58

In a TED talk "Why school should start later for teens", Wendy Toxel made an assertion that teenagers naturally fall asleep and rise up a couple of hours later than children OR adults (related to Melatonin production time).

I found one BBC article agreeing, but it didn't cite any research.

Does the research conclusively back that assertion up?

user5341
  • 31,075
  • 8
  • 130
  • 178
  • 4
    Unspoken: "Or are they just lazy?" – Ask About Monica May 24 '17 at 23:52
  • I'd also like to see how this lines up to the many, MANY articles about training your own circadian rhythm. Are teenagers unable to decide when they sleep? – AJFaraday May 25 '17 at 08:28
  • @AJFaraday from what I've read, it seems to be that the hormones (namely melatonin) responsible for sleep are produced later in teens. Of course it's possible to influence its production but it would be more difficult for a teen to change it to as early a time as an adult could, because adults already have the extra couple hours natural production time. – theonlygusti May 25 '17 at 15:50
  • 7
    When I first started a job where I got to set my own hours, I quickly fell back to the "teenage sleep cycle" and felt I was actually more productive for it. I have to wonder how much of the "adult" sleep cycle is just a constraint of having to hold down a job. – Chuu May 25 '17 at 16:26
  • 3
    @kbelder - if all teenagers are universally lazy, that too must be biological. – Davor May 25 '17 at 16:54
  • Note that questions like this would be appropriate on cogsci.stackexchange.com – Seanny123 May 26 '17 at 01:11
  • 2
    Personally I don't feel like getting up late is a privilege of teenagers only. I'm 36 and the only reason I wake up early is because of my 4 year old daughter. – Markus Malkusch May 26 '17 at 10:04

1 Answers1

61

There does appear to be fairly solid research backing this claim.

A summary page from UCLA: Sleep and Teens

One change in the body during puberty is closely related to how you sleep. There is a shift in the timing of your circadian rhythms. Before puberty, your body makes you sleepy around 8:00 or 9:00 pm. When puberty begins, this rhythm shifts a couple hours later. Now, your body tells you to go to sleep around 10:00 or 11:00 pm.

Neurology Times agrees:

In turns out that adolescents have a delayed release of regular daily melatonin, which causes them to become sleepy later at night, hours after nightfall. Given the fact that teenagers have an established need for 8-10 hours of sleep per night, the delayed melatonin release that allows teenagers to fall asleep late in the day has the expected effect of predisposing them to remain asleep for longer into the late morning or early afternoon, when it is feasible.

A relevant reference for the Neurology Times article:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2820578/

Other studies have been done that suggest better outcomes in teenage students when school starts later in the day, and not just in school:

Later high school start times are associated with positive outcomes among teens, including longer weekday sleep durations and reduced vehicular accident rates, research suggests.

Community
  • 1
Werrf
  • 741
  • 7
  • 11
  • 27
    I've always been skeptical of these studies, because its rare that they show or explore any mechanism connecting the two. The only thing I can think of is the Sun. So, what happens in areas with Daylight Savings time vs. areas that don't have it? Are the same effects seen at higher latitudes where the period of sunlight is lessened in the Winter? How about areas that are cloudy a lot of the year? They all seem to want to imply some magic relationship between health and the arbitrary numbers we assigned to timekeeping devices 4 millennia ago. – T.E.D. May 24 '17 at 20:18
  • I think those questions were outside the bounds of these studies, and this question. This was specifically about whether teenagers have different cicadian patterns from children or adults. What *I'd* really be interested in would be studies of teen behaviour in hunter-gatherer societies, to see if the same effect is observed there. – Werrf May 24 '17 at 22:13
  • 1
    @Werf is that relevant, though? I'm far more interested in how actual people today behave. The underlying reasons don't matter nearly as much. – user428517 May 24 '17 at 22:33
  • 14
    @sgroves It would help us determine whether it's something biological and inherent that can't be changed, or if it's something to do with our modern lifestyle that could be modified or ameliorated if we felt the need. – Werrf May 24 '17 at 22:36
  • Good luck getting everyone on the planet to adjust. Both are just as difficult to change, it seems to me. – user428517 May 24 '17 at 22:40
  • 15
    @sgroves It's not about getting everyone on the planet to adjust, it's about having the information to make an informed choice. We know that smoking can cause cancer. It's something to consider when making a choice. If my teenage son is struggling in school because he's short of sleep, I'd like to know if I can help him by cutting off video games sooner in the evening or by changing the lightbulbs, or if I should pressure the school into letting him start later in the day because there's nothing that can chance his sleep patterns. – Werrf May 25 '17 at 11:55
  • 6
    @T.E.D: studies are experimental, they're measuring what happens. So if indeed "adolescents have a delayed release of regular daily melatonin", that continues regardless of whether or not we can figure out how their pineal gland "knows" what time it is (but, yes, light has a lot to do with that). Rejecting experimental results because the experimentalist hasn't provided a complete theory seems... wrong. On the other hand, rejecting an alleged *theory* because it fails to propose a mechanism would be right. – Steve Jessop May 25 '17 at 14:30
  • @SteveJessop - That's kinda where I am. I'm not pushing back against the data, but rather against the inferences and suggested solutions that seem to be constantly made based on it without even having identified a mechanisim. For example, just last week my kid's high-school announced they are starting classes an hour later next year **because of these studies**. Misleading people by breathlessly reporting about studies with implied conclusions not supported by the data has real social consequences. – T.E.D. May 25 '17 at 14:38
  • 9
    @T.E.D.: to be fair to the school, even if the late melatonin-release turns out to be localised to Western teenagers making poor choices about when they go to bed, that fail to take into account their higher need for sleep compared with adults, then it *still* might be a rational response for the school to change hours. "We've tried asking you to send your kids to bed on time, studies show that didn't work and their melatonin release is too late, this is the next step". The school doesn't need to care *why* all their pupils are still asleep at 9am, just accept they've failed to prevent it :-) – Steve Jessop May 25 '17 at 14:42
  • 1
    @SteveJessop - More likely, the kids will just respond by going to bed another hour later. There's still the same # of hours in a day, no matter where you start and end it. If there was something special about the sun being an hour higher, then I could see it. But that's not what these particular studies are checking for. – T.E.D. May 25 '17 at 14:46
  • 1
    @T.E.D. If the kids did just respond by going to bed another hour later, we likely wouldn't see improved outcomes by shifting school start times, would we? We might see a short-term improvement as they adjusted, followed by a return to the previous numbers. – Werrf May 25 '17 at 14:55
  • @Werrf - Perhaps, but this isn't being sold as a scientific experiment, but rather a remedy. The people reporting on these studies are making implications not supported by the data (for the reasons I explained in the first comment). – T.E.D. May 25 '17 at 15:34
  • Look at it this way: Science isn't just about producing data. You need to be able to do [something predictive with the data](https://en.wikipedia.org/wiki/Scientific_method#Predictions_from_the_hypothesis). If there's no attempt whatsoever to identify a mechanism linking the data to the effect you are measuring, you can't be predictive with the data produced. That's simply not useful science (yet). – T.E.D. May 25 '17 at 15:40
  • 2
    @T.E.D. The evidence shows that it IS a remedy, since the studies do not show an immediate improvement followed by a return to normal. Why that remedy works or appears to work is up for debate, certainly, but whether it works because teenage hormones are all out of whack or because teens are naturally out of step with their parents or what have you, the evidence is that it does work. – Werrf May 25 '17 at 15:42
  • 2
    @Werrf - "the studies do not show an immediate improvement followed by a return to normal". Ahhh. Now *that's* the kind of thing I've been missing! Got a link for that? I couldn't find it in your linked material. What I found was [more like](https://www.sciencedaily.com/releases/2016/12/161215085930.htm): "the review found somewhat mixed support for the assertion that delaying school start times improved grades or standardized test scores" (IOW: Not really anything much). – T.E.D. May 25 '17 at 15:47
  • @SteveJessop are you assuming the school should adapt then? And when kids will be sleepy at 10, adapt again, and again until lectures start in the afternoon and kids are completely desynchronised? At that age I used to get up at 6:30 and start lectures at 8. Wasn't easy to get up so early but by the time it was 8 o'clock I never felt sleepy anymore. – Shautieh May 26 '17 at 03:22
  • @Shautieh: I don't know whether they should or not, I'm saying that whether it's rational for them to do so or not doesn't really depend on whether they know the mechanism. You predict that the kids will start to be sleepy an hour later. Werrf claims studies which observe this did not occur, so *regardless of mechanism* there is at least a measurable difference between the prediction that you and T.E.D. make, and the prediction the school makes. If the measurement is made for T.E.D's school, someone will be wrong. So we can proceed without a mechanism :-) – Steve Jessop May 26 '17 at 11:17
  • @T.E.D. from the linked PMC article"social factors cannot completely account for the adolescent delayed sleep onset typical of an evening chronotype. [...] suggests physiological underpinnings. Girls begin to show a delay in the timing of sleep 1 year earlier than boys, paralleling their younger pubertal onset. [...] In other cultures, similar developmental timing is observed, [...]. delay in the timing of sleep during the second decade of life has been observed in over 16 countries on 6 continents, in cultures ranging from pre-industrial to modern" (shortened to fit into comment) – cbeleites unhappy with SX May 26 '17 at 19:57