Tuesday, October 6, 2009

The Science of Decoupling

Those of you familiar with the training philosophies of Joe Friel (the guy decoupling big time in the shot above :-) will have no doubt come across the concept of ‘decoupling’, i.e. a shift in the power: heart rate relationship as a workout goes on.

An example of this, from one of the athletes I work with, in the form of a rise in heart rate and a drop in power as the session progresses is shown below.

Clearly, as time went on the gap between the athlete’s power and heart rate widened, to the point that by the end of the session, the difference in power:HR compared to the start is 26%. Or in other words, it is taking this athlete an extra 30 beats/min to generate the same power!!

Detailed info on the calculation of decoupling can be found here, but the general gist is; we take the power/heart rate for the first half of the session and divide it by the power/heart rate for the second half. E.g. if that athlete did 105 watts at 100bpm in the first half (power/HR = 1.05) and 100 watts at 100bpm in the second, i.e. he lost 5 watts (power/HR = 1.00), then his decoupling would be 5watts/100watts = 5%.

When you think about it, this is a pretty perplexing phenomena. We assume physiologically that a given effort requires a given amount of energy, which requires a given amount of oxygen, which in turn requires a given amount of heart beats, at least for a particular individual! So what are the causes and implications of a need for more heart beats at the same workload?

To illustrate, let’s start with a typical ex phys scenario:

Say that I start pedaling a bike at 260W, a level of power that on average requires approximately 3.5 L/min of Oxygen. As I start the exercise & my muscles figure out “we’re gonna need more O2 captain”, my body goes to work transporting O2 to the working muscles.

Let’s assume that I have 12g of hemoglobin per deciliter of blood (an average amount) Assuming 100% saturation, this 12g/deciliter carries 16ml of O2, so 160ml of O2 per liter of blood. But I need 3.5 L of O2, so it’s going to take me about 22 liters of blood per minute to keep up with the demand (3500/160). Assuming I have a cardiac stroke volume of 150ml, it will take my body 150 beats per minute to pump these 22 liters (to the ex-phys geeks, yes I’m ignoring the a-VO2 difference for the purpose of simplicity).

Pretty simple, eh? A given workload requires a given O2, which requires a given amount of heart beats. So, if the workload stays constant but the heart rate changes over time, what’s going on? At what step in the chain is the breakdown occurring?

The obvious one and the most commonly cited cause of increased heart rate for a given power is a change in stroke volume due to dehydration. If my cardiac stroke volume all of a sudden goes from 150ml down to 140ml my heart would need to beat 10 beats faster in order to get the same amount of blood per minute to the muscles. So, for my 260W, I would now be putting out 160bpm instead of 150bpm. The most common cause of this drop in stroke volume is a drop in blood volume via dehydration. For this reason, cardiovascular drift frequently occurs under hot conditions where some of the body’s fluids must be devoted to cooling rather than maintaining the integrity of the blood volume.

However, can an increase in heart rate for a given power reveal more?
Joe suggests that not only is decoupling of power and heart rate a sign of heat stress, he also uses it as an indicator of aerobic fitness. Is there a possible mechanism by which this metric could be used as a sign of not just heat tolerance, but also aerobic endurance?

Thomas and Chapman (2006) may be able to help answer the question of the validity of decoupling as a training metric. By observing VO2 during prolonged downhill walking on a steep grade, they saw a progressive rise in VO2 uptake with no change in body temperature or stroke volume. OK, you say, “the sweat thing made sense but what’s going on here?”

The break in the chain under these conditions occurs not in Oxygen transport, but Oxygen demand, i.e. at the top of the chain. During the eccentric exercise, as muscle damage occurs, the legs are forced to recruit larger, less economical muscle fibers. These fibers require a greater amount of O2 to exert a given level of power and the heart rate goes up for a given power output when the more economical fibers begin to fatigue.

In fact, type II fibers require ~twice the O2 for a given power output. Therefore, small fiber shifts result in relatively large differences in heart rate for a given power output (Coyle, 1992)

As we know, muscle damage isn’t the only cause of muscle fatigue. When a muscle fiber runs out of fuel (glycogen) it’s out of the game. Thus, decoupling can serve as an indicator of how our targeted muscle fibers are doing, both in terms of muscle damage and fuel stores.

As the targeted muscle fibers become stronger and more fatigue resistant, the time before the muscle fatigues to the point that it needs to call on it’s ‘big brother’ fibers increases. Therefore, as an athlete’s muscle fibers become more trained, decoupling over a training session decreases. In fact, the researchers above found that the effect disappeared when athletes were trained in downhill walking for a period of weeks. Or, in other words, as fitness for a given task increases, decoupling decreases.

Additionally, if we accept that HR:Power can indicate muscle damage and fuel depletion, we can also then use this metric to help determine if an athlete is adequately recovered for a key workout. If we know that typically an athlete takes 140bpm to run 7:00/mi (after warm up) we can use this number as a ‘check-in’ before key sessions. If the athlete takes 147bpm for the same pace (a difference of 5%) it may suggest that recovery is incomplete and the session should be postponed. Chuckie V wrote a great post on the practical implementation of this concept here.

Incidentally, a swing in the opposite direction can also indicate incomplete recovery via other mechanisms. In fact, over-reaching studies have typically found either decreased power/pace of ~5% for a given effort (e.g. Coutts et al. 2007, Jeukendrup et al., 1992), OR a decreased heart rate of 5% for a given power/pace (e.g. Hedelin et al, 2000). Therefore, ensuring athletes are within +/-5% of ‘normal’ power and HR is a good policy.

While it’s true that heart rate is subject to more confounding variables than other measures, it is not, as some coaches would suggest ‘useless’ as a training metric. The confounding variables can quite easily be accounted for by a good coach with effective communication and athlete knowledge. When used with a given athlete over a period of time, observing power:heart rate relationships offers the coach a fairly objective indicator of both the athlete’s base fitness and their readiness to work (two things that athletes notoriously over-estimate when left to their own devices). For this reason, in my opinion, decoupling is a key concept of science-based coaching.

Train Smart.



Chuckie V said...

Great stuff AC...

Prior to a given "quality" workout I incorporate what I call "ramp readiness tests". If the athlete fails to fall within what's considered a normal range (HR compared to power/pace, noting PE) or is simply feeling like crap (in spite of the numbers), we put the workout off a day. Actually, we "go-thru-the-motions" if it's an Ironman they're training for, and generally stick with the duration. But if it's a half-Ironman or an Oly-distance event, we cut the workout down drastically and focus on recovery straight afterward. Despite its relative simplicity I've had much success with it over the years. That same 5% figure is where coach takes note! See here, if you're interested...


Now I can link back to your blog somewhere within this one. (Readers might wonder how I can link to future blog write-ups of others!)

But about the theme of your blog: this decoupling awareness should be far more critical to the Ironman athlete than their FTP or 10K PB, but it seems most still fail to recognize this. Your blog should help!


Paul Fleuren said...

G’day Alan,

I use a similar approach with my weekly tempo run and long run. Decoupling is great way to track fitness and more importantly I use it as a benchmark to know when I’m ready to do more.
For example, I do an out and back 10km which has 1km markers along the way. I set my HR at 150-155bpm (zone 3) and after a good warm up off I go. I give it 1-2km to reach the HR target. I take my splits every km. Afterwards I compare my splits looking for decoupling in terms of pace. I want the HR line to be nice and flat BUT I do want to see some decoupling in terms of pace. If there is I know that is the benchmark I have set in terms of my weekly tempo run and I will stick with that run until the decoupling disappears. Once that occurs I know my body has adapted to the overload (I define the overload by assessing the amount of decoupling occurring) and I’m ready to do more. So I’ll move the run up to 12-14km. Then the following week I should see the decoupling effect occurring again between the 10-14km mark of the run. If not I’ll either increase the run further or quicken the pace.
Whenever you do set up BT session I feel there should be some elements of decoupling occurring with the goal of reducing the decoupling over a following weeks. This tells you that your body is overloading. If there is no decoupling occurring in your BT sessions then they are more likely maintenance sessions.
More often than not when I do this weekly tempo run my RPE used to drop well before there were signs of reduced decoupling. You need to be weary of this as I often made the mistake of doing more based on my RPE feeling easier without realising my body had not actually adapted to the session enough (ie not change in decoupling).
Quite often the guys and girls I coach will want to extend their sessions out further in terms of volume or intensity based on just RPE dropping. I’ll set up a test set with them first and if there is evidence of decoupling occurring in that session, I will generally hold them back for a few more weeks, because in most cases they want to do more based on just a drop in RPE. The drop in RPE is the first indicator of increased fitness followed a reduced decoupling effect as an adaptation to the increased levels of fitness.
It varies too how quickly someone adapts to a particular training session. Some weeks it may take just 4 weeks to see a reduction in decoupling while other times it could take up to 10-12 weeks with the change in session structure. Some people adapt quickly to the changes in volume while other people adapt quickly to the changes in intensity. The important thing to remember is knowing when to change and tracking decoupling can help you decide that.


Sven Barnes said...

Yes AC, a tired athlete could show an abnormal HR:Pace profile, meaning the session should be posponed. How many athletes would actually pospone a session? I don't know many. I agree this is a good way of guaging fatigue, but one must have the "head" to know when to stop. And isn't fatigue, at a safe level, just a sign of adaptation?

Sam, Australia.

Alan Couzens said...

Hey CV,

Great blog entry on the 'ramp readiness test'. You beat me to the punch. Great concept. I'll throw a link up. Though, as you mention, not nearly as cool as linking to a blog that is yet to be published. Who knew blogger could warp the time-space continuum :-)

Totally agree with your last point. Iron-fit athletes have 'rail track' power:HR charts irrespective of duration, heat etc. Training with this goal in mind (as opposed to FTP etc) for as long and until it takes to establish this relationship should be a high training priority.

Thanks for the comment.



Alan Couzens said...

Hey Paul,

Great comment. I agree 100% that eliciting decoupling in one form or another is one of the fundmental goals of training.

I follow the same approach with my athletes with one exception, particularly for more advanced athletes. When 'long term' decoupling is established I back-off the load with my more advanced guys in an effort to hasten the adaptation process. For novices or in advanced in the early season, I do what you suggest and rely on in-week recovery to elicit supercompensation from week to week. When the plateaus start getting longer, it's time to start the unloading weeks.

Thanks for the comment. While this concept isn't talked about a whole lot, it seems there are a lot of very good coaches using it on a regular basis.



Alan Couzens said...

Hey Sam,

Having 'the head' to stop is a big issue. That's where having a number guideline really helps. That said, RPE is still important.

For a relatively fit athlete, the 5% pace/power drop mark tends to coincide pretty well with an increase in RPE. At this point, a lot of athletes, particularly those without pace/power monitors will decrease the intensity in an effort to hold the RPE, often unconsciously. This is when training becomes 'dumb' and the athlete is getting tired rather than getting 'better'.

To be sure, fatigue AT A SAFE LEVEL is necessary. For mine, that is ~a 5%pace/power drop. However, while fatigue is necessary for adaptation it doesn't alone represent it. The true test of adaptation is getting tired (~5%), resting and then coming back at a higher performance level. This is the crux of effective training across all cycles.

Great comment.



runtilyoudrop said...

Nice post. One aspect not really explored is how much your stroke volume will vary during exercise and is affected by preious sessions. Since your cardiac output is a combination of both your heart rate and how much you pump with each heart beat (the stroke volume) it is clear that the stroke volume is important in these calculations.

Stroke volume is related to the preload (how much blood enters the heart chambers between contractions) therefore anything that reduces preload (such as dehydration) will impair stroke volume and thus explains the decoupling.

Perhaps of more interest and relevance is the role of hormones that affect the contractility of the heart (ionotropes) such as adrenaline and endogenous steroids. During prolonged exercise, or when experiencing fatigue following previous exercise, production of these endogenous ionotropes is impaired. This can lead to decoupling by reducing the stroke volume and therefore cardiac output for a given heart rate.

Alan Couzens said...

Thanks Runtilyoudrop.

Interesting take.

From my perspective, positive inotropes have a greater effect on HR than stroke volume in terms of mediating contractility, e.g. epinephrine or dopamine vs the opposite effect of beta blockers i.e. decreased HR.

Still, to some extent the logic holds. With increased catecholamine secretion during extended, stressful exercise, relative HR will increase in response to increasing stress rather than increasing workload and a similar effect is likely to occur independent of hydration.

Thanks for the food for thought.



AB@OZ said...

Hi Alan,

so what does it mean when you have a negative decoupling. Can this be a bad thing, ie showing excess fatigue etc? I am regularly seeing this on steady home trainer sessions of 90-180min duration, especially when cool down period is excluded from selected data. Varies from -1 up to -9%.
Just interested if this is a marker that I should be aware of.
Thanks AB

Alan Couzens said...

Hey AB,

Yes, a swing in either direction can be indiatie of tiredness. While there is still some argument over the exact mechanism a drop in HR for a given power output with endurance overtraining is a common phenomena. For this reason, keeping HR within a range for the bulk of sessions (not too high or too low) is a good plan.