Dear Sports Scientists: Will drinking fluids keep me cool?

Another look at fluid ingestion and temperature regulation

First, if you did not catch the NY Velocity interview with Ross, be sure to---Andy Shen and co do a great job over there and produce some excellent interviews.  Their site is a must read for any serious or enthusiastic cyclist, whether or not they reside in NYC.

Back in June I was very fortunate to present two sessions at the National Athletic Trainers Association annual meeting in Philadelphia, PA.  Both talks were about fluid ingestion, temperature regulation, and dehydration, and last week I received the audience feedback from the two sessions.  As usual the sessions produced polarized views on the subjects.  So I thought it might be a good time revisit this topic, one we have written about quite a bit on the site and in The Runner's Body.  After all, it is the end of summer, it is hot and humid, and plenty of people are training and racing in the heat.

The title of this post was not inspired by an email we received, but one of the core junctures where the two sides of this argument split is how much fluid is the right amount and why athletes should ingest it.  Nearly everyone will agree that ingesting fluid does have an effect on one's ability to regulate core temperature.  However one side of the argument is that athletes should try to maintain weight losses or lose less than 2% of their starting mass, while the other side feels ingesting fluid to thirst (which normally results in weight loss and some "dehydration") is the best practice.

Why ingest the fluid to prevent or minimize weight losses?  Well, some might argue that if we do not, we get too hot and this predisposes athletes to "heat illness."  The exact meaning and relevance of "heat-illness" is debatable and probably deserves its own post altogether, but the rationale for warding off dehydration by minimizing weight losses is that dehydration causes a rise in core temperature, and that causes heat illness, and that it might even cause heat stroke according to some.

The lit review (in brief!)


The evidence used to support the practice of replacing all or most of your weight losses comes from studies that control for the workload while asking subjects to run or cycle in hot and humid conditions.  The smaller scale studies measure weight losses (and sweat rates) and core temperature, the larger scale ones look at cardiac output and skin blood flow, among other variables.  This is good science, because if we permit our experimental subjects to speed up and slow down then suddenly we cannot determine what is affecting core temperature because now we have two independent variables (intensity and fluid volume) instead of one (fluid volume).  Therefore I am not slating those studies and authors and accusing them of bad science.  


The conclusions are that dehydration, as measured by weight losses, increases cardiovascular strain and results in an elevated core temperature at the same workload.  Fair enough, and as I mentioned earlier I do not think anyone, us included, will try to say that fluid ingestion has absolutely no effect on core temperature---it does, and these studies all demonstrate this effect.  And in fact their science is good, but it is the application of the conclusions that are bad.  In writing a scientific paper it is quite easy to wander off and begin to speculate about why you found what you found.  And it is at that point in your discussion that the reviewers let rip and often times sharply remind you to remain within the confines of your data and draw conclusions based only on what evidence you have available to you (i.e. your data)!


The issue with this topic of fluid and temperature is that the data are all collected within a strict set of conditions---as dictated by the scientific process--but then applied to every athlete (slow, fast, recreational, competitive, elite) in any situation (practice, race, fun) and any condition (cold, warm, hot).


Size counts


The size and magnitude of this effect is terribly small, however.  I try to teach my students in my stats class the difference between statistical and practical significance, and this is a classical exercise for this.  Take the absolute difference between the core temperatures at the end of a typical study, where the subjects exercise for up to two hours.  It is typically between 0.5-0.8 C, or maybe 38.x C in the fluid trial and <39.5 C in the no fluid trial.  Statistically significant?  Yes, most likely.  Practically significant or meaningful?  You are allowed to disagree, but I say "no."


And to follow up with that conclusion, the advice to replace fluids and prevent dehydration is dished out from this evidence even though none of the subjects in these trials ever report signs or symptoms of "heat illness."  So perhaps the real story is that even when we exercise in hot and humid conditions, our core temperatures rarely reach critical levels, and we cope with the additional stress just fine as evidenced by the lack of symptoms.  To me it begs the question, "Why are we telling people to follow this practice," because although there is a difference in temperature, it is small and not otherwise meaningful.


Ingesting fluid keeps does not keep you cool


Long ago, in an exercise lab far, far, away (ok, in Fort Worth, TX), some bored or motivated (or both) students were testing an athlete preparing for the Honolulu Marathon.  At the end of the heat acclimatization period, the runner did a performance run (80 min) at marathon pace (14-15 km/h, or  8.75-9.4 mph) and ingested quite a large volume of fluid (1.75 L) while we measured the rectal temperature response.  He did not ingest quite enough water to prevent weight losses, but came pretty close, losing only 1.6% of his body mass pre to post.
  

And by the way, the conditions in the heat chamber were 27.3 C and 60-65% relative humidity---the expected conditions on race day in December in Hawaii.  

So if the model is that you must prevent or minimize weight losses, and that you must do that to prevent an excessive rise in core temperature, and this model is based on the evidence I mentioned earlier, how does one explain the graph above?  According to that model, this athlete should be no where near 40 C since he lost only 1.6% of his body mass and was only minimally "dehydrated," yet after 80 min he is nearly to 41 C.  And herein lies a problem, because if some data do not support a particular conclusion, said conclusion must be scrapped and we must formulate a new one that incorporates all the available evidence.

Therefore it is not the fluid you ingest that keeps you cool, but as we have written here before it is your metabolic rate, or how hard you are exercising, that really predicts your core temperature during exercise.  Do not mistake what I am saying here, though---fluid plays a role, but only a very small one, and more importantly when we permit athletes to pace themselves they just slow down in the trials where they do not drink or receive very small volumes of fluid.  The result is that they reach the same core temperature at the end of the time trial, but take a little longer to finish.

For me the bigger message is that if performance is a desired outcome, if the runner wants to go as fast as they can, then they should drink to thirst.  Ingesting volumes that are larger than that have not been shown to produce faster performance times.  If performance is not important, the evidence from where I am sitting tells me that ignoring thirst and/or ingesting very small volumes of fluid will result in a miserable day out, but will not cause you to get heat stroke or collapse---two conditions that result from mechanisms other than changes in fluid balance.

If using a "hydration system" and lugging an additional 1-2 kg of mass during training runs makes you feel better, then please continue, but just know that you are not lowering your risk for anything by doing so.  The normal response is to replace less than what we lose, and it is perfectly normal and healthy to drink to thirst.

A look ahead:  Running economy and the marathon WR

Meanwhile, the Fall marathon season is upon us!   Berlin is around the corner and Chicago is boasting a super hot field that, under optimal conditions, is capable of a record.  Earlier we looked at the Joyner paper but left it before discussing what kind of running economy it would take to break two hours, so watch out for that analysis!

Jonathan

Sports drinks, sweat and electrolytes

The actual data from the lab and the field

In the last post we introduced you to Randy, our imaginary 70 kg average male runner, and we created some potential scenarios regarding his fluid and sodium losses and replacement. The biggest take home message was to listen to your body and to drink to thirst, as this has been shown again and again in the field and the lab to keep people from drinking either too little or too much. We have received tons of feedback and discussion, and as we stated in the comments to that post we are pleased that so many of you are participating in the discussion, sharing your stories, and asking relevant and insightful questions.

Just as we will admit that field studies are not the perfect experiment but play an important role, it is the same with prediction equations. We can predict all we want from imaginary scenarios, and some times the equations are pretty accurate, but there is no substitute for the real thing, and many readers wanted to see more actual data and references that demonstrated why sodium ingestion is not necessary, and how both sports drinks and water will can cause a fall in sodium concentration. So while it is important to go through the exercise in the prior post, the next logical step is to look at the actual data.

Baker LB, Munce TA, Kenney WL. "Sex differences in voluntary fluid intake by older adults during exercise." Med Sci Sports Exerc. 2005 May;37(5):789-96

This study examined ad libitum fluid ingestion in older adults during intermittent exercise in the heat. Basically, they had continual access to water in one trial and Gatorade in another, and they had to cycle for 15 min and then rest for 15 min. The total time of each trial was two hours (four work/rest bouts) followed by an additional 30 min recovery period. A number of variables were measured, but we will focus on the sodium concentrations. One limitation of this study for our purposes here is that it was performed in older adults, and there is a well-documented effect of age on the thirst mechanism so that as you become older you become less sensitive to thirst. That is, your plasma osmolality rises higher before thirst kicks in.

According to the authors, their main findings were (and we quote):

1. When cool palatable fluids were readily available, active adults aged 54–70 yr drank enough to match sweating rates and maintain their body mass;
2. Their fluid intake behavior was repeatable;
   
3. CES [note: Gatorade] promoted greater voluntary fluid intake and restored PV losses faster than water;
   
4. There were sex differences in the fluid intake behavior of older active adults, with women drinking more water per kilogram of BM than men

As we tried to explain in many of our prior posts, the ingestion of any hypotonic fluid in excess of thirst will cause a fall in the sodium concentration. In this case "in excess" means drinking more than to your thirst. This occurs even though sports drinks contain some sodium because they still have much less when compared to the body fluids. Therefore the end result is a fall in sodium concentration. The data from this study show that these older adults, even when drinking to thirst, experience a fall in sodium concentration when ingesting water or Gatorade:


What we see is the time on the x-axis and the sodium concentration on the y-axis. The black dots represent the Gatorade trial, and the white (open) dots represent the water trial. All the subjects started with a sodium concentration of 142 mM per Liter, and in both trials the average concentration fell over time to approximately 139-140. There were no differences between the groups, and the symbols you see on the graph means that those values are significantly (statistically) different from the baseline value. So water and Gatorade ingestion produced a similar effect, and so Gatorade did not prevent a fall in sodium concentration in these subjects.

However one female subject ingested 2.8 L of water and 2.7 L of Gatorade in the respective trials. She gained weight in both instances, indicating an excessive fluid ingestion, and here are the data that support the conclusion that ingesting Gatorade will not prevent hyponatremia:


The problem is that the authors herald this as proof that ingesting Gatorade is much better than water:

"Furthermore, this woman’s data support the notion that a CES [Gatorade] is superior to water in limiting reductions in serum sodium during exercise-heat stress. During the CES trial, this female subject consumed 2.7 L and had a final serum sodium of 131 mmol per L. Therefore, although she consumed similar amounts of CES and water, serum sodium was maintained above that of symptomatic hyponatremia during the CES trial."

While the authors are entitled to their interpretation of the data, we disagree and conclude that both fluids are producing a steady fall in sodium concentration, and that the 131 value in the Gatorade trial is just marginally outside the symptomatic range (< style="font-weight: bold;">ingesting either water or Gatorade produced a nearly identical fall (~2-3 mmol) in the sodium concentration.

The take-home part: Sports drinks do not prevent hyponatremia

In fact Jonathan tried to apply this finding to a more "real world" situation in a letter to the British Journal of Sports Medicine. In that letter he argued that since the mean finishing time of women marathoners in America is five hours, and if the ingestion of Gatorade at rates similar to those found in the study is advocated by races, coaches, scientist, etc., then there would likely be many women (and probably men, too) presenting with hyponatremia. These data demonstrate that sports drinks do not prevent this condition as their ingestion in these subjects and at these rates causes a fall in sodium concentration.

Again one limitation to this study is it was done on older subjects, who are less sensitive to thirst, and what we might see in younger subjects would be a slight rise in sodium concentration when ingesting Gatorade and a maintenance of sodium concentration when ingesting water. The evidence for that statement comes from a 1992 study by Robert Cade, the inventor of Gatorade who incidentally died this week at the age of 80. In that study three groups of runners completed a marathon. One ingested Gatorade, another "half-Gatorade" (50% water, 50% Gatorade), and the third group water:

So in fact ingesting Gatorade to thirst in younger subjects results in a rise in sodium concentration, which is why you drink more---you never lower your osmolality below the thirst threshold and therefore are thirstier when ingesting a sports drink, whereas with water you maintain the osmolality right around the thirst threshold and drink and abstain as your thirst comes and goes. With sports drinks you instead just get thirstier, which seems kind of ironic since their slogan is "The thirst quencher!"

So those are some of the data that support our conclusions, and we hope that helps to clarify some of our interpretation(s) and conclusions. We will still post a "wrap-up" for the series on cramping in which we will try to briefly summarize the main points but more importantly leave you with some practical advice on this complex topic!

Fluid intake, dehydration and exercise Part II

Today sees the second post in our series on fluid intake, dehydration and exercise. Yesterday we looked at the history of fluid intake and how radically our beliefs on the subject had changed. Today we turn our attention to the evidence that has accompanied this shift, beginning with the contention that runners who become dehydrated are likely to develop heat stroke.

It is the metabolic rate, not dehydration, that predicts core temperature

If you examine the scientific literature before the 1960's, you will not find much research on how fluid ingestion affects temperature regulation. The earliest thermoregulation study we know of is from 1938 and was performed by Marius Nielsen in Copenhagen. The title was, "Die Regulation der Korpertemperatur bei Muskelarbiet," which in English translates as "The regulation of the body temperature during muscular work." Nielsen performed an exhaustive series of experiments on several men, in which he demonstrated 1) the core temperature goes up as you exercise at higher power outputs and therefore exercise intensities; and 2) the core temperature is regulated at a higher level during exercise. In fact one figure shows that during a four-hour exercise bout at 180 Watts, the rectal temperature is maintained at about 38 C for the duration of the exercise.

Fast forward to 1960, when Sid Robinson published an article titled, "Temperature regulation in exercise." He began his summary of that paper with this:

"The central body temperature of a man rises gradually during the first half hour of a period of work to a higher level and this level is precisely maintained until the work is stopped...During prolonged work the temperature regulatory center in the hypothalamus appears to be reset at a level which is proportional to the intensity of the work and this setting is independent of environmental temperature changes ranging from cold to moderately warm."

Robinson's paper agreed with the findings of Nielsen---namely, that the core temperature during exercise is regulated at a higher level during exercise. This was precisely our point in our "When abnormal is normal" post: the body is quite happy with change during exercise, and many variables are regulated at different (higher) levels during exercise without any problems. In addition, these scientists showed that the metabolic rate is the thing that determines your temperature. In other words, the harder you exercise the hotter you get.

Introducing fluids - dehydration takes over

Up until this point, little mention was made of fluid - it was all about work rate. In 1970, Professor David Costill published the first lab study that investigated the effects of fluid ingestion on temperature regulation. In that experiment the main finding was that when the runners drank no fluid, their temperatures were higher at the end of the two-hour run, and therefore a relationship between the volume of fluid ingested and the rise in core temperature was established. Several other studies together with this one, then, appear to have shifted the paradigm. Gone was the idea that metabolic rate predicted the rectal temperature, and in its place came the concept that the level of dehydration was responsible for driving the temperature higher during exercise.


The paradigm shifts - Dehydration as the cause of the rise core temperature

We now move into the 1990's, which saw a more robust and detailed repeat of Costill's 1970 study. In this one, published in 1992, the cyclists were made to cycle on a bicycle in a laboratory for two hours, while drinking different volumes of fluid. The title of this study says it all: "The influence of graded dehydration on hyperthermia and cardiovascular drift during exercise."

The main finding, which is shown in the graph above (click to enlarge it) is that when the subjects drank no fluid (NF - the open circles) they had the highest rectal temperature, and when they drank more (LF - the solid circles), their temperatures were lower.

So then it is case closed? Two very good lab studies, published by very well known and respected Exercise Physiologists, seem to show very clearly that ingesting more fluid keeps your body cooler. The studies were well-controlled and the data are robust, right?

Good science - but can it be applied to YOU?

Well, yes, they were well-controlled and experimentally sound. However both of these studies have major limitations to the manner in which they can be applied, and here is why. First, the wind speeds that were moving over the runners and cyclists were not anything like what they would experience when exercising outside.

For example, the Costill study used elite level runners (average VO2max of 74 mL/kg!) exercising at 70% VO2max. This running speed corresponds to more than 15 km/hour. However the air blowing on them from a fan was only moving at 5.7km/h, which is no more like the speed of air when you're walking. Likewise, in the Coyle study, the cyclists were riding at power outputs corresponding to speeds of 30km/h, yet the air moving over them was only a mere 9km/h. Therefore:

• These atheltes were being made to exercise at a high intensity, which means that they are PRODUCING substantial heat, BUT...
• They were denied the opportunity to lose this heat because they were not given appropriate wind speeds

The importance of wind speed - far more significant than fluid intake

The effect of convective cooling is substantial. We won't go into the physics of the equations, but we can tell you that if you do a mathematical model, you discover that a change in wind speed of only 1km/h can change body temperature by more than 2 degrees Celsius over the course of a two hour trial. Now, if you take a study like that of Coyle's, where the small fan provided wind speeds of 9km/h, then you can see how vast the difference would be if the windspeed was just a little higher, let alone the realistic 30km/h!

Therefore, while these studies showed clearly an effect, it is arguable that this small effect was amplified by the lack of wind speed in these trials. Even in these two trials the differences in core temperature at the end of the two hours of exercise was less than 1 C. Even more important, the peak temperature were what we would call "normal" as they were only ~39 C. Even more still, the authors of these studies, although claiming the effects of dehydration on temperature, did not report that any of their subjects suffered any ill effects from not ingesting fluid, and they reported no signs or symptoms of any kind of "heat illness" after the exercise trials.

So does it come as any surprise that the 1992 study above was, and we quote, "supported by a grant from the Gatorade Sports Science Institue?"

Conclusion

To summarize this post, the important observations here are the following:

1. Metabolic rate is the best indicator of the core temperature
2. Any affect of fluid ingestion on the ability to regulate the core temperature is small (less than 1 C)
3. A major lack of convective cooling might have amplified this small effect so that it is even much less than 1 C
4. The workload in these studies was fixed, and the subjects were not allowed to pace themselves as they are in a real race situation.
   

The last point above is yet another limitation to how these studies are applied. . .however we will leave it at that for now as this post already is very lengthy---remember that we wrote theses on these topics! So to condense the concepts down into one post is a real challenge. Be sure to keep coming back for the next post in this series, where we will show you the real effects of dehydration on cycling performance and temperature regulation!

Fluid intake, dehydration and exercise: A new series of posts

First let us say a big "Thank You" to everyone who has asked questions, posted comments, and contributed to the debates here at the Science of Sport. We have had an incredibly positive reaction to our posts on the Chicago Marathon, and many relevant and good questions were asked in the wake of those posts. Therefore we are excited to bring you a new five-part series of posts on fluid replacement during exercise.

In this series we will examine the following concepts and ideas:

1. The history of fluid replacement in the marathon and endurance sports
2. Does dehydration really cause you to have a higher core temperature?
3. Evaluation of laboratory-based studies vs. field exercise: Is there a difference?
   
4. Hyponatremia: a disorder of fluid, not sodium, balance
5. The physiology of thirst: Why waiting until you are thirsty is NOT too late

We sincerely hope that this series generates as much interest as our previous two series on running technique and men vs. women in running. This debate is controversial, and has stirred serious debate in the past, as I'm sure it will now. We welcome this of course...

But at the heart of this debate is an eternal conflict between science and the commercial world. As we will see, the lines between marketing strategy and research strategy become blurred and confused, scientists begin to drive product sales through 'endorsements' from research, and a conflict of interests is created that cannot be tenable. The topic is worth discussing for this reason alone, but also affects every single person's training routine.

So tune in tomorrow for Part I, and keep on coming back for the full story. Again, because it's a series (in order to avoid writing a thesis in one go!), we'll move systematically through the topics, and so may seemingly omit points from intial posts. As always, we will rely on you, the reader, to ask the unasked questions, and to contribute to the debate with your insightful comments. We'll certainly do our best to bring a scientific analysis and explanation of these everyday concepts, and hope we can help you understand them better.

See you then!