HRM Calorie Counts --Some Real World Data (uber long post)

mirandan73189

Excellent post and very informative. You made a lot of this accessible for some who are confused about physiology and terms.

EvgeniZyntx

Suggested reading on VO2Max determined by HRM using firstbeat tech: https://www.firstbeat.com/app/uploads/2015/09/white_paper_VO2max_11-11-2014.pdf

Azdak

EvgeniZyntx wrote: »

I'm really busy this week and I'd like to give this thread a little more attention.

First comment - I assume your polar is using first beat technology algorithms - it is no longer the average HR alone but HR and change in HR. I wouldn't trust treadmill calories for a few reasons: because biomech gait changes quite a bit person to person, shifting from walking to run, etc...

I doubt very seriously if Polar is using any firstbeat technology--I think that's a Pepsi/Coke situation and Polar has never gone with outsiders or shared their technology with anyone as far as I know (which, I'll admit is limited about the corporate relationships).

There are definitely going to be inaccuracies in the treadmill numbers--after all, these are all estimates and there is a standard of error in everything. That being said, changes in gait are not that great, except at the extremes (there was a study in the past 2 yrs that looked at this and I believe my interpretation is correct, but my confidence in my memory is only about 50%. In any case, the subject has been studied before and the commonly accepted energy prediction equations do not use gait as a factor.

There have been a few--not many--studies that have looked at the ACSM energy-prediction equations that are the standard for this sort of thing. The one's I have seen suggest that the ACSM equations are very good for walking and for slower running, but start to overestimate as running speed increases, up to 15% with speeds greater than 7.5-8.0 mph. Add in the lack of resistance due to no wind, and faster runners are probably looking at a 15%-20% overestimate on a treadmill. That still is equal or better than the accuracy of an HRM. And since treadmill calorie numbers are based on the actual workload being performed, they are more consistent than HRM numbers which are subject to thermal stress, cardiovascular drift, etc.
I haven't posted that data yet, but I have a good graph from a 45 min steady-state workout in which I didn't change the workload at all, and my breathing did not change at all, but HR increased by 15-20 bpm during the workout.

The other potential inaccuracy with treadmill numbers is that, to my knowledge, there are no reliable equations for speeds from 4.2 to 5.0 miles per hour. So I am not sure how that is done. And if someone is running at speeds less than 5.0 mph, the numbers will be off as well, because the equations for walking and running are different.

All in all, however, these issues are tangential to the point of the posting, which is that, even though aerobic work intensity increased by 23%, the HRM numbers did not change at all. Somehow VO2 has to be factored in.

Azdak

ConicalFern wrote: »

Azdak wrote: »

blues4miles wrote: »

Azdak wrote: »

The key is looking at my average HR -- it was the same for each workout. Since the Polar only monitors HR, it assumed I was working at the same level, only for fewer minutes. But, based on what I was able to accomplish, my performance clearly improved.

<snip>

They have their own margin of error, but if my treadmill says I burned 15 cal/min on one workout (2/5) and 18.5 cal/min (2/24) on another, that represents a substantial increase in both aerobic work and calories burned, even if those actual numbers are not 100% accurate.

Yet that performance increase did not show up in my HRM numbers. Why?
HRM calorie counts are based on the fact that, during aerobic exercise, there is a consistent relationship between heart rate (HR) and oxygen uptake (VO2). If intensity and VO2 go up, so does heart rate. If they go down, HR goes down as well. (You can clearly see that in the HR graphs for my workouts, since they were interval-type workouts alternating higher effort and lower effort).

Just taking aside this little part, and maybe I am not understanding you correctly, isn't that how it's supposed to work?

-HRM measures your HR and uses some age-based calculation (and your weight) to determine calories. (My Garmin includes some extra info for speed/distance and elevation as well I believe, but is mostly based on my HR)
-Treadmill uses just your weight, and the distance you covered and the time you covered it.

So of course the treadmill would say you burned more calories when you ran faster on two same-distance workouts. But if your heart rate was the same, because your fitness was improving, why would you expect the HRM to report higher calories? It shouldn't. Your body has adapted to the workouts you were doing before. You've improved, but you're now burning fewer calories at the same level of effort you were before.

Again, sorry maybe I am not understanding. I track my average heart rates over runs. As I get better I expect them to either get lower or stay the same. If I were to keep running the same workout over and over, eventually my average heart rate should get lower for that same distance and speed. But because we get more fit, we can run faster at the same heart rate. The treadmill doesn't know your level of effort. And just my personal opinion, but I don't like any calorie burn that gives us mortals more than 10 cals/minute. Obviously people have different experiences with that. But I know if I tried to eat back 12, 15 or even your hypothetical 18 cals/minute I would gain weight. I do think a very well trained athlete can burn more, maybe you are that athlete.

That's all I got for now, I'll try to read over more later for a 2nd time. I do like the numbers game, am a fan of the numbers myself.

This isn't criticism, but your point about "efficiency" is exactly wrong. A lot of people--including people with fitness certifications--think the same thing. It's one of the reasons I write these posts--to correct that idea. The spread of HRMs has introduced this fallacy into the mainstream and it is difficult to correct.

To understand, you have to absorb the concept that the oxygen cost/calorie burn of an aerobic activity is built into the workload of the activity--speed, incline, watts, etc--NOT the heart rate. Heart rate is just a speedometer. I could give you a drug that raised your heart rate to 75% of maximum while you were sitting in a chair, but you won't not burn more calories, because you weren't doing any work.

So, for example, running 6 mph has an energy cost--about 10 METs. That's the same for everyone. You multiply 10 METs x your wt in kg and that's your calories/hour burn when you run 6 mph. Some who weighs 60kg will burn 600 cals/hr, an 80kg person will burn 800, and so on.

At the start of training, that 10 MET workload might represent 80% of your maximum oxygen uptake. So, when you run 6 mph and wear your HRM, the exercise HR will be 80% of your maximum heart rate (keeping it simple).

The whole point of training, however, is to INCREASE our aerobic maximum. So let's say you increase your max oxygen uptake 10%. Now, when you run at 6.0 mph, that workload of 10 METs only represents 75% of your aerobic max instead of 80%. Since your maximum heart rate has not changed, the exercise heart rate is now 75% of maximum, which will be lower.

Even though the heart rate is lower, the calorie burn is still the same (unless you lost weight). That's because the oxygen cost of 10 METs hasn't changed. It's just that now, 10 METs is a lower percentage of your aerobic maximum.

However, unless you can change it, the HRM doesn't know your maximum aerobic level has changed. HRMs can't measure the actual workload--they have to guess what you are doing based on the heart rate response. So the HRM assumes you are working at a lower workload and estimates fewer calories.

That has nothing to do with "efficiency". It's because your max fitness level has increased so that the same workload and same calorie burn requires less effort.

If you have a HRM that allows for manual input of VO2max and you change that number and increase it by 10%, your calorie burn will INCREASE for any given heart rate level. That's because you reset the scale.

Hope this helps a little. I understand that if one doesn't know the underlying physiology, it seems counterintuitive and the numbers on the HRM seem so tangible and "real" they're hard to ignore.

This is spot on in my opinion.

That said, I'm sticking with my Garmin calorie estimates as they are. I tend to eat back most of my exercise calories back, and I would rather under eat than over eat!

And, to repeat, this is not intended to criticize or question anyone's strategy, esp if that strategy is working for them. It's just fun geek stuff for anyone who is interested.

Azdak

blues4miles wrote: »

Azdak wrote: »

If you have a HRM that allows for manual input of VO2max and you change that number and increase it by 10%, your calorie burn will INCREASE for any given heart rate level. That's because you reset the scale.

Hope this helps a little. I understand that if one doesn't know the underlying physiology, it seems counterintuitive and the numbers on the HRM seem so tangible and "real" they're hard to ignore.

What V02max number does the HRM use to start with? How does it know the number it picked for you at the beginning is real? What if your V02max at the start is terrible because you are an unfit beginner but the calculation/average your HRM is using is what you eventually work up to? I am willing to believe efficiency has nothing to do with it but don't understand how you are coming to the conclusions that the early numbers are correct and the later numbers are bunk rather than the opposite.

Azdak wrote: »

Not really sure why you felt the need to become judgmental about the information presented (they aren't arguments, because this is settled science). Hard to understand how people can have strong opinions about things when they don't really understand the technology or the science.

I didn't know I was coming across as judgmental. I thought you were posting for discussion. If you were posting for me to simply agree with you, sorry, I'll move on. If this is settled science you can certainly post the research for future jerks like me who come in here and ask stupid questions. I'll see myself out.

I'm happy to answer questions, and I don't require that people be "experts" to contribute to the discussion. This is a posting for discussion. If you go back to your original question, you will see that I answered it in detail, and I was happy to do so.

However something like this:

"But I'm betting these formulas would give you guys much lower numbers than your HRM is giving you so that's why you don't like them. So you guys are saying you think the HRM is accurate at the beginning but then becomes less accurate over time because you don't like that effort/challenge affects your calorie burn."

And basically say that those who disagree with you are doing so because they want to create bogus numbers that make happy, that IS judgmental.

But, hey, not everything comes across as intended when written, so let's go back to your starting question.

You are introducing a different topic, which is fine, but it's not really the subject at hand. I am discussing relative data points, I.e. Fitness level on day A vs fitness level a week, month, etc later. The point is, whatever your fitness level is, if it improves and that improvement is not captured by the HRM, the decreased calorie count number is a) NOT due to "efficiency" and b) not actually decreased. The lower calorie count on the HRM is due to the inherent shortcomings of the HRM design, not because of any physiological change. If you weigh the same and you run 6.0 mph at a heart rate of 150, and a month later you run that same 6.0 speed, but at a heart rate of 142, you are burning the same number of calories, no matter what it says on the HRM.

Your question about how the HRM knows your initial fitness level and how that affects OVERALL calorie accuracy of the HRM is a good one. The answer is I don't really know. There are ways you can program an HRM to calculate those things and I'm sure Polar has some proprietary algorithm that does so. How accurate that is is anyone's guess. Again--and I don't mean this in a negative way at all--it's not relevant to this discussion.

EvgeniZyntx

Azdak wrote: »

EvgeniZyntx wrote: »

I'm really busy this week and I'd like to give this thread a little more attention.

First comment - I assume your polar is using first beat technology algorithms - it is no longer the average HR alone but HR and change in HR. I wouldn't trust treadmill calories for a few reasons: because biomech gait changes quite a bit person to person, shifting from walking to run, etc...

I doubt very seriously if Polar is using any firstbeat technology--I think that's a Pepsi/Coke situation and Polar has never gone with outsiders or shared their technology with anyone as far as I know (which, I'll admit is limited about the corporate relationships).

There are definitely going to be inaccuracies in the treadmill numbers--after all, these are all estimates and there is a standard of error in everything. That being said, changes in gait are not that great, except at the extremes (there was a study in the past 2 yrs that looked at this and I believe my interpretation is correct, but my confidence in my memory is only about 50%. In any case, the subject has been studied before and the commonly accepted energy prediction equations do not use gait as a factor.

There have been a few--not many--studies that have looked at the ACSM energy-prediction equations that are the standard for this sort of thing. The one's I have seen suggest that the ACSM equations are very good for walking and for slower running, but start to overestimate as running speed increases, up to 15% with speeds greater than 7.5-8.0 mph. Add in the lack of resistance due to no wind, and faster runners are probably looking at a 15%-20% overestimate on a treadmill. That still is equal or better than the accuracy of an HRM. And since treadmill calorie numbers are based on the actual workload being performed, they are more consistent than HRM numbers which are subject to thermal stress, cardiovascular drift, etc.
I haven't posted that data yet, but I have a good graph from a 45 min steady-state workout in which I didn't change the workload at all, and my breathing did not change at all, but HR increased by 15-20 bpm during the workout.

The other potential inaccuracy with treadmill numbers is that, to my knowledge, there are no reliable equations for speeds from 4.2 to 5.0 miles per hour. So I am not sure how that is done. And if someone is running at speeds less than 5.0 mph, the numbers will be off as well, because the equations for walking and running are different.

All in all, however, these issues are tangential to the point of the posting, which is that, even though aerobic work intensity increased by 23%, the HRM numbers did not change at all. Somehow VO2 has to be factored in.

Apparently the VO2Max estimator of the Polar V800 doesn't give great results for some people.
See the con section: http://fellrnr.com/wiki/Polar_V800

Thermal stress and CV drift ... oh, boy, big subjects.

robertw486

@Azdak

A great overall view of how HRM type devices could be skewed by variables. I personally think they are good for certain things, but as with most devices not good for the overall picture.

As a side, I'm not sure if the VO2max setting alone would fix the examples you gave. Though I do agree that HR is a great gauge and somewhat an indicator of load, I've seen a number of instances where at certain levels this doesn't hold very true, as the progression isn't as linear between the load and HR. For me due to all the variables in DIY testing methods, really haven't worried about nailing down a frequent VO2max number. Being that it can change some quickly and even day to day, I find it a good overall measure, but not an absolute.

On my elliptical it displays no HR average, but it does display calories per minute. What I find interesting is that at times and certain loads, I can bump load and thus calories per minute with very minimal impact on heart rate. At other times, seemingly somewhat related to cadence vs ramp vs resistance, the calories per minute change in overall load relates very closely to heart rate.


It would be interesting to have machines that have absolute measures of power, along with the charting capabilities that your HRM and software have. The two combined on a single chart would make it really easy to look at where the loads impact the most, and where HR can support higher loads without much change.