Export and analyze your own data in Excel

Question asked

I have a question about your export tool. I am wondering if the tool takes into account the type of workout, intensity of workout, or duration, or is it simply based on number of workouts logged? Or is this the reason for the different estimates for TDEE, to compensate for the different workout intensities?

That's a good question - I'll try to address that here but if it isn't clear please let me know.

It does use the workout/burn data but let me get to that.

Covering Total Daily Energy Expenditure(TDEE): For the calculation of observed TDEE or theoretical TDEE the tool does not need work out data. Let's look at these and see why.

First, what I call theoretical TDEE is just your TDEE from the standard equations - it is based on your activity level and this is why in the graph you see different curves for theoretical TDEE. Now observed TDEE is based on calculating your TDEE from the amount of food calories you consume and the average weight loss you get (and we assume that weight loss is 3500 cals per lb lost...). So neither of these use directly the workout/burn data. This calculation is consistent whether you enter every workout or not.

As of the latest version, I also report theoretical a new number [TDEE - exercise] - this takes the amount of calories burned reported from exercise, a fitbit, etc.. and subtracts them from the tTDEE. [TDEE-exercise] allows you to see what your daily need might be if you didn't have those reported calorie burns. In order to calculate that the program only looks at calories burned - there is no data on frequency, duration or intensity that is used. But, in general these factor affect the calorie burn - they are part of what makes the burn a burn.

How does this affect you and what can you do with the data?

If you find that the Activity Level result is relatively low (1.2-1.4 or less) it means that the estimated daily exercise level plus general activity *might* be low. (Under reporting calories eaten will also give you that.) If this is the case, consider upping your activity level during the day - this is lifestyle stuff not just more exercise sessions.

Your observed TDEE - if the diary data is accurate - tells you that this is your possible maintenance calorie level that you are experienced. Eating 5%-20% below that level results in the "reasonable" weight loss window.



The three top straight lines in the graph are just different activity level for theoretical TDEE (calculated from weight, gender, age and bf%), the fourth straight line is the theoretical Resting Daily Energy Expenditure (someone who is sedentary) vs your observed TDEE data, the dark orange line (from weight loss and diary calories reported) and with exercise subtracted, the light orange line.

Do you think it would be useful for the non-exercise TDEE that might help to decide to change your daily activity level (for those doing MFP method I'd gather) to use the MFP multipliers rather than the TDEE charts that include exercise?

It seems if you are going to take any TDEE and subtract the exercise back out, you are now getting a value that is really only going to be useful doing it MFP method, since you must be adding exercise back in.

Because their multiplier's are based on new study research and are different than the TDEE charts.

Sedentary - 1.25
Lightly Active - 1.4
Active - 1.6
Very Active - 1.8

Ok, I can update those. And yes, I think that TDEE - ex it really is useful only for MFP settings (a pseudo RMR if you are really logging all exercise)

If I get this correctly, you are also suggesting giving a table with

(oTDEE - exercise) * activity factor = MFP non-exercise TDEE

I could also include an x cut for y lbs per week subtractor.

I'll see what I can build out without creating confusion.

Do you think it would be useful for the non-exercise TDEE that might help to decide to change your daily activity level (for those doing MFP method I'd gather) to use the MFP multipliers rather than the TDEE charts that include exercise?

It seems if you are going to take any TDEE and subtract the exercise back out, you are now getting a value that is really only going to be useful doing it MFP method, since you must be adding exercise back in.

Because their multiplier's are based on new study research and are different than the TDEE charts.

Sedentary - 1.25
Lightly Active - 1.4
Active - 1.6
Very Active - 1.8

Ok, I can update those. And yes, I think that TDEE - ex it really is useful only for MFP settings (a pseudo RMR if you are really logging all exercise)

If I get this correctly, you are also suggesting giving a table with

(oTDEE - exercise) * activity factor = MFP non-exercise TDEE

I could also include an x cut for y lbs per week subtractor.

I'll see what I can build out without creating confusion.

Oh, I wouldn't go that far probably, but your point in description with graph about being perhaps more active than you thought you were, and selecting another activity level to better represent your non-exercise portion of day, I think is good point.


I'll also mention regarding the Fitbit adjustments that show up as exercise entries, those are not just the calorie burn from exercise, but the difference between MFP non-exercise maintenance, and Fitbit TDEE, which may or may not include exercise.

I've had days where really sedentary, except for the short intense workout. That was manually logged on Fitbit since it had bad low estimate on the burn (ie non-step based).

But when the compare with MFP happened, no adjustment, so it appeared no exercise. Or similar, big long workout, and indeed lazier rest of the day to recovery. The adjustment is easily 200 less than the workout was logged as.

Not sure if you handle those with caveats in the instructions, or figure someone with Fitbit is already getting their TDEE estimate there.

I have similar experiences in my fitbit logging of exercise and have given this some thought as well. Strangely enough, you would actually increase the accuracy of the overall estimate by *reducing* the MFP assumed activity level (as this hands off the calculation to fitbit more quickly/often which for this purpose are assuming is more accurate than MFP in the TDEE assumption (especially on no/low-exercise days).

I can't imagine the extraction tool would be able to do anything to deal with this problem without pulling the data directly from fitbit.

Although, I suppose allowing negative adjustments from fitbit on MFP would handle this, right?

Couple of comments:

- observed TDEE (oTDEE) doesn't use exercise or burn data; it's what you eat and gain/lose in weight that is used. It assumes a constant relationship of 3500 for each lb gained/lost (which isn't exactly true). Therefore the calculation isn't impacted at all by the fitbit adjustment or errors in evaluating or recording exercise burns.

- by subtracting burns from oTDEE to create a "non-exercise" oTDEE we introduce those possible sources of error - fitbit corrections as exercise, burn data variability

The resulting non-exercise oTDEE can be used to set-up MFP custom Net Calories Consumed* but what happens with those extra bitfit calories?

Let's say that the fitbit is adding on 200 extra cals and exercises give 300 cals per day average. Let's build up an sample situation:

Settings = 2000 cals
oTDEE = 2500 cals
burns=500 (200+ fitbit, 300 recorded)

There wouldn't be an issue because the fitbit is only adding the extra cals missed from exercise and MFP setting. If the person recorded 400 exercise cals, the fitbit will only add 100 as it estimates the total burn and eliminates the exercise period - if the user wrote the correct duration when recording the exercise burn.)

Does this make sense? Do you guys see it otherwise?