If HRMs are so unreliable for cals burned...

jjpptt2

... are we better off trusting the non-HRM calculations?

Example -
I've got a garmin watch. I can use a chest strap when I run/bike/whatever and it will monitor my HR, and use that along with a number of other data elements to calculate approximate calories burned.

-- OR --

I can use that same garmin watch without the chest strap, so there is no HR monitoring during my workout. The watch will use a number of data elements and approximate calories burned.

Assuming garmin uses some of the most reliable algorithms for calculating calories burned, is one going to be more reliable than the other?


I would assume HR-based would be more reliable for very even, steady state workouts. Agree? But what about non-steady state... for example runs where there are hills, but not specifically interval training.

I also assume this is going to vary greatly, and there probably isn't any 1 "right" answer. It was just something I thought about during my run this morning, and I thought it might make for a good conversation. There are some really knowledgeable people on this board about this topic, so...

MeanderingMammal

A lot of the value depends on the sophistication of the tracking devices. For a straightforward HR only device, I'd trust GPS based more than anything else, for a higher end Garmin that correlates HR with position and elevation I'm reasonably comfortable with the conclusions.

That's fine for something like running or cycling, but when you're talking about class based stuff when you don't have a consistent proxy to measure the isn't a good alternative. Low end devices aren't going to correlate elevated HR with vibration and come to a more informed view.

dewd2

There's a discussion on this subject here that I replied to (I'll have to dig it up - it was withing the past day or 2). It appears that Garmin is using some metrics from FirstBeat to better judge the energy use during runs when using the HRM. I was surprised to see the difference of my own runs.

In most cases I'd still argue that heart rate is nearly meaningless for counting calories. And even for Garmin this holds true since they are not really looking a the number of beats per minute to get the calculations. Plus, this only works for steady state cardio (in Garmin's case, only running).

dewd2

Here's the discussion along with a link I added to DCRainMaker's story about Garmin and Firstbeat.

http://community.myfitnesspal.com/en/discussion/comment/39686167#Comment_39686167

NorthCascades

It depends what you're doing. You gave two examples: cycling, and running (on a hilly course).

These are drastically different. With a bike, you can coast, even for dozens of miles at a time. You have access to a huge range of speeds that are not available to runners. (I can work really hard going 5 mph or 50 mph.) Garmins know absolutely nothing about the wind, which is the #1 thing slowing you down. GPS data alone doesn't have much to go on for calories on a bike, the other variables are too important.

When you run, you can't coast, and the wind can still help or hinder you but to a much lesser extent. Distance, elevation, and vertical oscillation make up a much bigger part of the picture for running.

XC skiing is another case where an HRM helps shed some light on what's going on. Snow conditions can be fast or slow, etc.

NorthCascades

dewd2 wrote: »

In most cases I'd still argue that heart rate is nearly meaningless for counting calories. And even for Garmin this holds true since they are not really looking a the number of beats per minute to get the calculations. Plus, this only works for steady state cardio (in Garmin's case, only running).

I did lots of tests with a modern Garmin running Firstbeat with a chest strap and comparing that to a power meter. The % error appears to be random and seems to have nothing to do with steady vs intervals.

jjpptt2

MeanderingMammal wrote: »

A lot of the value depends on the sophistication of the tracking devices. For a straightforward HR only device, I'd trust GPS based more than anything else, for a higher end Garmin that correlates HR with position and elevation I'm reasonably comfortable with the conclusions.

That's fine for something like running or cycling, but when you're talking about class based stuff when you don't have a consistent proxy to measure the isn't a good alternative. Low end devices aren't going to correlate elevated HR with vibration and come to a more informed view.

So given one of Garmin's better watches, and assuming it's an average person doing an average ride or run (not intervals, not hill repeats, not time trials, etc... just someone going out for a run or a ride) - would you more trusting of garmins calculations with or without the HRM?

jjpptt2

NorthCascades wrote: »

It depends what you're doing. You gave two examples: cycling, and running (on a hilly course).

These are drastically different. With a bike, you can coast, even for dozens of miles at a time. You have access to a huge range of speeds that are not available to runners. (I can work really hard going 5 mph or 50 mph.) Garmins know absolutely nothing about the wind, which is the #1 thing slowing you down. GPS data alone doesn't have much to go on for calories on a bike, the other variables are too important.

When you run, you can't coast, and the wind can still help or hinder you but to a much lesser extent. Distance, elevation, and vertical oscillation make up a much bigger part of the picture for running.

XC skiing is another case where an HRM helps shed some light on what's going on. Snow conditions can be fast or slow, etc.

I know they are different. We can break this conversation down into the finest of fine hairs, then split those hairs 10 different ways. But I was hoping to have a more, uh, "applicable to real life" conversation.

I'm a regular guy. I run a little bit, I bike a fair bit, I lift periodically, I do some other stuff here and there. But running and biking are my primary forms of exercise. If I wanted to make a reasonable effort to estimate calories burned, and I have one of Garmins better watches... given how unreliable every says HRMs are, am I better off just leaving the chest strap at home?

Some very preliminary tests suggest:

• road running without the chest strap gives me slightly higher calorie burn numbers, but still within reason if you assume the .65 * BW * distance formula to be the standard.
• road biking without the chest strap gives numbers much higher than expected... roughly 50% higher than what I would have expected.
• mountain biking without the chest strap gives me very low numbers, about 1/2 - 2/3 of what I would expect based on what I think I know after years of tracking/logging.

So my exceptionally early theory is that the HRM helps smooth out the error rate.

TavistockToad

jjpptt2 wrote: »

MeanderingMammal wrote: »

A lot of the value depends on the sophistication of the tracking devices. For a straightforward HR only device, I'd trust GPS based more than anything else, for a higher end Garmin that correlates HR with position and elevation I'm reasonably comfortable with the conclusions.

That's fine for something like running or cycling, but when you're talking about class based stuff when you don't have a consistent proxy to measure the isn't a good alternative. Low end devices aren't going to correlate elevated HR with vibration and come to a more informed view.

So given one of Garmin's better watches, and assuming it's an average person doing an average ride or run (not intervals, not hill repeats, not time trials, etc... just someone going out for a run or a ride) - would you more trusting of garmins calculations with or without the HRM?

in my experience, my Garmin 225 (wrist HRM) is pretty close to the bodyweight x 0.63 x distance calculation for running.

using real life data is the best way to calculate, regardless of which gadgets you have to play with.

Azdak

jjpptt2 wrote: »

NorthCascades wrote: »

It depends what you're doing. You gave two examples: cycling, and running (on a hilly course).

These are drastically different. With a bike, you can coast, even for dozens of miles at a time. You have access to a huge range of speeds that are not available to runners. (I can work really hard going 5 mph or 50 mph.) Garmins know absolutely nothing about the wind, which is the #1 thing slowing you down. GPS data alone doesn't have much to go on for calories on a bike, the other variables are too important.

When you run, you can't coast, and the wind can still help or hinder you but to a much lesser extent. Distance, elevation, and vertical oscillation make up a much bigger part of the picture for running.

XC skiing is another case where an HRM helps shed some light on what's going on. Snow conditions can be fast or slow, etc.

I know they are different. We can break this conversation down into the finest of fine hairs, then split those hairs 10 different ways. But I was hoping to have a more, uh, "applicable to real life" conversation.

I'm a regular guy. I run a little bit, I bike a fair bit, I lift periodically, I do some other stuff here and there. But running and biking are my primary forms of exercise. If I wanted to make a reasonable effort to estimate calories burned, and I have one of Garmins better watches... given how unreliable every says HRMs are, am I better off just leaving the chest strap at home?

Some very preliminary tests suggest:

• road running without the chest strap gives me slightly higher calorie burn numbers, but still within reason if you assume the .65 * BW * distance formula to be the standard.
• road biking without the chest strap gives numbers much higher than expected... roughly 50% higher than what I would have expected.
• mountain biking without the chest strap gives me very low numbers, about 1/2 - 2/3 of what I would expect based on what I think I know after years of tracking/logging.

So my exceptionally early theory is that the HRM helps smooth out the error rate.

For workouts where you cannot measure workload, and assuming the device is set up properly (esp true HR max and more accurate estimate of fitness level, and assuming the exercise is primarily aerobic, then yes, a chest strap HR sensor is more likely to improve accuracy.

MeanderingMammal

jjpptt2 wrote: »

MeanderingMammal wrote: »

A lot of the value depends on the sophistication of the tracking devices. For a straightforward HR only device, I'd trust GPS based more than anything else, for a higher end Garmin that correlates HR with position and elevation I'm reasonably comfortable with the conclusions.

That's fine for something like running or cycling, but when you're talking about class based stuff when you don't have a consistent proxy to measure the isn't a good alternative. Low end devices aren't going to correlate elevated HR with vibration and come to a more informed view.

So given one of Garmin's better watches, and assuming it's an average person doing an average ride or run (not intervals, not hill repeats, not time trials, etc... just someone going out for a run or a ride) - would you more trusting of garmins calculations with or without the HRM?

Given that it the becomes one of several data sources leading into the calculation then probably "with", although as a trail runner I'm always cautious anyway. One can see HR drift quite easily on a session of decent length.

NorthCascades

jjpptt2 wrote: »

I know they are different. We can break this conversation down into the finest of fine hairs, then split those hairs 10 different ways. But I was hoping to have a more, uh, "applicable to real life" conversation.

The point is sometimes you need a hammer and sometimes you need a screwdriver, and you get the best results if you can recognize when you use each tool. It's not a one-size-fits-all question.

jjpptt2 wrote: »

I'm a regular guy. I run a little bit, I bike a fair bit, I lift periodically, I do some other stuff here and there. But running and biking are my primary forms of exercise. If I wanted to make a reasonable effort to estimate calories burned, and I have one of Garmins better watches... given how unreliable every says HRMs are, am I better off just leaving the chest strap at home?

For cycling, wear an HRM. For running, I'd lean toward an HRM especially if it's one that measures running dynamics (HRM-Run, HRM-Tri). I'd trust running without an HRM with the new RD pod just as much, personally. For walking, no, I trust the numbers more without HR data.

The_Enginerd

The Garmin algorithm without the HRM is poor in my experience. It overestimates by quite a lot.

The following formula based on running energy expenditure studies has been reliable for me based on years of logging data at various mileages.:

Net Calories = 0.63 * Distance in Miles * Weight in Lbs

My Garmin 920 matches that on average pretty closely with a HRM, although the calorie counts for each individual run vary based on weather. Hot weather runs lead to overestimation of calories burned, and vice versa when it's cold.

BrianSharpe

The_Enginerd wrote: »

The Garmin algorithm without the HRM is poor in my experience. It overestimates by quite a lot.

The following formula based on running energy expenditure studies has been reliable for me based on years of logging data at various mileages.:

Net Calories = 0.63 * Distance in Miles * Weight in Lbs

My Garmin 920 matches that on average pretty closely with a HRM, although the calorie counts for each individual run vary based on weather. Hot weather runs lead to overestimation of calories burned, and vice versa when it's cold.

Agreed.....a good example (anecdotal, I know....) was my bike commute the other day. Riding to work (HRM transmitter used) recorded fewer calories expended than the ride home (forgot the transmitter at the office) despite the fact that the ride to work is slightly uphill & the ride home the opposite.

For my runs I find it very close to the runners world formula.