What happened to VO2max calculation method ?

Okay, so I know VO2 max calculation is a recurrent debate, and there are countless threads about it since Garmin and Firstbeat don’t give much details about how their algorithm really works. All we know (knew) is that it compares HR and pace data, and estimates the level of effort based on max HR percentage. But I feel there might be more to this. For the best, or worst.

What I am worried about is that in the beginning of summer 2022, Garmin made some changes in their algorithm. Before that, the VO2max was calculated after each activity. Plain and simple. Since that update, VO2max is now synced with Garmin Connect and saved into your profile, along with training status and some other information such as HRV, training load, etc. Good thing is that you can reset your watch, sync, and data will be there, just where you left it. But here’s my point : it looks like VO2max is now “weighed” by the previous recorded value, and as I suspect, other data (HRV?) which may lock or unlock quick VO2max change.

If you look at the table above, you’ll see my reported VO2max values after each run. What is clear is that since mid 2022, the value changes very little between two runs, and so it can take up to months of training to see a significant (+1) change in VO2max reported value. 

That brings up two questions :

  1. Is this the wanted behaviour ? Since the VO2max reflects physiological changes, which are slow by definition, it could make sense that important VO2max change is “filtered” by the software.

  2. If yes is the right answer to question 1, then has Training Status – which is mostly based on how quick VO2max rises or fades – become completely irrelevant ? I think so. because based on my personal experience, VO2max changes and training status have always been pretty much in line with how strong or fit I felt. This year, it happened quite often that I got “Maintaining” TS, whether I was improving PRs and training hard & consistent, or being sick and feeling like sh**. If you look at the last 3 months my VO2max is slowly decreasing, while my training load is getting higher and race prediction time is improving.

  3. Would there be a way to "reset" the value and start up with a fresh, new measurement that does not account for activity history ?

I would be interested to know if other users have noticed the same change in the VO2max calculation process. Not that it would be a life changer of course, but if VO2max is no longer a good indicator to follow, it’s better to be aware of it.

Also, it would be great if we could have someone from Garmin ( @Garmin_Chris ?) give at least some background information... VO2max changes were never mentioned in any backlog.
Thank you for reading this far !

  • Yes, I have noticed the same. For me VO2max used to be quite volatile. It typically dropped 1-2 points after every trail run and then rebounded on next run. I don't think the previous behavior was correct, but the new one seems even worse. It is now stuck for me at 51 and has been slowly (very slowly) descending over the last few months.Also VO2max no longer seems in sync with race time predictions.

    Regarding the VO2max estimation, I am under impression that it is now based on estimated power vs. speed rather than HR vs speed. I think that has changed at the same time when trail running VO2max was introduced. 

  • A made a comparison between Runalyze's Effective VO2Max and the VO2Max estimates from Fenix (mostly) devices over the years since 2016. Garmin always overestimates compared to Runalyze, but that is not important here. Garmin and Runalyze VO2 Max were equally variable until (approx.) the end of 2021, when I was using a Fenix 6X. After an update to the 6X, the VO2Max estimate variability suddenly decreased a lot. At the time, there were a number of people reporting on the (beta) forum that their VO2Max got "stuck".

    When I started using a Fenix 7 (mid 2022), the VO2Max estimate adjusted way too much upwards and then adjusted downwards, but the low variability remains to this day. It is probably the method used in the Firstbeat algorithm that was changed and introduced in updates for several devices at different times.

    Trail runs, leisurely walks, hard interval sessions, runs in sweltering heat or very high wind, it does not make much of a difference, while Runalyze is quite variable. I think Garmin may be using a moving average over a couple of weeks to avoid fluctuations after (nearly) every run, due to different environmental circumstances.

  • I think Garmin may be using a moving average over a couple of weeks to avoid fluctuations after (nearly) every run, due to different environmental circumstances.

    I agree. What would be utterly stupid since training status is completely based on those fluctuations. Now that VO2max barely moves, training status is stuck at "maintaining" or "recovery". When it changes to "productive" it gets back to "maintaining" the next day.

    I guess the Training Status indicator has lost its relevance... At least I'm not really giving it any credit anymore. But given the price of a Fenix 7 it's a quite expensive feature, especially if it's useless Smiley

  • Now that VO2max barely moves

    Once a level that is specific to an individual is reached VO2max will barely move. It’s physiologically impossible for it to keep climbing. Greatest gains are achieved by the less fit but then diminish as the max achievable is reached. Once that level is reached fitness gains then depend on how well the individual utilizes the oxygen the body absorbs. I’ve been active virtually all my life (6 decades worth). My VO2 max has barely moved for a very long while but my pace at VO2max has decreased as my fitness has dropped. 

  • You are right, there is a physiological limit to an individual's VO2max - but I think that is not the reason why Garmin's reported number is so steady. If you look at the curve, over the last 4 months (After I ran CCC last summer - the dot at 49.00 or so) my VO2max has been slowly and progressively decreasing, which got me "maintaining" TS most of the time - while the average value is, on the opposite, slowly increasing - and I can clearly feel I'm getting back in good shape as my HR/pace ratio on easy runs is widely improving.
    What I'm saying is, I think that Garmin has made some changes to the algorithm to make the figures look more consistent (which make sense, because trail activities VO2max, for example, are so difficult to evaluate that they need to use some sort of background data) - and now maybe the average calculation they are using also filters small fluctuations that used to be relevant.