Fitbit has 150 billion hours of "anonymized" health data

Originally published at:


What I wonder is how many people take their Fitbit off when they go to bed. I know I do. The screen would wake me up, either by turning on or clonking myself with it.


I mis-read that as “Fisting Heart Rate”

Which would be an extreme violation of privacy…


I have a Charge 2 and I wish I could sleep with it on because the silent alarm feature is really nice for when I need to get up and my wife doesn’t. It’s too uncomfortable for me to sleep with.

I leave my Alta HR on; it’s not too big and I want the sleep data :wink:

What I wonder is how representative this dataset is of the general population - a largely self-selecting cohort of wearers might have appreciably different characteristics to non-wearers.


I wore mine, and now wear my Apple Watch to bed. I love the sleep tracking.

David Pogue writes for Yahoo now? The mighty NYT tech writer must have really fallen on hard times.

ETA: Now that I’ve actually read the article. Wow. That was fascinating. I can’t help thinking that this data is skewed, though, because the people who would buy a fitbit are not representative of a country as a whole.

Also, it’s great for the users of fitbit, but from a scientific standpoint, I would have liked to have seen more correlations on personal data about the fitbit people, like eating habits and overall health. Like is it possible that those who get 9+ hours of sleep are more likely to be depressed, and so they would have higher heart rates because of that?


I’m just thinking anything in the cloud is bad news. Nothing is safe, everything is vulnerable and exploitable.


Fitbit can also tell how many people in its data set have 15-30 minutes of vigorous activity immediately before falling asleep.


What is surprising, though, is that the benefit tapers off after a couple hundred minutes of exercise.

It’s not that surprising is it? After all, the usual advice is to get regular, moderate, exercise. Not for everyone to go and do an iron man each weekend. (Not to mention the more exercise your do, the higher the risk of injury or illness caused by that exercise).


[Europe] have designed their cities so that there will be more physical activity. People have to walk more just do normal activities

European cities are not designed, they are adapted from how our ancestors built them. A better way to put it would be to say that “European cities are not designed for cars”, so instead people have to walk.

I do own a fitbit, (well, a Pebble) but I disabled the fitness side of it because I couldn’t see any use for it.


Having had a couple of fitbits, I’d be tempted to move the headline’s quotation marks over to “data.”

I haven’t been doing it. I think I found with the watch it would light up too. Also then you have to get into a different charging routine. I guess the morning hour before heading out would work.

But now that you mention it, where are THOSE studies?

Oh, yeah, THAT’s why I’m not…

1 Like

Are either of you considering a fitness band that has an blood oxygen saturation sensor for sleep tracking? Both Fitbit and Garmin have them (I assume Apple does or will) and are now including them in the Fitbit Charge 3 and the Garmin Vivosmart 4. Garmin can do sleep pulse ox tracking out of the box, but it seems like Fitbit hasn’t added the feature yet even though their new sensor hardware supports it. It’s yet another field of data for Fitbit to hoard, but I’d like the most sleep tracking info possible.

1 Like

I suspect that while many Fitbit owners really are heath nuts, manyare like the people who buy a big mac, large fries and a diet coke and are looking for a way to minimally improve and track fitness. People who actually live a healthy lifestyle probably don’t really need fitness tracking (not that they can’t use or possibly benefit from it). So I wonder if the data is skewed by a number of different populations.


My Apple Swatch Series 3 is only 3 months or so old, so I am going to stick with it until there is a new feature I find irresistible.

1 Like

Interesting data, but has it been established that differences of 3-5 bpm out of say 65 (5-10%) really has an effect on overall health?

Bad or misleading visualization techniques doesn’t help. Having a Y axis not start at 0 makes small differences seem more significant - sometimes innocently but often with an agenda (stock price deviations, etc).


A known issue with the Fitbit Versa is that the heart rate monitor is inaccurate. My Versa consistently detects my heart rate 30-40 bpm higher than what it actually is, and I am female. Since the sleep monitor uses heart rate as part of their formula to determine which stage of sleep you are in, that information is also inaccurate. It’s possible that the other Fitbit models have these inaccuracies. So we are talking about millions(at a minimum)of hours of flawed data.

This topic was automatically closed after 5 days. New replies are no longer allowed.