the math behind aerobic decoupling %

[9606]

As I understand it, AC% is (EF(1) - EF(2))/EF(1), and EF is NP/avg HR.

I have been manually calculating my AC% for my interval sessions. If I do a session with 6 x 5 min intervals I might find AC% for the 1st and 6th interval. If I do a 2 hr fairly steady state ride I might calculate AC% between the 1st 45 min and last 45 min 'intervals'.

1. Am I dealing with AC% correctly?

2. Is the design and use of a GC formula (which I imagine might be quite a formula) the only way for GC to show me such data?

[9606]

And 3. What is the math behind GC's AC%?

[Mark Liversedge]

Its a Joe Friel thing, they posted an explainer on the TP website:

http://home.trainingpeaks.com/blog/article/aerobic-endurance-and-decoupling

Mark

[Ale Martinez]

There is a builtin metric available in GC to do this calculation for rides and runs, it is called just "Aerobic Decoupling", you can add it to interval metrics if you like.

It is based on the Friel's article that Mark posted before, if you want to see the source code it is available here: https://github.com/GoldenCheetah/GoldenCheetah/blob/8d03e014ef7e74b29493161e2b03cc644afbbae9/src/AerobicDecoupling.cpp

[Mark Smith]

Am I missing something? Joe Friel's article calculates Decoupling based on EF using NP. If I am reading the GC code correctly it is using EF based on average power. That can make a big difference in the calculation.

For example on the ride I did today GC shows decoupling of 24% but if I manually calculate using the definition used by Joe Friel the decoupling is only 2.5%.

GC already calculates EF based on NP so why not use that in the code for decoupling?

[Steve Mansfield]

IIRC from his book, it also compares the first half of your ride with the second.

[9606]

Does that mean it is not appropriate to use to compare int (x) and int (y)? Or the second 45 min period and the seventh 45 minute period of a 300km brevet?

[9606]

[Mark Smith]

Agreed that in his book it compares first half to second half. GC does that but using average power instead of NP for the EF calculation.
For the standalone EF calculation GC does use NP so there is an inconsistency in GC between the standalone EF and the EF used for the decoupling calculation.

[Mark Liversedge]

On Sunday, 19 June 2016 19:29:21 UTC+1, Mark Smith wrote:

Agreed that in his book it compares first half to second half. GC does that but using average power instead of NP for the EF calculation.
For the standalone EF calculation GC does use NP so there is an inconsistency in GC between the standalone EF and the EF used for the decoupling calculation.

This is because we did not have permission from TrainingPeaks to use NP when aerobic decoupling was added.

We should fix this now.

Mark 

[Mark Smith]

Thanks for the explanation Mark.

[Nathan Townsend]

On Monday, 14 December 2015 20:30:35 UTC+3, Mark Liversedge wrote:

Its a Joe Friel thing, they posted an explainer on the TP website:

http://home.trainingpeaks.com/blog/article/aerobic-endurance-and-decoupling

Mark

Just a bit of background context here.... it's not actually a Joe Friel thing.  What Friel refers to as "aerobic decoupling" the rest of the known universe calls "cardiovascular drift" which is a well known physiological phenomenon the basics of which have been understood for decades.  Friel seems to have made up a catchy sounding technical phrase and along with that he also invented his own version of the underlying physiology instead of sticking with factual scientific evidence.  To his credit though, he does devote a tiny little factual paragraph right at the end of a whole bunch of mumbo jumbo.

So by all means, include a calculation of power versus HR divergence within GC, but wherever possible can't we give credit where it is due and stick with validated science and associated terminology rather than appealing to cheap marketing hype?

Please lets respect the many scientific researchers whom have contributed to understanding this phenomenon and call it by it's real name... cardiovascular drift.  

[SAM BUSH]

Nathan,

Can you please point me towards some published work on Cardio Drift? I'm keen to find out some more info.

Sam 

[Steve Mansfield]

https://en.m.wikipedia.org/wiki/Cardiovascular_drift seems helpful

[Nathan Townsend]

http://www.ncbi.nlm.nih.gov/pubmed/?term=cardiovascular+drift+review+exercise

yields as a starting point.....

Cardiovascular drift during prolonged exercise and the effects of dehydration

http://www.ncbi.nlm.nih.gov/pubmed/9694416

Cardiovascular drift during prolonged exercise: new perspectives

http://www.ncbi.nlm.nih.gov/pubmed/11337829

Cardiovascular drift during heat stress: implications for exercise prescription.

http://www.ncbi.nlm.nih.gov/pubmed/22410803

This study is particularly salient example of why Friel's blog should be viewed with caution.... 

Effect of ambient temperature on cardiovascular drift and maximal oxygen uptake

http://www.ncbi.nlm.nih.gov/pubmed/18461000

[SAM BUSH]

Thankyou. This will definitely get me started.

[Mark Smith]

Thanks for the links Nathan.

I live in Phoenix so I had wondered if heat would have an impact on cardiovascular drift. The articles look like they will be a useful read and helpful to interpreting and using training stats in a hot climate.

[Nathan Townsend]

Hi Guys, to give a little more context, Friel is actually on the right track, but in his blog he just doesn't explain it properly.  He doesn't give an underlying mechanism and secondly, he doesn't really even clearly define what he thinks "aerobic endurance" even means. So it leaves the reader just thinking that "endurance" can be estimated by cardiovascular drift, which can be misleading.

A key adaptation to endurance training is changes in blood volume, which is made up of both plasma volume and red cell mass. When blood volume is greater, it enables better matching between thermoregulation and oxygen delivery to the muscle, because either more blood can be diverted to skin (which cools us down) or a lower proportion of total blood volume needs to go to skin, which thus leaves more for oxygen delivery to muscle. Only the red cells carry oxygen though.  So if your red cell mass increases and you can deliver more oxygen to the muscle, this will tend to improve metabolic stability at higher level of energy demand = higher threshold power.  Increasing red cell mass from endurance training is a relatively slow process so there is a slow improvement in CV drift resulting from the slow increase in RCM. 

However, plasma volume can also be altered with both training and dehydration.  Training (esp in the heat) induces rapid plasma volume expansion.  For example, last year I piloted a heat acclimation protocol in our heat chamber.  In the space of 4-5 days I had huge increases in power for a given HR ie: I had big changes in "aerobic decoupling" however, clearly 4-5 days is not long enough to develop big changes in "aerobic endurance".  During stage racing, we generally always see quite large decrease in Hct which indicates rapid plasma volume expansion.  Plasma volume expansion (or reduction via dehydration) as a dominant effect on CV drift.

What I define as "aerobic endurance" though has more to do with fat versus CHO metabolism.  High volume aerobic base training induces improvements in fat metabolism.  This has a glycogen sparing effect.  Therefore, at a given percentage of threshold power, better "aerobic endurance" implies (at least to me), lower rates of glycogen depletion. This improves the ability to maintain high intensity power after a long day in the saddle.  The shift in substrate metabolism might be reflected in changes in CV drift, but any such effect would be masked by the much larger influence of changes in red cell mass and especially plasma volume. 

Years ago (when I was a grad student at the AIS), Dave Martin and various national (women's) cycling team coaches used a simple approach to estimate "aerobic endurance".... do a 30 min TT in a fresh state one day, and then on another day go out and ride 100km at aerobic base intensity, then immediately after, ride straight to the lab and do another 30 min TT. I personally assisted on data collection on many of these tests (which were implented during dedicated training camps).  If you could match your 30 min TT after 100km in the saddle, then endurance is good.  If not, more base volume is required.  For the majority of recreational (male) competitive cyclists, using international class female test protocols works pretty well, so I think 100km would be appropriate for most people here.  For more elite men (eg: cat 1 level and above), then probably 150 km might be more suitable.  I agree with this basic idea.... if you really want to know if your endurance is solid, then try doing a max effort (I don't think a 30 min TT is necessary, but definitely something in the range 10-15 min would be good) after at least 100 km at Z2 (with some Z3 also) and compare it to a recent PB over the same duration, or indeed the exact same stretch of road or hill.

[Drew Moffitt]

Nathan,

Thanks for the article links above and the information/anecdotes below. I appreciate them very much.

[Randolph Baral]

....sounds like my 20min power PB last week, 2.5hrs into my (already pretty solid) ride and the day after a TTT suggests I have decent 'aerobic endurance' :-)