AeT Heart Rate Drift Test – Negative Result

  • Creator
  • #49270
    Tyler Johnson

    Went out and tried to do the AeT Heart Rate Drift Test. Walked for a few minutes and then ran for 1:13. For the last hour of the run, my Pa:Hr ended up at -5.90% with an average heart rate of 131.

    This obviously isn’t accurate for my AeT, but before I go out and try again at a higher HR I wanted to mention that my previous “longest run” was around 45 minutes. I was running really slowly to keep my HR at 130-135 and a few times during the run I had to walk in order to bring my HR back down and my speed actually increased.

    I’m not sure if run/walking is allowed during the test or not.

    You must be logged in to view attached files.
  • Participant
    russes011 on #51684

    Thinking about it a bit more, you could theoretically determine your AeT HR by keeping your heart rate constant and measuring how much you have to adjust your pace to maintain the goal HR. This would be, in essence, the converse of a HR drift test, or in other words a Pace drift test. You could get similar info it seems and confirm your AeT HR this way, but it wouldn’t be fair to call it a HR drift test.

    — Steve

    Shashi on #51685


    There is no typo in the heart rate drift test article. It explains one way to conduct the heart rate drift test outdoors and it works.

    For a heart rate drift test, you try to maintain the HR or the pace. If you are able to maintain a constant pace for the duration of the run, then yes you can use that test as well.

    Training Peaks article on Pa:Hr calculation might be helpful.

    I think I understand why the article recommends maintaining the heart rate rather than pace for outdoor tests, but I will ask the coaches to clarify this.

    Shashi on #51691

    I reached out to Scott Johnston to understand why the heart rate drift test article recommends maintaining HR, rather than Pace. Here is the clarification he provided –

    The reason for trying to hold the HR constant is that it is easier than trying to hold a constant pace. However, if the course is flat enough the TP Hr:Pa calculation will account for variations in both HR and pace.

    russes011 on #51693


    Thank you for your posts, as well as your effort to help me understand what’s happening here.


    russes011 on #51699

    For those interested, this blog post on training peaks cleared this up for me and confirmed that I was wrong.

    That said, I think my (and perhaps others) confusion, in part, was/is related to calling the test a HR drift test when in fact the HR only drifts when the pace is kept constant. When HR is held constant, the pace drifts. Either way it gives the same results, and since, as Scott mentions, its easier to maintain HR (and not pace) when running outdoors this is the easiest way to measure the uncoupling of these two parameters. I may simply call it the drift test going forward.

    Off hand, however, I do not know if TP would give a negative value if you hold HR constant, and a positive value if you hold pace constant. If this is true then perhaps this explains why some users receive negative values. In the end its the deviation from zero that denotes drift, or decoupling, and not whether the value is positive or negative.

    Thanks Sashi and everyone else for ‘arguing’ with me, sometimes I need a little dose of the Socratic method to clear things up.

    LindsayTroy on #51741

    @steve- you can get a negative value in training peaks if your 2nd half is easier than your 1st half.

    That often happens if you are running up and back down a mountain. You could conceive going at a constant pace of say 15 min/mile with a HR of 150 and going down at a pace of 9 min/mile and a HR of 150. Likewise, you may also get a negative drift test if you go at a 15 min pace up AND down but your HR will vary from 150 on the up to maybe 100 on the down.

    I hope this concrete example helps.

    russes011 on #51764

    Thank you Lindsay,

    Thank you for your post.

    Bear with me if this is old news: based on the TP blog post I linked above the way TP calculates Pa:HR decoupling is pretty simple, it’s avg Pa/avg HR for the first half of your run minus avg Pa/avg HR for the second half of your run, with the result then divided by avg Pa/avg HR from the first half of your run. (note: one doesn’t need TP to determine decoupling, per TP)

    Your first example yields a decoupling of +40% and your second -53% (if my math is correct, using MPH for Pa).

    Nevertheless, I understand your point that the test can yield a negative result under certain circumstances (hills), and this result is likely suspect if the HR drifts down and/or the pace drifts up–specifically.


    LindsayTroy on #51795

    How did you convert pace to power? Trying to recreate your math but I don’t see the equations for pace, just for power.

    russes011 on #51797

    For running, the post says to substitute speed for power:

    “The AeT portion of the workout is split into halves. For each half the normalized power (cycling) or speed (running) is divided by the average heart rate to establish two ratios. The ratios are then compared by subtracting the first half ratio from the second half ratio and dividing the remainder by the first half ratio. This produces a power-to-heart rate-ratio percentage of change from the first half to the second half of the aerobic threshold ride. That percentage of change is your rate of decoupling.”

    (I used mph: 4mph = 15min pace; 6.6mph = 9min pace)

    As previously discussed, it goes on to state:

    “Note that there are two ways to do an aerobic threshold coupling workout. You can ride or run while keeping heart rate steady to see what happens to power or speed. Or you can maintain a steady power or speed and see what heart rate does.”

    Anonymous on #53344

    @russes011: The big-picture point is that you want to see how pace/power and HR behave relative to each other in an easy 60m time trial after a thorough warm up.

    On an outdoor flat course with moderate variation in pace and HR, the Training Peaks algorithm will calculate the Pa:HR drift.

    Without Training Peaks, you have two choices:

    1. Use a treadmill, keep the speed constant, and manually compare the first and second half averages in heart rate; or

    2. If outside with a GPS watch, keep HR relatively constant (within 5-ish beats ideally, definitely within 10), and then manually compare the first and second half averages in pace.

Viewing 10 replies - 16 through 25 (of 25 total)
  • You must be logged in to reply to this topic.