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.

Friday, May 02, 2008

Fatigue and Exercise Part I B

The pacing strategy continued: Setting the scene with two examples and a summary of models

In yesterday's post, we looked at the admittedly rather obvious question of why a pacing strategy exists. It's a question that is often dismissed as non-sensical and irrelevant with an obvious answer, but hopefully, it's possible to recognize that in fact, this question has profound implications for exercise physiology. First of all, it has no explanation according to the prevailing theory for fatigue. Secondly, it stimulates one to think a little more deeply about why such an obvious thing exists. In otherwords, is it conscious or sub-conscious, or perhaps even "pre-conscious"? Is it simply a function of training and experience? But then, how does training help you understand how to pace yourself? What is the mechanism?

In any event, the next step to take is to delve a little more into the issue, and I thought that two examples of fatigue-models might help to explain some lingering questions. Once again, I must stress that I have no definitive answer here - no one does. But it's a stimulating discussion, and we're working towards a model for fatigue which will be evidence-based (promise!)

So the two examples for today are the "leaky calcium channel" theory and the model of exercise in the heat.

The leaky calcium channel theory

Yesterday's post also introduced a theory that hit the news earlier this year, when scientists discovered that a possible cause for fatigue might be what they called "leaky calcium channels".
The one sentence summary of this theory is that exercised muscle becomes fatigued due to calcium channels which become progressively more and more "leaky", causing the force of contraction to go down. It was a landmark study, and caused much excitement in the field, thanks to a somewhat sensationalized article in the New York Times. (This is, incidentally, a great example of how people will find "silver bullet" explanations for complex issues. Fatigue is not a problem waiting to be "solved", but it's reported this way, for admittedly understandable reasons. We certainly would not claim to have any answer so exciting and definitive!)

Consider for a moment the implications of this theory for endurance exercise - the best example is if you are doing interval training on the track. Let's say you're doing 5 x 800m repeats. If you ran without a watch, and your goal was to run the TOTAL SESSION as hard as possible, and you were running alone (very important - we'll look at how "social factors" influence pacing later in the series!), then I can almost guarantee that your pace will go FAST-SLOWER-SLOWER-SAME-FASTEST for the five repeats. It's the same pattern as we saw yesterday from Haile Gebrselassie and just about any other athlete doing self-paced exercise.

But the leaky calcium theory is saying that the reason you slow down and develop fatigue is that your muscles becomes less and less able to exert their normal force. Now, the key requirement for this to be true, which I hope is obvious to people, is that all the muscle has to be active. Because if there is any muscle that is INACTIVE, then that muscle would surely not be affected by "leaky calcium channels"? The inactive muscle could simply be activated and the pace would be maintained.

"Ah", you say, "but that's not an intelligent pacing strategy!". To which my response is "Yes, but please tell me how the body knows this, when the problem is a leaky calcium channel in the muscle? How is it even remotely possible in this explanation that you can be aware of the fact that your slow down is being caused by a tiny channel in the muscle?" The point is, in this system, there is no allowance for your perception or "intelligence", and surely intelligent pacing requires that you somehow KNOW what is going on with your body?

Therefore, pacing during endurance exercise is also incompatible with the calcium channel theory. Note that this DOES NOT mean calcium channels are not somehow involved, for a I believe they are. I said in the first post of this series, the trick is to balance the extremes, and we'll hopefully manage to integrate all the information moving forward.

So here's the thing - we know that muscle is not 100% active during your 10km race. In fact, even when you do your best to exert maximal force for FIVE SECONDS, there is evidence that you still keep some "reserve" capacity. We know that because if someone is doing a maximal 5 second contraction, and you stimulate the muscle using an electric current, the force can go up, so clearly what the person thought was "everything" was actually still sub-maximal! So a reserve is a universal feature of any voluntary effort, regardless of how hard you try. Now, given that fact, one can appreciate that it is impossible to explain how any biochemical change - lactate, hydrogen, leaky calcium - can force you to slow down, either consciously (intelligent pacing) or unconsciously (acting on the muscle).

Exercise and pacing in the heat: The best comparison between "limitations" and "regulation"

To illustrate this point before we move onto Part II, perhaps the best example of how the pacing strategy comes into play is during exercise in the heat. This is a topic we'll devote an entire week to later on, because the heat is the best example of a "homeostatic failure" model compared to a model for "anticipatory regulation", because it changes the INPUTS, as we spoke about yesterday. Perhaps this example should have been used instead of the "endspurt" question, because it's a lot more logical to work through.

Basically, there are two lines of thought for why exercise performance in the heat is compromised:

The "Peripheral" model for fatigue in the heat - failure causes fatigue

This theory says that fatigue in the heat is the result of a failure to keep the body temperature down. When the body temperature rises, it causes fatigue because the overheated brain is less capable of activating muscle to keep exercise going. This theory was first developed through a series of very novel research studies by scientists in Denmark - Savard, Nybo & Nielsen are the common names, for those who are interested.

Basically, what they have done is find that once the body temperature hits 40 degrees, the athlete:

  1. Stops exercise - fatigue co-incides with this "limiting" temperature, hence the name "critical core temperature hypothesis for fatigue in the heat"
  2. Activates less muscle - muscle must be stimulated to contract, and what Nybo and Nielsen showed is that the activation of muscle by the "hot" brain is lower than that by the "cool" brain after exercise
  3. Has altered brain function - they measured brain waves during exercise and found that certain waves are altered, which suggests "reduced arousal" levels.
Their conclusion? Exercise is impaired in the heat because the body temperature rises until it reaches limiting values. At this point, the brain fails to activate the required muscle, and the athlete can no longer continue exercise.

A couple of key points: Firstly, there are a few details in the explanation of why the brain fails to activate muscle that we'll get into later. However, what is key to realise is that these studies, while excellent and crucial to our understanding of the heat, have failed to recognize that during any form of exercise, it is possible to slow down long before you stop! In other words, because these studies force people to exercise at a fixed power output until exhaustion, the conclusion they make is that fatigue is caused by some "failure". They then extend this finding to say that "impaired performance" is caused by the same thing, when in fact, they don't measure what happens BEFORE the limiting temperature is reached!

An anticipatory regulation model for exercise in the heat

The alternative model is that performance is regulated well before the limiting temperature is reached. For this to be true, it would require that the athlete slow down at sub-maximal body temperatures. And there is evidence for this - Frank Marino from Australia found it in runners, Stephen Cheung of Canada found it for small muscle groups, and I found it a couple of times in the heat in studies that have all been published (If you'd like these references, please let me know - I'm not going to include them in the text because it breaks the flow - there are some below, however)

Thus, the athlete should start slowing down even though the body temperature is not different from that in the cool condition. As a result of slowing down, the athlete would be producing less heat, and so the fall in work rate will ultimately produce body temperatures that are not different to those measured in the cool conditions! In otherwords, you don't slow down BECAUSE you are hot, you slow down in order to prevent yourself from getting hot!

Perhaps most interestingly, we've actually measured that LESS MUSCLE is activated during cycling in the heat than in the cool conditions. This was a study I did in 2004, and it will be discussed in detail later, but the key point was that cyclists in the heat slowed down very early on, when nothing measurable was different, and they did so by activating less muscle. That's completely incompatible with the "peripheral fatigue" model.

This will, I'm sure, be dismissed as obvious by many, but again, the crucial question is HOW is this achieved? To refer to yesterday's post, what are the inputs, how are they interpreted, and what is the output in response? These, and many more questions, are on the way.

Summarizing the models: A platform to move forward

The two diagrams below are concise summaries of the last two posts. The first diagram, directly below, shows what I have called "The homeostatic limitation model".

This model shows the following:
  1. Muscle contraction during exercise is responsible for producing changes, including biochemical ones (leaky channels, fall in pH, lactate, phosphates etc), as well as changes to the cardiovascular system, energy system (glycogen is depleted and blood glucose falls), and thermoregulatory (body temperature rises, as discussed). The figure shows, from top to bottom: A mitochondria, the liver for energy supply, the heart, and body temperature.
  2. These changes DIRECTLY inhibit exercise, either by:
    • Causing the muscle to lose its force generating ability. This is the theory for lactate, phosphates, oxygen supply (the "anaerobic" limit to exercise) and calcium ions; or
    • By acting on the brain to force the muscle activation levels down. This is the case with high body temperatures, as discussed briefly above.
The key to this model is that failure is responsible for fatigue. Something has gone "wrong", either with oxygen delivery, biochemistry, blood supply, or body temperature, and this has impaired the athlete's ability to exercise at the same pace, so they slow down. One very important point I must make is that this model makes absolutely no allowance for pacing - I've said this many times before, but there is no feedback from these systems that would allow the suggested "intelligent" pacing to take place!

The figure below shows the model for "Anticipatory regulation": Just a note to say that this is by no means the "definitive" model for anticipatory regulation (or the central governor, or anything else you may have heard it called!). This is just a summary of the last 3 posts, and as this series moves forward, I'll try to develop this in much, much more detail, until we have something hopefully more comprehensive

In this model, which was introduced yesterday, there are INPUTS and OUTPUTS.

The INPUTS are provided by the heart, liver (or energy supply), body temperature or rate of heat storage, biochemistry, and then, very importantly, the brain itself!

These INPUTS provide what is called afferent feedback to the brain, informing it of the situation. This feedback provides information on things like "How hot is it?", "How much energy is available today?", "What is the pH of the tissues?", "What is the heart doing?", and basically "Is it safe to keep going at this pace?".

In response, the brain integrates all this information, then evaluates it in the context of the exercise bout before enforcing some output on the system. Key to this evaluation is to know how far the athlete has gone, how far they still have to go, and a host of other inputs or "moderators". We'll discuss all these in turn. Regardless, the end result of this process is the OUTPUT - the activation or the inhibition of muscle. This is responsible for controlling the force output of the muscles, and hence the pacing strategy.

It's important to recognize the presence of the brain as an INPUT in this model. That is, the brain informs the brain (pardon the creative licence!) of certain key inputs before and during exercise. These include memory (the hippocampus, presumably, is the part of the brain involved), experience/training, and then crucially, social factors, or social facilitation. One cannot ignore these "conscious" cues that must also impact heavily on performance.

Also bear in mind that what you all know as true is that when you're in a race, you race with tactics and the presence of other athletes. There is no explanation for this, no mechanism, according to a "limitations" model...how would leaky calcium channels be integrated into racing? It can't be, because tactics and "intelligence" requires that the brain be involved...so you might actually be closer to this model than you realise!


So that is the model, which at this stage is just a theory, I admit! In the coming posts, we have to back-track a little, and look at pacing strategies, and then we'll move on to systematically evaluating and discussing each of the various INPUTS I've put forward. Importantly, we must gather the evidence, otherwise, this is all just conjecture! That evidence most definitely exists, and the next phase of this series will be to examine that evidence, logically and thoroughly. I'll be breaking it up into much smaller pieces, however, so don't worry, this is the last "mega-epic post" for a while!

Speaking of theory, we can now attempt to answer that most basic of questions that started this whole series: The "endspurt" is the result of an increase in muscle activation, controlled by the brain in response to numerous INPUTS during exercise. It occurs because the finish line is approaching, and the physiological changes are no longer deemed harmful or potentially limiting to continuing exercise. The reserve can thus be activated!

Stick with this, and hopefull that will evolve from being mere theory to fact!

Have a good weekend!


Some relevant references, for those interested:
  • Savard et al (1988), J Appl Physiol 64: 649 - 657
  • Nybo and Nielsen (2001), J Appl Physiol 91: 1055-1060
  • Nybo and Nielsen (2001), J Apply Physiol 91: 2017-2023
  • Nybo et al. (2002), J Physiol, 545: 697-704
  • Marino et al. (2004), J Appl Physiol, 96: 124-130
  • Marino (2004), Comp Biochem Physiol B Biochem Mol Biol 139, 561-569
  • Tatterson (2000), . J Sci Med Sport 3, 186-193
  • Tucker et al. (2006), J Physiol 574, 905-915
  • Tucker et al. (2004), . Eur J Physiol 448, 422-430


Andy Renfree said...

Another fascinating article. Just one comment - when you discuss the idea of leaky calcium channels you state that, "Because if there is any muscle that is INACTIVE, then that muscle would surely not be affected by "leaky calcium channels"? The inactive muscle could simply be activated and the pace would be maintained."

You seem to be working on the assumption that any additional muscle fibres that are recruited in order to maintain pace when the athlete becomes fatigued are equally as well trained as the muscle fibres originally recruited. It seems likely to me that you slow down because individual muscle fibres become fatigued, thereby forcing you to recruit additional fibres in an attempt to maintain pace which are unable to do so as they are not so well trained.

Unknown said...

Very interesting article again chaps.

My original understanding was that the type I (slowtwitch) muscles that were originally responsible for the bulk of aerobic work slowly fatigued, hence the downward slope from the start. Type IIa and IIb (Fast twitch) fibers would need to be recruited with what was left when the "horse smells the finishing gate", and because these muscles contract more forcefully but far less economically, this has to be held for the end. This would give rise to the finishing kick.

Anyways, I'll leave it to you two mythbusters to blow holes in my preconceived notions...

One interesting point that I wanted to mention is that a top international coach recently posted on his blog with regards to setting a pacing strategy for athletes, and he recommended to plan to spend 51% of the total race time on the first half with 49% on the second. This is based upon his analysis of numerous world records that have apparently been run in this fashion. He also mentions a few other reasons such as the body's ability to deal with muscle acidity at the various stages of the race etc.. Blah blah.

Any thoughts on the 51\49 planned split?


Anonymous said...

I agree with your final comment but even then you are not truly activating your whole reserve (ie. if you electronically stimulated the muscle at the end spurt it would produce more force than the voluntary contraction). So why does the 'central governor' hold back when there is no danger?!

Ross Tucker and Jonathan Dugas said...

Hi guys

Thanks for the comments, I'll respond briefly to each on in turn:

First, to Andy.

Fair point, the assumption is that the muscles activated towards the end are equally capable as those recruited early on. HOwever, there's no evidence or reason to suspect otherwise.

In fact, if you consider the point made by Chris below your comment, then the muscle fibres activated later would be EVEN BETTER, because they'd be the fast twitch fibres!

I deal with Chris' comment in a separate reply, but to get back to yours, there is no reason to suggest that the fibres activated later on would be inferior in any way. Remember, at the end of exercise, when you activate those fibres, running speed or power output increase - that would suggest that these fibres which are activated later on are perfectly capable of doing the same job as the others! So it doesn't really matter that they might be slightly less well trained (the issue of whether you train certain fibres is an interesting debate itself), the point is that because they were previously inactive, the recruitment of these fibres is able to increase work rate, as seen in the final sprint for the line.

If we are talking about a shorter duration (2 minutes or less) contraction, where the person is pretty much giving maximal effort for the whole 2 minutes, then I agree with you - the fibres activated later are probably not as strong. BUt this is not the case for sub-maximal exercise, such as during endurance running or cycling.

I must confess, however, that the issue of fibre type recruitment is a very shaky one, as mentioned in my reply to Chris below this.

Thanks for the comments!

Ross Tucker and Jonathan Dugas said...

HI Chris (or Jean!)

Interesting possibility.

The issue of fibre-type recruitment is a very tricky one. The basic premise (which you've mentioned) is that there is a logical progression of muscle fibre activation, with smaller Type I fibres activated first, and the larger, more powerful Type II fibres activated later.

There are those in the field who believe this is completely incorrect, and there is some confusion about what really happens. It will probably come up again in the future and we'll certainly dig into it a bit more.

As for the 51/49 split, it depends on the event being run. For every event above about 5,000m it's true. For any event below 1500m, it's not true, because in those events, you have to start faster, so it's more like 49/51. The 1500m event seems to fall exactly in the middle - 50/50.

For cycling and speed skating, it's also slightly different, because of the big benefit you get when you have a rolling start.

I actually did a study of all running records from 800m up to 10,000m, part of my PhD, and I'll definitely be including that analysis as a big part of this series on fatigue, because it's very relevant!

Staring with the next post, we'll start looking at what is optimal, and then move onto the physiology, so it's coming!


Ross Tucker and Jonathan Dugas said...

Final response for now, to "anonymous"

There is never zero danger. The reason the reserve is maintained even during very short (less than a few seconds) efforts is because it is protecting some mechanical component. It's not just biochemistry that must be regulated. So for example, if the muscle is over-stimulated, there is a better than average chance that the muscle and tendon would be placed under such great loads that you would rupture them completely. Therefore, one of the big "regulators" that prevents sprinters from going faster is a "MECHANICAL" type of feedback, which alerts the brain that excessively high force outputs might actually damage the body structurally.

In this sense, one of the inputs also comes from mechanoreceptors and proprioceptors, which inform the brain of loading, loading rates and position in space.

So that, in my opinion, is why you never fully activate muscle - it is possible (think of super-human feats of strength in emergency situations), but dangerous to do so.

So you caught me out - I should have said "allows the body to access MOST of the reserve!"


Anonymous said...

Thanks for the (very prompt) response. I look forward to the follow-up posts on the issue then.

I'll be considering the 51\49 splits with a bit more credibility then, although they won't be applicable to me on too many occasions other than a rare pan-flat half marathon with zero tactical intent.

As a matter of interest, where does one go to dig up lap splits for every world record from 800m to 10000m?

Jean (Chris)

Andy Renfree said...

Thanks for the quick response to my post.
I am going to respectfully disagree with you on this, although again I am guessing what is really happening.

My hunch would be that the additional fibers you recruit later in exercise will be less well trained, simply because of the fact that they are only recruited when you are already very fatigued. This would mean that it is only possible to train these fibres if you fatigue the preferentially recruited fibres first. If you go on a two hour run then you will only be training these fibres during the final 30 minutes or so when you are forced to recruit them to maintain pace as a result of glycogen depletion in the other fibres. This means that it is simply not possible to subject these fibres to the same total training volume as those that are recruited earlier in exercise.
With regards to the fast twitch fibres you activate later in exercise being BETTER, I suspect this depends on when fatigue occurs during a race. If it is relatively late, then I can believe that these fast twitch fibres can produce a higher power output for a short period of time allowing an end spurt. However, if you need to recruit them earlier on in the race then you are probably in trouble as the high fatiguability of these fibres means that they will be unable to sustain running speed for any significant period of time.

thanks for the debate

Andrew said...

If it takes you months to finish this topic, I certainly won't complain - this is fun stuff!

I really like your Anticipatory Regulation diagram, but I noticed that it's missing a couple arrows:

1) An arrow going from the leg back into all of the "inputs" since whatever the brain commands the muscles to do, it's going to affect those inputs.

2) An output arrow from the "top brain" to the "input brain" since as soon as a decision is made, it becomes direct feedback for future decisions.

I assume you left them out for simplicity, but they are why such systems are so difficult to analyze. When a regulator's outputs influence the inputs and are direct inputs themselves, it's a very tricky system to balance.

Ross Tucker and Jonathan Dugas said...

Great discussion to all, thanks to everyone for joining in here. . .

Just a note on the muscle fiber issue---we must be careful of narrowing down the issues too much. The recruitment of a large muscle mass such as that in the lower limbs is a bit more dynamic than simply activating or not activating Type I or Type II fibers.

I do not think The EMG frequency data are robust enough to describe with 100% confidence the nature of the muscle (fiber) recruitment.

It is likely that at any given time during sub-maximal exercise Type I, IIa, and IIb fibers are being activated at the same time, and it is not just a matter of switching off Type I fibers and activating Type II fibers for an end spurt.

Kind Regards,

Anonymous said...

Ross and Jonathan,

From what I understand the Anticipatory Regulation model acts like a race driver short on fuel. He drives "defensively" in order to be able to keep the car going, but when he sees the finish line he realizes that there is enough petrol left in the tank to cross the finish line and floors the accelerator pedal. Of course his actions cannot push the engine past the red zone of the tachometer otherwise the rev limiter (or governor in diesel engines) intervenes.

Keep the series going, I am thrilled!

Ross Tucker and Jonathan Dugas said...

hi Andy

I agree with Jonathan on the muscle fibre type issue. There is no evidence at all that fibers that are activated later are "weaker", and I stand by the evidence that power output can increase at the end of exercise thanks to that increase in muscle recruitment, which suggests that those fibres are still capable of increasing power output. They may be less well-trained, but that is guesswork on both our parts. Certainly, however, they have the capacity to increase work rate, which one would think impossible if they were "weak".

That evidence will come out in the series, in the next couple of weeks, where I'll look at those studies that have shown how power output and muscle activation levels are related. It's not a linear relationships, so in addition to what Jonathan has said about the EMG data being unable to distinguish between fibre types, the interpretation of EMG data is something of a point of contention.

However, in time, we will discuss it, and then hopefully this issue will come up again with more clarity.


Ross Tucker and Jonathan Dugas said...

Hi Colenso

Thanks for the comments. Without going through them one by one, I will say that many, perhaps five or six, of your questions will be answered in the series as we progress. I will, for example, spend a great deal of time on the extrinsic motivators - they are in fact already included in the basic model I've put across in this post. Also, questions 1, 2, 3, 4, 7, 8 and 10 all have explanations, and as I develop the series further, they'll be there. So while not wanting to skip over your questions, I'd ask that they be held back until the very end of the series, when (hopefully), they've been covered in more detail!


Ross Tucker and Jonathan Dugas said...

Hi Andrew

Thanks, you're quite right that the diagram is missing a couple of arrows. But this is really just an overview model, and right at the end of the series, I'll present a complete model for Anticipatory Regulation, which includes everything we'll cover in the next five or six weeks.

But you are correct, both those arrows should be there, along with an infinite number of others we don't even know about!


Anonymous said...

The endspurt is controlled by the brain, yes, but also by muscle physiology. Besides heat does not the accumulation of hydrogen ions affect the ability of muscles to contract even during an endspurt?
Once a tipping point is reached, the muscle shuts down, sub-maximally speaking, whether or not other inputs (will power, pain tolerance, more glucogen, calcium channels, etc) are factored in.
The interesting point to me is: can that tippng point be changed by training? I suspect yes.
And more interesting still: does the tipping point vary day to day and what factors influence it?
--Mike McGrath

Ross Tucker and Jonathan Dugas said...

hI MIke

You're almost certainly partly correct. I'm not sure of the hydrogen ions angle, because no study has ever found the cause of "peripheral fatigue" in human muscle while still in the body (if you take the muscle out, chemicals can affect it). Hydrogen is one of the options, though.

I believe, as you'll discover over the course of the series, that performance is regulated somewhere between the two extremes of the theories - not solely by the brain, but most definitely not in the muscle, as the textbooks inform us.

So perhaps the best description for it would be "an endspurt controlled by the brain, but MODERATED by biochemical factors in the muscle". Then, depending on the circumstances, you find different weightings, different interactions between these systems, and also the nature of the inputs changes.

As for a tipping point, in theory, yes, but I reckon it will be generations before we can establish where it is and explain it on a day to day basis, it's just too complex.

But muscle never "shuts down". There is no evidence whatsoever that muscle, even at its most fatigued, is completely incapable of contracting to generate force. That's why I referred earlier to MODERATION and not limitation. Peripheral factors may moderate muscle function, but it's never "poisoned" or shut down beyond all function.

But ultimately, fatigue is entirely contextual - it depends a great deal on the circumstances under which it is being examined, and that is the topic of the next post in this series, for Monday!


Unknown said...

Hey guys, very interesting, maybe I can figure out why my body shuts down at mile 22.