3 ms·

> The Fitbit Inspire, Fitbit Versa, and Oura Ring estimated light sleep with less bias than deep sleep. Conversely, the Withings Mat and Garmin Vivosmart estima

> The Fitbit Inspire, Fitbit Versa, and Oura Ring estimated light sleep with less bias than deep sleep. Conversely, the Withings Mat and Garmin Vivosmart estimated deep sleep with less bias than light sleep. Across devices, estimates for REM sleep were less biased than estimates of light and deep sleep. Despite low absolute biases, it should be noted that some estimates were still up to 250 min different than their PSG derived measures. In other words: these devices are all terrible, and shouldn't be trusted. Pretty different from the final conclusion, which is probably "soft" to get additional grants from the commercial entities selling these devices.

marcinzm 3y ago
I mean, given the study they shouldn't be trusted as a single night measure of absolute sleep quality however I don't know of anyone who uses them as such. You use them over long periods of time to either get aggregate statistics to compare against baseline or track personal changes over time. edit: Also, the last time I did a sleep study at home it had rather little correlation to how I normally sleep except in terms of apnea. Having wires, electrodes, a chest strap and a finger O2 monitor strapped to me did not lead to a pleasant night of sleep.
- pessimizer 3y ago
  > You use them over long periods of time to either get aggregate statistics to compare against baseline or track personal changes over time. Doesn't that assume that the errors will be consistent and precise enough that any trends don't get totally wiped out?
GIFtheory 3y ago
I think the conclusion is that /total sleep time/ is actually very accurate, while deep/light/rem measurements are crap.
- 2devnull 3y ago
  All they have to do is measure time in bed, which for most people most of the time will correlate closely with total sleep. Insomniacs are outliers so you don’t need to worry about measuring them and most are taught to just get up when they awaken anyway.
ilaksh 3y ago
You skipped the section where the measurement of total sleep for most devices was found to be more than adequate. Detecting light versus deep sleep accurately with wristwatches or devices under the mattresses might not be possible. Especially not when accuracy is supposed to be comparable to multiple electrodes places directly on the head to measure brainwaves.

marcinzm 3y ago

I mean, given the study they shouldn't be trusted as a single night measure of absolute sleep quality however I don't know of anyone who uses them as such. You use them over long periods of time to either get aggregate statistics to compare against baseline or track personal changes over time. edit: Also, the last time I did a sleep study at home it had rather little correlation to how I normally sleep except in terms of apnea. Having wires, electrodes, a chest strap and a finger O2 monitor strapped to me did not lead to a pleasant night of sleep.

pessimizer 3y ago
> You use them over long periods of time to either get aggregate statistics to compare against baseline or track personal changes over time. Doesn't that assume that the errors will be consistent and precise enough that any trends don't get totally wiped out?

GIFtheory 3y ago

I think the conclusion is that /total sleep time/ is actually very accurate, while deep/light/rem measurements are crap.

2devnull 3y ago
All they have to do is measure time in bed, which for most people most of the time will correlate closely with total sleep. Insomniacs are outliers so you don’t need to worry about measuring them and most are taught to just get up when they awaken anyway.

ilaksh 3y ago

You skipped the section where the measurement of total sleep for most devices was found to be more than adequate. Detecting light versus deep sleep accurately with wristwatches or devices under the mattresses might not be possible. Especially not when accuracy is supposed to be comparable to multiple electrodes places directly on the head to measure brainwaves.