Comparing Withings's sleep data to Oura's

The Oura ring was designed around the principle that the day begins with a sleep period. I have always thought of the day ending with crashing into bed and resting from the day’s labor. I am trying to shift my perspective to one of today’s sleep (midnight to morning) is preparation for today’s thinking and doing (morning to midnight). Both the Oura ring and the Withings watch report sleep as the beginning of today (e.g., the sleep period reported on 7/15/2019 is the sleep that began at 11:00 pm on 7/14/2019, basically midnight to morning of 7/15/19). Semantics, perhaps, but consciously thinking in new ways is a creative exercise.

Years ago I read Frans Johansson’s discussion of the power of reversing assumptions (see his book, The Medici Effect). I still use that practice today. 

For the past four days I wore the Withings watch to bed. Unlike the Apple Watch, the Withings watch does not have to be charged everyday, so wearing it to bed is seamless. I also wore the Oura ring, as usual. And, I kept a sleep journal of the time I went to bed, the time I woke up/got up during the night, and the time I got out of bed in the morning.

The ring is more comfortable; I don't even notice it. To be fair, I am getting used to the watch and find that using the leather band means I do not have to wear it so tightly. The Withings watch is more comfortable to wear overnight than the Apple Watch (less bulky than Apple's).

The Oura ring correctly recorded when I went to bed, because I take it off when I am getting ready for bed and put it on when I go to bed. It senses when the ring is off and then on and when I am vertical versus horizontal. The Withings watch can not. It never got the time I went to bed correct. It would be off by an hour or more. It would record as bedtime the time I took it off and got into the shower or the time I was walking around the house closing up for the night.

That last “feature” was particularly troubling. Two nights last week, that last 200 steps was the difference between making my goal for the day or not. The watch recorded the steps and set off fireworks and congratulations for making my goal, but the steps were never recorded in the app, even though I synced the app multiple times before going to bed and first thing in the morning. Needless to say, that was REALLY depressing.

The ring and the watch differ in how they stage sleep. The ring divides sleep into 4 categories: awake, light sleep, deep sleep (physically restorative), and REM sleep (mentally restorative). The watch divides sleep into awake, light sleep, and all other sleep. In the Health Mate app, the all other category is referred to as “restorative” or “deep” sleep. To make comparisons of the ring and watch I summed the ring’s measurement of deep and REM sleep into a single category of restorative sleep.

Each morning, there was a clear pattern of differences in the sleep analysis data from the two sources. Compared to the Oura ring, each day the watch over-estimated the total hours of sleep (by an average of 10%) and over-estimated the hours of restorative sleep (by an average of 30%). The average heart rate during the night was very close each day (the watch over-estimated by 1 bpm).

 

 

I found it interesting that the general form of Health Mate app’s plot of heart rate during the night matched the Oura app’s well. Heart rate is nicely rendered in Health Mate and clearly presents the variance in rate. [Variance in heart rate, is not the same as heart rate variability. Find a good explanation of the comparison here.]

Both the watch and the ring provide a sleep “score” that ranges from 0 to 100 and is based on multiple, multi-day factors. Those factors are different between the two sources, which resulted in an interesting comparison. Even though the watch said I had more hours of sleep and more hours of restorative sleep, it consistently gave me a lower overall score (e.g., this morning's watch score was 84, while the ring score was 89).

My general conclusion is (1) the ring has more sensors, (2) the sensors are more sensitive, and (3) the algorithms are more accurate than any watch I have tried. Comparing the ring to the Withings watch is comparing apples to oranges. Once I learned to use the ring’s additional detailed metrics, there was no going back. Still I need both ring and watch to do what I want to do.

The only monitor that might be close to the Oura ring, is the Whoop system, which is not a watch, nor a ring; it is a wrist strap. Its data and coaching may be a close second to the Oura ring. I did not try it, because I don't like the subscription business model. Oddly, I would still need a watch for telling time and getting daytime, real-time feedback.

The Withings app provides some coaching and promises a new Health Coach program will be offered in Summer 2019. It will be interesting to see how it compares to the coaching and detailed data of the ring.

I don't think the ring is for everyone. I am fascinated with data. Analyzing patterns helped me to focus on understanding how lifestyle affects sleep and vice versa. If I didn't have the time to spend 20 minutes to an hour with the data everyday, I think the ring would have been just another pedometer/heart rate monitor. Below are screenshots of sleep and readiness results from a recent day. I have come a long way from where I started. Amazing what small changes can do (e.g., shifting when I eat, how cool the bedroom is, and how much I move around during the day), when there is detailed feedback on the effect of making a change. The Oura app’s coaching has taught me to recognize patterns and understand their meaning.

   

My conclusions about the Oura ring (see earlier posts, especially July 9th, 2019)...

• Its measurement of activity during the day is too sensitive for some activities and not sensitive enough for others.
• I dislike its 10 minute delay in counting steps.
• I wish its app had GPS-connected route mapping for walking/hiking.
• I wish it magically had a display for steps throughout the day.
• No other watch or ring provides the same physiology data (body temp, respiration, HRV, etc.).
• No other watch or ring provides the data in as useful a way.
• It sets the standard for providing trend data with a single click and for access to raw data.
• It excels at providing detailed data and coaching about sleep, activity, and meditation that are hugely helpful in managing a healthy lifestyle.
• Its comfort level makes all other chest straps, watches, etc. pale by comparison.
• It has superb battery life and an easy to use charger.

My conclusions about the Withings watch (see earlier posts, especially July 9th, 2019)...

• Its measurement of activity during the day is not sensitive enough.
• Its measurement of sleep parameters is not sensitive enough.
• The delays in, and failures to, sync with the Health Mate app are more than annoying; they make me question the $200 purchase of the watch. 
• The app is well done. It is easy to read/use and it does provide interesting background articles.
• The ability to manually enter weight, percent body fat, and blood pressure and to have the app plot them and related derivatives (BMI, fat mass, hypertension, etc.) is great.
• It offers other integrated devices for measuring weight, percent body fat, blood pressure, and sleep, so users who want to invest in a whole system of devices can do so. 
• Its GPS-connected route mapping for walking/hiking, with heart rate, pace, and elevation is  well done. So good that it is a reason for keeping the watch.
• Its undercounting of steps and analog dial of goal achievement is strangely motivating, encouraging me to get up and walk around during the day, walk faster, and take a walk in the neighborhood at the end of a sedentary day. That feedback keeps me wearing it everyday.
• It has superb battery life and an easy to use charger.