Difference in calorie burn for different HRMs
Options
MarttaHP
Posts: 68 Member
Has anyone else seen a significant difference in the calorie burn reported by two different sports watches/activity trackers based on HR measurement? If so, how did you deal with it?
I realize HRMs only measure HR and not calorie burn, and the latter is only an estimate. However, I recently purchased a Garmin watch to replace my old Polar, both with wrist-based HR measurement. I've been wearing both of them on my runs just for comparison, one on each wrist. The Garmin gives me a considerably lower calorie burn estimate every time. For example, for a 1-hour, steady-state, low-effort run recently the Garmin gave me 463 kcal, while the Polar showed 627 kcal. That's a 164 kcal discrepancy!
I realize HRMs only measure HR and not calorie burn, and the latter is only an estimate. However, I recently purchased a Garmin watch to replace my old Polar, both with wrist-based HR measurement. I've been wearing both of them on my runs just for comparison, one on each wrist. The Garmin gives me a considerably lower calorie burn estimate every time. For example, for a 1-hour, steady-state, low-effort run recently the Garmin gave me 463 kcal, while the Polar showed 627 kcal. That's a 164 kcal discrepancy!
0
Replies
-
Did they both give the same (or very similar) average HR?
(To confirm they are both picking up your pulse correctly.)
I assume you have checked your stats/personal settings are identical in both?
If using the same estimating algorithm (they may not be, worth researching) is one perhaps a gross calorie estimate (Polar) and the other a net calorie estimate (Garmin). I know Garmins give net calories when using a power meter for cycling but not sure what they attempt to do for HR based estimates.
If you multiplied your weight in pounds X 0.63 efficiency ratio X miles run that should give a reasonable net calorie estimate for level ground, "normal speed" running.
May not be appropriate for your circumstances but I callibrated my over-estimating Polar FT60 HRM by validating against an accurate measure (power meter) and changed my stats/settings in the watch until the two got close to matching. That only worked for steady state training though - the HRM over-estimated by c. 25% for interval training.2 -
Did they both give the same (or very similar) average HR?
(To confirm they are both picking up your pulse correctly.)
Yes - Polar shows an average HR of 117 and Garmin 118. If anything, the Garmin seems to be giving me higher values throughout, especially toward the beginning of the run.I assume you have checked your stats/personal settings are identical in both?
I have indeed. Weight, height, age, gender, both are the same for each.If using the same estimating algorithm (they may not be, worth researching) is one perhaps a gross calorie estimate (Polar) and the other a net calorie estimate (Garmin). I know Garmins give net calories when using a power meter for cycling but not sure what they attempt to do for HR based estimates.
When I first got the Polar, I was trying to find information on what kind of calculations they actually use to arrive at all sorts of values, but wasn't able to find anything. Garmin, or rather Firstbeat on whose technology their devices are based, seems to be more open about their methods.
I'm quite sure both estimates are gross. Polar doesn't state this explicitly, but when I used to sync the calorie data with MFP, it showed up in the Exercise tab as a lower number than stated by my watch - i.e. I assumed what ended up on MFP were the net calories. The Garmin, on the other hand, syncs with MFP with the same values as estimated by my watch, but if I enable negative calorie adjustments, MFP subtracts another good chunk out of my overall Exercise calories by the end of the day. On Garmin Connect, it says that RMR calories are subtracted from the active calorie burn of the day.If you multiplied your weight in pounds X 0.63 efficiency ratio X miles run that should give a reasonable net calorie estimate for level ground, "normal speed" running.
I've wondered about this equation - does the effort of the run not come into play at all? Or is the difference between different running paces so small so as not to matter in the calorie calculation?May not be appropriate for your circumstances but I callibrated my over-estimating Polar FT60 HRM by validating against an accurate measure (power meter) and changed my stats/settings in the watch until the two got close to matching. That only worked for steady state training though - the HRM over-estimated by c. 25% for interval training.
Thanks for the tip. Do stationary bikes usually come with power meters? I've never used one, but my gym has them.0 -
Pace does have an impact but more pronounced at the extremes of very slow and very fast.
Overall, it's a reasonable estimate not an absolute but one based on physics rather than physiology to give a different metric for comparison. The effort in running distance would make a difference in the physics sense (more elevation, different conditions underfoot, running into a headwind...), effort in terms of feelings doesn't make a significant difference.
e.g. a fit, regular runner doing the same mileage as an unfit, new to running person of the same weight doing the same course. Obviously, people have some variation in their running efficiency but not typically a huge amount.
It's one of the failings of estimating energy by heartrate of course - a fit person will typically have a lower HR doing the same exercise. Personally, my calorie burns for the same HR went up by about a quarter over an extended period of time.
Power meters are becoming more and more common in gym bikes. Power based training has been used by the elite and high-end cyclists for quite a while but as costs have come down the popularity has gone up massively in the last ten years. Even old duffers like me use them!
Only caveat is that although the maths for converting power to energy are very well known and the numbers are very accurate due to the small variation in pedalling efficiency some gym manufactures still insist on playing games with the numbers ("Use my bike - burn more calories!") .
Average watts per hour X 3.6 = net calories with very high accuracy.
I often link my Garmin to the indoor trainer for just that reason and it uses that formula whereas the gym bike uses a most bizarre algorithm designed to flatter despite being a high end, high accuracy, dual power meter trainer.
Conversely some basic gym bikes with power meters give accurate calories.
1 -
Has anyone else seen a significant difference in the calorie burn reported by two different sports watches/activity trackers based on HR measurement? If so, how did you deal with it?
I realize HRMs only measure HR and not calorie burn, and the latter is only an estimate. However, I recently purchased a Garmin watch to replace my old Polar, both with wrist-based HR measurement. I've been wearing both of them on my runs just for comparison, one on each wrist. The Garmin gives me a considerably lower calorie burn estimate every time. For example, for a 1-hour, steady-state, low-effort run recently the Garmin gave me 463 kcal, while the Polar showed 627 kcal. That's a 164 kcal discrepancy!
I would be inclined to trust the lower value (I've been using Garmin GPS sports watches for over a decade). Is it just a HRM? How far did you run in that hour?
You can sanity check the estimates using the formula .63 x weight in lbs x distance in miles. This will give you an approximation of net calories expended.
1 -
I ran 7 miles, so at my weight of about 130, that's 573 kcal. So it's closer to the Polar estimate according to that formula.
I just realized that the Garmin on the other hand looks to be overly optimistic about my calorie burn when walking. During a 1-hour walk it claimed I burned 271 kcal. Siiiigh. Why can't technology be perfect?0 -
I don't know, but it's not perfect.
I had a Polar for a while, now I just use a flat number for walking and a little higher for hills.
I use 250 for a one hour flat walk, and I'm 140 lbs... that number has always been close enough to give me the weight control.
The other obvious issue is your food logging accuracy. Food logging is a lot easier to calculate than exercise, IMO.
Here's a good starter for that:
https://community.myfitnesspal.com/en/discussion/1234699/logging-accurately-step-by-step-guide/p10 -
Not sure if this has been mentioned already, but wrist HRM are not specifically measuring heart rate but blood flow/capillary flow; they are impacted by many things including how tight you wear it! I personally use a MyZone HRM which seem to be one of the best on the market at an affordable price. They use the following formula..... 211 - ( age x 0.64) for max heart rate and then a more detailed one for cal burn but it is the accuracy of the heart rate that is key ..... which works for most sports - I use mine predominantly for spinning and hiking.
MyZone is chest mounted and claims to be near- ECG level of accuracy; wrist monitors will never be that accurate and thus cal burn won’t be either.
I am a bit of a MyZone advocate so happy to chat off line if more detail needed.0 -
Not sure if this has been mentioned already, but wrist HRM are not specifically measuring heart rate but blood flow/capillary flow; they are impacted by many things including how tight you wear it! I personally use a MyZone HRM which seem to be one of the best on the market at an affordable price. They use the following formula..... 211 - ( age x 0.64) for max heart rate and then a more detailed one for cal burn but it is the accuracy of the heart rate that is key ..... which works for most sports - I use mine predominantly for spinning and hiking.
MyZone is chest mounted and claims to be near- ECG level of accuracy; wrist monitors will never be that accurate and thus cal burn won’t be either.
I am a bit of a MyZone advocate so happy to chat off line if more detail needed.
Anything that uses an equation for estimating calorie burns is just that: an estimate. 220-age is fine for about half of an average population, but the other half is at least one standard deviation away from those equations, and thus gets too high or too low burns. The equation you mention is a bit better for people with a higher maxHR, but bad news for a lower one.3 -
Polar had one of their early equations as part of a research study so it's publicly available.
They have tweaked it many times since then.
Garmin on some sport specific devices uses a FirstBeat algorithm which is also tweaked off a research study based equation.
But that is not used on all devices at all - I've only looked over info on a few daily trackers to get one - and didn't noticed FirstBeat mentioned on any of them.
Which means they could be using the base public study formula, but with none of the tweaks FirstBeat does.
So first off - yes they do use different ways of calculating calories.
@sijomial - my FR310XT still reports a calorie burn that has nothing to do with the power meter KJ or watts - it's always almost 2x higher - I have to manually correct the GC entry down to the better value.
Could that be different on bike specific computer? Or is that activity tracker paired with the bike specific devices?
I display the KJ & calories on 4th screen just for grins, always off the 2x.
@marttaHP - when devices use HR-based formula for calorie burn, the range (bottom of aerobic where walking could pop up into, and the upper moving into anaerobic) starts showing inflated calorie burn on the bottom and top.
Distance based would be better - but the walks are probably bumping into HR-based, hence the inflation.
Whereas for the harder run effort - it's in the middle of the range and can be more correct (with caveats to inflated HR due to stress, heat, dehydrated, ect), then better calorie burn.
The reason for it being lower than distance based burn calculated is probably because Garmin is guessing about your HRmax, and therefore the level of your effort.
You may have a lower HRmax than it's using, so it appears to be easier effort.
Whereas Polar may be doing double HR and distance based burns put together.
Haven't read up on them for a long while to see if they make any such claims - it sounds like you found nothing from them though.0 -
@heybales
I'm using a cycling specific Garmin (Edge 1000) which probably explains why it uses the power meter data correctly.
Garmin seem to have different selections of Firstbeat function for different platforms. I don't get the Fitness Age functionality which other Garmins do.0 -
MyZone is chest mounted and claims to be near- ECG level of accuracy; wrist monitors will never be that accurate and thus cal burn won’t be either.
I've been considering buying a chest strap monitor, mostly because the random fluctuations and inexplicable spikes in HR given by my watch are pretty annoying. But I just bought this Garmin, and can't really afford to get a strap right now, so maybe down the line.Anything that uses an equation for estimating calorie burns is just that: an estimate. 220-age is fine for about half of an average population, but the other half is at least one standard deviation away from those equations, and thus gets too high or too low burns. The equation you mention is a bit better for people with a higher maxHR, but bad news for a lower one.
Indeed; my max HR is at least 10 bpm lower than the age-based equation would suggest. If I went by the equation, I'd be constantly overexerting myself.Garmin on some sport specific devices uses a FirstBeat algorithm which is also tweaked off a research study based equation.
But that is not used on all devices at all - I've only looked over info on a few daily trackers to get one - and didn't noticed FirstBeat mentioned on any of them.
Which means they could be using the base public study formula, but with none of the tweaks FirstBeat does.
My watch is the Forerunner 245, which according to Firstbeat's website does use their algorithms.The reason for it being lower than distance based burn calculated is probably because Garmin is guessing about your HRmax, and therefore the level of your effort.
You may have a lower HRmax than it's using, so it appears to be easier effort.
I've manually set my max HR at 176, which is much lower than the 220-age formula would give me. I've been running with HRMs for a couple of years now, and my HR has never gone above that number on interval runs, so I believe it to be fairly accurate for me. So yes, if I hadn't set the max HR myself, I'd definitely expect Garmin to be underestimating my calorie consumption.0 -
cmriverside wrote: »I don't know, but it's not perfect.
I had a Polar for a while, now I just use a flat number for walking and a little higher for hills.
I use 250 for a one hour flat walk, and I'm 140 lbs... that number has always been close enough to give me the weight control.
The other obvious issue is your food logging accuracy. Food logging is a lot easier to calculate than exercise, IMO.
Here's a good starter for that:
https://community.myfitnesspal.com/en/discussion/1234699/logging-accurately-step-by-step-guide/p1
Thanks, I've been logging my food for a while now and think I've got it pretty well under control.
I'm not looking to lose weight, just to maintain. It's more that I want to be able to fuel my workouts properly, which means that even 150 kcal difference can end up being pretty significant. Plus, it's just plain irritating having all these fancy expensive gadgets and whatnot that turn out not to be 100% reliable 100% of the time! 0 -
Oh, googling this just now, I did find a white paper by Polar about their calorie calculation formula. https://www.polar.com/sites/default/files/static/science/white-papers/polar-smart-calories-white-paper.pdf
That was published after I first got my Polar, so that's why I must've missed it before. I guess I'll just need to compare this to the Firstbeat white paper on their method and maybe finally get some answers!0 -
Oh, googling this just now, I did find a white paper by Polar about their calorie calculation formula. https://www.polar.com/sites/default/files/static/science/white-papers/polar-smart-calories-white-paper.pdf
That was published after I first got my Polar, so that's why I must've missed it before. I guess I'll just need to compare this to the Firstbeat white paper on their method and maybe finally get some answers!
Cool, thank you.
They don't get into their specifics, their secret sauce, as I thought. Just as FirstBeat discusses the underlying public study info but stops at their tweaks.
But they do share what they find acceptable and what was shown in some confirmation studies.
"However, in practice it seemed that the
chosen experimental protocol, differences between
laboratories and devices, and human errors while
doing the experiments all caused errors. In total the
error in reference data was around 5–10%, and in
some cases even more. Therefore we agreed that
energy expenditure estimates with less than 10%
difference from the reference measurements can be
considered to represent good accuracy, and estimates
with less than 20% difference are still acceptable. "
1
Categories
- All Categories
- 1.4M Health, Wellness and Goals
- 391.4K Introduce Yourself
- 43.5K Getting Started
- 259.7K Health and Weight Loss
- 175.6K Food and Nutrition
- 47.3K Recipes
- 232.3K Fitness and Exercise
- 390 Sleep, Mindfulness and Overall Wellness
- 6.4K Goal: Maintaining Weight
- 8.5K Goal: Gaining Weight and Body Building
- 152.7K Motivation and Support
- 7.8K Challenges
- 1.3K Debate Club
- 96.3K Chit-Chat
- 2.5K Fun and Games
- 3.2K MyFitnessPal Information
- 22 News and Announcements
- 922 Feature Suggestions and Ideas
- 2.3K MyFitnessPal Tech Support Questions