Welcome to the Science of Sport, where we bring you the second, third, and fourth level of analysis you will not find anywhere else.

Be it doping in sport, hot topics like Caster Semenya or Oscar Pistorius, or the dehydration myth, we try to translate the science behind sports and sports performance.

Consider a donation if you like what you see here!

Did you know?
We published The Runner's Body in May 2009. With an average 4.4/5 stars on Amazon.com, it has been receiving positive reviews from runners and non-runners alike.

Available for the Kindle and also in the traditional paper back. It will make a great gift for the runners you know, and helps support our work here on The Science of Sport.

Wednesday, May 28, 2008

Fatigue Series Part 5: Exercise in the heat

Anticipatory regulation of exercise in the heat: Discussion continues

About a week ago, in our last post of our Series on Fatigue, we looked in some detail at a study by Frank Marino which found that African runners paced themselves differently to White runners during 8km time-trials in hot, but not cool conditions. Part of this difference was likely the larger body size of the white runners, which meant that their rate of heat storage would be higher than the African runners' at the SAME SPEED. Therefore, the theory put forward was that the RATE OF HEAT STORAGE mediates a reduction in running speed well before any potentially limiting level of hyperthermia is reached.

A couple of things arose out of this post. First, we got quite a few posts by people saying that they should have controlled for body size, and made sure that the two groups were equally large (or small, as they case may be). This is probably correct, technically speaking, but a little harsh and maybe missing the point of the study. The key here was not so much the mechanism for the different pacing strategies of Africans and White runners, but rather the fact that they did it at all. Perhaps it's genetic, perhaps size-related, perhaps metabolic, perhaps related to running economy? That's all for future work to establish, hopefully. But the point is that athletes pace themselves differently and the rate of heat storage is a very likely candidate that mediates this difference.

One reader said that they should have controlled for calf-size as well, which is also probably true, but if you go down that road, then you have to control EVERYTHING. And physiology is simply too complex to do this. That is why, as you may recall, we discussed how for many years, scientists used to limit themselves to these fixed work rate trials to exhaustion - they are simpler to manage. As soon as you allow pacing, the complexity becomes enormous, but it's the only realistic way to assess how PHYSIOLOGY works in the field.

Today, we look at further studies that have attempted to assess this, but this time, with a possible mechanism. For that, I get to summarize my own study, which is a little self-indulgent. It was not intended in this way, but was rather the result of the fact that five or six years ago, nobody was doing this kind of work. Still today, there are some problems with it (again, the motto is "Nobody can PROVE anything"), but it's worth looking at.

Anticipatory regulation of performance in the heat

Refresh your memory on the state of the knowledge prior to 2003. The thinking regarding exercise in the heat was that you fatigued because you were hot. That is:

  • Exercise increased heat production
  • In hot and humid environments, you are not abe to lose that heat
  • Your rate of heat storage is positive, so your body temperature rises
  • It rises until it reaches a critical limiting level of about 40 degrees Celsius
  • At that point, your brain fails to activate the muscle, your level of effort hits maximum, and you stop exercise
Impaired performance in the heat is thus the result of GETTING TOO HOT, to put it simply. This theory was borne out by studies that showed how brain activity was altered and EMG activity was lower when the body temperature was 40 degrees celsius. The problem was, these studies all fixed the cycling power output (or running speed), and so there was no room to slow down - it's either go or stop.

So, in 2002, I did a study in Cape Town that aimed to determine WHEN the decision is made to slow down or speed up, or, in the case of the existing theory, stop altogether?

This study, which was published in the European Journal of Physiology (Tucker et al. Eur J Physiol; 448: 422-430, 2004, for those interested), aimed to answer the following questions (in lay terms):
  • During exercise in the heat, WHEN does the athlete slow down? The current thinking was that they slowed down BECAUSE they got too hot. But Marino and some others were suggesting it happened before this.
  • What mechanism might exist to cause this slow down during exercise in the heat?

The study was relatively simple: 12 well-trained cyclists performed 20 km time-trials in the lab, either in the hot condition (35 degrees, 60% humidity), or cool (15 degrees, 60% humidity). During the trials, we measured something called EMG activity, which is basically the electrical signal sent from the brain, to the muscle to cause it to contract. This method, which is the same as was used previously to show how the brain activated less muscle when it reached 40 degrees, always ends up being the point of attack for people who don't buy into the whole regulation of exercise argument, but more on that later.

Things like heart rate, Rating of Perceived Exertion, skin temperature, body temperature were all measured during the trials as well. I'll sum up the two key findings below:

1. The pacing strategy differs, almost from the start of the trial

The graph below shows the power output measured through the trials. You'll not that in the heat, for the first 5 km, the power was the same as in the cold, and then it started dropping, whereas it was maintained in the cool trial. The result was that the overall power output was lower in the heat. Nothing unexpected there...

The mechanism - muscle activation and anticipatory regulation

But, what you should be asking is the following:

Why did the cyclists slow down after only 30% of the trial was completed?

There are two possible answers to that question:

You could say, based on the theory of heat LIMITING performance, that they slow down because their body temperature has risen quite high in those first 5km, and they slow down, because as was shown recently, a high body temperature directly prevents the brain from activating muscle;

OR, you might say

They slow down at this point so that they don't get hot later on during exercise. That agrees with the Marino theory for anticipatory pacing, and something other than high body temperature is responsible for reducing their power output.

The graph below shows the answer to this question:

What this graph shows is the EMG activity (as a % of maxium - we express it relative to some maximal value of muscle activity, measured before the trial when the cyclist pushes as hard as possible for 5 seconds) over the course of the trial.

You'll notice two key things:

1) First, the EMG activity is lower in the heat than in the cold, almost from the outset

2) The EMG activity increases significantly at the end of the trial - the "endspurt"

These changes in EMG activity EXPLAIN the changes in power output in our previous graph. That is, the power output in the heat is lower BECAUSE the activation of muscle is lower from very early on. Then, at the end of the trial, the power output increases substantially because the brain is activating more muscle. More muscle activation means more force, and that means more power.

But perhaps the key to all this comes from the tables I've inserted over the graph, which show that:

  1. The athlete slows down (the power output graph on top) and activates less muscle (the EMG graph below) even though their body temperatures, heart rates and even their Perception of Effort (the RPE) are THE SAME as in the cool condition. If you compare the HOT to the COOL conditions, you see that the body temperatures are "only" 38.4 degrees celsius, which is not different from the COOL condition, and nor is it anywhere close to the supposed "limit" to exercise of 40 degrees.
  2. Think for a moment about that for a moment - they "choose" to activate less muscle, to cycle at a lower power output, despite the fact that they are NOT HOT, and nowhere near the supposed "critical limiting temperature". This may strike you as obvious, but again, you need to ask HOW they could possibly know this, and based on what information is such a 'decision' made?
  3. Then, at the end of the trial, the athlete is able to SPEED UP in the cold trial, activate MORE MUSCLE, even though their body temperature is higher than it was before

Quite clearly, the decision to speed up or slow down has nothing to do with body temperature, which is what the textbooks say. These findings show that the activation of muscle, the power output and hence performance are regulated by something much more complex that simply the direct effect of body temperature.

The most amazing of all - you slow down, even though you feel the same!

What is perhaps most remarkable of all is that the cyclists slowed down in the heat even though their perception of effort was the same as in the cool condition. This perception of effort basically measures an overall Rating of Exertion, which is to say it's a mix of fatigue, effort and general perception. It's a highly complex measurement, and we'll come back to it later in this series.

Point is, it's not as though they felt worse, and therefore slowed down! That's what you might think, but the finding above suggests this is not the case. In other words:

  • you feel the same in terms of your effort and fatigue levels
  • you're equally as hot as you were in the cold condition
  • your heart is working at about the same level

yet you slow down through the activation of less muscle.

Now, there are many issues here that I won't get into for this post, but will gladly discuss in question and answer things (so do read the comments at the bottom of this post because your question may well come up there!). So yes, there are some grey areas, there are mechanisms still missing (what causes them to slow down, for example?) and there's much to be discovered still. But the take-home message here is that:

A model that says that you fatigue in the heat because you get too hot is clearly incorrect. Rather, fatigue in the heat is complex, and impaired performances happen long before athletes ever get hot. The regulation of exercise happens in anticipation of overheating, and it's mediated by factors that are still too complex to pin down exactly. However, there are theories, and that's what we will address next.

Join us then!



Anonymous said...

The question, of course, is can we adapt to allow more rapid heat dissipation and to alter this "anticipatory" regulation. What effect might this have on cold climate runners performing in hot climate events?

Unknown said...

Very interesting. I wonder how much "learned" behavior contributes to this anticipatory regulation of energy output and if there is an evolutionary connection to survival in hot conditions. Could this be a sort of subconscious mechanism for athletes in a higher temperature environment as a behavior adaptation so that they have something left in the end to beat out the competition?

Anonymous said...

Were you able to look at distribution of blood volume in the study?

I would think, that blood volumes would shift towards the skin in the hot condition. Thus, preventing normal flow to the muscles (as in cold condition), limiting muscle stimulation and thus less power.

If the athlete's were acclimated to heat, they would have adequate blood volume and distribution to prevent this from happening to the same degree. Any thoughts?

Great blog.

Amy@RunnersLounge said...

Hi Jonathon and Ross,

Great blog! Very interesting and educational articles. You do a nice job blending scientific information and "usefulness".

I was pointed to your blog by a fellow runner. She mentioned your great articles on running and heat. This week at Runners' Lounge, our theme is running in the heat. Each Thursday, we ask runners to drop off links to great articles to share with others on the chosen topic. I would be honored if you might participate and share one (or more) of your articles as links to for other runners. Check it out at blog.runnerslounge.com

Would you have an interest in sharing more of your articles with runners through the "know how" section in the Lounge (www.runnerslounge.com) and/or having your blog featured? If so, drop me a note at amy@runnerslounge.com

Great job!


Ross Tucker and Jonathan Dugas said...

Hi everyone

Thank you for the questions, which are really good ones - conferences often don't produce such insightful questions as yours! I'll do my best to respond to each below.

First, to anonymous:

That is of course the million dollar question. If you recall back to the very first two posts of this seres (an eternity ago), you will remember that I spoke about "Inputs" and "Outputs". Training will alter both.

The "Inputs" represent the rate of heat storage, the "Output" represents the brain's decision to alter muscle activation. both will change, because a trained athlete reduces their rate of heat storage, but also probably responds to the incoming signal differently, allowing them closer to the limit. The anticipatory regulation still happens, but it likely happens later and with less negative impact on performance.

Second, to Jen:

Good question. It most certainly would have something to do with survival in hot climates. Remember, humans in Africa used to hunt in the hottest part of the day, often running for hours in order to catch prey. Their ability to survive depending on pacing themselves appropriately, because the larger animals there were hunting would eventually simply stop running and give up. Why? Because those larger animals had reached a limiting core temperature value, whereas the smaller humans had not. So yes, I believe pacing played a crucial role in our survival.

Third, to Scott:

No, we didn't. It had been shown in a number of studies in the heat, from about 1979 onwards, that this redistribution of blood volume did not affect performance. So you're quite right that there is this circulatory adjustment, where blood is sent to the skin, but the research had shown pretty conclusively that this did little to impact on performance. We covered this in Part 3 of this series, when we looked at the limiting core temperature.

The key was that even though the body did send blood to the skin, it was more than able to adjust for this and maintain blood pressure and blood flow to the muscle. So the model that says that exercise in the heat is limited by insufficient blood to the muscle was proven incorrect, and therefore didn't need tob e considered.

Finally, to Amy:

Thank you very much for the support and the invitation. We'll definitely take you up on the offer. I have a bit of work to plough through today, but I'll get on it when I have some time. We did a whole series last year on running in the heat and heatstroke, and then of course, the last two articles we've done also apply. So there are probably four or five, I'll put one forward today.

Thanks again!


Phil said...

This series is fascinating. Could you comment on the significance (in the statistical sense) of the result... from the graph it looks like the difference in power was not that great compared to the standard deviation? Hard to recruit 1000 cyclists to get a bulletproof result I realize!

Ross Tucker and Jonathan Dugas said...

Hi Phil

Sure, sorry we left that out - it's a fine balance between being over-scientific and basically reproducing a journal article here, and writing it in a manner that "compromises" the science of the research...

But yes, the values are statistically significant, with P < 0.001. That's done with a Repeated Measures ANOVA, if that means anything

One of the problems in these physiological trials (apart from the issue of subject number to reduce the SD, as you mention), is that there is often large variability between subjects. The result is that Bob comes in and rides at 400W, whereas Fred rides at 250 W. The SD is enormous as a result, but the reason one can find quite strong significant differences is because the two guys show the same response when moved into the heat. In other words, Bob now rides 350W, Fred 200W - both lower, but still with large between subject variability.

So it does sometimes hinder the chances of finding difference, but it's saved by a very low intra-subject variability and what is a really predictable response of people to the "intervention" - in this study, for example, 11 out of 12 guys were slower in the heat by 10%, one of the 12 performed 2% slower.

Thanks for the question, hope that answers it!


Anonymous said...

G'day again Ross and Jonathan

I appreciate the very interesting direction you are coming from here: assuming I have understood you correctly, your sustained argument here is that it is not the intrinsic factors of core body temperature or other traditionally measured physiological factors such as heart rate that determine the athlete's response to thermal conditions (I have to say again, I don't think 'heat' is precise enough here): rather it is the athlete's perception of these conditions that then determine his or her pace.

My response to this fascinating and plausible thesis is:

1. What about then establishing a scale similar to the RPE, but instead analogously evaluating the athlete's RPETC - "Perception of Extrinsic Thermal Conditions"; and his or her RPITC "Perception of Intrinsic Thermal Conditions" (basically, how hot the athlete feels)?

2. Then, what about trying to correlate the RPETC and the RPITC with the intrinsic physiological measurements - including body mass, total surface area, exposed surface area; AND the extrinsic physical measurements of temperature, humidity, solar load, wind speed and (as many as possible) of all the other extrinsic factors that (from my personal experience) seem to be so critical when one races near to one's maximum effort in very hot and humid conditions?

3. Not sure how representative results for indoor cyclists (club, sub-elite, professional) are compared to outdoor professional cyclists in a sustained, real event such as the Tour de France.

4. Not sure that cycling on indoor ergos would necessarily gives same results as indoor running on a treadmill; or that running indoors on a treadmill is the same as running outdoors on an unshaded, open air course that heats up as the morning unwinds; undulates; is exposed to the sun and the wind.

5. The perception of being uncomfortably hot long before there is a discernible change in intrinsic parameters I have long thought of as the “Overcoat Syndrome”. I call it that because, before I began racing and competing in the tropics in the early 1990s, I lived for far too many years in the world’s ghastliest climate where one was almost always too cold or too hot no matter what one wore: too cold because of the usually inclement weather (back then); too hot because as soon as one tried to dress for the conditions one invariably overheated at the slightest change in the extrinsic thermal conditions. If one togged up on a cold, wet blustery day in January then as soon as one went from the street into a department store one immediately felt that one was about to explode (long before, I suspect, one actually started to overheat). If one wore a snug little semi-permeable jacket that did a great job keeping out the wind and rain on a long Sunday run then, Bobs your Uncle, as soon as one went round a corner and were out of the wind for an instant, one immediately felt stifled and was yanking urgently at the front zip.

My hunch is that the 'Overcoat Syndrome' is triggered by a combination of skin temperature and evaporation rates (evidently not by core temperature). A small increase in skin temperature and reduction in evaporation rates, particularly perhaps from certain areas of the body, leads quickly to the sensation of being stifled or overheated. Quite a lot of research has been done in this area on the perception of comfort of people inside badly heated rooms where the walls are at a different temperature to the surrounding air. Perhaps some of this is relevant here?

michael said...

Perhaps slowing down during exercise in the heat might be the a response to the ambient temperature (humidity?) instead of body temperature. Temperature receptors in the skin 'tell' the brain that it's hot outside, the information is processed in the brain, and as an output (reflex), muscle activation is reduced. I couldn't suggest the mechanism, of course, how such information is forwarded. But the point that the body reacts directly to ambient temperature could be checked by having subjects cycle in a room heated with infra-red lamps which won't raise the ambient temperature but only the subjects' body temperature. Possibly, they will not slow down at all, if body temperature has now role in anticipatory regulation.

Anonymous said...

Snap, Michael!

Anonymous said...

I might have overlooked it, but were the cyclists aware of their pace via speed or power reading? I'm just wondering if they knew they were on a slower pace in the heat and despite their best efforts were unable to go any faster? Or if they were going completely by perceived effort? Might tell us something (or might not).

The Sports Scientists said...

Hi Thom, and thanks for the valid question about the feedback to the subjects.

You have correctly identified that providing such feedback is a major confounder in a performance trial like this, as is the learning effect of performing one trial.

In our research unit at UCT we do three things to overcome these problems:

First, we make sure we know the repeatability of the performance test. We used this 20 km TT and also a different 100 km TT test for which we have performed repeatability studies and know the day-to-day variation in these tests. Briefly, these studies consisted of having trained cyclists complete the TT's on three different occasions with no interventions, but with repeating the same protocol (food, drink, time of day, etc) each time.

Next, we have each subject do a practice or familiarization trial first since we found in the repeatability studies that people improve from Trial 1 to Trial 2, but not from Trial 2 to Trial 3.

Finally, we blind the subjects from all feedback except distance. They may ask how far they have gone, and we also tell them at different time points (i.e., "You have completed five km"). However we do not tell them heart rate, speed, power, or anything else. In fact in a study such as this where each subject is completing multiple TT's, we do not even tell them their time from each TT until they complete the entire study.

Again, it can be a big confounder if you let the subjects have too much info, and that is why we blind them to everything except distance. Letting them know the endpoint, and more specifically how much work is left to do, is quite crucial, and I am sure we will get into this in a future post in this series as some of our recent work deals with this scenario (of not knowing how much work is left).

Thanks again for the comment!

Kind Regards,

Andrew said...

Have any studies evaluated the pacing effects of variables besides air temperature? For example, running uphill/downhill, cycling with a headwind/tailwind, swimming with/against a current, etc?

As you stated, we expect athletes to slow down with increasing "environmental adversity" despite equal RPE. It's the equal RPE that's really fascinating about your cycling study - clearly the athletes are somehow aware of their environment (or it's physiological effects) otherwise the RPE would be proportional to pace.

Anonymous said...

So the real question (and I hope you're getting into this at some point -- you promised to talk about what happens when the body is deceived...) from an athletic point of view is this: does the anticipatory regulation system (?do we call it a system?) lead to optimal pacing and performance results? Or should athletes be working to override this system to some degree; that is, to go faster in the middle of a race (or at the beginning of a race on hot days?) than the body is 'telling' us to?

Anonymous said...

Hi Jonathan & Ross
Extremely interesting. A couple of questions: Was the group mean HRmax found during the peak power trial any different than the final HR in each condition? Also, did you consider including split times and the change in power over each distance interval rather than %total time? How about a frequency spectrum for the EMG? Thanks.

Anonymous said...

Hi, so similar to Steve really..basically, what implications do your findings have for long-distance athletes?

Anonymous said...

so what happened to the rest of this series? I got to the end of Part 5 and it just ended. Can you post a link to the entire series or did you not complete it?

Ross Tucker and Jonathan Dugas said...

Hi Phil

Sure, sorry we left that out - it's a fine balance between being over-scientific and basically reproducing a journal article here, and writing it in a manner that "compromises" the science of the research...

But yes, the values are statistically significant, with P < 0.001. That's done with a Repeated Measures ANOVA, if that means anything

One of the problems in these physiological trials (apart from the issue of subject number to reduce the SD, as you mention), is that there is often large variability between subjects. The result is that Bob comes in and rides at 400W, whereas Fred rides at 250 W. The SD is enormous as a result, but the reason one can find quite strong significant differences is because the two guys show the same response when moved into the heat. In other words, Bob now rides 350W, Fred 200W - both lower, but still with large between subject variability.

So it does sometimes hinder the chances of finding difference, but it's saved by a very low intra-subject variability and what is a really predictable response of people to the "intervention" - in this study, for example, 11 out of 12 guys were slower in the heat by 10%, one of the 12 performed 2% slower.

Thanks for the question, hope that answers it!