Dehydration and endurance performance in competitive athletes


  • Eric DB Goulet

    Corresponding author
    1. Research Centre on Aging and Faculty of Physical Education and Sports, University of Sherbrooke, Sherbrooke, Québec, Canada
    Search for more papers by this author

EDB Goulet, Research Centre on Aging and Faculty of Physical Education and Sports, University of Sherbrooke, 1036 Belvédère Sud, Sherbrooke, Québec, Canada, J1H 4C4. E-mail: Phone: +1-819-780-2220, ext. 45226. Fax: +1-819-829-7141.


The field of research examining the link between dehydration and endurance performance is at the dawn of a new era. This article reviews the latest findings describing the relationship between exercise-induced dehydration and endurance performance and provides the knowledge necessary for competitive, endurance-trained athletes to develop a winning hydration strategy. Acute, pre-exercise body weight loss at or above 3% may decrease subsequent endurance performance. Therefore, endurance athletes should strive to start exercise well hydrated, which can be achieved by keeping thirst sensation low and urine color pale and drinking approximately 5–10 mL/kg body weight of water 2 h before exercise. During exercise lasting 1 h or less, dehydration does not decrease endurance performance, but athletes are encouraged to mouth-rinse with sports drinks. During exercise lasting longer than1 h, in which fluid is readily available, drinking according to the dictates of thirst maximizes endurance performance. In athletes whose thirst sensation is untrustworthy or when external factors such as psychological stress or repeated food intake may blunt thirst sensation, it is recommended to program fluid intake to maintain exercise-induced body weight loss around 2% to 3%.


The field of hydration and endurance performance research is going through a captivating but destabilizing time. In fact, recently published laboratory-, field-, and meta-analytic-type studies have questioned knowledge that was previously thought to be settled and set in stone.1–3 For instance, it has been believed for a while that exercise-induced dehydration that exceeds 2% body weight impairs endurance performance under almost all endurance-related exercise circumstances4,5 and that athletes should drink before they become thirsty to maximize endurance performance.6 Goulet2 recently demonstrated that exercise-induced body weight loss of up to 4% does not decrease cycling time-trial performance and that drinking to the dictates of thirst maximizes endurance performance. Moreover, Zouhal et al.3 published eye-opening results demonstrating that marathon performance times of 643 runners were inversely correlated to their end-of-exercise body weight loss. An update of this literature is, thus, necessary. This succinct review reports the latest findings regarding the impact of exercise-induced body weight loss on endurance performance and provides some basic guidelines enabling athletes to get the most out of their hydration needs. Throughout this article, exercise-induced body weight loss is taken as a representation of exercise-induced dehydration. The guidelines provided in this article are primarily aimed towards competitive endurance-trained athletes, but there is no reason to believe that they should not also extend to competitive, recreationally trained athletes.


Over the past 15 years, many scientific papers have extensively reviewed the effects of exercise-induced body weight loss on endurance performance and physiological functions.4,5 The general and common message conveyed by those papers is that exercise-induced body weight loss significantly impairs endurance performance. For instance, from 1996 to 2006, the American College of Sports Medicine's Position Stand on Exercise and Fluid Replacement recommended that “during exercise athletes should consume fluids at a rate sufficient to replace the water lost through sweating or consume the maximum amount that can be tolerated.”7 In their 2007 update of this position stand, the American College of Sports Medicine slightly altered their message and this time proposed that “dehydration >2% of body weight degrades aerobic exercise performance in temperate-warm-hot environments and that greater levels of dehydration will further degrade aerobic exercise performance.”5 It is, thus, a well-accepted concept, at least in North America, that endurance athletes should strive to prevent a body weight loss greater than 2% during exercise, otherwise, endurance performance will suffer.


The conclusion that exercise-induced body weight loss decreases endurance performance was mainly based on results of laboratory-based studies in which endurance performance was tested using either fixed-power output tests to exhaustion or short, high-intensity, time-trial type exercises preceded by exercises conducted at fixed-work rates.2 These testing methods have an inherently low ecological and external validity8 because under most training and racing situations affected by exercise-induced body weight loss, athletes adopt pacing strategies in which speed varies throughout either on a macro- or a micro-scale.9–12 In addition, compared with time-trial type exercise protocols, fixed-power output tests to exhaustion have been demonstrated to have poor reliability13 and an unclear sensitivity.14,15 Because fluid intake guidelines are primarily designed for, of interest to, and used by competitive athletes, it is questionable that results deriving from such research designs were used for the formulation of fluid intake guidelines.

Five laboratory-based studies have examined the effect of exercise-induced body weight loss upon time-trial type exercise performance, which emulates real-world exercise conditions.1,16–19 Of great importance, and with results completely opposing the well-accepted dogma, none of these studies demonstrate that exercise-induced body weight loss impairs endurance performance. Based on the results of these five studies, Goulet recently published a meta-analysis demonstrating that during cycling time-trial type exercise, exercise-induced body weight loss increased, albeit non-significantly, endurance performance (power output) by 0.06% compared with a well-hydrated state.2 Mean end-of-exercise body weight loss was 0.44% for the well-hydration condition compared with 2.20% for the dehydration condition (range, 1–4.3% of body weight). Moreover, it was demonstrated that there was no significant difference in the percentage change in power output between studies with an end-of-exercise total body weight loss of ≤2% or >2%.

However, similarly to the group of studies used to establish the 2% body weight loss concept, some of the studies that utilized time-trial exercise protocols to test the efficacy of exercise-induced body weight loss upon endurance performance contain limitations that may have confounded their findings. Hence, further studies are necessary before a more definitive conclusion can be reached about the effect of exercise-induced body weight loss under real-world exercise circumstances. Nevertheless, such results, albeit deriving from laboratory settings, lead one to believe that under out-of-door exercise conditions the effect of exercise-induced body weight loss may not be as great on endurance performance as previously thought by scientists.


Proof is rapidly accumulating to suggest that under real-world racing conditions, such as marathon, ultra-marathon, and ultra-triathlon races, exercise-induced body weight loss >2% does not further hinder endurance performance compared with exercise-induced body weight loss maintained at ≤2%. To the contrary, several field studies have demonstrated an inverse relationship between exercise-induced body weight loss and endurance performance, showing that the higher the exercise-induced body weight loss, the better the endurance performance.

In marathon runners, Zouhal et al.3 showed a significant linear relationship between the degree of body weight loss and race finish time, such that those with the greatest body weight loss had the best racing times. Similarly, Sharwood et al.20 observed a significant relationship between Ironman-triathlon race finishing time and body weight loss, such that athletes who finished the race with the highest body weight loss were also the fastest. Kao et al.21 examined the relationship between body weight loss and 24-hour ultra-marathon performance in 23 athletes. Again, there was a significant positive relationship between body weight loss and performance, with those having lost the most body weight running the greatest distance. In 1967, Pugh et al.22 reported that the winner (2 h, 38 min) of a national marathon held in England lost 5.23 kg, or in relative terms, 6.9% of his body weight. Moreover, it was reported that the first four athletes to finish that race had lost on average 5.8% of their body weight. More recently, it was demonstrated that the winner of the 2009 Dubai marathon, Haile Gebrselassie, completed the race with a body weight loss of 9.8%.23

The above examples should not necessarily be interpreted to suggest that exercise-induced body weight loss is ergogenic, but rather that the maintenance of body weight ≤2% is assuredly not critical to performance during endurance events.


Although the previous facts indicate that an aggressive defense of body weight during prolonged exercise is clearly not a prerequisite for the achievement of remarkable performance, they should not be taken to suggest that attention should not be paid to hydration during exercise. In fact, it cannot be denied that under particular circumstances, exercise-induced body weight loss impairs endurance performance. However, armed with basic knowledge, it is possible for endurance athletes to avoid potential traps and maximize the benefits of hydration on performance. The final part of this article provides guidelines that will help athletes, coaches, and health practitioners design efficient hydration plans.

Guideline 1: Make sure to be well-hydrated before exercise

Although for many this statement may be self-evident, it has been shown that some athletes enter an exercise session already hypohydrated.24 Research has shown that starting an exercise with a less than optimum body weight impairs subsequent aerobic exercise performance. In a meta-analysis, Gigou et al.25 demonstrated that pre-exercise hypohydration in the range of 2.6% to 5.6% body weight impairs short-duration (5–30 min), high-intensity exercise performance in a practical manner. Moreover, they observed that maximal oxygen consumption decreases by 2.9% for each percent loss in body weight above a threshold loss of 3.1%. Casa et al.26 recently demonstrated that an acute pre-exercise body weight loss of 2.3% significantly decreased time-trial running performance compared with a well-hydrated state. However, inferences based on this study should be limited, since athletes who started the exercise well hydrated were permitted to drink during the time-trial, whereas hypohydrated athletes were not.

Athletes can ensure they are starting exercise in a well-hydrated state by initially paying very close attention to their sensation of thirst, which should be kept as low as possible during the last hours prior to exercise by drinking fluids ad libitum.26 Moreover, in the last 2 h before exercise, athletes should make sure, by drinking whatever quantity of fluid needed (usually about 5–10 mL/kg body weight of water), to produce 2 micturitions that are very pale yellow to pale yellow in color, which is an indication that their body weight is within 1% of their well-hydrated baseline body weight.27 Such a level of hydration before exercise should ensure maintenance of normal plasma osmolality level, hence minimizing thirst at the start of exercise.26

Guideline 2: Drink according to your thirst sensation: no more, no less

Goulet2 demonstrated that cycling time-trial performances are maximized when athletes drink according to the dictates of their thirst. More specifically, it was shown that endurance performance decreased significantly for athletes drinking less and decreased nonsignificantly for athletes drinking more than thirst. These results are in sharp contradiction with the theory stating that during prolonged exercise, it is of critical importance to drink before thirst is experienced; otherwise, it will be too late and endurance performance will have already started to decrease. Moreover, the results suggest that endurance performance is maximized when plasma osmolality, not body weight, is optimally regulated during exercise. In fact, thirst sensation is primarily, but not entirely, regulated by plasma osmolality level and set at a plasma osmolality that is within the accepted normal range for this variable (280–296 mOsmol/kg H2O).28 Hence, repeated replenishment and satisfaction of thirst throughout exercise should preserve extracellular fluid homeostasis and maximize endurance performance. In athletes with high sweat sodium losses in whom osmotic-driven thirst sensation may be blunted, the expected greater relative plasma volume loss with exercise-induced body weight loss should serve as compensatory input to the thirst drive.29

Although thirst sensation is not easy to define, likely because it evolves through a graded continuum, osmotic-induced thirst has been characterized by a dry, sticky, and thick sensation in the mouth, tongue, and pharynx, which quickly vanishes when an adequate volume of fluid is drunk.30 It is possible that some athletes perceive their thirst sensation as being unreliable, potentially compromising adequate fluid replenishment and endurance performance. Alternately, or additionally, factors such as psychological stress or repeated food intake during exercise may blunt thirst sensation. In such situations, it may be wiser for athletes to preprogram their fluid intake during exercise such to maintain body weight loss around 2–3%, although there is no guarantee that this will optimize endurance performance.

Athletes may, at times, face exercise situations in which fluid access is very limited, is impossible, or is not practical, and in which meaningful body weight loss and thirst sensation could develop and hinder endurance performance. In such a scenario, the goal is to delay the onset of thirst for as long as possible after exercise start and this could be achieved through pre-exercise hyperhydration. In fact, Goulet et al.31 demonstrated that glycerol-induced hyperhydration reduces the feeling of thirst under prolonged exercise circumstances when fluid intake is limited. However, in 2010, the use of glycerol was banned by the World Anti-Doping Agency. Nevertheless, Goulet et al. recently observed that sodium-induced hyperhydration (26 mL/kg body weight of a 130 mmol/L sodium solution) produces a fluid retention (approximately 16 mL/kg body weight) and thirst-suppressing effects that are as important as when glycerol is used to induce hyperhydration.32

Guideline 3: Fluid intake should be limited during 1-hour high-intensity exercise

During high-intensity 1-hour exercise duration, research has shown that fluid intake is not important for maximization of endurance performance. In fact, during such exercise conditions, Goulet2 demonstrated that exercise-induced body weight loss improves endurance performance by 0.48% compared with a well-hydrated state, albeit in a nonsignificant manner. Trying to fully replace fluid losses during high-intensity exercise may lead to gastrointestinal problems, thereby degrading rather than improving performance.19,33

Over the past few years, several studies have examined the effect of mouth-rinsing sports drink solutions during 1-hour endurance performance. Rollo and Williams34 recently concluded that the available evidence suggests that mouth-rinsing with carbohydrate solutions during 1-hour exercise enhances endurance performance, especially when subjects have been fasted for 3–4 h before starting exercise. More specifically, athletes seeking to optimize performance are encouraged during 1-hour exercise periods to mouth-rinse approximately every 8–10 min for 5–10 s with approximately 20–25 mL of a 6% sports drink solution.


The present article addresses the latest findings on the effects of exercise-induced body weight loss on endurance performance and provides basic hydration guidelines needed for the design of efficient hydration strategies. However, readers must bear in mind that the depth and breadth of our knowledge in this research field is still limited and that much remains to be learned before reliable recommendations can be confidently made to competitive endurance athletes seeking the best performance. For instance, how exercise-induced body weight loss affects running time-trial performance is unknown. Moreover, we have no clue as to how exercise-induced body weight loss exceeding 4% influences endurance performance, or simply how it impacts endurance performance in women.


Funding and support.  This work was supported, in part, by ILSI North America. Any opinions, conclusions, or recommendations expressed here are those of the author.

Declaration of interest.  The author has no relevant interests to declare.