News

  • MyFitnessPal Overestimates, Garmin Connect Underestimates Calorie Burn

    It has been over 15 years since I've closely tracked calories and macronutrient percentages. About a month ago, I started tracking daily food intake on MyFitnessPal. More on why I began this investigation in the weeks to come.

    MyFitnesPal is a very handy tool with an enormous library of foods. I found that once I tracked just a few days of intake, it was super easy. That's because the program remembers your foods and you can store meals as well.

    However...

    It seemed to me that it was overestimating my calories burned during exercise. I didn't realize until yesterday (thanks Ron Kennedy!) that I could sync GarminConnect, Strava or TrainingPeaks and MyFitnessPal. When the programs are connected, exercise is automatically downloaded from one of the other applications into MyFitnessPal.

    Here are just a couple of examples of how MyFitnessPal overestimates exercise:

    1. I classified a solo bike ride I did Saturday as "Bicycling, 10-12 mph, light (cycling, biking, bike riding)" and MyFitnessPal credited the ride with burning 725 calories. It was a ride with a relatively tough climb and a big descent. My average mph for the ride was 14.6 mph. Garmin (based on my power meter, I believe, and not heart rate) gave me a calorie burn of 441 calories! Woah, a huge difference!
    2. MyFitnessPal gives me a calorie burn rate of 644 calories per hour for fast cycling (16-20mpg average). On my Sunday ride, yesterday, average speed was 17.1 mph. At 3.5 hours of fast cycling, MyFitnessPal credits me with 2254 calories. GarminConnect (again based on the powermeter) gave a calorie burn of 803 – a huge difference. *** Edit *** I looked at Strava for the same file and it credits 1,491KJ and 1,663 calories for this ride.)

    What is accurate?

    Any system is attempting to estimate calorie burn. By using only miles per hour, wind assistance, descending and drafting are not included. Using miles per hour and body weight overestimates calorie burn significantly.

    If we use only power output or heart rate, I believe calorie burn is underestimated. This is most apparent in mountain biking. While my heart rate might not be high (I don't run a power meter on the mountain bike) on some of the trail sections, I'm putting out a lot of work to maneuver these tricky sections of the trail. That is, a good deal of strength is needed for downhill riding and this effort isn't always captured by heart rate. Short, high effort pops on the trail don't drive high heart rates, but come at a high physical cost.

    The bottom line is neither system is 100-percent accurate and no system will be for some of the reasons I've explained. If you are basing your calorie budget off of either system, you will need to make conscious modifications so you aren't under or over consuming calories.

  • ← Next Post Previous Post →
  • Comments on this post (1 comment)

    • Scott Kleihege says...

      As I understand, Garmin devices aren’t calculating calorie burn on average heart rate but based on time between beats. Garmin claims to within 10% accuracy using the firstbeat method (more at http://www.dcrainmaker.com/2010/11/how-calorie-measurement-works-on-garmin.html). In my experience, garmin devices show more calories doing HIIT-type efforts like mountain biking as I would expect.

      The other side of the coin is calorie intake, which would be hard to get better than 10% accuracy without weighing everything I put into my body. By combining calorie tracking with periodic weigh-ins, I start to compensate for the inaccuracies on both sides, but loosing the last 5 pounds is where it gets really challenge for me.

      On April 20, 2015

  • Leave a comment