Fitbit is obsessed with sleep. The company has collected 6.5 billion nights worth of data now, and that has helped it observe different sleep trends amongst its users. But now it's doing something a little different, crossing this info with data from other Fitbit features to look for patterns.
In this case, it's taken anonymised data from 3,000 users who both use the Think Fast app for Ionic and Versa and also track their sleep. Think Fast is a two-minute game from Fitbit built on ideas from neuroscience and psychology to test cognitive ability and reaction time.
Read this: The best sleep trackers
Fitbit found that – on average – users who performed best on Think Fast slept between 5 hours and 50 minutes and 6 hours and 30 minutes per night. Don't get too excited, thinking you only need about six hours of sleep a night, though.
"This doesn't mean that you can go to bed and only be in bed for six hours and have optimal cognitive function," Jonathan Charlesworth, a staff research scientist at Fitbit, tells Wareable. "You basically want to add an hour to that." That added hour is for micro-wakes – all those times you wake up in the middle of the night for a couple of minutes.
Fitbit is quick to clarify that this isn't some extensive study it conducted. It simply got curious and decided to poke around its large data sets, finding some curious trends that were somewhat surprising.
Hot fitness tracker deals
Wareable may get a commission
For instance, the company found that women performed better on Think Fast with 30 minutes of more sleep a night than men. In fact, the relationship between REM sleep and Think Fast scores was higher than it was for men. Men, and women over 40, performed better on Think Fast with higher amounts of deep sleep. On the other hand, men who had higher time awake numbers had worse Think Fast scores, and that trend wasn't as strong for women.
Why the discrepancy? Charlesworth isn't sure, but it's a fascinating trend the company and its research scientist want to dive deeper into. One of the only ways to do that is to also look at some of the other data Fitbit collects, or can collect, including female health tracking, which the company says has about 2 million users thus far.
What is a little more sure is that Fitbit found that users over 30 saw a stronger link between sleep and better Think Fast scores. That mostly falls in line with other research that outlines that as you get older, sleep becomes more important.
How to recommend sleep
When you hear about sleep, you often hear about REM sleep, which is the sexy phase of sleep we associate with dreams. However, Charlesworth points out that it's good not to forget about deep sleep, which can be important for cognitive ability.
"In the last 10 years, researchers and neuroscientists have found that in the deeper stages of sleep the neurosystem seems to be cleaning itself of potentially toxic proteins and byproducts accumulated during the day due to your brain's activity," Charlesworth says. "And the researchers have a found link between this and cognitive decline and dementia in older people."
So what can you do? Fitbit recommends a time in bed of about 6 hours and 30 minutes to 7 hours and 20 minutes for men and 6 hours and 50 minutes and 7 hours and 50 minutes for women – based on its research so far. You'll want to aim for above 60 minutes of REM sleep and 85 minutes of deep sleep. Though you don't want to oversleep either, as Charlesworth says the trends show that oversleeping is just as bad for cognitive ability as undersleeping.
However, Fitbit also wants to keep its sleep recommendations a little more general, and that's because it recognises that its data set is looking at a wider population rather than drilling down into hard science that it can use to tell you exactly how much sleep you'll need. That's certainly possible, according to Charlesworth, but Fitbit isn't there yet.
"There are specific reasons why some people need more or less sleep. One of those appears to be age, one of those appears to be gender. As we go deeper into this data I fully expect us to find more and more explanations for why different people need more or less sleep."