Charge HR burned calories REALLY high

L_Master

heybales wrote: »

L_Master wrote: »

Probably as good an estimate as I can get for walking. It's hopeless for cycling though, underestimating by well over 40%. Rode for 2 hours the other day at HR avg 126 and the fitbit gave me just over 1300kcal for the ride, or 650 kcal/hr...an almost comical 36% lower figure than the actual 2,017 calories burned.

Honestly, it's odd to me fitbit tries to calculate calories from HR in the first place. HR by itself is a meaningless number when to calories burned, because you can't know someones output at a given HR. Either they are trying to estimate a VO2 (no clue how) or they are just assuming a very vanilla average fitness individual at a given height/weight.

All the HRM makers have had calorie estimates for a long while now - so that's not unusual.
Shoot - Polar has cheapened many of their models from nicer actual HRM functions and basically just left them calorie counters.

Because actually there is a good correlation between HR and calories burned for exactly the type of activity I mentioned - and only that.

But it appears to me Fitbit is using what Polar and Garmin uses (though different studies).

You estimate VO2max from resting HR, BMI, and level of weekly exercise. All stats the Fitbit knows. Along with estimated HRmax. Accuracy isn't half bad, better than sub-maximal VO2 tests actually.
www.ncbi.nlm.nih.gov/pubmed/16168867

But that's why it takes a week or two for both Fitbit and Garmin to adjust figures. Polar is self-selected exercise level.

Polar and Garmin adds on to it, which I doubt Fitbit can get through it's method of reading HR, is Heart Rate Variability (HRV), which is used to get better estimate of HRmax.

I've actually created a form in spreadsheet that does exactly the same thing, then it uses a Polar funded research study formula to get the calorie burn - again public domain to use since public study. Polar's current formula is not the same of course, being secret.
www.braydenwm.com/calburn.htm

But mine takes into account a potentially better estimate of HRmax.
HRM tab is self-contained, the Garmin tab shares their method and study. The whole spreadsheet was done before the daily activity trackers became popular, and better initial estimates were desired.
https://docs.google.com/spreadsheet/ccc?key=0Amt7QBR9-c6MdGVTbGswLUUzUHNVVUlNSW9wZWloeUE


And totally just as accurate for cycling. Because 1000 an hr for that low of a HR sounds like big inflated burn wherever that estimate came from.

@heybales
Those cycling calories are not an estimate. That data comes from my power meter, and from lab testing I know my efficiency as well. That number is within 2% accuracy, possibly 1%. Even if I didn't know my efficiency from laboratory testing it would be accurate within about 3-5% as efficiency on the bike ranges from about 21% to 27%, so worst case scenario at 244w average you're doing 903 kcal/hr. In my case I know I'm just under 24% efficient, which means pedalling at 244w average for me burns 1010 kcal/hr.

The resting HR and VO2 is interesting, as is the study you linked. I'm quite curious how VO2 is impacted by weight though. Relative VO2 certainly could, but for calories burned we would be interested in absolute and that doesn't change if you add or lose mass.

If fitbit attempts to take all that into account...that's pretty cool! I'll be following to see to what extent it adjusts my estimations over the next few weeks.

PS: Very cool spreadsheet! I'll be playing around with it a little out of curiosity.

heybales

Ok - so you just have a diesel heart it sounds like, perhaps not a high HRmax, but for your available range, great power even at what sounds like lower HR.
Compared to say a motorcycle heart with high HRmax and bigger range.

That's great you got your efficiency %, since it seems like most power meter head units either just display the KJ's as calories, or convert using 22%.

My first 2 VO2max tests was a difference of 21 lbs lost, and while per kg it improved from 48.2 to 59.5 mL or 23%, actual volume from 4.4 to 4.9 L/min, or only 11%.
My main interest was AT/LT from 171 to 177, which I base training HR on.

And correct, the formula used to get calories are for absolute not relative VO2 - and that's where the BMI inclusion breaks down.
(CO2 L/min / 02 L/min) x 1.2318 + 3.8151 = Cal/min

That BMI aspect along with resting HR is an attempt to figure out if fit or not. So bad BMI is assumed bad VO2 level to some degree - but we know that's not the case always. The cheaper Polar's really rely on that aspect since they have none of the other values.
When you look at blogs of people reporting their VO2max tests - there are people that manage good training and good VO2max despite overweight and not as good a BMI.
For them cheap Polars would be really off, the nicer ones like it appears Fitbit would at least buffer that BMI aspect with restingHR and amount of exercise weekly.

I've not seen any research studies on it, but they could have their own research that further breaks down that weekly exercise between some rough zones, to help see if vast majority of exercise time is very low level exercise that wouldn't impact VO2 much, or includes some really good workouts that would.