The guidelines regarding how much dehydration an athlete can tolerate before performance deteriorates has changed over the years, but what level of dehydration is acceptable and is it possible to avoid dehydration? 

Early guidelines for dehydration

Not all that long ago, the prevailing opinion in sports science was that you needed to replace 100% of your sweat losses to maintain your performance when exercising. In 1996 the American College of Sports Medicine stated that…

“During exercise, athletes should start drinking early and at regular intervals in an attempt to consume fluids at a rate sufficient to replace all the water lost through sweating (i.e., body weight loss), or consume the maximal amount that can be tolerated.”

Now, granted, these guidelines were not written specifically with long endurance events in mind. But if taken at face value the statement seems to imply that during an Ironman, ultra-running race or long sportive, some athletes should be aiming to drink as much as 2-3 litres an hour in order to replace 100% of their sweat output.

100% replacement of fluid losses during exercise just isn't realistic. That’s a hell of a lot of drinking. In fact, it’s beyond what is physically possible for most people.

That 1996 statement - along with the general “drink, drink, drink” marketing messages coming from the sports drink industry at around the same time - has been blamed for driving the worrying increase in cases of hyponatremia (ill health - or even death in extreme cases - caused by the over-consumption of fluid) seen in amateur sports. In fact, a study conducted at Ironman European Championships a few years ago found that as many as 10% of finishers had hyponatremia to some extent, which will have impacted upon their performance.

Revised guidelines

So, this advice soon came under fire from a lot of people in the sports medicine community and significant evidence contradicting it was unearthed. As a result the ACSM revised it’s guideline in 2007 to include the more conservative (if slightly vague) statement that…

“The goal of drinking during exercise is to prevent excessive (>2% body weight loss from water deficit) dehydration and excessive changes in electrolyte balance to avert compromised performance.”

Again, although the revised guidelines were undoubtedly more conservative and accept that 100% like for like replacement of sweat losses is not necessary during exercise, it still suggests that only dehydration of up to 2% (defined by body weight loss) is ‘acceptable’. 

The suggestion seems to be that dehydration beyond a 2% bodyweight loss must be detrimental to your performance (or health), meaning that in practice athletes ought to try to get pretty close to balancing ‘sweat out’ with ‘drinks in’ when training and competing.  

Just to calibrate your thinking on this, 2% dehydration defined by bodyweight loss would amount to an approximate 1.4L (~47oz) drop in mass for a 70kg (154lb) athlete.

Is the 2% dehydration threshold realistic?

There are a number of older lab-based studies in the scientific literature that do seem to point to an approximately 2% level of dehydration (by body weight loss) being detrimental to exercise performance and these are presumably what the 2007 ACSM guidelines are based on.

The issue is that quite a few of these studies have been criticised as being unrepresentative of what happens in the ‘real world’.

For example, some of the original lab studies into dehydration either used periods of induced sweating (saunas or hot baths) or diuretics (drugs that make you pee a lot) to artificially shed body water before exercise tests started, rather than letting dehydration accrue more gradually as it usually does during the course of normal training or racing.

This has procedural advantages in terms of how fast and accurately dehydration could be achieve in the subjects taking part, but it’s also artificial. Few athletes would deliberately try to do a hard training session or race having deliberately dehydrated themselves immediately before it! In the real world, fluid losses usually tend to build up over time once athletes have started exercising in a well-hydrated state.

Although these kind of criticisms of the early research are valid, this doesn’t mean there’s nothing to be learned from it. What they do show pretty conclusively is that uncorrected dehydration can massively impact physical performance in certain circumstances.

But, it’s fair to say that the consensus in sports science over the last couple of decades has shifted towards the idea that some level of uncorrected fluid loss is acceptable (maybe even desirable) as opposed to athletes having to aim for a 100% replacement of sweat whilst exercising.

That’s especially the case if they’re allowed to respond to the dictates of thirst and to start exercises well hydrated in the first place. Defining an exact level of dehydration that negatively effects performance (if one even exists) is still proving elusive though.

Dehydration studies

In the 2000s, a number of sports science researchers cast aside their lab coats and ventured outdoors to collect hydration-related data from athletes competing and training in the field.

Several studies looked at bodyweight changes in athletes taking part in endurance races (including Ironman triathlons) to estimate their sweat losses and rates of fluid replenishment, and compare that with performance data.

One of the most famous studies managed to obtain pre and post race bodyweight data from legendary distance runner Haile Gebrselassie. It demonstrated that he dropped a staggering 9.8% of his mass during the 2009 Dubai marathon - a race that he won in 2:05:29. 

This example in particular has often been used to attack the ‘2% dehydration impairs performance’ theory head on. After all, if someone can win a marathon in a world class time having basically dropped 10% body weight, how can dehydration credibly be said to be negatively impacting their performance?

Numbers were also crunched specifically on Ironman competitors, including large data sets from the 2000 and 2001 IM South Africa and the 2004 IM New Zealand races. What these seemed to show was that, whilst there was a huge spread of bodyweight change in these events, there was a general trend towards the faster competitors losing relatively large amounts of body weight (~4-8% loss for the top 5 finishers in SA and ~1-6% in NZ) when compared with the theoretical 2% benchmark. 

Most competitors lost weight in the races (as you’d expect), with around a 11% bodyweight loss being the most extreme drop reported in one case. It’s worth pointing out that - despite the demands of the arduous event - a small minority of athletes actually managed to gain weight from start to finish during both IM SA and IM NZ! Presumably by drinking more than they sweated out. Maybe they should get a special kind of medal for that?! 

But, when you look at the data as a whole, it’s clear that the majority of finishers are clustered in the 1-6% bodyweight loss area during an Ironman. Whilst that tells us little about their performance compared to their potential, it’s not unreasonable to say that something in that area would be considered an ‘average’ level of loss, for ultra distance events in climates similar to South Africa and New Zealand at the very least.

In addition to all the Ironman data, a 2013 study at the 100 mile Western States ultra-marathon running race showed that body weight changes of 1-6% were reported in the top finishers (and the Western States 100 is generally longer in duration than an Ironman, with the overall winning time in the region of 14-15 hours).

The relationship between body weight and sweat loss

Changes in body weight have long been used as a measure for estimating sweat losses in athletes. That’s because fluid loss through sweating accounts for the majority of any change in body mass during exercise. Body weight is also really easy to measure with a set of scales, certainly easier than running behind athletes collecting all the sweat that drips off them anyway!

So, body weight loss has almost become synonymous with sweat loss, with 1kg of weight loss being said to equal 1 litre of sweat output. But some scientists have cast doubt over this assumption, when it comes to longer duration exercise at least.

The problem with assuming that essentially all weight loss during exercise is fluid lost via sweating is that it fails to take into account the fact that you’re also burning through a lot of stored fuel when you’re working hard. 

You burn carbs, fats and some protein during exercise. The by-products of these biological processes are CO2 (that you breath out and so you become lighter) and water which is released and becomes available to the total pool of fluid in your body, so it can be assumed to offset some of what you lose in sweat.

But in an ultra distance event - where a lot of calories are burned (~7,000 - 11,000 Kcal during an Ironman for example) - the amount of fuel used up can be a significant confounding factor. So, assuming weight loss equals sweat loss is more problematic.

It’s also worth noting that your body stores somewhere around ~500g of glycogen in your muscles and liver and with every gram of glycogen something like 1-3g of water is also stored alongside it. As this glycogen is burned up (and you can burn through all of it in 90 minutes if you’re going at it hard) this process releases something like 0.5 to 1.5L of additional ‘free water’ for your body to re-use internally.

If you do the maths on what all the above could mean during a long event (as a group of researchers did at the 2016 Western States Run) it could theoretically mean that a runner could lose 4.5-6% of their bodyweight without seeing a significant change in total body water. In other words, they would maintain a good hydration status despite losing a hell of a lot of weight.

It’s worth pointing out that there are a lot of assumptions behind that calculation and that low numbers of athletes have been studied with it in mind. But it does open up some interesting questions about the level of bodyweight loss that ultra endurance athletes might be able to experience in long events without becoming seriously dehydrated.

Pre-exercise fluctuations in body weight

In my own experience of preparing for an Ironman (or other long events) I would tend to taper my training aggressively in the final few days, eat more food (especially carbs to top up my glycogen stores) and take on more sodium to help retain fluid. This would result in a significant weight gain in the final few days before my event. I’m also pretty sure I’m not alone in doing this either.

The result of this resting and feeding would typically leave me standing on the start line for a long distance event 1.5-2kg heavier than my ‘normal’ day to day weight. I always took this as a positive sign that I was sufficiently fortified for the long battle ahead and in numerical terms it would have meant that I was 2.2-2.8% heavier than usual.

Exactly how much of that additional weight was water (versus extra glycogen and body fat) would be hard to speculate, but even it if was 50% then this would presumably have a sizeable impact on the amount of sweat I could lose before becoming noticeably dehydrated. And what about training, where I’m unlikely to have followed the same pre-session ritual? 

I actually think that this might be a factor in case studies like the 9.8% bodyweight drop seen in Gebrselassie in his Dubai marathon win. Most marathon runners do a considerable amount of carb loading pre-race (as do triathletes) and so him losing so much mass on the day could well have been partly influenced by being somewhat above ‘normal’ bodyweight at the start of the event.

What level of dehydration is acceptable?

It seems clear that aiming to replace all of your sweat losses during exercise is unnecessary.

With so many case studies demonstrating that elite performers seem to lose some bodyweight throughout a race, you’d need an impressive lawyer to help build an argument that 100% like-for-like replacement is needed to support a high level of performance.

Although it’s not always possible to extrapolate what elite athletes can do down to “the rest of us”, the data seems to support the theory that most amateur athletes can finish an ultra distance event whilst experiencing a measurable degree of weight loss.

There are very few examples of athletes losing more than 10% bodyweight after finishing events under their own steam. It’s probably fair to say that something in the ~1-6% range seems to be ‘normal’ for ultra distance events (in hot climates at least).

There’s also highly likely to be a great deal of difference in the level of dehydration one athlete can tolerate to the next. Can Gebrselassie tolerate 10% bodyweight loss during a marathon because he’s adapted to that over years of training and competition at the highest level? Or has he got an innate ability to cope with high levels of sweat loss and is this something that contributes to him being a world class runner?

I think it’s clear that all athletes ought to be able to tolerate a measurable degree of dehydration when doing longer endurance events, assuming they start properly hydrated in the first place. No-one needs to aim for 100% fluid replacement on the move. In fact, trying to do so could lead to an increased risk of hyponatremia which is bad for your performance and health.

The sweet spot is likely to be highly individual and may even be slightly different for the same athlete on different days and in different climates. But, it probably falls somewhere between 2-4% body weight loss for most athletes

Ways to avoid dehydration

I’m hopefully preaching to the converted here but it’s definitely worth thinking about the role of sodium in retaining fluid in your quest to maintain your hydration status. Sodium helps you retain fluid in your bloodstream, useful when your sweat output is high.

Studies such as this recent one by researchers in Spain looked at fluid and sodium intake during a hot half-Ironman event in 2015. Their data showed that athletes who took in more sodium tended to lose less bodyweight in percentage terms than a control group who only took a placebo. 

The group that took the sodium in (and so maintained a better hydration status) outperformed those taking the placebo by an average of 26 minutes in the race. There seemed to be a clear and positive correlation between increased performance and losing less bodyweight overall by using sodium supplementation to absorb and retain more fluid. 

It’s well worth doing some testing in training and racing to get an idea of how much you’re sweating and how salty your sweat is, then you can take our free Fuel & Hydration Planner to help you understand what kind of fluid and electrolyte intake regime might work for you. This can then be refined through some good ol’ fashioned trial and error in training.

Or, if you want to know exactly how much sodium you’re losing in your sweat so you can more accurately replace a decent % of what you’re losing, it’s worth considering taking our Sweat Test.

Further Reading