occurs).
  • Noakes’ isn’t aware of a single case in the last 20 years of serious illness or death from dehydration per se in athletes, but there are at least 12 deathes from exercise-associated hyponatraemia (EAH) and encephalopathy (EAHE) (which Noakes’ & colleagues discovered 20 years ago and is only now becoming more seriously appreciated).  The 2007 ACSM position now finally advices drinking to thirst.
  • Ad libitum consumption maintains physiological hydration markers in the normal range.
  • Athletes who lose the most weight during races are generally faster finishers.
  • Even in races when athletes are given the same drinking advice, change in weight varies significantly.  Noakes writes: “It is clear that the factors determining drinking behavior during competitive exercise are poorly understood. Perhaps we should study this phenomenon more carefully before we produce dogmatic guidelines that conflict with common sense, the scientific evidence and even what athletes  actually do during exercise and competition.”
  • Drink to thirst, and don’t exceed 800 ml per hour.


Noakes, Timothy  (2011). Is Drinking to Thirst Optimum?
Annals of Nutrition & Metabolism : 10.1159/000322697

Hat tip to Steve Magness.

  • KatjaSawitsky

    I recently read something in Science Daily, I believe, that talked about this issue. The gist of the article is that drinking during exercise may be unnecessary given that through the oxidation of carbohydrates, metabolic water is released. Carbs plus oxygen yield CO2 plus water.

  • Pingback: Science and Exercise Roundup()

  • Pingback: Rethinking Hydration? « This Running Life()