The Importance of Sleep for Adolescents

There’s a wonderful study that you may have heard about, although it was only published a couple of weeks ago. It’s all about sleep and adolescents, and it has the best title that Tom Hanks would be proud of in “Sleepmore in Seattle”. Check it out here.

You see, sleep in teens is a little funky. They experience changes in what are called “circadian rhythms”, which is basically like the body’s internal clock guiding sleep and wake times. Teenagers have a shift in their natural internal clock which tells their body to go to bed later and wake up later. But in the real world this is a massive problem for them, because they have school to attend! If they go to bed later, they have to wake up earlier than their body wants to in order to get to school, and this results in less sleep. Less sleep is associated with all sorts of problems. Anyone reading this knows what it’s like to get sleep deprived. For a teenager going through important development, sleep deprivation for them is even worse. Aside from all the negative mental and physical consequences of sleep, it’s very bad for students because they have less time for one important mechanism of sleep, which is called “memory consolidation”. This is just a process of your brain creating new learning from whatever you did the day before.

Every adolescent therefore has a fight going on between their body’s internal clock and the school’s clock.

The Seattle school district recognized this problem and decided to do something about it. So they moved the start time of classes in High School from 7:50am to 8:45am. A smart group of researchers decided to examine the effects this would have on things like sleep, grades, and mood by studying kids in the Spring before the change, and in the Spring the following year after the change.

What Did They Do?

They gave kids at two different schools wrist monitors to wear. The monitors took recordings every 15 seconds about activity and light. Kids could also press a button on the watch to indicate they were going to sleep/waking up. They also had kids answer daily questions, including about sleep/wake time to compare against the data collected from their monitors. This is SO MUCH better than other studies because it’s actually collecting real-time recordings plus self-report. They had 3 different ways of determined sleep onset, end, and duration:

  • Watch monitoring activity
  • Watch button kids could press to “timestamp” sleep
  • Self-report

To combine all this information and figure out what the actual sleep pattern was they did some fancy adjustments which they describe in the paper but I’m not going to explain here.

Students took part in the study over a 2-week interval during the spring quarter. To try and spread out the data over a longer time period (which helps reduce the likelihood of getting biased data), they had some kids record data in the first 2 weeks, others in the following 2 weeks, and others in the 2 weeks after that.

  • They did this for kids in 2016 before the change in start time happened
  • They did this again with a different set of kids in 2017 after the change in start time happened

What Did They Find?

  1. Sleep was 34 minutes longer than is 2016 on school days, but there no difference on non-school days (weekends).
  2. Exposure to light was also significantly delayed in 2017 compared to 2016 (because they were sleeping longer), but in the evenings there was no change. I think this means that adolescents were not staying up later on screens, although they didn’t say this.
  3. Importantly, there was also some evidence that sleep deprivation correlated with tardiness/absences at school. They found this happened in one of the two schools, which was the most socio-economically disadvantaged one. More sleep was associated with fewer missing/tardy days. Specifically, absences in 2016 averaged 15.5 days per child whereas in 2017 it averaged 13.6. In addition, the average number of tardys in this school was 6.2 in 2016 and it was down to 4.3 in 2017. This doesn’t mean causality (i.e. more sleep doesn’t cause fewer absences), but it’s still an interesting association.
  4. Students were also significantly less sleepy during the day based on responses to a sleep scale called the “Epworth Sleepiness Scale”. Their grades were also significantly better in 2017 than in 2016, with a 4.5% increase in grades. However, again it’s very important to note that this is not a causal relationship.


Most studies just use self-report which is notoriously unreliable with sleep data. Not only that, they often base sleep information on 1 question or a parent report. The fact that they used self-report and device data with many kids over time is simply outstanding. I just can’t tell you enough how great the data they collected was!


How often is it that you get a chance to do a study like this? When does a school district ever make such a change?! Given what a unique opportunity this was, it’s a real shame how few people were included in the study. They had 92 people in 2016 and 82 in 2017. And not all of them could use the wrist monitors. I don’t know that this gives the study quite the power it needs to show us findings we can have real confidence in. It’s not their fault as they were clearly limited by the grant and not being able to afford more devices, something they alluded to. But it’s a big deal because this is an incredibly important study that could have implications for schools across the nation, and to not have a larger group means people will remain skeptical.

  • If you’re interested in the finer details: Looking at the exact data, this low N is more of a concern to me with the data on sleepiness and academic performance, with the p values in the range of .02 to .03. Given the low N, even though the results for sleeping and academic performance are “significant”, we should be very skeptical that if the study was done again that it would show the same “significant” results.

Take Away

This is a study based on excellent (and common sense) theory; that teens need more sleep and this can help them in school. The results back it up. As we often say with research, this study needs to be replicated and expanded. And how critical this is. If you’re part of a Parent-Teacher Association, or work in school districts, you should be taking notice!

Leave a Reply