Apple Watch sleep apnea detection gets FDA approval

We have an app which delivers medical care, and Apple let us on but required proof of medical licenses, malpractice insurance, etc. Medical apps are allowed, but there's a higher bar for evidence.

Question: were the "medical reasons" specific or a generic "no because it's medical related" reason?

As in, did their refusal say "no because the way you're tracking if the user is on their back is not accurate enough?" or did they say "we don't allow people to write watch apps related to medical diagnoses / treatments / etc?"

Just say it's a game. Add some social widgets ("Get your non-back sleeping scores on the leaderboard!"). Maybe throw in a T&C forbidding using it for medical purposes.

Any chance that you'd be interested in open-sourcing that app?

A very long time ago, I got the impression that sleeping on your back is healthy, and successfully "trained" myself into the practice… now I'm trying desperately to untrain the habit 9_9

Is this a sleep apnea thing because I've always been told you want to sleep on your back?

Everyone's talking about the app but I'm impressed by the tennis ball idea.

Interesting and frustrating story! There's an Android app I've used that does that: https://play.google.com/store/apps/details?id=com.koalative....

Great minds think alike! Here's my attempt using an ESP-based smart watch that clips onto your clothes at night: https://github.com/a1studmuffin/lilygo-twatch-sleepdevice

How does it detect your sleeping position? Is it based on the watch position or is there another trick?

Can you reverse the function to buzz you when you're NOT on your back if you use a cpap? (It only works most effectively when on your back.) Also, what's wrong with being on your back?

I decided to clarify why sleeping on your back can sometimes be bad. I found out that it's undesirable for pregnant women to sleep on their back. It must be a very useful app for many

If you’re based in Europe you could launch it on the AltStore.

Makes sense. Apple could be held liable for any false or misleading health claims made about an app.

Can you get your app to detect if your user's weight suddenly doubles while they're lying down?

> you can have apnea without snoring (sometimes the other way but not as common)

Got a source for the "not as common" part?

From my personal anecdotal experience snoring without apnea is excessively common. Most of people whom I had a pleasure hearing snore would just go on in a steady rhythm and that's it.

Tell me more about this implant. Is that the electric one that basically tenses up your tongue as you breathe in?

How long until you got it? Did you have to 'fail' CPAP & the mandibular device?

Are the patent issues just refusing to pay now or are they more complicated than that?

It’s only turned off in the US right? Would they use the data in the rest of the world since SpO2 data is available there?

I don't think the apple watch is as accurate as they claim, and it seems to need many nights of data to make a sleep apnea decision, so in that respect it is vastly inferior to polysomnography. But hey, if it helps more people seek a proper diagnosis, then I'm all for it. Sleep apnea sucks. It seriously ruins lives.

Also: SpO2 is less useful here than you might think. Most at-home sensors are pretty unreliable, and it's possible to have sleep disordered breathing without having significant desaturation.

For real; this stupid patent thing is just one greedy company trying to get royalties from another while an entire built solution that could tangibly help regular people hone down on medical issues is completely disabled. I’m not usually the type to trash talk capitalism but this is stupid. I hope the company chasing down Apple for this goes bankrupt. It’s not like they’ve offered something I want to go buy that does the same thing.

They're tracking movements. Apple has a doc on how it works here[0]

>Apple Watch tracks movement with triaxial accelerometer signals, which capture coarse motion of the body as well as fine movements including motion associated with breathing. Apple developed an algorithm that uses the accelerometer time series data to classify Breathing Disturbances that occur during sleep tracking, which are temporary interruptions in the breathing pattern.

> A total of 1499 participants were enrolled, with 1448 completing the study. The sensitivity was 66.3% (95% CI: 62.2% to 70.3%), and the specificity was 98.5% (95% CI: 98.0% to 99.0%), demonstrating that the feature meets the design objectives to confidently identify sleep apnea while minimizing false positives.

>It’s important to note that the specificity was 100% (95% CI: 99.7 to 100%) for the normal category, indicating that all participants with a “positive” algorithm result had at least mild sleep apnea. Also, sensitivity was higher in the severe category at 89.1% (95% CI: 83.7% to 93.2%), indicating that the large majority of severe cases were identified.

0:https://www.apple.com/health/pdf/sleep-apnea/Sleep_Apnea_Not...

> Can Apple do something about eliminating all these bureaucratic barriers that hurt the health of so many?

I sure hope so, because my experiences with sleep apnea made it seemed very quack-ish. I keep on hearing ads on the radio by a company called ADVENT with a catchy jingle promising that surgery is the solution to snoring. At the time I had my study (2013) I did not have an in-home test and had a sleep test (which was an awful experience). I was prescribed a Phillips CPAP machine and was very good about adhering to it and keeping up with regular doctor visits. During the 3 years I had it, I lost a significant amount of weight and each time my doctor re-programmed my CPAP SD card so it was a lower and lower setting. I always had very little face time with the doctor, I wish I knew about Sleepyhead earlier as I could have done this myself. The practice I was referred to would always try to sell me stuff like new masks, tubes, etc. when everything I had was in good condition and frequently cleaned. I even used distilled water in the CPAP itself. Eventually I went on a long vacation and didn't bring my machine and been off it since. I'm glad I did, it turns out Phillips was negligent about foam breaking down in their CPAP machines and even the replacement devices have this issue as well. At this point I don't know who to trust, and with something with Apple's name on it would give to some more legitimacy.

> They claim it is so they know your air pressure numbers but it is all just regulatory capture

I disagree. In lab titration is not about regulatory capture. It's the best place to get a good ballpark start. The rest should be up to the patient, but informing the patient of that is where things go wrong, alas.

> CPAP machines can auto adjust your pressure and people like to tweak numbers manually to their comfort level anyways.

They cannot. See my rant about APAP in the comments here, or my previous APAP rants in my history. Current generation APAP is awful and it needs to die. PAP machines are awful at detecting events. This could probably be partially mitigate with better software. But there is a reason why a sleep study is very much the gold standard. 12+ channels of valuable information. Versus 1 channel on the PAP device.

The largest barrier by far is actually starting the process to get a diagnosis. It's often a family member that notices and raises the alarm.

Note that you can be diagnosed with sleep apnea based on the home study alone, you don't always have to go to a sleep clinic. If you are really sure you have sleep apnea (and you can be pretty certain with a home sleep study + symptoms) you can just go online and get a prescription for free by talking to a doctor at one of the companies that resell CPAP machines. At least in the US, other countries can be more bureaucratic.

> Can Apple do something about eliminating all these bureaucratic barriers that hurt the health of so many?

Integrate with the existing providers (= making deals with all the existing players and provide the adequate SDKs) and bring more accurate and reliable measurements than dedicated, calibrated devices.

That sure could be possible, but has Apple ever done anything like that ?

They couldn't come to an agreement wit Massimo, I wouldn't be holding my breath on them sealing deals with the incumbents in the field.

It’s always disappointing when high-end tech doesn’t support features that seem like they should be standard.

Reddit is pretty great for cpap related help and hints.

I just did 3 nights with mine and conclusion: it's all about the type of mask + settings.

First 2 nights was terrible (and painful on the face) due to wrong masks and wrong settings. 3rd night was best I've slept in a longgggg time.

I can't tell you which masks or settings (it will be completely different per person), but reddit is pretty great at providing hints.

My case: disaster with nasal mask and full face mask size L and high pressure settings (lots of central apnea). Almost perfect with 4 to 10 pressure setting and face mask size M (which should be too small for me according to the manufacturer).

I use a Resmed Airsense 11, soon getting a heated tube so I can crank up the moisture setting without getting condensation and noises in the tube. Will try more different full face masks as well.

I’ve been using a CPAP for 15 years. A few items I had to adjust when getting that suffocating feeling a few times:

1. When getting a new CPAP I had to reduce the pressure from 15 to 14 or 13 to mitigate the suffocating feeling. I still get my normal 0.5 events per hour while I’m use. I also got a different type of mouth and nose mask which may have contributed to needing to adjust the pressure.

2. I was trying to mitigate the colder air being distracting by enabling the humidifier and tube heater. With the humidifier level too high at 5 I got a suffocating feeling, reduced it down to 2 or 3 and the air was warmer while not feeling suffocating. 3. I also turned off the dynamic pressure on breathe out as the timing felt off and would cause a hitch in breathing, so I have just plain constant pressure.

The name of the game is tweaking and experimenting.

Followup sucks in most part of the world, unfortunately, which leads to awful compliance.

Best thing to do is install Oscar [1], take screenshots of your data, and post it to Apnea Board [2]. The folks there will likely be able to offer some solid advice.

Without knowing anything about you or your data, all of this speculation, but: APAP sucks, it is always too slow to respond, it won't know to increase the pressure until after you needed more pressure. This is especially true for wide ranges which are often the default (like 5-20cm). If you need 15cm, it will take forever to get there, because you'll start at 5, start to drift off, get an apnea, the machine will increase the pressure, you'll wake up ever so slightly from the apnea, the machine will decrease the pressure because now you're breathing fine, and the cycle will continue.

[1] https://www.sleepfiles.com/OSCAR/ [2] https://www.apneaboard.com/forums/

> as soon as I try to go to sleep I feel like I’m being smothered and my heart starts racing

You may need your pressure adjusted. You are awake so whatever pressure the machine is using is fine. Then you start falling asleep and that's when your apnea kicks in.

If you are using the standard autoset 4-20 setting it may not be enough. Ideally this would be done by a medical professional and yada yada - insert standard disclaimer here. I would suggest increasing the lower pressure (say, to 7) to see if it helps you. I was using the default for years and, while it did help me, apparently it wasn't enough. After increasing, it's been night and day.

It _could_ be an issue with your mask but given what you described, probably not.

Things that have helped me:

- nose mask rather than full face mask

- nasal spray if my nose feels congested

- upping the pressure. In theory a professional is supposed to set your pressure but I do tweak mine. If the setting is not visible it's probably in a hidden menu - Google for how to enable it for your machine.

- sleeping pills

I hope you find a solution. Sleep apnea is terrible.

Sounds like you aren't getting enough pressure. Also sounds like there could be some mechanical reasons like gerd, allergies or inflammation that cause your apnea, especially if you start fine.

I went through the same thing.

- #1 - switch to Nasal pillows.

- I had to learn to breathe through my nose

- I had to learn to sleep on my back with a pillow that cradled my head without raising it

- I had to make sure the level of distilled water in the humidifier was not too high.

- make sure the tubes are not higher than your nose (otherwise condensation water drips into your nose)

Did you try different types of masks? I found that I felt like that during the initial phases of the titration study (where they determine the right settings for you) but once they dialed in the pressure, that feeling went away and now I actually quite enjoy it, since it makes it easier to breathe through my nose. I use a nose pillow.

This is a common and difficult problem. Has your sleep doctor helped you with it? You might also find an online Sleep Apnea forum has ideas for you. I learned a lot from https://www.apneaboard.com/

I was prescribed a CPAP (well, APAP as you mention) and the first week was miserable. I think I got 3 hours of sleep total that week, and obviously felt like crap. I talked to my primary doctor and she's like "well then go back to the specialist". That's what I'm doing.

My plan is to get a prescription I can use to buy a bunch of different masks (I don't care about the cost), and get a few days of sleep aids. That should sort it out.

The few hours I did sleep with a CPAP were crazy. I forgot that I stopped having dreams. I had so many dreams while I could breathe at night.

Your pressure is too low and you are suffocating. While awake you are manually breathing deeper which is why you are okay.

Set the minimum pressure to something higher, like at least 10, possibly more. If it’s an auto machine then it will adapt to you anyway. At the minute it’s probably set to something like 4 which isn’t any air at all.

Have you tried playing around with the humidity? If I have it set too low, that definitely makes a big difference. There are also other treatments for sleep apnea that you could look into which might work better.

That feeling of being "smothered" might simply be that you need a higher pressure setting. The pressure that feels comfortable while awake might not be sufficient when your body is asleep.

Have you looked into "elastic mandibular device"? You can get a mouth guard that keeps your jaw from falling away from alignment. It was enough for me to avoid a machine entirely.

EASE and/or MMA

https://sleepapneasurgery.com/about-dr-kasey-li/

Do you have a slow ramp up of pressure? I have to start on low pressure when I first put the mask on each evening.

Your sleep doc/clinic may be able to help.

That sounds like a huge problem! Which masks have you tried? There are many different mask options that you may respond differently to

Honestly? Find a new doc. I went to one for sorta similar issues and their advice was basically "oh well that sucks" and then years later went to another who said "okay let's try this instead"

Now I'm on a bipap and tolerate it through the night.

> I can breathe through the CPAP exclusively for hours but then as soon as I try to go to sleep I feel like I’m being smothered and my heart starts racing

I felt the same way when I was prescribed a CPAP about 20 years ago. I'd be set to go, get panicky, take the mask off and not feel like trying again for longer periods of time until I gave up. Then about 2 years ago I decided to get a doctor again, talk about my sleep apnea, get a new set of sleep studies done, try to get a machine (it took forever!). I got prescribed a similar machine called an ASV.

For me, the thing that works is a breathing technique the technician at the sleep study told me, which was breath in manually, but then just relax to breathe out rather than forcefully breathing out. Currently I do that by breathing in manually, and then relaxing while counting to three as I let it out. I don't know if it's just a mental way to relax or what, but it works for me. Also, when I was taking the sleep study, I was determined to complete the test, so taking it off wasn't an option, even if I got anxious. So I did the breathing thing, and they were able to get enough data to proceed.

I go to sleep when I'm very tired; that might help. As far as I remember, I fall asleep very quickly. If I wake up in the morning, I can't really get back to sleep since I just notice the mask too much while not falling asleep; I guess I'm not tired enough any more at that point. 7 hours is good for me, but often I'll get 5-6 hours.

Edit: oh yeah, my number of sleep events (when you stop breathing) went from 120 to ~1 per hour. For comparison, >5 is mild, >15 is moderate, >30 is severe.

The announcement said that the feature was pending FDA approval.

Most likely the decision was taken internally, but sometimes to make it official requires some procedural paperwork. Apple probably got the green light that the approval was approved on the way.

Sleep apnea is not completely due to being overweight.

> I wonder how many years are left before Ozempic-like drugs make sleep apnea a rarity.

Obstructive sleep apnea is not caused by being obese or old, although those things don't help. In some people, losing enough weight may allow for sufficient airflow despite their airway restrictions.

I wouldn't count the Ozempic chickens before they roost. So far it looks like it's a costly lifelong injectable drug with a lot of reported uncomfortable side effects. That doesn't bode well for adherence at the population level.

It's free if you buy a brand new model. Ha.

You don't lose your HealthKit data when your AppleID gets banned. The data is still on your phone.

You can also backup your phone outside of iCloud.

I agree – there is precedent in the medical industry: in the US, providers that take Medicare and Medicaid require retention of patient records for at least 5 years for both portability and audit purposes. If that principle's sound (at least for portability), it may be a good argument to extend that to healthcare applications that reach a certain threshold of substantial content.

I understand the argument that it's quite annoying that so much of the hardware you own relies on Apple permitting you to keep using it, but on the other hand what are you even doing to make them ban you? Their terms and conditions make it fairly clear that the only thing that would cause this is if you upload illegal content to their platform or attempt to chargeback purchases.

That is definitely one cause of OSA as is micrognathia and other craniofacial abnormalities, but it remains that the most common cause is fat mass constricting the upper airway - even if that person has completely normal craniofacial development and proper posture.

I have been really curious to this. In IRS Publication 502, it states: "Medical care expenses must be primarily to alleviate or prevent a physical or mental disability or illness."

You could make the argument that the Apple Watch can be used to manage your sleep apnea condition. On the off chance you get audited by the IRS you might just be better off paying for it with non-HSA funds rather than dealing with that mess.

PSA: HSA expenses are based on the honor system, but obviously don't go to Best Buy and buy a PS5 on a HSA debit card. You are allowed to reimburse yourself for medical purchases. For example, if you have a large ticket medical expense you can use a credit card that provides a perk like air miles and then just transfer the funds from your HSA checking account to your regular checking account. Just remember to keep detailed records/receipts.

No, the watch is not a medical device intended to treat something. Just a general fitness device. Even the built-in EKG doesn't qualify it for HSA, and that is a lot closer to being medical than a notification pop-up saying you might want to ask your doctor to test you for sleep apnea.

I can see it now: it's December and my FSA still has a balance, must be time to buy 12 Apple Watches and sell them on eBay.

"SE" denotes a cheaper, less capable model that is not part of the main sequence of product generation numbering. There have been two generations of Watch SE and three generations of iPhone SE; they're not updated as frequently as the flagship product line. The Watch SE does not have the sensors for EKG or blood oxygen, which are probably necessary for the sleep apnea detection.

SE is a line orthogonal to their flagship models. Usually it means they took the enclosure from a previous watch and shoved some improved guts in it. Apple Watch SE is similar to Apple Watch Series 8 in terms of processing power but externally (screen, sensors) it’s like Apple Watch Series 4 or thereabouts.

Steve Jobs must be turning in his grave over this dogawful naming scheme.

They announced they're targeting 150 countries, so there will likely be CE-mark approval coming (this is the EU equivalent of the FDA).

I did one of the take home studies. It seems like there’s so many ways to get inconclusive results - they depend on a microphone that might end up underneath you, the probes could fall off, the battery could die, the “UI” is weird. I think continual testing is a very decent solve for some of those shortcomings and pushing people to the actual clinic for the real test

I find Garmin's sleep tracking on my forerunner 255 quite good. I have no idea how it compares to an Apple Watch though.

Withings sells a sleep mat that connects to a wall socket. Disclosure: I work there.

There's Whoop, but it's pricey.

what a coincidence that they've added nap detection just in time then: https://www.macrumors.com/2024/06/17/watchos-11-supports-aut...

For Apple. Samsung Galaxy has sleep apnea and blood pressure as well.

The product isn't even on people's hands, and FDA approval only means it's not actively harmful and has a fighting chance of being somewhat beneficial.

Is there anything objectively groundbreaking at this point ?

>Can't wait to get this in more devices.

Samsung received FDA approval back in February:

https://aasm.org/samsung-galaxy-watch-sleep-apnea-feature-re...

It works on a different monitoring method and requires 2 sleep cycles over 4 hours each in a 10 day window to start detecting.

>The feature, a software-only mobile medical app, uses smartwatch built-in sensors to monitor the user’s sleep for significant breathing disruptions associated with OSA. Users may track their sleep twice for more than four hours within a 10-day period to utilize the feature.

Yeah I'm generally a tech skeptic, but getting FDA approval for a real, common problem is a fantastic use for this kind of device. I did an at-home sleep study a few months ago, and while it wasn't an unpleasant experience, it did involve a long wait for an appointment, two trips to & from a specialty clinic, and awkwardly strapping on this device to sleep with overnight. That's a lot of barriers. Being able to do get a useful result for a very impactful health condition, just with an affordable piece of hardware that many people already have is a game changer.

Great job to the team at Apple for getting this together & getting real approval for it.

[dead]

Yes, always has, always will. Unlike products of many other manufacturers, it keeps what it learns on your devices, and doesn't send it home to big brother.

That's precisely why I got one.

GLP-1s significantly improve the symptoms of sleep apnea, with tirzepatide reducing incidence by 2/3. Probably just stick to Lilly and Novo for a broader exposure ;)

https://www.healthline.com/health-news/glp-1-drugs-like-zepb...

Tim loves when you bounce on it.

Apple has their standards high (though I agree they sometimes can't meet it themselves).

They wouldn't want their brand to be associated with some half baked feature that either runs very slow, or draws too much power/heats up, or unreliable.

So even if another version is technically capable of that feature, if it runs slowly/inefficient, they'd not enable it to protect their brand value which makes sense: as a user if Apple announces a feature, I know it's more "done" compared to the competition like say Samsung/Android etc. and I trust that it's not half baked (which sometimes fails to meet the expectations, but still MUCH better than the competition)

The business reasons are just a plus _on top of_ these reasons.

If it was purely a business one, they would not have put it on the Apple Watch 9 which is no longer available. Putting it on the 9 doesn't do anything for sales.

There must have been some hardware difference.

The Series 9 has a better low-power background accelerometer than the Series 8, which is likely the reason.

What makes you so sure it’s not technical?

The Apple Watch 9 and Ultra 2 have a neural engine (GPU), which may be helpful if the sleep apnea detection algorithm uses a neural network. Those models also come with a third-generation heart rate sensor, although that part may not be required if the algorithm truly only relies on accelerometer data.

atleast due to processing/battery reasons.

>> detecting sleep apnea in flight crews to aid aeromedical certification.

24/7 medical monitoring via a wristwatch? Monitoring that may cost you your career? I can think of nothing pilots would resist more.

I’m on Garmin watches as well but this line of thinking is getting old. Partner has an Apple Watch and they charge it while in the shower/getting dressed and occasionally 10-15 minutes in the evening to make sure it gets through the night. Charger is in a convenient location and it’s dead easy to drop the watch on it. If you’re running ultras, then by all means stay on Garmin. But honestly for the day to day I’ve rarely seen it be an issue.

ETA: there’s plenty of valid reasons to prefer other watches (ANT+ compatibility being one of mine) but honestly the battery life thing is really not that bad for most users.

> With 36 hours being the absolute best-case scenario for battery life, it would seem your options are to either have two watches, or be perfectly fine only wearing your watch every other day?

That only makes sense if you think it takes 12+ hours to charge. 20 minutes on the charger in the morning while I'm taking a shower and getting dressed and it's fine for the next 24 hours. Oh no, I can't wear it in the shower. I wasn't going to anyways.

At the risk of sounding like a shill, you can get by on the Apple Watch ecosystem’s battery life. The charge time is really quick, so throwing it on the charger while showering each day is sufficient.

Your point that a 36 hour battery life on a watch is pretty inconvenient stands. I have an Ultra 2 almost entirely because it can be described as an Apple Watch with a slightly more reasonable battery life. That’s a price I don’t expect most folks to pay.

How is 36 hours different from 3600 hours, in your scenario? Either way, you need either need two watches, or not wear it sometimes.

The answer is, of course, that sometimes you don't wear it, but charge it. Like a few minutes in the morning and/or a few minutes in the evening while you shower or eat or whatever. Done.

> in practice this feature seems almost useless to me

Maybe to you, assuming you don't have sleep apnea.

Far too many people have it and it goes undiagnosed. Until that person goes on to have a stroke, heart attack or sleeps at the wheel.

It's a silent killer and I suspect we have an epidemic.

My Series 6 Apple Watch lasts from about 8am to midnight every day. If I wanted to sleep while wearing it I'd just take it off around 11pm and it'd be done charging almost every time by midnight, sometimes even sooner.

I used to religiously wear Pebble watches because they lasted forever between charges. But in all honesty I just don't find the one day battery life of AW to be that bad considering how quickly they charge. IMO they do have all-day battery.

Don't get me wrong, a better battery would be great. But the only time I ever run into issues with my apple watch dying is when I am traveling and my routine is messed up anyways.

Otherwise I am just in the habit of waking up, sticking it on the charger, getting ready for my day and putting it back on. Doesn't take long to get to a full charge.

I get access to all of the features and just loose it for 30 minutes to an hour.

You're not supposed to wear it while you shower (soap is generally bad for any waterproof seals...). Just slap it on the charger while you shower. Or slap it on the charger when you wake up, during your usual morning routine. It's a bit of a change of paradigm from the 'charge while you sleep' that we do with phones, but you get used to it.

When I come home from my morning walk, I'm usually a bit hot and sweaty, and it feels good to sit down and take the watch off and throw it on the charger. 10-30 minutes is enough for it to charge; I wish it would ding when full.

If it lasted 36 hours that would be enough for me to sleep with it.

How does this relate to sleep apnea detection?

I agree the battery life is shit compared to Garmin, but then you need to compare features and usability too. And as a long-time Garmin owner I am 100% certain that is not a fair comparison either.

Everyone sets their own routine. I charge it every morning when I start working. Takes about an hour.

It is manageable, but really sucks.

It is a bummer not to have Garmin like battery life, but it takes like 15 minutes to charge so you just sit it on the charger while Itake a shower or something or while im working and other than that I can wear it all the time. Battery life is mostly an issue if trying to do a long hike, need a garmin for that.

It takes like 20 minutes to charge the watch. Just don’t wear it for that time

I charge it when I shower.

I charge when I’m in the shower each day and this works fine for me.