My opinion only, but I think the sleep analysis is just a very vague guide. In the past I have started the sleep mode, then read in bed for an hour. The band shows that I fell asleep in just a few minutes. One night I got up to use the bathroom and the band indicated I never woke during the night. I still enjoy using it and it often reflects what I think happened during the night, for instance if I have a very restless night, it shows up accurately. I have restless leg syndrome and I sleep with CPAP, so my sleep is often interrupted.

UP tracks your movements, then uses an algorithm to decide whether you're awake, restless, or asleep. It's only good for studying trends. And "light" vs. "deep" sleep is something Jawbone invented to help you gauge the quality of your sleep: https://jawbone.com/kb/articles/soundsleeper.html

I have a screening appointment tomorrow for a sleep study at a hospital. I'll try and ask how UP & Fitbit track sleep. And if I get to do the study, I'll wear my UP and see how it compares to what the experts say.

Apparently I'm getting up and taking steps in the middle of the night multiple times, but I never remember doing it. I'm thinking, based on the number of steps, that I'm heading to the bathroom. Or the app is tracking awake times and steps when I'm just tossing and turning in bed. But the steps are almost always 30-60 steps.

The best sleep tracker apps are not fitness bands. For iPhone, the best is Sleep Cycle, for Android (and Pebble/Pebble Time w/Android phone), Sleep As Android, and the best "band" for sleep is the Basis Peak watch. Basis has performed best overall in studies.

They all perform better than fitness bands, but are no substitute for a laboratory sleep study.

I only have the Jawbone UP Purple app for my phone, which doesn't track sleep. But I have seen the graphs of others. And the graphs are not nearly as precise as my sleep tracker: Sleep As Android. It is now adding heartrate support and has added awake-detection algorithms. Still, it's only an indication of your activity at night. It cannot read EEG activity in your brain and so is a good approximation.

I also use a detailed widget-controlled diary called Sleepmeter. Android only. It matches up very well with my auto-sleep-activity tracker, and I use Sleep Cloud from Sleep As Android to fine-tune the diary. I know how much I slept. But I don't really know about deep, light, or REM sleep. You need a laboratory study for that.

The UP graphs have much overlap between deep and light, and I don't miss not having it.

Sleep well,
Robin

If you are wearing a band or watch, it should not pick someone else's movement.

Jawbone actually gives you much MORE "sound" sleep than dedicated sleep apps. (Only Jawbone calls it "sound" sleep. It's usually called deep sleep where you're not moving. In REM sleep, you're paralyzed by sleep. That happens only a few times a night. In very deep sleep you can still move quite a bit! Just not while dreaming.)

You should have a maximum of 70% deep or "sound" sleep but actually, levels in the 40 percentage and 50 percentage range are generally accurate.

It shouldn't go below 30% sound or deep sleep.

Robin

extremelygrumpycat wrote: »

To editorgrrl - that should be interesting to see how the bands hold up vs what the experts say.

I was able to wear my UP2 and Fitbit during the sleep study, but I haven't yet received the results. The technician said I toss & turn a lot.

My UP says 6h38 total sleep, 4h40 sound sleep, and 1h58 light sleep. (If I did the math right, that's 70% sound sleep.) I took 22 minutes to fall asleep, was awake for 46 minutes, and woke once.

For comparison, my Fitbit says I woke 4 times and was restless 14 times, for a total of 47 minutes awake/restless. (I remember waking at least 4 times.) The total sleep & time to fall asleep were the same, but Fitbit doesn't do "sound vs. light sleep."