How does Garmin Connect calculate VO2 Max?

17 watchers

Prev
1
2
3
4
5
6
7
Next
Last

Canute Aug 2016 12:05am, 29 Aug 2016 1,751 posts	0 Canute The FirstBeat technology (which Garmin use) computes VO2 max from HR at a measured submaximal pace and an estimated HRmax. A similar calculation can be doen from Daniels formula based on pace and HR/HRmax. There are two clever things about the way FirstBeat do it. 1) The select only those sections of the HR trace that appear to fit the expected relationship between HR and speed (eg they exclude section with artefact or where HR is atypical because of hills or terrain. 2) They use a non-linear algorithm for the relationships between HR, HRmax, pace and VO2max. This algorithm is based on data from a large number of athletes. Provided your HRmax estimate is accurate and you are a typical representative of the runners used to derive the algorithm, this non-linear algorithm will give an accurate answer for VO2max to within about 5%. If you are prepared to use a section of your HR trace which represents steady running at a measured pace on a level surface, and is free of visible artefacts, you can get a fairly good estimate from Daniels formula, which was also based on observations of many runners, but it uses a simple quadratic approximation rather than the algorithm used by FirstBeat. If you are familiar with Excel, it is fairly easy to prepare a spreadsheet. The required formulae are: VO2 = -4.6+0.182pace+0.000104pace^2 (pace in metres/min) %VO2max= (HR/HRmax -0.37)/0.64 VO2max=VO2 / %VO2 (VO2max in ml/min/Kg) In many cases the greatest uncertainly comes from the estimate of HRmax, but that applies to both the Daniels formula and First Beat.
Canute Aug 2016 9:06am, 29 Aug 2016 1,752 posts	0 Canute Sorry for the typos and other inaccuracies. ‘Done’ not ‘doen’ The first numbered paragraph should be: ‘1) They select only those sections of the HR trace that appear to fit the expected relationship between HR and speed. They exclude sections with artefacts or where the relationship between HR and pace is atypical because of hills, surface or wind’ However, the constant -4.6 in the equation for VO2 is correct. The combination of linear and quadratic terms slightly overestimates the best fit to the data across the full aerobic range. The fit is improved by subtracting a small constant amount. As a consequence, the formula is inaccurate at ‘tortoise paces’ less than 30 m/min (i.e. about 1 mile/hour).
fetcheveryone Jan 2021 3:06pm, 8 Jan 2021 25,338 posts	0 fetcheveryone I've updated our VO2Max calculator - you can put any training run in, along with your average heart rate on that run - and your age and weight - and it will give you an estimate. fetcheveryone.com/training-calculators-vo2max.php
Sigh Jan 2021 3:13pm, 8 Jan 2021 8,257 posts	0 Sigh Nice. Here's mine: Garmin told me 45, Runalyze provides that to two decimal places from the Garmin file at 45.09.
fetcheveryone Jan 2021 3:27pm, 8 Jan 2021 25,339 posts	0 fetcheveryone The calculator is scoring me at 51.2 for this run: fetcheveryone.com/t-18536043 Garmin is currently giving me 49. If I lop the first mile off that run, during which my HR was still working its way up, the calculator gives me 50.9. If I lop the second mile off as well, the remainder is at a largely consistent heart rate - and it gives me 49.6.
larkim Jan 2021 3:31pm, 8 Jan 2021 13,135 posts	0 larkim Mine seems less consistent with RA. Is it correct that VO2 max drops as weight increases? That may be physiologically correct, I just don't understand it! 5.54mile run, 144bp, same weight and age, comes out at 58.4 for a male. RunAnalyze would give 49.88 RA could well be wrong, though that's quite a big gap. I don't get VO2Max from my watch per run, so can't compare that. Is this estimating what your actual VO2Max should be if you were at full tilt? I suppose it is, if it's accepting a low HR; as the HR scales up the time should scale down in a fairly predictable way?
SPR Jan 2021 3:38pm, 8 Jan 2021 32,877 posts	0 SPR What is age doing in the calculator? Is it estimating a max HR or something else? In regards to weight I thought if your performance remains the same as weight goes up the VO2 max has stayed the same?
larkim Jan 2021 3:40pm, 8 Jan 2021 13,136 posts	0 larkim This fetcheveryone.com/t-17248828 earns me 62.9 with the first mile dropped for HR purposes, but RA only gave me 54.9! I like your numbers much more!! Though this brianmac.co.uk gives me 54.4 and as that activity was a race it should be closer to being "correct" in some sense, shouldn't it?
SPR Jan 2021 3:43pm, 8 Jan 2021 32,878 posts	0 SPR It's got to be assumptions based on weight and HR throwing it off. Runalyze calculates effective VO2 max, essentially based on HR but with a correction factor based on races you've done.
fetcheveryone Jan 2021 3:47pm, 8 Jan 2021 25,340 posts	0 fetcheveryone There's a link to a white paper in my blog from a few days ago.