Intended use of Fitbit and HeartRate tracking

[Peggy9970]

Yesterday I purchased a Versa 2, my very first fitbit device. 

My intended use is to monitor and track my heartrate while sleeping and associate that with "silent" atrial fibrillation events that might occur during sleep. 

One thing that I know for sure is that atrial fibrillation episodes are evidenced by significantly elevated heartrate that may last for several minutes to an hour or more.

So I reasoned that if a fitbit were able to monitor and track my heartrate overnight I might be able to (after the fact) detect AFIb events. 

So my first question is this: Have I purchased an appropriate Fitbit?

My second question is: Can the Versa2 Heartrate graph (BeatsPerMinute versus time of day) be expanded along the X-axis to be able to see the exact time that a heartrate spike occurred (say within a 2-3 minute window?  

If the answer to either of these is "NO", then is there another Fitbit model that is more appropriate with my intended use?

Thanks,

[LZeeW]

Hello @Peggy9970 and welcome to the Help forum.  I'm going to comment on your second question first.  When you look at your heart rate trace from the heart rate tile, you can expand it and rotate it onto landscape view.  You can also lightly press on the screen and slide your finger back and forth to see heart rate values.  But the number you see is your average heart rate over a five minute period.  It is unlikely that something that lasted 2-3 minutes could be detected .  People get heart rate changes simply changing positions in their sleep.  Premium users get a sleeping heart rate graph as part of the advanced sleep analytics.  This graph shows average heart rate over a one minute interval.  Again, there are spikes for different reasons.

As for your first question, NO Fitbit is considered to be a device that is intended to suggest the presence of atrial fibrillation after the fact.  As a matter of fact, my husband had exactly what you are describing -- paroxysmal atrial fibrillation -- where you are in atrial fibrillation some of the time, but not all of the time.  He never said anything about seeing heart rate tracings like the ones you are hoping to find.

The Sense and the Charge 5 have the hardware to record a 30 second ECG with the ECG app.  But this is done with you awake.  It isn't available in all countries.   (link) 

I hope this helps.  Please feel free to ask other questions if you need more information.

[Peggy9970]

Hi Laurie,

First I'm going by the name Peggy9970 because it is my wife's' fitbit and she has the Afib. 

With regard to your first comment. What I'm looking for is a method of detecting these silent Afib events during sleep and at what time they occur. I would ideally like an exact time (like 2:02:15 until 2:45:57) but would be content to have a "fuzzy" start and stop time that be about 2am until 2:50.  So a moving average of the previous 5 minutes is not a deal breaker at all. (BTW, I assume from what you've said that it is a moving average that gets reported and graphed.  Since these Afib events (for us) last more than ten minutes, it should pick those up, although the start and end times will be inexact.

 Now, to a actual event: this morning my wife awoke about 6:10am; at about 7am she felt like she was having afib, that lasted about 30 minutes or so; later a short bout that lasted 5 minutes at the most. 

Here is a screenshot of her heartrate today:

(Well apparently I have to learn how to insert a photo on this board)....get back to you after I figure it out!

[Peggy9970]

Well I still haven't figured out how to include a photo in a reply, but while I'm working on that. may I ask for confirmation about the heartrate averaging.

I think that you've said that the numerical value for Beats per minute that I see displayed on the Beats per minute graph are actually the average of five minutes of observed beats (which makes sense as the tin=me period between these numerical values is five minutes (as in 8:05, next is 8:10, next is 8:15). 

But my question is this: Is the BPM numeric at 8:05 the average of all the observations between 8:00 and 8:05, and completely independent of any observation taken seconds before 8:00? If so, how many observations (of BPM) are made during that 8:00:01 to 8:04:59 period?

I want to be sure that the numerical BPM is an absolute average and not a moving average.

Lastly, I saw some discussion on another Fitbit community board that seemed to suggest that what fitbit records as a beat is actually a change in capillary blood volume at the wrist and is not necessarily related to the beating of one's heart. The suggestion was that a fitbit BPM is not at all comparable to BPM from a ECG or a Kardia device (consumer ECG,  https://store.kardia.com/products/kardiamobile?gclid=EAIaIQobChMI-qfYs8OP9AIVDppmAh3xTQL_EAAYASACEgL...). Any update on that score?

[Adrixx]

Quick question, I have a Verse Lite. It doesn't even accurately track heart rate at rest,.awake or asleep. I have verified with constant monitor at the hospital, it's not close. Is the Sense or Charge 5 more accurate? I really need something more accurate to track vitals.

Thanks

[LZeeW]

@Peggy9970   I want to clarify that Community Council members are not Fitbit employees.  We have a lot of experience with Fitbit devices and try to help other users.

• I believe that the 8:00 to 8:05 time period is exactly as you describe it.
• Everything Fitbit does is proprietary and they do not share this information.
• You are correct about how Fitbit measures heart rate with the optical sensor (green LEDs).  It looks for pulsation in tiny blood vessels.
• You are correct about the Kardia.  It records an ECG.  The Fitbit Sense and the Charge 5 have the hardware to record an ECG.
• Fitbit conducted its own Heart Study last year and I was a participant.  Its goal was to look for the ability to detect atrial fibrillation using the optical sensor.  Fitbit never announced any results.  Apple did a similar heart study.
• The Versa 2, or any Fitbit measuring heart rate with the optical sensor, is not designed or intended to detect atrial fibrillation before, during, or after the fact.

[Peggy9970]

You said: "The Versa 2, or any Fitbit measuring heart rate with the optical sensor, is not designed or intended to detect atrial fibrillation before, during, or after the fact" and I am certain that that's what the lawyers will say (because Fitbit products are not FDA approved for that purpose).  

However, what I have found in actual practice is that the Fitbit Versa 2 is adequate to record significant change in heartrate (say a change from 60 to 120). When that increased rate is sustained for more than two to 3 minutes (enough time to overcome the averaging effect), the heart rate track can give a fairly accurate approximation of the time when heartrate began to increase. 

The very nature of atrial fibrillation is increased heart rate. 

What I have confirmed (by actual use) is that for my wife, the fitbit versa2 does a good job of detecting a "silent or sleeping AFIB" that occurs without any awakening, provided that the heart rate increases by a minimum of 60% and the episode has a duration of greater than 4-5 minutes.

(for example BPM shift from 58 to 93 within the 5 minutes averaging window and remains at the elevated rate for greater than 5 minutes)

In short, It works as I had hoped and within my expected limits to let us know that she has had a night free of obvious AFIB. 

BTW, one of the many reasons we want to see if we can "surmise" an silent AFIB episode is to help explain mysterious fasting blood glucose increases. We also know for a fact that AFIB will likely cause a liver glucose dump that artificially increases the blood glucose (that is to say, a glucose increase that is not associated with ingestion of food). Observation of Fitbit heartrate, in combination with fasting glucose monitoring over the next month or so may now help to explain "mysterious" FBG and also provide more input to her doctors. 

[SunsetRunner]

Yeah, same with Vera 2, used to be pretty good, but since the last update it doesn’t track the heart reset anywhere near accurate as I noticed when at the emergency department a few weeks ago.  It used to be so good my gp would check my heart rate with the Fitbit then check what he got, now he doesn’t bother because he knows it is no where near accurate now

[Adrixx]

I agree with you totally. My nurses and docs have gone over my Fitbit info. I can no longer use it to track my condition. I used to be able too. The sleep and step counter are not the reason I have the tracker.

[Peggy9970]

Sorry to hear that your device is failing to achieve desired accuracy. 

Perhaps it's because my Versa 2 is new, but the agreement of the watch display and my two reference devices is actually quite good for a consumer device. 

MY references are (1) Kardia 6 lead EKG that uses electrical sensors to detect heart rate and (2) a professional grade BP device that uses audio to detect heart beats (and the difference in tone between systolic and diastolic beats). Both of these devices agree with my GP and cardiologist's office devices. 

BTW, when I say "quite good" I mean within 5BPM.

Who knows, perhaps when my device celebrates it's 30 day usage the accuracy will go to pot. But in the meantime I happily satisfied that the 5 minute averages (from the graph) are sensitive enough for my detection of atrial fib.

[SunsetRunner]

No, we have three Vera 2’s that were new replacement for the ones that were just a few months old when this update screwed up, straight out of the box again these ones were giving inaccurate readings.

[Peggy9970]

I'm sorry, I can't explain your situation. My Versa2 (purchased 10 days ago) is working well. 

Just to confirm I'm comparing the BPM shown on the clock face with a Kardia ECG. The ECG will show individual heart beats and average them over a 30 second period (and by the way I'm comparing while in normal sinus rhythm). 

So today's test: Kardia reports 68 BPM, Versa2 watch display during that same 30 second period: 67, 66, 68, 67, 69, 66.

Sorry, you having difficulties, but mine (for the moment) is working fine.