Article Text

other Versions

To drink or not to drink? Explaining “contradictory findings” in fluid replacement and exercise performance: evidence from a more valid model for real-life competition
  1. Toby Mündel
  1. Correspondence to Toby Mündel, Institute of Food, Nutrition and Human Health, Massey University, Private Bag 11 222, Palmerston North, New Zealand; t.mundel{at}

Statistics from

Research into exercise and hydration is not new. Twenty-five years ago, White and Ford1 reported on the voluntary dehydration (hypohydration) during a competitive cycling road race; subjects' mean body mass losses were greater than 3%, and there were low rates of fluid ingestion despite fluid being freely available. It is notable that there was no relationship between fluid intake and finishing position. The BJSM published one of the first consensus statements on fluid replacement2 during and after exercise.

In this issue of BJSM, Marino and colleagues3 challenge the common belief that (full) fluid replacement is necessary to maintain/improve high-intensity exercise performance in moderate and warm conditions. Their data show neuromuscular adjustments according to hydration status allowing the attainment of similar peak and terminal core body temperatures and heart rates, and performance times. Therefore, subjects performed equally well during 60 min of cycling at 20°C when fully hydrated, as at 33°C without any hydration (with body mass losses of 2.1%). Interestingly, this is not the first study to demonstrate this occurrence. Robinson et al4 utilised a similar 60 min cycling protocol performed at 20°C and reported no benefit of replacing fluid losses during exercise when compared to complete fluid restriction (2.2% body mass loss); in fact, when receiving fluid, subjects actually performed worse!

Why does the sports medicine community still encourage drinking to improve performance?

So, why does the sports medicine and exercise science community advocate that “during prolonged exercise the ingestion of appropriate fluids will improve performance”2 and “dehydration >2% body weight degrades aerobic exercise”5, when fluid restriction3 4 or ad libitum ingestion1 6 has been shown to exert no deleterious performance effect?

Substantial literature demonstrates that prolonged exercise can cause considerable body water deficit (dehydration), which adversely affects blood (plasma) volume and therefore cardiac stroke volume and output, osmolality, mean arterial pressure, body temperature(s), perception of effort and exercise duration; fluid replacement can attenuate these effects.8 Heat stress usually accelerates this exercise-induced dehydration and therefore exacerbates the forementioned consequences. Most, if not all, of these observations were made employing fixed-intensity exercise protocols—an experimental model that controls numerous confounding factors. Ideally then, the investigators can confidently conclude that independent variable x (eg, fluid intake) exerts an effect on dependent variable y (eg, exercise performance).

Real competition—“as fast as possible” not “as long as possible”

While this approach is reasonable, and one on which much of our understanding of physiological responses to exercise has been built, there is one area where this model is falling out of favour—exercise “performance”. Traditionally, exercising to exhaustion, that is, volitional fatigue, has been the performance measure of choice, yet the ecological validity of such a test is low as most “real-life” sporting situations require us to complete a set distance as quickly as possible rather than for as long as possible. Importantly, exercise is usually self-paced—the athlete can adjust “pace” or workload as desired. Thus, while fixed-intensity exercise to exhaustion only allows an all-or-none response (exercise continuance vs termination), self-paced exercise of a set load (distance/work/time) allows continual behavioural adjustments that may affect performance. This explains the need to re-examine the literature on hydration and exercise performance, such as the improved exercise duration with adequate hydration observed by the same group who 1 year later observed no such performance improvement.4 7 The study by Marino and colleagues3 is an innovative foray into studies of self-paced exercise performance and its determinants, such as hydration. Importantly, this new study suggests that previous conclusions, based on fixed-intensity exercise models, cannot be assumed to apply to the more realistic setting of self-paced competition.


View Abstract


  • Competing interests None.

  • Patient consent Not obtained.

  • Provenance and peer review Commissioned; not externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.