Cyclists - Calculating calories on trainer

owendvn

Hey all. I've been doing a lot of indoor cycling over the Winter (began outdoor cycling a few months ago in the Fall 2017), but never had a good grasp of how many calories I was actually burning. I currently ride my road bike indoors and attach it to a somewhat decent trainer that has a power meter (claimed +/- 2.5% in measured power).

I ride in 40-minute sessions and each session generally falls within the following:
Speed: 20-23 mph
Power: 200-205 W/hr
Cadence: 80-85 rpm
Distance: 13-15 miles

The one missing component is heart rate. Will I need a heart rate monitor to accurately calculate calories burned? I've been following a formula crafted here (http://mccraw.co.uk/powertap-meter-convert-watts-calories-burned/) , which gives the following: energy (kcal) = avg power (W) X duration (hours) X 3.6

I try to underestimate the calculation and lower the multipliers, so if I was averaging 200 W/hr, I'd have 3.5 * (40min/60min) * (200 W/hr) = about 466 calories per 40-minute session.

Would this still be considered an overestimate? I've not trusted other online calculators as those numbers seemed wildly too high (in the 500-600s). Also, how much would EPOC factor in after these sessions? Would I also need to know my HR stats to have a better gauge of EPOC?

aokoye

Given that Garmin and Trainerroad also use that formula I suspect that it's fine. I should also note that MFP generally overestimates for me if I put in the average mph and duration of a workout on an indoor trainer.

Azdak

If you are measuring watts directly, then HR is irrelevant for calorie burn. EPOC is negligible, so I wouldn’t bother trying to calculate.

sijomial

Power is a far superior energy measure than heartrate which has huge personal and environmental variations.

Converting watts to calories is also giving you net calories so effectively the minimum number it could be, it's just about impossible to be an overestimate unless you have an unfeasibly extremely efficient pedalling efficiency ratio that is far better than pro riders.

Ignore EPOC, impossible to measure and likely insignificant even for intense workouts.

Why you are fiddling the numbers, which are well proven and a mathematical formula, by making a deliberate underestimate is a mystery to me!

owendvn

Thanks for the replies, guys. Me fiddling with the numbers is more poor research on my part. I’ve only really seen that formula from the website I linked and didn’t know it was widely used. I’m more confident in the counts I’m getting now, though. Thanks!

mjbnj0001

great thread, thanks. I was gearing up to ask the inverse question - could I figure my power (watts) from the info provided by the gym cycle. Now it's simple by using the above formula and a little algebra. I've been getting about 3x indoor vs. outdoor sessions since January 1st. I'm not using a power meter on my realworld bike (but I have started using an app, "Ride With GPS" that gives a power guesstimate). Now maybe I can begin typing the two types of riding together better. Thanks again.

mjbnj0001

tired fingers. that was, "... TYING the two types ..." above.

aokoye

mjbnj0001 wrote: »

great thread, thanks. I was gearing up to ask the inverse question - could I figure my power (watts) from the info provided by the gym cycle. Now it's simple by using the above formula and a little algebra. I've been getting about 3x indoor vs. outdoor sessions since January 1st. I'm not using a power meter on my realworld bike (but I have started using an app, "Ride With GPS" that gives a power guesstimate). Now maybe I can begin typing the two types of riding together better. Thanks again.

I suspect it's not that simple as different trainers/indoor spin bikes all have different power curves. If you're getting a calorie count from the gym cycle the question is, is it measuring power? If it's not measuring power then it's not an accurate count. Also I know at least at the gym I go to, the power meters on the spin bikes seem to seriously in need of calibrating. They were laughably high. I mean I could have been underestimating the effort I felt on them vs the power on my trainer at home, but I doubt it.

robertw486

owendvn wrote: »

Hey all. I've been doing a lot of indoor cycling over the Winter (began outdoor cycling a few months ago in the Fall 2017), but never had a good grasp of how many calories I was actually burning. I currently ride my road bike indoors and attach it to a somewhat decent trainer that has a power meter (claimed +/- 2.5% in measured power).

I ride in 40-minute sessions and each session generally falls within the following:
Speed: 20-23 mph
Power: 200-205 W/hr
Cadence: 80-85 rpm
Distance: 13-15 miles

The one missing component is heart rate. Will I need a heart rate monitor to accurately calculate calories burned? I've been following a formula crafted here (http://mccraw.co.uk/powertap-meter-convert-watts-calories-burned/) , which gives the following: energy (kcal) = avg power (W) X duration (hours) X 3.6

I try to underestimate the calculation and lower the multipliers, so if I was averaging 200 W/hr, I'd have 3.5 * (40min/60min) * (200 W/hr) = about 466 calories per 40-minute session.

Would this still be considered an overestimate? I've not trusted other online calculators as those numbers seemed wildly too high (in the 500-600s). Also, how much would EPOC factor in after these sessions? Would I also need to know my HR stats to have a better gauge of EPOC?

As stated by others, EPOC isn't worth trying to figure out. As for the heart rate, I'd say a way to measure HR would be great to measure fitness improvements, but certainly isn't of much help for your calorie burn calculations. As it is you have the only variables being the accuracy of your meter and the guesstimate on efficiency. And it doesn't get much better than that.

mjbnj0001 wrote: »

great thread, thanks. I was gearing up to ask the inverse question - could I figure my power (watts) from the info provided by the gym cycle. Now it's simple by using the above formula and a little algebra. I've been getting about 3x indoor vs. outdoor sessions since January 1st. I'm not using a power meter on my realworld bike (but I have started using an app, "Ride With GPS" that gives a power guesstimate). Now maybe I can begin typing the two types of riding together better. Thanks again.

Just keep in mind that depending on the type of bike and the outputs it gives you, you might be introducing more error. Many machines that spit out calorie numbers are outputting gross rather than net calories, and if you input an age and/or weight it's more likely calculating gross numbers.

Depending on the machine type, you might be able to find online info as to how they calculate, and how accurate or inaccurate it might be. From there you might be able to get a better real world power estimate.

As for apps vs machines vs meters, the meters take more variables into account. But given similar conditions and power estimates, the apps I've used show trends that are in line with the power changes. They may be less accurate as for the end number, but if you discount weather variables and such they will usually show the increase or decrease in power output over a set course.

dylanr422

If this is a smart trainer you should be able to record your workouts and have them uploaded to Strava. Strava will use the power data to calculate calories burned. I've found their calorie numbers to be extremely accurate when I'm using a power meter, less so when I'm not. This is based on eating maintenance calories and also 100% of the Strava workout calories and staying at exactly the same weight for extended periods of time.

https://support.strava.com/hc/en-us/articles/216917097-Calorie-Calculation

aokoye

Strava is generally wildly off base even when I'm feeding it data from my trainer's power meter (I have a Kickr Snap). Sometimes it will be within 50 calories, yesterday being a good example of that, but generally it's at least 100 calories higher than Garmin's calculation (using the above formula). This seems to be especially true if I'm doing intervals (which is most of my indoor training).

jlklem

Without knowing your bodies efficiency at creating energy even power meters can be 5-10% wrong. Strava gives 2 numbers...the larger is actually more accurate as most people are just not that efficient on the bike. A average rider’s efficiency on a bike is low....maybe 15-18%. The smaller Strava number assumes I think 23% efficiency which is rare. Pro numbers are 23-26. Highest ever I think was 29%.

mjbnj0001

Arrgh. Thanks, everyone. Gotta dig deeper, I guess. The gym uses a Life Fitness recumbent cycle, and I've been neglecting to enter my weight to calibrate the calories (I've done that amount of background research since last evening). If only things were simpler ... LOL:

saintor1

Exactly my problem I dealt with. With muy indoor bike, I use a wahoo chest monitor and felt that it was posting way too optimistic. So I cheated with my weight, at -30% to have something I trust

jlklem

mjbnj0001 wrote: »

Arrgh. Thanks, everyone. Gotta dig deeper, I guess. The gym uses a Life Fitness recumbent cycle, and I've been neglecting to enter my weight to calibrate the calories (I've done that amount of background research since last evening). If only things were simpler ... LOL:

Sorry if this stated above but does it give you a power number, if so weight does not matter

mjbnj0001

all ok, i wasn't being snarky or anything. i just fell fast from the "heights of wrong results," lol. i had great, if wrong, clarity for a moment.

Jthanmyfitnesspal

I am using a trainer as well, but it doesn't have a power meter (Kenitic). I let Wahoo do the calculation (based on HR, weight, and age) and I downrate it by about 20%. It just seems high based on how tired and hungry it makes me.

Now, I believe the calorie estimates from Wahoo for running. If I don't eat those calories back I starve!

jhanleybrown

I've found this to be accurate if you don't have a power meter.
[ { (.6309 x Ave HR) + (age x .2017) + (.09036 x weight (lb)) -55.0969 } x time (min) ] x (% HR is of max HR)

Obviously long but no biggie if you have a spreadsheet. I find it gets me numbers roughly equal to the power calc. Useful if you dont have a power meter or are on a trainer/gym bike with a suspiciously generous power reading.

sijomial

mjbnj0001 wrote: »

Arrgh. Thanks, everyone. Gotta dig deeper, I guess. The gym uses a Life Fitness recumbent cycle, and I've been neglecting to enter my weight to calibrate the calories (I've done that amount of background research since last evening). If only things were simpler ... LOL:

@mjbnj0001

The point of entering your weight escapes me for what is a non-weight bearing exercise.
Weight has some bearing cycling outside (affects the power needed for acceleration and hills) but not indoors.

dylanr422

aokoye wrote: »

Strava is generally wildly off base even when I'm feeding it data from my trainer's power meter (I have a Kickr Snap). Sometimes it will be within 50 calories, yesterday being a good example of that, but generally it's at least 100 calories higher than Garmin's calculation (using the above formula). This seems to be especially true if I'm doing intervals (which is most of my indoor training).

It's true that power meters are also sources of some degree of inaccuracy. My Stages power meter I use on my bike outside gives me around ~10 watts higher FTP than my Tacx Neo trainer. It is measuring from the crank vs the trainer which measures after drivetrain losses.

However they both are very consistent and I've found for me the calorie counts are quite close to the physiological cost. This is compared to Strava estimation with no power data which is often laughably off - especially for events like cyclocross where you can be going 8 mph but actually plowing through thick grass or mud.

I no longer trust heart rate as I find for me it's incredibly temperature dependent. My threshold heart rate in a race around 32F is almost 10bpm lower than at 70F.

mjbnj0001

Thanks, everyone for your insights, and @jhanleybrown for the above formula. I will see what it tells me with the data I have.

While I'm not the OP of this thread, I have enjoyed the extended discussion. One thing I didn't mention - I usually dont like sharing too many details of myself online, is that I'm newly-retiews, over 60 and have an older-guy ailment - atrial fibrillation (permanent). So, @dylanr422, I'm always measuring my heartrate as one of my basic datapoints. Two cardiologists - good ones - and my primary care physician are all thrilled that I've (re)taken up cycling (absence of about 20 yrs), but have given remarkably little guidance on what my exercise params/constraints should be, except "be conservative and what I can tolerate." The internet, and I've scoured as much as I can for stuff sitting outside of paywalls, is not much better. So I'm working up from essentially zero-base-of-cardiac-robustness effective January 1st on a gradual buildup program. I'm sticking to Training Zone 2 mostly, in the gym, with progressively adding 1-minute high RPM intervals (right now, I'm doing the highs at 8-minute intervals - 8/16/24/32/40); this drives me into Zone 3 for brief periods. I started at a 30-minute level-effort program, am now doing 45-minute varied-effort, and am looking to either add more high intensity (on the 7 minutes) or extend to 60 minutes on March 1st. My midyear goal is a 60-minute session with higher rpms and some (undetermined) number of higher-intensity intervals. Today's gym session resulted in a 9+ mile ride in 45 minutes, avg 72rpm with higher intervals in the 85-90rpm range, 290 cals.

Out in the realworld, I ride when the conditions are OK (down to about 30 degrees F, paths/trails clear of snow/ice, etc.). From zero on Christmas, I'm up to about 60-75 minutes actual cycling, 9-10 miles, on fairly level surface. I am working up a good sweat and feel invigorated after the gym rides and constant pedaling; my legs are more tired from the realworld rides where I am actually propelling myself across the landscape. Just as in the gym, I have marked up goals for longer/faster/more varied performance on the real bike building progressively during the year. A couple of decades ago, on my touring bike, average cadence was in the 90 range; I think those days are behind me now. On this new 29er hardtail, the gearing - plus me - seem to point to lower cadence goals. Since my outside hoses are shut until the season warms, I'm sticking to paved paths/trails (and mostly not roads) to stay away from as much road salt and grit as I can until I can clean the bike regularly.

Without being able to tie the numbers together between these two modes, I can only say that I believe that I am spending more calories in the gym rides than the realworld, but the two modes are somewhat comparable as exercise, and support each other. I feel "complete" when I wrap at the gym, but think I can do more time/dist in the realworld if I want (I'm keeping to a conservative program). My program has been to try and do 5 of 7 days per week on either the bike or in the gym, schedule driven by the weather. The other two days are rest days. The gym felt tougher today after two days away.

jjpptt2

FWIW...

MFP and Garmin (both my 920 and my 935, both chest strap and wrist-based HR) give very similar numbers for trainer work (I don't have power data). They are also reasonably close to what the stationary spin bikes at my gym say that do measure power.

I haven't logged my intake consistently enough to use the whole actual results vs expected results to verify the numbers in a long time, but when I was, I was losing/gaining/maintaining as expected based on the math... so that suggests the numbers are reasonable.

Are there lots of variables and assumptions being made when estimating cals based on anything other than power? Yep. But I think if you fall within any kind of reasonable averages/deviations, the numbers given are close enough to use as just exactly what they are - approximate estimations.

Jthanmyfitnesspal

jhanleybrown wrote: »

I've found this to be accurate if you don't have a power meter.
[ { (.6309 x Ave HR) + (age x .2017) + (.09036 x weight (lb)) -55.0969 } x time (min) ] x (% HR is of max HR)

Where does that formula come from? I haven't seen the correction for %MAXHR.

NorthCascades

dylanr422 wrote: »

aokoye wrote: »

Strava is generally wildly off base even when I'm feeding it data from my trainer's power meter (I have a Kickr Snap). Sometimes it will be within 50 calories, yesterday being a good example of that, but generally it's at least 100 calories higher than Garmin's calculation (using the above formula). This seems to be especially true if I'm doing intervals (which is most of my indoor training).

It's true that power meters are also sources of some degree of inaccuracy. My Stages power meter I use on my bike outside gives me around ~10 watts higher FTP than my Tacx Neo trainer. It is measuring from the crank vs the trainer which measures after drivetrain losses.

However they both are very consistent and I've found for me the calorie counts are quite close to the physiological cost. This is compared to Strava estimation with no power data which is often laughably off - especially for events like cyclocross where you can be going 8 mph but actually plowing through thick grass or mud.

I no longer trust heart rate as I find for me it's incredibly temperature dependent. My threshold heart rate in a race around 32F is almost 10bpm lower than at 70F.

Your Neo isn't measuring power, though. It's measuring wheel speed and predicting the power required to achieve that given the resistance. And your Stages is probably only measuring the left side.

A power meter has an accuracy spec. Most have a maximum error of about 1.5 %. Have a look at any DCR power meter review and see how well they all line up.

BrianSharpe

Azdak wrote: »

If you are measuring watts directly, then HR is irrelevant for calorie burn. EPOC is negligible, so I wouldn’t bother trying to calculate.

Installing a power meter on my bike was depressing in respect to calories expended but the watts do not lie......

dylanr422

NorthCascades wrote: »

Your Neo isn't measuring power, though. It's measuring wheel speed and predicting the power required to achieve that given the resistance. And your Stages is probably only measuring the left side.

A power meter has an accuracy spec. Most have a maximum error of about 1.5 %. Have a look at any DCR power meter review and see how well they all line up.

I guess from reading a bit the Stages power meter would be a form of transmission dynamometer while the Tacx Neo would be a type of absorption dynamometer. https://www.britannica.com/technology/dynamometer.

I do read DC Rainmaker. The Stages is not the most accurate power meter money can buy but for the price it's pretty decent. The Tacx Neo is generally considered to be the best trainer on the market today with accurate power measurement.

Earlier in the thread I posted that I had eaten back 100% of the Stages power derived calorie counts given by Strava along with maintenance calories for about 4 months and my weight didn't change at all despite wildly varying daily calorie intakes (2000 - 4500).

Apart from doing all your workouts in a lab to measure your oxygen consumption, I think cycling with power data is the most precise way you can burn calories. Errors in trying to estimate ingested calories probably outweigh any errors in the numbers due to power meter inaccuracy.

jlklem

BrianSharpe wrote: »

Azdak wrote: »

If you are measuring watts directly, then HR is irrelevant for calorie burn. EPOC is negligible, so I wouldn’t bother trying to calculate.

Installing a power meter on my bike was depressing in respect to calories expended but the watts do not lie......

But it clearly lets you know if your training is working. And you can really nail your calorie/weight loss goals. I have had 5 different power meters in my life. Trained with power since 2008. When I want to drop 5 pounds I can, I do it every year at this time. Race weight is 5-6% body fat (yes tested on a DEXA and in water 2 time for each). Could not get there, work out well without the power meter. The data power provides is spot on.

Now if my races could be as predictable....