Apple Watch may be able to detect coronavirus infection days before tests can

Posted:
in General Discussion
The Apple Watch may be able to detect if a wearer has coronavirus days before they are diagnosed or symptoms appear, a new body of research shows.

Credit: Andrew O'Hara, AppleInsider
Credit: Andrew O'Hara, AppleInsider


In some cases, wearable devices like Apple Watch or Fitbit devices can predict a COVID-19 infection even before a user becomes symptomatic or the virus is detectable by standard tests, according to studies from a number of leading medical institutions (via CBS News).

Medical researchers at Mount Sinai Health System in New York, for example, found that the Apple Watch can detect subtle changes in a user's heart rhythm up to seven days before an infection can be detected through testing.

The Mount Sinai study analyzed the variation in time between heartbeats, a metric known as heart rate variability. Researchers say it's a good measure of how a person's immune system is working.

"We already knew that heart rate variability markers change as inflammation develops in the body, and Covid is an incredibly inflammatory event. It allows us to predict that people are infected before they know it," said Rob Hirten, the author of the study and an assistant professor of medicine at the Icahn School of Medicine at Mount Sinai.

When it comes to COVID-19, individuals infected with the disease experience lower heart rate variability compared to those who tested negative. The study followed 300 health workers at Mount Sinai who wore Apple Watches for five months.

Notably, Apple highlighted the Mount Sinai study at its Apple Watch- and iPad-focused "Time Flies" event on Sept. 15, 2020.

Another study from Stanford University in California looked at a variety of activity and fitness trackers from Apple, Fitbit, Garmin, and other manufacturers.

That research, published in Nature Biomedical Engineering, found that 81% of those who tested positive for coronavirus experienced changes in their resting heart rates.

Like the heart rate variability metric, researchers found that the trackers could detect an infection up to nine and a half days before symptoms began.

According to Stanford University professor Michael Snyder, one of the main advantages of wearable trackers is that users wear them constantly throughout the day. That, Snyder said, could help do away with some of the disadvantages of standard coronavirus testing.

"The problem [with testing] is you can't do it on people all the time, whereas these devices measure you 24/7. The smartwatch gives you back the data right away, in real time, whereas if you're lucky you'll get your test back in a few days," Snyder told CBS.

The team also developed an alarm system that alerted users if their heart rate was elevated for a prolonged period of time. That could alert people to cancel going out or meeting others in-person, since they could be infectious. All of this research could help medical professionals and the public to stamp down coronavirus infections, since the majority of cases are spread by asymptomatic people.

"Right now, we rely on people saying they're sick and not feeling well, but wearing an Apple Watch doesn't require any active user input and can identify people who might be asymptomatic. It's a way to better control infectious diseases," Hirten said.

Comments

  • Reply 1 of 15
    And how do users selftest? What do we look for in Cardiogram? And how do we determine our baseline?
    bob nycpulseimagesbageljoey
  • Reply 2 of 15
    larryjwlarryjw Posts: 1,031member
    This is promising. My guess is many apps can be updated to perform this analysis. The question is what permissions would be necessary from the regulatory agencies before it can be rolled out, if any?
    rcfabageljoey
  • Reply 3 of 15
    rcfarcfa Posts: 1,124member
    There’s usually a long stretch of time between initial study results, verification of these results, creating a product, and getting FDA approval.

    Also, if these are indicators for inflammation, disambiguation is going to be a big thing.

    It’s one thing to look at COVID-19 patients and going back in time looking for indicators; it’s a different thing concluding based on indicators that someone’s got COVID and not some other infections, allergies, etc. that result in an inflammatory response.

    Best one could do relatively quickly is warn people that they have indicators typical for inflammation and tell them it might be worth reducing contacts until they can get tested; but likely even that would require disclaimers the size of a small phone book...
    bageljoeylongpath
  • Reply 4 of 15
    Benderclb said:
    And how do users selftest? What do we look for in Cardiogram? And how do we determine our baseline?
    Heart Rate Variability is recorded in the Breath app. I think the gist is that focused breathing decreases the heart rate, so you get a better sample of beat variations. 

    HRV is considered one of the best indicators of overall physical fitness. Like, an athlete can go from zero to sixty (figuratively) at a snap; that requires a heart that can kick into gear (and cool off) fast. 

    You can see your HRV in the iPhone Health app under "heart." Average HRV is very age dependent, a teenager will have an HRV approaching 100, middle age folk in the fifties.

    As for how this could be productized… Apple probably couldn't market any feature as "early covid detection," but they could perhaps put in a notification that reads "abnormal HRV, talk to your doctor" without going through too many FDA hoops. I think the potential for false positives might be pretty high, though.
    edited January 2021 qwerty52fred1longpathgeekmeemacgui
  • Reply 5 of 15
    Huge study already doing something similar w Garmin watch, Oura ring & daily symptom self-reporting. Hate the Garmin interface - would much prefer Apple Watch. 
  • Reply 6 of 15
    jgbO5 said:
    Huge study already doing something similar w Garmin watch, Oura ring & daily symptom self-reporting. Hate the Garmin interface - would much prefer Apple Watch. 
    Then why do you have a Garmin watch instead of an Apple Watch?
    longpath
  • Reply 7 of 15
    JFC_PAJFC_PA Posts: 932member
    So a simple prompt to seek further professional advice/testing? I’m not sure that requires any regulatory scrutiny as it’s not providing a diagnostic service. 
    PetrolDavelongpathcommand_f
  • Reply 8 of 15
    Benderclb said:
    And how do users selftest? What do we look for in Cardiogram? And how do we determine our baseline?
    Go to the Health app. Scroll down to Hear Rate Variability. Set the view to week. Is there a significant drop?
    Detnator
  • Reply 9 of 15
    rcfa said:
    There’s usually a long stretch of time between initial study results, verification of these results, creating a product, and getting FDA approval.

    Also, if these are indicators for inflammation, disambiguation is going to be a big thing.

    It’s one thing to look at COVID-19 patients and going back in time looking for indicators; it’s a different thing concluding based on indicators that someone’s got COVID and not some other infections, allergies, etc. that result in an inflammatory response.

    Best one could do relatively quickly is warn people that they have indicators typical for inflammation and tell them it might be worth reducing contacts until they can get tested; but likely even that would require disclaimers the size of a small phone book...
    An alert for inflammatory response would still be useful. As someone who, despite being an age group triathlete, has a low HRV, this is certainly something I’d want to know ASAP about.
  • Reply 10 of 15
    Interesting idea, but fairly useless as a predictor of Covid infection.  HRV is highly variable within the population- highly correlated with age, but more importantly is highly variable within each individual.  Mine varies from the teens to greater than 50 on any given night and is affected by an infinite number of influences - rest, exercise, sleep, alcohol and who knows what else.  Could probably add corona virus to the list, but would also need to add every other known or unknown infection......
    Respiratory rate (measured while in deep sleep) has been shown to be a more sensitive metric - it is remarkably consistent, so any significant rise (e.g. from 15 to 17+) would suggest that something is stressing your body greatly.
    GeorgeBMac
  • Reply 11 of 15
    GeorgeBMacGeorgeBMac Posts: 11,421member
    cabodoc said:
    Interesting idea, but fairly useless as a predictor of Covid infection.  HRV is highly variable within the population- highly correlated with age, but more importantly is highly variable within each individual.  Mine varies from the teens to greater than 50 on any given night and is affected by an infinite number of influences - rest, exercise, sleep, alcohol and who knows what else.  Could probably add corona virus to the list, but would also need to add every other known or unknown infection......
    Respiratory rate (measured while in deep sleep) has been shown to be a more sensitive metric - it is remarkably consistent, so any significant rise (e.g. from 15 to 17+) would suggest that something is stressing your body greatly.

    Thank you -- that saved me from having to type the same thing.   Plus, you said it better.
    The analogy might be:  "81% of people who drown ate ice cream in the weeks before drowning".   While that is obviously true, it does not fly the other way: because you ate ice cream doesn't mean you're going to drown.   Likewise, because your HRV dropped doesn't mean you are infected with the virus.  

    I would only add that I have heard some fitness experts claim that Apple's HRV cannot be used because they measure it at random intervals and HRV, like Blood Pressure and temperature, tend to vary throughout the day.  So it needs to be measured at the same time and in the same way each day.  Some recommend checking before you get out of bed in the morning.
  • Reply 12 of 15
    cabodoc said:
    Interesting idea, but fairly useless as a predictor of Covid infection.  HRV is highly variable within the population- highly correlated with age, but more importantly is highly variable within each individual.  Mine varies from the teens to greater than 50 on any given night and is affected by an infinite number of influences - rest, exercise, sleep, alcohol and who knows what else.  Could probably add corona virus to the list, but would also need to add every other known or unknown infection......
    Respiratory rate (measured while in deep sleep) has been shown to be a more sensitive metric - it is remarkably consistent, so any significant rise (e.g. from 15 to 17+) would suggest that something is stressing your body greatly.

    Thank you -- that saved me from having to type the same thing.   Plus, you said it better.
    The analogy might be:  "81% of people who drown ate ice cream in the weeks before drowning".   While that is obviously true, it does not fly the other way: because you ate ice cream doesn't mean you're going to drown.   Likewise, because your HRV dropped doesn't mean you are infected with the virus.  

    I would only add that I have heard some fitness experts claim that Apple's HRV cannot be used because they measure it at random intervals and HRV, like Blood Pressure and temperature, tend to vary throughout the day.  So it needs to be measured at the same time and in the same way each day.  Some recommend checking before you get out of bed in the morning.
    If you’re having a major inflammatory event, then your daily averages HRV should be impacted because all of the readings through the day should be impacted. Simply offering the option to be notified of a drop in HRV that impacts all readings within a given time frame, perhaps 18 hours, in order to make sure extra sleep doesn’t cause a false positive, shouldn’t require any regulatory requirements, as no diagnostic claims are made. Indeed, Apple could make similar notification options available for a range of Health metrics, based on whatever the end user wants to track.
    GeorgeBMac
  • Reply 13 of 15
    macguimacgui Posts: 2,350member
    I have no idea what this means but maybe six times over the last 1.5 months or so, my resting HR as dropped below 50bpm for over 10 minutes, according to my Watch. My resting HR is typically 55bpm give or take.
  • Reply 14 of 15
    GeorgeBMacGeorgeBMac Posts: 11,421member
    macgui said:
    I have no idea what this means but maybe six times over the last 1.5 months or so, my resting HR as dropped below 50bpm for over 10 minutes, according to my Watch. My resting HR is typically 55bpm give or take.

    It might mean that you're getting in better shape -- cardiorespiratory fitness has a lot to do with it. 
    But there also seems to be a natural variation as well:  according to Apple Health, mine varied from 45 to 55 over the past month.

    Another big variable is when Apple decides to record it:   There is typically an almost 10 point spread between my resting heart rate while I sleep versus while I'm awake (Which is possibly why Apple Health shows mine varying be so much).

    In any case, yours in the low 50's is better than most.   Congratulations!
  • Reply 15 of 15
    longpath said:
    Benderclb said:
    And how do users selftest? What do we look for in Cardiogram? And how do we determine our baseline?
    Go to the Health app. Scroll down to Hear Rate Variability. Set the view to week. Is there a significant drop?
    What constitutes a "significant" drop?  In the past week, mine has varied from a low of 8ms to a high of 69ms.  That 8 (which happened twice consecutively, about 2 hours apart on the day in question, where the other measurements were 19-41)  corresponds to a day when I wasn't feeling well (but no covid, I had a test because of a fever), but is that correlation even significant?

    cornchipGeorgeBMac
Sign In or Register to comment.