Planemaker

I always knew that my Garmin way overestimated the calories burned during a ride. This week I started riding with a HRM and saw first hand. The same miles and rides show less than half the calories burned using a HRM.

Good news is that I road 68 miles today averaging 16 mph with about half the miles into a stiff headwind. My average heart rate was 126 and my max was 167.

DMC707

what kind of Garmin --- i use an Edge 500 and find it reasonably accurate as long as i set it up correctly. Has a HRM also

Much more accurate than some of the fitness apps i have tried while riding

digibud

Impressive distance, HR and speed. Well done. I could never pull that off.

Beachgrad05

This.

At least not yet

jfries

I had a friend tell me that many apps overestimate calories burned by 25-30%, so what I've done in the past is "lied" about my weight on those apps and made myself thinner but I've still found them to be a little off.
I've started using a HRM in February (I use a Garmin Vivofit with a chest strap HRM). On Sunday I did a 52 mile route and MapMyRide said I burned 4102 calories. My HRM said I burned 3462. Average HR for the ride was 134, which is Z3 for me.
I don't know if HRM's are the best way to figure out calories burned but it seems better than relying on an app.

baron von trail 

That's probably close to a 3000 calorie burn.

sstorkel

Even the calorie number from the HRM sounds pretty optimistic to me. I've yet to find a HRM that didn't report far more calories than my power meter, though I haven't tried anything that was released in the last 5 years or so. In any event, I wouldn't put much stock in calorie estimates from anything other than a power meter. You'd probably do better, weight loss-wise, to use the old 30-35 calorie/mile guestimate.

baron von trail 

30 cal per mile sounds about right: 15mph, 450 cal burned. I can't see that effort burning much more than that unless it's all uphill.

LAE

I agree with most here to be honest, even with the HRM my 520 wayyy overestimates my calories burned, curiously Strava does ok, seems to be there or there abouts with my own estimates.

For example, a Z2 ride yesterday at around 18mi had my garmin (with HRM) measure 1702kcals!!! Strava on the other hand put it at a more reasonable 550kcal.

baron von trail 

300%

I like this graph. I've seen it on the forums before. It's probably as close to reliable as is possible when looking at generalities.

kc0bbq

That graph is probably fairly correct for a bike with wider tires like a cruiser or mountainbike. Gut feeling is that it's a bit high for skinny tires (my numbers on my least aggressive road bike are about 6 calories per mile lower, the other bikes are increasingly aero and aggressive and skew the graph in other ways).

Strava's estimates are pretty good assuming you are doing some sort of loop and not just sailing with the wind the entire route.

gsa103

Calorie estimates are about equivalent between HRM and power meters, unless you're doing hard short intervals. The problem with a power meter is that you need to know your (individual) metabolic efficiency (typically 18-23%).

Think of it this way:
1) A power meter measures how much mechanical work you performed.
2) A heart rate monitor measures how your body is responding to that physical effort.

Determining calories burned accurately requires knowing both pieces of information. Calorie estimates based on HR are generally considered to be ~10% accurate (FirstBeat 2, Garmin 510+). The problem with power meters is that you need to divide your Mechanical Work (measured) by your metabolic efficiency (18-23%), as an a result aren't any better at measuring calories burned. They are far better for training though (for a variety of reasons).

Consider the following scenario: Nominal ~1000 calorie effort
A) A good computer & HRM estimates will likely be in the 900-1100 range.
B) Power meter measures 200 units of work, assuming 20% efficiency ~1000 calories.
Assuming 18% = 200/0.18 = 1111 calories
Assuming 20% = 200/0.2 = 1000 calories
Assuming 23% = 200/0.23 = 869 calories

As you can see, both devices provide similar levels of accuracy, if used alone. A power meter can provide highly accurate calorie estimates, BUT it requires a lab test to determine your metabolic efficiency. Once your metabolic efficiency is known, the power meter is much more accurate than a HRM.

sstorkel

The calorie estimates from the 2-3 HRMs that I've tried were always 50-100% higher than the work estimate (KJ) provided by my power meter. If you've compared your HRM to a power meter and see numbers that are plus or minutes 10% then I would agree that they're equivalent. If you've found such a HRM, I'd love to know which brand and model you're using! I'd love to have an accurate HRM that I could use for running, kayaking, etc.

DrIsotope

Strava doesn't use HRM data to estimate work/calories, so my numbers are always absurdly high. The Garmin Express software I used with the 500, and the Bryton Explorer I use with the Rider 310 do use the HRM data, so the numbers tend to be a lot closer to what I would consider realistic.

35.9 miles, 1,270ft^, 1.50.15, avg. 19.6 mph
Strava: 2,103kcal.
Bryton: 1,599kcal.

29.6 miles, 1,480ft^, 1.36.25, avg. 18.5mph
Strava: 1,711kcal.
Bryton: 1,273kcal.

I mean, I wish I was burning 1,100kcal an hour, but both of those rides are tempo paced (avg. HR 145-150 for me) and I'm too aware that even the Bryton numbers are probably a little generous.

gsa103

My benchmark is an extended climb up a 6% grade (1000m vert), the Garmin 510 (w/ soft strap) estimates are basically dead-on with any average power metric. The calculation is done by the computer not the strap, so you'd need a watch or something.

What computer are you using to estimate calories for both the power meter and HR? And are you sure the power meter calorie estimate is correct?

sstorkel

Which power meter are you using?

With my PowerTap power meters, I assume KJ ~= kcal. My last ride was 28.5 miles with 1100 feet of elevation gain and an average speed of 15mph. I used a Garmin Edge 800 and Garmin HR strap. The power meter reported 838 KJ of work, while the Garmin estimated 1376 calories burned. Interestingly, my Apple Watch thought I burned 1401 calories. Back when I bought my first PowerTap and Garmin (an Edge 705, I think) I compared the numbers to the Polar HRM (watch + strap) I was using at the time. The Polar's calorie estimate were almost universally twice as large as the work (KJ) reported by the PowerTap. I also tried a higher-end Polar HRM a couple of years ago. Calorie estimates were slightly lower than with the original Polar HRM but IIRC still 50% higher than the numbers from the power meter.

Interesting fact: 30 calories/mile * 28.5 miles = 855 calories

gsa103

Your PowerTap is very likely under-estimating calories burned.

From a pure physics work definition:
1 kJ = 0.239 kcal

Assuming 1 kJ = 1 kcal burned is 1 kJ/0.239 = 1 kcal. Or a metabolic efficiency conversion of 23.9%, which is at the high end of the metabolic efficiency reported in scientific literature. That assumption may be valid for TdF rider, but likely doesn't hold for a clyde. Additionally, the PowerTap measures power at the wheel, rather than at the crank where it is delivered by the body so you need to account for drivetrain losses (~5%).

If you assume a mid-range 20% metabolic efficiency:
Your (838 kJ)/(0.95 drivetrain)= 882 kJ (crank)
kcal burned = (882*0.239)/0.2 = 1054 kcal

These are still lower than the Garmin, but the Garmin may be estimating high. The best approach is to think that your calorie burn is likely somewhere between the two estimates.

dr_lha

Citation? 25% is a typically accepted value, but I'd be interested in seeing your evidence for believing that a "normal"/clyde would be as low as 20%. We're generally considered to be more efficient than cars!

Also: Look ma I'm in a calorie estimation thread.

kc0bbq

Yeah, everything I've seen has put the "normal" range at 23-27% with the center of the curve essentially at 25%.

gsa103

Here's the paper I found. https://d3epuodzu3wuis.cloudfront.net/R060.pdf

The estimates do vary in the literature, so its possible there some test methodology dependence.

dr_lha

Do you actually understand that paper, out of interest?

I mean do you know the definitions or gross efficiency, delta efficiency and "economy"?

I would be concerned about cherry picking numbers from a single paper, unless you actually have a broad understanding of the field. I'll admit I don't.

gsa103

It's definitely not my field, so there's definitely some potential for that. Given the reported variances, I suspect there test protocol effects, and a ton of other factors that may contribute.

Looking through the paper, gross efficiency is clearly defined as (metabolic effort)/(mechanical work). The metabolic effort is measured using respiration which is what I've seen used in other papers. Mechanical work is the usual power meter. GE_final is the efficiency before the O2/CO2 balance exceeded 1.0 (nearing VO2max). GE_165 is the efficiency at 165W.

It looks like delta efficiency is the change in efficiency as you move from light to hard efforts. I'm not completely sure on the magnitude of the effect but it looks like the efficiency improves as you go up in effort, so a Z2 effort might be 18% but increase to 20% for Z4.

Economy is defined as Work/VO2 intake (kJ/L), which is basically how well is your body converting oxygen into mechanical effort.

I was also able to find this poster on the FirstBeat algorithm used by Garmin.
https://www.firstbeat.com/app/upload...5_congress.pdf

Summary: Power meters are great at figuring out mechanical effort. Getting accurate calories burned requires a lab test. Everything else is pretty some someone's best guess.

sstorkel

Doubt it. When I plug the PowerTap numbers into a weight-loss plan, I lose weight at almost precisely the rate that would be expected. When I plug the calorie estimates from the HRM into the same weight-loss plan, I fail to lose weight or even gain weight. That seems like a pretty good indication that the HRM is over-estimating the number of calories burned.

Doesn't the paper you linked suggest that there are No Differences in Efficiency Between World-Class and Recreational Cyclists? BTW, who said I was a Clyde?

dr_lha

Honestly anyone arguing that a HRM is a better calorie estimator than a power meter is just wrong. Sorry.

Masque

This has just inspired me to only bike downhill. I can drive up to the top of Larch Mountain and bike back down probably once per hour, at 40mph. I'll be burning calories like never before! Thank you!!!