Estimated Calories & HRM . . . a little insight, from my POV

CincyNeid

I started my weight loss journey in 2014, in which time I've Tampered with Endomondo [which runs the same logarithms as the MapMyFitness programs], Strava, Gamin Connect and Training Peaks.

I see a lot of people asking about HRMs, which online program/fitness tracker to use, and such. . . And I see a lot of people posting that these are only estimates. And they are. Purely Estimates. There is no way for them to tell you exactly what you burned.

And Since I've tampered with all of those various programs, I have them all sync'd together. So I can get different metrics, from different sites and compare them, purely for research purposes.

Well today I ran a 5K. In the Disabled American 5K in Cincinnati. It was the first 5K I ran in 3 years ago, so it has a special place in my heart.

AS far as My equipment used for said event I use my VivoActive HR, which is GPS and GLONASS equipped, and I used a Wahoo Fitness Tickr Run HRM, connected to my VivoActive HR via ANT+. And for what it's worth I did have ElectroGel on my HRM for optimal readings.. .. ..

According to Strava I burned 622 calories over the course of my 5K. I ran a 9m53s mile. Which isn't bragging worthy, nor is it shabby.

But according to Under Armour, Endomondo to be exact, I burned 563 calories, from the same .gpx file.


So, according to Strava I burned 200.64 kCal per mile, and according to UA I burned 181.61 kCal per mile ran. 19.03 kCal difference per mile between the two programs.

Now if really want to throw a wrench in the mix. According to Garmin Connect , I ran a 9m 48s mile. And burned 563 kCal.




So according to GC I burned the same amount of calories, but at a faster pace...

So when we say that Caloric burn is purely an estimate. It really is. Even when using a Chest Mounted HRM.


And all of the workouts are linked by name to their respective source for verification, so you know I'm not photoshopping anything.....

NorthCascades

Thanks for bringing your own experience and some real-world data to the table. That's always a lot more valuable than "this is how it is, just take my word for it."

Any idea why Garmin and Strava have different times? Is it an auto-pause kind of thing?

It looks like you burned 563 or 622 calories, from the calculations available to you. The difference is a little over 10 %, I feel like it's not that much.

rileysowner

Yes, they are estimates, but they are informed estimates. The danger is thinking they are way off, but even the difference between the highest calorie number and the lowest you are talking less that 100 calories difference. There will be more variation in the accuracy of the calorie numbers you eat. In other words, the estimates are not exact, but they certainly are good enough.

NorthCascades

I think high end Garmins and Suuntos do a pretty good job from HRV. I don't expect perfection but I think they get pretty close, enough to work with.

Here's some data I can share. These are two rides I did that felt very close in terms of perceived effort, and I feel like they took about the same amount of energy.

Last year I rode up and down Washington Pass (paved highway) on a road bike with a power meter. It took 2:54, and my average HR was 133 bpm. This was 1,683 kJ.
https://connect.garmin.com/modern/activity/913079039

The next day, I rode up and down Cutthroat Pass (singletrack) on a mountain bike. It took 3:17, and my average HR was 139 bpm. This was estimated at 1,650 kCal.
https://connect.garmin.com/modern/activity/914030386

I feel like that's pretty good. I can't prove it, but I'm satisfied. For the record, the estimate came from a Garmin Fenix 3 and I wore a chest strap HRM.

rileysowner

NorthCascades wrote: »

I think high end Garmins and Suuntos do a pretty good job from HRV. I don't expect perfection but I think they get pretty close, enough to work with.

Here's some data I can share. These are two rides I did that felt very close in terms of perceived effort, and I feel like they took about the same amount of energy.

Last year I rode up and down Washington Pass (paved highway) on a road bike with a power meter. It took 2:54, and my average HR was 133 bpm. This was 1,683 kJ.
https://connect.garmin.com/modern/activity/913079039

The next day, I rode up and down Cutthroat Pass (singletrack) on a mountain bike. It took 3:17, and my average HR was 139 bpm. This was estimated at 1,650 kCal.
https://connect.garmin.com/modern/activity/914030386

I feel like that's pretty good. I can't prove it, but I'm satisfied. For the record, the estimate came from a Garmin Fenix 3 and I wore a chest strap HRM.

If you have a power meter on a bike it is pretty much as accurate as you can get outside of a lab. If I could afford one for my bike, I would simply use its numbers. They are a little pricey for my income.

pedermj2002

For my money, I'm not very pleased with the VivoActive HR+. If I enable the heart rate monitoring on it (wrist worn, optical), then Garmin will claim I burned through about 4000 calories every day, and there's just no way that's right. I'm actually thinking of disconnecting Garmin from MFP, and just manually entering in step counting into MFP.

For the other parts: Strava (with no extra sensors) can't be 100% accurate, but it's been good enough for me. If I use it's calorie estimates, it manages to do pretty well for me in tracking my exercise, which means I'm losing weight at the rate I expect.

It can't be perfect, but it can be good enough, and it definitely is that.

MeanderingMammal

rileysowner wrote: »

The danger is thinking they are way off, but even the difference between the highest calorie number and the lowest you are talking less that 100 calories difference.

Prompted by this I took a good look at todays run. It was a reasonably technical run, only 12km of easy paced running with a reasonable elevation gain. The surface wa quite wet, and I was using Mud Claws.

I have a similar situation, My *.gpx file gets uploaded to Garmin Connect, and then distributed to Strava, Endomondo, Runkeeper, Training Peaks and MFP. There was significant difference across the data types, although with the HR trace there was a significant deviation. Being a trail run pace varied, elevation varied a lot and there is noticeable cardiac drift throughout the session.

Average pace had a deviation of 10secs per km,

Garmin Connect: Cals 821 Elevation 144metres Ave HR 156 Max HR 193
Strava: Cals 1180 Elevation 84 metres Ave HR 155Max HR 203
Runkeeper: Cals 980 Elevation 219 metres Ave HR 156 Max HR 190
Endomondo: Cals 821 Elevation 121metres Ave HR 156 Max HR 193
MFP: Cals 821

No big surprises about any of those, the nature of the elevation deviation reflects the terrain, an initial 40 metre climb followed by lots of 6-8 metre climbs. I'd also note that GPS elevation is the least reliable element of consumer grade devices, although military engineering kit can be much more accurate.

CincyNeid

rileysowner wrote: »

Yes, they are estimates, but they are informed estimates. The danger is thinking they are way off, but even the difference between the highest calorie number and the lowest you are talking less that 100 calories difference. There will be more variation in the accuracy of the calorie numbers you eat. In other words, the estimates are not exact, but they certainly are good enough.

Okay let's play a game.

You're building a house. You're off 5% here, 12% Off there.... What kind of results do you expect?

Yes they're estimated numbers. But many want to lay claim that they're gospel truth.... And you cannot do that.

rileysowner

CincyNeid wrote: »

rileysowner wrote: »

Yes, they are estimates, but they are informed estimates. The danger is thinking they are way off, but even the difference between the highest calorie number and the lowest you are talking less that 100 calories difference. There will be more variation in the accuracy of the calorie numbers you eat. In other words, the estimates are not exact, but they certainly are good enough.

Okay let's play a game.

You're building a house. You're off 5% here, 12% Off there.... What kind of results do you expect?

Yes they're estimated numbers. But many want to lay claim that they're gospel truth.... And you cannot do that.

But you are not building a house, and the variation in calories from the nutrition information can be up to 25%

CincyNeid

rileysowner wrote: »

CincyNeid wrote: »

rileysowner wrote: »

Yes, they are estimates, but they are informed estimates. The danger is thinking they are way off, but even the difference between the highest calorie number and the lowest you are talking less that 100 calories difference. There will be more variation in the accuracy of the calorie numbers you eat. In other words, the estimates are not exact, but they certainly are good enough.

Okay let's play a game.

You're building a house. You're off 5% here, 12% Off there.... What kind of results do you expect?

Yes they're estimated numbers. But many want to lay claim that they're gospel truth.... And you cannot do that.

But you are not building a house, and the variation in calories from the nutrition information can be up to 25%

I never questioned that. Not once. However there are many many people who quote their caloric burn, and intake as gospel. You have to have have give and take room.

lorrpb

So what is the value of HRM unless you're looking for serious HR training data? I've never see the point for calorie estimation only.

MeanderingMammal

lorrpb wrote: »

So what is the value of HRM unless you're looking for serious HR training data? I've never see the point for calorie estimation only.

To an extent, as long as one is using the same method consistently and is aware of the limitations, it can give a workable estimate. There are lots of "ifs, buts and maybes" around it but as long as one is thoughtful about it then it's not too bad.

Personally I just use the Garmin to MFP figure as a guideline, and compensate according to how things looked in the real. From my example above, the Garmin figure was probably a bit low so I could have added an extra 150 cals if I'd wanted to. Other times it'll feel a bit inflated.

For me, the benefits of HR are around recognising the underlying performance issues. Is it responding as I'd expect to the inputs, or is there something suggesting overtraining? To an extent it also plays into an individual session. It's easy to go out too fast, particularly in a race. I can sustain 170bpm for an hour but wouldn't be able to do that for the 4 hours of a marathon, so it's useful to keep an eye and wind back if I need to.

As upthread, the pool of error has a number of different contributors.

sijomial

lorrpb wrote: »

So what is the value of HRM unless you're looking for serious HR training data? I've never see the point for calorie estimation only.

@lorrpb

Getting a reasonably accurate (not entirely accurate) calorie estimate can be important for people doing a high volume of cardio. Especially a variable volume where your calorie needs vary enormously day to day or even season to season.
It obviously has flaws but with a bit of effort the estimates can certainly be reasonable/usable.

For indoor training I use a HRM calibrated to my personal VO2 max and tested max HR - it comes out very, very close to an expensive power meter equipped training bike under suitable conditions.
It can also vary wildly if those conditions aren't suitable, that's where common sense comes in.