Power vs Heart Rate Metering

I did some science today.  Like most ‘sports science’, it wasn’t very good science.  But it was an experiment of sorts.

I’ve been banging on to anyone that will listen about the benefits of power meters in general and the PowerTap in particular all year.  I’ve used mine this year to gauge my indoor training efforts which have largely consisted of threshold work (2x20s, 3x20s, hour of power, etc). 

One of the many reasons (don’t get me started…) for using a power meter is to accurately gauge the work that is being done during an interval.  Intervals consist of a target intensity, duration and rest period.  Traditionally the intensity has been monitored using a HRM (heart rate meter).  The problem with this is that heart rate is subject to many confounding factors – fatigue, dehydration, temperature, caffeine and so on, all have an impact on heart rate  which makes it only a so-so guide to the intensity of the effort that is being made.  The response of the heart to effort is also considerably ‘damped’; ie it takes a long time to catch up with any changes of intensity.

Anyway, to my wonderfully unscientific experiment.  My hypothesis is that the power, and therefore work, would vary considerably across an interval for a consistent heart rate and vice versa.  Initially raising the heart rate would take a deal of effort and as the interval proceeded fatigue would mean that, in order to maintain a consistent heart rate, the power would need to decline.

I was planning a 3×20 ‘sweet spot’ session.  Using power as a metric this corresponds to about 290W.  Using heart rate I estimate this would be about 160bpm.  My plan was to do the first interval based on power, and see what happens to the heart rate, the second according to heart rate (159bpm +/-1) and the final interval back to power again.  Here is the result…


(click to image to see full size).

NB Heart rate is shown in red, Power is yellow.

It is worth bearing in mind is that, even indoors with the absence of wind, hills, junctions and so on, it’s not possible to maintain perfectly consistent power output.  Despite this I hope that you can see that the first and third efforts are reasonably consistent.

The first interval is very interesting.  I only warmed up for 5 minutes (coz I’m ‘ard) but even so, you can see that my heart rate climbed throughout the entire 20 minutes and took 5 minutes before it even reached 150 bpm let alone 160.  So the first 20 minutes displayed a totally non-linear relationship which is what I had expected (and, yes, hoped for!).

The second interval was done with a consistent as possible heart rate, targeting 159 +/- 1 bpm.  I did not want to exaggerate the initial effort so I tried to slowly build for a minute or so into the interval.  Either way you can see that the initial power was high, and quite fatiguing, and did slowly drift downwards through the remainder of the session.  But not to the extent to which I thought it would have.  Given the non-linearity of the first interval, the second I found surprisingly consistent.

By the third interval I was back to consistent power and again the heart rate response was variable, slowly building throughout the entire 20 minutes but much less than the first interval  – presumably due to my inadequate warm-up.


It should be noted that this was a generous test from the perspective of heart rate – the intervals were neither short nor massively intense.

What I hope I have shown is that

  • There is not a linear relationship between effort and heart rate response
  • Even working reasonably hard (90% of FTP) it can take 20 minutes, or more, for the heart rate to level off
  • Due to this latency, it is likely that using heart rate as a guide to interval intensity is likely to lead to starting too hard leading to unnecessary fatigue. 
  • If you are using a HRM to gauge interval effort a proper warm-up is a good idea
  • I found the second interval the hardest, most probably because of the 310W + required to get my heart going

No Comments on “Power vs Heart Rate Metering”

  1. 1 Mark said at 2:14 pm on September 7th, 2009:

    You might want to take a look at Joe Friel’s “aerobic decoupling” concept http://www.trainingbible.com/joesblog/2009/08/intervals-tempo-and-decoupling.html. In a nutshell, he says that if you’ve got a very solid aerobic base, your heart rate and power should track like they’re on rails. Rising heart rate during an interval may be a sign you’re pushing beyond what your base can support.

    If you like data on your biking, you should take a look at PerfPro (http://jetall3.com/PerfPRO/) which gives you a ton of stats to help evaluate your workouts (including automatic calculation of aerobic decoupling) for a fraction of the cost of similar CyclingPeaks software.

  2. 2 admin said at 3:24 pm on September 7th, 2009:


    Interesting article. I was about to disagree but then I saw that the consistency is only at ‘tempo’ pace. If you look at the graph that supports the article you’ll notice that the 4×5 minute intervals are distinctly non-linear.

    I’ll check out the software though – but I’m committed to wko+ and I’ve spent some time setting up a google docs spreadsheet with manual CTL, etc, calculations.

    Thanks for posting.


  3. 3 Simon Roberts said at 4:40 pm on September 8th, 2009:

    I thought that you get a consistent heart rate and power correlation at a lower FTP level. If you have a good base of fitness, you should be able to do an hour of 75% FTP, without the heart rate creeping up. My experience is that, with 85%+ workouts for 10mins +, I always get a creep upwards in the heart rate. But that might be my inferior fitness.

    By the way, have you been impressed with your Dura Ace wheels (or underwhelmed)?