Benjamin J Gilbert

For the past few weeks, I’ve been wearing the BodyMedia Fit Link Armband. It provides a measurement of how many Calories per minute I’m burning. Using a combination of motion measurement with an accelerometer, steps taken, galvanic skin response (skin moisture), skin temperature, and heat flux (heat produced by the body that escapes), it can provide these measurements with a mean error less than 10%. It also provides data on when I was actually asleep at night, total sleep time, and brings in data from third party services like MyFitnessPal (where I’ve been tracking food intake) to provide net caloric balance. It’s got a mobile app that syncs over Bluetooth, so I can get graphs on-the-fly of my caloric burn during various parts of the day. I’ve noticed a few interesting trends, and I wanted to share them here.

This is a graph of Calories burned per minute on a day where I ran in the morning before work, and lifted weights after work.

This is a graph of Calories burned per minute on a day where I went skiing.

This is a graph of Calories burned per minute on a day where did an hour of hot yoga before work (6 AM – 7 AM).

This is a graph of Calories burned per minute on a day where I did a lot of weightlifting (short heavy sets) and a quick run, followed by some walking around to run errands afterwards. I also went out to bars the night before, which explains that 2 AM spike for the stumble home.

Here are a few of my most interesting observations from this data:

• Not surprisingly, running is the most time-efficient way to burn calories. I was running at roughly an 8:30 pace, and was right around 10 Calories burned / minute the entire time.
• When I’m at my desk, I may as well be asleep as far as my caloric burn is concerned. That’s a scary thought for those of us with desk jobs. Even scarier: most of the time I’m using a standing desk, which didn’t seem to affect much. Those spikes throughout the work day are just when I wander around to other people’s offices.
• Lifting weights burned about 3x the amount of Calories per minute as sleeping, but at 5 Cal/minute, it’s still only about half that of running. Lifting is obviously beneficial for other purposes though, like gaining strength.
• That skiing graph is pretty funny to look at. It’s obvious where I took lunch, and skiing seems to provide running-like caloric burn rates when on the hill, but low points at about ~2x sleeping while sitting on the ski lift between runs.
• In general, I’m amazed at how fast my caloric burn rate returns to that of sleeping after I work out. It seems to have little effect on my burn for the rest of the day.
• The yoga day is perplexing to me. The caloric burn during the yoga class was surprisingly low, never exceeding 6 Cal/minute. This can be attributed to one or both of the following:
  • All of that sweating really is just from the room being 104 degrees.
  • The BodyMedia Fit is poor at estimating in this case since there isn’t a lot of bipedal movement to make the accelerometer think I’m in motion.
• Another interesting takeaway from yoga: It took a long time for my caloric burn rate to return to my sleeping rate after I finished. The class was 6 AM to 7 AM, and despite sitting in a car driving home, it took about 45 minutes to decrease below the average rate from the class. My guess is that during this time, my nervous system was still pretty messed up from the temperature change and stressed body movement, and it takes a lot of Calories to recover.
• Merely walking around is pretty effective for burning Calories at about 5 Cal/minute. I guess that living 15 minutes from the bus stop to work really is a great lifestyle choice.
• Stumbling home from bars is surprisingly effective for burning Calories.

I’m going to continue testing the BodyMedia Fit for at least another couple months, until my free trial of the web service is up. I’d love to chat about my other experiences with the device, its sleep tracking, integration with the mobile app, and other devices you’ve used, so feel free to reach out on Twitter at @gilbert.