Example of Accuracy of MFP Exercise Calories

Thanks for this post!

Bump! And Thanks!!!

I always use the number the machine at the gym says I burned, Sometimes it is over and sometimes it is under what MFP says. MFP does not account for any variations of speed you may do while working out........Also when looking for a generic food I always search for the one with the highest calorie count even though I may feel that my meal was a little under what the database said. Everyone keep up the good work!!!!

Just wondering if anyone has contacted MFP to find out how and what information their calculations are based on?

I defintely don't go by MFP calories burned. I find for the most part they overestimate what is really burned. I use EA Active Sports for PS3 and you are hooked up to a heart monitor. That way I know for sure how many calories I am burning.

I thought I'd share this with you, but I do so with the caveat that it is one example and the lesson may not be directly accurate for others at different weights and fitness levels and with different activities.

I own and use a Garmin 310XT HR watch which is the only one (to my knowledge) that doesn't use an algorhythm to predict calories burned but actually calculates it on a person's actual heart rate through exercise (cardio only!). I believe it to be the most accurate calorie burn calculator on the market (that's why I paid the crazy $400 for the watch).

That said, I just finished an easy run, at 5 mph (to ensure I was comparing apples to apples re MFP exercise numbers) for 42 minutes. My TRUE calorie burn was 370 calories. But when I entered the 43 minutes (I had to add one minute because once I had "taught" MFP that 370 was correct, it wouldn't return to its own number and I neglected to write it down first), MFP told me that I had burned 460 calories.

That's an error of 90 calories -- not insignificant in my mind.

I attribute the lower actual number to my cardio fitness since I'm a regular runner and thus my heart doesn't pound as fast as maybe the average person's would during a run. So the error margin is going to be different for each person/exercise/duration.

I have found in the past that MFP seems to overestimate calorie burn by a MINIMUM of 20% for me.

I tell you that not to fault MFP because I believe that their numbers are probably the best that anyone could use for the average individual seeking to diet and lose weight. I share this only so that y'all don't take the numbers too seriously. If you seem to be hitting a plateau, consider shaving a bit of credit off the MFP exercise numbers and see if that helps at all.

It appears from your user manual that Garmin is using software from Firstbeat Technologies to estimate calories, which, from what I can tell, is one of the better ones out there. Suunto uses the same software. I don't think your device is quite at the top level (such as a Polar RS800x or Suunto T6d) because it does not appear to have to the capacity for beat by beat R-R interval analysis, but it is likely as good as anything else by Polar or Suunto (other than those two models).

However, the Garmin uses algorithms just like any other HRM. There really is no other way to do it. All HRMs measure heart rate, and they use algorithms to interpret the heart rate data and estimate calories based on their validation studies. (If you want to look at the studies, you can go to their website--it looks like they changed it to www.firstbeat,fi . It's in Finnish, but there is an "English" button in the upper part of the screen. If you want to spend $100, they have a more advanced software program for analysing performance--I'm just not sure it is compatible with the Garmin 310.

As far as MFP calorie tables go, I have never used them and never would. They are just not that accurate. The walking and running ones should be pretty close, since they are based on ACSM prediction equations, but the rest are not very useful.

Theoretically, your training status should not mean that much. Calories expended are based on the energy cost of the activity, not on heart rate per se. With training, a submaximal workload will represent a lower percentage of your VO2max (since VO2 max increases with training) so heart rate will be lower, however the actual energy cost of the activity remains the same. In other words, running 5 mph has an intensity of about 8.5 METs (per the ACSM prediction equations). That should be roughly the same for everyone. For a fit person, that might only be 50% of their maximum and an unfit person might find it 100% of their maximum, but it's still 8.5 METs.

If someone is training and their fitness level increases and they do not update the settings in their HRM, then, yes, the calorie number on the HRM will go down. But that's a mistake in the device setup--it doesn't represent what is going on physiologically.

I would be curious to know if they mentioned any of this at the Cooper Institute, and, if so, provided any actual research. Because I have searched PubMed high and low and can't find any. Quite frankly, in the exercise physiology field, it's not a subject of much importance.