Why you should use Apple Watch to track and analyze sleep

Sleep tracking isn’t new. But what’s unique about using Apple Watch?

General Introduction

Sleep is the single most important activity for a normal human being, measured by time spent. Having a good night of sleep usually means a jump start of the day. On the flip side, bad sleep (or no sleep at all) will lead to serious implication – both physically and mentally.

In this blog post, I am happy to share my story of using the Apple Watch + 3rd Party App to better measure the sleep.  Hopefully, it will inspire you to pick up a few things here and there and take action to make your own sleep better.

How everything started for me

First and most importantly, making the most out of my Apple Watch is fun.  I just can’t wait to milk everything out of my $500 investment 18-months ago.

Also, I had sleep problems in the past. Anxiety and parenting are the two big drivers.

Lastly, professional curiosity also pushes me to find more meaningful data to analyze.

Why should you do it (benefit)

Similar to my previous post on tracking Tennis and Swim, having sleep metrics available will allow us to see our sleep through a new lens.  “How you feel” after waking up is still an important signal, but no longer the only one.

What’s unique (compared to tracking tennis and swim) is, sleep tracking doesn’t require any extra button clicking. Just wearing the watch with you, making sure it has at least 20% battery, and the 3rd Party App (AutoSleep is my choice) will do the work in the background.

The biggest benefit is allowing you to ACTIVELY manage your own sleep:  Once the habit is established – checking the sleep pattern the 1st thing in the morning. In addition, it can provide:

  • Unbiased and more accurate protocol of sleep quality:  Is qualitative sleep data (asleep time, deep sleep time, quality score, average heart rate) matching quantitative feeling like “Am I feeling refreshed”?  If the answer is no, usually the quantitative data gives me a more unbiased indication.  For example – when I wake up from a bad dream feeling awful but the sleep data tells me the sleep quality is actually okay, my daytime energy is generally fine.
  • Give self an opportunity to reflect: Why today’s sleep is better or worse than yesterday?  This usually leads to an honest conversation with self on “Is there anything I do before the sleep that I can relate to?”   Over the time I have learned to avoid doing any work or exercise 2 hour before sleep.  Ideally, screen time should be avoided as well – it is just too hard to make it a reality.

How you analyze the sleep(case study)

There are four metrics available from the home screen of the AutoSleep app.

  • Sleep: the total hours of actually falling asleep recorded by the device.
  • Quality: a calculated metric based on total sleep time, how restless and the nocturnal dip in heart rate and determines how much quality sleep you had
  • Deep sleep: the number of hours where you met a sufficient nocturnal dip in heart rate combined with minimal movement.
  • BPM: average heart rate during the sleep
Home Screen – iPhone View
Home Screen – Watch view

After understanding the metric definition, we can easily export the data into a CSV file, dump it into Jupiter Notebook (Python) or R Studio(R) to wrangle, slice dice and visualize the data like crazy, and finally reach the following insights:

(I wish I can have the time to actually do the things above)

  • Deep sleep happens early: Get to bed early and sleep equal or less time will improve our sleep.   The concept might not be completely new, as research like this has made a similar point.  I never truly believe it, until I see my own data. Below is a screenshot of a “typical night” (Technically more data point is needed than just one typical example).   Being able to see my own sleep pattern is very powerful –  the purple bars, indicating deep sleep, clutters almost exclusively before 3AM.
How Deep sleep is distributed during the night
  • High resting heart rate indicatesillness or vice versa:  I don’t get sick very often. But when I am, my average heart rate went up significantly.  Apparently, many others have realized the same thing.   To be completely honest, this is a good-to-know metrics, and more likely we can only use it for “descriptive purpose” – a.k.a – “because of illness, the resting heart rate goes up. ” Personally, I am looking forward to the advancement of the technology, and we will be able to use it for prediction: “When our heart rate goes up by over 30%, the chances we are suffering from X disease is increasing by Y%”.    Additional metrics, like blood pressure and body temperature will be needed.  Will that happen in the near future? Let us wait and see.


Apple watch is not the only option in the market to track sleep.  I have used Misfit in the past, and Fitbit is another popular option. Both of these two are cheaper than Apple Watch, and the features are very similar.  However, those usually lack 3rd party analyzing tool.


Battery life: Putting on the watch when asleep means you won’t be able to charge it during the night.  Usually, it will cost around 20% of the battery to track a normal night of sleep.

Comfort: It does feel weird at the beginning to wear a watch to sleep. It is better to turn on the “theater mode”, otherwise the watch screen might turn on accidentally during the night.

Price: Apple watch isn’t cheap. I won’t recommend to use it to track sleep as the main purpose. Starting with a Misfit Flash (cost less than $30) might be a more economical choice.  However, if you already have one, the 3rd party app to track sleep is affordable: It cost less than $5.

Leave a Reply