Just got Fitbit: Counting steps on a elliptical

[Joy2]

Does the Fitbit count steps while doing elliptical

[SteveH · Accepted solution]

Not very well.

The problem is that fitbit depends on firm, distinct steps to be able to track properly and you do not get these with the elliptical. If it does track some steps the calories and active minutes are likely to be inaccurate.

The way round this is by manually logging the activity. This article describes how: 

http://help.fitbit.com/articles/en_US/Help_article/How-do-I-log-or-record-an-activity/

[ChristinaDi]

My elliptical step count is very accurate.  

[enny]

ChristinaDi, are you using the Flex or One. If you're using the One, where are you clipping this? Just got my One am anticipating my elliptical sessions.

[KC12]

I find Fitbit is pretty spot on with the elliptical. I use the One clipped to my waistband of my underwear.

[whome]

---

@ChristinaDi wrote:

My elliptical step count is very accurate.  

---

I have a question. How do you count steps when you don't take steps? 

Curious minds want to know

[KC12]

Your body is engaged in step-like motions on the elliptical. While on the elliptical (or bike), your lower body is moving similar to the way they move on a flat surface. Fitbit generally isn't advanced enough to figure out "okay, this leg movement is due to an elliptical, not a walk".

I suppose the placement of the Fitbit matters most. If you wear it around your wrist and leave your hands on the bars of the machine, the device will not detect movement. If it's on your ankle, the Fitbit might not think the circular motion is a step.

[Spyder95]

The flex comes with two straps - maybe with a second clasp they could be put together to fit around the ankle. This might also help with outdoor cold weather walking, when I keep my hands in my pockets.  Has anyone tried the flex on the ankle for elliptical or cycling?

[Heybales]

---

@Spyder95 wrote:

The flex comes with two straps - maybe with a second clasp they could be put together to fit around the ankle. This might also help with outdoor cold weather walking, when I keep my hands in my pockets.  Has anyone tried the flex on the ankle for elliptical or cycling?

---

The problem with that, especially with walking, is it'll only see the impact of 1 step, it won't see the other foot coming down.

So there is half the steps, there is half the distance, there is half the calories - or worse.

So depends on your purpose for doing it. To merely get a few more inaccurate steps than it would otherwise get from normal location - fine.

For accuracy of calorie count - terrible.

Especially because formula for calorie burn walking or running does not translate to elliptical or biking cadence.

[CourtB]

I never had any issue with my flex reading steps on the elliptical. My Charge Hr is fine too. Just make sure you are holding on to the moving arms for the entire stretch of the bars. I hold it at the very top of the bar above the HR sensors

[Spyder95]

Tested it on my ankle,  30 minutes on the elliptical.  Synched before and after.   Logged approx 2000 "steps".  Seems pretty accurate to me.   I don't need the calorie count anyway.  Next test,  stairs.   I frequently do 12 floors while on break, and much of the time I am either holding my cell phone, a cup of coffee (carefully), or have my hands in my jacket pockets.   Swinging my arms does not feel natural while walking stairs, so I would think the ankle would work better than the wrist.  Maybe I'm strange,  but my leg seems to move in rhythm with my arm, not at a double rate like some others.  I just don't 'get' all the anti-ankle sentiment on the forums!   I just enjoy seeing a graph of my activity,  which motivates me to do more. 

[James87]

I used mine on the elliptical on my left wrist and it seemed to be okay, not sure if its the exact steps but I was moving my arm/wrist each time so it would track a movement. 

[CpnT]

I received the FitBit Flex for Christmas and have been wearing it and logging activity pretty regularly.  One thing I've noticed is that when you manually record miles on the treadmill, it doesn't increase your total step count for the day.  I've also noticed that when I manually log activity on the elliptical, there is no option to add distance/miles, only time and calories.

Suggestion for FitBit:  Add the ability to track miles/distance for the elliptical, and calculate the pace just like manually logging time on the treadmill.  Additionally, it would be great if those manually logged miles updated your steps and miles recorded in the dashboard.

Community - what do you think?

[yldthng]

@CpnT I agree that this area needs some tweaking - unless I've missed something.

I purposely don't manually log for this reason. I do, however, activate activity mode before & after the elliptical so the activity and associated steps show up in my Activity Log. I do NOT go back and manually change the activity to say Elliptical either, because then you can't see the step breakdown. 

[Heybales]

---

@CpnT wrote:

I received the FitBit Flex for Christmas and have been wearing it and logging activity pretty regularly.  One thing I've noticed is that when you manually record miles on the treadmill, it doesn't increase your total step count for the day.  I've also noticed that when I manually log activity on the elliptical, there is no option to add distance/miles, only time and calories.

Suggestion for FitBit:  Add the ability to track miles/distance for the elliptical, and calculate the pace just like manually logging time on the treadmill.  Additionally, it would be great if those manually logged miles updated your steps and miles recorded in the dashboard.

Community - what do you think?

---------------------

When log running/walking, it will take your distance and calories as more accurate than what it estimated, and replace what it came up with for that time.

And that should include steps, which are calculated from your distance with your stride length.

If Steps didn't change, then Fitbit came up with that distance already.

If it goes down or up - your stride length is wrong in your personal stats.

For elliptical there is no standard way to translate action in to miles - despite the machines attempting to do it.

Too many variables without any true distance being done. The distance the foot pads move doesn't mean literal distance either, because your weight is supported and it's easier.

It's like the spin bikes trying to estimate distance - totally made up. Because they have no way of knowing what the friction from the pad is providing in the way of resistance, to match that up to a potential road speed that is based on road and air restistance.

[DrHI]

So I use a fitbit SURGE, and I use the exercise mode and specifically set the mode to "elliptical". I'm not sure this option is available on other fitbit devices.

In my experience, and in the gym I frequent, there is one elliptical machine that is VERY accurate. It is a precor type maching, with moving arms, and w/ an elliptical motion that is somewhat forceful w/ each step.

I make sure to use the swing arms, and hold the handles somewhat far up. 

It registers EXACTLY as I would expect, in sets of "two" for the motion of left then right steps. It seems to me to work best using the swinging arms.

That sad, there is another elliptical, simliar, but has a much  more gradual motion, almost like a long stride, which can vary depending on how fast you move. This stride does not have that foreful impact the other machine has, and the swinging arms are much slower and less distance covered w/ each movement.

Maching w/ forceful impact worked well, registered about 7K steps for 1 hour of intense work, where as the other maching measured 724 steps.. 

So it works, but certainly depends on the impact created w/ each step, not sure how much the arm movement part has to do, but may test this later!

Speaking to calorie burn, did a 1hr50 min stairmaster session (was motivated today for some reason), and Fitbit says i burned 1950 calories. Using the maching calculation (which including putting my age and weight), it turned out to be around 1200. The elliptical measures over 30% less than fitbit Surge.

In my measurements I usually just take another 10% off the elliptical and feel ok w/ that measurement. My heartrate stays over 160 for the whole workout, just not sure how accurate the calorie burn figures are at the moment, so being cautious until I feel more confident about fitbit measurements.

Anyhow, any other questions, ask away.

Good luck, and good stepping 🙂

[Heybales]

---

@DrHI wrote:

So I use a fitbit SURGE, and I use the exercise mode and specifically set the mode to "elliptical". I'm not sure this option is available on other fitbit devices.

In my experience, and in the gym I frequent, there is one elliptical machine that is VERY accurate. It is a precor type maching, with moving arms, and w/ an elliptical motion that is somewhat forceful w/ each step.

I make sure to use the swing arms, and hold the handles somewhat far up. 

It registers EXACTLY as I would expect, in sets of "two" for the motion of left then right steps. It seems to me to work best using the swinging arms.

That sad, there is another elliptical, simliar, but has a much  more gradual motion, almost like a long stride, which can vary depending on how fast you move. This stride does not have that foreful impact the other machine has, and the swinging arms are much slower and less distance covered w/ each movement.

Maching w/ forceful impact worked well, registered about 7K steps for 1 hour of intense work, where as the other maching measured 724 steps.. 

So it works, but certainly depends on the impact created w/ each step, not sure how much the arm movement part has to do, but may test this later!

Speaking to calorie burn, did a 1hr50 min stairmaster session (was motivated today for some reason), and Fitbit says i burned 1950 calories. Using the maching calculation (which including putting my age and weight), it turned out to be around 1200. The elliptical measures over 30% less than fitbit Surge.

In my measurements I usually just take another 10% off the elliptical and feel ok w/ that measurement. My heartrate stays over 160 for the whole workout, just not sure how accurate the calorie burn figures are at the moment, so being cautious until I feel more confident about fitbit measurements.

Anyhow, any other questions, ask away.

Good luck, and good stepping 🙂

---

Very nice analysis, and no, the other devices don't have the select exercise yet.

That would allow them to use other algorithms for calorie burn though, that could indeed be more accurate.

I'm thinking yours with HRM was using the HR formula for calorie burn, not step based though.

As such, there are several reasons why that HR method can be very badly accurate.

Several assumptions have to be made since there are no stats available to change or even view that I've heard of.

HRmax, and VO2max.

HRmax is usually calculated to 220-age. But that is very huge bell curve, more chance of being 10 bpm or more outside of range than in. My measured is 194, and hasn't decreased in years. My calculated says 174. Which is actually under my lactate threshold, or top of my aerobic zone before going anaerobic.

So you could have that issue of HRmax just being higher than they guess, either genetics or kept fitness.

So a high avgHR in a workout would appear you are burning a lot more.

For instance, using that 160 for my stats. From measured that's 82% of max, using calculated that's 92% of max. Guess which one would lead to higher calorie burn estimate.

VO2max is other stat, a few HRM's try to estimate it with more recent study formula, and decently up to a very fit level. But without the required stats (resting HR and workout frequency), it's totally estimated.

Most assume that if BMI (height & weight) is bad, then fitness level (age and gender) expressed as VO2max is bad too.

And that's a bad assumption, because you can get into cardio shape must faster than you can lose weight.

So if bad BMI, it assumes that high HR is big calorie burn for that low VO2max.

Another factor, you could have an inflated HR for other factors - genetics, medical, 5 espressos prior to workout, meds, dehydrated, stressed body from unrecovered prior workout, ect. Nothing to do with workout load, just inflated HR, so then inflated calorie burn.

I wonder if the slope on that 2nd machine could be changed to allow better impacts. Then again, some people use it differently.

That's why there is no good study formula's for elliptical - too many variables including how the person rides it.

edit - spelling corrections.

[DrHI]

Thanks so much for reply.

Although I do not know my v02 max, I do know my foced vital capicty as tested w/ a respirtator to be above average, clost to 7.6L I believe. I do have trouble with chest x-rays as the machine does not usually fit my entire chest on one go (in my limited experience w/ getting x-rays :).

Absolutely caffeine is a factor, I drink coffee like water, and I do know that it elevates my HR during day.

My resting HR since working out over 3 weeks however has decreased from 69 to about 62. 

I think certainly those factors are at play here....

I wonder if caffeine impacts exercise heart rate more so that resting rate? Is their a non-linear type relationship b/w these measurements? If one takes 200mg caffeine, does one have exponential increase in HR in exercise vs. resting state?

So to me I am getting an elevated caloric burn based on current rigid fitbit criteria. I think it would be great to have  modifying factors available (lung measurements...tidal volume, total lung capcity, etc.. VO2 max if availabe), those types of adjustments on as needed basis (and only for those that care to modify) could be made availabe I think w/ the abundance of current medical lit.

Finally, to above poster, is even the elliptical machines measurements deemed to be accurate at this point? (I do enter age and weight). 

Was hoping for a clear cut answer on what value to shoot for energy consumption, but it seems to me, that using the available online TDEE, subtract 1000 a day for a 2lb weight loss per week may be safest bet. And to disregrad the calorie expenditure from exercise?

Anyhow, using just TDEE measurements alone, and logging my calories in myfitnessp (goal at 1600 cal a day), given the fudge factor of counting calories, and fudge factor of exercise cals burned, seems to work w/ weight loss.

Either way, fitbtit has got me, only because I'm competetive and those step counts keep me moving!

Keep stepping,

HI

[Heybales]

Here's a study that looked at various amounts per kg of body weight, no increase during the exercise for this small group and the whole range of exercise they did. Not habitual coffee drinkers prior. So those amounts just depends on you. The 5 espressos I mentioned would be well above the amounts they took.

http://www.jissn.com/content/4/1/11

I know some meds cause bigger inflation in exercise, because heart is already under stress, and then with exercise added just adds more stress, so even higher yet.

Ya, a place to view and manually correct stats needed for HRM formula would be useful. They could even include the VO2max study formula, since they already have the restingHR in the morning. Just need to ask, or look at historical record, of workouts per week.

That would be really smart if they did automatically, even if they didn't allow user changes.

I guess only way to tell would be maintain same workout with same HR for couple weeks and see if calorie burn changes in a big way showing the influence of historical workouts.

But even Polar doesn't do it automatically, so not holding my breath Fitbit would.

Elliptical machine calorie burn formula's are mainly their own, with whatever level of research they cared to apply to it. Studies have found no good formula because you can adjust too much, slope, resistance, others depending on machine, and machines have different styles, and so do people.

That's why even the very accurate treadmill formulas for pace and weight for calorie burn start to get inaccurate as incline goes up, personal hill running style changes too much to be accurate.

I know one machine used the same research as was found on cyclergometers, get the watts put forth by the user, and figure the human body is about 24% efficient in making them. There's your calorie burn.

Does that machine show watts?

I'd suggest keep replacing your exercise calories with better estimates, since it seems your HRM is reading very high. Machine is better it sounds like.

But it for sure counts. Because 2 lbs weekly is already aggressive goal unless you have like over 60 to lose, and 1600 on rest days sure sounds low for a guy at that size.

No need to make the deficit even more unreasonable by making it bigger by 500 or calories on workout days. Bigger is not better in this case, unless muscle mass loss is the goal.

Let Fitbit correct that MFP estimated daily burn without exercise to increase your eating level.

You just need to decide best way to correct Fitbit manually.

[Sambre]

I wear my Charge 2 around wrist, and it still tracks my elliptical quite well