For the third week in a row, we’ve updated and improved the WHOOP Sleep Analysis. If you’ve been answering the app’s daily sleep survey questions, you may have noticed something new this morning--your responses are now incorporated into the feedback and recommendations we provide both in the app and on the web.
It should come as no surprise that certain activities prior to sleeping can lead to positive or negative impacts on sleep. While some behaviors seem to be obviously helpful or harmful to sleep, others affect users differently depending on who they are. This post focuses on high-level observations from the general WHOOP population, but the best way to discover how these behaviors impact you personally is to answer the survey each morning and let our upgraded WHOOP Coach work its magic.
For this report, we analyzed over three million sleeps recorded by our users alongside the survey answers they provided the next morning. Before we get into what we learned from their answers, let’s take a look at the questions we asked and how our users responded.
The first thing we looked at was how often users answered Yes/No to five of our most interesting survey questions:
Next, we looked into the demographic differences in these behaviors. In the set of images that follows, the y-axis shows the percentages of “Yes” responses for the indicated age.
Not surprisingly, age was strongly correlated with the frequency with which all questions were answered in the affirmative. As WHOOP users get older, they are more likely to self-report consuming both alcohol and caffeine in the evening, are less likely to use a screened device in bed, and are more likely to both read before bed and not to sleep alone.
The next step was to explore what benefits our users are getting out of doing or not doing those things. To learn more about this, we looked at what the average Sleep Performance, Sleep Efficiency, Slow Wave Sleep duration, REM Sleep duration, resting heart rate, and heart rate variability (HRV) were following affirmative and negative responses to the survey questions above.
Evening Caffeine Consumption
Evening caffeine use impacts younger WHOOP users more than it does older ones. We see this most strongly in the Sleep Performance, Sleep Efficiency, and Resting Heart Rate graphs. There is a large gap between the average when responding yes (blue lines) and no (black lines) for users below 40 years of age that seems to close as age increases towards 60.
Note that in all of the graphs that follow, an affirmative response is shown in blue while a negative response is shown in black.
Evening Alcohol Consumption
A similar pattern is seen in our responses to alcohol consumption. Younger WHOOP members see their metrics decline more significantly after self-reported evening alcohol consumption than do older WHOOP members.
We should note that the data we are presenting here has a few limitations. Most significantly, we don’t ask about how much users drank and we rely on self reporting. It is quite possible that when younger WHOOP members are drinking they are drinking more than older members, and that alcohol volume rather than age is the primary driver of the pattern seen here.
Working on Screened Devices in Bed
Screened device use, on the other hand, follows the opposite pattern. The sleep of older users is more significantly impacted than is that of younger users. This pattern is particularly evident with Sleep Performance. While it may seem counterintuitive that Sleep Efficiency increases in young WHOOP members following screen-device use, this can actually be a sign of acute sleep deprivation. Essentially, the more tired we are, the more “sleep pressure” we experience and the better we sleep.
Old-School Reading In Bed
Reading not on a screened device before bed had a neutral or positive impact on all the metrics analyzed, most notably on Resting Heart Rate (for all age groups) and on Sleep Performance for WHOOP members under 50.
A 2009 study, in which reading was shown to reduce stress levels by up to 68% after only 6 minutes, supports these findings. In the study, reading out-performed the stress reducing effect of listening to music (61%) and taking a walk (42%).
Last month, we released the results of a survey we sent to the 100 Best Sleepers on WHOOP. A large number of them said that reading is an important part of their bedtime routines.
The impact of sharing your bed on sleep has perhaps the most interesting dependency on age. The data suggests that sharing your bed has a positive effect on sleep for older WHOOP members, seen in REM duration and Sleep Performance, and a negative impact on sleep for younger WHOOP members, seen in sleep efficiency, SWS duration, and Resting Heart Rate.
One study from the University of Pittsburgh showed that women sharing their bed with a long-term partner had higher sleep quality than did women who slept alone. Many people report feeling a sense of security when sleeping with other people, the sense of safety provided by bed partners might contribute to reduced levels of stress hormones, which may explain the improved quality of sleep.
How do you think you stack up?
Let us know how you sleep, what you think about the WHOOP Sleep Analytics Platform, and what changes you would like to see us make by filling out our survey!