Calculating weekly MET minutes by fitbit data

I have a fitbit Charge 2 activity tracker which provides continuous HR monitoring data and daily energy expenditure estimates. Is there a way to convert & quantify this fitbit activity as “MET minutes” ?

http://journals.lww.com/acsm-healthfitness/Citation/2012/02000/Wouldn_t_You_Like_to_Know__How_Can_I_Use_METs_to.5.aspx

http://ergo-log.com/half-an-hour-exercise-a-day-make-that-three-hour.html

http://ergo-log.com/plaatjes/ischemicheartdiseasedoseresponse.gif

You can actually request MET values from the Fitbit API for your personal account.

See this github repo for instructions on setting up a google spreadsheet doc to get data. You may have to do some adjustments to get the MET values as they come from the /activities end point (I think).

Keep in mind that if you request MET values from the Fitbit API that they come with no decimal place, so a record of 10 is actually 1.0 (which is typical for sleeping).

You can learn more about the intraday data Fitbit provides here.

Here’s a quick example plot of my METs (also from a Charge 2) for last week.

1 Like

This is actually a bit better of a plot with each minute categorized at either sedentary, light, moderate or vigorous based on the MET categorization system (1-1.5, 1.5 - 3, 3-6, 6+, respectively).

1 Like

Wow, that’s great!

So, it looks like I can grab the JSON data at https://api.fitbit.com/1/user/ME/activities/calories/date/YYYY-MM-DD/1d/15min.json, add up the 96 x 15m METs intervals from returned activities-calories-intraday.dataset, divide by 10 for my daily METs, then iterate through the week for my weekly METs. Thanks!

Playing with the data, my math must be off here.
https://jsfiddle.net/02yzcxef/

Looking at my METs while sleeping, fitbit lists “150” for a 15m block which I’m guessing corresponds to 1.0 METs x 15m with a unit size of -1 (it would be nice if the fitbit API included a reference “unit” if they’re not going to send the decimal places.)

If the Physical Activity Guidelines for Americans recommends 500-1,000 MET minutes per week and there’s 1,440 minutes per day with a minimum MET score of 1.0… am I only counting METs greater than some arbitrary minimum or subtracting the basal 1440 METs? eg. https://jsfiddle.net/72u36sgm/

METs are inherently confusing as they’re expressed as a relative value.

For the guidelines, the 500-1000 MET minutes per week are usually based on time spent in MVPA, or at the very least, time spent above resting. That is all time spent above 1.5 METs (sleep, resting, sedentary behavior like sitting).

More details on METs and the guidelines are here: https://health.gov/paguidelines/guidelines/appendix1.aspx