Apple Watch’s Blood Oxygen Sensor Might Not Be All It’s Cracked Up To Be

Blood oxygen sensor

When the Apple Watch Series 6 was announced, many were excited at the addition of a blood oxygen sensor. Monitoring how much oxygen is packed into your red blood cells can be great for determining your overall lung health. Is the blood oxygen sensor really useful, though? Several reports are in, and the news so far isn’t exactly encouraging.

How a Blood Oxygen Sensor Might Help You

If you suffer from any chronic lung issue, like asthma, COPD, or bronchitis, it’s important to keep tabs on how oxygenated your blood is. This metric, reported as a percentage, reflects how well your lungs are transferring oxygen to your bloodstream. This is important to giving your body the fuel it needs to keep going. Your brain also depends on oxygen to keep you in tip-top mental shape. Most of the time, pulse oximeters measure your blood oxygen levels from your fingertip, shining a red light through your skin and analyzing the reflection. With Apple Watch Series 6, the sensor shines its red lights through your wrist.

pulse oximeter blood oxygen sensor
A pulse oximeter, a blood oxygen sensor, usually clips onto your fingertip to do its measurements

You can also use a blood oxygen sensor to help gauge how your sleep apnea is affecting you at night. Since breathing is interrupted by sleep apnea, you’d see your blood oxygen levels drop, sometimes dramatically, during a bad night. If you use a CPAP, it’s definitely a good idea to also keep track of your blood oxygen. That way, you’ll get an important clue when your mask might be ill-fitting or coming off.

Enter the Apple Watch Series 6

With the new Apple Watch, a blood oxygen sensor is right on your wrist at all times. The Watch checks your blood oxygen periodically, and records the data in the Health app. One thing that’s important to note, though, is the fine print Apple provides about this feature:

Blood Oxygen app measurements are not intended for medical use, including self-diagnosis or consultation with a doctor, and are only designed for general fitness and wellness purposes.

This means that you might not be able to really depend on the data to be very accurate. In fact, early tests are suggesting that it’s quite imprecise.

One Reporter’s Blood Oxygen Sensor Comparisons

Geoffrey Fowler, a tech columnist at The Washington Post, gave the Apple Watch 6’s blood oxygen sensor a test run. He found that the sensor occasionally says his “lungs and heart are the picture of health, pumping blood that’s 100 percent saturated with oxygen.” Other times, though, the sensor suggests he might have emphysema, his blood oxygen is so low.

Fowler purchased a $60 finger pulse oximeter, one that’s FDA-approved and reports an error rate of plus or minus two percentage points. He compared the results from this inexpensive medical device with those of the pricy Apple Watch Series 6. He found that the results sometimes matched, but most often were two to three percentage points different. At times, though, the Apple Watch differed from the finger pulse oximeter by 7 percent.

The Apple Watch Is Not a Medical Device

When Apple introduced the ECG function into its watch in 2018, it worked with researchers to publish accuracy studies. Cupertino also got FDA clearance (not the same as “approval”) for its app. Even so, the company was very forthcoming with its warning not to rely upon the irregular-rhythm notification if you’d previously been diagnosed with atrial fibrillation.

We’re not yet seeing the same diligence with the blood oxygen sensor in the Apple Watch Series 6. Apple is clear that the device isn’t meant to be used for diagnostic purposes, but it bears repeating. Do not purchase an Apple Watch Series 6 with the idea that it’s going to help you catch medical issues. You won’t be able to rely on its data to help identify COVID-19, sleep apnea, emphysema, pulmonary embolisms, or any other health problems.

3 thoughts on “Apple Watch’s Blood Oxygen Sensor Might Not Be All It’s Cracked Up To Be

  • Jeff:
    Just a quick follow up. Having played with my Series 6 over the weekend, and having done at least six workouts with it on, I have an early impression of the pulse ox. 
    First, having worked with a variety of pulse ox machines professionally over the years, this is the best execution of a consumer pulse ox using reflective technology that, in so small and portable a device, that I’ve seen. That Apple managed to miniaturise this device into a housing that fits on the wrist and that shares that space with so much additional functional tech is simply astonishing. Understanding how this works, what it is doing, and with such consistent performance is what will make a professional user take note. My readings have been consistent, all being within the normal range and all within an assumed standard error for this type of device (96% – 100%).
    Second, while John K found may have found his slow, my watch invariably gives me a reading within 15 seconds. Is it possible to give a faster reading? Quite likely, but the calculation it is doing is not simple, and is a consensus output. Particularly for abnormal readings, I’d prefer that it take the full 15 seconds than to come back sooner with a less accurate rendering on a less precise reading. As this technology gets further refined, the speed will likely increase, but 15 seconds for a reflective pulse ox is not bad at all. Indeed, the transmissive pulse ox machines we use in hospitals and clinics can, and often do, take longer.
    Finally, the fit is important. I find that the best band for consistently reliable readings has been one of my Nike sport bands with the Velcro clasp. It is not only snug, but does not change with moisture, or as either the material warms, is put through paces or the fit otherwise loosens (eg the Milanese band with the magnetic clasp). Assuming that the algorithms are adaptive, one might first start with a well fitting band for the first few readings before trying a looser fit. 
    Curious to see more data and feedback as the community try this out. 

  • Jeff:
    The discussion around pulse oximetry can be complex, speaking as one who has had to introduce the technology into new settings and train clinical staff on its use. Briefly, there are two different technological approaches, transmissive pulse ox which passes two different wavelengths of light through a thin body part (fingertip, toe, earlobe are favourites) to a sensor on the opposite receiving side, and calculates arterial blood oxygen saturation based on differential wavelength absorbance. 
    The other is reflective pulse ox, which is the one used by Apple, and which is both less common and often less reliable, although several innovations have improved the latter. Having just received my Apple Watch series 6 today, I am only just beginning to play with it, so cannot comment yet on its performance. 
    A good discussion can be found here
    That said, two things are worth noting. 
    First, even within the published limits of accuracy and variability of the AW pulse ox performance, it should be a useful aid to anyone in serious danger of hypoxaemia. Normal blood oxygen at sea level (an important qualifier) is ≥95%, but can be and often is lower at higher elevations, whose clinical staff are taught to use site-specific cutoffs for ‘normal’ (eg Nairobi). Hypoxaemia is usually graded into Mild, Moderate and Severe, with severe being <85% at sea level in a normal individual without a cardiovascular or obstructive pulmonary condition (eg COPD). For COVID-19, we are seeing people show up in the emergency department with O2 sats in the 70’s, 60’s and even 50’s – people who shouldn’t still be vertical, but they’re walking and talking. And although they report shortness of breath, they are unaware of being that hypoxic. So long as the sensor on the Apple Watch can get a reading on an individual, if it reports, consistently, something below 80-odd percent, that’s all one needs, combined with symptomatology, to seek professional help. An advantage of the reflective sensor is that it can read parts of the body that are not as susceptible to minor decreases in blood flow, like the fingertips, so should be able to get a reading under most circumstances. Again, we need more data on real world use case. 
    The second thing is that, apart from detecting abnormal state, another thing the watch can do for fitness enthusiasts is measure VO2 max, the maximum amount of oxygen your muscles can extract during exercise, which is a measure of cardiovascular fitness, and a parameter that competitive athletes monitor. This is a feature I look forward to testing out over the coming days with some of my intense workouts. The point being, that the pulse ox on the Apple Watch is not simply for those at risk for low blood oxygen states, but for those monitoring healthy states as well. 
    No pun intended, but how well the Series 6 Apple Watch performs both of these tasks, only time will tell. 

  • Also, the cheap pulse oximeters take a few seconds to get a good reading, whereas the Apple Watch takes forever. It’s basically a crap pulse oximeter on the Apple Watch.

    Where there is a cool use is while you’re asleep, the Apple Watch could take several readings a night and you can have a track record of readings. Over time, there might be analytic insights that could be made.

    However, the Apple Watch 5 battery was atrocious and it cant get through the night. You basically have to charge your Apple Watch twice a day if you want to sleep with it on, and that is just horrendous. The Apple Watch 6 seems not substantially better with battery life, probably because any gains it made in battery capacity and efficiency with renewed chips was destroyed by new pulse oximeter and real time complications added to the watch.

    The Apple Watch 4 was so far the best watch battery life time, but I hear watchOS7 ruined that. It’s not clear how the Apple Watch SE compares to the 4/5/6, but it stands a chance to be the best battery life out there since it doesn’t have the always on screen, the pulse oximeter etc.

    TLDR the pulse oximeter is junk and any small advantage it could provide is destroyed by the apple’ watches jank battery life. Apple needs to significantly increase the Apple Watch battery life to over 2 days of real use including sleep.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.