HornedFrog

background: i've been cycling for a few years now but recently really became interested in improving my fitness.

i just did a FTP test where i sustained maximum effort for 20 minutes without slowing down, recording splits every minute.

my average heart rate for this was ~172, and my bike computer gave a caloric burn estimate of ~14cal/min (i'm assuming this is actually kcal, and is based on HR measurements only I believe). i did the conversion and 14kcal/min = 58576 Joules/min = 976 Watts

My average speed on the trainer for this duration is ~20.25mph, which looking at kurt kinetic's published numbers should be ~270 Watts (16200 Joules/min)

i found this interesting that my body is only about 25% efficient at this power output. granted it has to perform more functions than just spinning the pedals of my bike at any given time. i imagine the better the shape you're in the higher your efficiency. anybody done a similar measurement?

edit: i think i could have gone another .25mph faster and still held on... but i was worried about cracking and wanted to finish strong.

silversx80

There really isn't a significant connection. Of course power goes up and HR will lag behind, eventually going up, but the estimates can be more than 20% off for energy burned. To get a better estimate based on body metrics, you'd need a calorimeter that will measure waste heat (but still make some significant assumptions).

As far as power from the trainer is concerned, don't use average speed. Power should be progressive, so even though you have an average of 20.25 mph, the power won't average out. At one point I had my Fluid2 mapped in an excel file, and I'd export from my Edge 500 with 1s readings. I would be able to get an estimated power based on the resistance curve. Eventually, I got tired of guessing and just got a power meter. The added benefit is that I had a specific number to hit, indoors and out.

RPK79

What you need to do is extrapolate the total wattage output and divide it by the total heart beats to come up with a watt per beat and you can use your WPB as a gauge to how you are progressing.

TrojanHorse

In fact they are connected but it's only really relevant if you're in great shape - it's called aerobic decoupling when you're not. Here's some light reading on the matter.

There is no direct connection between the two if that's what you meant, but they are directly related if you are suitably conditioned.

Seattle Forrest

My aerobic decoupling numbers are all over the place. One ride it's 12 %, the next it's -5 %, then 1.8 %, then 20 % then negative again and then ...

andr0id

270 watts for 20mph on a trainer is believable.

976 is just totally wrong but not unexpected because the "guessed" calories shown by most cycle computers tend to be 2 or 3 times high based on comparing real power meters against heart rate based calculations.

Seattle Forrest

Calories from heart rate (chest strap, not wrist monitor) can be up to 7 % off on a modern Garmin, Polar, or Suunto according to Firstbeat, who makes the algorithm they all share.

Calories from power can be off by up to 5 % according to the gods of power. Plus your power meter has an error of up to 1.5 % although I haven't been able to find any info on how that's determined.

That's pretty close.

(But 976 was watts, not calories.)

HornedFrog

i think it could be a fair estimate. i mean, if your BMR is 2000kcal/day while being a desk jockey, that's about 83kcal/hr or ~1.5kcal/min. if you go out an hammer as hard as you can doing rigorous exercise, you expect to burn what, 1000kcal/hr which is 16kcal/min? garmin is suggesting that i'm burning 14kcal/min. not crazy.

the calories put into the bike will always be less than the calories consumed by your body as a whole.

Seattle Forrest

It's also true that people are 20 to 25 % efficient on the bike. I think it's the case that efficiency in other sports varies a lot more than that, but running and swimming are pretty free form things, whereas almost all cyclists spend most of their time riding with their ass on the saddle and their hands on the bars, turning their feet in a circle anywhere from 320 to 355 mm across. All of your movement is pretty constrained on a bike and energy efficiency doesn't vary all that much as a result. At least that's what I've been told.

Drew Eckhardt 

The heart rate to power relationship is linear between aerobic and anaerobic (aka lactate) thresholds.



although heart rate lags power



and it can be influenced by factors like heat and over-reaching or over-training.

You can train to make more power at the same heart rate or loose that. Here's my peak 10 and 60 minute power/heart rate data which goes both ways:



223W in April and 235W in May come from FTP tests leading to a 2x20 workout instead of my usual 3x10 or 4x10 so heart rate had more time to catch up for a higher average with lower power.

The 239W at a high 166bpm in June followed riding too many endurance miles during my scheduled rest week which left me unable to finish 3x10, although progressing from over-reaching to over-training heart rate can be depressed.

April and May 60 minute numbers are from my weekly 1:00-1:15 steady threshold effort. There's a nice trend there with increasing power and decreasing heart rate.

It took me three weeks after crashing June 18th until I was comfortable enough to get on a trainer; that was catastrophic as reflected in a 50W 10 minute power drop at the same heart rate. 3x10 is OK although I'm not yet up for an hour threshold ride.

25% is often right for hard efforts, although bike computers are notoriously inaccurate except when approximating (1 kj = 1 Calorie) off measured energy output. My Garmin was often 20% high off heart rate, but could reach 80% until the last firmware upgrade which switched the data source from heart rate to power meter measured energy.

If you care about efficiency and/or your carb to protein ratio you need to have your respiration products measured by sports medicine people.

Drew Eckhardt 

The metric only works well at constant power output.

This is because the definition

((first half normalized power / first half average heart rate) - (second half normalized power / second half average heart rate)) / (first half normalized power / first half average heart rate)

does not take into account the zero intercept where you're on the bike producing no power.

That has more impact on the average at low intensities than high.

Psimet2001 

Use a power meter to measure power.

You do not produce 976 Watts in a 20 minute power test.

The 270 W number is more accurate but still an approximation.

Use a power meter to measure power. Anything else is guessing with varying degrees of confidence.

bmcer

Thanks for that explanation. I was beginning to think there was something significantly wrong with me since Golden Cheetah was reporting aerobic decoupling across a range from -5 to +45%. The vast majority of my rides involve a fair amount of stop and go plus a wide range of sustained power outputs. In other words, I don't do a lot of truly disciplined training since I don't compete.

So if I understand what you're saying about aerobic decoupling, I can pretty much ignore that metric unless I structure a ride specifically for the purpose of gauging it?

Drew Eckhardt 

You'd be faster with more structure to your training. That makes your tough rides hard enough to force adaptations in VO2max and your lactate system (strong efforts past an hour), and your gentle rides easy enough your body sticks with your aerobic system and that improves.

Seeing more scenery, loosing more weight, 20 MPH solo, etc. are all worthwhile if you don't compete.

Right.

bmcer

Understood. Thanks for the input and clarification.

gregf83 

Constant power at a tempo pace. Not really applicable if you're riding at FTP or above.

Drew Eckhardt 

Also Z1, Z2, and Z4.

In _The Power Meter Handbook: A User's Guide for Cyclists and Triathletes_ Friel writes

"For this metric to provide useful information, the workout or segment must have been fully aerobic (below the lactate threshold) and steady (Variability Index of 1.05 or less)"

where the Variability Index is defined as Normalized Power divided by Average Power.

gregf83 

Well Z4 is technically riding at lactate threshold and you'd expect HR to rise under those conditions. Z1 and Z2 aren't going to tell you as much about your fitness because most people won't have aerobic decoupling at low intensity.

chaadster

I'm sorry, but what was the question? "Anyone done a similar measurement?"