Friday, May 7, 2021

Low DFA a1 while running - a possible fix?

One of the most frustrating issues with analyzing DFA a1 during exercise has been the occasional athlete who exhibits the "low a1 while running syndrome".  The hallmark of this phenomenon is that the person will have reasonable behavior of a1 during an incremental ramp or series of progressive efforts while cycling, but have abrupt a1 suppression (with no further dynamic range of decrease) during even easy running.  Since a possible cause may be some sort of electromechanical artifact, a potential fix or partial mitigation could be through stabilization of the chest wall with a compression garment.  The following post will explore this in one individual who was kind enough to share his bike, run and compression shirt data.  All recordings were made with a Polar H10 over the past 2 months.

To start with, we need to document "normal" a1 behavior during cycling.  In the following, our friend cycled at different intervals of power and we compared HRV logger data to Kubios.  The HRVT was roughly at 210 watts or 132 bpm with good HRV logger agreement clinically to Kubios.

Low DFA a1 while running:

The next tracing was done while running at an easy pace until the end, where a sprint was done.  Since this was the first run done with HRV logger, our runner was perplexed as to why the a1 was so low with minimal effort and why it did not drop at the end during the sprint.

  • It is very evident that there is no dynamic range in the a1 - in fact it goes up with the sprint (blue circle)!
  • Ignoring the few a1 values above 1 (short rests) the average a1 was .61 with an average HR of 122.

Here is a histogram of the a1 values (minus the rests):

  • DFA a1 was .61 at a HR of 122, instead of about .75 at a 132 bpm during cycling.
  • Many DFA a1 values were .5 or below, normally associated with very high intensity efforts.

The possible fix - stabilization with a compression garment

The following tracing was done with the same HRM (H10) but our runner wore a compression shirt (Underarmour) and positioned the HRM belt higher to make sure the shirt compressed the belt with appropriate tension.  It was still an easy-moderate run but done at a higher average HR/pace than above and also followed by a sprint.

  • The inappropriate a1 suppression appears to be gone!
  • Over the course of the 20 minute constant pace, with an average HR of 130, the DFA a1 average was .77.
  • This is very similar to that obtained on the bike.
  • There is an expected DFA a1 drop (and HR elevation) during the sprint (red arrow)

Here is a histogram of DFA a1 values over the 30 minutes

  • We see a good "normal distribution" of values clustered around the mean of .77.
  • Yes, there are some low values, but that is to be expected when performing near the HRVT.


  • Thanks to a very helpful sports medicine physician, we may have a workaround to the "low a1 in runners" syndrome.
  • Stabilization of the chest/HRM belt interface by a compression garment may prevent or mitigate the inappropriate loss of correlation patterns in HRV.
  • Although this is very suggestive, reproducing this in others is important - please share your experiences including what kind of compression garment you are using.  It probably needs to be fairly tight, along the lines of Skins.
  • If this is repeatable in the majority of users, we will need to be aware of this oddity and how to fix it.
  • Be aware that DFA a1 can vary over time.  I presented the histograms to show that this type of behavior is common and to be expected. However, the mean values appear to be valid for HR comparisons.

Heart rate variability during dynamic exercise


  1. It's great if you are a woman! I wear my HRM belt tightly snugged under my running bra and it works great with the app. Maybe the men would want to try this too. :)

  2. Bruce, do you know what was the exact placement of the H10 ? Above nipples ?

    1. Yes, he placed it right across the nipples. We did that to maximize the compression effect. If your compression shirt is tight in the usual position, it might be fine to put the belt in the usual position.

  3. Is the purpose of the compression shirt mainly to stabilize the HRM itself or the contact points of the electrodes to the skin? thanks

    1. I don't think it has to be a loss of contact. Possibly related to vibration, so the compression hopefully will reduce that.

    2. Thanks! the reason I'm asking is because I've tried a compression t-shirt which didn't work, perhaps not tight enough. Then I tried stabilizing it with 2 pieces of KT tape cross taping just the HRM part and it still didn't work. I'll try moving my HRM up to my pecs + tapes next time to see if it works :D

    3. Let me know how you do. This is what I'm trying to achieve -

  4. Dr. Bruce, I came across your name on the Endurance Innovation podcast and have been intrigued/obsessively curious with determining aerobic threshold via DFA-1 for running. I've tried FatMaxxer and did a separate ramp test that I uploaded into Runalyze. Both times my DFA-1 drops below .75 before the testing ramp even begins. This seems really low to me since my critical pace is around 6:50 min/mile. I tried to troubleshoot using your blog and came across this article. Not sure if this is the reason for my low DFA-1. Any other instances you have come across? Is it possible to be very aerobically undertrained but still have threshold/anaerobic speed? I can send a copy of the file if you are willing to take a look

    1. Thanks for trying our method out. I think you have the classic "low a1 while running" issue, not from anything wrong aerobically. There are a few things to try - 1)do a ramp or a series of constant power intervals on a bike, 2)use a stepper or a very steep hill to rapidly walk and measure the a1 vs heart rate for the threshold. Let me know how it goes.

    2. Thanks for the feedback! I couldn't wait and did a bike ramp session during lunch. I used Fatmaxxer to view realtime DFA-1 values while I ran Richard Collier's DFA alignment workout from Xert which is a warm-up followed by 6 minute steps.

      The set-up was:
      Polar H10 connected to one phone running FatMaxxer
      Polar H10 connected to Garmin Fenix 6 (to record HRV)
      Bluetooth connection was used for everything
      Xert EBC Android app running on a second phone
      Garmin Fenix 6 (with HRV logging turned on) and Xert EBC connected to Wahoo Kickr Bike (power source)
      I went through the 6 minutes steps until FatMaxxer indicated a DFA-1 under .75 for a prolonged period. Interestingly, there was a dip below .75 early in the test but then it rebounded back up.

      After the test, the Garmin file with HRV was uploaded to Runalyze and it did an automatic calculation of AeT at 278 watts. So you are right that the "low a1 while running issue" is unfortunately applicable to me.

      In Xert, it has my Lower Threshold Power at 293 watts. The possible delta might be that I did a running session earlier this morning, about 4 hours prior.

      As a follow-up, if I'm looking to find my Running power (via Stryd) along with pace/heart rate via DFA-1 , do you have any recommendations? I do lift more than the normal endurance athlete so have larger pecs that prob cause extra noise (I wear my HR strap just under the pec). I also have the Polar Verity Sense optical HR which I've seen is possible to capture usable HRV values.

  5. Also, FatMaxxer showed a dip for a few minutes below .75 early on in the test but it rebounded and had a normal gradual decline again. Is that typical?

    1. Congrats on a successful test. Yes, the a1 can dip and undulate a bit during the ramp, especially with long intervals (see Use the HR found during the cycling test as your running threshold value - not the power equivalent (power run does not equal power cycling). So if the bike HR at a1=.75 is about 140 bpm, that's going to be close to the running HR threshold. You can also follow the threshold cycling power and HR over time, assessing improvements. I'm a long time lifter as well and I don't see the pectoral muscle size as a potential issue in running - I've seen the "low a1 issue running" in all body types. No optical HRM can capture accurate HRV while exercising to my knowledge.

    2. Thank you for being generous with your time answering my questions. Looking forward to reading more of your posts and hearing you speak on podcasts!

  6. Dr Bruce, I've been brainstorming ways to try to measure HRV on the run and workaround the low DFA values from chest vibration. Have you had any experience with the CorSense by EliteHRV as an alternative device to pick up HRV?

    1. Hi. No direct experience with that device. What I will say is that the low a1 issue happens with both the h10 and movesense ECG belt, both chest strap units. I imagine any chest strap device will behave similarly. It would be interesting to look at a 12 lead ECG in that person.

  7. What's the latest on the low a1 issue in runners? I find it to be quite reliable on a ramp on an exercise bike and I use it to guide me for running heart rate targets. But for actual running values I don't have any luck with it. I wonder if the research in this article was probably plagued by the issue without actually being aware of it?