Exposure to Backlit Displays Reduces Melatonin Production 192
alphadogg writes "Researchers have discovered that relatively little exposure to tablets and other electronics with backlit displays can keep people up at night by messing with their circadian rhythms. The study from the Lighting Research Center at Rensselaer Polytechnic Institute showed that a 2-hour exposure to electronic devices with such displays causes suppression of the melatonin hormone and could make it especially tough for teens to fall asleep. The study, funded by Sharp Laboratories of America, simulated usage of such devices among 13 people using special glasses/goggles and light meters"
Re:Explains a lot (Score:5, Informative)
Now we know why geeks are so pale.
You're thinking of Melanin, not Melatonin.
Re:Explains a lot (Score:5, Informative)
That's melatonin not melanin. Melatonin regulated sleep.
Re:Pretty Obvious + Plug for Awesomeness (Score:4, Informative)
Re:Sample size too small? (Score:5, Informative)
Surely 13 people is too few to draw meaningful conclusions?
Yes. Especially if not compared to people reading a book under a 60 watt incandescent light bulb.
Re:N = 13? (Score:4, Informative)
Because they wanted to control the exposure. People use tablets differently. Different distances, different brightnesses, different sized font (larger black letters means less light emitted), different tablets=different displays=different wavelengths emitted.
Too small of a group, but an interesting start.
Re:Pretty Obvious + Plug for Awesomeness (Score:5, Informative)
This.
Not that I use f.lux, but the GPL'd Redshift [jonls.dk] on my laptop. When switching it on, it feels like my eyes breathe a sigh of relief - it really is much easier to read off a red-orange-tinted surface at night.
Now if only they'd port it to Android.
Re:Any word? (Score:4, Informative)
Likewise. I have a Samsung TV with a bright red light that's on whenever the device is off (WHY?) and a Samsung with lighted buttons where the power button's light doesn't ever turn off (WHY?). Both are covered with electric tape.
I swear, every time I think electronics manufacturers can't get dumber, they prove me wrong. The worst part of it is the realization that a simple firmware fix would make the difference between these being great devices and making me want to fly to South Korea and smack all their engineers upside the head repeatedly with a clue-by-four....
Re:N = 13? (Score:4, Informative)
the omitted cases were due to not generating enough saliva for a melatonin assay. probably not much worry of confounding there.
it doesn't prove a whole lot, if anything. we already knew that blue light suppresses melatonin, and they give the predicted effect in the study along with their measurements. annoyingly, they don't give the two-hour theoretical effect, which is the regime in which they have statistical significance in their results. neither do they formally compare the tablet-only effect to high-blue-light (enforced by goggles) effect, but it's pretty obvious that the tablet isn't as bad. which, of course, isn't surprising since the lumens are lower.
conclusion: it's an almost completely useless study, but the statistics they give seem legit enough. they don't do multiple comparisons correction, but if they did, the two-hour effect would still be significant.
look, guys, if an experiment shows a statistically significant effect which also mostly conforms to the predicted effect (and there aren't blatant design errors), then there isn't much to complain about. i could, quite likely, have done this with n=6 (two for each treatment) and still gotten significance.
Re:Sample size too small? (Score:3, Informative)
Yes, this is really junk science, but I believe there are other studies that show similar results - see http://stereopsis.com/flux/research.html [stereopsis.com] for a list, including links to the full papers (the site is for F.Lux which I really recommend to adjust colour temperature to get more sleep, for Windows, Mac and Linux, and jailbroken iOS).
Re:N = 13? (Score:2, Informative)
fuck the paywall. here's the table. the intervals for measured melatonin suppression are +/- one standard error.
http://tinypic.com/view.php?pic=33bfl8o&s=6 [tinypic.com]