Frédéric Grappe analysis of Froome data

[Fearless Greg Lemond]

We have this :
http://3.bp.blogspot.com/-QlklucnWg...AAARA/5yfDjR-GNkY/s1600/froome_ITT_vuelta.jpg

We should all copy it before it disappears from the web.

Are people really to believe the Froomster rode the TT of his life, beating Wiggins/Cancellara, with that heart rate?

Also, 405W divided by 5.8W/k makes 69.82 kg for the Froomster that day. Yet Grappe has stated the Froomsters weight was always steady around 68 kg with a 900 gram margin?

Looks like Sky even alter SRM files to fool the grand public? And let them dissapear from the interwebz of course. Just as shady as Katusha hiding their medical staff from their website all of a sudden.

 

[meat puppet]

HR's are highly individual. moreover, HR just measures how many beats the heart makes per minute. nothing more, nothing less. without a good set of power and heartrate data and coupling them, a lone HR figure is basically useless. if we could plot its drop, then it would be at least something

also, it is not unheard of that a top athlete has their max around 170. if i remember correctly, dave zabriskie has his max way into the 220-230s. if froome's threshold HR, lets say when fatigued, is around 145 then so be it. apart from the casual observation that it is quite low, this datapoint indicates nothing either way.

in the vuelta mountain stage dataset posted here froome had 122 avg for the whole stage, peaking at 165 (presumably at the end).

power profiling i'm all for. heart rate estimations go right into the tinfoil area. not everyone has a "normal" BPM. and it's quite clear that we cannot rely on anything "normal" with froome.

 

[The Hitch]

I remember how outraged I was when I read his LA apology in l'Equipe in those days.
I guess he was on LA's payroll as an advisor of some sort.

Anyway, Brailsford would never have shown the data to Vayer, He is much safer with Grappe who seems to always have had trouble with numbers.

Would you by any chance remember where Froome's power profile was shown?.
Either on this website or on a link I found on cyclingnews. I printed it but can't find it.

We have this :
http://3.bp.blogspot.com/-QlklucnWg...AAARA/5yfDjR-GNkY/s1600/froome_ITT_vuelta.jpg

We should all copy it before it disappears from the web.

What is the source of that?

 

[Brooks Fahey Baldwin]

I've had days (TTs) like that as well, yet my power has been unimpaired. That doesn't necessarily mean that cardiac output is reduced, however, since SV may very well increase to compensate.

acoggan, you'll have to forgive my lack of expertise with the details of heart rate and power decoupling. However I struggle to understand your hypothesis that the stroke volume would increase significantly.

In trained athletes, stroke volume increases from around 105ml at rest to 163ml at maximal exercise (a 55% increase). This is compared to a heart rate increases that go from 50 odd bpm to 185 or so bpm (270%) increase.

Furthermore, heart rate and cardiac output are governed by the Frank-Starling law and the Bainbridge reflex. As cardiac output increases, more blood returns to right atrium, stretching the walls of the heart chambers and causing the heart to reflexly contract and eject blood in to the systemic or pulmonary circulation.
All the while, this increased stretch causes increased activation of baroreceptors in the sinus node and causes an increase in heart rate. Then of course stretching of the right atrium also causes the upregulation of the Bainbridge reflex which also increases heart rate.

So any significant increase in stroke volume that is not accompanied by HR rises would have to overcome these reflexes. I suppose its possible via vagal stimulation though? Perhaps due to fatigue...

However for cardiac output to increase significantly enough to supply someone at maximal exercise without heart rate going up much....?
After all, cardiac output and oxygen consumption are dirrectly related. I also assume someone riding at a high power output in a TT will also require significant oxygen consumption, no matter how far you are in to a grand tour. Correct me if I am wrong here.

So if we just assume Chris Froome is an average elite athlete for a moment (by that I mean resting HR of 50 and max of 185) and then we can see what is required for a high cardiac output (lets say 30L per minute) with a low heart rate.

At rest, we can say he is doing 105ml SV X 50bpm which = 5.25L per minute.

At maximal exercise we would have a scenario of 163ml SV X 185bpm = 30.15L per minute.

If he were to have a maximal cardiac output of 30L and average HR of 150bpm, stroke volume would need to be a significant 200ml. Thats a huge variation from a resting stroke volume. While evidence shows that changes in heart rate are responsible for 75% of the increase in oxygen uptake from rest to maximal exertion.

Personally, on days where my HR doesnt get up to the >170bpm range in time trials, I have rather poor power. Conversely, my TTs riden at >190bpm range are usually very good performances. Of course the plural of anecdote is 'no data' though.

I'm not saying its impossible at all, but for your hypothesis to be correct; an increadible variation in stroke volume would need to occur along with the suppression of the bainbridge reflex. I'm not saying you're wrong, I just find this... shall we call it "inverse cardiovascular drift" unlikely.

Or am I not seing the forrest through the trees?

 

[vrusimov]

Do we know if the data actually contain 20-min maximal and 60-min maximal efforts?

All we know is that a comparison between the highest 20-min and 60-min power was made from the data supplied, but we don't know if either or both are actually maximal efforts over those durations. We'd need additional context to make sense of it.

e.g. if a race deciding climb takes 40-minutes, you'll get neither a rider's 20-min nor 60-min maximal power capability from that day's file. Hence the ratio between them may not actually be based on true maximal capability for both durations.

The Salamanca file shows either a non-maximal effort or some weird CV physiology (stroke volume, EDV, ESV, venous return, ejection fraction). His heart rate is super low [~141-143bpm] for what looks like 6-8 min in the first part of the trial, yet he is still putting out 410-420W. I don't think that fatigue alone can explain such a low rate because his power is quite high...he is on basically a 60 minute effort and will average 405W for the duration. Again, weird!

Just watched 5 minutes of highlights from Stage 15 and all I can say is that I have a very difficult time believing that Froome is clean. I've seen all of Armstrong's attacks, including the scintillating ones on Ullrich in 2001 and I am at a loss to explain the cadence, the recovery and the rope-a-dope techniques that Froome uses. THIS MAN ACTS VERY WELL! I noticed this in last years tour but thought it was an aberration. Now I see it again and it seems to be a pattern. The bob and weave, feigning difficulty and then an explosion of power. Like I figured, he toyed with Quintana rather mercilessly. I don't know what else to say but my eyes can't decipher this man without some serious misgivings and suspicions. Like I figured after watching Tirreno-Adriatico, this man was hiding something vicious and it shows.

 

[Fearless Greg Lemond]

HR's are highly individual. moreover, HR just measures how many beats the heart makes per minute. nothing more, nothing less. without a good set of power and heartrate data and coupling them, a lone HR figure is basically useless. if we could plot its drop, then it would be at least something.

Bike pure would have a nice dataset we might guess?

 

[Cobblestones]

German newspaper calling it PR stunt

For Grappe there's a 'physiological limit' while acoggan maintains that no physiological parameter can be construed to indicate doping.

In fact Grappe's analysis is useful since it 'predicts' that Froome has to have a VO2max at the physiological limit. Now, that would be easy to test, wouldn't it?

Actually it's even simpler. VO2max doesn't change all that much; it's notoriously difficult to improve with training. So let's just go back in time and look at all the manifestations of such an extraordinary VO2max in Froome's career. Aww shucks, there aren't any. You do the math...

 

[Alex Simmons/RST]

...and what % difference would you expect between the two durations?

That depends on information which we simply don't have.

Still doesn't make much sense. Say Grappe is handed files from a 10 min climb a 30 and a 40 min one. He takes a wild guess as to what the 20 min and 60 min power are, gets an improbable difference, then issues a statement saying "yup, normal".

Well that's my point. We don't actually know, hence attempting to draw such conclusions from one snippet of reported information without any context about what it means is pointless.

e.g. 60W might indeed be right on the money, who knows, we might be comparing a somewhat fatigued hour in an aggressive TT position or perhaps when not really going au bloc versus a 20-min smashfest ascent with fresh legs.

Without actual context, we just can't say.

Does everyone here put out exactly same maximal effort power on different days, while carrying different levels of fatigue, on flats versus climbs, on TT rig versus road bike, on a stinking hot day versus a cool one etc etc?

 

[meat puppet]

Bike pure would have a nice dataset we might guess?

dont know about that, but in either case it would be nice to glance thru the data

re: vo2max it is my understanding that the room for development among top athletes is next to nothing but us mortals can up it some 20% via proper training (from the starting point, that is).

 

[vrusimov]

acoggan, you'll have to forgive my lack of expertise with the details of heart rate and power decoupling. However I struggle to understand your hypothesis that the stroke volume would increase significantly.

In trained athletes, stroke volume increases from around 105ml at rest to 163ml at maximal exercise (a 55% increase). This is compared to a heart rate increases that go from 50 odd bpm to 185 or so bpm (270%) increase.

Furthermore, heart rate and cardiac output are governed by the Frank-Starling law and the Bainbridge reflex. As cardiac output increases, more blood returns to right atrium, stretching the walls of the heart chambers and causing the heart to reflexly contract and eject blood in to the systemic or pulmonary circulation.
All the while, this increased stretch causes increased activation of baroreceptors in the sinus node and causes an increase in heart rate. Then of course stretching of the right atrium also causes the upregulation of the Bainbridge reflex which also increases heart rate.

So any significant increase in stroke volume that is not accompanied by HR rises would have to overcome these reflexes. I suppose its possible via vagal stimulation though? Perhaps due to fatigue...

However for cardiac output to increase significantly enough to supply someone at maximal exercise without heart rate going up much....?
After all, cardiac output and oxygen consumption are dirrectly related. I also assume someone riding at a high power output in a TT will also require significant oxygen consumption, no matter how far you are in to a grand tour. Correct me if I am wrong here.

So if we just assume Chris Froome is an average elite athlete for a moment (by that I mean resting HR of 50 and max of 185) and then we can see what is required for a high cardiac output (lets say 30L per minute) with a low heart rate.

At rest, we can say he is doing 105ml SV X 50bpm which = 5.25L per minute.

At maximal exercise we would have a scenario of 163ml SV X 185bpm = 30.15L per minute.

If he were to have a maximal cardiac output of 30L and average HR of 150bpm, stroke volume would need to be a significant 200ml. Thats a huge variation from a resting stroke volume. While evidence shows that changes in heart rate are responsible for 75% of the increase in oxygen uptake from rest to maximal exertion.

Personally, on days where my HR doesnt get up to the >170bpm range in time trials, I have rather poor power. Conversely, my TTs riden at >190bpm range are usually very good performances. Of course the plural of anecdote is 'no data' though.

I'm not saying its impossible at all, but for your hypothesis to be correct; an increadible variation in stroke volume would need to occur along with the suppression of the bainbridge reflex. I'm not saying you're wrong, I just find this... shall we call it "inverse cardiovascular drift" unlikely.

Or am I not seing the forrest through the trees?

I would think that Froome would have a RHR lower than 50. Mine is 40 bpm at rest so I'm almost certain that Froome's should be under that. I would guess mid 30's at a minimum. I'm nearly at his physique [5'10 68kg] if that means anything. I would also think that sarcomere length would prohibit any significant increase in compensatory stroke volume [contraction force]. It may be that he simply sandbagged it.

 

[acoggan]

I believe this must be the link I had found concerning Froome's power-duration curve (which I called power profile yesterday), published I believe soon after the Vuelta 2011.

http://teamsky.cyclingnews.com/wp-content/uploads/2011/09/VueltaPowerDistFroome2011.png

I has disappeared from the web.

Maybe some web forensic expert can retrieve it before Brailsford intervenes further?

I have no doubt that Andy Coggan has that in his myriads of data files, but I don't count on him to reveal it to us, he would say it's privileged information he is not allowed to reveal

Sorry, but Team Sky hasn't shared any data with me, and I don't know any more than anyone else here re. his power-duration relationship.

 

[acoggan]

trainingpeaks-analysis-froome-powers-up-sierra-nevada-to-support-wiggins/

http://web.archive.org/web/20120824085112/http://www.trainingpeaks.com/av/OGWG6HAU5XQLNINSED3OOFS55Q

"Load failed"?

 

[meat puppet]

"Load failed"?

worked just fine an hour or so ago, but it was without graphics tho. pretty usual-looking numbers for a mountain stage, really.

 

[acoggan]

1) I thought Froome's weight hadn't changed in 2 years?

According to Grappe, yes.

2) the TT was 10 days into the Vuelta. And I thought that Froome didn't show that decline according to the guy who just looked at his power data?

<shrug> All I'm saying is that it isn't necessarily representative of what he can (could generate) when fresh and when the chips are down.

you are correct though, the FTP entered into the software definitely shows a higher number than 405. Without Sky releasing actual data, I can only look at what he did though, not what they say he can do.

Then you're being rather selective, as you're claiming a 10% increase based on what Hunter says he can now do.

 

[acoggan]

405W divided by 5.8W/k makes 69.82 kg for the Froomster that day. Yet Grappe has stated the Froomsters weight was always steady around 68 kg with a 900 gram margin?

1. I thought Grappe only analyzed 18 files? Perhaps this wasn't one of those provided to him?

2. W/ only two significant digits for power/mass, the proper conclusion is that his mass as entered into the software that day was +/- 1 kg on either side of 70 kg. IOW, there's some rounding error in getting to 5.8 W/kg, which can't be undone by reversing the calculation.

 

[acoggan]

End Diastolic Volume is limited by the pericardium. Cutting the pericardium increases EDV and SV at rest. I am uncertain of the effects during exercise.

It's true during exercise as well (as those in dog-racing circles have long recognized).

I know AC hypothesizes increases in efficiency. I'll admit that I am not up to date on the latest research, but I am pretty skeptical of this concept explaining significant changes in aerobic performance.

Just pointing out the mathematical necessities.

AC: I think that there is something interesting in Froome's hemodynamics. His SV has got to be significantly different that the rest of the pro peloton. It has to be with HR that low. Again I am not suggesting doping, but mearly an interesting study.

At that level, they're all well outside the norms.

 

[Alex Simmons/RST]

1. I thought Grappe only analyzed 18 files? Perhaps this wasn't one of those provided to him?

That's what was reported by Velonews.

And if so people are trying to infer the 60-min : 20-min mean maximal power ratio from power data covering only 18 climbs of unknown duration over a 2 year period.

Right.

 

[vrusimov]

According to Grappe, yes.



<shrug> All I'm saying is that it isn't necessarily representative of what he can (could generate) when fresh and when the chips are down.



Then you're being rather selective, as you're claiming a 10% increase based on what Hunter says he can now do.

After what I'm seeing, reading and Grappe's opinion of Froome being "at the physiological limits of known science" where VO2 max is concerned, I think it might be reasonable to assume a 440-445W FTP, given what that would mean for a "theoretical" VO2 max at the "limits", which undoubtedly insinuates a figure approaching 100. Allen's 6.5w/kg pronouncements probably confirm this. But in an era of "slower times representing clean performances" (and at the risk of seeming conspiratorial) it seems odd that Allen and Brailsford are setting the parameters for the future performances of Froome as the new normal. The last time that someone averaged 6.5w/kg for an entire Tour, his name was Bjarne Riis (per Ross Tucker).

 

[thehog]

Bike pure would have a nice dataset we might guess?

They would indeed!

We contacted Team Sky’s Chris Froome on this issue some weeks ago, asking if he would be willing to produce his data during or after the Tour. We didn’t receive a direct response however he did pass the email to his team. We received a phone call on 20th June from Fran Millar, Head of Business operations for Team Sky who said Froome wouldn’t be making any of his data public. Bike Pure also asked if Froome would be willing to make data available after the Tour and we have not received a response as yet.

Many will know that Froome aligned with our organisation some years ago whilst riding for Team Barloworld. We have asked for clarification from Chris on a number of occasions in the last 18 months via email and direct message on Twitter if he still wished to form part of our organisation. As a result of not receiving such clarification from Chris or Team Sky in recent days we have made the difficult decision to remove his bio page from our website. This in no way insinuates that Froome is a suspicious rider but we feel that if riders do not support our organisation then there is no reason for us to promote them as such.

For those who ask ‘Why should riders release their data?’, our reply would simply be ‘Why wouldn’t you?’. World opinion is that pro cycling is tainted and as such, many cycling fans have become non believers. Bike Pure are campaigning for more transparency and emphasis on the publication of data which over time, will help bring the non believers back to the sport.

http://bikepure.org/2013/06/transparency-grand-tour-contenders/

 

[sniper]

Race Radio ‏@TheRaceRadio 17 Jul
Guess it is not "pseudo science" @fredgrappe had already done his own estimates and Sky's number's came in within 2.5% of his estimates

seems RR suddenly forgot all about Grappe's flawed analysis of the armstrong data.

are we ever gonna see the old RR back?

 

[acoggan]

acoggan, you'll have to forgive my lack of expertise with the details of heart rate and power decoupling. However I struggle to understand your hypothesis that the stroke volume would increase significantly.

In trained athletes, stroke volume increases from around 105ml at rest to 163ml at maximal exercise (a 55% increase). This is compared to a heart rate increases that go from 50 odd bpm to 185 or so bpm (270%) increase.

Furthermore, heart rate and cardiac output are governed by the Frank-Starling law and the Bainbridge reflex. As cardiac output increases, more blood returns to right atrium, stretching the walls of the heart chambers and causing the heart to reflexly contract and eject blood in to the systemic or pulmonary circulation.
All the while, this increased stretch causes increased activation of baroreceptors in the sinus node and causes an increase in heart rate. Then of course stretching of the right atrium also causes the upregulation of the Bainbridge reflex which also increases heart rate.

So any significant increase in stroke volume that is not accompanied by HR rises would have to overcome these reflexes. I suppose its possible via vagal stimulation though? Perhaps due to fatigue...

However for cardiac output to increase significantly enough to supply someone at maximal exercise without heart rate going up much....?
After all, cardiac output and oxygen consumption are dirrectly related. I also assume someone riding at a high power output in a TT will also require significant oxygen consumption, no matter how far you are in to a grand tour. Correct me if I am wrong here.

So if we just assume Chris Froome is an average elite athlete for a moment (by that I mean resting HR of 50 and max of 185) and then we can see what is required for a high cardiac output (lets say 30L per minute) with a low heart rate.

At rest, we can say he is doing 105ml SV X 50bpm which = 5.25L per minute.

At maximal exercise we would have a scenario of 163ml SV X 185bpm = 30.15L per minute.

If he were to have a maximal cardiac output of 30L and average HR of 150bpm, stroke volume would need to be a significant 200ml. Thats a huge variation from a resting stroke volume. While evidence shows that changes in heart rate are responsible for 75% of the increase in oxygen uptake from rest to maximal exertion.

Personally, on days where my HR doesnt get up to the >170bpm range in time trials, I have rather poor power. Conversely, my TTs riden at >190bpm range are usually very good performances. Of course the plural of anecdote is 'no data' though.

I'm not saying its impossible at all, but for your hypothesis to be correct; an increadible variation in stroke volume would need to occur along with the suppression of the bainbridge reflex. I'm not saying you're wrong, I just find this... shall we call it "inverse cardiovascular drift" unlikely.

Or am I not seing the forrest through the trees?

The latter, I'm afraid.

To give you a personal example: when I was younger, I had a VO2max of 5.4-5.45 L/min (at a body mass of 67-68 klg). My maximal heart rate was 181-183 beats/min, my hemoglobin concentration was 13.5-14 g/100 mL, and (not surprisingly, given my VO2max) I suffered from moderate exercise-induced arterial hypoxemia (let's say an SaO2 of 95% at VO2max). I'll let you work out what my stroke volume had to have been...then recognize that my VO2max and FTP were some 10-20% lower than found in the creme-de-la-creme of elite endurance athletes.

 

[Le breton]

I believe this must be the link I had found concerning Froome's power-duration curve (which I called power profile yesterday), published I believe soon after the Vuelta 2011.

http://teamsky.cyclingnews.com/wp-content/uploads/2011/09/VueltaPowerDistFroome2011.png

I has disappeared from the web.

..............

I FOUND MY PRINTED VERSION OF
http://teamsky.cyclingnews.com/wp-content/uploads/2011/09/VueltaPowerDistFroome2011.png
Unfortunately some details are hard to see because I had left it on a desk not far from a Velux roof window and it rained a bit on it.

The "x-axis" of the histogram is watts : 10-15 to 780-785 watts. 5 watts bins.
The y-axis goes up to 50 minutes in 50 seconds increments.

As published, the graph shows 2 line fits in red.
I can look at it more closely later.
But the top line LINE TOP goes from
~306 watts @ 44:10 duration
to 400 watts @ 34-35 min. duration

The bottom line LINE BOTTOM goes from
~417 watts @ ~31 min.duration
to 522 watts @ 7:30 duration.

The histograms VALUEs BETWEEN 306 AND 400 WATTS TEND TO BE ABOVE THE LINE (durations longers by a few minutes up to 3-4 minutes.

Between 417 and 522 watts the histograms values are close to the red LINE
BOTTOM.

^From 520 watts to about 800 watts the histogram values drop smoothly

800 watts <==> 50 s.
680 watts <==> 1:40
630 watts <==> 2:30
605 watts <==> 3:20
585 watts <==> 4:10

From there you can interpolate with the 552 watts / 7:30 point.

AND there is a comment on the graph :
Drop in power distribution at 420 watts. This is a good indication of threshold power

 

[acoggan]

I FOUND MY PRINTED VERSION OF
http://teamsky.cyclingnews.com/wp-content/uploads/2011/09/VueltaPowerDistFroome2011.png
Unfortunately some details are hard to see because I had left it on a desk not far from a Velux roof window and it rained a bit on it.

The "x-axis" of the histogram is watts : 10-15 to 780-785 watts. 5 watts bins.
The y-axis goes up to 50 minutes in 50 seconds increments.

As published, the graph shows 2 line fits in red.
I can look at it more closely later.
But the top line LINE TOP goes from
~306 watts @ 44:10 duration
to 400 watts @ 34-35 min. duration

The bottom line LINE BOTTOM goes from
~417 watts @ ~31 min.duration
to 522 watts @ 7:30 duration.

The histograms VALUEs BETWEEN 306 AND 400 WATTS TEND TO BE ABOVE THE LINE (durations longers by a few minutes up to 3-4 minutes.

Between 417 and 522 watts the histograms values are close to the red LINE
BOTTOM.

^From 520 watts to about 800 watts the histogram values drop smoothly

800 watts <==> 50 s.
680 watts <==> 1:40
630 watts <==> 2:30
605 watts <==> 3:20
585 watts <==> 4:10

From there you can interpolate with the 552 watts / 7:30 point.

AND there is a comment on the graph :
Drop in power distribution at 420 watts. This is a good indication of threshold power

I'm afraid that I'm a bit confused - some of what you wrote makes it sound like a power distribution histogram, but others make it sound like a power-duration graph ("mean maximal power chart" in WKO+-speak). Any chance you could scan the print-out and post the image?

 

[taiwan]

That's what was reported by Velonews.

And if so people are trying to infer the 60-min : 20-min mean maximal power ratio from power data covering only 18 climbs of unknown duration over a 2 year period.

Right.

That depends on information which we simply don't have.



Well that's my point. We don't actually know, hence attempting to draw such conclusions from one snippet of reported information without any context about what it means is pointless.

e.g. 60W might indeed be right on the money, who knows, we might be comparing a somewhat fatigued hour in an aggressive TT position or perhaps when not really going au bloc versus a 20-min smashfest ascent with fresh legs.

Without actual context, we just can't say.

Does everyone here put out exactly same maximal effort power on different days, while carrying different levels of fatigue, on flats versus climbs, on TT rig versus road bike, on a stinking hot day versus a cool one etc etc?

To be clear, my point was that part of the analysis is clearly wrong/meaningless. It's Grappe who saw fit to draw conclusions in saying that 60W was normal (or abnormal) when it's just an arbitrary number.

 

[V3R1T4S]

I'm afraid that I'm a bit confused - some of what you wrote makes it sound like a power distribution histogram, but others make it sound like a power-duration graph ("mean maximal power chart" in WKO+-speak). Any chance you could scan the print-out and post the image?

I don't have the image but it is a histogram containing the total time throughout the entire (or most of) the Vuelta spent at certain powers, binned in 5w increments. The red line and comment exist, as Le Breton said.

If we assume the ~3% inflation factor from the funny rings, a threshold of 420w on them is ~408w or 6 w/kg at 68kg.