[brandonb]

Hey HN! I'm not the OP, but I am the burrito-eating, Game of Thrones watching researcher mentioned in the article.

This story focuses on a fun application of the Apple Watch's heart rate sensor, but there's a lot we can learn about the human body through this type of data.

For example, on the super burrito causing my heart rate to spike to 123bpm, there's very sparse medical literature on what this means. First off, what type of foods cause heart rate spikes? Does it relate to macronutrients (carbohydrates, protein, fats)? Ion balance (e.g., sodium, potassium, calcium, and their effects on the cardiac action potential)? Something else? Is this behavior universal across people? Are there clusters of people, e.g., do people with prediabetes respond to a different set of foods? Most importantly, what does this mean for health? Am I more likely to get diabetes? To have a heart attack? To die?

And that's not even mentioning more acute conditions like cardiac arrhythmias:

  http://qz.com/472522/how-the-apple-watch-may-have-saved-this-mans-life/

We've seen the same thing--a person who discovered they have atrial flutter through their Android watch.

Finally, if you have an Apple Watch and want to play around with this data, we just started a public beta of Cardiogram on the App Store:

  http://cardiogr.am

If anybody has questions about Apple Watch + healthcare, or anything else, happy to chat on this thread.

[rrss1122]

I thought I read somewhere that the heart rate goes up after eating due to the stomach requiring more blood to digest. I also read that foods that cause blood sugar to rise quickly will also cause the heart rate to rise. Does this sound like it makes sense?

Also, would a super burrito contain enough minerals to affect cardiac action potential?

[brandonb]

I think blood sugar => insulin => adrenaline => raised heart rate is as good a hypothesis as any!

For cardiac action potential, I'd like to find that out. The avocado (potassium) would be my leading candidate, but cheese (calcium) and the meat (marinated, so sodium) could also have an effect.

[collias]

Do you do any accounting or normalizing for clearly erroneous heart rate readings from the watch?

Just anecdotally, the heart rate monitor seems to be wildly inaccurate at times. Over the course of a workout at the gym, my heart rate jumped to 170bpm while benching (probably close to correct), and dropped to 45bpm while deadlifting (definitely incorrect). My resting heart rate is ~55bpm.

After a few days of seeing my heart rate have a delta of ~100bpm in the matter of seconds, I've given up on the heart rate monitor being good for much.

Maybe I'm an outlier and it's accurate for other people, but for me, it was very disappointing.

[brandonb]

We'd love to do a validation study to get a definitive answer to this question.

Anecdotally, the Apple Watch is within about 5bpm of my AliveCor (an FDA-approved medical device), even while running. And it's been similarly accurate on most (but not all) patients in arrhythmias.

That said, we know accuracy will vary across skin tone, body mass index, pulse pressure, sweat, and we need to understand that more precisely before the AW can start to have real medical applications.

[rrss1122]

You can try getting a tighter fit and see if that helps.

I've noticed with any consumer heart rate monitor, if the fit is not tight and the monitor moves even a little bit, the heart rate the monitor reports will be off. I'm assuming the same will apply to the Apple watch.

[brandonb]

That's good advice. When I had a Moto 360, I also sometimes did the opposite: loosen the strap one notch, and slide it up my arm slightly till it was taught again. Your forearm is a slightly easier place to measure pulse than on a bony wrist.

[collias]

Interesting thought. I'm going to try to place it further up my arm and see if it helps.

Thanks!

[kevhsu]

I haven't worked with the Apple Watch, but for the Microsoft Band, the SDK will return a confidence level for the heart rate. I think both devices usually display the number regardless of confidence level to the user, sometimes with a signification that it is still acquiring your heart rate.

[hollerith]

I'd like to teach myself to be less tense while programming, and by "tense" I refer to my sympathetic nervous system. What's the best sufficiently cheap way for me to use gadgets, software and sensors to measure my tension minute by minute? An Apple watch would qualify as sufficiently cheap for the purposes of this question.

Thanks in advance.

[Sven7]

I don't get this. Sure the watch makes it easy to collect your heart rate date, but where is the data on whether they are eating a burrito or watching game of thrones going to come from?

[brandonb]

The app lets you add a title to each cardiogram, as well as optional annotations for particular moments.

[tbirdz]

Any thoughts on privacy concerns with collecting this data? Do you think it should have the same requirements as regular patient medical data (eg HIPAA), or should it be treated differently?

[brandonb]

I think the intent of HIPAA is good, but in practice, it's a big barrier to a lot of medical advancements. It's less the letter of the law itself, to be honest, and more the penumbra of fear that develops around any complex law. We saw the same thing on healthcare.gov.

There was a great article about this a while ago, but I can't seem to find it now.

Overall, I'm pretty happy with the approach the FDA and other regulatory bodies are taking to digital health -- mostly hands-off, but stepping in to stop pure snake oil apps that really have the potential for harm. That seems like a good balance.

[dsjoerg]

Thanks for using the wonderfully evocative phrase "Penumbra of Fear". Googling the phrase led me to read "Terms of Service" (http://projects.aljazeera.com/2014/terms-of-service), a great read.

And yes, as someone working in the highly regulated finance industry, I know from personal experience that complex laws directly create a penumbra of fear you can't ignore.

[wasd]

Would you consider adding a fitbit integration as well? I think many more people own a fitbit charge hr (which has a heart rate monitor).

[brandonb]

I'd love to show Fitbit data! If they reverse their decision not to integrate with HealthKit, then I think we'll be able to integrate without too much trouble: http://www.cnet.com/news/dont-expect-fitbit-to-integrate-wit...