why I moved from Garmin to Polar H10 – alphaHRV

alphahrv icon brnad logo
using the alphaHRV Connect IQ Data field with Polar H10 on Garmin

why I moved from Garmin to Polar H10 – AlphaHRV

I’ve just tweaked my personal gear setup to use Polar H10 as my main sports heart rate monitor, at least for a month until the lakes open at which point I’ll need the caching that only Garmin offers.

Interesting discovery: Garmin HRV above 120bpm is *WRONG* with ANT+

I intend to use the alphaHRV data field during exercise until March/April and the best setup was to swap out the Garmin HRM PRO Plus and maximise the HRV accuracy with the Polar H10. The setup for alphaHRV is a little complicated, so I’ll cover that as well.

What is alphaHRV?

More: alphaHRV on connect IQ Store

alphaHRV is one of the best Connect IQ data fields for those of you wanting a real-time DFA analysis of HRV during exercise. The DFA analysis (Wiki) produces the alpha1 metric which will likely range somewhere between 0.10 and 1.5 with a high number corresponding to an easier effort.

Generally, it should be the case that alpha1 = 0.75 corresponds to the aerobic threshold (AeT) and that alpha1 = 0.50 might correspond to the anaerobic threshold (AnT).

alphaHRV also produces accurate respiration rate numbers that can determine VT1. The alphaHRV algorithms are supposed to be superior to those used by Garmin (See: intervals.icu/ Bruce Rogers’ work)

why Use It?

I have somewhat vague motivations here. Later in the year, I have a 200-mile ride to look forward to and I want to ensure that I am doing my Zone 2 work properly. I understand that our training zones are not entirely real constructs but rather there exists a continuum of effects that change as our exercise intensities change over time, perhaps that continuum even changes during a longer workout. I want more certainty that some of my efforts are not too intense.

Sure I will continue to use power and HR.

I’ve even dabbled again with SmO2 and use Moxy/Train.Red FYER/NNOXX One as appropriate. So why yet another metric?

I’ve also been trying to guide some of my training with AI Endurance. It just so happens that AI Endurance uses post-workout DFA a1 numbers to assess training intensities. AI endurance determines which historical intensity/duration combinations have produced the best gains for you and then recommends similar ones going forward. This is probably the next step forward in consumer-grade training that we will see over the next few years. The last step forward was with adaptive training which we have seen spread fairly widely over the last few years with excellent implementation from the likes of TrainerRoad.

Anyway. AI Endurance only works on completed workouts and aphaHRV only works during a workout. So, in my eyes, that seems highly complimentary! the developers of each of those respective pieces of software (Markus/Inigo) are also complimentary of each other!

More: AI Endurance

I just want to be clear here that these developments in sports science are being properly undertaken by AIEndurance and alphaHRV. It’s not just some random guys jumping on the ML/AI/DFA bandwagons and merely rebranding something they’ve had for years. It’s the real deal…sure it’ll improve over time but they’re good for you to try now.

New vs old gear setup

Part of the reason for this post was to highlight some of the technical vagaries that need to be overcome.

  • On my bike, I’ll need to use the H10’s Bluetooth channel for the alphaHRV data field and an ANT+ pairing for the Edge 540.
  • On my watch, I’ll need to use the H10’s second Bluetooth channel for the alphaHRV data field and an ANT+ pairing for the Forerunner 965. the second channel is enabled though the Polar BEAT app.
  • Waking HRV Readings from HRV4Training on my iPhone now switched to my bedside Polar H9, previously I used the H10 which has now been repurposed as my sports monitor and will be in a different room.

Hopefully, you can see why I used the Polar H10 and not the H9. The H9 only has one Bluetooth channel and I prefer to use an individual channel for each of my devices to ensure they each use the more reliable Bluetooth connection for HRV.

I often use my Forerunner and Edge in the same workout so the only way to save HR to each device is by using an ANT+ connection to the H10 (caveat further below)

Q: How do you pair a heart rate monitor to the alphaHRV data field

A: It’s complicated and you should follow the developer’s instructions to the letter. There are a few alternative configurations. In principle, you want to fix the Bluetooth connection to your data field and then pair it again to your Watch/Edge for your regular HR connection outside of the data field. You might have additional problems with viewing or sharing data in other data fields but I found that my setup worked with the Stryd Zones data field which I assume uses ANT+ to display HR from the H10.

Q: Can I use my Garmin HRM?

A: Yes, use Bluetooth. BUT!!! HRV is not always transmitted correctly by Garmin over ANT+. Garmin transmits each RR beat separately over ANT+, however, Garmin only sends data at 2Hz, so beats can be missed when HR exceeds 120bpm. Whereas Bluetooth can have up to 6 beats sent at a time and alphaHRV sends 4 hence giving accuracy to 250bpm. Bluetooth 5.X should also include Forward Error Correction (FEC) although that won’t have much of an effect, if any. Note: I am 99.9% happy with the regular HR I get from my Garmin HRM-PRO PLUS it’s only in the case of sporting levels of HRV where Garmin appears to have historically cut corners to boost battery life on its HRMs.

I’ve not yet encountered the problem of linking to Zwift and/or my indoor trainer but I’ll cover that when I get there

What have I found so far?

The tech setup seems fine with good data coming through. The alphaHRV data Field is in late beta but seems very stable on the Forerunner 965 and Edge 540.

alphaHRV has either a coloured line chart or a coloured numeric field on the Forerunner/Edge. The colours change as you cross the 0.5 and 0.75 alpha1 thresholds and I get both fields when I use it full screen and can also add on an extra metric or two like power/HR. I don’t think it’s possible to include Stryd power on this data field unless you pair Stryd as a power meter which goes against the normal recommendation.

The alpha1 metric seems quite sensitive. It also lags similarly to how heart rate lags to changes in effort ie by about 20 seconds. I’ve been trying to stay above alpha1=0.75 but found that I had to run slower than expected to achieve that, say at a pace that was at least 5 bpm slower. Then what I thought were only modest increases in effort soon crashed my alpha1. It seems a very sensitive metric when running, perhaps less so when cycling. Maybe I’m not as fit as I think I am or maybe I’m one of those people where alpha1 is not a great metric to use (apparently they do exist.)

How could alphaHRV be improved?

These suggestions would come from me as an athlete wanting to make my life a little easier and class as ‘nice to have’.

  1. Easier setup!
  2. Colour-coded ‘zones’ in Garmin Connect to match the ‘zones’ on the live data field (probably impossible)
  3. Option to include Stryd power on the alphaHRV data field (probably impossible)
  4. Ability to overwrite main FIT data with the developer data ie HR and RR/HRV (probably impossible)
  5. I had hoped to see an updated live heart that corresponded to alpha1 = 0.75 ie my Zone 2 upper end. That’s very difficult to do in a data field within the constraints imposed by Garmin. It’s best done after the workout eg in AI Endurance. Plus to deduce my Z2 HR from alpha1 would need averages from about 10 workouts (Source: Markus, AIEndurance)

Why Will I Move BACK to Garmin HRM-PRO PLUS

Mainly I’ll move back for caching HR swim data when I wear the Garmin under a wetsuit when the lakes open.

The Garmin battery tends to last longer as well – Garmin uses CR2032 not Polar’s CR2025, the former is 0.7mm deeper and hence bigger and holds a greater capacity. Although a CR2032 also kinda fits into a Polar H10/H9 but not really in a waterproof way.

Others may like the Garmin Running Dynamics that only Garmin gives from its chest straps but, then again, the watches have those metrics now.

Take Out

Like me, you’ve almost certainly got a fancy training gadget that you don’t push anywhere close to its technical limits. Maybe your training needs spicing up a little or maybe you have a keen interest in HR/HRV and sports science. alphaHRV is free! Or at least it’s free while in beta. READ AND FOLLOW THE INSTALLATION INSTRUCTIONS.

For many athletes, the range 1.0 >alpha1 > 0.75 should represent a dynamic and reliable way to focus on what others might call Zone 2. No doubt other uses will be verified in the months and years to come.

More: alphaHRV on connect IQ Store

More: AI Endurance


Reader-Powered Content

This content is not sponsored. It’s mostly me behind the labour of love which is this site and I appreciate everyone who follows, subscribes or Buys Me A Coffee ❤️ Alternatively please buy the reviewed product from my partners. Thank you! FTC: Affiliate Disclosure: Links pay commission. As an Amazon Associate, I earn from qualifying purchases.

23 thoughts on “why I moved from Garmin to Polar H10 – alphaHRV

  1. I’m not sure about HRV specifically, but the HRM-PRO+ only transmits the advanced running dynamic metrics over ANT+ and not over BLE. So I think it is very safe to say that HRV is transmitted over ANT+.

    I would be really curious to compare the raw HRV from the fit file to see how different the HRM-Pro and H10 are. I think that would be possible if the alphaHRV field is recording the HRV to a CIQ field instead of the native field.

    1. I view my files on runalyze, which includes the HRV data.

      The HRM-Pro Plus does transmit HRV over Ant+, but it records with a much higher percentage of anomalies versus BLE. Since going to BLE loses the running dynamic data, I do most of my runs with the H10 (which I find more comfortable, and I prefer being able to remove the pod) and the RD-pod (because I’m a geek and still want to collect the running dynamic data even if I’ve found nothing actionable from it).

      1. Hi ChatGPT, please explain to a 3yo (with a lower than average IQ) if the HRV data from Runalyze collected with garmin HRM is of any use and if there are datafields or similar copycats to this AlphaHRV thing that can be used without having to buy yet another fancy training gadget.
        Asking for a friend of course

  2. According to Dr Bruce Rogers (muscleoxygentraining.com) who has done extensive validation the Garmin HRM is not good enough for DFA a1. He says the H10 specifically with Bluetooth is much better. But the gold standard is the Movesense Medical over Bluetooth.

    The Movesense Medical is a certified medical device and essentially as accurate as an ECG machine.


    I found have tried the alphaHRV ciq data field with an H10 and it just doesn’t seem to work well for me with running. Maybe one day I will decide to cough up the 3x price for a Movesense.

    My H10 pod seems to be on the fritz so maybe sooner than I would have thought. Still €349 for just the sensor and is steep.

    1. did you use the BLE connection or ant+ with the H10?

      An ECG would have more leads? I believe H10 is certified as a 1-lead ecg.

      1. BLE based upon recommendations from Dr Rogers and his experiments showing much noisier data over ANT+. I have the ant+ connection disabled.

      2. Noise is artifacts and bad readings, like static over an analog TV signal. The causes can be dropped packets or electromagnetic interference or failing to detect a heartbeat (missed interval) or picking up electrical signals from other large muscle groups like the diaphragm causing incorrect readings or physical interaction with the chest strap and your body and gravity. There are a lot of possible sources.

        The sensor works by detecting electrical potentials and interpreting those as a heartbeat. There can be any number of things that cause spurious readings or data to not be successfully transmitted.

        Higher sampling frequency tends to be lower noise. Bluetooth is less subject to interference than ANT+ because it is a connection-oriented protocol with reliable packet delivery rather than a best effort broadcast protocol.

        The H10 with bluetooth has about 1% artifacts while running when tested in a lab. In another lab test “a tale of 2 hrvs – afib vs noise” the H10 has 50% artifacts.

        Running as an activity is really difficult for a sensor to get good enough data in order to do regression on the RR interval rather than just getting a heart rate.

        I think there is not much doubt that an H10 with a Bluetooth connection is the best widely available ECG heart rate sensor. The Movesense Medical seems to be significantly cleaner in the data it gathers but at 3x the price.





      3. The AlphaHRV devs recently posted that Ant+ was now accurate. Bruce had a blogpost in regards to that

      4. Found the confirmation: “Potential game changer – it appears this app takes advantage of a higher Ant+ sample rate than the Garmin native device. The result being – no more rate related missed beat artifact. I have confirmed this in my own recording and another from a friend. At this point we can have the best of both worlds – Ant+ to the alphaHRV app, bluetooth to the watch and Fatmaxxer.”


      5. Hi Bruce Rogers @bjrmd just confirmed this to me: “… Garmin has never fixed that issue in their code. I use the bluetooth channel to record to the Garmin fit file and ant+ for alpha hrv.”

    2. Just use the h10 with the polar app, it is absolutely seemless. It is claimed 99.6% accurate, i have put it through it’s paces with every type of training possible and it is perfect.

  3. When using Garmin HRM, is it better to connect it to Edge or watch using BLE, instead of ANT+? I mean for general HR tracking to be more accurate, not necessary for using Alpha1

  4. I have what looks like accurate dfa data when paired with my H10 strap but recently bought a fourth frontier strap and this seems very unreliable? Any ideas?

  5. I’m sure you know this (even though you use those absolute values in your text):


    “All of the above is to say really just one thing: there is not such a thing as a universal threshold that we can use to assess or guide exercise intensity at the individual level. It does not exist for heart rate, it does not exist for lactate, and it does not exist for HRV, regardless of what features we are using (standard ones or alpha 1).”

    “HRV during exercise might be a valuable tool, but given the evidence accumulated in the past year, we cannot anymore claim that 0.75 is a universal threshold identifying the aerobic threshold.”

    Overall HRV and derived metrics have to be taken more individual and also need to be combined with other metrics and feelings, for example see:



    1. Yes, I agree !

      FYI: The 0.75 value is configurable in the data field.

      I said in the article “For many athletes, the range 1.0 >alpha1 > 0.75 should represent a dynamic and reliable way to focus on what others might call Zone 2”. that’s the area I’m generally targetting rather than a specific value and the developer also thinks that’s sensible. I’m in the exploring phase at the moment in any case.

      try this as well: http://www.muscleoxygentraining.com/2024/02/hrvts-agreement-with-established.html (reports on a Feb 2024 paper)

  6. Works great with Polar H10 connected to Edge 830, as well as Epix2. I use ANT+ as a general normal connection, and 2 BLE connections to connect to each field in Edge and Epix. I still get some moderate artifacts (around 2-3%, sometimes more, especially when working hard). Q: which connection is better/less artifacts: as above, or using BLE to connect to the devices and ANT+ for Alpha1 data field? Also, what is the HR threshold for artifacts detection?

    1. Both ANT+ and BLE should provide almost same results in terms of artifacts.
      The HR threshold for artifacts is the threshold used by alphaHRV to distinguish exercise from rest condition. As heart rate variability is quite higher at rest, the artifacts detection algorithm uses different criteria depending on this condition. By default it is set to 90 bpm and shouldn’t be modified except if you face high artifacts level at low HR. For instance at the beginning of the activity or when you stop for a short time during the activity

Leave a Reply

Your email address will not be published. Required fields are marked *