Hacker News Reader

Comment by bicijay | Hacker News Reader

But how do i know my sleep quality?

If you don't want to have to worry about charging things, or wearing them to bed, Google Nest Hub is great: https://store.google.com/us/product/nest_hub_sleep_sensing?h... .

It can also track sleep stages, etc. This has been on sale for < $75, which is way less than a wearable, etc.

https://ai.googleblog.com/2021/03/contactless-sleep-sensing-...

https://ai.googleblog.com/2021/11/enhanced-sleep-sensing-in-...

That tells Google about your sleep patterns (and more) which in my opinion is a no-go. This type of data should only be shared on a consensual basis and only with those who use it for your benefit or (truly anonymised) fit research purposes. Building a profile for advertising and whatever other nefarious purposes is not for your benefit. The same is true for other "connected, individualized" devices like Apple's iThings and others.

> Building a profile for advertising and whatever other nefarious purposes is not for your benefit.

"Your sleep data is handled responsibly. Your coughing and snoring sound data doesn't leave the Nest Hub, only sounds events are sent to Google servers. Consistent with our commitment to privacy in the home, your sleep data is not used for ad personalization."

https://support.google.com/googlenest/answer/9415830

And going further:

"This also includes answering clearly if your home sensor data sent to Google is used to show you ads. And so we commit to you that for all our connected home devices and services, we will keep your video footage, audio recordings, and home environment sensor readings separate from advertising, and we won’t use this data for ad personalization. When you interact with your Assistant, we may use those interactions to inform your interests for ad personalization. For example, if you ask, “Hey Google, what’s the weather today?” we may use the text of that voice interaction (but not the audio recording itself) to show you personalized ads."

https://support.google.com/googlenest/answer/9849262

> For example, if you ask, “Hey Google, what’s the weather today?” we may use the text of that voice interaction (but not the audio recording itself) to show you personalized ads."

That is to say, they use sound data - in processed, not raw form - to build a profile for advertising and whatever other nefarious purposes not for your benefit. That's on top of interns listening to some of the audio to ensure the service works as expected, that is to say recordings of you having sex will be passed around the office.

OK, I think you didn't understand the whole statement or are twisting it to meet your narrative. It sounds like you have an axe to grind, so I won't really go further.

For others, your answer is available at https://support.google.com/assistant/answer/10176224. I suggest people read the privacy policy themselves.

If I untangled it right it reads:

“the data from the sensors don’t leave the device. Instead we extract events and interactions from this data and send back home to do whatever we wish with it.”

The policy you link makes statements about “hey Google” but for me it’s still very unclear what apply to this sentence vs others it records. May be my far from perfect english understanding.

Does this work if you are sleeping in bed with a partner?

Yes. It can even tell you when you're snoring (instead of your partner).

Skeptical it can glean all of that without EEG, let alone without heart rate.

Heart rate can be discerned remotely using infrared cameras [1] so that would be doable.

[1] https://www.thinkmind.org/articles/smart_2017_2_20_48002.pdf

Check out some details in the second blog link.

Of course a sleep clinic with a lot of sensors will give you much better results.

The analogy would be a massage chair vs masseuse. Different price/convenience tradeoffs.

Even if I determine my sleep quality…then what? If it’s bad, and I am eating healthy, workout and don’t have apnea what am I supposed to do with this information?

You could try different combinations of beds, pillows, blankets, white noise, blackout curtains-- there are lots of different ways to adjust your sleep environment to do a bit of trial-and-error to find out what helps.

I found that I sleep best with blackout, loud white noise, a very soft pillow under my head, and a firm pillow beside my legs so I can sleep on my back with one leg bent and resting on the leg pillow. The bent leg thing is not something I've ever heard of anyone else doing. I just had to figure it out by trying a ton of different ways of arranging my limbs in bed-- previously I had really struggled with hip pain during the night no matter how I slept.

also ambient temperature, etc.

and sunlight during the day, darkness during the night (or filtering out blue light with glasses / apps)

There are a lot more things to check/do besides "eating healthy, workout and not having apnea".

For example, the times when you eat matter (even if you eat healthy to begin with). The light levels matter - is your bedroom pitch dark? Noise levels matter. Temperatures matter a lot. Things like sleep position could play a role. Pets sleeping with you waking you up at night. And several other things besides.

Or you might have a condition that prevents you from sleeping well, even if it's not apnea. Might be stress. Or something that needs medical treatment, or a vitamin deficiency. You might also need supplements such as melatonin.

There are professionals (medical staff) that specialize in this stuff.

I'm happy with my Oura ring. Apparently its sleep analysis is on par with gold standard polysomnography (PSG) sleep lab test. https://ouraring.com/blog/new-sleep-staging-algorithm/

Independent studies show that it's among the best but not quite on par with medical grade devices. I suggest to have a look at this channel

https://www.youtube.com/watch?v=7m_oBiYk8Rk

Over 300 EUR and a subscription on top of that?!

How does that compare to the cost of a PSG test? Where I live a sleep lab test is pretty expensive so the Oura is an absolute bargain.

My personal theory / opinion in how I know I slept deeply is whether I had a dream that felt like a fairly long period. There was a point in my life where I literally had no dreams and slept like I was passing out. I realized that this was greatly affecting my long term memory. I went and made incremental steps in getting my sleep better and thankfully was able to start dreaming again, and even got to some points where my dreams became lucid, like being in inception. Anyways that's just my personal indicator on whether ive been sleeping deeply.

You're talking about REM sleep, not deep sleep. The article is talking about (I believe) the N3 stage of non-rem sleep, which is when the surrounding space around cerebral arteries expand and cerebral spinal fluid outside those arteries with flush out toxins

During this stage of sleep, neurons also fire in unison at a slow frequency on the order of 1Hz

I don't think that happens in REM sleep, which actually looks more like wakefullness than the N3 stage, but with certain aspects of cognition shut off (like the rational part, and the motor activation part)

Thanks for pointing this out! This definitely makes. me wonder if there's a way for us to tell whether we had adequete deep sleep without using an app or some external machine. My points about REM/Dreaming was that I could notice these things naturally without aid.

get a minimal activity monitor, without any "smart" functions. you can get ones that charge for 15m every 5 days now. (trying not to promote a particular product)

Use a smartwatch, or ask your doc to have a sleep study done

Is there something I can just keep next to my pillow? Or something like that?

I don't like wearing (and don't wear) watches or any kind of bands/rings/etc at all - they feel like a constant weight and then there's sweat, and pressure, and all that.

Before Apple bought them and discontinued the product last year, the Beddit sleep monitor was the best option for me. I’ve had mine for years and use it religiously. It’s a small sensor that goes on your mattress. You may be able to find one still for sale. It’s far more practical than wearing an Apple Watch to bed.

I have not personally tried it but I have heard good things about this: https://www.withings.com/dk/en/sleep-analyzer

The Eight mattress will measure sleep stages without you having to wear anything.

My experience with them is mixed. I’ve had a couple of theirs and they’ve been markedly inaccurate in some categories. For example, reporting I was not in bed when I actually was, or vice verse on other nights. As a result, the sleep hours were off. I still love it for the temperature control (set to manual, not their ‘auto’ algorithms), but if it can’t reliably report something as simple as whether I’m in or out of bed, then accurately judging sleep stages should be even less likely.

My oura ring has been far more helpful in identifying patterns leading to less quality sleep. For example , by highlighting elevated resting heart rates and asking me if I ate later that night, consumed spicy foods or alcohol, etc. That said, it’s not terribly durable. After a couple years it refused to accept firmware updates. Then a few weeks ago (after maybe 3 years of use) it stopped connecting to the app altogether and support tickets are going unanswered. This is greatly disappointing but I may spring for a replacement anyway because of how helpful it’s been in correlating lifestyle choices to sleep parameters (& quite consistently so).

Agreed on preferring my oura vs my eight bed for sleep tracking. Just found out that their web interface gives you a lot more analytic tools than the app, check that out if you have not already.

You can also get a less accurate but potentially still useful reading with only the phone in some apps. You'll get false negatives because it can only really measure movement but if it says your sleep is shit it probably is.

Color me septal…

Can you cite a medical study that evaluates the accuracy of your smart watch’s assessment of your quality of sleep?

Preferably one not sponsored by the manufacturer?

Your skepticism is justified. Oura publish their data, and to quote them:

>Oura ring was 96% accurate in detecting sleep compared to polysomnography, 48% accurate in detecting wakefulness, 65% agreement in detecting “light sleep,” 51% agreement in detecting “deep sleep” and 61% agreement in detecting REM

So it's basically wrong with sleep categorization about half the time.

oura ring hardware is horribly limited in terms of precision. let it not be a barrier to invest in a good wrist/arm/head band

"Performance of seven consumer sleep-tracking devices compared with polysomnography" Sleep. 2021 May PMC8120339

Seems to accurately analyze them. Conclusion "Device sleep stage assessments were inconsistent."

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8120339/

This research was funded by the Office of Naval Research, Code 34. (see also Disclosure Statement)

From the study: "Most devices (Fatigue Science Readiband, Fitbit Alta HR, EarlySense Live, ResMed S+, SleepScore Max) performed as well as or better than actigraphy on sleep/wake performance measures".

Great news.

The Garmin one is discontinued.

Also, is there a more "layman readable" version of this paper/report?

The Garmin watches have been replaced by newer versions with upgraded sensors. Garmin's tracking isn't perfect, but each version has made small improvements.

Yes, the study evaluates the accuracy, and the results are not good!

Many have already pointed that there have been studies to measure the accuracy of sleep analysis by wearables. They aren't that great. If you have a concern, you should definitely seek out a polysomnograph, and wearables should not be a replacement for that. However, you can't get a polysomnograph every night, and wearables are probably the best option for regular analysis. Just because they're not that great doesn't mean they're useless, and doesn't mean they won't improve.

they can be

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8120339/

Most devices (Fatigue Science Readiband, Fitbit Alta HR, EarlySense Live, ResMed S+, SleepScore Max) performed as well as or better than actigraphy on sleep/wake performance measures, while the Garmin devices performed worse.

Garmin publishes their accuracy data: https://www.garmin.com/en-US/blog/health/garmin-health-annou...

heh, I recently got a GW4. it's telling me that i get approx 22 mins a night of "Deep sleep" (average of 7 hours total asleep)

this sounds like i'm screwed...

There's the Dreem2 headband, but I don't think its available to consumers in the USA anymore. Maybe it is in other countries

Its basically a reusable EEG device connected to your phone, aka a sleep study every night, but maybe slightly lower quality

I'm sure with persistence and ingenuity there should be a way to get it

A sleep study is the gold standard, but sleep trackers like Oura reportedly getting pretty good and a lot cheaper and easier.

My son just had one, and they said they found nothing. But looking around, apparently most sleep studies in USA just look for sleep apnea. By default they don't actually look for anything else.

You need at least an EEG to have even a remote chance of getting good data. No smartwatches or rings will do the trick. Zeo was a good one, but discontinued 10 years back.

https://www.semanticscholar.org/paper/Consumer-Sleep-Trackin...

Try the Sleep Cycle app, at least on iOS. It’s an orange icon with an old analoge, bell alarm clock image.

My fitbit tracks it well, cheap thing I've had for years.

Reply on news.ycombinator.com